Gothenburg Intestinal Transplant Endoscopy Score: a proposal for a grading system of the endoscopic findings after intestinal transplantation
Jonas Varkey1, Mihai Oltean2, Magnus Simrén1, Per-Ove Stotzer1, Gustaf Herlenius2.
1Department of Internal Medicine and Clinical Nutrition, Institute of Medicine, Sahlgrenska University Hospital, Gothenburg, Sweden; 2Transplant Institute, Sahlgrenska University Hospital, Gothenburg, Sweden
Microscopic examination of endoscopic biopsies is the mainstay of acute rejection (AR) surveillance after intestinal transplantation. However, biopsies require time for processing and may entail risks. The endoscopic features encountered during AR are quite well defined yet there are no established endoscopic criteria for diagnosing AR or any grading system for the severity of the endoscopic findings. In theory, a grading system could allow a faster preliminary diagnosis of AR based on endoscopic findings. Moreover, such scoring system would reduce interobserver variability and allow for more objective intra- and interpatient comparisons.
We designed a five-stage grading score based on previously described endoscopic features of AR (erythema, edema, ulcerations, friability, villous alterations). Two investigators retrospectively reviewed the endoscopy reports of 26 patients (19 adults and 7 pediatric) receiving 27 intestine containing grafts and graded the examinations using the score. Results were correlated with the pathology findings with regard to the presence and degree of AR.
Five hundred four ileoscopies were reviewed: 374 examinations (74%) were considered normal (G0), 59 had mild alterations (erythema, edematous villi- G1), moderate changes (erosions, blunted villi-G2) were found in 36 endoscopies, 17 ileoscopies revealed advanced changes (ulcerations, villus loss - G3) while in 18 occasions the changes were estimated as severe (mucosal loss- G4). Inter-reviewer agreement was very good (kappa=0,815, p<0.001). AR was found in biopsies from 86 endoscopy sessions (17%) and in 63 cases AR was considered moderate or severe. For the cases with low-degree AR the sensitivity of this new grading score was 66% and the specificity was 86% whereas the positive and negative predictive values were 45% and 93% respectively. During advanced AR the sensitivity of the score was 90% and the specificity was 86% whereas the positive and negative predictive values were 40% and 98% respectively.
We herein suggest a novel grading score that both summarizes the findings and allows comparisons between intestinal graft endoscopies. Its prospective evaluation is ongoing.