Background: The prediction of conversion surgery in patients with technically unresectable colorectal liver metastases has not been generalized or well-established. We developed a predictive model for conversion surgery and assessed the long-term outcomes of patients with technically unresectable colorectal liver metastases.
Methods: In this single-center, retrospective study, we analyzed the perioperative parameters and outcomes of 892 consecutive patients (2014-2021). Conversion surgery was indicated when the chemotherapy response allowed the complete resection of colorectal liver metastases with negative margins and adequate remnant liver volume.
Results: Of the 892 patients, 122 had technically unresectable colorectal liver metastases; 61 underwent conversion surgery (conversion surgery group) and 61 did not (nonconversion surgery group). The median overall survival was significantly higher in the conversion surgery group than in the nonconversion surgery group (5.6 vs 1.8 years, P < .001). After univariate and multivariate analyses, the predictive model for conversion surgery was constructed using 4 predictive factors: Rat sarcoma viral oncogene homolog status (mutant, +2 points), tumor number (≥15, +1), hepatic vein contact (≥2 hepatic veins, +1), and the presence of preservable sections (absence of preservable sections, +2). The area under the curve for conversion surgery was 0.889. Patients were graded according to the scores (A [0-2], B [3-4], and C [5-6]), and the conversion rates were 91.5% (A), 32.6% (B), and 10.3% (C) (P < .001). Grade A patients (median survival time, 5.7 years) had significantly better overall survival than grade B and C patients (median survival time, 2.2 and 1.6 years, respectively; P < .001).
Conclusion: Patients who underwent conversion surgery for technically unresectable colorectal liver metastases had better prognoses, and our novel predictive model was useful in predicting conversion surgery and prognosis.
Copyright © 2022. Published by Elsevier Inc.