Optimizing surgical indication in patients with borderline resectable pancreatic ductal adenocarcinoma to prevent futile pancreatectomy.

BACKGROUND

Borderline resectable pancreatic ductal adenocarcinoma (BR-PDAC) carries a high risk of early recurrence (ER) despite neoadjuvant chemotherapy (NAC) and surgery. Identifying ER predictors is essential to optimize surgical indication.

METHODS

We retrospectively analyzed 70 patients with BR-PDAC, of whom 48 underwent resection. ER was defined as recurrence within 8 months. Pre- and post-NAC carbohydrate antigen 19-9 (CA19-9) levels were assessed using Cox regression and receiver operating characteristic (ROC) curve analysis.

RESULTS

ER occurred in 18 patients (38%). Both pre-NAC (median, 699 vs. 71 U/mL; P = 0.010) and post-NAC CA19-9 levels (149 vs. 27 U/mL; P = 0.002) were significantly higher in ER patients. ROC curve analysis identified a post-NAC CA19-9 cutoff of 100 U/mL (area under the curve, 0.77) predicting ER. Patients with post-NAC CA19-9 ≥100 U/mL had significantly worse progression-free (hazard ratio [HR], 5.84; P < 0.001) and overall survival (HR, 6.36; P = 0.002). Notably, patients with ER had a similar OS to those who did not undergo surgery (HR, 0.93; P = 0.87).

CONCLUSIONS

Persistently elevated CA19-9 after NAC predicts ER and poor survival, suggesting limited benefit from resection. Post-NAC CA19-9 may help prevent futile pancreatectomy.