|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 9 (March 20), 2006: pp. 1326-1328 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.04.8561
The Ongoing Search for the Sources of the Breast Cancer Survival DisparityDepartment of Health Studies, The University of Chicago, Chicago, IL In the search for explanations for the persistent disparity in outcomes after breast cancer diagnosis between African American and white American women, there has always been something of a divide in the research community. Social and economic differences between the groups are too glaring to ignore, and have significant health consequences across a spectrum of diseases, including cancer. On the other hand, in a disease as heterogeneous as breast cancer, biology is destiny, in ways that we sometimes understand and in others that remain elusive. Thus, when tumor characteristics associated with poor prognosis are noted in greater frequency among black women, an obvious explanation for the disparity is presented. Researchers overly favoring either socioeconomic or biologic explanatory pathways do so at some peril, and consensus has built around the conclusion that both of these factors, and others, must be considered concurrently. However, it has proven difficult to adequately account for the complex constellation of factors that contribute to cancer disparities in any one study. The report by Newman et al in this issue of the Journal of Clinical Oncology attempts to bring the role of socioeconomic status (SES) to the forefront by synthesizing results from studies of black/white disparities that either included socioeconomic measures or were conducted in "equal-access" health care situations.1 The investigation is an update of an earlier investigation that includes additional studies published since 2001.2 The primary conclusion, that accounting for SES measures does not completely explain disparities, seems to be supported by many of the individual studies, as well as the combined estimate from the meta-analysis. Some reservations may be in order, however, based on the methodologic challenges and data limitations inherent in the individual studies summarized. Also, whether biologic explanations therefore gain more credence based on these findings is an inference that must be considered with some caution. Meta-analysis is a valuable research tool, as it forces critical evaluation of the published literature and a quantitative summary that hopes to move beyond seemingly informed opinion. However, even in the case of randomized trials, individual study idiosyncrasies are such that the whole may be less than the sum of its parts, and the state of the art has moved to analysis of patient-level data to avoid some of these difficulties. In the case of disparate types of studies over a broad time span and reliance only on summary effect measures, the analytic task and the interpretation of the results are even more challenging. The authors do apply the appropriate methodology to obtain their summary estimate, which one can either take at face value, or if preferred, choose to examine and weigh the individual studies less formally (Figure 1 of Newman et al).1 Roughly half of the studies show a relative mortality excess below the summary estimate of 27%, and half find a greater excess. Interestingly, another meta-analysis by Bach et al,3 containing some of the studies included here and some others not included, examined black/white differences in mortality among cancer patients in "equal-care" circumstances, usually without additional adjustment for SES. For breast cancer, their overall and cancer-specific mortality hazard ratios were remarkably similar. The majority of the studies in the meta-analysis rely on proxy measures of SES, such as census tract level income. While these measures are the only ones available in many cases and often do contain prognostic information, concerns about misclassification error and other problems4 must lead us to temper our confidence in conclusions regarding the (in)ability of SES measures to explain outcome disparities. Similarly, the limits of prognostic modeling with many covariates must be recognized. That is, we must ask how much observed black/white hazard ratios in the range of 1.25 to 1.50 (the excess mortality in numerous studies that adjust for multiple factors but not SES) can be moved toward the null value, using imperfect SES measures in a heavily parameterized model. The failure of "adjustment" for SES to make the disparity disappear is at least in part a function of limitations in the measure and the method, necessitating a more circumspect interpretation. It is regrettable that in large clinical trials—an excellent data source for separating different components that contribute to disparities in outcomes—we lack even rudimentary patient-level social and economic variables. With recent interest in cancer survivorship, the cooperative groups are considering how best to prospectively collect demographic information that may contribute to prognosis. In most studies of race/ethnicity and breast cancer, overall (all-cause) mortality differs in favor of whites. This is perhaps not surprising, given the notable difference in life expectancy between blacks and whites (approximately 3 years among women surviving to 45 years of age).5 In theory, SES adjustment should reduce this effect, but it may not be adequate to fully account for the health effects of resource disparities throughout life. In the study of Newman et al in this issue of the Journal, breast cancer–specific mortality differences are smaller, with a 19% relative excess overall, suggesting that a portion of the disparity is indeed due to other chronic diseases. It is interesting that one of the studies included, which specifically accounted for chronic disease comorbidities that are more prevalent in blacks, showed essentially the same overall mortality between black and white patients.6 In breast cancer, comorbidity has a significant influence in older patients,7 and in a recent study, was a major explanatory factor in all-cause mortality differences between black and white patients.8 Eleven of the studies did incorporate treatment history variables, a key factor since effective adjuvant therapies have increasingly become available during the time span of the studies in question. While differences in care have been documented over a variety of cancers, a large survey of breast cancer adjuvant therapy use between 1987 and 1995 showed relatively minor differences between black and white patients.9 However, recent research suggests that a more detailed examination may be necessary. Specifically, black/white differences in adverse events, adherence to planned therapy, and regimen completion have been documented.10-12 Adjuvant chemotherapy for breast cancer is sensitive to departures from the intended delivery schedule, so that patients who are apparently treated equally may experience differential benefits, even in controlled settings such as clinical trials. In fact, randomized clinical trials have shown largely similar outcomes between blacks and whites, though there exist small to moderate overall mortality excesses for black patients, which are not statistically inconsistent with those of the Newman study.13,14 More recent studies confirm that disparities may persist in the clinical trial setting in some cases.15,16 These differences tend to be smaller than those noted in population-based comparisons,17 but are nonetheless larger than expected, and require further investigation to determine their source. Technology and knowledge gains in molecular characterization of breast tumors have revitalized efforts to identify particularly aggressive phenotypes. The longstanding observation that certain tumor features, such as lack of estrogen-receptor expression, are more prevalent in black women, has led researchers to ask whether additional unfavorable characteristics are also over-represented in blacks. This might explain differences in prognosis after accounting for well-known clinical and pathologic tumor features and other factors. As Newman et al discuss, recent research does suggest that multiple additional tumor features indicative of unfavorable prognosis are more frequent in tumors of black women.18-20 Furthermore, frequency of the recently defined basal-like tumor subtype21 seems greater in women with African ancestry.22,23 Whether these characteristics will prove explanatory for the roughly 20% excess in breast cancer–specific mortality estimated by Newman et al and reported in clinical trials,16 is yet to be determined. Along the way, we must avoid the creation of tautologies between race/ethnicity and certain breast cancer "types," and always account for the gross confounding influences of stage and other well-known prognostic indicators before attributing differences to these newer markers. As the study by Newman et al aptly illustrates, there are no easy solutions or ready explanations for breast cancer survival disparities. They soundly recommend a multifaceted biologic and sociologic approach to future research. This work critically needs to lead us beyond explanation, to discovery of the intervention points that might close the survival gap. Author's Disclosures of Potential Conflicts of Interest The author indicated no potential conflicts of interest. Author Contributions
REFERENCES
1. Newman LA, Griffith KA, Jatoi I, et al: Meta-analysis of survival in African American and white American patients with breast cancer: Ethnicity versus socioeconomic status. J Clin Oncol 24:1342-1349, 2006 2. Newman LA, Mason J, Cote D, et al: African-American ethnicity, socioeconomic status, and breast cancer survival. Cancer 94:2844-2854, 2002[CrossRef][Medline] 3. Bach PB, Schrag D, Brawley OW, et al: Survival of blacks and whites after a cancer diagnosis. JAMA 287:2106-2113, 2002 4. Geronimus AT, Bound J: Use of census-based aggregate variables to proxy for socioeconomic group: Evidence from national samples. Am J Epidemiol 148:475-486, 1998 5. Hoyert DL, Kung HC, Smith BL: Deaths: Final data for 2003—National vital statistics reports (Volume 53, No. 15). Hyattsville, MD, National Center for Health Statistics, 2005 6. Du W, Simon MS: Racial disparities in treatment and survival of women with stage I-III breast cancer at a large medical center in metropolitan Detroit. Breast Cancer Res Treat 91:243-248, 2005[CrossRef][Medline] 7. Yancik R, Wesley MN, Ries LG, et al: Effect of age and comorbidity in post-menopausal breast cancer patients aged 55 years and older. JAMA 285:885-892, 2001 8. Tammemagi CM, Nerenz D, Nesland-Dudas C, et al: Comorbidity and survival disparities among black and white patients with breast cancer. JAMA 294:1765-1772, 2005 9. Harlan LC, Abrams J: Warren JL, et al: Adjuvant therapy for breast cancer: Practice patterns of community physicians. J Clin Oncol 20:1809-1817, 2002 10. Hershman D, Weinberg M, Rosner Z, et al: Ethnic neutropenia and treatment delay in African-American women undergoing chemotherapy for early-stage breast cancer. J Natl Cancer Inst 95:1545-1548, 2003 11. Hershman H, McBride R, Jacobson JS, et al: Racial disparities in treatment and survival among women with early-stage breast cancer. J Clin Oncol 23:6639-6646, 2005 12. Griggs JJ, Sorbero ME, Stark AT, et al: Racial disparity in the dose and dose intensity of breast cancer adjuvant chemotherapy. Breast Cancer Res Treat 81:21-31, 2003[CrossRef][Medline] 13. Roach M 3rd, Cirrincione C, Budman D, et al: Race and survival from breast cancer: Based on Cancer and Leukemia Group B trial 8541. Cancer J Sci Am 3:107-112, 1997[Medline] 14. Dignam JJ: Efficacy of systemic adjuvant therapy for breast cancer in African-American and Caucasian women. J Natl Cancer Inst Monogr 30:36-43, 2001[Medline] 15. Albain KS, Unger JM, Hutchins LF, et al: Outcomes of African Americans on Southwest Oncology Group (SWOG) breast cancer adjuvant therapy trials. Breast Cancer Res Treat 82:S12, 2003 16. Dignam JJ, Wieand K, Johnson K, et al: Effects of obesity and race on prognosis in lymph node-negative, estrogen receptor-negative breast cancer. Breast Cancer Res Treat 2005 [Epub ahead of print Dec 6] 17. Chu KC, Lamar CA, Freeman HP: Racial disparities in breast carcinoma survival rates: Seperating factors that affect diagnosis from factors that affect treatment. Cancer 97:2853-2860, 2003[CrossRef][Medline] 18. Mehrotra J, Ganpat MM, Kanaan Y, et al: Estrogen receptor/progesterone receptor-negative breast cancers of young African-American women have a higher frequency of methylation of multiple genes than those of Caucasian women. Clin Cancer Res 10:2052-2057, 2004 19. Porter PL, Lund MJ, Lin MG, et al: Racial differences in the expression of cell cycle-regulatory proteins in breast carcinoma. Cancer 100:2533-2542, 2004[CrossRef][Medline] 20. Jones BA, Kasl SV, Howe CL, et al: African American/white differences in breast carcinoma: p53 and other tumor characteristics. Cancer 101:1293-1301, 2004[CrossRef][Medline] 21. Sotiriou C, Neo SY, McShane ML, et al: Breast cancer classification and prognosis based on gene expression profiles from a population-based study. Proc Natl Acad Sci U S A 100:10393-10398, 2003 22. Carey LA, Perou CM, Dressler LG, et al: Race and poor prognosis basal breast tumor (BBT) phenotype in the population-based Carolina Breast Cancer Study (CBCS). J Clin Oncol 22:837s, 2004 (suppl, abstr 9510) 23. Ikpatt OF, Xu J, Kramtsov A, et al: Hormone receptor negative and basal-like subtypes are overrepresented in invasive breast carcinoma from women of African ancestry. Proc Am Assoc Cancer Res 46:2550, 2005
Related Article
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
