Rahul Tendulkar, MD from the Cleveland Clinic explains the ASTRO 2020 abstract Impact of Post-Mastectomy Radiation Therapy (PMRT) in Women with T1-2N1M0 Breast Cancer: A Multi-Institutional Analysis.
Body
Purpose/Objective(s): To explore the impact of PMRT on women in the modern age with pT1-2N1M0 breast cancer.
Materials / Methods: In order to include women who underwent mastectomy without neoadjuvant therapy with pT1-2N1M0 disease from 1995-2015, data from 5 North American institutions were pooled. Patients were divided into cohorts for discovery and confirmation. Using several imputations with chain equations, missing covariate values were imputed and combined using Rubin’s law. Regression of competing risks was conducted to classify factors associated with locoregional recurrence (LR), remote metastases (DM), and mortality from breast cancer (BCM).
Results: 639 (12%) of 5197 patients in the discovery cohort received sentinel node biopsy alone, 2848 (55%) received PMRT, and 3641 (70%) received adjuvant chemotherapy (CHT). Median 11 (LN) lymph nodes have been dissected. The median follow-up period was 9.8 years. Larger tumors, more positive LN (LN+), lymphovascular invasion (LVI), extracapsular extension (ECE), HER2 + disease, positive margins, and more frequent CHT (all p<0.001) were substantially younger in patients who received PMRT. The 10-year unadjusted combined incidence rates of LR, DM, and BCM were 2.8 percent, 17 percent and 14 percent with PMRT, and 7 percent, 17 percent, and 14 percent without PMRT, respectively, despite more negative features. Increased LR was correlated with younger age (HR 1.04, p<0.001), greater tumor size (HR 1.19, p=0.006), grade 3 vs. 1 (HR 2.51, p=0.002), 3 vs. 1 LN+ (HR 1.54, p=0.02), LVI (HR 1.48, p=0.005), and inner vs. outer tumor position (HR 1.56, p=0.02) for multivariable competing risk regression. With the receipt of PMRT (HR 0.31, p<0.001) and CHT (HR 0.44, p<0.001), decreased LR was correlated. Increased DM was correlated with younger age (HR 1.01, p<0.001), greater tumor size (HR 1.32, p<0.001), grade (grade 2 vs. 1 HR 2.01, p<0.001; grade 3 vs. 1 HR 3.07, p<0.001), 3 vs. 1 LN+ (HR 1.40, p<0.001), N1 vs. N1mic (HR 1.40, p=0.01), LVI (HR 1.31, p<0.001), ER / PR negative (HR 1.28, p=0.02), and location of internal vs. external tumor (HR 1.49, p<0.001). Decreased DM was correlated with PMRT (HR 0.84, p=0.04), CHT (HR 0.64, p<0.001), HER2 +, trastuzumab vs. HER2- (HR 0.60, p=0.001), and subsequent care (HR 0.85, p=0.04) administration. Increased BCM was correlated with younger age (HR 1.01, p=0.01), greater size of tumor (HR 1.33, p<0.001), grade (grade 2 vs. 1 HR 2.61, p<0.001; grade 3 vs. 1 HR 4.06, p<0.001), 3 vs. 1 LN+ (HR 1.39, p=0.002), N1 vs. N1mic (HR 1.69, p=0.03), LVI (HR 1.28, p=0.002), ER / PR negative (HR 1.59, p<0.001), and location of inner vs. outer tumor (HR 1.49, p<0.001). CHT (HR 0.55, p<0.001) and HER2 + trastuzumab vs. HER2- (HR 0.56, p=0.001) were correlated with decreased BCM. Margin status and ECE were not correlated substantially with any endpoints.
Conclusion: PMRT for pT1-2N1 breast cancer is associated with reduced LR and DM at a median follow-up period of 10 years. In a validation cohort, nomograms will be developed and tested to assist clinicians to individualize risk estimates in contemporary practice.