Contemporary toxicity profile of breast brachytherapy versus external beam radiation after lumpectomy for breast cancer

Purpose We compared toxicities after brachytherapy versus external beam radiation therapy (EBRT) in contemporary breast cancer patients. Methods and Materials Using MarketScan healthcare claims, we identified 64,112 women treated from 2003 to 2012 with lumpectomy followed by radiation (brachytherapy vs EBRT). Brachytherapy was further classified by multichannel versus single-channel applicator approach. We identified the risks and predictors of 1-year infectious and noninfectious postoperative adverse events using logistic regression and temporal trends using Cochran-Armitage tests. We estimated the 5-year Kaplan-Meier cumulative incidence of radiation-associated adverse events. Results A total of 4522 (7.1%) patients received brachytherapy (50.2% multichannel vs 48.7% single-channel applicator). The overall risk of infectious adverse events was higher after brachytherapy than after EBRT (odds ratio [OR] = 1.21; 95% confidence interval [CI] 1.09-1.34, P<.001). However, over time, the frequency of infectious adverse events after brachytherapy decreased, from 17.3% in 2003 to 11.6% in 2012, and was stable after EBRT at 9.7%. Beyond 2007, there were no longer excess infections with brachytherapy (P=.97). The overall risk of noninfectious adverse events was higher after brachytherapy than after EBRT (OR=2.27; 95% CI 2.09-2.47, P<.0001). Over time, the frequency of noninfectious adverse events detected increased: after multichannel brachytherapy, from 9.1% in 2004 to 18.9% in 2012 (Ptrend =.64); single-channel brachytherapy, from 12.8% to 29.8% (Ptrend<.001); and EBRT, from 6.1% to 10.3% (Ptrend<.0001). The risk was significantly higher with single-channel than with multichannel brachytherapy (hazard ratio = 1.32; 95% CI 1.03-1.69, P=.03). Of noninfectious adverse events, 70.9% were seroma. Seroma significantly increased breast pain risk (P<.0001). Patients with underlying diabetes, cardiovascular disease, and treatment with chemotherapy had increased infectious and noninfectious adverse events. The 5-year incidences of fat necrosis, breast pain, and rib fracture were slightly higher after brachytherapy than after EBRT (13.7% vs 8.1%, 19.4% vs 16.0%, and 1.6% vs 1.3%, respectively), but the risks were not significantly different for multichannel versus single-channel applicators. Conclusion Toxicities after breast brachytherapy were distinct from those after EBRT. Temporal toxicity trends may reflect changing technology and evolving practitioner experience with brachytherapy.