Smouldering
hepatitis B virus replication in patients with chronic liver disease and hepatitis delta virus superinfection
Hepatitis B wrus deoxyribonucleic acid (HE\‘-DNA) was studied by Southern blat analysis in liver biopsy specimens from 75 HBsAp-positive patients with chronic liver disease livinc in southerr. Italv. Twentv-seven of the oatients were henatitis delta virus (HDV) kperinfected. Intrahepatic HBV-DNAks detected in 54 (72%) patients, 32 (59%) of them with replinative forms. The presence of replicative forms was directly related to liver HBcAg and inversely related to liver HDAg, as shown by multwxiate analysis. However. 14 patients with intrahepatic HBV-DNA non-replicalive pattern and about half of HDV-infecicd patients were liver IJBcAg and/or serum HBV-DNA positive, mostly ic ‘wv amount: Histnhogical inflammatory actwity was strongly related to liver HBcAg expression wgardkss of HDV superinfection, as confirmed by mutt~variate analysis. Our resclts confnm previous studies about the concordance between intrahepatic HBV-DNA replicative partern and liver HBcAg expression and about inhibition by HDV of high-level HBV replication. However, they suggest that low-ieve tiBV replication may have an important role in causing liver damage also among HDV-infected patients. in a population
Hepatitis
where tt.e spreading
B virus (HBV)-re!ated
are common in soutbem
of HBV and HDV is a naturally occurring event.
chronir liver diseases
Italy, accounting
for about 20%
casts some doubts on the direct cytopathogenicity of HDV. Thus, the relative role of the two rixs.es :n the
of all chronic liver disease in hospitals (I). Hepatitis D vi. rus (HDV) superinfection is present in about one-third of these patients (2). Chronic HBV carriers may show normal liver morphology despite ongoing virus replication in liver cells. Thus, HBV is believed tJ have no direct cytopathic action. Immwx response to HBcAg probably plays
pathogenesis of liver cell necrosis awing chronic HDVIHBV infection is still w&x. Intrahepatic HBV-DNA ~equenccs have been extensively studied by Southern blot analys;s (12-16). In patients with chronic liver disease it has been shown that the typical intrahepatic HBV-DNA replicative pattern is CDT.
a major role in causing liver damage (3-S). Since liver disease is almost always present in HDV-infected subjects
related to the preseoce of serum HBV-DNA, liver HBcAg and tissue inflammatory activity (14-16). There are limited data about the influence of HDV superinfection on liver HBV genamr pattern and on the severity of underlying liver disease cxrently available (16). Tbe aim of our work was to assess the replicative status of HBV in HBsAg-positive patieuts with HDWHBV-re-
(6.7). HDV is considered to be directly cytopathic. This latter hypothesis is atsa supported by morphological (8,9) and immunological evidence (10). However. the recent demonstration of early HDAg expression after liver transplnntarion without evidence of liver damage (11) -Corr77andwr
Dr SilViOF&grin. Divirionc dl Medrina tnrerna. apedale V Cerveuo. VinTrabucco ,*o, 1.90146r%,ermo, tta,y
65
lated chronic iiver disease, as compared 10 ‘pure’ HBVrelated chronic liver disease, by studying the intrahepatic HBV genome pattern and its relationship to conventional serumand liver markers of HBV and HDV infection, We also evalualed the influence of HBV and HDV on the severity of liver disease.
replication
nitj was preservedior HBv-DN.4 analysis.Tissue HBcAg and HDAg were evaluated by immu,,a”,,orescence on 4 wn frozen liver sections fixed in chIors:orm/ acetcme (5050. v/v) using a mouse monoclonal anti_HBc and P human polyclonal FITC-conjugated anti-HD (Cam-
20-30
bridge Biomedical,
U.K.).
Liver HBV-DNA MaterlaIr
and Metbods
Cellular DNA
was purified from biopsies after tissue ly-
sis in a buffer containing
100 mM NaCI.
50 mM Trir (pH
P&m We studied 75 chronic HBsAg carriers (54 males; mean age 32.2 yrs, range 5-72) who had undergone liver biop sv. All subiecfs had had abnormal AST values (more than twice the upper normal limit) during the previous 6 months. None of them admitted significant alcohol intake (>50 g/day). Twenty-seven patients bad chronic HDV supcrinfection (Le., serum anti-HD positive for at least 6 months and liver HDAg in 5-RO% of all hepatocytes). Four patients were anti-HIV positive (two of them drug addicts and two homosexuals). None had ever received antiviral or immunosuppressive treatments. Htstological diagnoses were: (1) chronic persistent hepatitis (9 p’s); (2) chronic lobular hetmtiiis (2 ots): (3) mild chronic active hemdills (6 pts); (4j se& chro% &tiw hwatitis + ctrrhoiis (40
8.0),
pts); and (5) inactive cirrhosis (18 pts). Diagnosis 1. 2 and
Data acre analyzed using tlve j-test with Yates’ correction and. when appropriete, exact Fisher iest. for proportional comparison. Yariables :ignifIcan: ivah uniyariate analysis were tested by multivariate step-dcwn anal-
3 were pooled and identified as ‘mild disease’, and dingnoses 4 and 5 as .severe disease’. Ail patents had compensated iiver disease, i.e., no liver failure, asches, gastrointestinal haemorrhage, encephaiopathy, hepatocelluiar carcinoma.
Sera, taken at the time of liver biopsy. were tested for
HBsAg,
HBeAg
by RIA (Abbott Laboratories, Chicago, IL, U.S.A.). Anti-HD and anti-HIV were also analysed by RIA (Sarin. Salu&~, Italy). Anti-
IO mM EDTA,
ase K. followed
0.5% SDS and 2OOpglml oiprolein-
by incubation
for 12 h at 37 “C. Several
extractions were lhen performed
with phcnolichloroform
(1: 1, WV) saturated with TE buffer (10 mM Trls-HCI
(pH
and then with chloroform. The aqueous phase was digested with 50 @ml cf D&se-free RNAse for 1 h at 37 “C, rpcxtrarted as above, prccipiratcd with 2 YOIPof ethanol and resuspended m TE buffer. Aliquots of 20rg of purified liver DNA were digested with EcoRI restrtction endcnucleese under the conditions recommended by the supplier, electraphorescd through 1% agarose gel at 30 V for I2 h, then transferred to nylon membrane according to Southern (20).
8.0).
1 mM
EDTA)
ysts. All data were processed by TABSUR and AMS packages implemented an the IBM 370 computer.
Results
and anti-HBe
HIV positive sera were confirmed by ilnmwtcblotting (Du Pant De Nemours, Cologne Monzese. Italy). HBVDNA was detected in serum by a molecular hybridization assay, as described previously (17). The results were cxpressed by comparison of the standard signal intensity. The lowest amount detectable ~~0.5 pe/& Himpathology Blind liver biopsies taken with a 1.6 mm Menghini needle were processedby standard techniques. Histologic diagnosis was established b.u pre-set criteria (lS,ll). fragments of the b,opsy mre were immediately
Two
snap-f:o
ren in liquid nitrogen-cooled isaprnthane. One WPSem-
bedded in O.C.T. (Miles Laboratories, Naperville, IL, U.S.A.) for cryostat sectioning. while the other (about
Serum HBV-DNA
was cbseiy lmked to liver H&kg
regardless of HDV superinfection (Table I). Among HDV-patients 34/48 (71%) were liver HBcAp-posdive
ABCDEF
(ns wtlt less than 10% positive hepatocytes) and 37148 (77SiI were strum HBV-DNA vositive (nine with less
1
than p&,1). Among HDV.posit/vcpatients 11/27(41%) were liver HBcAg positive (five with less than 10% positive hcpntwytcs) nnd 14/27 (52%) were serum HBVDNA positive (six with lcssthcn 1 p&l) (Table 1). Liver HBV.DNA sequencer. as assessed by Southern blot analysis. were detectable in one of the patterns shown in Fig, 1. Thirty-two pdtients had bands betwern 3.5 and 3.2 kb and a smear below (lanes C and E). six (one anti-HIV positive) also showing integrated forms. These
G
212 b
patterns were considered replicative forms of HBVDNA. Twenty-two patients had a band at 3.2 kb withou! any smear (lane F). two also showing integrated forms (lane A), and three others (two anti.HIV positive) a hand at about 2.5-2.8 kh (lane B). These patterns were considered as non-replicative forms of HBV-DNA. In 21 patients (one anti-HIV positive) HBV-DNA was undetectable (lane D). lntrahepatic HBV-DNA was closely related to HBcAg expression. Tbc majority of the patients with rcplicative forms were HBcAg-positive. while most patients with non-rcplicative forms or with undetectable HBV-DNA were HBcAg-negative (Table 2) However, seven patients with noareplicative fonm (sli serum HBV-DNA positive. four with leas then 1 p&l and ih;er HDV positive) were liver HBcAg positive. Thrcr patients had the patwm shown in Fig,. 1, lane B, and HBcAg in more than 10% of hepatocytes. Four patients had the pattern shown in Fig. 1, lane F. and HBcAg in !PCS!hzn 10% of hepatccytcs. In the other scvcn patients (six HBV-DNA posilive. three with Icu than I p&l and 3 HDV positive), intrahem& HBV-DNA was undetectable. while HBcAe was present in the liver. although in small amounts (six out of seven with less than 10% positivc hepatosytes). Active HDV replication. as rsyrjred by liver HDAg, wns inversely related to intrahepatic HBV replication. In
sis confirmed that lfiveerHBcAg and HDAg were the vatiables best related to intraheparic HBV-DNA replicative
fxt. only 5132 (16%) patients with HBV-DNA replicative forms were HDAg-positive (Table 3). Multivariate anaty-
logic activity) evaluated by univariate analysis. Five of these (age. HBeAg, serum HBV-DNA, HBcAg and
forms among the eight items (sex, age, HBeAg, serum HBV-DNA, HBcAg, HDAg, disease severity and histo-
67
HDAg)
were significant
with univariate
analysis
(Table
4). Subjects without HDV superinfection
had mild (14148.
29%) or severe (34/48. 71%) disease, and intrahcpatic HBV-DNA replicative forms were frequently found in both groups (43 and 62%. respectively) (Fig. 2). In HDVinfected subjects liver disease was almost always scvsrc (24I27, 89%) and rarely asswiated with intrahepatic HBV replication (5127, 18%) (Fig. 2). Histologic inflammatory activity was only significantly related to liver HBcApl (Table 5A). The exclusion of patients with mild disease from the analysis did not moany this latter finding (Table 58). Moreowr, the correlation between liver HBcAg, and histologic activity among patients with severe chronic liver disease. was confimnrd by multivariate analysis of the five vanables (sex, age, HBeAg, serum HBV-DNA, inrrahepatic HBV-DNA and histological activity) which were signifacanrly related to HBcAg by univariate analvsis. Histologic activity. was found to be one of the three best predictors pression (Table 6).
HDV-
of HBcAg ex-
WI”+
DIscussfan We found that intrahepatic HBV-DNA replicawe forms are strlctty related to bver HBcAg expression. However, I4 patients, mostly with low amounts of HBcAg. showed episomal forms or undetectable HBVDNA. The cmall amounts of serum HBV-DNA found in all but one of these 14 patients confirm the prrsence of low-level HBV replication in these subjects and show that liver HBcAg and serum HBV-DNA still remain. in our opinion. the most sensitive and reliable markers ir! assessicg HBV replication in patients with chronic liver disease. Tbe presence of both liver HBcAg and serum HBVDNA, without the typical intrabepatic replicative forms, may be due to sampling error in detecting HBV replication nr to different sensitivity in the methais used for detection of the three parameters of HBV-DNA replication. HDV supennfection
was found to he inversely related
P COLOMBO
68
tu iotrnhcpatic HBV replicanon, BS expected (21-25). However. 41% of HDV-related chronic liver diseasepaticns showed HBcAg in the liver and 52% HBV-DNA in the serum, altbough the malority in small rmtounts. This suggeststhat HDV can diminish HBV replication. often mnking the detection of HBV.DNA repticative intcrmediatcs in ibr liver impossible. A consistentproportion of HDV supcnnfectea p&scts, however. stilt show signsof ongoing HBV replication (i.e !iv-r HBcAgandlorserum HBV.DNA). We performed on:y a cross+%tionsl study, but b is possible that during the follow-up some patients who were negative for liver HBcAg and serum HBVDNA might become positive. Certainty. the finding that about half of tbz HDV-CLD patients have at least one of the known signs of active HBV replication raises some doubts concerning the widely scccpted nssump:ion that HBV does not generally replicate in HDV-infected pa-
,
tients. Only four out of the 75 patients studied were HIV infected. None of the other patients had B causefor immunosuppression.known to facilitate the active replication of both HB;’ and HDV viruses (26). As a result, the above finding appears to be a naturally occurring event. This migbi br pariisuiarly importam since the subjens studled are representativeof a population of patients with HBsAg positive chronic liver disease. In fact, in southern Italy HBV and HDV infections spread for the most pan through intrsfamitial contact within the first decade of life (27). In northern Europe and the U.S.A., however, most sobjezts become infected during their adult life due to a lifestyle (drug addiction, homosexuality) associatedwith high risk ior acquiring HIV infection. The high proportion of HDV-superinfected patients
et al.
with concomitant active HBV replication makes the recognition of the relative role of the two viruses in causing liver damage difficult. However, if we assumethat ongoing liver damage is reflected by inflammatory histological activity, liver HBcAg seems the strongest predictor of liver damage. It is noteworthy that the correlation between liver HBcAg and histological activity increases if we exclude patients with mild disease (chronic persistent hepatitis or chronic lobular hepatitis) from analysis. For this reason it is unlikely to be a casual finding. Liver HBcAg correlated to histological activity more ctoscty than intrahepatic HBV replication as a result of the prcsence of patients without HBV-DNA replicative forms, but with small amounts of HBcAg in the liver, and with inflammatory liver cell damage. Thus, in HDV-infected subjectsatso, tow-level HBV replication might frequently cause liver cell damage, as it does in HDV-negative patients. This evidence should encourage new therapeutic approachesin the future, since preliminary investigations suggest a worse prognosis in the subgroup of HDV-infccted subjects witb concomitant HBV replication (28). Further studies, including a larger population of patients and their fallow-up, are necessaryto confirm the importance of the role of ‘smoutdering’ HBV replication in HDV-related chronic active disease.
Acknowlcdgcments This work was supported by a grant from Assessorato alla Ssnitd della Regione Sicilia (No. 75/M), and by funds from the Ministero delta Pubblica Istruzione.
6 Rlzzctto M, Verme 0, C&in GL, et al. Hepatitkdeltavirusdisease.Prog Liver Dis 1986;8: 417-X
7 Arka s, AragonaM. ltiuetro M, et at. ClinicatSignificancE ot antibodyto hepatitisdeltaYinISin symptomtess HBr,Q carriers. Lancet1985;ir356-8. Popper H. Pathotaglcd observation Cm dells antigen inkctian. In: “yar GN, oienrrag IL. Hwbagle JH, 4s. viral andLiverDisease.Orlando:GnmeandStratton,t9B4:381-4. 9 popperH, PurcellRH, GermGL. HisIotogyottlepa.dna andde,. la virus-induced hepatitisand hepatoeettulnr carcinoma.In: RobinsonW, Koikc K. Wilt H. eds. HepadnaVimras. New York Alan R. Liss.1987;373-86. 10 Magrins, craxt A, Caini c, et at. Interleukin-2,imerteukia-: receptorand gamma-interferon syntbcri.by peripberatblood mononuclear cellsin chronicbep~ns deltavirusinfection.J Hep&i ist+.,1I: a: 35g-66.
8
HqmhiP
HBV REPLICAT[ON
IN HDV-RELATED
69
CLD
13 Pontisso P, Bortolott~ F, Fattovich G, et al. Hepatitis B virus DNA lormr m the liver of chronically infec!ed individua!s.J Heoatoli989: 9: 29-35. 14 bmata M. Yokoruka 0. lmveki ;j, CI.al. Correlation of hepatilis B virus DNA and antiwns in the liver. Gartroenteroloev 1987; 92: 192-6. 15 Ramalho F, Brunerto MR, Roccu G. ct at. Serum muikz-s of hepatitis B vitw replication, liver hixdogy and lnlrahepatrc cxpressionof B core antigen. J Hepatol 1988;7: 14-20. 16 VillarID. Raimondo G,Smedilc V.ct al. HcpatitisB-DNA replication and hiatologzcal ~&terns in liver bmpy spectmensof chronic HBsA~ ooririve ~atientr wilh and without henatitis delta
I.
hepatitis
18 19 20
21
don asray for the detection if hepatitis B virus DNA in wum. Factors determining iIs sensitivity and specificity Hepatolagy 1987;7: 557-62. Scheuer PI. Liver Biopsy Interpretation. London: Bailliere Tindall. 1980. Review by an In~ernatiollnl Group. Acuts and chwnrc hepstms rcvisitcd. Lancet 1977: ii: 914-9. Southern EM. Detection of specificxqucncer among DNA fragment~ seoarated bv Mel electroshowis. J Mol 0101 1975: 8% . 503-17. . Rizzetto M, Ponzetto A, Bon& F, et al. Superimposedhepatitis and lhc clfcct on vimt replicalion in chronic hepatitis B. J Hepa-
to1 19wsz: 35-41. 22 Covindarajan 9. Valinluck B. Serum beps!l!is B vima DNA tn chronic hepatitis B and delta infection. Arch Pathol Lab Med 1985: 1G9:x%9. 23 Krogsgaard K. Kntgg~ P, Alderphtile 1, et al. Delta infection and suppressionof HE virus replication in chronic HEsAg car. rierr. Henatology WV: 7: 42-5. 24 Gcoesa*I. Jardi 1, Bud M. exal. Hepativ, it ,iru rcplicaton m acute hcpabbs B. acute hcprtitn Bwzute delta vine minfecrion and acute hepanrn deba suupcrinfection.Hepalology ,981; 7: 572-96. 25 Chen P. Chen D. Chen C, et al. Delta infection in asymptomatic carriers of hepabbsB surlacc antigen:low orcvalcnccof delta activity and cficctrve suuppres&no? hcpatit& B vim%reptiration. Hepatology l9SS;L1:1121-4. 26 Cnxl 4, Di Marco Y, Ma@ S, et al. Hepatitis B virus-induced chronic liver direax:: features of hspmnia D and B virus rephcatioo. In: RobinsonW, Koike K, WiIIH. cds. HepndnnVirunes. New York: Alan R. Liss, 1987; @U-IO. 27 Bonino F, Capaso N, Dentico P, et al. Familiar clusteringand spreading of hepatitis delta virus infection. J Hepatol 1985: 1: 221-6. 28 Smedile A, Chiabcrge E, Piantino I’. et dl. Hcpatitts D virus (HDV)Meparirlr B virus interactions. u mnleeulur and wrologrc study(Ahrtr h Flepatol 1989:9: S85.
I
hepatitis B virus replication in patients with chronic liver disease and hepatitis delta virus superinfection
Hepatitis B wrus deoxyribonucleic acid (HE\‘-DNA) was studied by Southern blat analysis in liver biopsy specimens from 75 HBsAp-positive patients with chronic liver disease livinc in southerr. Italv. Twentv-seven of the oatients were henatitis delta virus (HDV) kperinfected. Intrahepatic HBV-DNAks detected in 54 (72%) patients, 32 (59%) of them with replinative forms. The presence of replicative forms was directly related to liver HBcAg and inversely related to liver HDAg, as shown by multwxiate analysis. However. 14 patients with intrahepatic HBV-DNA non-replicalive pattern and about half of HDV-infecicd patients were liver IJBcAg and/or serum HBV-DNA positive, mostly ic ‘wv amount: Histnhogical inflammatory actwity was strongly related to liver HBcAg expression wgardkss of HDV superinfection, as confirmed by mutt~variate analysis. Our resclts confnm previous studies about the concordance between intrahepatic HBV-DNA replicative partern and liver HBcAg expression and about inhibition by HDV of high-level HBV replication. However, they suggest that low-ieve tiBV replication may have an important role in causing liver damage also among HDV-infected patients. in a population
Hepatitis
where tt.e spreading
B virus (HBV)-re!ated
are common in soutbem
of HBV and HDV is a naturally occurring event.
chronir liver diseases
Italy, accounting
for about 20%
casts some doubts on the direct cytopathogenicity of HDV. Thus, the relative role of the two rixs.es :n the
of all chronic liver disease in hospitals (I). Hepatitis D vi. rus (HDV) superinfection is present in about one-third of these patients (2). Chronic HBV carriers may show normal liver morphology despite ongoing virus replication in liver cells. Thus, HBV is believed tJ have no direct cytopathic action. Immwx response to HBcAg probably plays
pathogenesis of liver cell necrosis awing chronic HDVIHBV infection is still w&x. Intrahepatic HBV-DNA ~equenccs have been extensively studied by Southern blot analys;s (12-16). In patients with chronic liver disease it has been shown that the typical intrahepatic HBV-DNA replicative pattern is CDT.
a major role in causing liver damage (3-S). Since liver disease is almost always present in HDV-infected subjects
related to the preseoce of serum HBV-DNA, liver HBcAg and tissue inflammatory activity (14-16). There are limited data about the influence of HDV superinfection on liver HBV genamr pattern and on the severity of underlying liver disease cxrently available (16). Tbe aim of our work was to assess the replicative status of HBV in HBsAg-positive patieuts with HDWHBV-re-
(6.7). HDV is considered to be directly cytopathic. This latter hypothesis is atsa supported by morphological (8,9) and immunological evidence (10). However. the recent demonstration of early HDAg expression after liver transplnntarion without evidence of liver damage (11) -Corr77andwr
Dr SilViOF&grin. Divirionc dl Medrina tnrerna. apedale V Cerveuo. VinTrabucco ,*o, 1.90146r%,ermo, tta,y
65
lated chronic iiver disease, as compared 10 ‘pure’ HBVrelated chronic liver disease, by studying the intrahepatic HBV genome pattern and its relationship to conventional serumand liver markers of HBV and HDV infection, We also evalualed the influence of HBV and HDV on the severity of liver disease.
replication
nitj was preservedior HBv-DN.4 analysis.Tissue HBcAg and HDAg were evaluated by immu,,a”,,orescence on 4 wn frozen liver sections fixed in chIors:orm/ acetcme (5050. v/v) using a mouse monoclonal anti_HBc and P human polyclonal FITC-conjugated anti-HD (Cam-
20-30
bridge Biomedical,
U.K.).
Liver HBV-DNA MaterlaIr
and Metbods
Cellular DNA
was purified from biopsies after tissue ly-
sis in a buffer containing
100 mM NaCI.
50 mM Trir (pH
P&m We studied 75 chronic HBsAg carriers (54 males; mean age 32.2 yrs, range 5-72) who had undergone liver biop sv. All subiecfs had had abnormal AST values (more than twice the upper normal limit) during the previous 6 months. None of them admitted significant alcohol intake (>50 g/day). Twenty-seven patients bad chronic HDV supcrinfection (Le., serum anti-HD positive for at least 6 months and liver HDAg in 5-RO% of all hepatocytes). Four patients were anti-HIV positive (two of them drug addicts and two homosexuals). None had ever received antiviral or immunosuppressive treatments. Htstological diagnoses were: (1) chronic persistent hepatitis (9 p’s); (2) chronic lobular hetmtiiis (2 ots): (3) mild chronic active hemdills (6 pts); (4j se& chro% &tiw hwatitis + ctrrhoiis (40
8.0),
pts); and (5) inactive cirrhosis (18 pts). Diagnosis 1. 2 and
Data acre analyzed using tlve j-test with Yates’ correction and. when appropriete, exact Fisher iest. for proportional comparison. Yariables :ignifIcan: ivah uniyariate analysis were tested by multivariate step-dcwn anal-
3 were pooled and identified as ‘mild disease’, and dingnoses 4 and 5 as .severe disease’. Ail patents had compensated iiver disease, i.e., no liver failure, asches, gastrointestinal haemorrhage, encephaiopathy, hepatocelluiar carcinoma.
Sera, taken at the time of liver biopsy. were tested for
HBsAg,
HBeAg
by RIA (Abbott Laboratories, Chicago, IL, U.S.A.). Anti-HD and anti-HIV were also analysed by RIA (Sarin. Salu&~, Italy). Anti-
IO mM EDTA,
ase K. followed
0.5% SDS and 2OOpglml oiprolein-
by incubation
for 12 h at 37 “C. Several
extractions were lhen performed
with phcnolichloroform
(1: 1, WV) saturated with TE buffer (10 mM Trls-HCI
(pH
and then with chloroform. The aqueous phase was digested with 50 @ml cf D&se-free RNAse for 1 h at 37 “C, rpcxtrarted as above, prccipiratcd with 2 YOIPof ethanol and resuspended m TE buffer. Aliquots of 20rg of purified liver DNA were digested with EcoRI restrtction endcnucleese under the conditions recommended by the supplier, electraphorescd through 1% agarose gel at 30 V for I2 h, then transferred to nylon membrane according to Southern (20).
8.0).
1 mM
EDTA)
ysts. All data were processed by TABSUR and AMS packages implemented an the IBM 370 computer.
Results
and anti-HBe
HIV positive sera were confirmed by ilnmwtcblotting (Du Pant De Nemours, Cologne Monzese. Italy). HBVDNA was detected in serum by a molecular hybridization assay, as described previously (17). The results were cxpressed by comparison of the standard signal intensity. The lowest amount detectable ~~0.5 pe/& Himpathology Blind liver biopsies taken with a 1.6 mm Menghini needle were processedby standard techniques. Histologic diagnosis was established b.u pre-set criteria (lS,ll). fragments of the b,opsy mre were immediately
Two
snap-f:o
ren in liquid nitrogen-cooled isaprnthane. One WPSem-
bedded in O.C.T. (Miles Laboratories, Naperville, IL, U.S.A.) for cryostat sectioning. while the other (about
Serum HBV-DNA
was cbseiy lmked to liver H&kg
regardless of HDV superinfection (Table I). Among HDV-patients 34/48 (71%) were liver HBcAp-posdive
ABCDEF
(ns wtlt less than 10% positive hepatocytes) and 37148 (77SiI were strum HBV-DNA vositive (nine with less
1
than p&,1). Among HDV.posit/vcpatients 11/27(41%) were liver HBcAg positive (five with less than 10% positive hcpntwytcs) nnd 14/27 (52%) were serum HBVDNA positive (six with lcssthcn 1 p&l) (Table 1). Liver HBV.DNA sequencer. as assessed by Southern blot analysis. were detectable in one of the patterns shown in Fig, 1. Thirty-two pdtients had bands betwern 3.5 and 3.2 kb and a smear below (lanes C and E). six (one anti-HIV positive) also showing integrated forms. These
G
212 b
patterns were considered replicative forms of HBVDNA. Twenty-two patients had a band at 3.2 kb withou! any smear (lane F). two also showing integrated forms (lane A), and three others (two anti.HIV positive) a hand at about 2.5-2.8 kh (lane B). These patterns were considered as non-replicative forms of HBV-DNA. In 21 patients (one anti-HIV positive) HBV-DNA was undetectable (lane D). lntrahepatic HBV-DNA was closely related to HBcAg expression. Tbc majority of the patients with rcplicative forms were HBcAg-positive. while most patients with non-rcplicative forms or with undetectable HBV-DNA were HBcAg-negative (Table 2) However, seven patients with noareplicative fonm (sli serum HBV-DNA positive. four with leas then 1 p&l and ih;er HDV positive) were liver HBcAg positive. Thrcr patients had the patwm shown in Fig,. 1, lane B, and HBcAg in more than 10% of hepatocytes. Four patients had the pattern shown in Fig. 1, lane F. and HBcAg in !PCS!hzn 10% of hepatccytcs. In the other scvcn patients (six HBV-DNA posilive. three with Icu than I p&l and 3 HDV positive), intrahem& HBV-DNA was undetectable. while HBcAe was present in the liver. although in small amounts (six out of seven with less than 10% positivc hepatosytes). Active HDV replication. as rsyrjred by liver HDAg, wns inversely related to intrahepatic HBV replication. In
sis confirmed that lfiveerHBcAg and HDAg were the vatiables best related to intraheparic HBV-DNA replicative
fxt. only 5132 (16%) patients with HBV-DNA replicative forms were HDAg-positive (Table 3). Multivariate anaty-
logic activity) evaluated by univariate analysis. Five of these (age. HBeAg, serum HBV-DNA, HBcAg and
forms among the eight items (sex, age, HBeAg, serum HBV-DNA, HBcAg, HDAg, disease severity and histo-
67
HDAg)
were significant
with univariate
analysis
(Table
4). Subjects without HDV superinfection
had mild (14148.
29%) or severe (34/48. 71%) disease, and intrahcpatic HBV-DNA replicative forms were frequently found in both groups (43 and 62%. respectively) (Fig. 2). In HDVinfected subjects liver disease was almost always scvsrc (24I27, 89%) and rarely asswiated with intrahepatic HBV replication (5127, 18%) (Fig. 2). Histologic inflammatory activity was only significantly related to liver HBcApl (Table 5A). The exclusion of patients with mild disease from the analysis did not moany this latter finding (Table 58). Moreowr, the correlation between liver HBcAg, and histologic activity among patients with severe chronic liver disease. was confimnrd by multivariate analysis of the five vanables (sex, age, HBeAg, serum HBV-DNA, inrrahepatic HBV-DNA and histological activity) which were signifacanrly related to HBcAg by univariate analvsis. Histologic activity. was found to be one of the three best predictors pression (Table 6).
HDV-
of HBcAg ex-
WI”+
DIscussfan We found that intrahepatic HBV-DNA replicawe forms are strlctty related to bver HBcAg expression. However, I4 patients, mostly with low amounts of HBcAg. showed episomal forms or undetectable HBVDNA. The cmall amounts of serum HBV-DNA found in all but one of these 14 patients confirm the prrsence of low-level HBV replication in these subjects and show that liver HBcAg and serum HBV-DNA still remain. in our opinion. the most sensitive and reliable markers ir! assessicg HBV replication in patients with chronic liver disease. Tbe presence of both liver HBcAg and serum HBVDNA, without the typical intrabepatic replicative forms, may be due to sampling error in detecting HBV replication nr to different sensitivity in the methais used for detection of the three parameters of HBV-DNA replication. HDV supennfection
was found to he inversely related
P COLOMBO
68
tu iotrnhcpatic HBV replicanon, BS expected (21-25). However. 41% of HDV-related chronic liver diseasepaticns showed HBcAg in the liver and 52% HBV-DNA in the serum, altbough the malority in small rmtounts. This suggeststhat HDV can diminish HBV replication. often mnking the detection of HBV.DNA repticative intcrmediatcs in ibr liver impossible. A consistentproportion of HDV supcnnfectea p&scts, however. stilt show signsof ongoing HBV replication (i.e !iv-r HBcAgandlorserum HBV.DNA). We performed on:y a cross+%tionsl study, but b is possible that during the follow-up some patients who were negative for liver HBcAg and serum HBVDNA might become positive. Certainty. the finding that about half of tbz HDV-CLD patients have at least one of the known signs of active HBV replication raises some doubts concerning the widely scccpted nssump:ion that HBV does not generally replicate in HDV-infected pa-
,
tients. Only four out of the 75 patients studied were HIV infected. None of the other patients had B causefor immunosuppression.known to facilitate the active replication of both HB;’ and HDV viruses (26). As a result, the above finding appears to be a naturally occurring event. This migbi br pariisuiarly importam since the subjens studled are representativeof a population of patients with HBsAg positive chronic liver disease. In fact, in southern Italy HBV and HDV infections spread for the most pan through intrsfamitial contact within the first decade of life (27). In northern Europe and the U.S.A., however, most sobjezts become infected during their adult life due to a lifestyle (drug addiction, homosexuality) associatedwith high risk ior acquiring HIV infection. The high proportion of HDV-superinfected patients
et al.
with concomitant active HBV replication makes the recognition of the relative role of the two viruses in causing liver damage difficult. However, if we assumethat ongoing liver damage is reflected by inflammatory histological activity, liver HBcAg seems the strongest predictor of liver damage. It is noteworthy that the correlation between liver HBcAg and histological activity increases if we exclude patients with mild disease (chronic persistent hepatitis or chronic lobular hepatitis) from analysis. For this reason it is unlikely to be a casual finding. Liver HBcAg correlated to histological activity more ctoscty than intrahepatic HBV replication as a result of the prcsence of patients without HBV-DNA replicative forms, but with small amounts of HBcAg in the liver, and with inflammatory liver cell damage. Thus, in HDV-infected subjectsatso, tow-level HBV replication might frequently cause liver cell damage, as it does in HDV-negative patients. This evidence should encourage new therapeutic approachesin the future, since preliminary investigations suggest a worse prognosis in the subgroup of HDV-infccted subjects witb concomitant HBV replication (28). Further studies, including a larger population of patients and their fallow-up, are necessaryto confirm the importance of the role of ‘smoutdering’ HBV replication in HDV-related chronic active disease.
Acknowlcdgcments This work was supported by a grant from Assessorato alla Ssnitd della Regione Sicilia (No. 75/M), and by funds from the Ministero delta Pubblica Istruzione.
6 Rlzzctto M, Verme 0, C&in GL, et al. Hepatitkdeltavirusdisease.Prog Liver Dis 1986;8: 417-X
7 Arka s, AragonaM. ltiuetro M, et at. ClinicatSignificancE ot antibodyto hepatitisdeltaYinISin symptomtess HBr,Q carriers. Lancet1985;ir356-8. Popper H. Pathotaglcd observation Cm dells antigen inkctian. In: “yar GN, oienrrag IL. Hwbagle JH, 4s. viral andLiverDisease.Orlando:GnmeandStratton,t9B4:381-4. 9 popperH, PurcellRH, GermGL. HisIotogyottlepa.dna andde,. la virus-induced hepatitisand hepatoeettulnr carcinoma.In: RobinsonW, Koikc K. Wilt H. eds. HepadnaVimras. New York Alan R. Liss.1987;373-86. 10 Magrins, craxt A, Caini c, et at. Interleukin-2,imerteukia-: receptorand gamma-interferon syntbcri.by peripberatblood mononuclear cellsin chronicbep~ns deltavirusinfection.J Hep&i ist+.,1I: a: 35g-66.
8
HqmhiP
HBV REPLICAT[ON
IN HDV-RELATED
69
CLD
13 Pontisso P, Bortolott~ F, Fattovich G, et al. Hepatitis B virus DNA lormr m the liver of chronically infec!ed individua!s.J Heoatoli989: 9: 29-35. 14 bmata M. Yokoruka 0. lmveki ;j, CI.al. Correlation of hepatilis B virus DNA and antiwns in the liver. Gartroenteroloev 1987; 92: 192-6. 15 Ramalho F, Brunerto MR, Roccu G. ct at. Serum muikz-s of hepatitis B vitw replication, liver hixdogy and lnlrahepatrc cxpressionof B core antigen. J Hepatol 1988;7: 14-20. 16 VillarID. Raimondo G,Smedilc V.ct al. HcpatitisB-DNA replication and hiatologzcal ~&terns in liver bmpy spectmensof chronic HBsA~ ooririve ~atientr wilh and without henatitis delta
I.
hepatitis
18 19 20
21
don asray for the detection if hepatitis B virus DNA in wum. Factors determining iIs sensitivity and specificity Hepatolagy 1987;7: 557-62. Scheuer PI. Liver Biopsy Interpretation. London: Bailliere Tindall. 1980. Review by an In~ernatiollnl Group. Acuts and chwnrc hepstms rcvisitcd. Lancet 1977: ii: 914-9. Southern EM. Detection of specificxqucncer among DNA fragment~ seoarated bv Mel electroshowis. J Mol 0101 1975: 8% . 503-17. . Rizzetto M, Ponzetto A, Bon& F, et al. Superimposedhepatitis and lhc clfcct on vimt replicalion in chronic hepatitis B. J Hepa-
to1 19wsz: 35-41. 22 Covindarajan 9. Valinluck B. Serum beps!l!is B vima DNA tn chronic hepatitis B and delta infection. Arch Pathol Lab Med 1985: 1G9:x%9. 23 Krogsgaard K. Kntgg~ P, Alderphtile 1, et al. Delta infection and suppressionof HE virus replication in chronic HEsAg car. rierr. Henatology WV: 7: 42-5. 24 Gcoesa*I. Jardi 1, Bud M. exal. Hepativ, it ,iru rcplicaton m acute hcpabbs B. acute hcprtitn Bwzute delta vine minfecrion and acute hepanrn deba suupcrinfection.Hepalology ,981; 7: 572-96. 25 Chen P. Chen D. Chen C, et al. Delta infection in asymptomatic carriers of hepabbsB surlacc antigen:low orcvalcnccof delta activity and cficctrve suuppres&no? hcpatit& B vim%reptiration. Hepatology l9SS;L1:1121-4. 26 Cnxl 4, Di Marco Y, Ma@ S, et al. Hepatitis B virus-induced chronic liver direax:: features of hspmnia D and B virus rephcatioo. In: RobinsonW, Koike K, WiIIH. cds. HepndnnVirunes. New York: Alan R. Liss, 1987; @U-IO. 27 Bonino F, Capaso N, Dentico P, et al. Familiar clusteringand spreading of hepatitis delta virus infection. J Hepatol 1985: 1: 221-6. 28 Smedile A, Chiabcrge E, Piantino I’. et dl. Hcpatitts D virus (HDV)Meparirlr B virus interactions. u mnleeulur and wrologrc study(Ahrtr h Flepatol 1989:9: S85.
I
