|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 18 (June 20), 2006: pp. 2697-2699 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.05.4742
Adjuvant Chemotherapy Use and Outcomes in Older Women With Breast Cancer: What Have We Learned?
Boston University Medical Center, Boston, MA In both the basic science and clinical trial worlds it is not uncommon for two or more research groups to make similar discoveries or the same discovery simultaneously, or for two similarly designed clinical trials to obtain similar or the same results. However, this phenomenon is less common in the clinical epidemiology/outcomes research world. As editorial board member/reviewer (R.A.S.) and associate editor (P.A.G.) of the Journal of Clinical Oncology, we encountered this issue. We felt that the story, its outcomes, and the research and clinical implications would be of interest to the readership. In April 2005, Giordano et al1 submitted an article addressing temporal changes in adjuvant chemotherapy use in older women, including changing patterns of specific agent use. The article also focused on patient factors associated with the receipt of chemotherapy and with putative toxicities of these agents. The analyses presented relied on the Surveillance, Epidemiology, and End Results (SEER)-Medicare data set,2 with the strengths of large numbers of cases across 14 geographic sites and over time, and the weaknesses of reliance on administrative claims data (and therefore lack of clinical data) for comorbidity, chemotherapy use, and toxicities. Although the reviewers and editor were enthusiastic about the topic, the lack of linkage of treatment to outcomes and the limitations of the data sources for toxicity information dampened their enthusiasm. A revised article received in June 2005 addressed the relationship between adjuvant chemotherapy treatment and breast cancer-specific and all-cause mortality and deleted the toxicity analyses. However, several remaining concerns of the reviewers led to the request for additional revisions. While these revisions were underway, Elkin et al3 submitted an article addressing temporal changes in adjuvant chemotherapy use in older women, also focusing on patient factors associated with receipt of chemotherapy, relating receipt of chemotherapy to all-cause mortality. The authors also used propensity scores and sensitivity analysis to address the methodologic challenges of measured and unmeasured confounders in observational studies. The additional strengths associated with these analytic approaches, the fact that the striking similarity between the two articles was a result of the editorial process, and the fact that both articles were simultaneously under review led to the decision to publish them both. The tangible outcomes of this process are found elsewhere in the Journal of Clinical Oncology.1,3 To complicate things additionally, while both research groups were finalizing their articles, the Journal of Gerontology: Medical Sciences published a similar paper by Du et al,4 again based on analyses of the SEER-Medicare data set. For comparative purposes, the key features of all three articles are displayed in Table 1. As might be expected from experienced investigators using the same data, all three groups observed increasing use of adjuvant chemotherapy over time; substantial geographic variation in use (from a low in San Francisco to a high in Hawaii); decreasing use in patients with increasing age and comorbidity; and increasing use in patients with poor prognostic indicators. The varying association of race across studies may well be related to different strategies for socioeconomic status measurement which are known to not be comparable,5 and differences in tumor characteristics (eg, stage and receptor status) selected for sampling across the three studies.
It is the differences in tumor characteristics coupled with differences in outcome that are most likely responsible for the apparent lack of consistency of findings across the three studies. Giordano et al1 were least restrictive in their initial sampling, including all women with stage I-III disease, both estrogen receptor (ER) –positive and –negative, who received either breast-conserving surgery or mastectomy. However, in the end, after stratification by ER and lymph-node status, they identified lymph node–positive, ER–negative patients as the subgroup benefiting from chemotherapy (hazard ratio, 0.72; 95% CI, 0.54 to 0.96). In contrast, Elkin et al restricted their sample to all stage I-III women with ER–negative tumors. Their initial analyses considered both lymph node-positive and lymph node-negative patients, yielding a hazard ratio of 0.85 (95% CI, 0.77 to 0.95). In response to the Giordano paper, analyses restricted to the lymph node-positive group, but with the additional propensity score adjustment, yielded results very similar to those of Giordano: hazard ratio, 0.76; 95% CI, 0.65 to 0.88; suggesting that most of the overall observed effect was restricted to this subgroup. Both sets of investigators observed that the benefits of adjuvant chemotherapy did not vary by age, and specifically were similar for those women younger than 70 years of age versus those 70 years of age or older. This latter finding is of particular importance because the overview of trials of adjuvant chemotherapy only included 1,044 women (approximately 1%)6 and the four clinical trials summarized by Muss et al only included 159 (2%)7 in this age group. However, the observations of Giordano et al1 and Elkin et al3 are in conflict with Du et al’s conclusions of no protective effect of adjuvant chemotherapy in those 70 years of age or older.4 One possible explanation for the discrepancy is that Du included both ER–negative and –positive patients. Indeed, 58% of those who were included in his analyses were ER positive. In addition, he restricted his analyses to stage II and IIIA patients. A final consideration relates to the end points analyzed. While Elkin and Du considered only all-cause mortality, Giordano also used breast cancer-specific mortality as an end point, although the details of time frame and data sources were not provided. In their analyses, the impacts on breast cancer-specific and all-cause mortality were quite similar. This could be because breast cancer deaths represented the majority of deaths (which is unlikely8) or because patients receiving adjuvant chemotherapy truly are different in unmeasured ways from those who do not. Candidate ways in which those receiving chemotherapy may differ from those who do not include having lower comorbid disease burden and severity, better functional status, and better health behaviors. In addition, they may also receive higher quality nonbreast cancer care from their physicians. Understanding the potential for these selection biases is particularly critical, so as not to misattribute the etiology of improved survival. What are the key messages from these studies? First, although the SEER-Medicare data set is a wonderful resource for addressing many important questions, limitations of the data elements and their sources necessarily mean that duplicative analyses are likely to become more commonplace. Integrated data sets that include more detailed clinical information are critically needed. Secondly, even in the absence of a robust body of scientific evidence, the use of chemotherapy in older women with breast cancer is increasing. The sparse evidence from the Giordano and Elkin studies and the clinical trial synthesis6 suggests that this increase may be warranted in light of improved survival and the lack of age-related differences in benefit. However, considerable uncertainty remains regarding the subgroups of older women most likely to benefit, based on tumor characteristics as well as on comorbidity burden, functional status and reserve, and future life expectancy. In this regard, new clinical trial designs that allow for enrollment of more heterogeneous samples of older adults, as well as observational studies of treatment effectiveness, are needed if we are to meet the challenge of providing evidence-based high quality care for the growing numbers of older adults with cancer. Authors’ Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest. Author Contributions
REFERENCES
1. Giordano SH, Duan Z, Kuo Y-F, et al: Use and outcomes of adjuvant chemotherapy in older women with breast cancer. J Clin Oncol 24:2750-2756, 2006 2. Warren JL, Klabunde CN, Schrag D, et al: Overview of the SEER-Medicare data: Content, research applications, and generalizability to the United States Elderly population. Med Care 40:IV3-IV18, 2002 (suppl) 3. Elkin EB, Hurria A, Mitra N, et al: Adjuvant chemotherapy and survival in older women with hormone receptor-negative breast cancer: Assessing outcome in a population-based, observational cohort. J Clin Oncol 24:2757-2764, 2006 4. Du XL, Jones DV, Zhang D: Effectiveness of adjuvant chemotherapy for node-positive operable breast cancer in older women. J Gerontol: Med Sci 60A:1137-1144, 2005 5. Braveman PA, Cubbin C, Egerter S, et al: Socioeconomic status in health research: One size does not fit all. JAMA 294:2879-2888, 2005 6. Early Breast Cancer Trialists’ Collaborative Group. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: An overview of the randomized trials. Lancet 365:1687-1717, 2005[CrossRef][Medline] 7. Muss HB, Woolf S, Berry D, et al: Adjuvant chemotherapy in older and younger women with lymph node-positive breast cancer. JAMA 293:1073-1081, 2005 8. Yancik R, Wesley MN, Ries LAG, et al: Effect of age and comorbidity in postmenopausal breast cancer patients aged 55 years and older. JAMA 285:885-892, 2001
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
