Evaluation of cosmetic outcome following breast-conserving therapy in trials: panel versus digitalized analysis and the role of PROMs.

Cosmetic outcome is an important quality of life-related end point following breast-conserving therapy (BCT). To advise on a gold standard, we compare cosmetic outcome evaluated by panel and an objective evaluation (BCCT.core software). Second, patient-reported outcome measures (PROMs) are compared to cosmetic outcome evaluation by panel and BCCT.core. Sixty-eight breast cancer patients were included following BCT between 2007 and 2012. Two independent 6-member panels and two observers using the BCCT.core evaluated cosmetic outcome. First, reproducibility, repeatability, and relatedness of panel and BCCT.core were analyzed using the interclass correlation coefficient (ICC). Second, the association between panel/BCCT.core with PROMs (EORTC-QLQ-C30/BR23, EQ-5D-5L, and BREAST-Q) was analyzed with a linear regression and the goodness of fit by the R2 . Both panel and BCCT.core evaluations showed "excellent" intraobserver agreement (ICC 0.93 [95% CI: 0.83; 0.97] and 0.93 [95% CI: 0.84; 0.97]) for respectively panel 1 and BCCT.core 1 and "excellent" interobserver agreement (ICC 0.94 [95% CI: 0.90; 0.96] and 0.85 [95% CI: 0.77; 0.91]) respectively for panel and BCCT.core. Association between panel and BCCT.core varied, ICC 0.59-0.69. Only the PROM BREAST-Q showed a significant association with both panel evaluations and BCCT.core observers (panel 1 and BCCT.core 1; R2 of .157 [P = .002] and .178 [P = .001]). Both panel and BCCT.core showed comparable "excellent" intraobserver and interobserver agreement. For future trials evaluating cosmetic outcome following BCT, one of those can be chosen. Solely, the PROM BREAST-Q showed a significant association with panel and/or BCCT.core evaluation. To enable standardized cosmetic outcome evaluation and corresponding patient satisfaction in future trials, at least the BREAST-Q should be combined with a panel or BCCT.core evaluation.