Objective To investigate the value of liver fibrosis serum markers in predicting esophagogastric variceal re-bleeding (EGVR) after laparoscopic splenectomy and azygoportal discon-nection (LSD).

Methods The prospective study was conducted. The clinical data of 155 cirrhotic portal hypertension patients with EGVR after LSD in the Clinical Medical College of Yangzhou University from September 2014 to January 2017 were selected. Observation indicators: (1) grouping situations of the enrolled patients; (2) risk factors analysis for postoperative EGVR; (3) prediction of postoperative EGVR; (4) follow-up. Follow-up was conducted using telephone interview, outpatient examination and hospitalization. Patients were followed up once every 3 months after operation to detect occurrence of EGVR and survival of patient up to January 2018. Measurement data with normal distribution were represented as Mean±SD, and comparison between groups was conducted using the t test. Measurement data wite skewed distribution were represented as M(range), and comparison between groups was conducted using the non-parameter test. Count data were described as absolute numbers, and comparison between groups were conducted using the chi-square test or Fisher exact probability. Logistic regression model was used for multivariate analysis. The area under curve (AUC) of receiver operating characteristic (ROC) curve was used to estimate the diagnostic efficiency. The Youden index was used to determine the optimal cut-off point.

Results (1) Grouping situations of the enrolled patients. A total of 155 patients were selected for eligibility. There were 106 males and 49 females, aged (53±11)years. Of the 155 patients, there were 21 cases with EGVR in the postoperative 1 year and 134 cases without EGVR in the postoperative 1 year. The protein expression of laminin and collagen Ⅳ were 100.3(range, 16.1‒712.2)µg/L and 68.4(range, 35.0‒198.8)µg/L in patients with EGVR, vs 35.5(range, 2.0‒521.2)µg/L and 43.5(range, 4.3‒150.4)µg/L in patients without EGVR, showing significant differences between them (Z=‒4.55, ‒4.52, P<0.05). (2) Risk factors analysis for postoperative EGVR. According to the Youden index, the optimal cut-off point of protein expression of laminin and collagen Ⅳ were 64.0 µg/L and 65.0 µg/L, respec-tively. Results of multivariate analysis showed that the protein expression of laminin ≥64.0 µg/L and the protein expression of collagen Ⅳ ≥65.0 µg/L were independent risk factors for postoperative EGVR (odds ratio=9.69, 8.16, 95 confidence intervals as 3.05‒30.82, 2.65‒25.15, P<0.05). (3) Prediction of postoperative EGVR. Results of ROC curve showed that the AUC of laminin and collagen Ⅳ in predicting postoperative EGVR was 0.79 (95% confidence interval as 0.66‒0.92), with sensi-tivity as 0.62 and specificity as 0.96. (4) Follow-up. All the 155 patients were followed up for 12(range, 1‒12)months. During the follow-up, there were 21 of the 155 patients (13.55%) with post-operative EGVR, including 3 cases died of EGVR. Of the 21 patients with postoperative EGVR, there were 6 cases with postoperative EGVR during the first month after operation including 2 cases died, 5 cases with postoperative EGVR at postoperative 1‒3 month, 6 cases with postoperative EGVR more than 3 month and less than 6 month after operation and 4 cases with postoperative EGVR at postoperative 6‒12 months including 1 case died at postoperative 12 month.

Conclusions Laminin and collagen Ⅳ show satisfactory ability to predict EGVR after LSD.