Predicting High Grade Lesions of Sinusoidal Obstruction Syndrome Related to Oxaliplatin-Based Chemotherapy for Colorectal Liver Metastases

Abstract

Oxaliplatin-based chemotherapy induces sinusoidal obstruction syndrome (SOS) lesions in the nontumorous liver parenchyma, which may increase the risk of liver resection for colorectal liver metastases. The objective of this study was to evaluate the accuracy of aspartate aminotransferase to platelet ratio index (APRI) and FIB-4 scoring systems to predict chemotherapy-associated liver injury and to correlate the severity of sinusoidal injury with postoperative outcome. Between 1998 and 2007, 78 patients were operated for colorectal liver metastases after preoperative oxaliplatin-based chemotherapy. Grading of steatosis and SOS in the nontumorous liver parenchyma was obtained in these patients. Univariate analysis of 18 preoperative factors to predict SOS occurrence was performed as well as multivariate analysis. Relevance of preoperative platelet count level, transaminase levels, and fibrosis scoring systems were evaluated to predict high grade lesions of SOS using a receiving operative curve analysis. Ninety-day mortality and morbidity were studied according to SOS severity in 51 patients who underwent major liver resection. Overall, pathologic examination showed high-grade lesions of SOS (SOS 2/3) in 46 (59%) patients. Univariate analysis showed that a low preoperative platelet count, elevated preoperative aspartate aminotransferase, short interval between chemotherapy and surgery were significant factors associated with high-grade lesions of SOS. Multivariate analysis showed that only the APRI score was an independent predictive factor for severe SOS. Receiving operative curve analysis revealed that the cut-off value predicting high-grade lesions of SOS with the best accuracy was an APRI score of 0.36 (area under the curve, 0.85; sensitivity, 87%; specificity, 69%). After major liver resection (n = 51), SOS 2/3 (n = 38) was associated with postoperative hepatic dysfunction (26/38 in SOS 2/3 vs. 3/13 in SOS 0/1; P = 0.004) and ascites (P = 0.03). A low preoperative platelet count and high APRI score seem to be the most reliable indicators to predict SOS severity.