Predictors of Pathological Complete Response and Patient-Reported Outcomes During Neoadjuvant Chemotherapy in Early Breast Cancer: A Single-Center Retrospective Cohort Study

Background: Neoadjuvant chemotherapy (NAC) is the standard of care for biologically aggressive early breast cancer, and pathological complete response (pCR) is a robust surrogate of long-term outcome. Real-world cohorts integrating clinical, pathological, and patient-reported variables remain limited. Methods: We retrospectively analyzed 149 consecutive women with stage I-III invasive breast cancer treated with NAC followed by curative surgery at Memorial Hospital, Bucharest, Romania. The primary endpoint was overall pathological complete response (pCR), defined as the absence of residual invasive disease in both breast and axillary lymph nodes (ypT0/Tis ypN0; residual ductal carcinoma in situ permitted). Clinicopathological variables, NAC regimen, residual cancer burden (RCB), stromal tumor-infiltrating lymphocytes (TILs), and patient-reported outcomes (European Organisation for Research and Treatment of Cancer {EORTC} global health, Hospital Anxiety and Depression Scale {HADS}) at baseline and end of NAC were analyzed. Predictors of pCR were identified by univariate analysis and multivariable logistic regression with internal bootstrap validation. Results: Median age was 54 years (IQR: 46-62). Subtype distribution was luminal B 68/149 (45.6%), human epidermal growth factor receptor 2 (HER2)-positive 49/149 (32.9%), and triple-negative 32/149 (21.5%). Overall, pCR was achieved in 41/149 patients (27.5%) as follows: 6/68 (8.8%) in luminal B, 26/49 (53.1%) in HER2-positive, and 9/32 (28.1%) in triple-negative breast cancer (TNBC) (χ²=27.95, df=2, p<0.001). pCR rates also differed markedly across NAC regimens (χ²=27.55, df=6, p<0.001), ranging from 1/17 (5.9%) for TC to 12/19 (63.2%) for taxane/trastuzumab/pertuzumab (THP). On multivariable analysis, clinical subtype, stromal TILs, and clinical nodal status were the principal independent predictors of pCR; the integrated model achieved an area under the receiver operating characteristic curve (AUC) of 0.93 (bootstrap-corrected: 0.92), with adequate calibration (Hosmer-Lemeshow χ²=4.90, df=6, p=0.557). EORTC global health declined by a median of four points between baseline and end of NAC (Wilcoxon W=2615.5, p<0.001), with 51/149 patients (34.2%) experiencing a clinically meaningful decline. At baseline, 30/149 patients (20.1%) had clinically significant anxiety, and 57/149 (38.3%) had borderline depressive symptoms; clinical depression emerged de novo in 18/149 (12.1%) by the end of NAC. Baseline patient-reported outcomes did not significantly predict pCR after adjustment for biological variables (likelihood-ratio χ²=4.71, df=3, p=0.194). Conclusions: In this Romanian single-center cohort, pCR was independently predicted by clinical subtype and a limited set of biological variables, in line with contemporary literature. Although baseline patient-reported outcomes did not improve prediction of pathological response, the high prevalence of pre-treatment anxiety and the emergence of clinical depression during NAC argue for systematic psychosocial screening as part of standard neoadjuvant care.