The international surgical journal with global reach

This is the Scientific Surgery Archive, which contains all randomized clinical trials in surgery that have been identified by searching the top 50 English language medical journal issues since January 1998. Compiled by Jonothan J. Earnshaw, former Editor-in-Chief, BJS

Systematic review and meta‐analysis of liver resection for metastatic melanoma. BJS 2013; 100: 1138-1147.

Published: 17th June 2013

Authors: J.‐M. Aubin, J. Rekman, F. Vandenbroucke‐Menu, R. Lapointe, R. J. Fairfull‐Smith, R. Mimeault et al.

Background

The multidisciplinary management of metastatic melanoma now occasionally includes major hepatic resection. The objective of this work was to conduct a systematic review of the literature on liver resection for metastatic melanoma.

Method

MEDLINE, Embase, the Cochrane Library and Scopus were searched (1990 to December 2012). Studies with at least ten patients undergoing liver resection for metastatic melanoma were included. Data on the outcomes of overall survival (OS) and/or disease‐free survival (DFS) were abstracted and synthesized. Hazard ratios (HRs) were derived from survival curves and subjected to meta‐analysis using random‐effects models.

Results

Twenty‐two studies involving 579 patients (13 per cent weighted resection rate) who underwent liver resection were included. Study quality was poor to moderate. Median follow‐up ranged from 9 to 59 months. Median DFS ranged from 8 to 23 months, and median OS ranged from 14 to 41 months (R0, 22–66 months, R2, 10–16 months; R0 versus R1/R2: HR 0.52, 95 per cent confidence interval (c.i.) 0.37 to 0.73). The OS rate was 56–100 per cent at 1 year, 34–53 per cent at 3 years and 11–36 per cent at 5 years. Median OS with non‐operative management ranged from 4 to 12 months. Comparison of OS with resection and non‐operative management favoured resection (HR 0.32, 95 per cent c.i. 0.22 to 0.46).

Conclusion

Radical resection of liver metastases from melanoma appears to improve overall survival compared with non‐operative management or incomplete resection, but this observation requires future confirmation as selection bias may have confounded the results.

Full text