Radiologic Occult Metastases in Pancreatic Cancer: Analysis of Risk Factors and Survival Outcomes in the Age of Contemporary Neoadjuvant Multi-agent Chemotherapy.

BACKGROUND

Radiologic occult metastatic disease (ROMD) in patients with pancreatic ductal adenocarcinoma (PDAC) who undergo contemporary neoadjuvant chemotherapy (NAC) has not been well studied. This study sought to analyze the incidence, risk factors, and oncologic outcomes for patients who underwent the NAC approach for PDAC.

METHODS

A retrospective review analyzed a prospectively maintained database of patients who had potentially resectable PDAC treated with NAC and were offered pancreatectomy at our institution from 2011 to 2022. Multivariable regression analysis was performed to assess risk factors associated with ROMD. Kaplan-Meier curves with log-rank analyses were generated to estimate time-to-event end points.

RESULTS

The study enrolled 366 patients. Upfront and borderline resectable anatomic staging comprised 80% of the cohort, whereas 20% had locally advanced disease. The most common NAC regimen was FOLFIRINOX (n = 274, 75%). For 55 patients (15%) who harbored ROMD, the most common site was liver-only metastases (n = 33, 60%). The independent risk factors for ROMD were increasing CA19-9 levels during NAC (odds ratio [OR], 7.01; confidence interval [CI], 1.97-24.96; p = 0.008), indeterminate liver lesions (OR, 2.19; CI, 1.09-4.39; p = 0.028), and enlarged para-aortic lymph nodes (OR, 6.87; CI, 2.07-22.74; p = 0.002) on preoperative cross-sectional imaging. Receipt of palliative chemotherapy (p < 0.001) and eventual formal pancreatectomy (p = 0.04) were associated with survival benefit in the log-rank analysis. The median overall survival (OS) of the patients with ROMD was nearly 15 months from the initial diagnosis, with radiologic evidence of metastases occurring after a median of 2 months.

CONCLUSIONS

Radiologic occult metastatic disease remains a clinical challenge associated with poor outcomes for patients who have PDAC treated with multi-agent NAC.