Impact of body composition on toxicity and pathological complete response in locally advanced breast cancer (BC).

Abstract

593 Background: Obesity, defined as body mass index (BMI) > 30 kg/m2, has been associated with inferior outcomes in localized breast cancer (BC), including lower pathological complete response (pCR) to neoadjuvant chemotherapy (NAC). High BMI is usually associated with excess adipose tissue (AT), but also reflects skeletal muscle (SM) mass. Low SM mass (sarcopenia) has also been associated with inferior outcomes and more chemotherapy-associated toxicity. We aimed to evaluate the association of BMI, AT and SM tissue with pCR and toxicity after NAC for stage II-III breast cancer (BC). Methods: 191 patients with stage II-III BC received NAC, had information regarding baseline BMI, toxicity and pCR to NAC at surgery, and, had abdominal computerized tomography (CT) prior to NAC. Total AT (TAT), visceral AT (VAT), subcutaneous AT (SAT), SM area (SMA) and SM density (SMD) were measured by CT at the L3 level using the TOMOVISION software. SM index (SMI) was calculated (SMA/height) to assess for sarcopenia (SMI < 40). Association linking BMI, SAT, VAT, SMA and SMD to pCR and severe toxicity (ST ≥ grade 3) was evaluated using logistic regression models. Results: Patients were predominantly black (51%) with a median age of 54 years (interquartile range = 45-63). pCR was achieved in 31% (60/191) of patients. Of those, 47% (n = 28/60), 40% (n = 24/60) and 13% (n = 8/60) corresponded to HER2(+), triple negative, and hormone receptor-HR(+)/HER2(-) tumors, respectively. ST occurred in 38% (n = 73/191) of patients. Obesity and sarcopenia were present in 52% (n = 100/191) and 14% (n = 27/191) of patients, respectively. There was a statistically significant association between VAT and pCR (median VAT 95.3 cm2 vs. 121.8 cm2 in the pCR vs. no-pCR groups, respectively, p = 0.03). This association remained after adjusting for age, race, tumor grade, stage, BMI, SMD, HR and HER2 status (p = 0.04). There was a statistically significant association between SMA and ST (mean SMA 123 cm2 vs. 130 cm2 in the ST vs. no-ST groups, respectively, p = 0.03). This association disappeared after adjusting for age, race, BMI, VAT, SAT, and SMD (p = 0.21). Conclusions: This study provides evidence that in patients with stage II-III BC receiving NAC, excess VAT is associated with significantly lower pCR rates, and low SMA is associated with ST. Additional research is needed to elucidate the pathophysiologic mechanisms contributing to these associations.