RESUMEN
BACKGROUND: Multiple factors can affect cosmetic outcome after breast-conserving therapy. We investigated which combination of factors could optimally predict patient reported cosmetic outcome. METHODS AND MATERIALS: Dutch patients treated with BCT between 2008 and 2013 were retrospectively analysed. Demographics, tumour characteristics, and treatment-related factors were obtained from available digital medical records. Patients were asked to indicate their degree of cosmetic satisfaction using the Harvard scale. Both univariable and multivariable logistic regression analyses were performed to create a prediction model. RESULTS: The present study based on 220 patients shows that univariable analysis radiotherapy boost (OR 4.01 [1.85-8.70], p ≤ 0.001) and adjuvant chemotherapy (OR 2.67 [1.45-4.92], p = 0.002) were significantly associated with unsatisfactory cosmetic outcome. In multivariable analyses, only a radiotherapy boost remained significantly associated (OR 4.08 [1.76-9.49], p = 0.001) with poor cosmetic outcome, while adjuvant chemotherapy was no longer associated with it (OR 1.61 [0.81-3.24), p = 0.18). These two factors, together with tumour histology and lymph node staging, showed the highest predictive values, although the predictive property of the overall model was limited (AUC = 0.69). CONCLUSION: In the present study, a variety of factors is explored. However, a radiotherapy boost was the only factor that was independently associated with a poor cosmetic outcome after BCT. This finding can be used for counselling in daily clinical practice. Creation of a prediction model will require further investigation.