Individual benefits from adjuvant anthracycline-based chemotherapy can be predicted from a predictive nomogram created from primary chemotherapy (PC) data: Results of a randomized trial

Publisher Website
Google Scholar
Abstract
623 Background: The benefit provided by chemotherapy (CT) for breast cancer patients is based on the risk of recurrence at a population level and not on an individual benefit. Pathologic complete response (pCR) to primary chemotherapy (PC) has been demonstrtated to be a surrogate marker of survival, demonstrtaing that patients that achieve a pCR are the most likely to benefit from CT. We hypothesized that individuals who benefit from adjuvant CT share the same clinico-pathologic characteristics as patients with pCR after PC and therefore that the same nomogram can be used to evaluate the probability of pCR and the benefit from receiving adjuvant CT after initial surgery. Methods: Two cohorts of patients treated at the IGR were studied. The first cohort (n=502) has received anthracycline-based PC. The second cohort (n=788) has been treated with initial surgery and has been randomized to receive either an adjuvant anthracycline-based CT or no CT. This trial demonstrated a disease-free survival benefit in the CT arm. We developed a nomogram to determine the probability of pCR in first cohort. For each patient of the second cohort, we calculated the theoretical probability of pCR. These patients were then categorized according to the probability of achieving pCR. Survivals were compared within each category according to the treatment arm. Results: A nomogram, based on age, tumor size, histologic grade and estrogen receptor status provided a good estimation of the probability of pCR (area under the curve=0.72). Achieving a pCR was associated with a 15 to 25% survival benefit. When applied to the cohort of patients treated with initial surgery, 49, 372 and 366 patients theoretically had less than 5%, between 5 and 10% and more than 10% chance to achieve pCR, respectively. The absolut survival benefits of adjuvant CT at 10 years were 0, 4 and 7%, respectively for the three categories (p: 0.05, linear trend). Conclusions: Probability of pCR after PC and individual benefit from adjuvant anthracycline-based CT are correlated. Our study suggests that indication of adjuvant CT could be based on individual benefit and not on a population risk. No significant financial relationships to disclose.

This publication has 0 references indexed in Scilit: