Adjuvant endocrine therapy with tamoxifen in young women with breast cancer: determinants of interruptions vary over time.

BACKGROUND: In premenopausal women with hormone receptor-positive breast cancer (BC), 5 years of tamoxifen is recommended. Little is known about reasons for interruption in this population. The aim was to estimate the incidence of tamoxifen interruption and its correlates among younger women. PATIENTS AND METHODS: Using a prospective cohort Elippse 40 of women with BC aged ≤ 40 diagnosed between 2005 and 2008, we studied 196 women. Tamoxifen interruption was defined as two or more consecutive months without dispensed prescription of tamoxifen, based on pharmacy refill database. Two periods were studied: between tamoxifen initiation and 16 months after BC diagnosis, and between 16 and 28 months. RESULTS: Among women treated with tamoxifen, 42% interrupted within the first 2 years of treatment. During the first period, treatment interruptions were associated with a lack of understandable information about endocrine treatment and insufficient social support. During the second period, another set of factors were associated with interruption: treatment side-effects, no longer fearing cancer relapse, lack of social support, no opportunity to ask questions at the time of diagnosis, and fewer treatment modalities. CONCLUSIONS: Improving information and patient-provider relationship might prevent interruption. Particular attention should be paid to women with little social support.

Analysis of the effect of age on the prognosis of breast cancer.

To explore the effect of age at diagnosis on relative survival from breast cancer at different cancer stages and grades, using appropriate statistical modeling of time-varying and non-linear effects of that prognostic covariate. Data on 4,791 female invasive breast cancers diagnosed between 1990 and 1997 were obtained from a French cancer registry. The effect of age on relative survival was studied using an approach based on excess rate modeling. Different models testing non-linear and non-proportional effects of age were explored for each grade and each stage. In the whole population, the effect of age was not linear and varied with the time elapsed since diagnosis. When analyzing the different sub-groups according to grade and stage, age did not have a significant effect on relative survival in grade 1 or stage 3 tumors. In grade 2 and stage 4 tumors, the excess mortality rate increased with age, in a linear way. In grade 3 tumors, age was a time-dependent factor: older women had higher excess rates than younger ones during the first year after diagnosis whereas the inverse phenomenon was observed 5 years after diagnosis. Our findings suggest that when taking into account grade and stage, the time-varying impact of young age at diagnosis is limited to grade 3 tumors, without evidence of worst prognosis at 5 years for the youngest women.