icon-folder.gif   Conference Reports for NATAP  
  42nd ICAAC Meeting
San Diego, Sept 27-31, 2002
Back grey_arrow_rt.gif
Hepatitis C: Can You Replace the Liver Biopsy With Less Invasive Tests
Reported by Jules Levin
  End stage liver disease appears to be the leading cause of death in HIV. Many patients are afraid of the biopsy and it is an invasive procedure, despite that it remains the best way to assess liver disease. Since treatment for HCV in coinfected patients is becoming more common, more patients are in a position to need liver biopsy. Margaret Hoffman-Terry reported on a study to see if she could predict which patients do not need a liver biopsy because other testing could predict if they had more advanced liver disease. She performed a retrospective chart review for 45 coinfected patients undergoing liver biopsy from 1/97 through 10/01 at Lehigh Valley Hospital. Liver biopsies were graded on the basis of the Scheuer system (0=no inflammation/fibrosis to 4=necrosis/cirrhosis). Surrogate markers were compared to biopsy results if done 3 months or less prior to biopsy: platelets, ALT, AST, international normalized ratio (INR), abdominal ultrasound (US), alpha fetoprotein, Cd4 counts, HIV viral loads. 82% had HCV RNA over 1 million copies and 80% had genotype 1. 56% had <50 copies HIV RNA. 76% of the coinfected patients haf mild-moderate liver diagnosis by biopsy (grade 0-2.5), and 24% had advanced diagnosis/cirrhosis (grade 3/4). 67% were male; 33% white, 47% Latino, 20% African-American. Average age 43. Mean CD4 count 419.
RESULTS. Using univariate analysis, you were more likely to have a high grade biopsy (grade 3/4; low grade 0-2.5) if the AST was elevated (avg 140 vs 78 for low grade biopsy), p=0.01. Patients with lower AST/ALT ratio were more likely to have high grade biopsy (0.8 vs 1.2), p=0.02. The INR was 1.2 if you had more advanced liver disease vs 1.0 for patients with lower grade biopsy, p=0.002. Patients having an abnormal ultrasound were more likely to have more advanced liver disease (p=0.003); 81% of patients with advanced liver disease had abnormal ultrasound vs 32% of patients who had abnormal biopsy and had low grade biopsy. Patients with higher platelet counts were more likely to have low grade liver disease (p=0.0007). The average platelet count for patients with low grade liver disease was 196 vs 165 for patients with high grade liver disease. Using a multivariate model they found the ALT/AST ratio (odds ratio 0.8), INR (odds ratio 2.3), and abnormal ultrasound (odds ratio 3.8) were associated with advanced liver disease.
CONCLUSIONS. INR, ALT/AST ratio, and ultrasound can be used to predict those patients least likely to benefit from a liver biopsy. The author is suggesting that a doctor can use these tests to exclude having to do a biopsy. The author is saying that if the ALT/AST ratio is low, if the INR is high and if the ultrasound is abnormal it is more likely that the biopsy will show more advanced liver disease. It is important to bear in mind that these tests talk about probability: it is more likely that a patient will have more advanced liver disease if they have these test results. It still is generally accepted among key hepatitis doctors that the liver biopsy is the only way to most accurately assess a person's stage of liver disease. The author listed the limitations of her study: the study is retrospective; small numbers of patients; there may be other markers we have not looked at yet; the findings may not be generalized to other patient populations. The author concluded that larger studies are needed to see if liver biopsies can be avoided by using less invasive methods to identify low vs high risk patients. And a predictive model should be tested to determine specificity, sensitivity, and predictive value of model. Again, the biopsy remains the best way to assess the stage of liver disease.
End Stage Liver Disease Leading Cause of Death in HCV/HIV Coinfected in France; 46% Had Cirrhosis; 37% had >200 Cd4s & <500 copies/ml viral load
Abstract title: "Characteristics and Causes of Death in HIV-HCV Co-Infected Patients (French Survey Mortalité 2000)". Abstract H-1719
S. DOMINIQUE, D. SALMON-CÉRON1, C. LEWDEN2, T. MAY3, F. BONNET4, L. HERIPRET1, S. BEVILACQUA3, J. BOILEAU5; Cochin Hospital 27, Rue du Fbg St Jacques, Paris, France
Among HIV+ adults deceased in France in 2000, we aimed at describing the main characteristics of those coinfected with HCV. 185 French hospital wards involved in the management of HIV infection prospectively notified all deaths occurring in 2000 (among about 65,000 HIV-infected adults followed). Underlying causes of deaths were documented from standardized questionnaires. A total of 975 deaths were notified. Data of 774 cases were available for analysis. HCV serology testing was positive in 240 patients (31%). In HIV-HCV coinfected pts, end stage liver disease was the most frequent underlying cause of death (33%). At the time of death, cirrhosis was present in 46% of cases (n=105) and hepatocarcinoma in 4% (n=10). AIDS was the underlying cause of death in 28%: NHL (n=14), LEMP (n=11), oesophageal candidiasis (n=11), cerebral toxoplasmosis (n=9). Other causes included non AIDS-related cancer (6%), bacterial infection (6%) cardiovascular disease (8%), drug overdose (4%), suicide (4%). At time of death, CD4+ count was >200/mm3 in 38% pts, HIV-RNA was < 500 cop/mL in 37%. When comparing causes of death in HCV+ and HCV- pts, HCV+ pts were more likely to be in poor socio-economic conditions (45% vs 28%), alcohol abusers (52% vs 19%), active drug abusers (31% vs 3%), to have had a psychiatric disease (34% vs 20%), to be diagnosed earlier for HIV (median: 10.9 vs 5.7 years), to have a higher CD4 cell count (142/mm3 vs 65/mm3) and a lower HIV-RNA (3.8 vs 4.7 log10 cop/mL). More pronounced trends were found for patients with notified positive HCV RNA (60.5%). The study authors concluded that HCV is the leading cause of death in HCV+HIV+ patients. Nearly half of death cases occurred at a relatively preserved immunological condition. Therapies directed towards eradication of HCV virus or slowing progression of fibrosis are urgently needed. The high proportion of unfavorable socioeconomic and psychiatric conditions needs also special attention.
Abnormal Lipids (low cholesterol & triglycerides, high glucose) May Predict Poor Outcome in HCV/HIV Coinfected
Abstract title: "Hypolipidemia Impacts Mortality in Patients with HIV and Hepatitis C (HCV) Co-infection". Abstract H-1723; C. HSIAO, J. MIQDADI; University at Buffalo, SUNY, Buffalo, NY
Hypolipidemia (low levels of cholesterol & triglycerides) is frequently found in severe chronic hepatic insufficiency because the liver is the most active site of lipid metabolism. As well, diabetes is more prevalent in persons with hepatitis C. The study authors say that a low baseline serum cholesterol level is associated with higher mortality risk in patients with liver cirrhosis. One of the major causes of mortality in HIV infected patients is liver related, especially, for those who had HIV HCV co-infection. The cohort study is to see if hypolipidemia impacts mortality in patients with HIV HCV co-infection. A retrospective review was conducted of records of HIV infected patients during 12/98 until 9/01 in an urban HIV Clinic at Buffalo, New York. Statistical analyses were performed using EpiInfo2000 and Prism 3.0 software to see if patient lipid abnormalities impact risk of mortality. The entire study population is 640 patients. Of which 253 patients had HCV (median age: 43; 80 Females, 173 Men; 122 Black, 75 Hispanic, 52 white and 4 others). 41 deaths had occurred during this study period with median time of 385 days (12-833 days) from the date of data collection. Analyses of lipid pr files indicated that serum cholesterol < 150 mg/dl, decreased VLDL, HDL level < 35 mg/dl or triglyceride level < 120 mg/dl were significantly associated with an increased risk of mortality with Odd Ratios (ORs) of 2.79, 1.83 and 3.80 respectively. A higher total cholesterol level (> 200 mg/dl) did not impact mortality; however, a higher triglyceride level (> 320 mg/dl) reduced mortality (P = 0.047, OR: 0.31). Hypotriglyceridemia is found to be the most weighted factor that correlates mortality in HIV HCV co-infected patients; however, it does not impact mortality in HIV-infected patients without HCV infection. The authors concluded that hypolipidemia is a poor prognostic indicator for patients with HIV HCV co-infection. According to previous literature, hypolipidemia may be associated with patients poor nutritional status or advanced stage of liver cirrhosis. In talking with the study authors he feels that elevated glucose and low cholesterol and triglycerides may be a sign that the liver is impaired and may be a sign of more advanced liver disease. On the other hand, patients who undergo HCV therapy successfully be eliminating HCV viral burden may see an increase in cholesterol & triglycerides because the liver is functioned better. The authors suggest that hypobetalipoproteinemia in patients with HCV suggests hepatic steatosis and this can be a sign of advanced liver cirrhosis. In a second study presented here by the same authors, they suggest that in HCV/HIV coinfected patients with either serum glucose <120 mg/dL or triglycerides <120 mg/dL with the combination of any other 2 factors or the combination of any 4 factors can predict more than 50% risk for mortality within 3 years (anemia <12 g/mL, HIV viral load >50,000 copies/ml, age >49, platelets <130,000, LDH >700 U/L (1.25 times upper limit of normal).Also, patients with hypolipidemia may be more susceptible to drug toxicity. Further studies to explore the pathogenesis of hypolipidemia in patients with HIV HCV co-infected patients are warranted.
GGT >100 Ul/ml Predicts More Advanced HCV
Abstract title (H-1733): Features and Biochemical Markers of Histological Severity in HIV-HCV Coinfected Patients
The purpose of this study was to assess potential biochemical markers of histological severity (fibrosis and necroinflamatory activity) in HCV/HIV coinfected patients. They did a prospective analysis of the histological activity index (HAI) and fibrosis (F) in 99 liver biopsies (LB) from HCV/HIV-coinfected, excluding chronic carriers of HBsAg. We evaluated factors associated with HCV (genotype, VL, time of HCV, age at acquisition, sex), HIV (CDC stage, CD4 and VL) and alcohol abuse. Mild-moderate hepatitis was defined for HAI between 1-9, severe for HAI >10. Fibrosis > 10 was considered severe. Most patients were male (77%), mean age 38 years (range 29-51). Prior IVDU reported in 88%, and alcohol abuse in 30 (32%). Mean time of HCV infection, 17 years (range, 3-34). Median CD4 and HIV-RNA at the time of LB were 494 cells/µl (range, 36-1260) and 2.1 log10 (range, 1,7-5,4). HCV genotype 1 was the most frequently observed (55%), and mean HCV-VL 2x106 UI/ml (range, 3x103-15x106). Hepatitis activity was mild in 29%, moderate in 58% and severe in 12% cases. LB showed a lack of fibrosis in 6%, F I in 38%, F II in 18%, F III in 26%, and cirrhosis in 12%. By univariate analysis, a GGT plasmatic level >100 UI/ml predicted a severe HAI (OR 6,9; 95% CI [1,6-29,6], p=0,007). This was also confirmed by multivariate analysis performed including potential confounding factors