Registration Dossier

Administrative data

Endpoint:
epidemiological data
Type of information:
experimental study
Adequacy of study:
supporting study
Reliability:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
study well documented, meets generally accepted scientific principles, acceptable for assessment

Data source

Referenceopen allclose all

Reference Type:
publication
Title:
Hazardous air pollutants and breast cancer risk in California teachers: a cohort study
Author:
Garcia et al
Year:
2015
Bibliographic source:
Environmental Health 14:14 (2015)
Reference Type:
other: Personal communication
Title:
E. Delzell COMMENTS ON Garcia et al. Environmental Health, 2015; 14:14. Hazardous air pollutants and breast cancer risk in California teachers: a cohort study
Author:
Delzell, E
Year:
2018

Materials and methods

Study type:
cohort study (prospective)
Remarks:
California Teacher Study (CTS)
Endpoint addressed:
carcinogenicity
Test guideline
Qualifier:
no guideline followed
Principles of method if other than guideline:
Prospective cohort study of estimated inhalation exposure to 24 substances considered to be potential mammary gland carcinogens (MGC) with incidences of breast cancer in California teachers

See method details below
GLP compliance:
no

Test material

Reference
Name:
Unnamed
Specific details on test material used for the study:
No details on test substance; atmospheric concentrations of propylene oxide modelled for study purposes

Method

Type of population:
occupational
Ethical approval:
other: reviewed by the Human Subjects Research Committees of five Agenices and found to be in compliance with their standards and US Code of Federal Regulations, Title 45, Part 46 on the Protection of Human Subjects.
Details on study design:
The CTS is an ongoing prospective cohort started in 1995/6 by returned basline questionnaires from current and retired female teachers and administrators enrolled in the California State Teachers Retirement System.
Excluded from the cohort of 133,479 for the purposes of this study were women who:
- were non resident in California (8,867)
- had an unknown prior cancer history (139)
- had a prior history of invasive or in situ breast cancer (6,212)
- asked to be removed (1)
- had an address that could not be geocoded (5,882)
Resulting in a cohort of 112,378 women which is followed annually for information on cancer diagnosis (provided by linkage between the California Cancer Registry and cohort membership), death and change of address.
The study defined a case as any woman diagnosed with invasive breast cancer (ICD-03 site codes C500-C509 with histology codes 9050-9055, 9140 and 9590-9992 excluded) between the dates of questionnaire completion and December 31st 2011.
Exposure assessment:
estimated
Details on exposure:
The National-Scale Air Toxics Assessment (NATA) produced every 3 years by the United States Environmental Protection Agency (EPA) is designed to identify and prioritise up to 180 hazardous air pollutants (HAPs) with respect to potential health risks. Dispersion models incorporating emissions data from the National Emissions Inventory and meteorological data are used in the Assessment System for Population Exposure Nationwide (for area, on-road and non-road emission sources) and the Human Exposure Model (for major emission sources) to estimate annual ambient air concentrations of HAPs at a geographic census tract level.

Cohort members census tracts were determined by geocoding their address at the time of baseline questionnaire completion and they were assigned the corresponding modelled annual average ambient HAP concentration from NATA to create a proxy of inhalation exposure concentration.

Modelled concentration estimates were compared with available monitored data for 12 HAPs, with years 2002 and 2005 showing the best agreement between modelled and monitored data. The 2002 NATA estimates chosen for use in the study represented an approximate mid-point in the follow-up period to the baseline questionnaire (1995-2011) and were generally reflective of the 1996, 1999 and 2005 estimates (median Spearman correlation coefficients 0.69, 0.76 and 0.88 respectively). Using a single year estimate over combining multiple years of estimates was preferred due to concern that inconsistent methodical approaches and temporal variations in the level of agreement could introduce exposure misclassification.

Exposure estimated for 24 substances thought to be mammary gland carcinogens (including propylene oxide) were used in this study.
Statistical methods:
Spearman correlation
Cox proportional and regression models
False Discovery Rates

Results and discussion

Results:
The cohort experienced 5,676 incident invasive breast cancers during follow-up, 76.7% of which were estrogen or progesterone receptor-positive. Cohort members with a breast cancer diagnosis tended to be older, more likely to have a family history of breast cancer and more likely to be a non-mover (geographic).

Analyses of all forms of breast cancer combined, without adjustment for multiple personal risk factors for breast cancer or for multiple comparisons, found evidence of increased risks (HRs for some quintiles elevated at p < 0.05) for several substances including propylene oxide (other substances were: acrylamide, carbon tetrachloride, chloroprene, 4,4’-methyl bis(2-chloro aniline) and vinyl chloride).

After adjustment for multiple comparisons, only results for propylene oxide (and vinyl chloride) remained statistically significant. However, results for these (and other) compounds were not statistically significant after adjustment for multiple potential confounders and multiple comparisons.

For propylene oxide, the largest HR (1.11; 95% confidence interval, 1.02-1.20) occurred in the third quintile of exposure concentration and remained statistically significant after adjustment for multiple comparisons but not after adjustment for multiple confounders; and the p-value for trend was 0.18 (not statistically significant).
Confounding factors:
Some attempts were made to include potential confounding factors in the analyses (see 'any other information on materials and methods' above) however, it is not clear that these confounding factors were adequately applied to all analyses (see 'strengths and weaknesses' below).
Strengths and weaknesses:
Strengths
- large, prospective cohort study with high-quality data on personal risk factors for breast cancer and on incident invasive breast cancer cases occurring during follow-up
- use of census tract level data on ambient HAPs.
- ability to adjust for multiple potential confounders.

Limitations
These stem mainly from the approaches used to estimate exposure to MGC HAPs and to control for confounding and from the presentation of results.
- potential for exposure misclassification: assumed that 2002 air pollution data were representative of the entire follow-up time
- exposure estimates assume that the available data provided good estimates of inhalation exposure
- exposure from indoor air pollutants or exposure routes other than inhalation
- exposure prior to baseline questionnaire not considered; cancers can take years to develop and risk cannot be assumed to start at cohort initiation

- confounding factors; although some of the presented analyses attempted to control for personal risk factors for breast cancer, not all analyses did so
- confounding; analyses of a particular HAP that adjusted for exposure to other HAPs apparently were not done.

- results; no results of analyses that adjusted for personal risk factors for breast cancer are presented
- results; results for subtypes of breast cancer and for cohort subgroups do not appear to have been adjusted for multiple breast cancer risk factors.

Applicant's summary and conclusion

Conclusions:
This study provides, at most, weak evidence of an association between exposure to outdoor MGC HAPs, including propylene oxide, and invasive breast cancer. HRs by quintile of exposure concentration were not markedly elevated for any compound (most “elevated” HRs were in the range of 1.1 to 1.2 and none was above 1.5); trends of increasing HR with increasing quintile of exposure were, for the most part, irregular; quintile-specific HRs and trend tests were not statistically significant after adjustment for multiple comparisons and for multiple confounders; subgroup analyses were not adjusted for confounding by personal breast cancer risk factors; and no analyses were adjusted for multiple HAP exposures. Moreover, there is little external support for the possible associations observed in the study. Thus, chance, exposure misclassification and confounding cannot be ruled out as alternative explanations of the weak associations reported in the study.