Effectiveness of Adjuvant Chemotherapy for Node-Positive Operable Breast Cancer in Older Women.

XLL Du,DV Jones,D Zhang
DOI: https://doi.org/10.1093/gerona/60.9.1137
2005-01-01
The Journals of Gerontology Series A
Abstract:BACKGROUND Randomized clinical trials have shown the efficacy of adjuvant chemotherapy in treating node-positive operable breast cancer in women aged < or = 69 years, but the benefit of chemotherapy in women aged > or = 70 is questionable. This study was to examine if adjuvant chemotherapy is effective for these women with breast cancer. METHODS We studied a cohort of 5464 women diagnosed with node-positive operable breast cancer at age > or = 65 in 1992 through 1996 with last follow-up of December 31, 1999 in five states and six metropolitan areas. Hazard ratio (HR) for all-cause mortality was used for survival analysis with adjustment for patient and tumor characteristics; propensity analysis was used to control for observed factors; and sensitivity analysis was used to estimate potential effects of unmeasured confounders. RESULTS After adjusting for propensity to receive chemotherapy, the chemotherapy-treated and untreated groups were not statistically significantly different for covariates except for age and hormone receptor status. Mortality was significantly reduced in women aged 65-69 who received adjuvant chemotherapy compared to those who did not, after adjusting for patient and tumor characteristics (HR = 0.70, 95% confidence interval [CI], 0.57-0.88) or after adjusting for propensity scores (HR = 0.76, 95% CI, 0.62-0.94). HR did not significantly differ between the treated and untreated women aged > or = 70 (HR = 0.96, 95% CI = 0.83-1.09, and HR = 0.99, 95% CI, 0.87-1.14). These results were relatively insensitive to changes in unmeasured confounders. CONCLUSIONS Adjuvant chemotherapy is associated with improved survival in women with node-positive operable breast cancer aged 65-69 living in the community, but not in women aged > or = 70. These findings are consistent with those found in randomized controlled trials.
What problem does this paper attempt to address?