Accounting for Multiple Comparisons in Statistical Analysis of the Extensive Bioassay Data on Glyphosate.

Glyphosate is a widely used herbicide worldwide. In 2015, the International Agency for Research on Cancer (IARC) reviewed glyphosate cancer bioassays and human studies and declared that the evidence for carcinogenicity of glyphosate is sufficient in experimental animals. We analyzed 10 glyphosate rodent bioassays, including those in which IARC found evidence of carcinogenicity, using a multiresponse permutation procedure that adjusts for the large number of tumors eligible for statistical testing and provides valid false-positive probabilities. The test statistics for these permutation tests are functions of p values from a standard test for dose-response trend applied to each specific type of tumor. We evaluated 3 permutation tests, using as test statistics the smallest p value from a standard statistical test for dose-response trend and the number of such tests for which the p value is less than or equal to .05 or .01. The false-positive probabilities obtained from 2 implementations of these 3 permutation tests are: smallest p value: .26, .17; p values ≤ .05: .08, .12; and p values ≤ .01: .06, .08. In addition, we found more evidence for negative dose-response trends than positive. Thus, we found no strong evidence that glyphosate is an animal carcinogen. The main cause for the discrepancy between IARC's finding and ours appears to be that IARC did not account for the large number of tumor responses analyzed and the increased likelihood that several of these would show statistical significance simply by chance. This work provides a more comprehensive analysis of the animal carcinogenicity data for this important herbicide than previously available.

Ratcheting up cancer potency estimates.

The current paradigm for cancer risk assessment in the United States (U.S.) typically requires selection of representative rodent bioassay dose-response data for extrapolation to a single cancer potency estimate for humans. In the absence of extensive further information, the chosen bioassay result generally is taken to be that which gives the highest extrapolated result from the "most sensitive" species or strain. The estimated human cancer potency is thus derived from an upper-bound value on animal cancer potency that is technically similar to an extreme value statistic. Thus additional information from further bioassays can only lead to equal or larger cancer potency estimates. We here calculate the size of this effect using the collected results of a large number of bioassays. Since many standards are predicated on the value of the cancer potency, this effect is undesirable in producing a strong counter-incentive to performing further bioassays.

Application of quantitative microbial risk assessments for estimation of risk management metrics: Clostridium perfringens in ready-to-eat and partially cooked meat and poultry products as an example.

The U.S. Department of Agriculture, Food Safety and Inspection Service is exploring quantitative risk assessment methodologies to incorporate the use of the Codex Alimentarius' newly adopted risk management metrics (e.g., food safety objectives and performance objectives). It is suggested that use of these metrics would more closely tie the results of quantitative microbial risk assessments (QMRAs) to public health outcomes. By estimating the food safety objective (the maximum frequency and/or concentration of a hazard in a food at the time of consumption) and the performance objective (the maximum frequency and/or concentration of a hazard in a food at a specified step in the food chain before the time of consumption), risk managers will have a better understanding of the appropriate level of protection (ALOP) from microbial hazards for public health protection. We here demonstrate a general methodology that allows identification of an ALOP and evaluation of corresponding metrics at appropriate points in the food chain. It requires a two-dimensional probabilistic risk assessment, the example used being the Monte Carlo QMRA for Clostridium perfringens in ready-to eat and partially cooked meat and poultry products, with minor modifications to evaluate and abstract required measures. For demonstration purposes, the QMRA model was applied specifically to hot dogs produced and consumed in the United States. Evaluation of the cumulative uncertainty distribution for illness rate allows a specification of an ALOP that, with defined confidence, corresponds to current industry practices.

Risk assessment for Clostridium perfringens in ready-to-eat and partially cooked meat and poultry products.

An assessment of the risk of illness associated with Clostridium perfringens in ready-to-eat and partially cooked meat and poultry products was completed to estimate the effect on the annual frequency of illnesses of changing the allowed maximal 1-log growth of C. perfringens during stabilization (cooling after the manufacturing heat step). The exposure assessment modeled stabilization, storage, and consumer preparation such as reheating and hot-holding. The model predicted that assuming a 10- or 100-fold increase from the assumed 1-log (maximal allowable) growth of C. perfringens results in a 1.2- or 1.6-fold increase of C. perfringens-caused illnesses, respectively, at the median of the uncertainty distribution. Improper retail and consumer refrigeration accounted for approximately 90% of the 79,000 C. perfringens illnesses predicted by the model at 1-log growth during stabilization. Improper hot-holding accounted for 8% of predicted illnesses, although model limitations imply that this is an underestimate. Stabilization accounted for less than 1% of illnesses. Efforts to reduce illnesses from C. perfringens in ready-to-eat and partially cooked meat and poultry products should focus on retail and consumer storage and preparation methods.

A Weight-of-evidence analysis of the cancer dose-response characteristics of 2,3,7,8-tetrachlorodibenzodioxin (TCDD).

Cancer risk assessment for TCDD and other compounds must focus on the cancer dose-response relationship and corresponding potency for the range of human doses before it can have relevance to the human exposure environment. Major differences of opinion exist over whether the dose-response curve for TCDD and other dioxin congeners is non-linear (incorporating a threshold dose region below which tumors are unlikely to be elicited) or linear (implying that any exposure has a statistical likelihood of causing cancer). The World Health Organization and others strongly support a non-linear dose-response relationship for TCDD and cancer, whereas USEPA characterizes the dose-response function as linear. This review critically summarizes the available information on TCDD dose-response relationship for cancer utilizing a weight-of-evidence approach. This assessment concludes that the available data support a non-linear dose-response relationship as being most likely and appropriate for human cancer risk assessment, i.e., the evidence suggests that a biological threshold exists in the dose-response. While proof of a threshold is not absolute, and never can be, the level of certainty for TCDD is substantial because of the concordance of many lines of evidence and the consistency of repeated observations pointing to non-linearity.

Uncertainty in octanol-water partition coefficient: implications for risk assessment and remedial costs.

The assessment of human health and ecological risks at chemically contaminated sites often includesthe use of models to assess chemical transport, fate, and exposure/toxicity. These models require input data on a variety of physical and chemical properties for each compound of concern. Small changes in some of these parameters may result in significant differences in estimated human health or ecological risks and in the extent of required remediation efforts. The octanol-water partition coefficient (Kow) for hydrophobic organic compounds is one such parameter, particularly because it is often used to estimate additional partitioning and bioaccumulation parameters. Unfortunately, there is considerable variability among tabulated Kow values for many compounds of concern. This paper assesses the implications of using various values of Kow to calculate health-protective polychlorinated biphenyl (PCB) sediment quality objectives (SQOs) in a case study using a simplified food chain model and the range of Kow values available from or recommended by the U. S. EPA. For the site and Kow values considered in this study, which are a snapshot of values available in the spring of 2004, the SQOs differ by as much as a factor of 5. This range of SQOs is estimated to correspond to a difference in remediation costs of $48 million.

What's wrong with the National Ambient Air Quality Standard (NAAQS) for fine particulate matter (PM(2.5))?

Associations between airborne concentrations of fine particulate matter (PM(2.5)) and mortality rates have been investigated primarily by ecologic or semiecologic epidemiology studies. Many investigators and regulatory agencies have inferred that the weak, positive association often observed is causal, that it applies to all forms of airborne PM(2.5), and that current ambient levels of PM(2.5) require reduction. Before implementing stringent regulations of ambient PM(2.5), analysts should pause to consider whether the accumulated evidence is sufficient, and sufficiently detailed, to support the PM(2.5) National Ambient Air Quality Standard. We take two tacks. First, we analyze the toxicologic evidence, finding it inconsistent with the notion that current ambient concentrations of all forms of fine particulate matter should affect pulmonary, cardiac, or all-cause mortality rates. More generally, we note that the thousands of forms of PM(2.5) are remarkably diverse, yet the PM(2.5) NAAQS presumes them to be identical toxicologically, and presumes that reducing ambient concentrations of any form of PM(2.5) will improve public health. Second, we examine the epidemiologic evidence in light of two related examples of semiecologic associations, examples that both inform the PM-mortality association and have been called into question by individual-level data. Taken together, the toxicologic evidence and lessons learned from analogous epidemiologic associations should encourage further investigation of the association between particulate matter and mortality rates before additional regulation is implemented, and certainly before the association is characterized as causal and applicable to all PM(2.5).