A l 1 2 0 AASLD ABSTRACTS GASTROENTEROLOGY, Vol. 108, No. 4

• ALPHA INTERFERON TREATMENT PREVENTS HEPATOCELLULAR CARCINOMA IN HCV RELATED LIVER CIRRHOSIS. G. Mazzella, E. Accogli, S. Sottili, O. Festi, M. Orsini, A. Salzetta, V. Novelli, A, Cipolla, C. Fabbd, A. Pezzoli, E. Roda. Cattedra di Gastroenterologia - University of Bologna - Bologna Italy,

The aims of ~z-interferon treatment for Chronic viral liver infections are clearance of the virus and healing of the disease. Hepatocellular carcinoma is a complication of viral cirrhosis; but it is not yet known whether treatment of viral cirrhosis with (z-interferon prevents this complication. The incidence and the risk (Cox regression analysis) of developing hepatocellul~r carcinoma were calculated in 347 patients with hepat ic cirrhosis, 227 (34 HBV and 193 HCV related) of whom were treated with c¢-interferon and 120 (28 HBV and 92 HCV) who did not receive this treatment, in order to evaluate the efficacy of ~x-interferon in the prevention of hepatocellular carcinoma. In all patients, the cirrhosis was well compensated (Child A). Over mean follow-up periods of 49 months for HBV and 32 months for HCV, 20•347 patients (6•62 HBV and 14/285 HCV) developed hepatocellular carcinoma. The risk of developing this tumor was significantly greater in males (p<0.004) and in patients not treated with c¢-interferon (p<0.01), The Relative Risk of developing hepatocallular carcinoma increased significantly (p<0.0001) with each passing year. In patients with hepatic cirrhosis secondary to HBV infection, risk did not seem to be modified by (x-interferon treatment, even though a greater, but not significant, risk (RR=5.2; p=0.27) was calculated for untreated patients; in contrast, in HCV related cirrhosis,, this risk was reduced by a factor of 3.8 (p<0.04). The tumor developed only in non,responder patients regardless of virus type. After adjustment for confounding factors, a statistically significant (p<0.025) effect of interferon treatment in preventing hepatocellular carcinoma was still demonstrated when responders were matched with controls, but not when responders where compared with non-responders. These results show that, in addition to its ability to halt the progression of viral induce liver disease, (x-interferon also benefits the patients with HCV Cirrhosis who respond to this treatment by lowering their risk of developing hepatocellular carcinoma.

Q A PROSPECTIVE STUDY ASSESSING RISK FACTORS FOR VENO-OCCLUSIVE DISEASE (VOD) IN PRE-BONE MARROW TRANSPLANT (BMT) PATIENTS. BM McGuire, ME Herman, JS Miller, DK Freese, and JR Rank. The University of Minnesota Hospital, Minneapolis, MN and Mayo Clinic, Rochester, MN. Background: VOD is common in patients (reported rates up to 54%) who have undergone BMT for hematologic malignancy. It is characterized by tender hepatomcgaiy, jaundice, and fluid retention. Criteria by McDonald and co-workers for the diagnosis of VOD include two of the following events within 20 days of BMT: hyperbilirubinemia, hcpatomegaly or right upper quadrant pain of liver origin, and sudden weight gain and fluid accumulation (Ann Int Med 118:255-67). Prior hepatitis, abnormal pre- BMT liver enzymes, second BMT, or past history of abdominal radiation are suggested pre-BMT "high risk" factors predicting VOD. Aim: The purpose of this study was to review prospectively "high risk" pre-BMT markers for the prediction of VOD post-BMT. A >50% rate of VOD was expected in these patients, given the current literature. Methods~ 20 patients undergoing BMT for txeatment of a malignancy with any "high risk" pre- BMT features were prospectively entered in this study. Patients were evaluated clinically and echographieaily for development of VOD. Results: Patients pre-BMT risks were: 3 with previous BMT, 4 with previous abdominal radiation, 14 with elevated liver enzymes, and 15 with a history of hepatitis. The table shows the pre-BMT "high risk" factors per patient and post-BMT results. The "uncertain" group met VOD criteria, but had other potential causes of liver disease (graft vs. host disease {GVHD)or sepsis). Only one patient clinically had VOD by McDonald, et al criteria. One "uncertain" ~atient had GVHD and at auto , had focal VOD.

# of RiSK CLINICAI~ VOD FACTORS No I Uncertain Yes

1 7 2 2 6 i 3 2 1 4 1

IX TOT*LS 115 I 4 C inclusions: Using accepted edictive pre-BMT criteri= high risk development of VOD POst-BMT, a >50% rate was anticipated. Our study shows a significantly lower occurrence of VOD using these criteria, with only one clinical VOD identified. Reasons for these clinical differences are not clear, but may be related to pre-BMT induction or post-BMT treatment differences. Consideration of prophylactic therapy of patients at high risk for VOD will require better pre-BMT predictors than are currently available. This research was funded in part by Baxter Healthcare Corp., Deerfield, IL.

Q A NON-INVASIVE METHOD OF DETERMINING PORTAL HYPERTENSION: CORRELATION OF HEPATIC ARTERIAL DOPPLER RESISTANCE VALUES WITH HEPATIC WEDGE PRESSURES AND PORTAL FLOW PATTERNS IN THE CIRRHOTIC LIVER. DE McGuire, KP Etzkorn, J Hibbeln, R Aizenstein, TE Wiley, C Owens, D Warner, TJ Layden Departments of Medicine and Radiology, University &Illinois at Chicago, Chicago , Illinois 60612.

Introduction: The liver's dual blood supply renders perfusion evaluation difficult, and often normal hepatopetal portal flow is present in the cirrhotic patient When portal flow is preserved in cirrhosis, increasing hepatic arterial resistance should be seen and may correlate with elevated corrected sinusoidal vein pressures (CSVP). Materials & Mcth.ods: 6 patients being investigated for cirrhosis underwent transjugular liver biopsy and CSV'P by hepatic vein catheterization. While still fasted and sedated, portal doppler imaging demonstrated normal portal flow without collaterals. Hepatic arterial resistive index (HARI), (systolic-diastolic velocity)/systolic, velocity was calculated. Biopsy disclosed 3 patients positive for cirrhosis, 2 without cirrhosis having chronic active hepatits and 1 with alcoholic hepatitis HARPs were obtained in 6 normal patients. Results:

Patient Group (No,) Mean CSVP Mean HARI

Bx pos. cirrhosis (3) 21.6 0.85

Bx neg. cirrhosis (3) 19.7 0.72

Normal controls (6) N/A 0.75

Conclusion~: l) Patients with biopsy-proven cirrhosis had higher elevations in HARI and CSVP compared to non-cirrhotics and normal controls. 2) All patients in the cirrhotic group had HAKI greater than 0 8, while no patient in the non-cirrhotic group and only one patient in the normal group had a HARI greater than 0.8 3) Although the numbers in this study are small, RI may prove to be predictive of CSVP in cirrhotic patients with preserved portal flow who require hepatic vein catheterization

LIVER BIOPSY FEATURES OF GVHD AND CHRONIC VIRAL HEPATITIS IN BONE MARROW TRANSPLANT PATIENTS. R.M. McMahon, C. V. Natchvolodoff. Division of Gastroenterology, Univ. of Arkansas Medical Sciences, Little Rock, AR

PURPOSE: Histopathology of viral chronic active hepatitis (CAH) and chronic Graft-versus-host disease (GVHD) demonstrated in liver biopsies of patients with both.

METHOD: Lists of liver biopsies and bone marrow transplants (BMT) performed between 1989 and 1994 identified 37 BMT patients who had liver biopsie& 18 of these patients received allogeneic transplants. Two of these 18 had features of both GVHD and viral CAH (one with Hepatitis B and CMV; the other with Hepatitis B: Group I). One autologous BMT patient had viral CAH (Hepatitis C) without GVHD (Group II). Viral CAH was defined by serology in addition to liver biopsy. Several patients had evidence of GVHD alone (Group Ill), Other patients had evidence of hepatic injury presumably from idiosyncratic drug reaction or their underlying disease (e.g.. features of amyloid secondary to myeloma). This latter group of patients was used as the contro (Group IV) Representative microscopic specimens of each group were compared in a non-bfinded review.

RESULTS: Microscopic specimens from patients in group IV were most easily distinguished from group I, II, or III specimens; specimens from group I and group lU were most difficult to distinguish from each other. Specimens from group I and II were intermediate in difficulty. Features of GVHD included bile duct alteration and portal inflammation with little parenchymal inflammation; features of CAH included piecemeal necrosis and portal inflammation. Scant fibrosis and mild Iobular inflammation was also seen in CAH specimens.

CONCLUSION: The underlying disease and medical treatment leading to BMT did not reduce the ability to interpret liver biopsy specimens. However, discrimination between GVHD and CAH may be difficult when both are present in BMT patients, The findings in GVHD may be subtle and the concurrent presence of CAH makes the diagnosis of chronic GVHD by liver biopsy difficult. Biopsies from patients with features of both illuminate the site and nature of hepatic injury that may occur with each,