Regulations (Preambles to Final Rules) - Table of Contents Regulations (Preambles to Final Rules) - Table of Contents
• Record Type: Occupational Exposure to 1,3-Butadiene
• Section: 5
• Title: Section 5 - V. Health Effects

V. Health Effects

A. Introduction

The toxicity of BD was long considered to be low and non-cumulative. Thus, the OSHA standard for BD was 1,000 ppm on the basis of its irritation of mucous membranes and narcosis at high levels of exposure. However, in the 1980s, carcinogenicity studies indicated BD is clearly a carcinogen in rodents. In 1986, the American Conference of Governmental Industrial Hygienists (ACGIH) was prompted by these studies to lower the workplace threshold limit value (TLV) from 1,000 to 10 ppm. (Ex. 2-5) Rodent studies are now conclusive that BD is an animal carcinogen.

Further, a consistent body of epidemiologic studies have also shown increased mortality from hematopoietic cancers associated with BD exposure among BD-exposed production and styrene/BD rubber polymer workers. Complementary studies of metabolic products and genotoxicity support these cancer findings. OSHA was also concerned about evidence that BD affects the germ cell as well as the somatic cell, and what potential reproductive toxicity might result from exposure to BD. Since BD itself does not appear to be carcinogenic, but must be metabolized to an active form, OSHA also reviewed studies on the metabolism of BD to determine wether they might help explain the observed differences in cancer incidence among species.

The following sections discuss the effects of BD exposure, both in human and animal systems.

B. Carcinogenicity

1. Animal Studies

In the proposed BD rule, OSHA discussed the results of two lifetime animal bioassays, one on the Sprague-Dawley rat and one in the B6C3F(1) mouse. (55 FR 32736 at 32740) Both studies found evidence of BD carcinogenicity, with the greater response in the mouse. The rat study involved exposure levels of 0, 1000, or 8000 ppm BD, starting at five weeks of age, to groups of 100 male and 100 female Sprague-Dawley rats for 6 hours per day, five days per week, for 105 weeks. Mortality was increased over controls in the 1,000 ppm exposed female rats and in both of the male rat exposure groups. Significant tumor response sites in the male rats included exocrine adenomas and carcinomas (combined) of the pancreas in the highest exposure group (3, 1, and 11 tumors in the 0, 1000, and 8000 ppm groups, respectively); and Leydig-cell tumors of the testis (0, 3, and 8 in the same groups, respectively). In the female rats, the significantly increased tumor response also occurred in the highest exposure group; cancers seen included follicular-cell adenomas and carcinomas (combined) of the thyroid gland (0,4, and 11 tumors in the three exposure groups, respectively), and benign and malignant (combined) mammary gland tumors (50, 79, and 81 in the same exposure groups). To a lesser degree there were also sarcomas of the uterus (1, 4, 5 tumors in the three exposure groups), and Zymbal gland (0, 0, 4 tumors in the same exposure groups, respectively). While only high exposure group tumor response for some of these sites was statistically significant, trend tests were also significant.

In contrast to the generally less than 10% increase in tumor response seen in the Sprague-Dawley rat at levels far above BD metabolic saturation, the carcinogenic response to BD in the B6C3F(1) mouse in the National Toxicology Program study (NTP I) was extensive. (Ex. 23-1) In this study, groups of 50 male and 50 female mice were exposed via inhalation to 0, 625 or 1250 ppm BD for 6 hours per day, 5 days per week in a study originally designed to last 2 years. However, the high carcinogenic response included multiple primary cancers, with short latent periods, and led to early study termination (60-61 weeks) due to high cancer mortality in both the 625 ppm and 1250 ppm exposure groups of both sexes. This mortality was due mainly to lymphocytic lymphomas and hemangiosarcomas of the heart, both of which were typically early occurring and quickly fatal. This large and rapidly fatal carcinogenic response led to both the NTP and industry to undertake additional studies to better understand the mechanisms involved.

Some commenters have associated qualitative or quantitative differences in mouse and rat BD carcinogenicity with the differences in rat and mouse BD metabolism. Many studies published and submitted to the BD record since the proposed rule have sought to better characterize the metabolic, distributional, and elimination processes involved, and some have attributed species differences (at least in part) to the metabolic differences. These will be addressed separately in the metabolism section.

Another factor hypothesized to account for differences between mouse and rat BD carcinogenicity was the role of activation of ecotropic retrovirus in hematopoietic tissues on tumor response in the B6C3F(1) mouse. This virus is endogenous to the B6C3F(1) mouse and was hypothesized to potentiate the BD lymphoma response in this strain. To study this hypothesis Irons and co-workers exposed both (60) B6C3F(1) male (those with the endogenous virus) and (60) NIH Swiss male (those without the endogenous virus) mice to either 0 or 1250 ppm BD, for 6 hours./day, 5 days per week for 52 weeks. (Ex. 32-28D) A third group of 50 B6C3F(1) male mice received 1250 ppm for 12 weeks only and was observed until study termination at 52 weeks. The results of the study showed significantly increased thymic lymphomas in all exposed groups but significantly greater response in the B6C3F(1) mouse--1 tumor/60 (2%) in the control (zero exposure) group, 10/48 (21%) in the 12 week exposure group, and 34/60 (57%) in the 52 week exposure group--vs. the NIH Swiss mice, which developed 0 tumors/60 in the control group, and 8 tumors/57 (14%) in the BD exposed group. Hemangiosarcomas of the heart were also observed in both strains exposed to BD for 52 weeks--5/60 (8%) in the B6C3F(1) mice vs. 1/57 in the NIH Swiss mice. (Ex. 32-28D). The B6C3F(1) response was very similar to the NTP I high exposure group response, verifying that earlier study. The qualitatively similar lymphoma responses of the two strains also confirmed that the mouse hematopoietic system is highly susceptible to the carcinogenic effects of BD, although quantitatively the strains may differ. The 21% 1-year lymphoma response in the 12-week stop-exposure B6C3F(1) group also increased concerns about high concentration, short duration exposures.

NTP II Study

Concurrent with the industry studies, the NTP, in order to better characterize the dose-response and lifetime experience, conducted a second, much larger research effort over a much broader dose range. (Ex. 90; 96) These toxicology and carcinogenesis studies included a 100-fold lower (6.25 ppm) low exposure group than NTP I, several intermediate exposure groups, a study of dose-rate effects using several high-concentration partial lifetime (stop-) exposure groups, and planned interim sacrifice groups. Other parts of the study included clinical pathology studies (with the 9- and 15-month interim sacrifices, metabolism studies, and examination of tumor bearing animals for activated oncogenes).

For the lifetime carcinogenesis studies, groups of 70 B6C3F(1) mice of each sex were exposed via inhalation to BD at levels of 0, 6.25, 20, 62.5, 200, or 625 ppm (90 of each sex in this highest group) for 6 hours per day, 5 days per week for up to 2 years. Up to 10 randomly selected animals in each group were sacrificed after 9 and 15 months of exposure, and these animals were assessed for both carcinogenicity and hematologic effects.

For the stop-exposure study, different groups of 50 male mice were exposed 6 hours per day, 5 days per week to concentrations of either 200 ppm for 40 weeks, 625 ppm for 13 weeks, 312 ppm for 52 weeks, or 625 ppm for 26 weeks. Following the BD exposure period, the exposed animals were then observed for the remainder of the 2-year study. The first two stop-exposure groups received a total exposure (concentration times duration) of 8,000 ppm-weeks, while the latter two groups received approximately 16,000 ppm-weeks of exposure. For the analysis discussed below, groups are compared both with each other for dose-rate effects and with the lifetime (2 year) exposure groups for recovery effects.


Male mice were 6-8 weeks old and female mice were 7-8 weeks old when the exposures began. Animals were exposed in individual wire mesh cage units in stainless steel Hazelton 2000 chambers (2.3 m(3)). The exposure phase extended from January, 1986 to January, 1988. Animals were housed individually; water was available ad libitum; NIH-07 diet feed was also available ad libitum except during exposure periods. Animals were observed twice daily for moribundity and mortality; animals were weighed weekly for the first 13 weeks and monthly thereafter. Hematology included red blood cell count (RBC), and white blood cell count (WBC). The study was conducted in compliance with the Food and Drug Administration (FDA) Good Laboratory Practice Regulations with retrospective quality assurance audits.

The results of the study are presented below for the two-year and stop-exposure study. Between study group comparisons are made where it is deemed appropriate. Emphasis is placed on the neoplastic effects.


Two-Year Study

While body weight gains in both exposed male and female mice were similar to those of the control groups, exposure related malignant neoplasms were responsible for decreased survival in all exposure groups of both sexes exposed to concentrations of 20 ppm or above. Excluding the interim sacrificed animals, the two-year survival decreased uniformly with increasing exposure for females (37/50, 33/50, 24/50, 11/50, 0/50, 0/70), and nearly uniformly for males (35/50, 39/50, 24/50, 22/50, 4/50, 0/70). As with the earlier NTP study, all animals in the 625 ppm group were dead by week 65, mostly as a result of lymphomas or hemangiosarcomas of the heart. The 200 ppm exposure groups of both sexes also had much higher mortality, but significantly less than that of the 625 ppm group. The survival of the lowest exposure group (6.25 ppm) was slightly better than controls for the male mice, slightly less for the female mice. Mean survival for the males was an exposure-related 597, 611, 575, 558, 502, and 280 days; for the females it was similarly 608, 597, 573, 548, 441, and 320 days. This decreased survival with increasing exposure was almost totally due to tumor lethality.


Nine different sites showed primary tumor types associated with butadiene exposures, seven in the male mice and eight in the female mice. These were lymphoma, hemangiosarcoma of the heart, combined alveolar-bronchiolar adenoma and carcinoma, combined forestomach papilloma and carcinoma, Harderian gland adenoma and adenocarcinoma, preputial gland adenoma and carcinoma (males only), hepatocellular adenoma and carcinoma, and mammary and ovarian tumors (females only). These are shown in Table V-1 adapted from Melnick et al. (Ex. 125) From this table it is seen that six of these tumor sites are statistically significantly increased in the highest exposed males and five were statistically significantly increased in the highest exposed females. Two additional sites which showed significant increases at lower exposures showed decline at the highest exposures because other tumors were more rapidly fatal. At 200 ppm preputial gland adenoma and carcinoma combined were significantly increased in males (p< .05; 0/70 (0%) control vs. 5/70 (7%) in the 200 ppm group) and hepatocellular adenoma and carcinoma were increased for both exposed males and females. At the lowest exposure concentration, 6.25 ppm, only female mouse lung tumors (combined adenoma and carcinoma) showed statistical significance (p< .05; 4/70 (6%) in controls vs. 15/70 (21%) in the 6.25 ppm group); these tumors in female mice showed a monotonic increase with increasing exposure up to 200 ppm. At 20 ppm female mouse lymphomas and liver tumors also reached statistical significance (lymphomas, p< .05; 10/70 (15%) in controls vs. 18/70 (26%) in the 20 ppm group; liver tumors, p< .05; 17/70 (24%) in controls vs. 23/70 (33%) in the 20 ppm group), and at 62.5 ppm, tumors at several other sites were also significantly increased. In general, while there were some differences in amount of tumor response between the male and female mice, there is fairly consistent pattern of tumor type in mice of both sexes for the six non-sexual organ sites.

                      [Adapted from Ex. 125]
                       |   |   Exposure concentration (PPM)   |
                       |   |__________________________________|
        Tumor          |Sex|    0     |    6.25   |    20     |
                       |   |__________|___________|___________|
                       |   |  I  |R(c)|  I  |  R  |  I  |  R  |
Lymphoma...............| M | 4/70|  8 | 3/70|  6  |8/70 |  19 |
                       | F |10/70| 20 |14/70| 30  |(a)18|  41 |
                       |   |     |    |     |     | /70 |     |
Heart--                |   |     |    |     |     |     |     |
  Hemangiosarcoma......| M | 0/70|  0 | 0/70|  0  | 1/70|   2 |
                       | F | 0/70|  0 | 0/70|  0  | 0/70|   0 |
Lung--Alveolar-        |   |     |    |     |     |     |     |
  bronchiolar adenoma  |   |     |    |     |     |     |     |
  and carcinoma........| M |22/70| 46 |23/70| 48  |20/70|  45 |
Forestomach--Papilloma |   |     |    |     |     |     |     |
  and carcinoma........| F | 4/70|  8 |15/70|(a)32|19/70|(a)44|
Harderian gland--      |   |     |    |     |     |     |     |
  Adenoma and          |   |     |    |     |     |     |     |
  adenocarcinoma.......| M | 1/70|  2 | 0/70|  0  | 1/70|   2 |
                       | F | 2/70|  4 | 2/70|  4  | 3/70|   8 |
Preputial gland--      |   |     |    |     |     |     |     |
  Adenoma and carcinoma| M | 6/70| 13 | 7/70| 15  |11/70|  25 |
                       | F | 9/70| 18 |10/70| 21  | 7/70|  17 |
Liver--Hepatocellular  |   |     |    |     |     |     |     |
  adenoma and carcinoma| M | 0/70|  0 | 0/70|  0  | 0/70|   0 |
Mammary glan--         |   |     |    |     |     |     |     |
  Adenocareinoma.......| M |31/70| 55 |27/70| 54  |35/70|  68 |
Ovary--Benign and      |   |     |    |     |     |     |     |
  malignant granulosa- |   |     |    |     |     |     |     |
  cell tumors..........| F |17/70| 35 |20/70| 41  |23/70|(a)52|
                       | F | 0/70|  0 | 2/70|  4  | 2/70|   5 |
                       | F | 1/70|  2 | 0/70|  0  | 0/70|   0 |

                       |   |    Exposure concentration (ppm)   |
                       |   |___________________________________|
      Tumor            |Sex|    62.5   |    200    |    625    |
                       |   |___________|___________|___________|
                       |   |  I  |  R  |  I  |  R  |  I  |  R  |
Lymphoma...............| M |11/70|(a)25| 9/70|(a)27|69/90|(a)97|
                       | F |10/70|   26|19/70|(a)58|43/90|(a)89|
Heart--Hemangiosarcoma.| M | 5/70|(a)13|20/70|(a)57| 6/90|(a)53|
                       | F | 1/70|   3 |20/70|(a)64|26/90|   84|
Lung--Alveolar-        |   |     |     |     |     |     |     |
  bronchiolar adenoma  |   |     |     |     |     |     |     |
  and carcinoma........| M |33/70|(a)72|42/70|(a)87|12/90|(a)73|
Forestomach--Papilloma |   |     |     |     |     |     |     |
  and carcinoma........| F |27/70|(a)61|32/70|(a)81|25/90|(a)83|
Harderian gland--      |   |     |     |     |     |     |     |
  Adenoma and          |   |     |     |     |     |     |     |
  adenocarcinoma.......| M | 5/70|  13 |12/70|(a)36|13/90|(a)75|
                       | F | 4/70|  12 | 7/70|(a)31|28/90|(a)85|
Preputial gland--      |   |     |     |     |     |     |     |
  Adenoma and carcinoma| M |24/70|(a)53|33/70|(a)77| 7/90|(a)58|
                       | F |16/70|(a)40|22/70|(a)67| 7/90|   48|
Liver--Hepatocellular  |   |     |     |     |     |     |     |
  adenoma and carcinoma| M | 0/70|   0 | 5/70|(a)17| 0/90|   0 |
Mammary glan--         |   |     |     |     |     |     |     |
  Adenocarcinoma.......| M |32/70|   69|40/70|(a)87|12/90|   75|
Ovary--Benign and      |   |     |     |     |     |     |     |
  malignant granulosa- |   |     |     |     |     |     |     |
  cell tumors..........| F |24/70|(a)60|20/70|(a)68| 3/90|   28|
                       | F | 6/70|(a)16|13/70|(a)47|13/90|(a)66|
                       | F | 9/70|(a)24|11/70|(a)44| 6/90|   44|
  Footnote(a) Increased compared with chamber controls (0 ppm), p <  0.05,
based on logistic regression analysis.
  Footnote(b) The Working Group noted that the incidence in control males
and females was in the range of that in historical controls (Haseman et
al., 1985).
  Footnote(c) Mortality adjusted tumor rates are adjusted for competing
causes of mortality, such as death due to other tumors, whose rates differ
by exposure group.

Hemangiosarcoma of the heart, with metastases to other organs was first observed at 20 ppm in 1 male (the historical controls for this strain are 1/2373 in males and 1/2443 in females), in 5 males and 1 female at 62.5 ppm and in 20 males and 20 females at 200 ppm; at 625 ppm these tumor rates leveled off as other tumors, especially lymphomas became dominant. Lymphatic lymphomas increased to statistical significance first in females at 20 ppm and were usually rapidly fatal, the first tumor appearing at week 23, most likely preempting some of the later appearing tumors in the higher exposure groups. Because of the plethora of primary tumors and the different time patterns observed to onset of each type, several tumor dose-response trends do not appear as strong as they would otherwise be.

Non-Neoplastic Effects

Several non-cancer toxic effects were noted in the exposed groups, reflecting many of the same target sites for which the neoplastic effects were seen. (Ex. 90; 96; 125).

Although the reported numbers differ slightly in the different exhibits, generally dose-related increases in hyperplasia were observed in the heart, lung, forestomach, and Harderian gland, both in the two-year study (both sexes) and in the stop-exposure study (conducted in males only). In addition, testicular atrophy was observed in both the two-year and stop-exposure male mice, but remained in the 6%-10% range except for the 2-year, 625 ppm group where it was 74%. Ovarian germinal hyperplasia (2/49 (control), 3/49 (6.25 ppm), 8/48 (20 ppm), 15/50 (62.5 ppm), 15/50 (200 ppm), 18/79 (625 ppm), ovarian atrophy (4/49, 19/49, 32/48, 42/50, 43/50, 69/79), and uterine atrophy (1/50, 0/49, 1/50, 1/49, 8/50, 41/78) were also dose related, with ovarian atrophy significantly increased at the lowest BD exposure of 6.25 ppm. These toxic effects to the reproductive organs are discussed in greater detail in the reproductive effects section of this preamble. Bone marrow atrophy was noted only in the highest exposure groups, occurring in 23/73 male mice and 11/79 female mice.

Stop-Exposure Study

As with the 2-year study, the body weights of the four treated groups in the stop-exposure study were similar to controls. All exposure groups exhibited markedly lower survival than controls, and only slightly better survival than that of the comparable full lifetime exposure groups. Mortality appeared to be more related to total dose than to exposure concentration. Most deaths were caused by tumors.

Neoplastic Effects

All of these stop-exposure groups exhibited a very similar tumor profile to that of the lifetime high exposure groups, with the lone exception of liver tumors, which were increased only in the lifetime exposure group; all the other multiple primary tumors were observed at significantly increased levels in both the stop-and lifetime-exposure groups, Table V-2. (Ex. 125) In addition, the 625 ppm, 26 week exposure group had higher rates for several of the tumor types compared to the lifetime 625 ppm group, possibly because of the shorter exposure group's slightly better survival. The most prevalent tumor type, lymphoma, also showed a dose-rate effect, as the tumor incidence was greater for exposure to short-term higher concentrations compared with a lower long-term exposure (p=.01; 24/50 at 625 ppm for 13 weeks vs. 12/50 at 200 ppm for 40 weeks: p< .0001; 37/50 at 625 ppm for 26 weeks vs. 15/50 at 312 ppm for 52 weeks). The same pattern was seen with forestomach tumors and preputial gland carcinomas. Conversely, the hemangiosarcomas of the heart and alveolar-bronchiolar tumors showed an opposite trend, as lower exposures for a longer time yielded a significantly higher incidence of these tumors than the same cumulative exposures over a shorter time (survival-adjusted, as opposed to the raw incidence lung tumor rates actually suggest no dose-response trends). These inconsistent trends with the different tumor sites may be the result of multiple mechanisms of carcinogenicity or partially due to the rapid fatality caused by lymphocytic lymphomas in the short-term high-exposure groups. As with the lifetime study, angiosarcomas of the heart and lymphomas presented competing risks in the highly exposed mice.

                       STUDY AT 104 WEEKS)
                     [Adapted from Ex. 125]
                      |             Exposure             |
                      |     0    | 200 ppm,  | 625 ppm,  |
         Tumor        |          |   40 wk   |   13 wk   |
                      |  I  |R(c)|  I  |  R  |  I  |  R  |
Lymphoma..............| 4/70|  8 |12/50|(a)35|24/50|(a)61|
Heart--Hemang-        |     |    |     |     |     |     |
  iosarcoma...........| 0/70|  0 | 7/50|(a)47| 7/50|(a)31|
Lung--Alveolar-       |     |    |     |     |     |     |
  bronchiolar adenoma |     |    |     |     |     |     |
  and carcinoma.......|22/70| 46 |35/50|(a)88|27/50|(a)87|
Forestomach--Squamous-|     |    |     |     |     |     |
  cell papilloma and  |     |    |     |     |     |     |
  carcinoma...........| 1/70|  2 | 6/50|(a)20| 8/50|(a)33|
Harderian gland--     |     |    |     |     |     |     |
  Adenoma and         |     |    |     |     |     |     |
  adenocarcinoma......| 6/70| 13 |27/50|(a)72|23/50|(a)82|
Preputial gland--     |     |    |     |     |     |     |
  Carcinoma...........| 0/70|  0 | 1/50|   3 | 5/50|(a)21|
Kidney--Renal tubular |     |    |     |     |     |     |
  adenoma.............| 0/70|  0 | 5/50|(a)16| 1/50|   5 |

                           |         Exposure          |
          Tumor            |   312 ppm,  |   625 ppm,  |
                           |    52 wk    |    26 wk    |
                           |  I   |  R   |  I   |  R   |
Lymphoma...................| 15/50| (a)55| 37/50| (a)90|
Heart--Hemang-iosarcoma....| 33/50| (a)87| 13/50| (a)76|
Lung--Alveolar-bronchiolar |      |      |      |      |
  adenoma and carcinoma....| 32/50| (a)88| 18/50| (a)89|
Forestomach--Squamous-cell |      |      |      |      |
  papilloma and carcinoma..| 13/50| (a)52| 11/50| (a)63|
Harderian gland--Adenoma   |      |      |      |      |
  and adenocarcinoma.......| 28/50| (a)86| 11/50| (a)70|
Preputial gland--Carcinoma.|  4/50| (a)21|  3/50| (a)31|
Kidney--Renal tubular      |      |      |      |      |
  adenoma..................|  3/50| (a)15|  1/50|    11|
  From Melnick et al(1090).
  Footnote(a) Increased compared with chamber controls (0ppm),
p< 0.05mm based on logistic regression analysis.
  Footnote(c) Mortality adjusted tumor rates are adjusted for
competing causes of mortality, such as death due to other tumors,
whose rates differ by exposure group.

Activated Oncogenes

The presence of activated oncogenes in the exposed groups which differ from those seen in tumors in the control group can help in identifying a mechanistic link for BD carcinogenicity. Furthermore, certain activated oncogenes are seen in specific human tumors and K-ras is the most commonly detected oncogene in humans. In independent studies, tumors from this study were evaluated for the presence of activated protooncogenes. (Ex. 129) Activated K-ras oncogenes were found in 6 of 9 lung adenocarcinomas, 3 of 12 hepatocellular cancers and 2 of 11 lymphomas in BD exposed mice. Nine of these 11 K-ras mutations, including all six of those seen in lung tumors, were G to C conversions in codon 13. Activation of K-ras genes by codon 13 mutations has not been detected in lung or liver tumors or lymphomas in unexposed B6C3F(1) mice, but activation by codon 12 mutation was observed in 1 of 10 lung tumors in unexposed mice. (Ex. 129)


All of the four animal bioassays (one rat, three mouse) find a clear carcinogenic response; together they provide sufficient evidence to declare BD a known animal carcinogen and a probable human carcinogen. The three mouse studies, all with a positive lymphoma response, further support a finding that the mouse is a good model for BD related lymphatic/hematopoietic and other site tumorigenicity. The most recent NTP II study confirms and strengthens the previous NTP I and Irons et al. mouse studies, and presents clear evidence that BD is a potent multisite carcinogen in B6C3F(1) mice of both sexes. (Ex. 23-1;32-28D, Irons) The finding of lung tumors at exposures as low as 6.25 ppm, 100 fold lower than the lowest exposure of the NTP I study and a level that is in the occupational exposure range, increases concern for workers' health. Two other concerns raised by both the second NTP and the Irons et al. studies are, (1) substantial carcinogenicity is found with less-than-lifetime exposures (as low as 12 or 13 weeks) for lymphomas and hemangiosarcomas, at least at higher concentrations, and, (2) for lymphomas and at least two other sites, there appears to be a dose-rate effect, where exposure to higher concentrations for a shorter time yields higher tumor response (by a factor of as much as 2-3) than a comparable total exposure spread over a longer time. These findings suggest that even short-term exposures should be as low as possible. Positive studies for genotoxicity and the detection of activated K-ras oncogenes in several of these tumors induced in mice, including lymphomas, liver, and lung, suggest a mutagenic mechanism for carcinogenicity, and support reliance on a linear low-dose extrapolation procedure (on the basis of the multistage mutagenesis theory of carcinogenicity), at least for these tumor sites. The finding of activated K-ras oncogenes in these mouse tumors may also be relevant to humans, because K-ras is the most commonly detected oncogene in humans.

The different dose-rate trends for different tumor sites suggest that different mechanisms are involved at different sites. The observation of a highly nonlinear exposure-response for lymphomas at exposure levels of 625 ppm and above suggests a secondary high-exposure mechanism as well, not merely a metabolic saturation, as is suspected with the high-exposure saturation seen in Sprague-Dawley rats. (Ex. 34-6, Owen and Glaister) The picture emerges of BD as a potent genotoxic multisite carcinogen in mice, far more potent in mice than in rats.

With respect to appropriate tumor sites for risk extrapolation from mouse to humans, Melnick and Huff have presented information comparing animal tumor response for five known or suspected human carcinogens--BD, benzene, ethylene oxide, vinyl chloride, and acrylonitrile. (Ex. 117-2) BD, benzene, and ethylene oxide all have strong occupational epidemiology evidence of increased lymphatic/hematopoietic cancer (LHC) mortality and all three cause both LHC, lung, Harderian gland, and mammary gland tumors in mice, plus several other primary tumors (see Table V-3). Only BD and vinyl chloride cause mouse hemangiosarcomas, BD in the heart and vinyl chloride in the liver. In rats, while all five carcinogens cause tumors at multiple sites, only brain and Zymbal gland tumors are associated with as many as four of the compounds. In general mice and rats are affected at different tumor sites by these carcinogens. LHC, lung, Harderian gland, mammary gland and, possibly hemangiosarcomas are sites in mice which correlate well with human LHC. This suggests that mice, rats and humans may have different target sites for the same carcinogen, but that compounds which are multisite carcinogens in the mouse and rat are likely to be human carcinogens as well. Based on BD's strong LHC association in humans, and its multisite carcinogenicity in the mouse, including occurrence at several of the same target sites seen with other carcinogens, OSHA concludes that the mouse is a good animal model for predicting BD carcinogenesis in humans.

                             [From Ex. 117-2]

                       | 1,3-Butadiene |   Benzene   | Ethylene oxide |
        Site           |  Mice | Rats  | Mice | Rats |  Mice |  Rats  |
Lymphatic/hematopietic.|   *   |       |   *  |      |   *   |   *    |
Lung...................|   *   |       |   *  |      |   *   |        |
Heart..................| (f)*  |       |      |      |       |        |
Liver..................|   *   |       |   *  |      |       |        |
Forestomach............|   *   |       |   *  |   *  |       |        |
Harderian gland........|   *   |       |   *  |      |   *   |        |
Ovary..................|   *   |       |   *  |      |       |        |
Mammary gland..........|   *   |   *   |   *  |      |   *   |        |
Preputial gland........|   *   |       |   *  |      |       |        |
Brian..................|       |   *   |      |      |       |   *    |
Zymbal gland...........|       |   *   |   *  |   *  |       |        |
Uterus.................|       |   *   |      |   *  |   *   |        |
Pancreas...............|       |   *   |      |      |       |        |
Testis.................|       |   *   |      |      |       |        |
Thyroid gland..........|       |   *   |      |      |       |        |

                       | Vinyl chloride |  Acrylonitrile |
        Site           |  Mice |  Rats  |  Mice  |  Rats |
Lymphatic/hematopietic.|       |        |    NS  |       |
Lung...................|   *   |        |        |       |
Heart..................|       |        |        |       |
Liver..................|(a)*   | (a)*   |        |       |
Forestomach............|       |    *   |        |    *  |
Harderian gland........|       |        |        |       |
Ovary..................|       |        |        |       |
Mammary gland..........|   *   |        |        |    *  |
Preputial gland........|       |        |        |       |
Brain..................|       |    *   |        |    *  |
Zymbal gland...........|       |    *   |        |    *  |
Uterus.................|       |        |        |       |
Pancreas...............|       |        |        |       |
Testis.................|       |        |        |       |
Thyroid gland..........|       |        |        |       |
  NS, not studied.

2. Epidemiologic Studies

(i) Introduction. OSHA has concluded that the epidemiologic studies contained in this record, as well as the related hearing testimony and record submissions, show that occupational exposure to BD is associated with an increased risk of death from cancers of the Lymphohematopoietic (LH) system. However, in contrast to the available toxicologic data, our understanding of BD epidemiology is based on observational studies, not experimental ones. In other words, the investigators who conducted these epidemiologic studies did not have control over the exposure status of the individual workers. They were, nonetheless, able to select the worker populations and the observational study design.

Cohort and case control studies are two types of observational study designs. Each of these designs has strengths and weaknesses that should be considered when the results are interpreted. Cohort studies, for example, have the advantages of decreasing the chance of selection bias regarding exposure status and providing a more complete description of all health outcomes subsequent to exposure. The disadvantages of cohort studies include the large number of subjects that are needed to study rare diseases and the potentially long duration required for follow-up. By comparison, case control studies are well suited for the study of rare diseases and they require fewer subjects. The disadvantages of case control studies, however, include the difficulty of selecting an appropriate control group(s), and the reliance on recall or records for information on past exposures. Regardless of the selected observational study design, the greatest limitation of occupational epidemiologic studies is their ability to measure and classify exposure.

In spite of the inherent limitations of observational epidemiologic studies, guidelines have been developed for judging causal association between exposure and outcome. Criteria commonly used to distinguish causal from non-causal associations include: Strength of the association as measured by the relative risk ratio or the odds ratio; consistency of the association in different populations; specificity of the association between cause and effect; temporal relationship between exposure and disease which requires that cause precede effect; biologic plausibility of the association between exposure and disease; the presence of a dose-response relationship between exposure and disease; and coherence with present knowledge of the natural history and biology of the disease. These criteria have been considered by OSHA in the development of its conclusion regarding the association between BD and cancer of the LH system.

As stated previously, each type of epidemiologic study design has strengths and weaknesses. Since epidemiologic studies are observational and not experimental, each study will also have inherent strengths and weaknesses; there is no perfect epidemiologic study. The most convincing evidence of the validity and reliability of any epidemiologic study comes with replication of the study's results.

There are six major epidemiologic studies in the record that have examined the relationship between occupational exposure to BD and human cancer. These studies include: A North Carolina study of rubber workers (Ex. 23-41; 23-42; 23-4; 2-28; 23-27; 23-3); a Texaco study of workers at a BD production facility in Texas (Ex. 17-33; 34-4; 34-4); a NIOSH study of two plants in the styrene-butadiene rubber (SBR) industry (Ex. 2-26; 32-25); the Matanoski cohort study of workers in SBR manufacturing (Ex. 9; 34-4); the nested case-control study of workers in SBR manufacturing conducted by Matanoski and Santos-Burgoa (Ex. 23-109); and a follow-up study of synthetic rubber workers recently completed by Delzell et al. (Ex. 117-1). Several comments in the record have concluded that these studies demonstrate a positive association between occupational exposure to BD and LH cancers. However, OSHA has been criticized by the Chemical Manufacturers Association (CMA) and the International Institute of Synthetic Rubber Producers, Inc. (IISRP) for its interpretation of these studies as showing a positive association; the chief criticisms will be discussed below. (Ex. 112 and 113) OSHA's final consideration of the BD epidemiologic studies is organized and presented according to what have been identified as key issues. These are the epidemiologic issues that were raised and considered throughout the rulemaking. They are also the issues most pertinent to OSHA's conclusions. These key issues surrounding BD exposure and LH cancer are: Evidence of an association; observation of a dose-response relationship; observation of short latency periods; the potential role of confounding exposures and the observed study results; the biological basis for grouping related LH cancers; relevance of subgroup analyses; and appropriateness of selected reference populations.

(ii) Evidence of an Association Between BD and LH Cancer. Each of the studies listed above contributes to the epidemiologic knowledge upon which OSHA's conclusion regarding the relationship of BD exposure and LH cancer has been developed.

(a) North Carolina Studies. This series of studies was undertaken to examine work-related health problems of a population of workers in a major tire manufacturing plant. They were not designed to look specifically at the health hazards of BD. (Tr. 1/15/91, p. 117) However, in a work area that involved the production of elastomers, including SBR, relative risks of 5.6 for lymphatic and hematopoietic malignancies and 3.7 for lymphatic leukemia were found among workers employed for more than five years. The International Agency for Research on Cancer (IARC) evaluation concluded that this study suggests an association between lymphatic and hematopoietic malignancy and work in SBR manufacturing. (Tr. 1/15/91, p. 117) However, the IISRP asserted that these studies do not provide "meaningful evidence of an association between butadiene and cancer." (Ex. 113, p. A-23) OSHA recognizes that the researchers who conducted these studies acknowledged that the workers may have had exposures to organic solvents, including benzene, a known leukemogen, as pointed out by the IISRP. (Ex. 113, p. A-24) (b) Texaco Study. The two Texaco studies examined mortality of a population of workers in a BD manufacturing facility in Texas. (Ex. 17-33; 34-4 Vol. III, H-2; Divine 34-4, Vol. III, H-1) A qualitative method of exposure classification, based on department codes and expert consensus judgement, was used in the Downs study. (Ex. 17-33; 34-4, Vol. III, H-2) From this methodology four exposure groups were defined: Low exposure, which included utility workers, welders, electricians, and office workers; routine exposure, which included process workers, laboratory personnel, and receiving, storage and transport workers; non-routine exposure, which included skilled maintenance workers; and unknown exposure, which included supervisors and engineers. It is OSHA's opinion that although this is a crude approach to exposure classification, there are important findings in this study that contribute to our understanding of BD epidemiology.

In the Downs study (Ex. 34-4, Vol. III, H-2) the standardized mortality ratio (SMR) for all causes of death in the entire study cohort was low (SMR 80; p < .05) when compared to national population rates. However, a statistically significant excess of deaths was observed for lymphosarcoma and reticulum cell sarcoma combined (SMR 235; 95% confidence interval (CI) = 101,463) when compared with national population rates. (The issue of reference population selection is discussed below in paragraph (viii).) When analyzed by duration of employment, the SMR for the category of all LH neoplasms was higher in workers with less than five years employment (SMR = 167) than for those with more than five years employment (SMR = 127). (Ex. 34-4, Vol. III, H-2) However, neither of these findings was statistically significant. Alternatively, it has been suggested that perhaps the short-term workers were wartime workers, and that these workers were actually exposed to higher levels of BD, albeit for a shorter time. (Tr. 1/15/91, p. 119) Analyses of the four exposure groups also showed elevated but not statistically significant SMRs. The routine exposure group had a SMR of 187 for all LH neoplasms, explained primarily by excesses in Hodgkin's disease (SMR = 197) and other lymphomas (SMR = 282). (Ex. 34-4, Vol. III, H-2) Those workers in the non-routine exposure group also had an elevated SMR for all LH neoplasms (SMR = 167), with excess mortality for Hodgkin's disease (SMR = 130), leukemias (SMR = 201), and other lymphomas (SMR = 150) (Ex. 34-4, Vol. III, H-2).

These data were updated by Divine by extending the period of follow-up from 1979 through 1985. (Ex. 34-4, Vol. III, H-1) The SMR for all causes of mortality remained low (SMR = 84, 95% CI = 79,90), as it did for mortality from all cancers (SMR = 80, 95% CI = 69,94). (Ex. 34-4, Vol. III, H-1) However, the SMR for lymphosarcoma and reticulosarcoma combined was elevated and statistically significant (SMR = 229, 95% CI = 104,435). This finding was consistent with the previous analyses done by Downs. (Tr. 1/15/91, p. 120).

Exposure group analyses were also consistent with the previous findings by Downs. The highest levels of excess mortality from lymphatic and hematopoietic malignancy were again seen in the routine and non-routine exposure groups. The routine exposure group that was "ever employed" had a statistically significant excess of lymphosarcoma (SMR = 561, 95% CI = 181,1310), that accounted for most of the LH excess. (Ex. 34-4, Vol. III, H-1) The cohort of workers employed before 1946 (wartime workers) also demonstrated a statistically significant excess of mortality due to lymphosarcoma and reticulosarcoma combined (SMR = 269, 95% CI = 108,555). (Ex. 34-4, Vol. III, H-2) In summary, the Texaco study provides several notable results. The first of these is the consistently elevated mortality for lymphosarcoma. This finding is consistent with excess lymphomas observed in experimental mice. (Ex. 23-92) Second, the excess risk of mortality was found in the routine and non-routine exposure groups. Based on the types of jobs held by workers in these two exposure groups, this finding suggests that the incidence of lymphatic malignancy is highest in the groups with the heaviest occupational exposure to BD. (Tr. 1/15/91, p. 121) The third notable result of this study was the significantly elevated rate of malignancy in workers employed for fewer than 10 years.

(c) NIOSH Study. The NIOSH study was undertaken in January 1976 in response to the report of deaths of two male workers from leukemia. (Ex. 2-26; 32-25) These workers had been employed in two adjacent SBR facilities (Plant A and Plant B) in Port Neches, Texas. The hypothesis tested by this study is that:

Employment in the SBR production industry was associated, specifically, with an increased risk of leukemia and, more generally, with an increased risk of other malignancies of hematopoietic and lymphatic tissue. (Ex. 2-26) This study did not specifically examine the association between BD and all LH cancers. Thus, OSHA agrees with the criticism that this study by itself did not demonstrate that occupational exposure to BD causes cancer. (Ex. 113, pp. A-13, A-19) However, the findings in this study are consistent with the patterns observed in the other epidemiologic studies discussed in this section. In Plant A, the overall mortality was significantly decreased (SMR=80, p< 0.05). (Ex. 2-26) The SMR for all malignant neoplasms was also decreased (SMR=78), but this result was not statistically significant. (Ex. 2-26) The SMR for LH cancers was elevated (SMR=155), as it was for lymphosarcoma and reticulum cell sarcoma (SMR=181) and leukemia (SMR=203), but none of these results was statistically significant. (Ex. 2-26) The pattern of mortality for a subgroup of wartime workers was also examined for the Plant A population. For this subgroup of white males, employed at least six months between the beginning of January 1943 and the end of December 1945, there was an elevated SMR for lymphatic and hematopoietic neoplasms (SMR = 212) that was statistically significant at the level of 0.05< p< 0.1. (Ex. 2-26) Likewise, the SMR for leukemia was increased (SMR=278), also with statistical significance at the level of 0.05< p< 0.1. (Ex. 2-26) At Plant B, the overall mortality was low (SMR=66, p< 0.05), as was death from all malignant neoplasms (SMR=53, p< 0.05). (Ex. 2-26) The SMR for LH cancer was also low (SMR=78), but this finding was not statistically significant. (Ex. 2-26) When this study was updated, the mortality patterns remained unchanged. (Ex. 32-25) The most remarkable findings of the NIOSH study are the excess mortality for malignancies of the LH system, and the excess of these cancers in workers employed during the wartime years.

(d) Matanoski Cohort Study. The cohort study conducted by Matanoski et al. is comprised of two follow-up periods: In the original study, completed in June 1982, the cohort was followed from 1943 to 1979; and in the update, completed in March 1988, the cohort follow-up period was extended to 1982. (Ex. 9; 23-39; 34-4, Vol. III, H-3 and H-6, respectively) The original study analyzed mortality data for 13,920 male workers employed for more than one year in eight SBR production plants in the United States and Canada. Although historical quantitative exposure data were not available, creation of a job dictionary made it possible to designate four general work activities as surrogates for exposure: Production; utilities; maintenance; and a combined category of all other jobs. The work activities with the highest BD exposures were production and maintenance. (Ex. 16-39) The total duration worked was measured by the dates of first and last employment.

The mortality experience for the original study cohort, as compared with death rates for males in the United States, was low for all causes (SMR=81) and all cancers (SMR=84). (Ex. 9; 23-39) The SMR for all LH cancers was also low (SMR=85). (Ex. 9; 23-39) The mortality rate for Hodgkin's disease was slightly elevated (SMR=120), but it was not statistically significant. (Ex. 9; 23-39) In fact, there were no statistically significant excesses in mortality from cancer at any site found in this original cohort study.

These initial data were also analyzed according to major work area. There were not any elevations of mortality rates for the category of all LH cancers. (Ex. 9; 23-39) For production workers, the SMR for other lymphatic cancers was elevated (SMR=202), but it was not statistically significant. (Ex. 9; 23-39) The SMR for leukemia in the utilities work group was also elevated (SMR=198), but it was based on only two deaths and was not statistically significant. (Ex. 9; 23-39) Slight excesses, none of which was statistically significant, were seen for Hodgkin's disease in each of the four work group categories. (Ex. 9; 23-39) OSHA has been criticized for its opinion, expressed in the preamble of the BD proposed rule, that the original Matanoski cohort study did not have sufficient power to detect a difference in the cancer SMR if one actually existed. (Ex. 113, pp. A-10-11) Statistical power of at least 80% is the accepted rule-of-thumb for epidemiologic research study design. Calculations provided by Matanoski indicate that, for the outcomes of greatest concern to OSHA, statistical power was often below the 80% level. (Ex. 9) For leukemia, statistical power to detect 25% and 50% increases in mortality was only 27% and 62%, respectively. (Ex. 9) The power to detect a 25% increase in mortality for all lymphohematopoietic cancers was only 49%. (Ex. 9) However, the study did have a statistical power of 93% to detect a SMR of 150 for all LH cancers. (Ex. 9) Thus, for the cancers of most interest to OSHA, this study had limited statistical power to detect mortality excesses that were less than two-fold. OSHA does not consider this to be an "unrealistically strict standard of acceptability," as alleged by the IISRP, but rather part of a thorough critique of an epidemiologic study with purportedly "negative results." (Ex. 113, p. A-11) The update of Matanoski's original study extends the period of cohort follow-up from 1979 to 1982, providing a full 40 years of mortality experience for analysis. The update study cohort differed from the original cohort in two additional ways: Canadian workers with relatively short-term exposure were removed from the cohort; and the proportion of workers lost to follow-up was reduced. The extension of follow-up resulted in findings of excess mortality from lymphatic and hematopoietic cancers that had not been observed in the original analyses. (Ex. 34-4, Vol. III, H-6) The SMR for all causes of mortality remained low (SMR=81, 95% CI=78,85), as it did for death from all cancers (SMR=85, 95% CI=78,93). (Ex. 34-4, Vol. III, H-6) For lymphatic and hematopoietic cancers, the overall SMR for white males was not increased (SMR=92, 95% CI=68,123). (Ex. 34-4, Vol. III, H-6) However, for black males, the SMR for all LH cancers was elevated (SMR=146, 95% CI=59,301). (Ex. 34-4, Vol. III, H-6) Specific increases were also found for lymphosarcoma (SMR=132), leukemia (SMR=218, 95% CI=59,560), and other lymphatic neoplasms (SMR=116, 95% CI=14,420). (Ex. 34-4, Vol. III, H-6) These increases were based on small numbers of observed deaths.

Analyses conducted on the four exposure groups also produced some evidence of excess mortality. For the total cohort of production workers, an elevated SMR was observed for all lymphopoietic cancers (SMR=146, 95% CI=88,227). (Ex. 34-4, Vol. III, H-6) For white production workers, the SMR for that category was 110, explained principally by excess mortality from other lymphatic neoplasms (SMR=230, 95% CI=92,473). (Ex. 34-4, Vol. III, H-6) Although based on small numbers, the results for black production workers were more pronounced and statistically significant: The SMR for all lymphatic and hematopoietic cancers was 507 (95% CI=187,1107). (Ex. 34-4, Vol. III, H-6) That overall increase in black workers reflected excess mortality from lymphosarcoma (SMR=532), leukemia (SMR=656, 95% CI=135,1906), and other lymphatic cancers (SMR=482, 95% CI=59,1762). (Ex. 34-4, Vol. III, H-6) A pattern of excess mortality for all LH cancers was also seen in utility workers (SMR=203, 95% CI=66,474). (Ex. 34-4, Vol. III, H-6) That elevated SMR may be explained by elevated rates for leukemia (SMR=192, 95% CI=23,695) and other lymphatic cancers (SMR=313, 95% CI=62,695). (Ex. 34-4, Vol. III, H-6) No increases in LH malignancy were seen in the other exposure groups, i.e., maintenance or other workers.

From these study results Matanoski et al. concluded:

Deaths from cancers of the hematopoietic and lymphopoietic system are higher than expected in production workers with significant excesses for leukemias in black workers and other lymphomas in all (production) workers. (Ex. 34-4, Vol. III, H-6, p. 116)

In response to criticism from the IISRP that OSHA placed too much emphasis on the findings in the group of black production workers, OSHA is aware of the statement offered by the researchers that because of the potential for bias from misclassification of race: "* * * the total SMRs are probably the most correct representation of risk." (Ex. 34-4, Vol. III, H-6) However, OSHA also agrees with the authors that the risk of death from LH cancers seems to be higher in this SBR industry population than in the general population, and these causes of death seem to be associated with different work areas. These cohort study findings stimulated the design and implementation of the Santos- Burgoa and Matanoski nested case-control study.

(e) Santos-Burgoa and Matanoski Nested Case-Control Study. To further investigate the findings of the cohort study, Santos-Burgoa and Matanoski et al. designed and conducted a case-control study of LH cancers in workers in the styrene-butadiene polymer manufacturing industry. (Ex. 23-109; 34-4, Vol. III, H-4) The specific questions addressed by this research study are: "Is there a risk of any lymphatic or hematopoietic cancer which is associated with exposure to (BD) or styrene or both?"; and "is there a risk of these cancers related to exposure to jobs within the industry?" (Ex. 34-4, Vol. III, H-4) This is the first study to specifically investigate the association between LH cancers and individual worker exposure to BD, which is why, contrary to the opinion of IISRP, OSHA places so much "weight" on these results. (Ex. 113, pp. A-25-34) The subjects in this case-control study were "nested," or contained, within the population of the original cohort study. "Cases" in this study were defined as males who worked one year or more at any of eight synthetic rubber polymer producing plants and who died of or with a lymphopoietic cancer. These cancers included: Lymphosarcoma and reticulum cell sarcoma, Hodgkin's lymphoma, non-Hodgkin's lymphoma, all leukemias, multiple myeloma, polycythemia vera, and myelofibrosis. Sixty-one cases were identified, but two cases were omitted from data analyses, resulting in a total of 59 cases. One case was omitted because he could not be matched to controls, and the other case lacked job records from which exposure could be identified.

Eligible "controls" included workers who were either alive or had died of any cause other than malignant neoplasms, who had been employed at one of the eight SBR plants, and who had not been lost to follow-up. These controls were individually matched to cases on the following criteria: Plant; age; hire year; employment as long or longer than the case; and survival to the death of the case. The study aim was to select four controls per case. Even though this was not always possible, there were, on average, just over three controls per case in each group of lymphopoietic cancer. The total number of controls was 193.

Unlike the previous studies, in this research study an exposure measurement value for BD (and also for styrene) was determined for each case and control. This value was determined by a multi-step process. First, the job records of each subject were reviewed and the number of months that each job was held was determined. Second, the level of BD (and styrene) associated with the job was estimated by a panel of five industrial experts, i.e., engineers with long term experience in SBR production. The exposure level for BD (and styrene) for each job was based on a scale of zero to ten, with ten being the rank given to the job with the highest exposure. The next step in the development of each individual job-exposure matrix was to add all of the exposures to the chemicals for all the months a specific job was held and then sum the exposures over a working lifetime. This procedure resulted in a cumulative BD exposure value for each case and control.

The distribution of the cumulative exposure estimates for the study population was not normally distributed, i.e., there were some extreme values. In order to approximate a normal distribution, a required assumption for many statistical analyses, a logarithmic transformation of these values was done. (Ex. 34-4, Vol. III, H-4) Exposure was analyzed as a dichotomous variable, i.e., ever/never exposed. "Exposed" workers were defined as those with a log rank cumulative exposure score above the mean of the scores for the entire population of cases and controls within a cancer subtype; "non-exposed" workers were those with a score below the mean.

There were several important findings in this study. First, in the unmatched analysis of cases and controls, the leukemia subgroup had a significant excess risk of 6.8 fold for exposure to BD among cases compared to controls (Odds Ratio (OR)=6.82, 95% CI=1.10,42.23). (Ex. 34-4, Vol. III, H-4) The results were even stronger in the matched-pair analyses. In that analysis for exposure to BD, the OR was 9.36 (95% CI=2.05,22.94) in the leukemia subgroup. (Ex. 34-4, Vol. III, H-4) This result can be interpreted to mean that cases with leukemia were more than nine times as likely as their controls to be exposed to BD. Additionally, the data in this analysis indicate that BD exposure above the group mean is 2.3 times (OR=2.30, 95% CI=1.13,4.71) more common among cases with all lymphopoietic cancers when compared to a similar exposure in the controls. (Ex. 34-4, Vol. III, H-4) This case-control study has been the subject of criticism that has centered on both validity and reliability. (Ex. 23-68; 113) For example, the data from this study have been criticized as being "inconsistent" with the results of the Matanoski cohort study. (Ex. 23-68; 113, p. A-25) Further, it has been suggested that "the study results are not reliable and should not be relied upon by OSHA.: (Ex. 113, p. A-25) OSHA rejects these criticisms for the reasons discussed below.

First, regarding the issue of inconsistency, a nested case control study does not test the same hypotheses or make the same comparisons as a cohort study. (Ex. 32-24; Tr. 1/15/91, p. 161; Tr. 1/16/91, p. 347) In fact, as presented in the above discussions of the studies, they ask and answer different research questions. For example, the cohort study asked whether all of the SBR workers have a different risk of leukemia from the general population, and the case control study asked whether workers with leukemia have different exposures within the industrial setting from workers without leukemia. (Ex. 32-24) Thus, the criticism that the results of these two studies are incompatible, and therefore invalid, is not relevant. (Ex. 32-24) Second, the challenge directed at the reliability of the case-control study does not hold up under close scrutiny. This criticism is based on four issues: Log transformation of the exposure data; instability of the results; irregular dose-response pattern; and selection criteria for "controls." (Ex. 113, A-29-34) Regarding the log transformation of the exposure data, the IISRP asserts that there is not a sound rationale for this approach to data analyses. (Ex. 113, A-29-30) However, Santos-Burgoa offered the following explanation of this procedure in his testimony:

For analysis, exposures were categorized in advance above and below the mean of the cumulative exposure for the study subjects. This cutpoint was defined from the very beginning of the analysis design as follows. The total cumulative exposures, as happens in most environmental exposures, showed a skewed distribution with many observations at the low levels and few at the high levels. Since the geometric mean is the best estimate of the central tendency point in log normal data, such as exposure data, the cumulative exposures were transformed by the logarithm, and then the mean was calculated. (Ex. 40, pp. 12-13) It is OSHA's opinion that, given the log normal distribution of the exposure data, Santos-Burgoa chose the best approach for data analyses.

The case-control study has also been criticized for producing "highly unstable and therefore unreliable" results. (Ex. 113, A-30) For example, the leukemia subgroup (matched-pair analysis) OR of 9.36 with a 95% confidence interval of 2.05-22.94 has been used to illustrate statistical instability of the data. (Ex. 113, A-31) However, as previously discussed, the disease category of "all lymphopoietic cancers" (matched-pair analysis) had an OR of 2.30 with a confidence interval of 1.13-4.71. Thus, it is OSHA's opinion that, while some specific odds ratios may have wide confidence intervals, the study results as a whole are not "unreliable."

The IISRP has also criticized the case-control study for "* * * fail(ing) to demonstrate a dose-response relationship * * *" (Ex. 113, A-32) However, the test for linear trend, i.e., test for dose-response, shows a statistically significant, but irregular, trend in the odds of leukemia with increasing levels of exposure to BD. Specifically, as exposure levels increase the pattern of odds ratios is: 7.2; 4.9; 13.0; 2.5; and 10.3. (Ex. 23-109, Table 10) Although this is not a compelling linear dose-response, in OSHA's opinion, it is suggestive of a pattern of increasing disease risk at increasing exposure levels.

Inconsistent application of the control selection criteria is the final criticism directed at the case-control study by the IISRP. (Ex. 113, A-33) However, careful review of docket exhibits related to the case-control study reveals this criticism to be unfounded. In his dissertation, Santos-Burgoa clearly states the protocol for control selection:

All cohort subjects were arranged into groups by plants, date of birth, date of hire, duration of work and duration of follow-up. A two and a half year period around each time variable was relaxed in a few instances when no more controls were available. One lymphosarcoma case was lost since no match was found for his date of birth, even allowing for three and a half years around the date. This was the only case lost to analysis because of lack of a matched control. (Ex. 32-25, p. 80)

With only 59 cases, Santos-Burgoa was correctly concerned about loss of valuable data should any additional cases need to be eliminated due to lack of a match. Also, regarding the potential for bias, abstractors were blinded to case or control status when employment data were being collected. (Ex. 34-4, Vol. III, App. H-5) Thus, it is most likely that any misclassification bias would be nondifferential, biasing the study results towards the null.

(f) Delzell et al. Follow-up Study for the IISRP. The most recent study of synthetic rubber workers was conducted by Delzell et al. (Ex. 117-1) This study updated and expanded the research on SBR workers conducted by NIOSH, Matanoski et al., and Santos-Burgoa. More specifically, the Delzell et al. study consists of workers at seven of eight plants previously studied by The Johns Hopkins University (JHU) investigators, and the two plants included in the NIOSH study.

This retrospective cohort study evaluated the associations between occupational exposure to BD, styrene, and benzene and mortality from cancer and other diseases among the SBR workers. There were five study objectives:

(1) To evaluate the overall and cause-specific mortality experience of SBR workers relative to that of the USA and Canadian general populations;

(2) To assess the cancer incidence experience of Canadian synthetic rubber workers relative to that of the general population of Ontario;

(3) To determine if overall and cause-specific mortality patterns vary by subject characteristics such as age, calendar time, plant, period of hire, duration of employment, time since hire and payroll status (hourly or salaried);

(4) To examine relationships between work areas within the SBR study plants and cause-specific mortality patterns;

(5) To evaluate the relationship between exposure to BD and [styrene] and the occurrence of leukemia and other lymphopoietic cancers among SBR workers. (Ex. 117-1 p. 10)

The study population for this investigation included 17,964 male synthetic rubber workers employed in one of eight plants in either the USA or Canada. In order to be eligible for inclusion, a worker had to be employed for a total of at least one year before the closing date of the study, January 1, 1992. Additional eligibility criteria were developed for selected plants due to limitations in availability of plant records and follow-up of subjects. The eligibility criteria in this study were considered by the investigators to be more restrictive than in either the JHU or NIOSH studies. (Ex. 117-1, p. 13) Most of the exclusions were based on less than one year of employment. During the study period of 1943 through 1991, there were 4,665 deaths in the study population.

The methods used in this study included development of work history information and retrospective quantitative exposure estimates for individual members of the study population. Complete work history information was available for approximately 97% of the study cohort. There was a total of 8,281 unique "work area/job" combinations for all of the plants combined, with a range of 199 to 4,850 in specific plants. Additionally, 308 work area groups were defined based on individual plant information regarding production, maintenance, and other operations, as well as jobs and tasks within each type of operation. Five "process groups" and seven "process subgroups" were derived from the work area groups. The process groups include: Production of SBR, solution polymerization (SP), liquid polymerization (LP), and latex production; maintenance; labor; laboratories; and other operations.

Six plants had sufficiently detailed individual work history information for use in development of retrospective quantitative exposure estimates for BD and styrene. The process used to produce these exposure estimates included: In-depth walk-through surveys of each plant; meetings with plant management; interviews with key plant experts, such as individuals with long-term employment. The interviews were used to collect information regarding the production process, specific job tasks, and exposure potential. Additionally, the results of industrial hygiene monitoring from these plants were obtained. The actual exposure estimation was based on:

Specification of the exposure model; the estimation of exposure intensities for specific tasks in different time periods; the estimation of exposure intensities for generic (nonspecific) job titles (e.g., "laboratory worker") in different time periods; validation of exposure intensity estimates; the computation of job- and time period specific summary indices; and the compilation of job-exposure matrices (JEMs) for BD, [styrene], and [benzene] and linkage with subjects' work histories. (Ex. 117-1, pp. 27-28)

A limited validation of the quantitative exposure estimations was conducted, which resulted in revision of the estimates used in analyses presented in the Delzell et al. study. (Ex. 117-1)

The major findings of this study have been reported by Delzell et al. in five categories: General mortality patterns; mortality among USA subjects compared to state populations; cancer incidence; mortality patterns by process group; and mortality patterns by estimated monomer exposure. Key results from each of these categories, especially as they relate to leukemia and other LH cancers, are briefly presented.

First, regarding general mortality patterns, there were deficits in both all causes (SMR=87, 95% CI=85,90) and all cancers (SMR=93, 95% CI=87,99) for the entire cohort. (Ex. 117-1, p. 53) Of the LH cancers, excess mortality was only observed for leukemia (SMR=131, 95% CI=97-174). (Ex. 117-1, p. 53) In a cohort subgroup having 10 or more years of employment and 20 or more years since hire, the excess of leukemia deaths was even greater (SMR=201, 95% CI=134,288). (Ex. 117-1, p. 54) Analyses were also conducted to explore the possibility of racial differences in the general mortality patterns. Regarding mortality from leukemia, the SMRs were higher for blacks than for whites. In a subgroup of "ever hourly" workers with 10 or more years of work and 20 or more years since hire, the SMRs for leukemia were 192 (95% CI=119,294) for whites and 436 (95% CI=176,901) for blacks. (Ex. 117-1, p. 55) Additionally, analyses were done by specific groups of LH cancers:

Lymphosarcoma; leukemia; and other lymphopoietic cancer. For the overall cohort, there was an excess of mortality from lymphosarcoma in those members who died in 1985 and beyond (SMR=215, 95% CI=59,551). (Ex. 117-1, p. 116) This excess was observed in "ever hourly" white men; there were no lymphosarcoma deaths in blacks. (Ex. 117-1, p. 119) In the "other lymphopoietic cancer" category, the overall cohort had a slight deficit of mortality (SMR=97, 95% CI=70,132). (Ex. 117-1, p. 116) When analyzed according to racial groups, whites were also observed to have a deficit of mortality from this group of cancers (SMR=91, 95% CI=63,127). (Ex. 117-1, p. 118) Blacks, however, had an increase in mortality from "other lymphopoietic" cancers (SMR=142, 95% CI=61,279). (Ex. 117-1, p. 120) The analyses for leukemia mortality in the overall cohort showed a modest increase (SMR=131, 95% CI=97,174). (Ex. 117-1, p. 116) The increase in mortality was found primarily in the subgroups of workers who died in 1985 or later, those that worked for 10 or more years, and those with 20 or more years since hire. A dose-response type of pattern was observed among "ever hourly" subjects in the analysis of the relationship of leukemia and duration of employment: Less than 10 years worked, the SMR=95 (95% CI=53,157); 10-19 years worked, the SMR=170 (95% CI=85,304); and 20 or more years worked, the SMR=204 (95% CI=123,318). (Ex. 117-1, p. 117) Leukemia mortality was also analyzed for racial difference among "ever hourly" men. Overall, the SMR was higher for black subjects (SMR=227, 95% CI=104,431) than for white (SMR=130, 95% CI=91,181). (Ex. 117-1, p. 122) In fact, there were statistically significant elevations in the leukemia SMR for black "ever hourly" men with 20 or more years worked (SMR=417, 95% CI=135,972), and 20 to 29 years since hire (SMR=446, 95% CI=145,1042). (Ex. 117-1, p. 122) Second, Delzell et al. analyzed the mortality data of the USA cohort subgroup using both state general population rates and USA general population rates for comparison. The overall pattern of these analyses was that of "slightly lower" SMRs when the state general population rates were used. (Ex. 117-1, p. 60) For example, in the analysis for leukemia mortality, the SMR using the USA rates was 131 (95% CI not provided), and it decreased to 129 (95% CI=92,176) when state rates were applied. (Ex. 117-1, pp. 61, 136) Third, the results of the Delzell et al. study include an analysis of the cancer incidence in the Canadian plant (plant 8). Regardless of whether the cancer experience of terminated workers was included or excluded, the overall cancer incidence was not elevated in this cohort subgroup (SIR=105, 95% CI=93,117; SIR=106, 95% CI=94,119, respectively). (Ex. 117-1, pp. 61-62) However, analysis of this cohort subgroup, with the terminated workers included, "revealed an excess of leukemia cases before 1980 (overall cohort, 6 observed/3.0 expected; ever hourly, 6 observed/2.9 expected)" (further data were not provided). (Ex. 117-1, p. 62) Fourth, Delzell et al. examined mortality patterns by work process group. These analyses produced elevated SMRs for both lymphosarcoma and leukemia. There was excess lymphosarcoma mortality in field maintenance workers (SMR=219, 95% CI=88,451), production laborers (SMR=263, 95% CI=32,951), and maintenance laborers (SMR=188, 95% CI=39,548). (Ex. 117-1, pp. 65-66) However, these results were not statistically significant, and may be due to chance. For leukemia, the results were more striking: Polymerization workers had a SMR of 251 (95% CI=140,414); workers in coagulation had a SMR of 248 (95% CI=100,511); maintenance labor workers had a SMR of 265 (95% CI=141,453); and workers in laboratories had a SMR of 431 (95% CI=207,793). (Ex. 117-1, pp. 66,151) It should be noted that excess mortality by work process group was also observed for other cancers, i.e., lung cancer and larynx cancer.

Fifth, the final set of analyses performed by Delzell et al. was designed to examine mortality patterns by estimated monomer exposure, i.e., BD, styrene, and benzene. Poisson regression analyses conducted to explore the association between "BD ppm-years" and leukemia indicated a positive dose-response relationship, after controlling for styrene "ppm-years", age, years since hire, calendar period, and race. Specifically, in the cohort group that included all person-years and leukemia coded as either underlying or contributing cause of death, the rate ratios (RRs) were: 1.0, 1.1 (95% CI=0.4,5.0), 1.8 (95% CI=0.6,5.4), 2.1 (95% CI=0.6,7.1), and 3.6 (95% CI=1.0,13.2) for BD ppm-year exposure groups of 0, >0-19, 20-99, 100-199, and 200+, respectively. (Ex. 117-1, pp. 68-69; 158) Poisson regression analyses were also conducted using varying exposure categories of BD ppm-years. These analyses demonstrated a stronger and more consistent relationship between BD and leukemia than between styrene and leukemia. (Ex. 117-1, p. 69, 159) Although a clearly positive relationship between BD "peak-years" and leukemia was observed from additional Poisson regression analyses, even after controlling for BD ppm-years, styrene ppm-years, and styrene peak-years, the dose-response relationship was less clear. (Ex. 117-1, pp. 71, 162) In summary, one of the most important findings of the research of Delzell et al. was strong and consistent evidence that employment in the SBR industry produced an excess of leukemia. In the authors own words:

This study found a positive association between employment in the SBR industry and leukemia. The internal consistency and precision of the result indicate that the association is due to occupational exposure. The most likely causal agent is BD or a combination of BD and [styrene]. Exposure to [benzene] did not explain the leukemia excess. (Ex. 117-1, p. 85)

(g) Summary. These studies provide a current body of scientific evidence regarding the association between BD and LH cancers. As previously discussed, two of the criteria commonly used to determine causal relationships are consistency of the association and strength of the association. The consistency criterion for causality refers to the repeated observation of an association in different populations under different circumstances. Consistency is perhaps the most striking observation to be made from this collection of studies: "[E]very one of these studies to a greater or lesser extent finds excess rates of deaths from tumors of the lymphatic and hematopoietic system." (Tr. 1/15/91, p. 129) Strength of the association is determined by the magnitude and precision of the estimate of risk. In general, the greater the risk estimate, e.g., SMR or odds ratio, and the narrower the confidence intervals around that estimate, the more probable the causal association. In the nested case-control study, although the confidence intervals were wide, the odds ratios provide evidence of a strong association between leukemia and occupational exposure to BD.

(iii) Observation of a Dose-Response Relationship. A dose-response relationship is present when an increase in the measure of effect (response), e.g., SMR or odds ratio, is positively correlated with an increase in the exposure, i.e., estimated dose. When such a relationship is observed, it is given serious consideration in the process of determining causality. However, the absence of a dose-response relationship does not necessarily indicate the absence of a causal relationship.

OSHA has been criticized for its conclusion that the epidemiologic data suggest a dose-response relationship. (Ex. 113) The IISRP offers a different interpretation of the data. In their opinion, the data provide a "consistent finding of an inverse relationship between duration of employment and cancer mortality." (Ex. 113, A-34) This observation is further described by John F. Acquavella, Ph.D., Senior Epidemiology Consultant, Monsanto Company, as "the paradox of butadiene epidemiology." (Ex. 34-4, Vol. I, Appendix A) This interpretation assumes that cumulative occupational exposure to BD will increase with duration of employment, and, thus, cancer mortality will increase with increasing duration of employment. (Ex. 113, A-35-39) In OSHA's opinion, this is an erroneous assumption; the epidemiologic data for BD tell a different story. For the workers in these epidemiologic studies, it is unlikely that occupational exposure to BD was constant over the duration of employment. According to Landrigan, BD exposures were most likely higher during the war years than they were in subsequent years. (Tr. 1/15/91, p.146) It is logical that exposures would be especially intense during this time period because of wartime production pressures, the process of production start-up in a new industry, and the general lack of industrial hygiene controls during that phase of industrial history. Unfortunately, without quantitative industrial hygiene monitoring data, the true levels of BD exposure for wartime workers cannot be ascertained. In the absence of such data, however, OSHA believes it is reasonable to consider wartime workers as a highly exposed occupational subgroup. (Tr. 1/15/91, p. 121; Tr. 1/16/91, pp. 225-227) Thus, the excess mortality seen among these workers provides another piece of the evidence to support a dose-response relationship between occupational exposure to BD and LH cancers.

Additional support that excess mortality, among workers exposed to BD, is dose-related can be found in the analyses of the work area exposure groups. The studies by Divine, Matanoski, and Matanoski and Santos-Burgoa all provide evidence that excess mortality is greatest among production workers. (Ex. 34-4, Vol. III, H-1; 34-4, Vol. III, H-6; 23-109, respectively) Production workers are typically the most heavily exposed workers to potentially toxic substances. (Ex. 34-4) The most compelling data that support the existence of a dose-response relationship for occupational exposure to BD and LH cancers are those in the study by Delzell et al. (Ex. 117-1) Analysis of the cumulative time-weighted BD exposure in ppm-years indicates a relative risk for all leukemias that increases positively with increasing exposure. This relationship is present even with statistical adjustment for age, years since hire, calendar period, race, and exposure to styrene. It is OSHA's opinion that identification of a positive dose-response in an epidemiologic study is a very powerful observation in terms of causality.

(iv) Observation of Short Latency Periods. Short latency periods, i.e., time from initial BD exposure to death, were seen in two epidemiologic studies. In the NIOSH study, three of the six leukemia cases had a latency period from three to four years. (Ex. 2-26) Additionally, five of these six workers were employed prior to 1945. (Ex. 2-26) In the Texaco study update, a latency of less than 10 years was seen in four of the nine non-Hodgkin's lymphoma (lymphosarcoma) cases, and seven of these workers were also employed during the wartime years. (Ex. 34-4, Vol. III, H-1) According to OSHA's expert witness, Dr. Dennis D. Weisenburger,

these findings are contrary to the accepted belief that, if a carcinogen is active in an environment, one should expect the * * * SMRs to be higher for long-term workers than for short-term workers (i.e., larger cumulative dose). (Ex. 39, p. 9)

Thus, it has been argued that these findings appear to lack coherence with what is known of the natural history and biology of LH cancers. (Ex. 113, A-40-42) Furthermore, these findings have been interpreted as evidence against a causal association between BD and these LH cancers. (Ex. 113, A-42) In OSHA's opinion, there are other possible explanations for these observations. First, as proffered by Dr. Weisenburger, a median latency period of about seven years has been found for leukemia in studies of atomic bomb victims, radiotherapy patients, and chemotherapy patients who have received high-dose, short-term exposures. (Ex. 39) In contrast, Dr. Weisenburger points out that low-dose exposure to an environmental carcinogen, such as benzene, has a median latency period for leukemia of about 15-20 years. (Ex. 39) He concludes that short-term, high-dose exposures may be associated with a short latency period, whereas long-term, low-dose exposures may be associated with a long latency period.

Second, the occurrence of short latency periods for LH cancer mortality in these two studies was concentrated in workers first employed during the wartime years. As previously discussed, it is possible that exposure to BD during the wartime years was greater than in subsequent years. (Ex. 39; Tr. 1/15/91, p. 121) Dr. Weisenburger suggests that the "short latency periods for LH cancer in these studies may be explained by intense exposures to BD over a relatively short time period." (Ex. 39, p. 10) In his testimony, Dr. Landrigan, another OSHA expert witness, makes the point that "duration of employment is really only a crude surrogate for total cumulative exposures, not itself a measure of exposure." (Tr. 1/15/91, p. 121) In other words, it is possible that short-term workers employed during the wartime years may have actually had heavier exposures to BD than long-term workers. (Tr. 1/15/91, pp. 115-205) On cross-examination, Dr. Landrigan cautioned against "assuming that duration of exposure directly relates to total cumulative exposure." (Tr. 1/15/91, p. 180) He also emphatically stated that an increased cancer risk in short-term workers would not be inconsistent with a causal association. (Tr. 1/15/91, p. 204) (v) The Potential Role of Confounding Exposures and Observed Results. In epidemiologic studies "confounding" may lead to invalid results. Confounding occurs when there is a mixing of effects. More specifically, confounding may produce a situation where a measure of the effect of an exposure on risk, e.g., SMR, RR, is distorted because of the association of the exposure with other factors that influence the outcome under study.

For example, the IISRP has suggested that confounding exposures from other employment were responsible for the LH cancers observed in the studies of BD epidemiology. (Ex. 113, A-43) This argument is based on the past practice of using petrochemical industry workers, who may have also been exposed to benzene, to start up the SBR and BD production plants. The IISRP finds support for this position in the observation of elevated SMRs in short-term workers employed during the wartime years, precisely those most likely to be cross-employed. (Ex. 113, A-43) However, there are a number of research methods in occupational epidemiology that are available to control potential confounding factors. Research methods that eliminate the effect of confounding variables include: Matching of cases and controls; adjustment of data; and regression analyses. In the nested case-control study, for example, cases and controls were matched on variables that otherwise might have confounded the study results. In the testimony provided by Santos-Burgoa, he states that the "matching scheme allowed us to control for potential confounders and concentrate only on exposure variations." (Ex. 40, p. 12) On cross-examination, Landrigan also addressed the potential role of confounding exposures and the observed study results. First, he observed that Dr. Philip Cole, Professor, Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, one of the outspoken critics of OSHA's proposed rule, found no evidence for confounding in his review of the Matanoski study. (Tr. 1/15/91, p. 178) Second, Dr. Landrigan dismissed the notion of previous exposure to benzene as the causative agent for the observed results in the short-term workers. (Tr. 1/15/91, p. 178-179) In their analyses of mortality patterns by estimated monomer exposure, Delzell et al. used Poisson regression to control for potential confounding factors. (Ex. 117-1) As previously stated, the analyses conducted to determine the association between BD ppm-years and leukemia indicated a positive dose-response relationship, even after controlling for styrene ppm-years, age, years since hire, calendar period, and race. In the opinion of the investigators, benzene exposure did not explain the excess of leukemia risk, and BD is the most likely causal agent. (Ex. 117-1, p. 85) (vi) The Biological Basis for Grouping Related LH Cancers. The epidemiologic studies that have examined the association between occupational exposure to BD and excess mortality have grouped related LH cancers in their analyses. This approach has been criticized as evidence of a lack of "consistency with respect to cell type" which "argues against a common etiologic agent." (Ex. 113, A-45) In other words, these critics suggest that the relationship between BD and excess mortality does not meet the specificity of association requirement for a causal relationship. This requirement states that the likelihood of a causal relationship is strengthened when an exposure leads to a single effect, not multiple effects, and this finding also occurs in other studies.

More specifically, OSHA has been criticized for its position that "broad categories such as `leukemia' or `all LHC' should be used to evaluate the epidemiologic data." (Ex. 113, A-46) Dr. Cole, for example, commented that:

It is a principle of epidemiology--and of disease investigation in general--that entities should be divided as finely as possible in order to maximize the prospect that one has delineated a homogeneous etiologic entity. Entities may be grouped for investigative purposes only when there is substantial evidence that they share a common etiology. (Ex. 63, p. 11) It is Dr. Cole's opinion that LH cancers are "distinct diseases" with "heterogeneous and multifactorial" etiologies. (Ex. 63, p. 47) Dr. Weisenburger, OSHA's expert in hematopathology, provided testimony to the contrary. (Ex. 39, pp. 7-8) According to Dr. Weisenburger, "LH (cancer) cannot be readily grouped into `etiologic' categories, since the precise etiologies and pathogenesis of LH (cancer) are not well understood." (Ex. 39, p. 7) In his opinion, because LH cancers are "closely related to one another and arise from common stem cells and/or progenitor cells, it is valid to group the various types of LH (cancer) into closely-related categories for epidemiologic study." (Ex. 39, p.7) The issue of grouping related LH cancers to observe a single effect was also addressed by Dr. Landrigan in his testimony. (Tr. 1/15/91, pp. 131-133) The first point raised by Dr. Landrigan is that the "diagnostic categories [for LH cancers] are imprecise and * * * overlapping." (Tr. 1/15/91, p. 131) For example, he explained that in clinical practice transitions of lymphomas and myelomas into leukemias may be observed. In such a case, one physician may record the death as due to lymphoma and another may list leukemia as the cause of death. (Tr. 1/15/91, p. 131-132) Additionally, Dr. Landrigan testified that "some patients with lymphomas or multiple myeloma may subsequently develop leukemia as a result of their treatments with radiation or cytotoxic drugs." (Tr. 1/15/91, p. 132) These recordings of disease transition are further complicated by the historical changes that have occurred in nomenclature and The International Classification of Diseases (ICD) coding. According to Dr. Landrigan, certain lymphomas and * * * leukemias, such as chronic lymphatic leukemia are now considered by some investigators * * * to represent different clinical expressions of the same neoplastic process. There have been recent immunologic and cytogenetic studies which indicate that there are stem cells which appear to have the capacity to develop variously into all the various sorts of hematopoietic cells including T-lymphocytes, plasma cells, granulocytes, erythrocytes, and monocytes. (Tr. 1/15/91, p. 132)

Dr. Landrigan summarized his testimony on this issue by stating that "these different types of cells share a common ancestry * * * there is good biologic reason to think that they would have etiologic factors in common." (Tr. 1/15/91, pp. 132-133)

OSHA maintains the opinion, which is well supported by the record, that there is a biological basis and a methodologic rationale for grouping related LH cancers. Furthermore, OSHA rejects the criticism that the observation of different subtypes of LH cancers argues against the consistency and specificity of the epidemiologic findings.

(vii) Relevance of Worker Subgroup Analyses. OSHA has been criticized for focusing on and emphasizing the "few positive results" seen in the results of worker subgroup analyses. (Ex. 113, A-48) It has been pointed out, for example, that in the update of the Matanoski cohort study "there were hundreds of SMRs computed in that study and it's not surprising that one or two or even more would be found to be statistically significant even when there is in fact nothing going on." (Tr. 1/22/91, p. 1444) Additionally, it has been suggested that OSHA has ignored the "clearly overall negative results" of the epidemiologic studies. (Ex. 113, A-48) OSHA agrees with the observation that when many statistical analyses are done on a database, it is possible that some positive results may be due to chance. However, OSHA rejects criticism that the Agency has inappropriately concentrated on the positive results and disregarded the negative results. It is OSHA's opinion that there is a compelling pattern of results in the epidemiologic studies.

Furthermore, a reasonable explanation for the elevated SMR for black production workers in the update of the Matanoski cohort study is that this subset of the population actually had heavy exposure to BD. Support for this explanation can be found in the industrial hygiene survey results of Fajen et al. (Ex. 34-4) In this case, then, the risk for excess mortality would be concentrated in a small subset of otherwise very healthy and unexposed workers that would be diluted when analyses are based on the entire group being studied. The only way to observe the risk in the most highly exposed subset would be to analyze the data by subgroups of the population.

(viii) Appropriateness of Selected Reference Populations. OSHA also has been criticized for "ignor[ing] the fact that most of the epidemiologic studies of butadiene-exposed workers only used U.S. cancer mortality rates for comparison to worker mortality." (Ex. 113, A-49) The significance of this criticism is based on the observation by Downs that "use of local (mortality) rates (for comparison) tended to bring the SMRs closer to 100." (Ex. 17-33, p.14) This finding results from cancer rates along the Texas Gulf coast that are higher than national rates. (Ex. 17-33) In other words, it has been argued that national comparison rates artificially inflate the SMRs, while local rates provide a more accurate picture of the mortality experience of workers with occupational exposure to BD. (Ex. 113, A-50) Dr. Landrigan captured the essence of this issue in his testimony on cross-examination,

This is a perennial debate in epidemiology of whether to use local comparison rates or regional or national, and there's [sic] arguments [to] go both ways. (Tr. 1/15/91, p. 154)

He presented several arguments for using national rates. First, U.S. mortality rates are based on the entire population, so they are more stable. Second, national rates are more commonly used, so it is easier to compare results from different studies.

On the other hand, the argument in favor of using local rates centers on the fact that people in a local area may truly be different from the total population or a regional population(s). Thus, comparing a local subpopulation with the entire local population may provide more accurate results. However, the weakness in this argument was highlighted by Dr. Landrigan when he said that,

* * * if there are factors acting in the local population, such as environmental pollution that may elevate rates in the local area so that they are closer to the rates in the occupationally exposed population, then theoretically at least one could argue that the local population is overmatched, too similar to the employee population and that the use of the national comparison group actually give [sic] a better reflection of reality. (Tr. 1/15/91, p. 155)

In fact, he went on to point out that the BD plants have been identified by the Environmental Protection Agency (EPA) as "major" polluters of the local environment with BD. (Tr. 1/15/91, p. 155) OSHA acknowledges that there are pros and cons to both approaches of reference population selection. However, in the study by Delzell et al. mortality data of the USA cohort subgroup were analyzed using both state, i.e., local, general population rates and USA general population rates. (Ex. 117-1) As previously stated, there was little difference in the overall pattern of these analyses. (Ex. 117-1, p. 60) Additionally, the Santos-Burgoa and Matanoski nested case control study used the most appropriate comparison group of all: Those employed at the same facilities. (Ex. 23-109 and 34-4, Vol. III, H-4) Thus, given the available data in the record, OSHA is of the opinion that it cannot ignore the findings of excess mortality that are based on national comparison rates.

(ix) Summary and Conclusions. (a) Summary. Table V-4 lists the criteria that can be used to judge the presence of a causal association between occupational exposure to BD and cancer of the lymphohematopoietic system. When the available epidemiologic study results are examined in this way, there is strong evidence for causality. The data fulfill all of the listed criteria: Temporal relationship; consistency; strength of association; dose-response relationship; specificity of association; biological plausibility; and coherence.

In his testimony, OSHA's epidemiologist expert witness agreed that there is "definite evidence for the fact that occupational exposure to 1,3-Butadiene can cause human cancer of the hematopoietic and lymphatic organs." (Tr. 1/15/91, p. 133) Dr. Weisenburger, OSHA's expert witness in hematopathology, also concluded that "it would be prudent to treat BD as though it were a human carcinogen." (Ex. 39, p. 11)

           Criterion for causality           |     Met by BD
Temporal relationship....................... |       Yes.
Consistency................................. |       Yes.
Strength of association..................... |       Yes.
Dose-response relationship.................. |       Yes.
Specificity of association.................. |       Yes.
Biological plausibility..................... |       Yes.
Coherence................................... |       Yes.

(b) Conclusion. On the basis of the foregoing analysis, OSHA concludes that there is strong evidence that workplace exposure to BD poses an increased risk of death from cancers of the lymphohematopoietic system. The epidemiologic findings supplement the findings from the animal studies that demonstrate a dose-response for multiple tumors and particularly for lymphomas in mice exposed to BD.

C. Reproductive Effects

In addition to the established carcinogenic effects of BD exposure, various reports have led to concern about the potential reproductive and developmental effects of exposure to BD. The term reproductive effects refers to those on the male and female reproductive systems and the term developmental refers to effects on the developing fetus.

Male reproductive toxicity is generally defined as the occurrence of adverse effects on the male reproductive system that may result from exposure to chemical, biological, or physical agents. Toxicity may be expressed as alterations to the male reproductive organs and/or related endocrine system. For example, toxic exposures may interfere with spermatogenesis (the production of sperm), resulting in adverse effects on number, morphology, or function of sperm. These may adversely affect fertility. Human males produce sperm from puberty throughout life and thus the risk of disrupted spermatogenesis is of concern for the entire adult life of a man.

Female reproductive toxicity is generally defined as the occurrence of adverse effects on the female reproductive system that may result from exposure to chemical, biological, or physical agents. This includes adverse effects in sexual behavior, onset of puberty, ovulation, menstrual cycling, fertility, gestation, parturition (delivery of the fetus), lactation or premature reproductive senescence (aging).

Developmental toxicity is defined as adverse effects on the developing organism that may result from exposure prior to conception (either parent), during prenatal development, or postnatally to the time of sexual maturation. Developmental effects induced by exposures prior to conception may occur, for example, when mutations are chemically induced in sperm. If the mutated sperm fertilizes an egg, adverse developmental effects may be manifested in developing fetuses. Mutations may also be induced in the eggs. The major manifestations of developmental toxicity include death of the developing fetus, structural abnormality, altered growth and function deficiency.

To determine whether an exposure condition presents a developmental or reproductive hazard, there are two categories of research studies on which to rely: Epidemiologic, or studies of humans, and toxicologic, or experimental studies of exposed animals or other biologic systems.

Many outcomes such as early embryonic loss or spontaneous abortion are not easily detectable in human populations. Further, some adverse effects may be quite rare and require very large study populations in order to have adequate statistical power to detect an effect, if in fact one is present. Often, these populations are not available for study. In addition, there are fewer endpoints which may be feasibly measured in humans as compared to laboratory animals. For example, early embryonic loss is difficult to measure in the study of humans, but can be measured easily in experimental animals. There are no human studies available to address reproductive and developmental effects of BD exposure to workers. Thus, evidence on the reproductive and developmental toxicity of BD comes from toxicologic studies performed using primarily mice.

Animal studies have proved useful for studying reproductive/developmental outcomes to predict human risk. A very important advantage to the toxicological approach is the ability of the experimenter to fully quantitate the exposure concentration and conditions of exposure. Although extrapolation of risk to humans on a qualitative basis is accepted, quantitative extrapolation of study results is more complex.

In his testimony, OSHA's witness, Dr. Marvin Legator, an internationally recognized genetic toxicologist from the University of Texas Medical Branch in Galveston, cautioned that in assessing risk "humans in general have proven to be far more sensitive than animals * * * to agents characterized as developmental toxicants." (Ex. 72) He also noted that "of the 21 agents considered to be direct human developmental toxins, in 19 * * * the human has been shown to be more sensitive than the animal * * *" He also pointed to the possibility that sub-groups of the human population may be even more highly sensitive than the population average.

OSHA believes that the animal inhalation studies designed to determine the effect of BD on the reproduction and development of these animals indicate that BD causes adverse effects in both the male and female reproductive systems and produces adverse developmental effects. These studies are briefly summarized and discussed below.

Toxicity to Reproductive Organs

In the first NTP bioassay, an increased incidence of testicular atrophy was observed in male mice exposed to BD atmospheric concentrations of 625 ppm. (Ex. 23-1) In female mice, an increased incidence of ovarian atrophy was observed at 625 and 1,250 ppm. These adverse effects were confirmed in reports of the second NTP study, which used lower exposure concentrations. The latter lifetime bioassay exposed male and female B3C6F1 mice to 0, 6.25, 20, 62.5, 200, and 625 ppm BD. (Ex. 114, p 115) See Table V-5. Testicular atrophy in males was significantly increased at the highest dose tested, 625 ppm, and reduced testicular weight was observed from BD exposures of 200 ppm. (Ex. 96) These latter data are not shown in the Table. In female mice at terminal sacrifice, 103 weeks, ovarian atrophy was significantly increased at all exposure levels including the lowest dose tested, 6.25 ppm, compared with controls. Evidence of ovarian toxicity was also seen during interim sacrifices, but in these cases was the result of higher exposure levels. After 65 weeks of exposure, 90% of the mice exposed to 62.5 ppm experienced ovarian atrophy.


                              Exposure concentration (ppm)
Lesion         of      0      6.25     20      62.5      200       625

                                         Incidence (%)

Testicular   40   0/10(0)  NE       NE       NE        0/10(0)   6/10(60)
  atrophy    65   0/10(0)  NE       NE       NE        0/10(0)   4/7(57)
            103   1/50(2)  3/50(6)  4/50(8)  2/48(4)   6/49(12) 53/72(74)


Ovarian      40   0/10(0)  NE       NE        0/10(0)   9/10(90)  8/8(100)
atrophy      65   0/10(0)  0/10(0)  1/10(10)  9/10(90)  7/10(70)  2/2(100)
            103  4/49(8) 19/49(39) 32/48(67) 42/50(84) 43/50(86) 69/79(87)
  NE, not examined microscopically.
  Source: Ex. 114.

Extensive comments on the BD induced ovarian atrophy were received from Dr. Mildred Christian, a toxicologist who offered testimony on behalf of the Chemical Manufacturers Association. She questioned the relevance of using the data from studies of mice to extrapolate risk of ovarian atrophy to humans because most of the evidence was observed among the animals who were sacrificed after the completion of the species reproductive life and only after prolonged exposure to 6.25 ppm and 20 ppm (Ex. 118-13, Att 3, p. 4) On the other hand, Drs. Melnick and Huff, toxicologists from the National Institute of Environmental Health Sciences stated that: "Even though ovarian atrophy in the 6.25 ppm group was not observed until late in the study when reproductive senescence likely pertains, the dose-response data clearly establish the ovary as a target organ of 1,3-butadiene toxicity at concentrations as low as 6.25 ppm, the lowest concentration studied." (Ex. 114, p. 116) In addition, it should be noted that an elevated incidence of ovarian atrophy was observed at periods of interim sacrifice of female mice exposed to 20 ppm that took place at the 65 week exposure period, a time prior to the ages when senescence would be expected to have occurred. NIOSH also accepted Dr. Melnick's view that mice exposed to 6.25 ppm BD demonstrated ovarian atrophy. (Ex. 32-35) OSHA remains concerned about the ovarian atrophy demonstrated at low exposure levels in the NTP study. Thus, OSHA concludes that exposure to relatively low levels of BD resulted in the induction of ovarian atrophy in mice.

Sperm-Head Morphology Study

NTP/Battelle investigators also described sperm head morphology findings using B6C3F(1) mice exposed as described in the dominant lethal study mentioned below, e.g., exposures to 200, 1000 and 5000 ppm BD. The mice were sacrificed in the fifth week post-exposure and examined for gross lesions of the reproductive system. (Ex. 23-75) The study authors chose this interval as having the highest probability for detecting sperm abnormalities. Epididymal sperm suspensions were examined for morphology. The percentage of morphologically abnormal sperm heads was significantly increased in the mice exposed at 1,000 ppm and 5,000 ppm, but not for those exposed to 200 ppm. The study authors concluded that "these significant differences in the percentage of abnormalities between control mice and males exposed to 1000 and 5000 ppm [BD] indicated that their late spermatogonia or early spermatocytes were sensitive to this chemical." (Ex. 23-75, p. 16) In reviewing this study, Dr. Mildred Christian stated that these results are not necessarily correlated with developmental abnormalities or reduced fertility and are "reversible in nature" and that the observed differences are "biologically insignificant." (Ex. 76, p. 14) In its submission, the Department of Health Services of California said: "A conclusion as to the reproductive consequences of these abnormalities cannot be made from this study." (Ex. 32-168) In reviewing Dr. Christian's comments, OSHA is in agreement that the observation of a significant excess of sperm head abnormalities as a result of BD exposure is not necessarily correlated with the development of abnormal fetuses or of reduced fertility; however, the Anderson study, which did evaluate fetal abnormality and reduced fertility, demonstrated a significant excess of both fetal abnormality plus early and late fetal mortality as a result of male mice exposure to BD. (Ex. 117-1, P. 171) These observations of fetal mortality could only occur as a result of an adverse effect on the sperm. In response to Dr. Christian's comment that the sperm head abnormality observed in the study is reversible, the reversibility would be dependent upon cessation of exposure. Since workers may be exposed to BD on a daily basis, the significance of reversibility may be moot.

Developmental Toxicity

Dominant Lethal Studies

A dominant lethal study was conducted by Battelle/NTP to assess the effects of a 5-day exposure of male CD-1 mice to BD atmospheric concentrations of 0, 200, 1,000 and 5,000 ppm BD for 6 hours per day on the reproductive capacity of the exposed males during an 8-week post-exposure period. (Ex 23-74) If present, dominant lethal effects are expressed as either a decrease in the number of implantations or as an increase in the incidence of intrauterine death, or both, in females mated to exposed males. Dominant lethality is thought to arise from lethal mutations in the germ cell line that are dominantly expressed through mortality to the offspring. In this study, the only evidence of toxicity to the adult male mouse was transient and occurred over a 20 to 30 minute period following exposure at 5,000 ppm. Males were then mated to a different female weekly for 8 weeks. After 12 days, females were killed and examined for reproductive status. Uteri were examined for number, position and status of implantation. Females mated to the BD-exposed males during the first 2 weeks post-exposure were described as more likely than control animals to have increased numbers of dead implantations per pregnancy.

For week one, the percentage of dead implantations in litters sired by males exposed to 1,000 ppm was significantly higher than controls. There were smaller increases at 200 ppm and 1000 ppm that were not statistically significant. The percentage of females with two or more dead implantations was significantly higher than the control value for all three exposure groups. For week two, the numbers of dead implantations per pregnancy in litters sired by males exposed to 200 ppm and 1000 ppm were also significantly increased, but not for those exposed to 5000 ppm. No significant increases in the end points evaluated were observed in weeks three to eight. These results suggested to the authors that the more mature cells (spermatozoa and spermatids) may be adversely altered by exposure to BD. (Ex. 23-74) The State of California Department of Health Services concluded that the above mentioned study showed no adverse effect from exposure to BD, with the possible exception of the increase in intrauterine death seen as a result of male exposures to 1000 ppm BD at the end of one week post exposure. (Ex. 32-16) Since values for the 5000 ppm exposure group were not significantly elevated for this same period of follow up, the California Department of Health thought the biological significance of the results of the 1000 ppm exposure was questionable. (Ex. 32-16) On the other hand, Dr. Marvin Legator stressed the low sensitivity of the dominant lethal assay which, he felt was due to the endpoint-lethality. He expressed the opinion that the studies were "consistent with an effect on mature germ cells." (Ex. 72) He felt that since an effect was observable in this relatively insensitive assay that only the "tip of the iceberg" was observed, and that "[t]ransmissible genetic damage, displaying a spectrum of abnormal outcomes can be anticipated at concentrations (of BD) below those identified in the dominant lethal assay procedure." (Ex. 72, p. 17) The dominant lethal effect of BD exposure was more recently confirmed by Anderson et al. in 1993. (Ex. 117-1, p. 171) They studied CD-1 mice using a somewhat modified study design. Two exposure regimens were used. In the first, "acute study," male mice were exposed to 0 (n=25), 1250 (n=25), or 6250 (n=50) ppm BD for 6 hours only. Five days later they were caged with 2 untreated females. One female was allowed to deliver her litter and the other was killed on day 17 of gestation and examined for the number of live fetuses, number of early and late post-implantation deaths and the number and type of any gross malformation. The authors stated that sacrifice on day 17 (rather than the standard days 12 through 15) allowed examination of near-term embryos for survival and abnormalities. The mean number of implants per female was reduced compared with controls at both concentrations of BD, but was statistically significant only at 1250 ppm. Neither post-implantation loss nor fetal abnormalities were significantly increased at either concentration. The authors concluded that "a single 6-hour acute exposure to butadiene was insufficient to elicit a dominant lethal effect." (Ex. 117-1, p. 171) In the second phase of the study, the "subchronic study," CD-1 mice were exposed to 0 (n=25), 12.5 (n=25), or 1250 (n=50) ppm BD for 6 hours per day, 5 days per week, for 10 weeks. They were then mated. The higher 1250 ppm BD exposure resulted in significantly reduced numbers of implantations and in significantly increased numbers of dominant lethal mutations expressed as both early and late deaths. See Table V-6. Non-lethal mutations expressed as birth abnormalities were also observed in live fetuses (3/312; 1 hydrocephaly and 2 runts).

The lower exposure (12.5 ppm) did not result in decreases in the total number of implants, nor in early deaths; however, the frequencies of late deaths and fetal abnormalities (7/282; 3 exencephalies in 1 litter and one in another, two runts and one with blood in the amniotic sac) were significantly increased.

The authors felt that their finding of increased late deaths and fetal abnormalities at a subchronic, low exposure of 12.5 ppm was the main new finding of the study. They noted that these adverse health effects were increased 2-3 fold over historical controls. In evaluating these latter two studies OSHA notes that while there was no demonstrable effect on dominant lethality as a result of a single exposure to 1250 ppm BD, subchronic exposure to 12.5 ppm, the lowest dose tested, resulted in the induction of dominant lethal mutations and perhaps non-lethal mutations. (Ex 117-1, p 171) OSHA has some reservations about whether or not the fetal abnormalities observed in the Anderson et al. "subchronic" study were actually caused by non-lethal mutations or by some other mechanism because they were observed in only a few of the litters produced by the mice. (Ex. 117-1, p. 171)

    |             |             |             |             |
    |Implantations|   Early     |    Late     |Late deaths  | Abnormal
    |             |   deaths    |   deaths    | including   | fetuses
    |             |             |             |dead fetuses |
    |    |        |   |         |   |         |   |         |    |
    | No.|Mean    |No.| Mean(a) |No.|Mean(a)  |No.|Mean(a)  |No. |Mean(a)
    |    |        |   |         |   |         |   |         |    |
Con-|278 | 12.09  |13 | 0.050   |0  |         |2  |  0.007  |0   |
trol|    |+/-1.276|   |+/-0.0597|   |         |   |+/-0.0222|    |
    |    |        |   |         |   |         |   |         |    |
12.5|306 | 12.75  |16 | 0.053   |7  |0.23**   |8  |0.026    |7(b)|0.024*
    |    |+/-2.507|   |+/-0.0581|   |+/-0.038 |   |+/-0.0424|    |+/-0.062
    |    |        |   |         |   |         |   |         |    |
1250|406 | 10.68**|87 | 0.204***|6  |0.014*** |7  |0.016    |3(c)|0.011**
 ppm|    |+/-3.103|   |+/-0.161 |   |+/-0.0324|   |+/-0.339 |    |+/-0.043
    |    |        |   |         |   |         |   |         |    |(< /=)
  Footnote(*) Significantly different from control at: *p< /=0.05;
**p< /= 0.01; ***p< /= 0.001 (by analysis of variance and least
significance test on arc-sine transform data).
  Footnote(a) Per implantation.
  Footnote(b) Four exencephalies (three in one litter), two runts (< /=
70% and 60% of mean body weight of others in litter; total litter
sizes 7 and 9, respectively one fetus with blood in amniotic sac but
no obvious gross malformation (significance of difference not altered
if this fetus is excluded).
  Footnote(c) One hydrocephaly, two runts (71% and 75% of mean body
weight of others in litter; total litter sizes; 2 and 11,

A dominant lethal test was also performed by Adler et al. (Ex. 126) Male(102/E1XC3H/E1)F(1) male mice were exposed to 0 and 1300 ppm BD. They were mated 4 hours after the end of exposure with untreated virgin females. Females were inspected for the presence of a vaginal plug every morning. Plugged females were replaced by new females. The mating continued for four consecutive weeks. At pregnancy day 14-16 the females were killed and uterus contents were evaluated for live and dead implants. Exposure of male mice to 1300 ppm BD caused an increase of dead implants during the first to the third mating week after 5 days of exposure. The dead implantation rate was significantly different from the concurrent controls only during the second mating week. Adler et al. concluded that dominant lethal mutations were induced by BD in spermatozoa and late stage spermatids and that these findings confirmed the results of the Battelle/NTP study which showed effects on the same stages of sperm development. (Ex. 23-74) The authors were of the opinion that BD may induce heritable translocations in these germ cell stages.

The earliest reproductive study reported on BD was conducted by Carpenter et al. in 1944. (Ex. 23-64) In this study, male and female rats were exposed by inhalation to 600, 2,300 or 6,700 ppm BD, 7.5 hours per day, six days per week for an 8-month period. Although this study was not specifically designed as a reproductive study, the fertility and the number of progeny were recorded. No significant effects due to BD exposure were noted for either the number of litters per female animal or for the number of pups per litter.

In the Hazelton study, Sprague-Dawley (SD) rats were exposed by inhalation to 0, 200, 1,000 or 8,000 ppm BD on days 6 though 15 of gestation. (Ex. 2-32) There were dose-related effects on maternal body weight gain, fetal mean weight and crown-to-rump length. Post-implantation loss was slightly higher in all BD-exposed groups. In addition, there were significant increases in hematoma in pups in the 200 and 1,000 ppm exposure groups. In the 8,000 ppm exposure group, a significantly increased number of pups had lens opacities and there was an increased number of opacities per animal. According to the authors, the highest exposure groups also had a significantly increased number of fetuses with skeletal variants, a higher incidence of bipartite thoracic centra, elevated incidence of incomplete ossification of the sternum, higher incidence of irregular ossification of the ribs, and "other abnormalities of the skull, spine, long bones, and ribs." The authors concluded that the fetal response was not indicative of a teratogenic effect, but was the result of maternal toxicity.

In the Battelle/NTP study, pregnant Sprague-Dawley (SD) rats and pregnant Swiss mice were exposed to 0, 40, 200, or 1,000 ppm BD for 6 hours per day from day 6 through day 15 of gestation. (Ex. 23-72) Animals were sacrificed and examined one day before expected delivery. In the rat, very little effect was noted; in the 1,000 ppm exposure group only there was evidence of maternal toxicity, i.e., depressed body weight gains during the first 5 days of exposure. No evidence of developmental toxicity was observed in the SD rats evaluated in the study, e.g., the number of live fetuses per litter and the number of intrauterine deaths were within normal limits.

In the mouse, exposure to the above mentioned concentrations did not result in significant maternal toxicity, with the exception of a reduction in extra-gestational weight gain for the 200 ppm and 1000 ppm BD exposed dams. In the female mice, there was a significant depression of fetal body weight only at the 200 and 1,000 ppm exposure levels. Fetal body weight for male pups was reduced at all exposure concentrations, including the 40 ppm exposure level, even though evidence of maternal toxicity was not observed at this exposure concentration. No significant differences were noted in incidence of malformations among the groups. However, the incidence of supernumerary ribs and reduced ossification of sternebrae was significantly increased in litters of mice exposed to 200 and 1,000 ppm BD.

In reviewing these data, Drs. Melnick and Huff noted that since maternal body weight gain was reduced at the 200 and 1000 ppm exposure levels and body weights of male fetuses were reduced at the 40, 200, and 1000 exposure levels "[t]he male fetus is more susceptible than the dam to inhaled 1,3-butadiene." (Ex. 114, p. 116) They further stated that "the results of the study in mice reveal that a toxic effect of 1,3-butadiene was manifested in the developing organism in the absence of maternal toxicity." On the basis of this study, the authors concluded that "1,3-butadiene does not appear to be teratogenic in either the rat or the mouse, but there is some indication of fetotoxicity in the mouse." (Ex. 23-72) On the other hand, Dr. Mildred Christian was of the opinion that the significant decrease in male mouse fetal weight gain in the 40 ppm exposure group was not a selective effect of BD on the conceptus, but rather was a result of the statistical analysis used which she considered inappropriate. (Ex. 118-13, Att. 3, p. 6) She was also of the opinion that the larger litter sizes in the 40 ppm exposure group as compared with the control group contributed to the statistical finding. Dr. Christian, however, did not present any specific information on the type of analysis used for statistical testing that she thought made the results inappropriate. In general, one would expect that the evaluation of data from larger litter sizes would give one more confidence in the statistical findings.

In reviewing the same study, the State of California, Department of Health Services was more cautious. It stated that "The increased incidence of reduced ossifications and the fetal weight reductions in the absence of apparent maternal toxicity in the 40- and 200-ppm groups is evidence of fetotoxicity * * * in the Swiss (CD-1) mouse." After reviewing the study results and arguments about the study, OSHA concluded that the NTP study provides evidence of fetotoxicity in the mouse. (Ex. 23-72)

Mouse spot test

Adler et al. (1994) conducted a spot test in mice. (Ex. 126) The spot test is an in vivo method for detecting somatic cell mutations. A mutation in a melanoblast is detected as a coat color spot on the otherwise black fur of the offspring. Pregnant females were exposed to 0 or 500 ppm BD for 6 hours per day on pregnancy days 8, 9, 10, 11 and 12. They were allowed to come to term and to wean their litters. Offspring were inspected for coat color spots at ages 2 and 3 weeks. Gross abnormalities were also recorded. Exposure to a concentration of 500 ppm did not cause any embryotoxicity, nor were gross abnormalities observed. The BD exposure, however, significantly increased the frequency of coat color spots in the offspring. This study demonstrates that BD exposure is capable of causing transplacentally induced somatic cell mutations that can result in a teratogenic effect in mice.

Summary of Reproductive and Developmental Effect

OSHA has limited its discussion on reproductive and developmental hazards to a qualitative evaluation of the data. This approach was chosen because no generally accepted mathematical model for estimating reproductive/developmental risk on a quantitative basis was presented during the rulemaking. For example, the CMA Butadiene panel disagreed with OSHA's findings in the proposal regarding the potential reproductive and developmental risks presented by BD exposure using an uncertainty factor approach. (See Ex. 112) They cited Dr. Christian's conclusion that the mouse possessed a "special sensitivity" to BD and should not be used as a model on which to base risk estimates.

The agency has determined, however, that animal studies, taken as a whole, offer persuasive qualitative evidence that BD exposure can adversely effect reproduction in both male and female rodents. The Agency also notes that BD is mutagenic in both somatic and germ cells. (Ex. 23-71; Ex. 114; Ex. 126) Some evidence of maternal and developmental toxicity was seen in rats exposed to BD, but the concentrations used were much higher than those that elicited a response in mice. (Ex. 118-13, Att. 3, p. 2) In mice, evidence of fetotoxicity was observed in either the presence or absence of maternal toxicity, the latter evidence being provided by decreased fetal body weight in male mice whose dams were exposed to 40 ppm BD, the lowest dose tested in the study. In addition, a teratogenic effect was observed in mice (coat color spot test) as a result of transplacentally induced somatic cell mutation.

OSHA is also concerned about the observation of a significant excess of sperm head abnormalities as a result of BD exposure, even though this expression of toxicity is not necessarily correlated with the development of abnormal fetuses or of reduced fertility. The Anderson study, which did evaluate reduced fertility and fetal abnormality, demonstrated a significant excess of both early and late fetal mortality and perhaps fetal abnormality as a result of male mice exposure to BD. (Ex. 117-1, P. 171) This observation could only occur as a result of an adverse effect on the sperm. Two additional studies also provide evidence of dominant lethality as a result of male exposure to BD. (Ex. 23-74; Ex. 126) The observation of germ cell effects is supported by additional evidence of genotoxicity in somatic cells, as demonstrated by positive results in the micronucleus test and in the mouse spot test. (Ex. 126) Some of the adverse effects related to reproductive and developmental toxicity in the mouse, e.g., ovarian atrophy, testicular atrophy, reduced testicular weight, abnormal sperm heads, dominant lethal effects, were acknowledged by Dr. Christian, but she urged the Agency not to rely on these findings because of negative study results in other species, or because positive findings in other species required much higher exposure levels. (Ex. 118-13, Att. 3, p. 1) For example, a CMA witness has argued that the diepoxide is responsible for the ovarian atrophy observed in relation to low level BD exposure (6.25 ppm). (Ex. 118-13, Att. 3) However, the monoepoxide could also play a role in the ovarian atrophy and evidence indicates that humans can form the monoepoxide of BD and that humans have the enzymes present that could cause conversion to the diepoxide. Therefore on a qualitative basis, the observation of ovarian atrophy in the mouse is meaningful in OSHA's view. In addition, the metabolic factors related to testicular atrophy, malformed sperm and dominant lethal mutations in the mouse are not known. (See section on in vitro metabolic studies.) These observations further support the findings in mice as being meaningful for humans on a qualitative basis. The mouse spot test which demonstrates a somatic cell mutation leading to a teratogenic effect inconsistent with data showing the ability of BD to cause adverse effects on chromosomes and hprt mutations in humans exposed to BD.

OSHA also notes that studies of workers exposed to low concentrations of BD demonstrated a significant excess of chromosomal breakage and an inability to repair DNA damage. Thus, BD exposure seems capable of inducing genetic damage in humans as a result of low level exposure. Therefore, the mouse studies which demonstrate genetic damage (mutations) in both somatic and germinal cells seem to be a better model on a qualitative basis than the rat for predicting these adverse effects in humans.

D. Other Relevant Studies

1. Acute Hazards

At very high concentrations, BD produces narcosis with central nervous system depression and respiratory paralysis. (Ex. 2-11) LC(50) values (the concentration that produces death in 50 percent of the animals exposed) were reported to be 122,170 ppm (12.2% v/v) in mice exposed for 2 hours and 129,000 ppm (12.9% v/v) in rats exposed for 4 hours. (Ex. 2-11, 23-91) These concentrations would present an explosion hazard, thus limiting the likelihood that humans would risk any such exposure except in extreme emergency situations. Oral LD(50) values (oral dose that results in death of 50 percent of the animals) of 5.5 g/kg body weight for rats and 3.2 g/kg body weight for mice have been reported. (Ex. 23-31) These lethal effects occur at such high doses that BD would not be considered "toxic" for purposes of Appendix A of OSHA's Hazard Communication Standard (29 CFR 1910.1200), which describes a classification scheme for acute toxicity based on lethality data.

At concentrations somewhat above the previous permissible exposure level of 1,000 ppm, BD is a sensory irritant. Concentrations of several thousand ppm were reported to cause irritation to the skin, eyes, nose, and throat. (Ex. 23-64, 23-94) Two human subjects exposed to BD for 8 hours at 8000 ppm reported eye irritation, blurred vision, coughing, and drowsiness. (Ex. 23-64)

2. Systemic Effects

In the preamble to the proposal, OSHA reviewed the literature to discern the systemic effects of BD exposure. (55 FR 32736 at 32755) OSHA discussed an IARC review which briefly examined several studies from the former Soviet Union. In these, various adverse effects, such as hematologic disorders, liver enlargement and liver and bile-duct diseases, kidney malfunctions, laryngotracheitis, upper respiratory tract irritation, conjunctivitis, gastritis, various skin disorders and a variety of neurasthenic symptoms, were ascribed to occupational exposure to BD. (Ex. 23-31) OSHA and IARC have found these studies to be of limited use primarily due to their lack of exposure information. Except for sensory irritant effects and hematologic changes, evidence from studies of other exposed groups have failed to confirm these observations.

Melnick and Huff summarized the observed non-neoplastic effects of BD exposure in the NTP I and NTP II mouse bioassays. They listed the following effects associated with exposure of B6C3F(1) mice to BD for 6 hours per day 5 days per week for up to 65 weeks:

* * * epithelial hyperplasia of the forestomach, endothelial hyperplasia of the heart, alveolar epithelial hyperplasia, hepatocellular necrosis, testicular atrophy, ovarian atrophy and toxic lesions in nasal tissues (chronic inflammation, fibrosis, osseous and cartilaginous metaplasia, and atrophy of the olfactory epithelium.) (Ex. 114, p. 114)

They noted that the nasal lesions were seen only in the group of male mice exposed to 1250 ppm BD and that no tumors were observed at this site. Further, Melnick and Huff suggested that some of the proliferative lesions observed in the bioassay might represent pre-neoplastic changes.

The findings of testicular and ovarian atrophy are discussed more fully in the Reproductive Effects section of this preamble,.

Nephropathy, or degeneration of the kidneys, was the most common non-carcinogenic effect reported for male rats in the Hazelton Laboratory Europe (HLE) study in which rats were exposed to 1000 or 8000 ppm BD for 6 hours per day, 5 days per week for up to 2 years. Nephropathy was one of the main causes of death for the high dose males. (Ex. 2-31, 23-84) The combined incidence of marked or severe nephropathy was significantly elevated in the high dose group over incidence in the low dose group and over incidence in the controls (p< .001). HLE's analysis of "certainly fatal" nephropathy shows a significant dose-related trend (p< .05), but when "uncertainly fatal" cases were included, the trend disappeared.

The HLE study authors concluded that the interpretation of the nephropathy incidence data was equivocal. They stated that "an increase in the prevalence of the more severe grades of nephropathy, a common age-related change in the kidney, was considered more likely to be a secondary effect associated with other unknown factors and not to represent a direct cytotoxic effect of the test article on the kidney."

Upon reviewing the HLE rat study for the proposed rule, OSHA expressed concern that only 75% of the low-dose male rats in the HLE study exhibited nephropathy, while 87% of the control rats had some degree of nephropathy, suggesting low-dose male rats were less susceptible to kidney degeneration than control rats, thereby decreasing the comparability between rats in the low-dose and control groups. (55 FR 32736 at 32744) Dr. Robert K. Hinderer, in testifying for the CMA BD Panel, countered that the NTP I mouse study also had "selected instances where the response in the test group (was) lower than that in the controls" and that "* * * (o)ne cannot look at single or a few individual site responses to evaluate the health status or overall effect of the chemical." (Ex. 51) OSHA agrees that there may be some variability in background response rates for specific outcomes. However, the Agency believes that it is important to assess the impact of the variability in background response rates when drawing conclusions about dose-related trends in the data. This was not done in the HLE study nephropathy analysis.

Other non-carcinogenic effects observed in the HLE rat study were elevated incidence of metaplasia in the lung of high dose male rats at terminal sacrifice as compared with incidence in male controls at terminal sacrifice, and a significant increase in high dose male rat kidney, heart, lung, and spleen weights over the organ weights in control male rats.

3. Bone Marrow Effects

There was a single study of BD-exposed humans discussed in the proposal--a study by Checkoway and Williams that examined 163 hourly production workers who were employed at the SBR facility studied by McMichael et al.. (described more fully in the Epidemiology Section of this Preamble.) (Ex. 23-4, 2-28).

Exposure to BD, styrene, benzene, and toluene was measured in all areas of the plant. BD and styrene concentrations, 20 (0.5-65) ppm and 13.7 (0.14-53) ppm, respectively, were considerably higher in the Tank Farm than in other departments. In contrast, benzene exposures, averaging 0.03 ppm, and toluene concentrations, averaging 0.53 ppm, were low in the Tank Farm. The authors compared the hematologic profiles of Tank Farm workers (n=8) with those of the other workers examined.

The investigation focused on two potential effects, bone marrow depression and cellular immaturity. Bone marrow depression was suspected if there were lower levels of erythrocytes, hemoglobin, neutrophils, and platelets. Cellular immaturity was suggested by increases in reticulocyte and neutrophil band form values.

Although the differences were small, adjusted for age and medical status, hematologic parameters in the Tank Farm workers differed from those of the other workers. Except for total leukocyte count, the hematologic profiles of the Tank Farm workers were consistent with an indication of bone marrow depression. The Tank Farm workers also had increases in band neutrophils, a possible sign of cellular immaturity, but no evidence that increased destruction of reticulocytes was the cause.

While acknowledging the limitations of the cross-sectional design of the study, the authors felt, nevertheless, that their results were "suggestive of possible biological effects, the ultimate clinical consequences of which are not readily apparent." OSHA finds any evidence of hematological changes in workers exposed at BD levels well below the existing permissible limit (1000 ppm) to be of concern since such information suggests the inadequacy of the present exposure limit. However, this cross-sectional study involved only 8 workers with relatively high levels of exposure to BD and low levels of exposure to benzene, so it is quite insensitive to minor changes in hematologic parameters.

In a review of BD-related studies, published in 1986, an IARC Working Group felt the study of Checkoway and Williams could not be considered indicative of an effect of BD on the bone marrow (Ex. 2-28). In 1992, IARC concluded that the "changes cannot be interpreted as an effect of 1,3-butadiene on the bone marrow particularly as alcohol intake was not evaluated." (Ex. 125, p. 262) In light of the more recent animal studies that were not available to IARC, however, OSHA believes that the bone marrow is a target of BD toxicity. Furthermore, the fact that changes in hematologic parameters could be distinguished in workers exposed to BD at 20 ppm indicates that such measurements may prove a sensitive indicator of excessive exposure to BD.

In testimony for the CMA BD Panel, Dr. Michael Bird stated his conclusion that the hematological differences between the 8 tank farm workers and the lesser exposed group of workers was not "statistically significant by the usual conventional statistics." (Tr. 1/18/1991, p. 1078) He believed that although the raw data were not available, the reported means were within the historical and expected range for these parameters. (Tr. 1/18/1991, p., 1078) In contrast, OSHA concludes from this study that the hematologic differences observed in BD-exposed workers, although small, are suggestive of an effect of BD on human bone marrow under occupational exposure conditions.

Thus OSHA considers the Checkoway and Williams study to be suggestive of hematologic effects in humans, but does not regard it as definitive. No other potential systemic effects of BD exposure on this population were addressed in the Checkoway and Williams study.

In 1992, Melnick and Huff reviewed the toxicologic studies of BD exposure in laboratory animals. (Ex. 114) Only slight to no systemic effects were observed in an early study of rats, guinea pigs, rabbits and a dog exposed to BD up to 6,700 ppm daily for 8 months. (Ex. 23-64) The study of Sprague Dawley rats exposed to doses of BD up to 8,000 ppm daily for 13 weeks also did not result in hematologic, biochemical, neuromuscular, nor urinary effects. However, there were marked effects seen in exposed mice.

Epidemiologic studies of the styrene-butadiene rubber (SBR) industry suggest that workers exposed to BD are at increased risk of developing leukemia or lymphoma, two forms of hematologic malignancy (see preamble section on epidemiology). Consequently, investigators have looked for evidence of hematopoietic toxicity resulting from BD exposure in animals and in workers. For example, Irons and co-workers at CIIT found that exposure of male B6C3F(1) mice to 1,250 ppm of BD for 6-24 weeks resulted in macrocytic-megaloblastic anemia, an increase in erythrocyte micronuclei and leukopenia, principally due to neutropenia. Bone marrow cell types overall were not altered, but there was an increase in the number of cells in the bone marrow of exposed mice due to an increase in DNA synthesis. (Ex. 23-12) Melnick and Huff also reviewed the available information on bone marrow toxicity. (Ex. 114, p. 114) Table V-7 represents the reported findings of a study of 10 B6C3F(1) mice sacrificed after 6.25-625 ppm exposure to BD for 40 weeks. The authors concluded that these data demonstrated a concentration-dependent decrease in red blood cell number, hemoglobin concentration, and packed red cell volume at BD exposure levels from 62.5 to 625 ppm. The effects were not observed at 6.25 and 20 ppm exposure levels. Melnick and Kohn also noted the increase in mean corpuscular volume in mice exposed at 625 ppm, and suggested that this and other observations (such as those of Tice (Ex. 32-38D)) who observed a decrease in the number of dividing cells in mice and decreased rate of their division), suggested that BD exposure led to a suppression of hematopoiesis in bone marrow. Melnick and Huff concluded that this, in turn, led to release of large immature cells from sites such as the spleen, which was considered indicative of macrocytic megaloblastic anemia by Irons. They concluded that these findings "(establish) the bone marrow as a target of 1,3-butadiene toxicity in mice." (Ex. 114, p. 115)

              6 HOURS.DAY, 5 DAYS/WEEK FOR 40 WEEKS
            |             |             |             |
            | Red blood   | Hemoglobin  |  Volume     |     Mean
BD exposure | cell count  |   conc.     | packed RBC  |   corpuscular
   (ppm)    | (x10(6)/ul  |  (g/dl)     | (ml/dl)     |     vol
            |             |             |             |
0...........| 10.4+/-0.3  | 16.5+/-0.4  | 48.1+/-1.5  |   46.3+/-0.8
6.25........| 10.3+/-0.3  | 16.4+/-0.5  | 47.8+/-1.7  |   46.4+/-1.0
20..........| 10.4+/-0.4  | 16.7+/-0.7  | 48.2+/-2.2  |   46.3+/-0.8
62.5........|(a)9.9+/-0.4 |(a)15.9+/-0.6|(a)45.9+/-2.1|   46.7+/-1.2
200.........|(a)9.6+/-0.5 |(a)15.6+/-0.9|(a)45.4+/-2.7|   47.2+/-1.0
625.........|(a)7.6+/-1.2 |(a)13.5+/-1.8|(a)39.9+/-5.3|(a)53.2+/-2.9
  Adapted from Melnick and Huff, Exhibit 114.
  Footnote(a) Different from chamber control (0 ppm), P< 0.05. Results
of treated groups were compared to those of control groups using
Dunnett's t-test.

4. Mutagenicity and Other Genotoxic Effects

OSHA discussed the genotoxic effects of BD exposure in some detail in the proposal. (55 FR 32736 at 32760) Briefly, BD is mutagenic to Salmonella typhimurium strains TA 1530 and TA 1535 when activated with S9 liver fraction of Wistar rats treated with phenobarbital or Arochlor 1254. These bacterial strains are sensitive to base-pair substitution mutagens. Since the liver fraction is required to elicit the positive mutagenic response, BD is not a direct-acting mutagen and likely must be metabolized to an active form before becoming mutagenic in this test system. IARC published an extensive list of "genetic and related effects of 1,3-butadiene." (Ex. 125) They noted in summarizing the data that BD was negative in tests for somatic mutation and recombination in Drosophila, and that neither mouse nor rat liver from animals exposed to 10,000 ppm BD showed evidence of unscheduled DNA synthesis.

As OSHA described in the proposed rule, and Tice et al. reported in 1987, BD is a potent in vivo genotoxic agent in mouse bone marrow cells that induced chromosomal aberrations and sister chromatid exchange in marrow cells and micronuclei in peripheral red blood cells. (55 FR 52736 at 52760) Some of these effects were evident at exposures as low as 6.25 ppm (6 hours/day, 10 days). However, similar effects were not observed in rat cells exposed to higher levels of BD (10,000 ppm for 2 days). Sister chromatid exchange is a recombinational event in which nucleic acid is exchanged between the two sister chromatids in each chromosome. It is thought to result from breaks or nicks in the DNA. Irons et al. described micronuclei as "* * * chromosome fragments or chromosomes remaining as the result of non-dysjunctional event. Their presence in the circulation is frequently associated with megaloblastic anemia." (Ex. 23-12).

In a subsequent study, Filser and Bolt exposed B6C3F(1) mice to the same 3 concentrations of BD, 6.25, 62.5 or 625 for 6 hours/day, 5 days/week, for 13 weeks. (Ex. 23-10) Peripheral blood samples were taken from 10 animals per group and scored for polychromatic erythrocytes (PCE) and micronucleated normochromatic erythrocytes (MN-NCE). The MN-NCE response, which reflects an accumulated response, was significantly increased in both sexes at all concentrations of BD, including 6.25 ppm.

Certain metabolites of BD also produce genotoxic effects. These are detailed in a number of reviews (see for example, Ex. 114, 125). Briefly, epoxybutene (the monoepoxide) is mutagenic in bacterial systems in the absence of exogenous metabolic activation. Epoxybutene also reacts with DNA, producing two structurally identical adducts and has been shown to induce sister chromatid exchanges in Chinese hamster ovary cells and in mouse bone marrow in vivo.

IARC in its review concluded that the diepoxide, 1,2,:3,4-diepoxybutane, induced DNA crosslinks in mouse hepatocytes and, like epoxybutene, is mutagenic without metabolic activation. As discussed below, BD diepoxide also induced SCE and chromosomal aberrations in cultured cells.

A human cross-sectional study involving a limited number of workers in a Texas BD plant indicated genotoxic effects. (Ex. 118-2D) Peripheral lymphocytes were cultured from 10 non-smoking workers and from age- and gender-matched controls who worked in an area of very low BD exposure (0.03 ppm). Production areas in the plant had a mean exposure of 3.5 ppm BD, with most exposed workers in this sample experiencing exposure of approximately 1 ppm BD.

Standard assays for chromosomal aberrations and a gamma irradiation challenge assay that was designed to detect DNA repair deficiencies were performed. The results of the standard assay indicated that the exposed group had a higher frequency of cells with chromosome aberrations and higher chromatid breaks compared with the control group. This difference was not statistically significant. In the challenge assay, the exposed group had a statistically significant increased frequency of aberrant cells, chromatid breaks, dicentrics (chromosomes having 2 centromeres) and a marginally significant higher frequency of chromosomal deletions than controls. Au and co-workers concluded that cells exposed to BD are likely to have more difficulty in repairing radiation induced damage. (Ex. 118-2D) To determine the mutagenic potential of both BD and its three metabolite epoxides, Cochrane and Skopek studied effects in human lymphoblastoid cells (TK6) and in splenic T cells from exposed B6C3F(1) mice. (Ex. 117-2, p. 195) TK6 cells were exposed for 24 hours to epoxybutene (0-400 uM), 3,4-epoxy-1,2-butanediol (0-800 uM), or diepoxybutane (0-6 uM). All metabolites were mutagenic at both the hprt (hypoxanthine-guanine phosphoribosyl transferase) and tk (thymidine kinase) loci, with diepoxybutane being active at concentrations 100 times lower than epoxybutane or epoxybutanediol.

They also studied mice exposed to 625 ppm BD for 2 weeks and found a 3-fold increase in hprt mutation frequency in splenic T cells compared with controls. They also intended to give daily IP doses of epoxybutene (60, 80 or 100 mg/kg) or diepoxybutane (7, 14, or 21 mg/kg) every other day for three days. However, only animals given the lowest dose of the diepoxide received three doses because of lethality. After two weeks of expression time, cells were isolated for determination of mutation frequency. Both exposure regimens resulted in increased mutation frequency. For example, at the highest exposure to epoxybutene, the average mutation frequency was 8.6 x 10(6), while the diepoxide exposed group had a frequency of 13 x 10(6), compared to a control mutation frequency of 1.2 x 10(6).

Cochrane and Skopek used denaturing gradient gel electrophoresis to study the nature of the splenic T cell hprt mutants in the DNA. They found about half were frameshift mutations. A potential "hotspot" was also described in which a plus one (+1) frameshift mutation in a run of six guanine bases was observed in four BD-exposed mice, in four expoxybutene-exposed mice and in two mice exposed to the diepoxide. They observed both G:C and A:T base pair substitutions in the epoxide treated group; however, similar to the findings of Recio, et al. (described below), A:T substitutions were observed only in the BD-treated group. The authors offered no hypothesis for this observation. These researchers also noted a significant correlation of dicentrics with the presence of a BD metabolite, (1,2-dihydroxy-4-(N-acetyl-cysteinyl-S)butane) in the urine of exposed workers. They further concluded that:

This study indicates that the workers had exposure-induced mutagenic effects. Together with the observation of gene mutation in a subset of the population, this study indicates that the current occupational exposure to butadiene may not be safe to workers. (Ex. 118-2D)

An abstract by Hallberg submitted to the Environmental Mutagenesis Society describes a host-cell reaction assay in which lymphocytes transfected with a plasmid with an inactive chloramphenicol acetyl transferase (CAT) reporter gene were challenged to repair the damaged plasmid and reactivate the CAT gene. No effect was noted among cells of workers exposed to 0.3 ppm benzene; however, BD-exposed workers (mean exposure 3 ppm) had significantly reduced DNA repair capacity (p=0.001). The authors believed that this finding confirmed the DNA repair defect due to BD exposure observed in the Au et al. study's challenge assay. (Ex. 118-2D) Ward and co-workers reported the results of a preliminary study to determine whether a biomarker for BD exposure and a biomarker for the genetic effect of BD exposure could be detected in BD-exposed workers. (Ex. 118-12A) The biomarker for exposure was excretion of a urinary metabolite of BD, (1,2-dihydroxy-4-(n-acetylcysteinyl-S)butane. The genetic biomarker was the frequency of lymphocytes containing mutations at the hypoxanthine-guanine phosphoribosyl transferase (hprt) locus. Study subjects included 20 subjects from a BD production plant and 9 from the authors' university; all were verified non-smokers. Seven workers were in areas or at jobs that were "considered likely to expose them to higher levels of butadiene than in other parts of the plant." Ten worked in areas where the likelihood of BD exposure was low. Three "variable" employees worked in both types of jobs or areas. hprt assays of 6 of the 7 high exposure group and 5 of the 6 non-exposed groups were completed at the time of the report. Air sampling was used to estimate exposure. In the production area, the mean was approximately 3.5 ppm, with most samples below 1 ppm. In the central control area (lower exposure) the mean was 0.03 ppm. The frequency of mutant lymphocytes in the high-exposure group compared with either the low- or no-exposure group was significantly increased. The low- and non-exposed groups were not significantly different from each other in mutant frequencies.

Similarly, the concentration of the BD metabolite in urine was significantly greater in the high exposure group than in the lower- or non-exposed groups. There was a strong correlation among exposed subjects between the level of metabolite in urine and the frequency of the hprt mutants (r=0.85). (Ex. 118-2A) Another study of humans for potential cytogenetic effects of BD exposure was reported recently by Sorsa et al. in which peripheral blood was drawn from 40 BD production facility workers and from 30 controls chosen from other departments of the same plants, roughly matched for age and smoking habits. (Ex. 124) Chromosome aberrations, micronuclei and sister-chromatid exchanges were analyzed. No exposure related effects were seen in any of the cytogenetic endpoints. The typical exposure was reported as less than 3 ppm. (Ex. 124) Among the limited number of human studies involving BD exposed workers is that of Osterman-Golker who evaluated post-exposure adduct formation in the hemoglobin of mice, rats, and a small number of workers. (Ex. 117-2, p. 127) Mice and rats were exposed at 0, 2, 10, or 100 ppm for 6 hours per day, 5 days per week for 4 weeks and their blood tested for the presence and quantity of the BD metabolite, 1,2-epoxybutene, forming an adduct with the N-terminal valine of hemoglobin. The result was a linear response for mice at 2, 10 and 100 ppm; and, for rats at 2 and 10 ppm, with the 100 ppm dose group deviating from linearity. In addition, while the adduct level per gram of globin in the 100 ppm rats was about 4 times lower than the level observed in mice exposed to 100 ppm BD, at lower exposures, the adduct levels were similar.

In the portion of the study dealing with effects on humans, blood was taken from four workers in two areas of a chemical production plant with known BD exposure, and five workers from two non-production areas where BD concentrations were low. In the higher exposure area, the mean BD exposure was about 3.5 ppm, as determined by environmental sampling. The lower exposure areas had a mean BD level of about 0.03 ppm. On a mole of adduct per gram of hemoglobin level, the adduct levels in the higher BD exposed workers were 70 to 100 times lower than those of either the rat or mice exposed at the 2 ppm level discussed above. Production workers had adduct levels ranging from 1.1 to 2.6 pmol/g globin. Most controls in the study were below the level of detection of the assay (0.5 pmol adduct/g globin). (Two heavy smokers reported from a previous study had higher adduct levels than non-smokers; their levels approached those observed in BD exposed workers and were consistent with the amount of BD in mainstream smoke.) Similar results for mice and rats exposed to BD were reported by Albrecht et al. (Ex. 117-2, p. 135) In this study which exposed the rodents to 0, 50, 200, 500 or 1300 ppm for 6 hours/day, for 5 consecutive days, BD monoepoxide adduct levels in the hemoglobin of mice were about five times that of the rat at most BD exposure concentrations. Humans were not studied in this report.

Another observation pertaining to human cytogenetics with potentially important implications for BD-induced human disease is contained in a report by Wiencke and Kelsey. (Ex. 117-2, p. 265) These researchers studied the impact of the BD metabolite, diepoxybutane, exposure on sister chromatid exchange (SCE) frequencies in several groups of human blood cell cultures (n=173 healthy workers). They discovered that the study populations were bimodally distributed according to their sensitivity to induction of SCEs when cell cultures were exposed to 6 uM diepoxybutane. Wiencke and Kelsey reported that they had observed in earlier studies that "genetic deficiency of glutathione S-transferase type u leads to bimodal induction of SCEs by epoxide substrates of the isozyme" and that cells from individuals with the deficiency had SCE induction scores that were significantly higher than those observed in the general population. (Ex. 117-2, p. 271) Approximately 20% of the tested groups were sensitive to induction of SCE and the remaining 80% were relatively insensitive.(1) Subsequent testing indicated that the sensitive population was also sensitive to induction of chromosomal aberrations by diepoxybutane with significant increases in the frequencies of chromatid deletions, isochromatid deletions, chromatid exchanges and total aberrations. The relevance of these findings in not yet clear; however, they may indicate that certain subsets of the population are more highly susceptible to the effects of this mutagenic metabolite of BD.


Footnote(1) For example, in the 58 newspaper workers tested, 24% had greater than 95 SCE/cell, while the remaining 76% had fewer than 80 SCE/cell.

Recio et al. used transgenic mice containing a shuttle vector with a recoverable lac 1 gene to study in vivo mutagenicity of BD and the spectrum of mutations produced in various tissues. (Ex. 118-7D) Mice were exposed to 62.5, 625 or 1250 ppm BD for 4 weeks (5 days/week, 6 hours/day). The investigators extracted DNA from bone marrow and determined mutagenicity at the lac 1 transgene.

The mutant DNA was sequenced. Dose-dependent mutagenicity--up to a 3-fold increase over air controls--was observed among mice exposed at 625 or 1250 ppm. Although a number of differences in patterns were noted, the most striking was that sequence analysis indicated an increased frequency of in vivo point mutations induced by BD exposure at adenine and thymine (A:T) base pairs following inhalation.

In further studies of BD-exposed transgenic mice, Sisk and co-workers exposed male B6C3F(1) mice to 0, 62.5, 625, or 1250 ppm, BD for 4 weeks (6 hour/day, 6 days/week). (Ex. 118-7Q) Bone marrow cells were isolated and mutation frequency and spectrum evaluated. Lac 1 mutation frequencies were significantly increased at all 3 exposure levels and were dose-responsive in the 62.5 and 625 ppm BD-exposed mice, compared to controls. A plateau in mutation frequencies was observed at 1250 ppm BD-exposed mice, perhaps indicating saturation or mutant loss due to the effects of high level exposure.

When the mutants were sequenced, several from the same animal were found to have identical mutations. Although they might have arisen independently, Sisk et al. felt that this was likely due to clonal expansion of a bone marrow cell with a mutated lac 1 gene.

As had Recio et al., Sisk et al. observed a higher frequency of mutations at A:T sites in the exposed mice DNA, compared with controls. A:T to G:C transitions comprised only 2% of the background mutations, but made up 15% of those in the exposed mice.

Sisk et al. concluded that their observation coupled with in vitro studies " * * * suggest that BD may mutate hematopoietic stem cells." (Ex. 118-7Q, p. 476) As discussed in the animal carcinogenicity section in this preamble, BD-induced mouse tumors have been found to have activated proto-oncogenes. Specifically, the K-ras oncogene is activated and is the most commonly detected oncogene in humans. (Ex. 129) OSHA concludes that BD is mutagenic in a host of tests which show point and frameshift mutations, hprt mutations, chromosome breakage, and SCEs in both animals and humans. The data suggest that mice are more susceptible than rats to these alterations. In addition, certain subsets of the human population may be more susceptible to the effects of BD exposure than others (based on the Wiencke and Kelsey study of human blood cell cultures, Ex. 117-2, p. 265). OSHA further notes with concern the fact that the data suggest that BD exposure at relatively low levels adversely affects DNA repair mechanisms in humans and is associated with mutational effects.

5. Metabolism

In vitro genotoxicity studies have shown that BD is mutagenic only after it is metabolically activated. Biotransformation is probably also important to the carcinogenicity of this gas. It is thought that the formation of epoxides, specifically epoxybutene, also termed the "monoepoxide" and 1,2:3,4-diepoxybutane, termed the "diepoxide," is required for activity and that the reaction is cytochrome P450 mediated(2). Both the mono- and diepoxide are mutagenic in the Salmonella assay, with the diepoxide being more active. The reactive epoxides can bind to DNA, and formation of DNA adducts is hypothesized to initiate a series of events leading to malignancy.


Footnote(2) Cytochrome is defined as any of a class of hemoproteins whose principal biologic function is electron transport by virtue of a reversible valency change of its heme iron. Cytochromes are widely distributed in animal and plant tissues.

As described earlier, for most cancer sites, mice are more sensitive than rats to the carcinogenic effects of BD exposure. Studies of the metabolism of BD have been undertaken in an attempt to elucidate the contributions of dose-metric factors for the observed differences in carcinogenicity between the species.

Much of the research in this area has been performed at the Chemical Industry Institute of Toxicology and in German laboratories. Work on metabolism of BD was described by OSHA in the 1990 proposal. (55 FR 32736 at 32756) OSHA reviewed the current literature in the record and concluded:

1. The rate of metabolism of BD in mice is approximately twice that in rats;

2. Mice accumulate more radiolabelled BD equivalents in a 6 hour exposure than do rats at the same concentration;

3. Mice have about twice the concentration of the metabolite (1,2-epoxy-3-butene) (BMO) in blood as rats exposed at similar concentrations;

4. Over a wide range of exposures, mice received a larger amount of inhaled BD per unit body weight than rats, and had a higher concentration of BMO in the blood than rats (As expected, because of body size differences and breathing rates, and some enzymology);

5. BD is readily absorbed and widely distributed in tissues of both mice and rats, with tissue concentrations per umole BD inhaled higher in mice than in rats, by factors of 15-fold or more;

6. While there are species differences in the amount of BD metabolism at various sites, both mice and rats metabolize BD to the same reactive metabolites suspected of being ultimate carcinogens.

In comments on OSHA's proposal, Dr. Michael Bird of Exxon testified on behalf of the CMA BD Task Group that the mouse "will attain a significantly higher amount of the epoxides over a longer period of time than the rat. . . or primate when exposed to butadiene." (Ex. 52, p. 27) Dr. Bird concluded that the differences in metabolism of BD in the species help "explain the greater sensitivity of the mouse to BD carcinogenic activity." He further concluded that the differences in rates of enzyme mediated processes indicate non-human primates have lower internal concentrations of BD or BMO, and "man is more similar to the primate with respect to 1,2-epoxy-3-butene formation than the rat or mouse." (Ex. 52, p. 22) He argued that the mouse may be "uniquely sensitive " to BD carcinogenicity due to its greater uptake, faster BD metabolism and "elimination of the epoxide 1,2-epoxy-3-butene is saturable in mice but not in rats." (Ex. 52, p. 21) He felt this observation correlated well with the observed cytogenetic and bone marrow response (seen in mouse, but not rats.) Others hold an opposing view, e.g., Melnick and Kohn argued that "[b]ecause the rat appears to be exceptionally insensitive to leukemia/lymphoma induction, the mouse must be considered as the more appropriate model for assessing human risk for lymphatic and hematopoietic cancers." (Ex. 130, p. 160) Dr. Bird urged OSHA to use the monkey data of Dahl, et al. which indicated that the retention rate for BD in primates is over 6 times lower than that for the mouse, in "drawing any firm conclusions about the cancer risk to humans." (Ex. 52, p. 36) During the public hearing, the work of Dahl was presented as a preliminary report. (Ex. 44) Dahl exposed 3 cynomolgus monkeys to BD and measured uptake and metabolism. Each animal was exposed to three concentrations of C(14)-labeled BD, progressing from 10,300 to 8000 ppm with at least 3 months separating the re-exposure of each monkey. Post-exposure blood was taken. Each animal's breathing frequency and tidal volume was measured.

Dahl and co-workers found BD uptake to be lower in monkeys than in rats. The reported blood levels of the epoxides were also lower in the monkey than the levels reported by Bond et al. in rats and mice.

Dahl et al. attempted to quantitate total BD metabolites through collection of feces, urine and exhaled material though use of cryogenic traps. Measurement of residual labeled material retained in the animals at the end of the 96 hour post exposure period was not determined. HPLC (high-performance liquid chromatography) identification of the trapped material (at 95 C) indicated that only 5 to 15% of the radioactivity was present as monoepoxide.

Melnick and Huff, in reviewing this study, found its significance "clouded" because only three animals of unknown age were studied and there was uncertainty about the ability of vacuum line cryogenic distillation alone to identify and quantitate BD metabolites. (Ex. 114, p. 133) In testimony at the public hearing, Dr. James Bond of CIIT acknowledged the limitations of the use of vacuum-line cryogenic distillation as follows:

* * * there will be some material no matter what kind of vacuum you apply to it * * * simply will not move into the traps. That's referred to as non-volatile material.

We don't know what that material is and I think that's an important component of this study, because, in fact, in many cases it can represent 70 to 80 percent of the material that actually distills out. (Tr. 1/22/91, p. 1553)

Melnick and Huff were also concerned that only the monkeys, not the mice or rats, were anesthetized during exposure and question what impact that might have had on respiratory rates and cardiac output and what the influence might be on inhalation pharmacokinetics of BD. (Ex. 114, p. 133) In their 1992 review, Melnick and Huff concluded that studies to date have not revealed species pharmacokinetic differences of sufficient magnitude "to account for the reported different toxic or carcinogenic responses in one strain of rats compared to two strains of mice." (Ex. 114, p. 134) In post hearing comments Dr. David A. Dankovic of NIOSH reviewed this topic and concluded "* * * the most prudent course is to base 1,3-butadiene risk assessments on the external exposure concentration, unless substantial improvements are made in the methodology used for obtaining `internal' dose estimates." (Ex. 101, Att. 2, p. 5)

Recent Studies

Recent studies have focused on the metabolism of BD to the epoxides, epoxybutene and diepoxybutane, and their detoxification by epoxide hydrolase and glutathione. Bond et al. recently reviewed BD toxicologic data. (Ex. 118-7G) Epoxybutene and diepoxybutane were reported to be carcinogenic to mice and rats via skin application and/or subcutaneous injection, with the diepoxide having more carcinogenic potency. Bond et al. also concluded that the diepoxide is more mutagenic than the monoepoxide by a factor of nearly 100 on a molar basis. The diepoxide also induces genetic damage in vitro mammalian cells (Chinese hamster ovary cells and human peripheral blood lymphocytes). These studies are summarized in this preamble discussion of reproductive effects.

In vitro metabolic studies

In 1992 Csanady et al. reported use of microsomal and cytosolic preparations from livers and lungs of Sprague-Dawley rats, B6C3F(1) mice and humans to examine cytochrome P450-dependent metabolism of BD. (Ex. 118-7AA) The preparations were placed in sealed vials and BD was injected by use of a gas-tight syringe. Air samples were taken from the head space at 5 minute intervals and analyzed by gas chromatography for epoxybutene.

Cytochrome P450-dependent metabolism of the monoepoxide to the diepoxide was examined. Enzyme mediated hydrolysis of BMO by epoxide hydrolase was measured. (Non-enzyme mediated hydrolysis was determined using heat-inactivated tissue and none was observed.) Second order rate constants were determined using 100 mM monoepoxide and 10 mM GSH. The human samples were quite variable, with rates ranging from 14 to 98 nmol/min/mg protein.

The maximum rates for BD oxidation to monoepoxide (Vmax) were determined to be highest for mouse liver microsomes(3) (2.6 nmol/mg protein/min); the Vmax values for humans were intermediate, at 1.2 nmol/mg protein/min; the Vmax values for rats was 0.6 nmol/mg protein/min. For lung microsomes, the Vmax in the mouse was found to be similar to the mouse liver rate, but over 10-fold greater than that of either humans or rats.


Footnote(3) A microsome is defined as one of the finely granular elements of protoplasm, resulting from fragmentation (homogenization) of the endoplasmic reticulum.

From these data Csanady et al. calculated a ratio of activation to detoxification for each species tested. These values, expressed as mg cytosolic protein/gm liver [glutathione-S-transferase is a cytosolic enzyme], resulted in the determination of an overall activation:detoxification ratio of 12.3 for the mouse, 1.3 for the rat, and 4.4 for the human samples.

If these in vitro liver microsomal studies can be extrapolated to the whole animal in vivo, then this implies, as pointed out by Kohn and Melnick, that the mouse produces 2.8 times as much BMO per mol of BD as the human and that the human activation:detoxification ratio is 3.4 times that of the rat. However, the Csanady et al. study demonstrated a wide variability in BD metabolic activity among the 3 human liver microsomes, and a 60-fold variation was found in 10 human liver samples by Seaton et al. (Ex. 118-7N) Kohn and Melnick noted that this human variability in CYP2E1, the P450 enzyme primarily responsible for the activity, suggests that a "* * * fraction of the human population may be as sensitive to butadiene as mice are." (Ex. 131, p. 620).

A study similar to that of Csanady et al., reported by Duescher and Elfarra in 1994, determined that the Vmax/Km ratios for BD metabolism in human and mouse liver microsome were similar and were nearly 3 to 3.5 fold higher than the ratio obtained with rat liver microsomes. (Ex. 128) Duescher and Elfarra suggest that differences between their results and those of Csanady et al. may have been due in part to experimental methodology differences, such as incubation and assay methods. Duescher and Elfarra found that two P450 isozymes, 2A6 and 2EI, were most active in forming BMO of the 7 isozymes tested. They concluded that since human liver microsomes oxidized BD at least as efficiently as mouse liver microsomes (and much more so than rat liver microsomes), this "suggests that if [BMO] formation rate is the primary factor which leads to toxicity, humans may be at higher risk of expressing BD toxicity than mice or rats, and that the mouse may be the more appropriate animal model for assessing toxicity." Duescher and Elfarra felt that since P450/2A6 appears to play a major role in BD oxidation in human liver microsomes, and that it is more similar to that of mouse P450/2A5 than to rat P450/2A1, the mouse may be a better model to use in assessing human risk.

In 1994 Himmelstein et al. hypothesized that "[S]pecies differences in metabolic activation and detoxification most likely contribute to the difference in carcinogenic potency of BD by modulating the circulating blood levels of the epoxides." (Ex. 118-13, Att 3) To address this, Himmelstein and colleagues looked at the levels of BD, BMO, and BDE in blood of rats and mice exposed at 62.5, 625, or 1250 ppm BD. Samples were collected at 2, 3, 4, and 6 hours of exposure for BD and BMO and at 3 and 6 h for the BDE. Blood was collected from mice by cardiac puncture and from rats through an in-dwelling jugular cannula. Melnick and Huff criticized earlier studies which failed to use in-dwelling cannulae.

Because steady state levels of [monoepoxide] are lower in rats than in mice and because the metabolic elimination rate for this compound is 5 times faster in rats than in mice, any delay in obtaining immediate blood samples would have a much greater effect on analyses in blood samples obtained from rats than those obtained from mice. (Ex. 114, p. 133)

Himmelstein et al. found that the concentration of BD in blood was not directly proportional to the inhaled concentration of BD, suggesting that the uptake of BD was saturable at the highest inhaled concentration. In both rats and mice BD and the BMO blood levels were at steady state at 2, 3, 4 and 6 hours of exposure and declined rapidly when exposure ceased. This is consistent with exhalation being the primary route of elimination of BD. (Ex. 118-7B) Genter and Recio used Western blot and immunohistochemical analyses to detect P450/2E1 in bone marrow of B6C3F(1) mice. (Ex. 118-7T) Although both methods detected the presence of the protein in livers of both male and female mice, non was seen in the bone marrow. The limits of detection were not stated in the report. The author hypothesized the BD might be converted to the monoepoxide in the liver prior to uptake by the bone marrow or that another pathway (e.g., myeloperoxidase) is responsible for BD oxidation in the marrow. Recio and Genter suggest that the greater sensitivity of mice to BD-induced carcinogenicity can be explained in part by the higher levels of both epoxides in the blood of mice compared with that of rats.

Himmelstein et al. furthered this work in 1995 in a report in which they determined levels of the epoxides in livers and lungs of mice and rats exposed to BD. (Ex. 118-7/O) Animals were exposed at 625 or 1250 ppm of BD for 3 or 6 hours. Himmelstein et al. found that in mice exposed to this regimen, the monoepoxide levels were higher in lungs than in livers. Rats at 625 and 1250 ppm had lower concentrations of BMO in lungs and livers than mice. When rats were exposed to 8000 ppm BD, the maximum concentration of BMO in the lung and liver was nearly the same. The diepoxide levels in lungs of mice exposed at 625 and 1250 ppm were 0.71 and 1.5 nmol/g respectively. The diepoxide was not detected in livers or lungs of rats exposed at any tested level.

Himmelstein et al. also observed depletion of glutathione in liver and lung samples from both rodent species. Following 6 hours of exposure, the lungs of mice exhibited greater depletion of GSH than mouse liver, rat liver or rat lung at all concentrations of BD tested. The conclusion reached by the study authors was that their data indicate that GSH depletion is associated with tissue burden of the epoxides and that this target organ dosimetry might help explain some of the non-concordance of cancer sites observed between the species. OSHA notes, however, that while % GSH depletion was highest in the mouse lung, the major increase in depletion was at 1250 ppm BD, while lung tumor incidence was increased in the female mice at 6.25 ppm and in male mice at 62.5 ppm. Depletion of glutathione was dependent on concentration and duration of BD exposure.

Himmelstein et al. stressed the importance of the fact that the diepoxide was detected in the mouse lung but was not quantifiable in the mouse liver, and stated that if the diepoxide was formed in the liver, it is rapidly detoxified or otherwise moved out of the liver. They also found that depletion of glutathione was greater in mouse than rat tissues for similar inhaled concentrations of BD and concluded that conjugation of the monoepoxide with glutathione by glutathione S-transferase is an important detoxification step.

In contrast to rats and mice, lungs and livers from humans had much faster rates of microsomal monoepoxide hydrolysis by epoxide hydrolase compared to cytosolic conjugation with glutathione by the transferase. (Ex. 118-7AA) Thornton-Manning et al. in 1995 examined the production and disposition of monoepoxide and diepoxide in tissues of rats and mice exposed at 62.5 ppm BD. (Ex. 118-13, Att. 3) They found monoepoxide was above background in blood, bone marrow, heart, lung, fat, spleen and thymus tissues of mice after 2 or 4 hours of exposures to BD. In rats, levels of monoepoxide were increased in blood, fat, spleen and thymus tissues. No increase in monoepoxide in rat lung was observed. The more mutagenic diepoxide was detected in all tissues of the mice examined immediately following 4 hours of exposure. It was detected in heart, lung, fat, spleen and thymus of rats, but at levels 40- to 160-fold lower than those seen in mice.

In mice, the level of diepoxide exceeded the monoepoxide levels immediately after exposure in such target organs as the heart and lungs. Thornton-Manning et al. concluded that the high concentrations of diepoxide in heart and lungs they observed suggested to them that this compound may be particularly important in BD-induced carcinogenesis.

The study authors noted that neither epoxide was detected in rats' liver and was present only in quite low concentrations in the livers of mice. Thornton-Manning et al. found this surprising since epoxides present in blood in the liver should have yielded values greater than those observed in the liver samples. They hypothesized that it might be due to prior metabolism of the epoxides before reaching the liver or it might be an artifact due to post-exposure metabolism of the epoxides in the liver.

Thornton-Manning et al. did not detect the monoepoxide in rat lungs, and found the diepoxide level to be quite low. In contrast, in the mice they found both epoxides present in lung tissue, with the monoepoxide level present at a concentration less than expected using blood volume values, and the diepoxide level agreeing with that expected as a function of blood volume. Thornton-Manning et al. concluded that these results "* * * suggest that the lung is capable of metabolizing BDO, but perhaps is less active in metabolizing BDO(2). (Ex. 118-13, Att. 3) Moreover, Thornton-Manning et al. believed that although BD is oxidatively metabolized by similar metabolic pathways in the rats and mice, the quantitative differences in tissue levels between species may be responsible for the increased carcinogenicity of BD in mice.

                    EXPOSURE TO 62.5 PPM BD BY INHALATION

                          |     Epoxybutene     |    Diepoxybutane
           Tissue         |_____________________|_____________________
                          |    Rats  |    Mice  |   Rats   |  Mice
Blood.....................|    36+/-7|  295+/-27|     5+/-1| 204+/-15
Heart.....................|   40+/-16|  120+/-15|   3+/-0.4| 144+/-16
Lung......................|        ND|    33+/-9| 0.7+/-0.2| 114+/-37
Liver.....................|        ND|     8+/-4|        ND|   20+/-4
Fat.......................|  267+/-14|1302+/-213| 2.6+/-0.4|  98+/-15
Spleen....................|     7+/-6|   40+/-19| 1.7+/-0.5|  95+/-12
Thymus....................|12.5+/-3.2|  104+/-55| 2.7+/-0.7| 109+/-19
Bone marrow(1)............| 0.2+/-0.1| 2.3+/-1.5|        ND|1.4+/-0.3
  ND = Not Detected.
  Footnote(1) Bone marrow data are presented as mean pmol/mg protein
+/_, n=3 or 4 each determination. Adapted from Ex. 118-13, Att.3.

These data are shown in Table V-8. Seaton et al. examined the activities of cDNA-expressed human cytochrome P450 (CYP) isozymes for their ability to oxidize epoxybutene to diepoxybutane. (Ex. 118-7N) They also determined the rate of formation of the diepoxide by samples of human liver microsomes (n=10) and in mice and rat liver microsomes. Seaton et al. found that two of the cytochrome P450 isozymes, CYP2E1 and CYP3A4, catalyzed oxidation of 80 uM of monoepoxide to detectable levels of diepoxide, and that CYP2E1 catalyzed the reaction at higher levels of monoepoxide (5mM), suggesting the predominance of 2E1 activity at low substrate concentrations. Hepatic microsomes from all 3 species formed the diepoxide when incubated with the monoepoxide. Seaton et al. hypothesized that the difference between these results and those of Csanady et al. (who did not detect the diepoxide when the monoepoxide was substrate in a similar microsomal assay) was due to differences in experimental methodology.

Seaton et al. noted a 25-fold variability in Vmax/Km among the 4 human livers. They reported that Vmax/Km for oxidation of the monoepoxide to the diepoxide for the 4 human samples was 3.8, 1.2, 1.3 and 0.15, while that of the pooled rat samples was 2.8, and the mouse ratio was 9.2.

The authors, using available data, calculated an overall activation/detoxification ratio (Vmax/Km for oxidation of BD to the monoepoxide) taking into account hydrolysis of the monoepoxide by epoxide hydrolase and conjugation with glutathione. The activation/detoxification ratio was estimated at 1295 for the mouse, 157 for rats and 230 for humans. However, Melnick and Kohn point out that "when yields of microsomal and cytosolic protein content and liver size were considered, the activation to detoxification ratio was only 2.8 times greater in mice than in humans and 3.4 times greater in humans than in rats. These ratios do not take into account inter-individual variability in the activities of the enzymes involved." (Ex. 131) Recently, Seaton et al. studied production of the monoepoxide in whole airways isolated from mouse and rat lung. (Ex. 118-7C) They explained the impetus to use fresh intact tissue by stating that lung subcellular fractions, as employed in experiments by Csanady et al., described above, contained mixtures of cell type "so that the metabolizing capacities of certain cell populations may have been masked." They anticipated that use of airway tissue would allow more precise quantitation of differences in lung metabolism of BD.

Whole airways or bronchioles isolated from both male B6C3F(1) mice and male Sprague-Dawley rats were incubated for 60 min with 34 um BD. Levels of 10.4+/-5.6 nmol epoxybutene/mg protein were detected in mouse lungs, while 2-3 nmol/mg protein was observed in rat lung airway regions. Seaton et al. noted that while the species differences "are not dramatic," they may in part contribute to the differences in carcinogenicity observed in mice and rats.

To characterize conjugation of BD metabolites with glutathione (GSH), Boogard et al. prepared cytosol from lungs and livers of rats and mice and from 6 human donor livers and incubated them with 0.1 to 100 mM diepoxide and labeled glutathione (GSH). (Ex. 118-7J) NMR (nuclear mass resonance) and HPLC techniques were used to characterize and quantitate conjugate formation.

Non-enzymatic reaction was concluded to be negligible. The conjugation rates (Vmax) in mouse and rat livers were similar and 10-fold greater than those observed in the human samples. The initial rate of conjugation (Vmax) was much higher in mouse than rat lung. Both rodent species exhibited higher initial rates of conjugation than human. This led Boogard et al. to conclude that the higher diepoxide levels observed in BD-exposed mice compared with rats "are not due to differences in hepatic or pulmonary GSH conjugation of BDE (the diepoxide)," and further that since humans oxidize BD to the epoxides at a low rate, the low activity of GSH conjugation of the diepoxide in human liver cytosol demonstrated in this study "will not necessarily lead to increased BDE (diepoxide) levels in humans potentially exposed to BD." They also pointed out the need to determine the rate of BDE detoxification by other means, specifically by epoxide hydrolase in all three species.

Studies of Urinary Metabolites of BD

Two metabolites of BD have been identified in urine of exposed animals by Sabourin et al. (Ex. 118-13 Att. 3) These are 1,2-dihydroxy-4-N-acetylcysteinyl-S-)-butane, designated MI, and MII, which is 1-hydroxy-2-N-acetylcysteinyl-S-)-3-butene. (Ex. 118-13-Att. 3) These mercapturic acids are formed by addition of glutathione (GSH) at either the double bond (MI) or the epoxide (MII). MI is thought to form by conjugation of GSH with butenediol, the hydrolysis product of the monoepoxide, while MII is thought to form from conjugation of the monoepoxide with GSH.

Sabourin et al. measured MI and MII in urine from rats, mice, hamster and monkeys. Mice were observed to excrete 3 to 4 times as much MII as MI, while the hamsters and rats produced about 1.5 times as much MII as MI. The monkeys produced primarily MI.

The ratio of formation of metabolite I to the total formation of the two mercapturic acids, MI and MII, correlated well with the known hepatic epoxide hydrolase activity in the different species, suggesting that the monoepoxide undergoes more rapid conjugation with glutathione in the mouse than in the hamster or rats, and that the least rapid conjugation occurs in the monkey. The epoxide availability is inversely related to the hepatic activity of epoxide hydrolase, which removed the epoxide by hydrolysis.

In 1994, Bechtold et al. published a paper describing a comparison of these metabolites between mice, rats, and humans.(4) In workers exposed to historical atmospheric concentrations of 3 to 4 ppm BD, Bechtold measured urine levels of MI and MII by use of isotope-dilution gas chromatography, and found MI, but not MII, to be readily detectable. Bechtold et al. found that employees who worked in production areas (having 3-4 ppm BD exposure) could be distinguished by this assay from outside controls and that low level human exposure to BD resulted in formation of epoxide.


Footnote(4) A preliminary study on the human population of this study is described in the section of this preamble dealing with the genetic toxicology of BD exposure.

Bechtold et al. stated in their abstract that since monkeys displayed a higher ratio of MI to MI + MII than mice did, and "because humans are known to have epoxide hydrolase activities more similar to those of monkeys than mice, we postulated that after inhalation of butadiene, humans would excrete predominantly MI and little MII." (Ex. 118-13 Att. 3) Their observations suggested that the predominant pathway for clearance of the monoepoxide in humans is by hydrolysis rather than conjugation with glutathione.

Bechtold et al. found when mice and rats were exposed to 11.7 ppm BD for 4 hours and the ratio of the two metabolites was then measured, for mice, the ratio of MI to MI +/- MII (or the % of total which is MI) was 20%, that of rats was 52%, while humans exhibited more than 97% MI. These data also indicate the predominance of clearance by hydrolysis pathways rather than GSH conjugation in the human.

Nauhaus et al. used NMR techniques to study urinary metabolites of rats and mice exposed to ([(1,2,3,4)-(13)C]-butadiene). (Ex. 118-7I) They characterized metabolites in mouse and rat urine following exposure by inhalation to approximately 800 ppm BD for 5 hours. Urine was collected over 20 hours from exposed and control animals, centrifuged and frozen.

The findings of this study are quite extensive and are briefly summarized as follows. Nine metabolites were detected and chemically identified in mouse urine and 5 in that of rats. Five were similar in the 2 species, though differing markedly in concentration. One was unique to the rat and four to the mouse. Nauhaus et al. observed that "when normalized to body weight (umol/kg body weight), the amount of diepoxide-derived metabolites was four times greater in mouse urine than in rat urine." They further hypothesized that "the greater body burden of (diepoxide) in the mouse and the ability of rats to detoxify [it] though hydrolysis may be related to the greater toxicity of BD in the mouse." Nauhaus et al. found that both mice and rats conjugated the monoepoxide with glutathione, but the rat preferentially conjugated at the two carbon, while the mouse preferentially conjugated at the one carbon. Additionally, the finding of a metabolite of 3-butenal, a proposed intermediate in the oxidation of BD to crotonaldehyde, an animal carcinogen, is suggestive of an alternative carcinogenic pathway for BD. In general, this study supports the in vitro findings of Csanady et al. who reported similar rates for BMO conjugation with glutathione between rats and mice. (Ex. 118-7AA)

Interaction of Butadiene With Other Chemicals

Bond et al. described use of available data to simulate the potential interaction of BD with other workplace chemicals. (Ex. 118-7V) Specifically they modeled potential interaction assuming competitive inhibition of BD metabolism by styrene, benzene and ethanol. The model predicted that co-exposure to styrene would reduce the amount of BD metabolized, but that because of its relative insolubility, BD would not effectively inhibit styrene metabolism. Benzene, which, like BD, is metabolized by P450/2E1, was also predicted to be a highly effective inhibitor of BD metabolism because of its solubility in tissues. The models predicted that ethanol would have only a marginal effect on BD metabolism at concentrations of BD "relevant to human exposure."

BD and styrene co-exposures often occur in the SBR industry and both are metabolized by oxidation to active metabolites, in major part, by cytochrome P450/2E1. To determine the metabolic effect of joint exposure to BD and styrene, Levans and Bond developed and compared two PBPK models, one with one oxidative pathway and competition between BD and styrene and the other with two oxidation pathways for both BD and styrene. (Ex. 118-7E) For model validation, Levans and Bond exposed male mice to mixtures of BD and styrene of 100 or 1000 ppm BD and 50, 100 or 250 ppm styrene for 8 hours. They used chamber inlet and outlet concentrations to calculate uptake and, when steady-state was reached, calculated the rate of metabolism. They analyzed blood for styrene, styrene oxide, epoxybutene and diepoxybutane by GC-MS.

Leavens and Bond found BD metabolism was inhibited when mice were co-exposed to styrene. The inhibition approached maximum value at co-exposure concentrations of styrene above 100 ppm.

The report also described the preliminary development of pharmacokinetic models to simulate the observed rate of BD metabolism in co-exposed mice. Their results supported the hypothesis that "more than one isozyme of P450 metabolized BD and styrene and competition does not occur between BD and styrene for all isozymes." They were unable to accurately predict blood concentrations of styrene following exposure, and felt that " perhaps the diepoxide may inhibit metabolism of styrene by competing for the same P450 enzyme."

Although preliminary in nature and reflecting effects of relatively high exposures, these observations of interactions between styrene and BD exposure may have implications for the observed pattern of BD-induced effects in human populations jointly exposed. Specifically, the cancer effects seen in SBR production workers may underestimate the effects of BD with no styrene or benzene exposure.

Pharmacokinetic Modeling of BD Metabolism

In a recent publication, Bond et al. reviewed the results of application of a number of physiologically-based pharmacokinetic (PBPK) dosimetry models. (Ex. 118-7M) They noted that three of the models which included monoepoxide disposition (Kohn and Melnick, Johanson and Filser, Medinsky) predicted that, for any BD exposure concentration, steady-state monoepoxide levels will be higher for mice than for rats. Bond et al. further observed that "while the three models accurately predict BD uptake in rats and mice, they overestimate the circulating blood concentrations of (monoepoxide) in these species compared to those experimentally measured by Himmelstein." Their results also led Bond et al. to conclude that the disagreement between model predictions for the monoepoxide and experimental data suggests that the structure and/or parameter values employed in these models are not accurate for predicting blood levels of BD epoxides, and conclusions based on model predictions of BD epoxide levels in blood or tissue may be wrong." (Ex. 118-7M, p. 168) OSHA agrees with these authors that BD epoxide levels should not be used in assessing risk. In the discussion, the authors pointed to the need for inclusion of diepoxide toxicokinetics (as well as that of the monoepoxide) in future modeling exercises, since they believe the diepoxide to be the ultimate carcinogenic metabolite of BD.

Kohn and Melnick, in a recent publication, used available data and attempted to apply a PBPK model to see whether it was consistent with observed in vivo uptake and metabolism. (Ex. 131) The model included compartments for rapidly and for slowly perfused tissues. Rate equations for monoepoxide formation, its hydrolysis, and for conjugation with glutathione were included.

Kohn and Melnick acknowledged numerous sources of uncertainty in applying the model to the data (in which there are many gaps), necessitating various assumptions. Their calculations led them to conclude that the "model reproduces whole-body observations for the mouse and rat" and that it predicts that "inhalation uptake of butadiene and formation and retention of epoxybutene are controlled to a much greater extent by physiological parameters than by biochemical parameters. . . " (Ex. 131) When Kohn and Melnick interchanged the biochemical parameters in the mouse and human models to see if "the differences in calculated net uptake of butadiene among the three species were due to differences in metabolic activity," they found that use of human parameters in the mouse model decreased the level of absorption of BD, but not to a level as low as that of the human. Kohn and Melnick noted that the model predictions of epoxybutene levels in the heart and lung of mice and rats failed to account for the observation that mice, but not rats, develop tumors at these sites. Kohn and Melnick suggested that factors other than epoxybutene levels, not accounted for in the model, are probably crucial to induction of carcinogenesis.


Many metabolism studies have been conducted both in vitro and in vivo, mostly in mice and rats, to determine the BD metabolic, distribution, and elimination processes, and these studies have been extended in attempts to explain, at least in part, the greater carcinogenic potency of BD in the mouse, whether the mouse or the rat is a better surrogate for human cancer and reproductive risk assessment, and what is the proper dose-metric to use in dose-response assessments. The question of whether the mouse or the rat is a better model for the human on the basis of tumor response is partly addressed in the risk assessment section of this preamble. This section more specifically considers whether these metabolic studies in total can explain the different cancer responses and potencies observed in the mouse, rat, and human. What is clear throughout the record is that most scientists who study the topic consider not BD itself, but the major epoxide metabolites of BD, BMO and BDE and 1, 2-epoxybutane-3,4-diol, to be the putative carcinogenic agents. Most of this research has focused on the relative species production of BMO and BDE. Both BMO and BDO have been reported in early studies to be carcinogenic to mice and rats via skin application and/or subcutaneous injection, with BDO being somewhat more potent. (Ex. 23-88, Ex. 125).

Metabolism of BD to BMO in both the liver and lung of mice, rats and humans is by the P450 oxidation pathway, with CYP2E1 and CYP1A6 being the major enzymes. Based on the studies reviewed by OSHA, overall the mouse metabolizes BD to the monoepoxide and the diepoxide in these organs at a faster rate than do the rat and human. This is supported by the following evidence: (1) The mouse has higher BMO and BDE levels in blood, lung, and liver (i.e., see Ex. 118-7S, Ex. 118-7D, and Ex. 118-13), which are the target organs for cancer in the mouse but not the rat; (2) the mouse has higher in vitro lung and liver microsome Vmax/Km ratios for both BD and BMO metabolism than do rats or humans (Ex. 118-7AA); and (3) the mouse has higher hemoglobin-BMO adduct levels than rats and much higher levels than humans. (Ex. 118-7Y) A major exception to the findings of these studies is the study by Duescher and Elfarra, who found the in vitro BD Vmax/km ratios to be the same in mice and human liver microsomes and 3-4 times higher than they were in rats, suggesting that mice and humans have similar BD metabolic potential, at least in the liver. (Ex. 128) Large variations, about 60 fold, were found among 10 human liver microsome BD metabolic activities. (Ex. 118-7N) A recent BD in vitro metabolism study by Seaton et al. on whole rat and mouse lung airway isolates found that the mouse produced about twice the amount of BMO as the rat (this difference could not explain the difference between mouse and rat tumor incidence). (Ex. 118-7C) BMO and BDE were also measured in heart, spleen, thymus, and bone marrow (target sites for mouse but not rat tumors) following 4 hour BD inhalation exposure (62.5 ppm) to mice and rats. (Ex. 118-13) In these tissues, mouse BMO and BDE levels were 3 to 55 fold higher than rat levels for the same metabolites, although the mice organ levels of these metabolites correlated poorly with the mouse target organ cancer response at this exposure level. Only high BDE levels in the mouse lung were consistent with the mortality adjusted cancer incidence (see hazard identification--animal studies section, Ex. 114). This suggests that BD metabolite tissue levels can, at best, only partly explain differences in carcinogenic response. Differences in both species and tissue sensitivity must also be accounted for.

The Thornton-Manning and other studies also provided information about BD elimination. (Ex. 118-7I) With higher experimental exposure levels, the major route of elimination of BD is via expiration. Elimination of BMO occurs by different pathways in different species and different organs. At higher BD exposure concentrations, some BMO is expired. The mouse liver and lung appear to eliminate BMO predominantly by direct conjugation with GSH(5). For the rat there is approximately equal elimination by the GSH and EH mediated pathways, while for the human and monkey hydrolysis to butanediol is the major pathway for excretion. ( Ex. 118-13 Att. 3) This species elimination pathway difference is a partial explanation for the higher levels of both BMO and BDE seen in the mouse, assuming that most of the BD metabolism takes place in the liver. With respect to the bone marrow BD distribution and metabolism, mouse levels of the BD metabolites in the bone marrow were lower than at any of the other target organs studied. (Ex. 118-13) In vitro studies by Gentler and Recio have found no detectable P4502E1 in the bone marrow of B6C3F(1) mice. (Ex. 118-7T) These authors conclude that this "suggests that BD is converted to BMO outside of bone marrow and is subsequently concentrated in bone marrow, or that the conversion of BD to BMO occurs by an alternate enzymatic pathway within the bone marrow." The latter appears to be the more likely since Maniglier-Poulet and co-workers showed that in vitro BD metabolism to BMO in both B6C3F(1) mouse and human bone marrow occur by a peroxidase-mediated process and not via the P450 cytochrome system. (Ex. L-133) Since in their system both human and mouse bone marrow generated about the same amount of BMO/cell, this suggests that both BD distribution to bone marrow and local metabolic reactions should be considered in species-to-species extrapolations and in PBPK modeling.


Footnote(5) One exception: Seaton et al. found evidence "that in mouse airways hydrolysis of BMO by epoxide hydrolase (EH) contributes to BMO detoxification to a greater extent than does glutathione conjugation." (Ex. 118-7C)

Inclusion of bone marrow local reactions becomes even more important when considering the animal species to use for modeling human cancer. BD is genotoxic in the bone marrow of mice, but not in rats. (Tice et al. 1987; Cunningham et al. 1986, reported in Ex. 131) BD and BMO have been implicated as affecting primitive hematopoietic bone marrow stem and progenitor cells related to both T-cell leukemia and anemia in the mouse. (Irons et al., 1993, in Ex. 117-2) BD causes lymphoma in mice, but no lymphoma or leukemia in rats even at 8,000 ppm. Furthermore, the body of epidemiologic evidence strongly indicates that BD exposure poses an increased risk of human leukemia (see the epidemiologic section and especially Ex. 117-1).

Fat storage of BD during exposure, and release following cessation of exposure, is also a major concern, both in estimating target organ levels and in determining species differences. There is little in the record on the effect of fat storage and release. In the Thornton-Manning study discussed above, both mouse and rat fat levels of both BMO and BDE declined rapidly following cessation of exposure, suggesting little lingering effect. However, Kohn and Melnick present a model in which post-exposure release of BD from the fat would result in extended epoxide production in humans in contrast with the mouse. (Ex. 131) Bond et al. suggest that the more rapid metabolism of BD to BMO in the mouse, and the more rapid EH BMO elimination pathways in the rat and human may be an explanation for lower, if any, BDE levels seen in rat and human liver microsomes and why BD will not be carcinogenic to humans at exposure levels seen in the environment or the workplace. (Ex. 130) They also conclude that "Since significant tumor induction in male rats occurs only at 8000 ppm BD, BMO levels are probably not predictive of a carcinogenic response." Thornton-Manning et al. characterize the peak levels of BDE in the mouse lung and heart as being either greater than or equivalent to peak levels of BMO, and suggest "that the formation of BDE may be more important than the formation of BMO in the ultimate carcinogenicity of BD." (Ex. 118-13) However, BMO levels in these organs were also quite high, and were higher than BDE levels in blood and bone marrow, target organs for hematopoietic system cancers. OSHA believes that the evidence is not sufficient to dismiss the potential contribution of BMO to mouse, rat or human carcinogenicity; to conclude that BDE should be considered more actively carcinogenic than BMO; or to find that BDE levels are sufficiently characterized in either mouse or human tissue to be used as the dose metric for BD human risk assessment.

Thus, OSHA concludes, based on the body of metabolic and other evidence presented, and the above discussion, that the mouse is a suitable animal model for the human for BD cancer risk assessment purposes, and that metabolism of BD to active metabolites is probably necessary for carcinogenicity. However, while the uptake, distribution, and metabolism of BD to active carcinogenic agents are important, local BD metabolic reactions and specific species sensitivities appear to have at least as large an impact on BD potency in the various species. This is likely to be especially true in the human, whose metabolic processes appear to be much more variable with respect to BD. Thus, although the metabolism studies provide insight into BD's metabolic processes in various species and organs (with the possible exception of mouse lung tumorigenicity related to lung BDE levels and protein cross linking), OSHA finds that too many questions remain unanswered, both with PBPK modeling efforts and with actual in vivo measurements (and the lack of such measurements in humans) to base a quantitative risk assessment on BD metabolite level equivalence between mice and humans. (Ex. L-132)

Regulations (Preambles to Final Rules) - Table of Contents Regulations (Preambles to Final Rules) - Table of Contents