Surveillance of Allograft Rejection After Intestinal Transplantation Using Endoscopy

Study design

This cohort study was conducted prospectively and approved by the ethics committee of Far Eastern Memorial Hospital (IRB No. FEMH-105023-F). Consecutive patients who received IT between November 2016 and September 2017 were enrolled. The inclusion criteria were as followings: post-IT patients older than 7-year-old, and with informed consent for endoscopy and biopsy. Those with bleeding tendency (platelet count <80k/ul, or prothrombin time international normalized ratio >1.5), without informed consent, or leukopenia (absolute neutrophil count <1,500/ul) were excluded.

Endoscopic and histological grading for allograft rejection Endoscopic surveillance was carried out after IT twice a week during the first month, once a week in the second month and then once per month until ileostomy closed. IEE examinations were performed with magnifying endoscopy (ME) which has powerful 80 times optical magnification under narrow-band imaging (NBI) system (Evis Lucera CLV-260NBI, GIF-H260Z endoscopy, Olympus Medical Systems Corp, Tokyo, Japan).14 ME was intubated through chimney ileostomy to anastomosis site to evaluate the changes of superficial mucosal surface and vascular architecture. Biopsy samples were taken from endoscopically suspicious rejection areas for histological examination. All the endoscopic procedures were done by one experienced endoscopist (C.S-C.) who was blinded to the histological reports.

All the histological reports were carried out by one experienced pathologist (C.C.-T.). The severity of histological rejection was graded according to followings:11 1) Indeterminate:

minimal histological changes including localized inflammatory infiltrate, crypt epithelial injury, increased crypt epithelial apoptosis but less than 6 apoptotic bodies/10crypts, no to minimal architectural distortion, no mucosal ulceration, or changes insufficient for the diagnosis of mild rejection; 2) Mild: mild localized inflammatory infiltrate with activated lymphocytes, mild crypt epithelial injury, increased crypt epithelial apoptosis > 6 apoptotic bodies/10crypts, mild architectural distortion but no mucosal ulceration; 3) Moderate: widely dispersed inflammatory infiltrate in lamina propria, diffuse crypt epithelial injury, increased crypt apoptosis with focal confluent apoptosis, more prominent architectural distortion, possible mild to moderate intimal arteritis, no mucosal ulceration; 4) Severe: features of moderate ACR, possible ulceration and denudation of mucosa, severe intimal arteritis or transmural arteritis may be seen. We classified those with apoptotic body≦1/10crypts and without other findings associated with indeterminate, mild, moderate to severe rejection as no rejection.

We used a novel endoscopic scoring system to record the architectural changes of microsurface and microvascularity of mucosa (Table 1). The observed endoscopic findings were recorded by the following five components ("V-E-N-C-H") graded 0 to 2 (Figure 1): 1) "V" (villi appearance): string-like (V-0), tongue-like (V-1), dome-like (V-2); 2) "E" (erythema of mucosa): no redness (E-0), sporadic (E-1), diffuse (E-2); 3) "N" (network of capillary): normal capillary loops (N-0), meandering caliber of capillary loops and loss of capillary loops in <50% villi under one maximal ME view (N-1), loss of capillary loops >50% villi under one maximal

ME view (N-2); 4) "C" (crypt widening): narrow (C-0), slightly widened within a width of one villi (C-1), markedly widened beyond a width of one villi (C-2); 5) "H" (heterogeneity of mucosal change): minimal (H-0), patch (H-1), diffuse (H-2).

Statistical analysis The continuous variables were expressed as mean ± standard deviation, whereas categorical variables were summarized as count (%). Pearson's correlation coefficient was used to analyze the severity of rejection between endoscopic scoring system and histological grading system. The significance levels were defined as p value <0.05 and the statistical analysis was performed using STATA software (version 11.0; Stata Corp, College Station, TX, USA).