METRO OCTAVO
1 Ver, por ejemplo, Séneca: De vita beata
2.4.1 Introduction
T he population exam ined in this study was infected in 1977 w ith genotype lb from a single source, and was not diagnosed until 1994. Therefore, this cohort had 17 years duration o f disease prior to the initiation o f any therapies. D espite this, it has been dem onstrated that the viral clearance rate was higher than to be expected for genotype lb and that the severity o f liver inflam m ation and fibrosis was less than expected for both genotype lb and duration o f infection [Sheehan et al, 1997; K enny-W alsh et al, 1999]. T his leads to the assum ption that in this specific cohort, host factors m ust play a significant role in d etennining outcom e o f hepatitis C infection.
2.4.2 Class 11 alleles in HCV infection
MHC class I and II m olecules present peptide fragm ents o f protein antigens to CDS and CD4 lym phocytes respectively, leading to their activation. In acute infection w ith hepatitis C, viral clearance is associated with a potent lym phocyte response to hepatitis C antigens [D iepolder et al, 1995; M issale et al, 1996]. In subjects w ho do not clear the virus and develop chronic liver disease, intrahepatic and peripheral C D 4 lym phocyte responses to hepatitis C peptides, especially core, have been shown to be m ore frequent and m ore vigorous; this is associated with a predom inant T helper cell 1 -like cytokine response [Napoli et al, 1996; M inutello et al, 1 9 9 3 ;L o h re t al, 1996]. T herefore, H LA class II restriction o f the im mune response, as an inherited host factor controlling peptide recognition and hence potentially the potency o f the im m une response, m ay in part determ ine outcom e in hepatitis C infection.
2.4.3 Unique study population
C om parison o f the broad DR phenotype frequencies in our subject population com pared w ith a population o f healthy unrelated blood donors revealed no differences. T his suggests that the subject population is derived norm ally from the overall population. T he 2 study groups w ere age and sex m atched and had no confounding variables contributing to outcom e. T he study group was exposed to a single HCV, genotype lb . It is an unselected population derived from that described by W alsh et al [K enny-W alsh et al, 1999] and as such has a low rate o f progression o f liver disease and a high rate o f norm al biochem ical m arkers o f liver disease in the presence o f biopsy-proven chronic active hepatitis.
2.4.4 H L A -D R B l, -D Q B l associations w ith viral outcom e
W e have deinonstraled that D R B l *0101 and D Q B l *0501 are significantly associated with transient infection and D R B l *03011, and D Q B l *0201 w ith persistent infection in our population. The allele D R B l*0401 occurs m ore frequently in those with transient infection and D R B l *1301, D Q B l *0202 and D Q B l *0603 occur m ore frequently in persistent infection, although not significantly after correction for the num ber o f alleles tested.
In the Irish population, DRB 1*0101 -D Q B 1 *0501 and DRB 1*03011 -D Q B 1 *0201 are known to be inherited in strong linkage disequilibrium . T he frequencies o f these haplotypes
segregate to disease outcom e as do the individual alleles and we are unable to report which m em ber o f the haplotype contributes more significantly to outcom e.
The role o f the M HC in HCV infection may not be lim ited to antigen presentation by class II alleles. O ther genetic loci carried on a haplotype m ay be responsible for viral clearance either independently or due to an interaction with class II alleles. To elucidate this further studies exam ining the M HC are required.Indeed, 2 studies have show n such an association with the TA P genes located w ithin the class II region o f the MHC and biochem ical and histological m arkers o f liver disease [K uzushita et al, 1999; Asti et al, 1999].
2.4.5 A ntigen specificity
A lthough the association o f DRB 1*0401 w ith viral clearance is not statistically significant after correction for the num ber o f alleles tested, this association may in fact be biologically significant. It is know n that P-chain o f D R l and DR4 share am ino acids 67-74 and this is located in the antigen-binding cleft o f the P4 pocket [Stem et al, 1994; D essen et al, 1997]. In R heum atoid A rthritis, a D R l - and D R 4-associated disease, the autoantigenic peptide that binds this shared region has recently been described as residues 261-273 o f collagen II [Fugger et al, 1996; R osloniec et al, 1997; A ndersson et al, 1998]. In our subject population, D R B l * 0101 and / o r *0401 alleles are present in 55% o f those who sustain viral clearance, com pared w ith only 23% o f chronically infected subjects (P < 0.0001). Hence it is feasible that both o f these alleles are selected for in the hepatitis C-negative population by im m une responses to a com m on im m unodom inant antigen. Shirai et al [Shirai et al, 1999] have dem onstrated that T cell recognition o f the H V R l region, a m ajor neutralising antibody epitope, o f different HCV isolates is C D 4+m ediated and possibly restricted by D R l and DR4 class II antigens. W e propose that in hepatitis C, D R B l *0101 and D R B l *0401- restricted T-cell recognition o f an im m unodom inant hepatitis C peptide may direct the T-cell response to favour viral clearance.
2.4.6 C om parison w ith other H LA class II studies
Barrett et al [Barrett et al, 1999] have recently published an association betw een DRB1*01 and spontaneous viral clearance in Irish fem ales infected w ith m ultiple hepatitis C
genotypes through anti-D Ig. The authors failed to identify further associations betw een viral and histological outcom es and D R B l alleles, as in our study. T his m ay be explained by several differences betw een the 2 studies. Barrett et al perform ed a lim ited class II analysis o f the D R B l locus only (n=14), in a sm aller population infected w ith m ixed HCV genotypes at m ultiple tim e points, including subjects w ith co-m orbid high alcohol intake. Tw o studies to date have shown an association betw een D Q B l *0301 and transient infection [Alric et al, 1997, Cram p et al, 1998]. C ram p et al conclude that D Q B l*0301 has a greater bearing on disease outcom e because D Q B l *0301 has been associated with transient infection in tw o studies involving different populations and in w hich D Q B l *0301 exists in linkage w ith different D R B l alleles. O verall, the frequency o f this allele is com parable in the three populations: 0.309 in Irish, 0.336 in B ritish and 0.337 in French. H ow ever, we did not dem onstrate any association with disease outcom e: D Q B l *0301 w as present in 27.7% with persistent infection and 35.8% w hh spontaneous clearance (P = 0.18). In the British and French studies, the D Q B l *0301 allele was present in 53.1% and 84% o f those with transient infection com pared with I 8.2% and 30.8% , respectively, o f those with viral persistence. In the French population study an association betw een clearance and
D R B 1*1101, which is in linkage with D Q B l *0301, was identified. A lthough D R B 1*1101 occurs in linkage with D Q B l *0301, the lack o f association with disease outcom e in our study may be explained by the low er frequency o f D R B l *11 in the Irish (0.084) than French (0.252) populations. In the B ritish population study, D R B l *04 was associated with transient infection and it occurred in strong linkage disequilibrium w ith D Q B l *0301. In our population both D R B l *04 and a subtype D R B l *0401 appeared to occur m ore frequently in transient infection. At low er resolution, D Q B l *03, w hich is linked to D R B l *0401, did occur m ore frequently in those subjects w ith transient infection in this study (60.4% versus 42.2% , Pc= 0.23) but the frequency o f D Q B l *0301 (0.309) w as low er and D Q B l *0302 (0.177) was higher than in the B ritish population studied (0.346 and 0.125 respectively). A lthough these differences are not large, they m ight be sufficient to sw ay the im portance o f the D Q B l *0301 allele w ith respect to clinical outcom e in a population study. A nother possible reason for the discrepancy betw een these studies lies in the viral genotype. The F rench and G en n an (H ohler et al, 1997) studies include 4 genotypes and the English study fails to give detail on the genotype(s) included, and our study exam ined a population e.xposed exclusively to genotype lb.
2.4.7
Conclusion
T he binding o f antigenic peptides to M HC class II m olecules is dependent on am ino acid sequences interacting specifically with residues in the peptide binding groove. T he various genotypes o f hepatitis C m ay alter the peptides presented, w ith subsequent changes in configuration and thus differences in their ability to bind the various MHC antigens. T herefore, differences in individual studies exam ining M HC association with outcom e in the highly variable HCV w ould be expected to depend not only on the population
characteristics o f the study but also on the viral sequences involved. This study differs from others in that exposure to a single inoculum , at a single tim e point, lim iting viral diversity, perm its a m ore rigorous assessm ent o f the role o f the M HC in outcom e.
In conclusion, w e show the associations o f several class II antigens with clinical outcom e in hepatitis C in a single genotype cohort. B ased on this and other studies w e propose that outcom e in hepatitis C is dependent on the nature o f the T cell response to the virus and that host im m unogenetic factors contained w ithin the MHC greatly influence this response.