Patterns of recurrence after liver transplantation for nonresectable liver metastases from colorectal cancer

Abstract

Background: Surgical resection is the only curative modality for colorectal liver metastases (CLM), and the pattern of recurrences after resection affects survival. In a prospective study of liver transplantation (Lt) for nonresectable CLM we have shown a 5-year overall survival rate of 60 %, but 19 of 21 experienced recurrence. This study reports the pattern of recurrences after Lt for CLM and the effect on survival.

Methods: Characterization of metastatic lesions in a prospective study for Lt for nonresectable CLM was performed (n = 21). The study included reexamination of chest computed tomographic scans taken before Lt.

Results: At the time of first recurrence, 16 were a single site, and three were multiple sites. Thirteen of the single sites were pulmonary recurrences. The pulmonary recurrences appeared early and were slow growing, and several were accessible to surgical treatment. When chest computed tomographic scans were reexamined, seven patients had pulmonary nodules at the time of Lt without an effect on survival. There was no first single-site hepatic recurrence. Six of the seven patients who developed metastases to the transplanted liver died from metastatic disease.

Conclusions: The pulmonary recurrences after Lt for CLM were of an indolent character, even those that were present at the time of Lt. This contrasts with the finding of metastases to the transplanted liver, which was prognostically adverse. The lack of single hepatic first-site recurrences and hepatic metastases only as part of disseminated disease is different from the pattern of recurrence after liver resection. This suggests two distinct mechanisms for hepatic recurrences after resection for CLM.

Fig. 1

a Distribution of first-site recurrences after liver transplantation (Lt) for colorectal cancer for nonresectable liver metastases. Nineteen patients had experienced recurrence at the end of follow-up. b Surgical interventions and outcome according to distribution of first and second site of recurrence after Lt. Of the 21 patients included in study, six were dead, all because of disseminated cancer disease; seven were with no evidence of disease (NED); and the remaining eight were alive with disease (AWD). No patient was lost to follow-up

Fig. 2

Nineteen of 21 patients who experienced relapse after liver transplantation for colorectal liver metastasis were assessed. All deaths were due to disseminated cancer disease. a Thirteen patients had pulmonary-first metastases, and the 5-year survival from the time of Lt was 72 % (95 % confidence interval 44–99 %). b Three patients had two metastatic sites (two patients had liver and lung, and one patient had liver and ovary) as first sites (green line), whereas the remaining 16 patients had single sites (pulmonary or lymph node metastases or recurrent rectal cancer; yellow line). The survival plot shows time from Lt. c Time from diagnosis of hepatic metastases to death (purple line). The orange line represents patients who experienced pulmonary metastases but no hepatic relapse. From the diagnosis of pulmonary metastases to the end of follow-up, there were no deaths in this group. d Overall survival from time of recurrence. The log-rank test was used for calculation of p-values in panels (b, c)

Fig. 3

Previous chest CT scans on 17 patients who experienced lung metastases were reassessed. Tracing back from obvious metastatic manifestations, prior metastatic nodules were found in several patients. In seven patients they were present at or even before liver transplantation. The plot shows the course of individual patients. Each line is the time from the earliest manifestation of metastases seen retrospectively on CT, the vertical line at point 0 is the time of liver transplantation, and the red bar is the time of diagnosis of hepatic metastases. The patients alive are in upper part of the plot, and the dead, below