HEPATOLOGY Vol. 22, No. 4, Pt. 2, 1995 AASLD ABSTRACTS 341A

937 THE SPONTANEOUS FLUCTUATION OF VIREMIA OVER TIME IN UNTREATED PATIENTS WITH CHRONIC TYPE C HEPATITIS. SC Gordon. BA Khan. CF Dmuchowski. VP Kodali. AL Silverman. Division of Gastroenterology-Hepatology, William Beaumont Hospital, Royal Oak, MI

The natural history of chronic hepatitis C remains undefined. It is also not known whether significant changes in viremle levels occur over time. Serial serum HCV RNA levels were measured in 18 patients with chronic type C hepatitis. 16/! 8 patients had at least 3 samples obtained and frozen according to standard protocol. No patient was treated with antiviral agents. HCV RNA was detected quantitatively by bDNA (Quantiplex", ChironL Baseline HCV RNA levels (genomic eq/ml X 10 s) ranged from 3.5 to 117.5 with a median of 36.5 (mean, 44.1; S.D. =37.7). Follow-up times ranged from 15,4 to 55.4 months (mean, 33 months, S.D.=411. Levels at last follow-up ranged from 3.5 to 130.5 median, 37.6 (mean, 46.1; S.D. = 40.2). Eleven of ! 8 patients (61.1 %) indicated an increase in follow-up HCV RNA level from base ine (95% C.I. = 36% - 83%). Two patients with consistently low HCV RNA levels over time (<3.5), both PCR positive, did not change from baseline. The change from baseline HCV RNA level to last follow-up ranged from a decrease of 60 to an increase of 52 with a median increase of 5.4 (mean increase, 2.7, S.D.=32; paired t-test, p = .79). The HCV RNA parallel plot is given below, showing baseline to last

o . . . . . . . . . . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 10 lS 2O 2S 3O 2O 4O 4S 50

Monlw Fr0m ~ Controlling for time (p = .366) in a two variable muRivariate linear model, baseline HCV RNA levels alone significantly predict follow-up viral levels (p= .002). Baseline HCV RNA levels are positively correlated with follow-up levels measured by Pearson's r=.68, p=.002 at 1 D.F. Conclusionsi 11 HCV RNA levels remain relatively constant over time. 2) The possibility that most patients with chronic HCV display slight increases in viremia from baseline overtime cannot be excluded. 3) The baseline HCV RNA level is a significant predictor of follow- up HCV RNA levels.

recorded value. ,n

~2o

938 S P O N T A N E O U S F L U C T U A T I O N S O F V I R A L L O A D IN U N T R E A T E D PATIENTS W I T H C H R O N I C H E P A T I T I S C. M. Bourli~re *°, Ph Halfon**°,P Berthezene***, H Khiri**, l Portal °, C Wartelle °, V. Gerolami, G Cartouzou °, AP Gauthier. Liver unit, H6pital Saint Joseph*, H6pital de la Conception ° Alpha Bit**, INSERM U 315, Marseille, France. HCV viraemia is an established predictive factor of response to IFN therapy. Aim; To prospectively study fluctuations of HCV RNA level in untreated patients with chronic hepatitis C. Methods: Serum was collected from 6 patients daily for 7 days, weekly for 2 weeks and monthly for 2 months. All patients had biopsy proven chronic hepatitis with a Knodell score ranging from 2 to 7, Genotypes were as follows: lb (3 pts), 2a (2 pts) and la (1 pt). ALT was determined and HCV RNA was measured using both quantitative HCV RNA branched (bDNa) amplification assay( Chiron) and quantitative PCR (Amplicor HCV Monitor, Roche) on each serum. Resul ts : (1) There was no significant correlation between HCV RNA level and ALT values. (2) HCV RNA assessed by bDNA and Monitor may be related by the formula :

bDNA= 5.12x Monitor+l.06. ( r= 0.66 p<0.001 ;Pearson).

HCV RNA {bDNA) (10 6/ml) 6.51±3.9 3.13+1.26 1.58-1~0.7 6.69+1.98 2.09-+0.82 2.54_+1.05 Range 0.7-11 9 1.7- 5.3 0.6-3.06 2 .7-9 .6 1.3-3.8 0.8-4.8 Range of change (%) -891184 -45/170 -60/193 -591144 -35/184 -67/189 HCV RNA (Monitor) (1.06/ml) 0.82±0.5 0.61i'O.16 0.18"1"0.11 0.66!-0.23 0.41±0.39 0.4±0.22 Range 0.4-1.7 03-0.9 0.0%0.35 0.46-L2 0.13-1.5 0.13-0.78 Range of change (%) - 50/206 -47/153 -611194 -301181 -681360 - 66 / 194

Values shown as mean+ SD (3) Spontaneous fluctuations of viral load exists in each patients whatever the methods used, with a 2.6 to 17 fold change throughout the study. Conclusions: (1)HCV RNA levels are not related with ALT values. (2)A correlation exists between HCV RNA assessed by bDNA and Monitor but the values are differents.(3) Significant fluctuatiorLs o f viral load are observed when measured dally, weekly and monthly in untreated chronic hepatitis C patients. This raises the issue wether a single assessment of pre-therapeutic viraetoia can be considered a reliable tool in determining IFN posology.

939 ~ c x o, ASSOCIATION BETWEEN TYPE OF HEPATITIS C VIRUS, SERUM LOAD AND SEVERITY OF LIVER DISEASE R.Romeo r M. Colombo t M.G. Rumi. R. Soffredini. E. Del Ninno, M.F. Donato. A. Ruseo* and P. Simmonde**. Institute of Internal Medicine, University of Milan, Italy; *Epidemiology Unit, Aviano Cancer Center, Aviano, Italy; **Department of Medical Microbiology, University of Edinburgh, United Kingdom.

Chronic infection with hepatitis C virus (HCV) may lead to a variety of hepatic lesions from benign inflammation to liver cancer, but the relationships between infection and development of liver disease are poorly understood. To assess whether virus type and load are of pathogenetic importance, 197 Italian carr£ers with various hepatic lesions were investigated consecutive!y. They were 129 males (x age 50 yr, 21-77) anti'HCV'positive (EIA2) with biopsy ?roven chronic liver disease: 36 chronic persistent hepat~tls (CPH), 29 mild chronic active hepatitis (CAH), 73 CAH, 32 active cirrhosis and 27 HCC. HCV-RNA was detected by nested PCR (5" UTR primers) and quantitated by branched DNA assay (b-DNA, Chiron Co). HCV eerotypee were detected using an ELISA assay based on branched peptidee corresponding to the hypervariable antigenic regions of the NS-4 protein of HCV. HCV genotypes were detected by reverse hybridization assay (Inno-Lipa HCV, Immunogenetics). 187 (95%) patients had serum HCV-RNA by reverse transcribed polymerase ~hain reaction (RT-PCR) with a median level of 1,003x10 genomio equivalents/ml according to a branched-DNA assay (b-DNA). One hundred and seven patients(54%) had serotype i, 22 (11%) had serotype 2, 9 (5%) had seroty~e 3, 17 (9%) had mixed serotypes and 42 (21%) hadno speolfied serotype. One hundred andthirty four ~atientu were also tested for genotype. The genotype distrlbution was as follows: 17 (13%) ha~genotype la;~7 (50%) lb; 29 (22%) 2a; 12 (9%} 3a; 3 (2%) had genotype 1 not classified (NC); 3 (2%) had2NC; 2 (1,4%)had genotype 4 and 1 (1%) had mixed genotype la+3a. No virus was associated with any particular hietologic d~agnosis and all were equally distributed between groups with progreeeive and non-progressive liver disease groups. Serum HCV-RNA levels were similar in the liver diseases groups. In analogy with hepatitis B, there was a lack of direct correlation between type and level of viremia and the severity of the underlying liver damage.

940 WIDE FLUCTUATION OF PLASMATIC HCV GENOME LEVELS IN SERIAL SAMPLES OF PATIENTS BEFORE INTERFERON (IFN) TREATMENT. F Gioetra. R Francesconi, P Groff, *A Manzin, *L Solforosi, F Lari, G Ballardini, M Lenzi. *M Clementi. FB Bianchi. Medicine Interne II, Universit& di Bologna. Istituto di Virologia; Universit~ di Ancona. Italy.

While data are growing on HCV genome level as a possible clue to response to IFN treatment, no definite informations are available on its variations in serial samples of the same patient.

For this purpose we studied 9 consecutive HCV RNA positive patients, 9 of which with monthly HCV plasma level determianations for 6 months, and one with weekly sampling for 6 weeks using a novel competitive RT-PCR. ALT value was determined at the time of each sampling. Intrinsic variation of this method has been demonstrated to be lower than 20%. Maximal/minimal viremia ratio in each patient, as a determinant of oscillation width, was evaluated. Results are summarized in the table: Pte HCV molecules# Max/min 1 4564 (175-12170) 69.5 2 676 (61-7884) 128 3 4159 (2799-7093) 2.5 4 1162 (10-2874) 267 5 928 (472-2674) 5.6 6 409 (193-883) 4.5 7 810 (399-2136) 5.4 8 2775 (2061-29492) 4.6 9* 2 9 8 3 (1285-3771) 14 # xlO 3 median (range) * weekly sampling.

These data show a wide fluctuation of the number of the HCV molecules in several samples of the same patient on a monthly base; this pattern is maintained, even in lower proportion, employing a weekly follow up. No linear correlation could be determined between ALT and viremia levels,

It has been reported that low viremic levels are better correlated to response to IFN treatment. Our data suggest that a single HCV genome quantitation may be inadeguate as a predictor of IFN treatment outcome. Patients with high initial determinations could take advantage of a longer follow up for starting therapy at lower viremic levels.