Abader, Peter

Renal dysfunction is common in liver transplantation (LT) candidates, but differentiating between reversible and irreversible renal injury can be difficult. Kidney biopsy might be helpful in differentiating reversible from irreversible renal injury, but it is associated with significant complications. We aimed to identify pre‐LT predictors of potentially reversible renal injury using histological information obtained on pre‐LT renal biopsy. Data on 128 LT candidates who underwent pre‐LT kidney biopsy were retrospectively collected and correlated with renal histological findings. Indications for kidney biopsy were iothalamate glomerular filtration rate (iGFR) ≤40 mL/minute, proteinuria >500 mg/day, and/or hematuria. According to the biopsy diagnosis, patients were grouped into the following categories: normal (n = 13); acute tubular necrosis (ATN; n = 25); membranoproliferative glomerulonephritis (n = 19); minimal histological changes (n = 24); and advanced interstitial fibrosis (IF) and glomerulosclerosis (GS) (n = 47). Compared with patients having advanced IF/GS, patients with normal biopsies and those with ATN had lower systolic blood pressure (SBP) and diastolic blood pressure (DBP) and higher international normalized ratio and total bilirubin levels (<0.05 for all). Both SBP and DBP directly correlated with the degree of IF and GS (R = 0.3, P ≤ 0.02 for all). SBP ≤90 mm Hg was 100% sensitive and 98% specific in correlating with normal biopsies or ATN, whereas SBP ≥140 mm Hg was 22% sensitive and 90% specific in correlating with advanced IF/GS. Model for End‐Stage Liver Disease score, serum creatinine, iGFR, urinary sodium excretion, and renal size did not correlate with biopsy diagnosis or degree of IF or GS. In conclusion, SBP at the time of LT evaluation correlates with renal histology, and it should be included along with other clinical and laboratory markers in the decision‐making process to list patients with renal dysfunction for LT alone versus simultaneous liver‐kidney transplantation.

INTRODUCTION: Mucosal healing is a widely accepted target for ulcerative colitis (UC) treatment. Linked color imaging (LCI) is a novel endoscopic enhancement system that utilizes Light Emitting Diode (LED) sources to spectrally enhance subtle erythema and mucosal surface patterns which may improve endoscopic classification of IBD associated inflammation. We aimed to assess the accuracy of LCI in classifying UC-associated mucosal healing compared to a validated histological reference standard (Geboes score). METHODS: 22 patients with ulcerative colitis were prospectively enrolled and evaluated by LCI colonoscopy using the Fujifilm’s digital video Processor VP-7000 and LED light source BL-7000 (Fujifilm, Tokyo Japan). Interobserver agreement was reported as kappa values. Geboes score was reported as score range of 0 to 5.4. A Geboes score of ≥3.1 (presence of surface intraepithelial neutrophils or < 5% crypts) was used to define histologic activity and a score of < 2B (no increase in neutrophils in the lamina propria) was used to define histologic remission. Geboes score has been previously validated to predict clinical recurrence. We calculated the overall accuracy, positive predictive value (PPV) and negative predictive value (NPV) of the LCI classification identifying mucosal healing against the Geboes score using the Fisher’s exact test. RESULTS: Total of 62 endoscopic images were obtained for different parts of the colon, 43 of which were matched to a corresponding biopsy. Endoscopic images were scored by 2 endoscopists into LCI-A (no redness), LCI-B (redness with visible vessels) and LCI-C (redness without visible vessels). A final LCI score was given for images with discrepant scores based on a consensus vote from both endoscopists. The interobserver agreement between the two endoscopists was 81%, kappa of 0.71 (95% confidence interval [CI] 0.57- 0.86). Among regions scored ≥3.1 on Geboes score, 9.5%, 28.6% and 61.9 % were classified as LCI-A, LCI-B and LCI-C, respectively ( P < 0.0001). Among images scored < 2B on Geboes score, 81.8%, 13.6% and 4.6% were classified as LCI-A, LCI-B and LCI-C, respectively ( P < 0.0001). Analyzing data against histology with LCI-A representing mucosal healing and LCI-B and C representing active mucosal disease, the overall accuracy becomes 86%, with NPV of 82.6% and PPV of 90% for identifying UC patients with histological and endoscopic evidence of mucosal healing. CONCLUSION: LCI technology has a potential role in assessing mucosal healing in UC.