Loco-regional therapy and the risk of breast cancer-related lymphedema: a systematic review and meta-analysis

Background This meta-analysis was designed to assess the association between two loco-regional therapies, regional nodal irradiation (RNI) and axillary lymph node dissection (ALND), and breast cancer-related lymphoedema (BCRL). Methods We searched PubMed, Science Direct, Embase, and BMJ databases for clinical studies published between January 1, 2010 and January 1, 2020, which assessed risk factors and incidence/prevalence of BCRL. Two investigators independently selected articles to extract relative data and calculate corresponding exact binomial 95% confidence intervals (CIs). In total, 93 articles were reviewed, from which 19 studies were selected. The extracted data were pooled using a random-effects mixed model. Results The incidence of lymphedema in the selected studies ranged from 3% to 36.7%, with a pooled incidence of 14.29% (95% CI 13.79-14.79). The summary odds ratio/risk ratio (OR/RR) of ALND vs. no-ALND was 3.67 (95% CI 2.25-5.98) with a heterogeneity (I-2) of 81% (P < 0.00001). After excluding the studies with an abnormally high risk of lymphedema from self-reporting, the summary hazard ratio (HR) was 2.99 (95% CI 2.44-3.66) with no heterogeneity (I-2 = 0%, P = 0.83). The summary OR/RR of patients with vs. without radiotherapy (RT) was 1.82 (95% CI 0.92-3.59), but the RR of RT to breast/chest vs. both axillary and supraclavicular areas was 2.66 (95% CI 0.73-9.70). Conclusion Regional nodal irradiation has a significantly higher risk for developing lymphedema than irradiation of the breast/chest wall. Axillary dissection and axillary RT have a similar risk for early-onset of breast cancer-related lymphoedema, although the risk trends higher for axillary dissection.