Breast cancer-related lymphedema (BCRL) is a feared outcome of breast cancer treatment, yet the push for early screening is hampered by a lack of standardized quantification. We sought to determine the necessity of preoperative baseline in accounting for temporal changes of upper extremity volume. 1028 women with unilateral breast cancer were prospectively screened for lymphedema by perometry. Thresholds were defined: relative volume change (RVC) ≥10 % for clinically significant lymphedema and ≥5 % including subclinical lymphedema. The first postoperative measurement (pseudo-baseline) simulated the case of no baseline. McNemar's test and binomial logistic regression models were used to analyze BCRL misdiagnoses. Preoperatively, 28.3 and 2.9 % of patients had arm asymmetry of ≥5 and 10 %, respectively. Without baseline, 41.6 % of patients were underdiagnosed and 40.1 % overdiagnosed at RVC ≥ 5 %, increasing to 50.0 and 54.8 % at RVC ≥ 10 %. Increased pseudo-baseline asymmetry, increased weight change between baselines, hormonal therapy, dominant use of contralateral arm, and not receiving axillary lymph node dissection (ALND) were associated with increased risk of underdiagnosis at RVC ≥ 5 %; not receiving regional lymph node radiation was significant at RVC ≥ 10 %. Increased pseudo-baseline asymmetry, not receiving ALND, and dominant use of ipsilateral arm were associated with overdiagnosis at RVC ≥ 5 %; increased pseudo-baseline asymmetry and not receiving ALND were significant at RVC ≥ 10 %. The use of a postoperative proxy even early after treatment results in poor sensitivity for identifying BCRL. Providers with access to patients before surgery should consider the consequent need for proper baseline, with specific strategy tailored by institution.
Keywords: Breast cancer; Lymphedema; Quantification; Standardization; Temporal baseline; Treatment morbidity.