F-fluorodeoxyglucose positron emission tomography-computed tomography for predicting pathological complete response to neoadjuvant chemotherapeutic in breast cancer patients.

BACKGROUND

Accurately predicting pathological complete response (pCR) to neoadjuvant chemotherapy (NACT) in breast cancer remains a clinical challenge. Current imaging-based models are limited in their ability to integrate key metabolic parameters to enhance prediction accuracy. This study aimed to develop and validate a nomogram using F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) parameters, including maximum standardized uptake value (SUV), metabolic tumor volume (MTV), and total lesion glycolysis (TLG), to improve pCR prediction. These parameters, representing both tumor metabolic burden and activity, were hypothesized to collectively provide a robust means of predicting pCR.

METHODS

This retrospective cohort study enrolled 95 breast cancer (BC) patients who underwent F-FDG PET/CT before and after NACT. Patients were categorized into pCR (n=46) and non-pCR (n=49) groups based on postoperative pathological outcomes. Clinical and pathological characteristics, as well as changes in SUV, MTV, and TLG, were compared between the two cohorts. Logistic regression identified independent predictors of non-pCR. The dataset was then randomly divided into training (n=66) and validation (n=29) cohorts for nomogram construction and validation. The model's performance was evaluated using the area under the receiver operating characteristic (ROC) curve, calibration curves, and decision curve analysis.

RESULTS

Relative to the non-pCR cohort, the pCR group exhibited smaller tumor diameters, lower Ki-67 expression, fewer lymph node metastases, and higher proportions of HER2+ molecular subtype (P<0.05). Pretreatment SUV, MTV, and TLG levels in the pCR group were significantly lower than those in the non-pCR group, and showed a marked decrease after treatment (P<0.05), whereas no significant changes were observed in the non-pCR group (P>0.05). SUV, MTV, TLG, and molecular subtype were identified as independent predictors of non-pCR through logistic regression analysis. A nomogram constructed using these predictors achieved area under the ROC curve (AUC) of 0.9003 and 0.9363 in the training and validation cohorts, respectively. The model demonstrated good calibration (Hosmer-Lemeshow test, χ=6.412, P=0.60) and clinical utility through decision curve analysis, effectively stratifying patients at high risk of non-pCR based on a cutoff value of 0.8230.

CONCLUSIONS

F-FDG PET/CT demonstrates significant clinical value in predicting pCR to NACT in BC patients. By integrating metabolic parameters such as SUV, MTV, and TLG into a nomogram, this approach enables accurate prediction of treatment efficacy, aiding in the early identification of patients unlikely to benefit from NACT. This facilitates timely adjustments to personalized treatment plans, optimizing clinical outcomes and resource allocation.