Abstract
Background Breast-conserving surgeries have significantly advanced breast cancer treatment, offering favorable oncological outcomes, enhanced cosmetic results, reduced postoperative morbidity, and better psychological acceptance compared to mastectomy. The introduction of neoadjuvant therapy has expanded the applicability of breast conservation surgery to include locally advanced tumors. Tumor response to neoadjuvant chemotherapy is evaluated using imaging modalities such as breast ultrasound, breast magnetic resonance imaging (MRI), and positron emission tomography/computed tomography (PET/CT). Accurate prediction of therapeutic response facilitates the planning of surgical and adjuvant treatments. This study aims to compare the diagnostic accuracy of MRI and PET/CT in predicting treatment response to neoadjuvant chemotherapy in breast cancer patients. Methods This retrospective study was conducted at a tertiary care center in Bahrain. A total of 138 patients with locally advanced breast cancer or human epidermal growth factor receptor-2 (HER2) positive, hormone receptor-negative cancers who underwent breast-conserving surgeries between June 2018 and December 2022 were included. The inclusion criteria focused on patients achieving a complete pathological response following neoadjuvant systemic therapy, ensuring a homogenous study population. Patients with hormone receptor-positive early breast cancers or metastatic tumors, ineligible for neoadjuvant chemotherapy, were excluded. Non-responders and partial responders were also excluded from the study. Statistical analysis was performed using IBM SPSS v26.0 (IBM Corp., Armonk, US). Response rates for the imaging modalities and histopathology results were assessed. Agreement between histology and imaging modalities was computed using kappa statistics. Diagnostic performance for predicting "no residual" disease was evaluated using the McNemar Test. All tests were two-tailed, with a p-value <0.05 considered statistically significant. Results The study included 138 patients, of whom 73 (52.9%) had an incomplete response or residual disease, while 65 (47.1%) had a complete response or no residual disease according to histology reports. There was slight agreement between post-neoadjuvant MRI and histology results (Cohen's kappa 0.172, p=0.010), while substantial agreement was observed between post-neoadjuvant PET/CT and histology results (Cohen's kappa 0.614, p=0.000). PET/CT demonstrated a higher sensitivity of 93.8% (p<0.001) and a specificity of 68.5%. Although MRI was more specific, the positive predictive value was comparable for both PET/CT and MRI. Conclusion PET/CT shows higher sensitivity and can serve as an early marker for predicting complete pathological response in post-neoadjuvant breast cancer patients. However, the prediction of residual disease is optimized by combining both MRI and PET/CT as diagnostic modalities.