Conclusions

• The subcommittee agrees with DPR's selection of the toxicological endpoints and the NOAELs used to derive the inhalation reference concentrations (RfC). The subcommittee considers the NOAELs to be protective and conservative.
• The subcommittee agrees that it is appropriate to use a developmental study for the derivation of an acute RfC for the general population.
• DPR's selection of the dog study (Newton 1994b) with a neurotoxicity endpoint is appropriate for developing a subchronic RfC, but the subcommittee is concerned about whether the decrease in responsiveness seen at exposure to 5 ppm of methyl bromide in two of eight dogs is a true LOAEL or even an effect at all. Nevertheless, the subtle neurological deficits observed in occupationally exposed humans (Anger et al., 1986) supports these animal data that neurotoxic responses can occur at low exposure concentrations. Therefore, the subcommittee concurs with the conservative assignment of the 5 ppm value in this dog study as a LOAEL.
• The rabbit developmental study had toxicity endpoints of gallbladder agenesis and fused sternebrae, which are not considered major malformations; however, the subcommittee feels that these are indicators of developmental toxicity, and therefore, are appropriate endpoints for the developmental RfC (Breslin et al. 1990b).
• The subcommittee agrees with DPR's selection of nasal epithelial hyperplasia as the toxicity endpoint for the chronic RfC, but notes that the effect is mild and might be closer to a NOAEL than a LOAEL.
• In general, DPR's presentation of the toxicological information is clear and easy to follow and permits the reader to follow DPR's logic in selecting critical studies and NOAELs.

Recommendations

• Methyl bromide is a methylating agent that is a direct-acting mutagen in vitro. However, there are good animal studies that indicate it is not carcinogenic. DPR should review the literature for any discussion on methyl bromide and other methylating agents as to why an in vitro mutagen is not an in vivo carcinogen. This could aid in understanding the mechanism of methyl bromide toxicity and lend confidence when extrapolating from the animal data to humans.
• The dog study from which the 6-week subchronic RfC is derived (Newton 1994b) had a small number of animals and some subjective observations that led to a LOAEL of 5 ppm. The subcommittee recommends that a new study be conducted to verify the neurotoxicity endpoints of decreased responsiveness at 5 ppm.
• Further developmental studies on methyl bromide would help to clarify several major issues
  • —Does in utero or early postnatal exposure to methyl bromide affect adult reproductive function? This question arises from the observation of apparently reduced fertility in the F1 offspring, but not the F0 parents in a two-generation study (American Biogenics Corporation 1986; Hardisty 1992; Busey 1993).
  • —What are the critical exposure periods for expression of reduced pup weights found during lactation and decreased offspring brain weights and dimensions (i.e., are they due to gestational or lactational exposure to methyl bromide?)
  • —Is methyl bromide excreted in breast milk? This question could be answered by measuring methyl bromide concentrations in the breast milk of lactating animals exposed to methyl bromide by inhalation.
  • —Does gallbladder agenesis occur following a single exposure to methyl bromide during the critical period for gallbladder development?

Conclusions

• Although the exact levels of exposure for workers and residents are not known, DPR has collected a substantial amount of data that indicate that some of these exposures are significant, exceeding current regulatory limits, and therefore are of concern.
• The measures of exposure are frequently based on a single value with no accompanying information on ambient air temperature, relative humidity, and wind conditions. The lack of representativeness of the measurements is not assessed in the main text of the DPR report and is only acknowledged as a possible confounder in an appendix.
• In general, the subcommittee is highly critical of the analysis and presentation of the available exposure data, finding it seriously deficient in understanding and application of modern concepts of variability and uncertainty, and in the fair evaluation of the magnitude and distribution of existing exposures relative to exposure levels intended to be achieved by current regulatory controls.
• There is considerable room for improvement in the methods used by DPR to obtain monitoring data, particularly with regard to good measurement techniques and sampling strategies that assess variability of actual exposure.
• Information is lacking on exposures to residents living near application areas and exposures for populations subject to aggregate applications (e.g,, those living in basin area where multiple fields have been treated). Available data and modeling suggest that for some populations, exposures might exceed regulatory limits.
• A substantial ambiguity exists for current methods used to measure methyl bromide in air, particularly with respect to recovery values and the field conditions during air sampling. As a result, actual measured air concentrations of methyl bromide and potential exposure levels are uncertain.
• DPR's use of 24-hr averaging for 8-hr exposures adds a further uncertainty to the protectiveness of the regulations.
• DPR's documentation of their exposure assessment is difficult to follow and requires searching through numerous appendices and other documents (many of which were requested by the subcommittee at a later date) to determine the data sources used by DPR and the approach that was used to evaluate and model the data. A roadmap of the information in the appendices and a more systematic presentation of the data would be helpful to the reader. In particular, DPR's discussion of buffer zones and the measurements taken at them, is confusing and appears to be missing important pieces of information.

Recommendations

• DPR should explicitly state what populations or subpopulations are expected to be represented by the scenarios.
• Identify the best analytical methods for determining methyl bromide concentrations in air under a variety of field conditions. The entire risk assessment process is fundamentally dependent on the quality of the analytical information on exposure conditions. A substantial effort is needed to develop rigorous and robust field analytical methods for determining concentrations of methyl bromide. This will require a complete multilaboratory series of tests that can allow a determination of the reliability of analytical information from field samples.
• Conduct systematic recovery analyses of field and laboratory air samples under a variety of air temperature, wind, and relative humidity conditions.
• Establish a new sampling program to determine the representativeness of exposure data with an emphasis on residential (including house fumigations) and high-exposure occupations.
• DPR should consider quantifying—at the very least—the potentially exposed populations in its occupational categories, and if possible, the number of residents near fields, fumigation facilities, and residents returning to fumigated homes.
• DPR should evaluate its exposure data using modern distributional concepts—including both variability and uncertainty to quantify how accurately the observed measurements represent the real distributions of exposure concentrations and durations. The subcommittee believes that analyses intended to support regulations should frankly disclose the expected degree of confidence the public should have that real exposures will be kept below regulatory levels for defined percentiles of exposed populations.

Conclusions

• The subcommittee overall agrees with the risk characterization for inhalation exposure of methyl bromide. The subcommittee believes that the toxicity endpoints used might be overly conservative due to their equivocal nature, but also believes that the exposure assessments might understate the actual exposures, particularly for residents living near fields where methyl bromide is applied.
• The subcommittee agrees that DPR's use of factors of 10 for intraspecies variation and for animal to human variation, as well as a benchmark margin of exposure (MOE) of 100, is consistent with traditional risk management practices.
• The subcommittee believes that the uncertainties associated with DPR's exposure levels call into question the validity of its MOE values. Given the likelihood that the error in the measurements will underestimate some exposures, the subcommittee anticipates that some MOEs will be lower than those calculated by DPR, some of which already indicate there is a cause for concern (i.e., they are currently less than 100).
• Given the lack of information on methyl bromide drift off-site from fumigated fields, it is unclear to the subcommittee how DPR can develop a coherent and protective plan for buffer zones and injection times for field fumigation as specified in Section 6450 of Title 3 of the California Code of Regulations.
• The subcommittee concludes that DPR has failed to conduct a true risk assessment in that it does not combine both exposure assessments and dose-response assessments to estimate the probability of specific harm to exposed individuals or populations. Furthermore, DPR does not characterize the distribution of risk to the exposed populations.

Recommendations

• Buffer zones should be derived so that they appropriately protect those persons who might spend appreciable amounts of time near treated areas (e.g., residential, schools, offices). These buffer zone distances will need to be based on reasonable worst-case scenarios. Additional field studies should be undertaken to validate these buffer zones.
• At the very least, DPR should characterize occupational and residential exposures with distributions, that is, estimate how many people are likely to be exposed at what levels relative to levels of concern for a given duration of exposure. DPR should also conduct some uncertainty analyses to determine what level of confidence in the exposure values is appropriate given the existing data.
• More neurological testing among those occupationally exposed, particularly at various time intervals after methyl bromide exposures have occurred (instead of during exposures), would enable DPR to look for possible long-term or permanent effects.
• To protect workers and residents from the adverse effects of methyl bromide, DPR must be more explicit about linking its methodology for exposure and MOE analysis to the regulatory levels that are based upon the risk assessment or MOE values. The subcommittee recommends that DPR state at the beginning of its risk characterization document the regulatory goals it hopes to achieve and how its risk characterization will meet them.

In conclusion, the subcommittee recognizes that conducting additional toxicity testing and exposure monitoring is somewhat problematic given the phase-out of methyl bromide in the United States by 2005. Nevertheless, the subcommittee believes that extensive use of this pesticide at this time in California and elsewhere warrants an acknowledgment of existing data gaps that must be addressed to ensure that agricultural workers and residents living near areas where methyl bromide is used are protected against the short-term and long-term health effects of this pesticide. These data gaps might require the combined efforts of regulatory agencies such as DPR and the methyl bromide industry, including manufacturers and pesticide applicators.