Hepatitis C virus (HCV) infects >4 million persons in the United States. In young children, perinatal transmission accounts for new infections, 1 with rates between 4–7%. 2 HCV will produce swarms of closely related clones or quasispecies; the most extensively documented area is an 81-bp region in the N-terminus of the env 2 region called the hypervariable region 1 (HVR-1), a major target for neutralizing antibodies. Analysis of HVR-1 allows establishment of linkage between different sequences. 3 Perinatal transmission can be verified by demonstrating close linkage between maternal and infant HCV isolates. Weiner et al 4 reported a case of an HCV-infected newborn whose unique isolate was different from those of the mother and raised the issue of whether the transmitted virus was an escape mutant or whether selection occurred at the time of transmission. Subsequent studies in children have demonstrated diverse patterns of transmission including transmission of multiple clones, transmission of dominant or subdominant clones, or a mixture of these. 5–7
Evolution of the quasispecies is hypothesized to be due to ongoing selection of viruses that are most fit for a particular host. 8 Selective pressure can be related to several factors but host immune pressure is thought to be a main factor driving diversification. 9–11 Failing to detect a correlation between HCV evolution and the strength of the host immune response to HVR-1 epitopes, Allain et al 12 have suggested that variation can result from genetic drift occurring independently of immune pressure. Comparison frequencies of synonymous nucleotide substitutions per synonymous site (dS) and nonsynonymous nucleotide substitutions per nonsynonymous site (dN) can be used to evaluate the process of natural selection. In the absence of selection, dS exceeds dN in most protein-coding genes. 13 A pattern of dN > dS suggests positive selection. 14
There have been relatively few studies of viral diversification in perinatally HCV-infected children, mainly in a handful of immune-competent infants and in 1 child with hypogammaglobulinemia. Manzin et al 6 have suggested that HCV diversification in immunocompetent infants occurs as a result of selective pressure. Viral diversification resulting from host immunologic pressure is thought to be a factor that contributes to viral persistence. 15 In immunocompromised hosts, diminished host immune pressure is more likely to result in less HCV diversification, as has been observed in HIV-coinfected adults 16,17 and in hypogammaglobulinemia. 18–20
This report evaluates the relationship of HCV transmitted from 2 HIV/HCV-coinfected mothers to their offspring and the quasispecies changes that occur in these and an additional infected infant.
PATIENTS AND METHODS
Three perinatally HIV- and HCV-coinfected infants and the coinfected mothers of 2 of these infants were enrolled. They represent an example of our HCV-infected population selected by virtue of having samples near birth. All were infected with HCV genotype 1b. Infants C1 and C2 were treated with nucleoside reverse transcriptase inhibitors soon after birth with nevirapine added at 4 months and 1 month of age, respectively. C3 was treated with reverse transcriptase inhibitor at 5 months and a protease inhibitor was added 3 years later. C1 and C2 had detectable antibody to HCV that developed in the 1st year of life after the attrition of transplacentally transmitted antibody. C3 had no antibody detected to HCV until well after his 2nd birthday. HCV quantitation was performed using the Roche COBAS Monitor v.2.0 assay (Indianapolis, IN). HIV quantitation was performed using the Roche Amplicor PCR assay. Virologic and immunologic characteristics of the patients are shown in Table 1.
HCV Sequence Analysis
HCV RNA Extraction and RT-PCT
HCV RNA was extracted by the single-step method from plasma (QIAamp viral RNA purification kit, QIAGEN, Valencia, CA) and subjected to reverse transcription polymerase chain reaction (RT-PCR) to amplify HRV-1. The RNA was reverse transcribed at 42°C for 50 minutes with random hexamer primer using superscript II (GibcoBRL, Rockville, MD). cDNA was amplified in a nested PCR reaction by rTth DNA polymerase (PE Applied Biosystems, San Jose, CA) for 35 cycles in an automated thermocycler 9600 (Perkin Elmer, San Jose, CA) with cycles of 94°C for 1 minute, 56°C for 1.5 minutes, and 72°C for 1.5 minutes. The outer primers for the PCR were sense primer 5′-TGG GAC ACA TGA TGA TGA ACT GGT, antisense primer 5′-GAT GTG CCA GCT GCC ATT GG. The inner primers were sense primer 5′-TAC TAC TCC ATG GTG GGA GAC TGG GC, antisense prime 5′-GAT GTG CCA GCT GCC ATT GG.
Amplicon Purification and Cloning
The amplified product was purified with the QIAquick gel extraction protocol (QIAGEN). The product was cloned into vector pCR-Blunt II-TOPO and was transformed into TOPO 10 Escherichia coli using the Zero Blunt TOPO PCR cloning kit (Invitrogen, Carlsbad, CA). Fifteen colonies were sampled each time and plasmid DNAs were isolated (MiniPrep kit, QIAGEN). All plasmids were confirmed by restriction analysis using Eco RI (GibcoBRL) and PCR reaction with inner primers.
Sequencing and Analysis
Sequencing was performed by the dideoxynucleotide chain-termination method with the Applied Biosystems automated DNA sequencer (model 373A). Simple sequence similarity comparisons were performed with BLAST 2.0. Alignments of both nucleotide and amino acid sequences were performed using the GCG program (Wisconsin package). Phylogenetic trees (neighbor-joining method 21) were generated on the basis of the number of nucleotide substitutions per site (d), estimated by the Jukes and Cantor 22 method. The reliability of branching patterns in the trees was assessed by bootstrapping. 23 One thousand bootstrap samples were used. The number of synonymous nucleotide substitutions per synonymous site (dS) and the number of nonsynonymous nucleotide substitutions per nonsynonymous site (dN) were estimated by the method of Nei and Gojobori. 24 We use the term “genetic distance” (Gd) to refer to the mean of “d” computed for all pairwise comparisons between 2 samples of sequences. Standard errors of mean d, dS, and dN were estimated by the bootstrap method. 25 This method takes into account the covariance among related sequences when means are computed for all pairwise comparisons. 25 The equality of mean dS and dN was tested by z-test, which provides a conservative test of this hypothesis. 25
HVR-1 Sequences and Phylogenetic Trees
HVR-1 sequences for the each of the 3 children and 2 mothers are presented in Table 2. Sequences are arranged from the first to last sampling time. Amino acid substitutions are noted for each position and point of time with respect to the initial dominant clone from the first sample.
For C1, plasma was negative for HCV and HIV at birth but positive at 1 month of age. Six variants were isolated (Table 2A), one of which contained almost half of the isolates. The consensus sequence was most closely related (differing by only 1 amino acid) to a minor variant present in the mother (M1m0) 3 weeks before delivery and also present in the mother 7 months later. This variant was also present as a minor variant in the infant at birth. None of the 14 other isolates present in the infant at 1 month of age was present in the mother close to delivery or at any other timepoint. One month later (C1m2), the most prevalent variant was replaced with one that was identical to the most prevalent variant found in the mother around the time of birth. This differed from the original major variant by 3 amino acids (D384E, N397S, and K410N) and was found in >50% of the clones, while the 410N motif, absent 1 month before, was found in 12 of 15 clones. Five months later (C1m7), the 410N motif was found in only 2 of 15 clones and the most prevalent variant (6/15) was identical to the consensus sequence at birth. This then was replaced by variants that contain almost exclusively 384E, 386H, and 405S changes, the most prevalent motif found in the mother at 7 and 19 months postpartum.
C2 was the product of an extrauterine pregnancy, with placental implantation on the peritoneal cavity, and was both HIV and HCV (>5 × 105 copies/mL) viremic at birth. Several variants were present at in the first sample (Table 2B). The dominant strain, accounting for 11 of 15 sequences, was identical to the dominant maternal virus. None of the minor variants present in the mother was present in the child at birth. The mother had 2 clusters of virus: the dominant variant that was transmitted to the infant and a cluster of closely related minor variants that share <50% of HVR-1 amino acids with the dominant variant. Relatives of these maternal minor variants emerged in the infant and completely replaced the dominant infant variant at the next sample time at 12 months to make up a relatively homogeneous population. Seven months later, only 20% of the variants retained elements of this motif and additional variants emerged. By 23 months, a new variant had almost completely replaced the previous ones.
C3 was first diagnosed and studied at 5 months of age. No maternal blood samples were available for analysis. Seven variants were isolated initially and were closely related, with no isolate differing by >2 amino acids (Table 2C). They can be grouped into 4 or 5 closely related clusters including the dominant variant that accounted for >50% of the isolates. Almost all isolates obtained after the first timepoint retained a Q to R change at amino acid 412 that was present initially in only several minor variants. This change to R at amino acid 412 accompanied by an F to L change at amino acid 397 was a motif that persisted until 54 months, with the 397L gradually replaced with an S. After the first sample, a striking homogeneity of viral sequences was seen with little variation until age 54 months. At that time, a V to M change was seen at amino acid 392 and a G to D change occurred at position 401. Fourteen months later, a major change occurred with the appearance of one homogeneous cluster that incorporated 4 new amino acid changes and the reversion of 401D and 412R to that found in the initial dominant variant seen at 5 months of age.
The overall shift of the quasipecies can be more easily visualized by construction of phylogenetic trees for each of these children (Fig. 1A–C) that show substantial diversity in HCV sequences from both mother and child and corroborate the individual sequence results. In the case of M2 and her child C2, the sequences formed 2 major clusters, each of which included sequences from both mother and child. The internal branch separating these 2 clusters received highly significant bootstrap support (99%) (Fig. 1B), supporting the hypothesis that the 2 clusters represented major clades of HCV that were present in M2 and transferred to her offspring C2. Interestingly, several of the initial maternal sequences did not appear in the child at birth but were evident at the child’s plasma at month 12 and month 19. Because this clade was present in the mother and eventually appeared in the offspring, it seems most likely that it was present in the child at birth but at a frequency too low to detect.
The phylogenetic tree of sequences from M1 and her child C1 (Fig. 1A) showed a pattern similar to that seen in M2 and C2. However, the clusters in the phylogeny of sequences from M1 and C1 were not as strongly supported as those from M2 and C2.
Genetic diversity (Gd) is a simple measure of the genetic distance or diversity within a population of viruses that allows us to place some of the changes described above in perspective. Despite extensive differences in the quantity and composition of viral variants, C1 and C2 show only a 5.7 and 3.5% divergence (Gd), respectively, from maternal variants at birth. This compares with 40.7, 51.9, and 40.7% divergence between C2 and C1, C2 and C3, and C3 and C1, respectively.
Over the total period of observation, C1 diverged the most. However, between intervals the greatest amount of diversity was seen in C2; a shift in dominant clusters from birth to 12 months and then back to the beginning at 19 months resulted in less overall diversification than in C1. C3 showed the least amount of variation and diversification with the tightest clustering at each timepoint and overall (Fig. 1C).
Comparisons of numbers of synonymous nucleotide substitutions per synonymous site (dS) and the number of non-synonymous nucleotide substitutions per nonsynonymous site (dN) were used to test the hypothesis that natural selection favors amino acid changes in the hypervariable region of the protein (Table 2).
In the case of pairwise comparisons within samples taken from C1, dN was significantly greater than dS in the hypervariable region in the 1st, 19th, and the 27th months, supporting the hypothesis that natural selection acted to diversify the hypervariable region. In the case of samples taken from C2, dN was significantly greater than dS in the hypervariable region in the 19th and the 23rd months and also in the overall comparison of all samples. The strongest selection was seen in C2. In C3, dS and dN did not differ among all sequences sampled.
Table 3 shows means of pairwise comparisons between each sample and the previous sample. In C1, dN was significantly greater than dS in the hypervariable region in the comparison between the 19th and 13th months and between the 27th and 19th months. In C2, dN was significantly greater than dS in the hypervariable region in every comparison. In C3, there was no significant difference between dS and dN in these comparisons.
The child with the most intact immune system, C2, had the most evidence of viral diversification, while C3, who demonstrated the most immune suppression and HIV replication, at least until highly active antiretrovirals were initiated at 41 months, had the least amount of diversity. However, there was no direct correlation between CD4 status at any specific time and the amount of diversity at that specific timepoint. No association between viral diversity and alanine amino transferase (ALT) or HCV viral load was observed in any of the children.
These results, obtained in children coinfected with HIV and HCV, confirm the findings of other studies suggesting that infants of HCV-infected mothers may be infected with multiple HCV clones from the mother, representing either dominant or minor populations of variants. The appearance of new variants in the child that were closely related to variants found in the mother at other timepoints suggests transmission of an even larger number of maternal variants in low frequency and not identifiable at earlier timepoints. The number of transmitted variants in these children is considerably higher than reported for children with only HCV infection. 5–7 Whether this is merely a sampling issue or whether it reflects transmission of greater numbers of variants or less initial selection of transmitted variants when the infant is also HIV infected is unclear. Transmission rates of HCV to their children are 5- to 7-fold higher in women who are also HIV coinfected. 26 This could result from higher HCV viral loads or greater viral diversity in HIV-coinfected women or could be a function of the infant’s ability to clear infection. HCV viremia is 10-fold higher in HIV-infected individuals. 27 Although HCV transmission has been correlated with the maternal viral load, 28 we have previously reported 26 that the increase in HCV transmission was limited to HIV infected and not related to maternal viral load. In both children with paired maternal results, dominant and minor variants were transmitted; in one case (C1) predominantly minor variants were transmitted, whereas in the 2nd child (C2) the large majority of isolates corresponded to the dominant maternal variant.
Diversification of HCV occurred over time but with different rates and evolution in each of the 3 children. Differential HCV phylogenetic evolution of common source–infected individuals suggests that individual host selective pressures are at play in determining quasispecies transmission and evolution. 29 The effect of immune pressure, or adaptive selection, not only increases viral diversification but may also result in escape from humoral or cellular immunity. A number of studies have documented that the hypervariable region (HRV-1) in the E2 region of the virus is a dominant neutralization epitope. 30–33 The carboxyterminal end of HRV-1 also contains epitopes for both T-helper and cytotoxic responses. The presence of immune pressures could play a role in determining which viral clones are passed from mother to child and could also influence the evolution of the viruses in both the mother and child over time. 6 Additional factors that have been evoked have been the mode of transmission (ie, intrauterine vs. perinatal) and potentially different quasispecies populations (blood, placenta vs. blood, vaginal fluid, and the amount of inoculum). 34 The high rate of changes in the 1st year of life, a time when infant immunity is generally diminished, 35 reflects possible viral escape as a result of declining titers of passively acquired maternal HCV antibody rather than newly acquired HCV-specific immune responses in the infant. After the 1st year of life, changes in infant quasispecies populations are less dramatic and suggest a more slowly evolving process.
In each infant consistent early changes were noted in the amino acid residues of the particular regions of HVR-1, especially after amino acid 390, that may correspond to critical anti-HCV epitopes. Zibert et al 36,37 have shown that in chronic HCV infection, antibodies are directed mainly against the C-terminus of the HVR-1, whereas antibodies to the N terminus of HVR-1 are associated with acute self-limited HCV infection. A major binding site has been mapped at position 390–405. 38 Antibodies binding to the 390–410 region were found to correlate with neutralizing activity. 31
The dramatic early shifts in these particular variants, especially the rapid alternating changes in C1 and C2 within the 1st year of life, might be explained by changes expected in the levels of transplacental maternal antibody followed by the appearance of the child’s humoral immune response. In the case of C3, HCV antibody production was delayed beyond the age of 2 years and may account for the absence of the alternation in viral variants seen in the other 2 infants. Viral diversification became more prominent only after antibodies to HCV developed and his HIV infection was under better control.
It would be impossible to accurately correlate viral diversification and immune status studying only 3 individuals. However, in these HIV-coinfected children, a trend was observed between HCV viral diversity and evolution over time and the degree of HIV-associated immunosuppression. C2, the child who was aggressively treated within the 1st month of life resulting in the maintenance of normal CD4 counts and complete and durable HIV suppression, had the most diversification of HCV. C3, who was treated at a much older age and with poor HIV suppression and greater immunoattrition, demonstrated the least amount of HCV diversification. In part this may have been related to his failure to produce HCV antibody until after 2 years of age. When finally begun on more effective antiretroviral agents, he achieved excellent HIV suppression and evidence of more extensive HCV diversification. No direct effect of HIV therapy on HCV viral load that could account for these differences was observed. In a study of quasi-species diversification (QSD) in HIV/HCV-coinfected adults, no contribution was noted by HIV except in patients with CD4 counts <50 cells/μL in whom the QSD index was one-quarter of that seen in those with higher CD4 counts or in those without HIV infection. 17 Moreover, those with the lowest CD4 had both lower cell-mediated immunity (CMI) and decreased antibody to the C100-3 antigen of HCV, underscoring the fact that CMI and humoral immunity play a role in quasispecies selection of regions in the HRV-1 region. A report of 10 adults with chronic HCV infection who had recently become coinfected with HIV showed a trend toward lower QSD and lower nonsynonymous substitutions in 7 of the individuals. 39 The overall magnitude of the effect was small but was most marked in the 5 individuals with rapid HIV disease progression.
Some studies suggest that as many as 50% of HCV-infected children (all HIV uninfected) will resolve infection by the end of the 2nd decade of life, although it is not clear at what age resolution occurs. 40 We have not seen any HIV-coinfected infants who have resolved their HCV infection in a decade of follow-up. HIV coinfection in HCV-infected adults results in a more rapidly evolving HCV infection with faster progression to cirrhosis than in HIV-uninfected individuals, 41 but there is little information available in a comparable population of children. More direct measures of humoral and cell-mediated immune responses to the HVR-1 region will be necessary to understand sequence changes that occurred in these children. In all 3 infants, HCV-specific cytotoxic T-lymphocyte responses to other regions of the HCV genome, outside of the HVR-1 region, were previously examined by tetramer analysis and were found to be relatively weak. 42
In conclusion, the analysis of HVR-1 changes is particularly complicated in perinatally infected children because of the frequent occurrence of multiple founder clones and the confounding effect of passively acquired maternal anti-HCV antibody, which dissipates over the first 12–15 months of life, on viral selection. Additionally, the presence of multiple transmitted maternal clones makes it difficult, at times, to distinguish new variants that arise in the infant from variants transmitted at birth at low frequency that only emerge at later time-points. Despite these difficulties we were able to demonstrate evidence of the selection and evolution of HCV variants even among HIV-coinfected infants before and after a year of age.
1. Schwimmer JB, Balistreri WF. Transmission, natural history, and treatment of hepatitis C virus infection in the pediatric population. Semin Liver Dis
2. Wasley A, Alter MJ. Epidemiology of hepatitis C: geographic differences and temporal trends. Semin Liver Dis
3. Weiner AJ, Christopherson C, Hall JE, et al. Sequence variation in hepatitis C viral isolates. J Hepatol
4. Weiner AJ, Thaler MM, Crawford K, et al. A unique, predominant hepatitis C virus variant found in an infant born to a mother with multiple variants. J Virol
5. Katayama Y, Tajiri H, Tada K, et al. Follow-up study of hypervariable region sequences of the hepatitis C virus (HCV) genome in an infant with delayed anti-HCV antibody responses. Microbiol Immunol
. 1998;42: 75–79.
6. Manzin A, Solforosi L, Debiaggi M, et al. Dominant role of host selective pressure in driving hepatitis C virus evolution in perinatal infection. J Virol
7. Rapicetta M, Argentini C, Spada E, et al. Molecular evolution of HCV genotype 2c persistent infection following mother-to-infant transmission. Arch Virol
8. Goto J, Nishimura S, Esumi M, et al. Prevention of hepatitis C virus infection in a chimpanzee by vaccination and epitope mapping of antiserum directed against hypervariable region 1. Hepatol Res
9. Gretch DR, Polyak SJ, Wilson JJ, et al. Tracking hepatitis C virus quasi-species major and minor variants in symptomatic and asymptomatic liver transplant recipients. J Virol
10. Ni YH, Chang MH, Chen PJ, et al. Decreased diversity of hepatitis C virus quasispecies during bone marrow transplantation. J Med Virol
. 1999;58: 132–138.
11. Sullivan DG, Wilson JJ, Carithers RL Jr, et al. Multigene tracking of hepatitis C virus quasispecies after liver transplantation: correlation of genetic diversification in the envelope region with asymptomatic or mild disease patterns. J Virol
12. Allain JP, Dong Y, Vandamme AM, et al. Evolutionary rate and genetic drift of hepatitis C virus are not correlated with the host immune response: studies of infected donor-recipient clusters. J Virol
13. Kimura M. Preponderance of synonymous changes as evidence for the neutral theory of molecular evolution. Nature
14. Hughes A. Adaptive Evolution of Genes and Genomes
. New York: Oxford University Press; 1999.
15. Farci P, Shimoda A, Coiana A, et al. The outcome of acute hepatitis C predicted by the evolution of the viral quasispecies. Science
. 2000;288: 339–344.
16. Mao Q, Ray SC, Laeyendecker O, et al. Human immunodeficiency virus seroconversion and evolution of the hepatitis C virus quasispecies. J Virol
17. Toyoda H, Fukuda Y, Koyama Y, et al. Effect of immunosuppression on composition of quasispecies population of hepatitis C virus in patients with chronic hepatitis C coinfected with human immunodeficiency virus. J Hepatol
18. Adams G, Kuntz S, Rabalais G, et al. Natural recovery from acute hepatitis C virus infection by agammaglobulinemic twin children. Pediatr Infect Dis J
19. Booth JC, Kumar U, Webster D, et al. Comparison of the rate of sequence variation in the hypervariable region of E2/NS1 region of hepatitis C virus in normal and hypogammaglobulinemic patients. Hepatology
. 1998;27: 223–227.
20. Kumar U, Monjardino J, Thomas HC. Hypervariable region of hepatitis C virus envelope glycoprotein (E2/NS1) in an agammaglobulinemic patient. Gastroenterology
21. Saitou N, Nei M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol
22. Jukes TH, Cantor CR. Evolution of protein molecules. In: Munro HN, ed. Mammalian Protein Metabolism
. New York: Academic Press; 1969:21–132.
23. Felsenstein J. Confidence-limits on phylogenies: an approach using the bootstrap. Evolution
24. Nei M, Gojobori T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol
25. Nei M, Kumar S. Molecular Evolution and Phylogenetics
. New York: Oxford University Press; 2000.
26. Papaevangelou V, Pollack H, Rochford G, et al. Increased transmission of vertical hepatitis C virus (HCV) infection to human immunodeficiency virus (HIV)-infected infants of HIV- and HCV-coinfected women. J Infect Dis
27. Sanchez-Quijano A, Andreu J, Gavilan F, et al. Influence of human immunodeficiency virus type 1 infection on the natural course of chronic parenterally acquired hepatitis C. Eur J Clin Microbiol Infect Dis
. 1995; 14:949–953.
28. Ohto H, Terazawa S, Sasaki N, et al. Transmission of hepatitis C virus from mothers to infants. The Vertical Transmission of Hepatitis C Virus Collaborative Study Group. N Engl J Med
29. McAllister J, Casino C, Davidson F, et al. Long-term evolution of the hypervariable region of hepatitis C virus in a common-source-infected cohort. J Virol
30. Bassett SE, Thomas DL, Brasky KM, et al. Viral persistence, antibody to E1 and E2, and hypervariable region 1 sequence stability in hepatitis C virus-inoculated chimpanzees. J Virol
31. Farci P, Alter HJ, Wong DC, et al. Prevention of hepatitis C virus infection in chimpanzees after antibody-mediated in vitro neutralization. Proc Natl Acad Sci U S A
32. Shimizu YK, Hijikata M, Iwamoto A, et al. Neutralizing antibodies against hepatitis C virus and the emergence of neutralization escape mutant viruses. J Virol
33. Zibert A, Schreier E, Roggendorf M. Antibodies in human sera specific to hypervariable region 1 of hepatitis C virus can block viral attachment. Virology
34. Fujii K, Hino K, Okazaki M, et al. Differences in hypervariable region 1 quasispecies of hepatitis C virus between human serum and peripheral blood mononuclear cells. Biochem Biophys Res Commun
35. Wilson CB. The ontogeny of T lymphocyte maturation and function. J Pediatr
36. Zibert A, Meisel H, Kraas W, et al. Early antibody response against hypervariable region 1 is associated with acute self-limiting infections of hepatitis C virus. Hepatology
37. Zibert A, Schreier E, Roggendorf M. Antibodies in human sera specific to hypervariable region 1 of hepatitis C virus can block viral attachment. Virology
38. Mondelli MU, Cerino A, Segagni L, et al. Hypervariable region 1 of hepatitis C virus: immunological decoy or biologically relevant domain?Antiviral Res
39. Mao Q, Ray SC, Laeyendecker O, et al. Human immunodeficiency virus seroconversion and evolution of the hepatitis C virus quasispecies. J Virol
40. Vogt M, Lang T, Frosner G, et al. Prevalence and clinical outcome of hepatitis C infection in children who underwent cardiac surgery before the implementation of blood-donor screening. N Engl J Med
41. Soto B, Sanchez-Quijano A, Rodrigo L, et al. Human immunodeficiency virus infection modifies the natural history of chronic parenterally-acquired hepatitis C with an unusually rapid progression to cirrhosis. J Hepatol
42. Lechner F, Sullivan J, Spiegel H, et al. Why do cytotoxic T lymphocytes fail to eliminate hepatitis C virus? Lessons from studies using major histocompatibility complex class I peptide tetramers. Philos Trans R Soc Lond B Biol Sci