CHRONIC LIVER ALLOGRAFT-REJECTION - A NATIONAL INSTITUTE OF DIABETES AND DIGESTIVE AND KIDNEY-DISEASES INTERINSTITUTIONAL STUDY ANALYZING THE RELIABILITY OF CURRENT CRITERIA AND PROPOSAL OF AN EXPANDED DEFINITION

A study was conducted to assess the inter and intrarater agreement for the histopathologic features and diagnosis of chronic rejection (CR) and several other important causes of late liver allograft dysfunction . On two occasions, five pathologists, experienced with liver transpla ntation, reviewed a set of 49 slides representing a range of diagnoses , without knowledge of the clinical history or liver injury test resul ts. The readings were correlated with the original histopathologic dia gnosis, liver injury tests, and clinicopathologic follow-up. Assessmen t of biopsy adequacy (kappa = 0.69) and portal tract counts (kappa = 0 .79) showed good to excellent intrarater agreement, whereas interrater agreement for these variables was moderate to good, respectively (kap pa = 0.44 and 0.65). Likewise, the intrarater agreement for the diagno sis of CR (kappa = 0.68), hepatitis (kappa = 0.77), and obstructive ch olangiopathy (kappa = 0.55) showed good to excellent agreement, wherea s the interrater agreement for these same diagnoses ranged from fair t o good (kappa = 0.58, 0.46, and 0.25, respectively). In 18 specimens, there was a near unanimous diagnosis of CR across both readings. These biopsies were obtained at a median of 7.1 months (range, 42 days to 4 .9 years) after transplantation, and the average number of portal trac ts was 8.4 (range, 4-15). Interestingly, only 13 of these 18 specimens showed bile duct loss in >50% of the portal triads; the remaining cas es showed atrophy/pyknosis of the biliary epithelium in a majority of small bile ducts. Clinicopathologic correlation showed that these 18 b iopsies were obtained from 16 grafts from 15 patients, 14 of whom ulti mately required retransplantation or died of or with CR, whereas two o f the grafts/patients recovered. A high rate of sensitivity (92%) and a somewhat lower, but acceptable, rate of specificity (71% to 80%) was found for the diagnosis of CR. Chronic rejection and other causes of late liver allograft dysfunction can he diagnosed reliably by a group of pathologists experienced with liver transplantation. and the diagno sis of CR correlates with clinical course and liver function abnormali ties. Expanded criteria for the diagnosis of CR are presented, and pot ential problem areas for practicing pathologists are discussed.