In the context of multimodal therapy liver resections (LR) for metastases of colorectal cancer (CRC) are frequently performed after chemotherapy (CTx). Several CTx-medications may result in liver damage up to fibrosis or cirrhosis. Recent studies reported conflicting results regarding complications after LR in correlation to prior CTx. We, therefore, analyzed our experience with postoperative complications after LR for CRC metastases.
Methods: Since 1998 199 primary LR were performed for CRC metastases (wedge resections 14%, segmental 32% and hemi- or extended hemihepatectomy 54%). Following neoadjuvant CTx LR was performed not earlier than four weeks after CTx. CTx was performed in 62% before LR either as adjuvant CTx of the primary CRC or as neoadjuvant therapy for liver metastases. During our analyses we also differentiated between the first and second part of the study period (P1/P2) and between limited (wedge/segmental) and extended (at least hemihepatectomy) LR. In P2, under new direction of the hepatobiliary surgery team, refined surgical techniques and perioperative management strategies were applied. In P2 the frequency of LR increased 3-fold compared to P1. Seven clinical, multiple laboratory parameters and preoperative CTx were evaluated on their prognostic influence on the occurrence of complications. Multivariate risk factors were analyzed by binary logistic regression.
Results: Any complication occurred in 48%, surgical complications in 32% and hepatic insufficiency in 6.5%. Mortality was 2.5%. In P1 four patients died perioperatively, in P2 only one (mortality < 1% in P2). Four of those five patients were diabetic. Multivariate risk factors for any complication were a diabetes (p<0.05; relative risk RR 3.1) and extended resections (p<0.05; RR 2.5). Risk factors for surgical complications were again extended resections (p<0.001; RR 3.3) and male gender (p<0.01; RR 3.2). Independent risk factors for hepatic insufficiency were a diabetes (p<0.02; RR 2.0) and extended resections (p<0.02; RR 2.3). A diabetes (p<0.01; RR 3) and resections performed during P1 (p<0.01; RR 2.6) were independent risk factors for mortality. Age, BMI, laboratory values and blood transfusions did not influence any of the various complication rates. Preoperative CTx did not correlate with complications (even tendency of fewer complications after CTx)
Conclusions: The extent of resections and diabetes are relevant risk factors for complications following hepatic resection for CRC metastases. An increasing centre experience clearly decreases perioperative mortality. In our experience a preoperative chemotherapy did not increase postoperative morbidity.