Optimal duration of adjuvant tamoxifen treatment in elderly breast cancer patients: Influence of age, comorbidities, and various effectiveness hypotheses on life-expectancy and cost

The aim of this study was to explore the effects of the duration of adjuvant tamoxifen on life-expectancy of elderly patients with breast cancer and to study the cost-effectiveness of prolonging tamoxifen treatment. Our data sources were the medical literature published prior to May 1, 1995, and the H. Lee Moffitt Cancer Center's tumor registry. Decision analysis models were created with three arms: tamoxifen for 2 years (tam2), 5 years (tam5), or lifetime (tamL), each simulated as a Markov process. Patients 65, 75, and 85 years old were simulated. The basic model presumed constant effect, lasting 2 years after cessation, and no other protective effect. Alternative models included influence of tamoxifen on nonbreast cancer mortality, decreasing effect with time and no residual effect. In the basic model, the gain in life-expectancy from tamL versus tam2 was 0.79 years at a marginal cost-effectiveness (C/E) of $8,800/year gained at 65 years, 0.42 years and $9,600 at 75 years, and 0.12 year and $16,200 at 85 years. Gains from tamL versus tam5 were 0.54 year, $8,500; 0.26 year, $9,300; and 0.06 year, $16,000, respectively. Noncancer mortality strongly influenced effectiveness and C/E. Its reduction by tamoxifen had a stronger impact on life-expectancy than similar reductions in relapse, especially in the oldest patients. Lifelong tamoxifen may lead to a substantial gain in life-expectancy. In the oldest old, a nononcologic effect of tamoxifen might be the major cause of this gain. In the elderly, any study of malignancies with cancer-specific survival greater than 2 years should analyze noncancer mortality and comorbidity.

Duke Scholars

Author Gary Herbert Lyman Medicine, Medical Oncology