Revising the National Ambient Air Quality Standards: The integration of science and policy

Under the Clean Air Act, Congress granted discretionary decision making authority to the Administrator of the Environmental Protection Agency (EPA). This discretionary authority involves setting standards to protect the public's health with an "adequate margin of safety" based on current scientific knowledge. The Administrator of the EPA is usually not a scientist, and for the National Ambient Air Quality Standard (NAAQS) for particulate matter (PM), the Administrator faced the task of revising a standard when several scientific factors were ambiguous. These factors included: (1) no identifiable threshold below which health effects are not manifested, (2) no biological basis to explain the reported associations between particulate matter and adverse health effects, and (3) no consensus among the members of the Clean Air Scientific Advisory Committee (CASAC) as to what an appropriate PM indicator, averaging period, or value would be for the revised standard. ^ This project recommends and demonstrates a tool, integrated assessment (IA), to aid the Administrator in making a public health policy decision in the face of ambiguous scientific factors. IA is an interdisciplinary approach to decision making that has been used to deal with complex issues involving many uncertainties, particularly climate change analyses. Two IA approaches are presented; a rough set analysis by which the expertise of CASAC members can be better utilized, and a flag model for incorporating the views of stakeholders into the standard setting process. ^ The rough set analysis can describe minimal and maximal conditions about the current science pertaining to PM and health effects. Similarly, a flag model can evaluate agreement or lack of agreement by various stakeholder groups to the proposed standard in the PM review process. ^ The use of these IA tools will enable the Administrator to (1) complete the NAAQS review in a manner that is in closer compliance with the Clean Air Act, (2) expand the input from CASAC, (3) take into consideration the views of the stakeholders, and (4) retain discretionary decision making authority. ^

Relationships between the alterations of tumor suppressor genes, risk factors and clinical outcomes of gliomas

The relationship between MMAC/PTEN, DMBT1 and the progression and prognosis of glioma, and the association between the alterations of MMAC/PTEN, p53, p16, and Rb and some cancer risk factors, such as smoking, exposure to radiation, family cancer history, and previous cancer history, were assessed in 4 studies. ^ By allelic deletion analysis, MMAC/PTEN locus was shown to be frequently lost in glioblastomas multiforme (GM) but maintained in most lower-grade astrocytic tumors. DMBT1 locus, however, was frequently lost in all grades of gliomas examined. The potential biological significance of these two regions was frontier assessed by examining microcell-hybrids that contained various fragments of 10q. Somatic cell hybrid clones that retained the MMAC/PTEN locus have less transformed phenotypes, exhibiting an inability to grow in soft agarose. On the other hand, the presence or absence of DAMT1 did not correlate with any in vitro phenotype assessed in our model system. Further, Cox proportional hazards regression analysis, adjusted for age at surgery and histologic grades (GM, and non-GM), showed that without LOH at the MMAC/PTEN locus had a significantly better prognosis than did patients with LOH at MMAC/ PTEN (hazard ratio = 0.5; 95% Cl = 0.28–0.89; P = 0.018). Furthermore, status of LOH at MMAC/PTEN was found to be significantly associated with age, while that for DMBT1 was not. These results suggest that the DMBT1 may be involved early in the oncogenesis of gliomas, while alterations in the MMAC /PTEN may be a late event in the oncogenesis related with progression of gliomas and provide a significant prognostic marker for patient survival. ^ The associations between 4 cancer risk factors and 4 tumor suppressor genes were assessed. The expression of p16 was observed to be associated with current smoking (adjusted OR = 1.9, 95% CI = 1.02–3.6) but not the former smoking (adjusted OR = 1.1, 95% Cl = 0.5–3.5). The expression of p53 was found to be associated with the family cancer history (OR = 3.5, 95% Cl = 1.07–11 for patients with first-degree family history of cancer). MMAC/ PTEN was associated with the histologic grade (OR = 2.8, 95% CI = 1.2–6.6) and age (P = 0.035). Also, the OR for LOH around MMAC/PTEN in patients with a family history of cancer was elevated (OR = 1.9, 95% CI = 0.8–4.6 for patients with first-degree family history of cancer). The associations between exposure and the alterations of tumor suppressor genes, between smoking and p16, between family history of cancer and p53 and MMAC/PTEN, provide suggestive evidences that those exposures are related to the development of gliomas. ^