Hepatitis C virus reinfection and superinfection among treated and untreated participants with recent infection

Authors


  • Potential conflict of interest: G.D., G.M., and J.K. have received research support from Roche Pharmaceuticals. G.D. is on the speaker's bureau for Roche Pharmaceuticals. G.D. and G.M. are members of advisory board for Roche Pharmaceuticals. G.D. and B.Y. have received travel grants from Roche Pharmaceuticals. G.D. is a consultant/advisor for Merck, Tibotec, and Abbott. J.G. is a member of an advisory board for Merck. G.M. is a consultant/advisor for Bristol-Myers Squibb, Gilead, Merck, and Roche.

Abstract

The purpose of the study was to evaluate reinfection and superinfection during treatment for recent hepatitis C virus (HCV). The Australian Trial in Acute Hepatitis C (ATAHC) was a prospective study of the natural history and treatment of recent HCV. Reinfection and superinfection were defined by detection of infection with an HCV strain distinct from the primary strain (using reverse-transcription polymerase chain reaction [RT-PCR] and subtype-specific nested RT-PCR assays) in the setting of spontaneous or treatment-induced viral suppression (one HCV RNA <10 IU/mL) or persistence (HCV RNA >10 IU/mL from enrollment to week 12). Among 163 patients, 111 were treated, 79% (88 of 111) had treatment-induced viral suppression, and 60% (67 of 111) achieved sustained virological response. Following treatment-induced viral suppression, recurrence was observed in 19% (17 of 88), including 12 with relapse and five with reinfection (4.7 cases per 100 person-years [PY], 95% confidence interval [CI]: 1.9, 11.2). Among 52 untreated patients, 58% (30 of 52) had spontaneous viral suppression and recurrence was observed in 10% (3 of 30), including two with reinfection. Following reinfection, alanine aminotransferase (ALT) levels >1.5× the upper limit of normal were observed in 71% (5 of 7). Among 37 with persistence, superinfection was observed in 16% (3 of 19) of those treated and 17% (3 of 18) of those untreated. In adjusted analysis, reinfection/superinfection occurred more often in participants with poorer social functioning at enrollment and more often in those with ongoing injecting drug use (IDU). Conclusion: Reinfection and superinfection can occur during treatment of recent HCV and are associated with poor social functioning and ongoing IDU. ALT levels may be a useful clinical marker of reexposure. (HEPATOLOGY 2012)

The majority of new cases of hepatitis C virus (HCV) infection in the developed world occur among injecting drug users (IDUs).1 Pegylated-interferon (PEG-IFN) and ribavirin (RBV) combination therapy is effective in ≈55% of patients,2 but treatment uptake among active IDUs is extremely low.3-5 Although converging patient, provider, organizational, and structural barriers complicate HCV care among IDUs,6 response to therapy in this group is similar to that observed in clinical trials.7 However, there is still a general reluctance to provide treatment, largely driven by concerns of adherence, social instability, side effects, comorbid psychiatric disease, and the perceived risk of reinfection following successful treatment.6 Although the concern of reinfection is often cited as a reason for not offering treatment to IDUs,6 little is known about reinfection in the setting of treatment.

Reinfection following successful treatment for HCV can occur.8-12 Few studies have evaluated the incidence of reinfection following successful HCV treatment and the incidence of reinfection has been reported as less than five cases per 100 person-years (PY).8-12 Studies of reinfection following treatment-induced clearance are limited by the small sample sizes (n < 40), incomplete longitudinal follow-up, insensitive detection methods, and the few cases of reinfection that have been detected.

Ongoing risk behaviors can also lead to HCV reexposure in the setting of persistent viremia (no period of aviremia). The detection of multiple unique HCV viral strains at a single timepoint is often referred to as HCV mixed infection and could be the result of simultaneous exposure or superinfection. HCV superinfection is used to define the presence of a new infection with a strain different from the initial infection in participants with persistent HCV infection.13, 14 There are no studies of HCV mixed infection or superinfection in the setting of treatment for HCV infection. HCV reexposure and superinfection during the early stages of HCV treatment could impact virological suppression and treatment outcomes.

The Australian Trial in Acute Hepatitis C (ATAHC) is a prospective trial of the natural history and treatment of recently acquired HCV infection, consisting mainly of IDUs.15 The aim of the current study was to investigate the epidemiological, virological, and clinical characteristics of HCV reinfection and superinfection among treated and untreated participants with recent HCV infection.

Abbreviations

ALT, alanine transaminase; ATAHC, Australian Trial in Acute Hepatitis C; CI, confidence interval; E1/HVR1, envelope glycoprotein-1 and hypervariable region 1 on E2; HCV, hepatitis C virus; HIV, human immunodeficiency virus; IDU, injecting drug user; nRT-PCR, nested reverse transcription - polymerase chain reaction; PEG-IFN, pegylated-interferon; PY, person-year; RBV, ribavirin; SVR, sustained virological response; ULN, upper limit of the normal range.

Participants and Methods

Study Participants.

ATAHC was a prospective cohort study of the natural history and treatment of recent HCV infection.15 Recruitment of human immunodeficiency virus (HIV)-infected and uninfected participants was from June 2004 through November 2007. Recent infection with either acute or early chronic HCV infection with the following eligibility criteria: first positive anti-HCV antibody within 6 months of enrollment; and either: (1) Acute clinical hepatitis C infection, defined as symptomatic seroconversion illness or alanine aminotransferase (ALT) level greater than 10 times the upper limit of normal (ULN, >400 U/L) with exclusion of other causes of acute hepatitis, at most 12 months before the initial positive anti-HCV antibody; or (2) Asymptomatic hepatitis C infection with seroconversion, defined by a negative anti-HCV antibody in the 2 years prior to the initial positive anti-HCV antibody. All participants with detectable HCV RNA during screening were assessed for HCV treatment. Participants unwilling to undergo treatment and those with undetectable HCV RNA at screening continued to be followed. From screening, participants were followed for up to 12 weeks to allow for spontaneous clearance, and if HCV RNA remained detectable were offered treatment. Participants were then seen at baseline and 12-weekly intervals for up to 144 weeks (participants receiving treatment were also seen at 4-weekly intervals up to week 12).

All study participants provided written informed consent. The study protocol was approved by local Ethics Committees. The study was registered with clinicaltrials.gov registry (NCT00192569).

HCV Treatment.

Participants who began HCV treatment received PEG-IFN α-2a 180 μg weekly for 24 weeks. Due to nonresponse at week 12 in the initial two participants with HCV/HIV coinfection, the study protocol was amended to provide PEG-IFN and RBV combination therapy for 24 weeks in this group. RBV was prescribed at a dose of 1,000-1,200 mg for those with genotype 1 and 800 mg in those with genotype 2/3.

Detection and Quantification of HCV RNA.

HCV RNA assessment was performed at all scheduled study visits, initially with a qualitative HCV RNA assay (TMA assay, Versant, Bayer, Australia, lower limit of detection 10 IU/mL) and if detectable repeated on a quantitative assay (Versant HCV RNA 3.0 Bayer, lower limit of detection 615 IU/mL). HCV genotyping using a commercial assay (Versant LiPa2, Bayer) was performed on all participants with detectable HCV RNA at screening for clinical decision making. Additional HCV genotyping using sequencing methodologies were also performed on longitudinal samples as described below.

HCV Virological Suppression, Persistence, Reinfection, and Superinfection.

Participants with HCV virological suppression (spontaneous or treatment-induced) were defined by one detectable HCV RNA test followed by ≥1 undetectable qualitative HCV RNA test(s) (<10 IU/mL). The estimated date of HCV virological suppression was calculated as the midpoint between the dates of the last detectable HCV RNA test and the first undetectable qualitative HCV RNA test. Participants with HCV virological persistence (in the presence or absence of treatment) were defined by one detectable HCV RNA test followed by ≥1 detectable HCV RNA test(s), with no undetectable qualitative HCV RNA tests (<10 IU/mL).

Participants with HCV RNA recurrence (detectable HCV RNA following HCV virological suppression) were identified. HCV RNA sequencing was performed (details below) on the first available detectable HCV RNA sample and the first available detectable HCV RNA sample following HCV RNA recurrence. HCV reinfection was defined by the detection of infection with an HCV strain that was distinct from the primary infecting strain among participants with either spontaneous or treatment-induced HCV virological suppression (as described above, ≥1 undetectable qualitative HCV RNA test(s) <10 IU/mL).

Participants with HCV virological persistence from the screening/baseline timepoint to the first available sample with detectable HCV RNA at ≥12 weeks were identified and HCV RNA sequencing was performed on these samples. Participants with HCV mixed infection were defined by the detection of two distinct HCV strains at the same timepoint (<94% envelope glycoprotein-1 and hypervariable region 1 on E2 [E1/HVR1] sequence match between viruses, as described in Pham et al.16). Participants with HCV superinfection were defined by the detection of an HCV strain that was distinct from the primary infecting strain in those with HCV virological persistence (<94% E1/HVR1 sequence match between both timepoints as described in Pham et al.16). Among those in whom HCV superinfection was identified, strain-specific nested reverse-transcription polymerase chain reaction (nRT-PCR)16 was used to exclude the possibility that the viral strain at the time of superinfection was present at earlier timepoints. Participants with HCV virological persistence and no HCV superinfection were defined by the detection of infection with an HCV strain that demonstrated close sequence identity to the primary infecting strain in those with HCV virological persistence with E1/HVR1 sequence data matching >97.6% between both timepoints, as described in Pham et al.16

The estimated dates of reinfection and superinfection were calculated as the midpoint between the dates of the last undetectable qualitative HCV RNA test (or detectable test for those with superinfection) and the first detectable HCV RNA test at the time of HCV reinfection (or superinfection).

Detection of HCV Reinfection, Mixed Infection, and Superinfection.

The methods for this work have been previously described.16 Briefly, RNA was extracted from 140 μL of sera using the QIAmp Viral RNA extraction kit (Qiagen, Hilden, Germany).16 The E1/HVR1 region encoding the last 171 basepairs (bp) of Core, E1, and the HVR-1 of E2 (840 bp, nucleotides 744 to 1583, with reference to HCV strain H77, GenBank accession number AF009606) was amplified by a real-time nRT-PCR using reagents and reaction conditions as described.16 Mixed infection at a single timepoint was detected by performing four Core real-time subtype-specific nRT-PCRs for HCV 1a, 1b, 2a, and 3a.16

Statistical Analyses.

Rates of HCV reinfection were calculated using person-time of observation. Confidence intervals (CI) for rates were calculated using Poisson distribution. The time at risk for participants with sustained viral suppression and no viral recurrence was calculated from the estimated date of viral suppression until the date of the last HCV RNA test during follow-up. The time at risk for participants with viral recurrence (viral relapse or reinfection) was calculated from the estimated date of viral suppression until the first positive HCV RNA test of viral recurrence.

Logistic regression was used to estimate crude and adjusted odds ratios (OR) and corresponding 95% CI to identify factors associated with HCV reinfection/superinfection. In unadjusted analyses, potential predictors were determined a priori and included sex, age, social functioning score at enrollment (median), current depression at enrollment, HIV infection at enrollment, IDU (ever, previous 6 months at enrollment, and previous 30 days at enrollment), and IDU during follow-up (previous 30 days). Social functioning was calculated using a validated scale from the Opiate Treatment Index17 and addresses employment, residential stability, and interpersonal conflict as well as social support (higher score reflects poorer social functioning, range score 0-48). Further, in a separate model among those with a history of IDU we hypothesized that peak IDU frequency, syringe borrowing, and methamphetamine injecting were associated with reinfection/superinfection. All variables with P < 0.20 in bivariate analysis were considered in multivariate logistic regression models using a backwards stepwise approach. Statistically significant differences were assessed at P < 0.05; P-values are two-sided. All final adjusted models included only factors that remained significant at the 0.05 level. All analyses were performed using the statistical package Stata v. 10.1 (College Station, TX).

Role of Funding Source.

The funding sources for the study did not contribute to study design, data collection, analysis, or interpretation and had no role in writing of the article or decision to submit the article for publication.

Results

Participant Characteristics.

Between June 2004 and February 2008, 163 participants with recent HCV were enrolled in ATAHC.15 The study characteristics and outcomes have been published.15 Briefly, the majority were male (72%) and had a history of IDU (76%). Overall, 35% reported IDU within the last month at enrollment (Table 1).

Table 1. Enrolment Characteristics of Participants (n=163)
 Overall n (%)HCV Reinfection n (%)HCV Superinfection n (%)No Reinfection or Superinfection* n (%)
  • *

    Excludes 13 participants with inadequate follow-up to assess virological suppression or persistence; IDU, injecting drug use.

Total participants, (n)16376137
Male, n (%)117 (72)6 (86)4 (67)100 (73)
Age (yrs), mean/SD34.4 ± 10.036.8 + 7.224.3 ± 5.434.7 ± 10.0
Tertiary education or greater, n (%)66 (40)4 (57)1 (17)56 (41)
Accommodation, n (%)    
 Rent101(62)3 (43)3 (50)86 (63)
 Privately owned39 (24)3 (43)1 (17)34 (25)
 Other23 (14)1 (14)2 (33)17 (12)
Full-time or part-time employment, n (%)63 (39)2 (29)1 (17)57 (42)
Social functioning score    
 ≤1484 (52)1 (14)1 (17)77 (56)
 >1460 (37)4 (57)4 (67)45 (33)
 Missing19 (12)2 (29)1 (17)15 (11)
Current depression, n (%)25 (16)2 (29)2 (33)19 (14)
HIV infection, n (%)50 (31)3 (43)0 (0)45 (33)
IDU ever, n (%)124 (76)6 (86)6 (100)101 (74)
IDU in previous 6 months, n (%)99 (61)6 (86)5 (83)79 (58)
IDU in previous 30 days, n (%)55 (35)4 (57)1 (17)44 (32)

At enrollment, 17 participants were negative for HCV RNA. The remaining 146 participants were positive for HCV RNA at enrollment and were thus eligible to receive HCV treatment. Of these, 111 initiated HCV treatment. Among those initiating treatment, 74 were infected with HCV alone (received PEG-IFN monotherapy) and 37 were HCV/HIV coinfected (received PEG-IFN alpha-2a monotherapy, n = 2; received PEG-IFN alpha-2a +RBV combination therapy, n = 35).

HCV Reinfection Among Treated Participants with Recent Infection.

Of 111 treated in the ATAHC study, 79% (88 of 111) had treatment-induced virological suppression (Fig. 1) and 60% (67 of 111) achieved a sustained virological response (SVR). Participants were followed for a mean of 1.2 years following treatment-induced virological suppression (median, 1.1 years; interquartile range [IQR], 0.7, 1.9 years; range, 0.1, 2.9 years) and 0.5 years following SVR (median, 0.5 years; IQR, 0.0, 1.7 years; range, 0, 2.0 years).

Figure 1.

Overview of study population at risk of HCV reinfection.

Following treatment-induced virological suppression, HCV RNA recurrence was observed in 19% (17 of 88, Fig. 1) and HCV RNA sequencing identified viral relapse in 12 participants with homologous viruses at both timepoints (14%; 12 of 88) and five participants with reinfection (6%; 5 of 88; four confirmed by sequencing and one possible). Among the five with reinfection, three were infected with a single heterologous genotype and ID810 demonstrated mixed reinfection with both HCV 1a and 3a viruses (Fig. 2A). The one participant with possible reinfection (ID2303) had an HCV RNA recurrence following SVR which could not be sequenced due to low-level viremia (1,463 IU/mL) and subsequently spontaneously cleared infection (four HCV RNA negative tests after clearance, with the last test 1.4 years following the first negative test). HCV reinfection occurred 19 weeks (ID1101), 28 weeks (ID637), 31 weeks (ID2303), 41 weeks (ID810), and 68 weeks (ID2601) following the end of treatment. The demographic characteristics of participants with reinfection are shown in Tables 1 and 2.

Figure 2.

Natural history of HCV reinfection in (A) treated (n = 5), and (B) untreated (n = 2) participants with recent HCV infection and virological suppression. Circles indicate HCV RNA (light gray shading indicates persistent viremia) and squares indicate ALT levels. Subtypes are color-coded circles as shown in the key (sizes of the circle correlate to log of viral load [IU/mL] of each HCV population at each timepoint). Treatment with PEG-IFN and/or RBV is indicated by the dark gray box. HCV genotype is indicated with either an arrow (indicating genotype at a single timepoint) or a line with hashes indicating that all samples between that specific time period were of that same genotype. Greek superscripts indicate infection with a genetically different HCV subtype.

Table 2. Detailed Demographic, Behavioral, and Virological Characteristics of Participants with HCV Reinfection and Superinfection
IDAge (years)SexHIVMode of Primary InfectionIDU EverIDU Frequency at EnrolmentMode of Reinfection/ SuperinfectionHCV Genome Region AnalyzedPrimary Subtype (Estimated Days Since Infection)Reinfection/ Superinfection Subtype (Estimated Days Since Infection)Time Between Primary and Secondary Infection (Weeks)Total Number of Viruses DetectedConfirmed/ Possible
  1. M, male; F, female; N, No; Y, Yes; IDU, injecting drug use. Greek superscripts indicate infection with a genetically different HCV subtype.

Reinfection
Treated            
 63738MNIDUYes≥dailyIDUE1/HVR11a (365)3a (1093)872 (1a/3a)Confirmed
        Core1a (365)3a (1093)87  
 81046MYSexualNoNoneSexualCore2a (217)1a (784)813 (2a/1a/3a)Confirmed
        Core2a (217)3a (784)81  
 110131MNIDUYes≥dailyIDUE1/HVR13a (213)1a (556)492 (3a/1a)Confirmed
 230335MNIDUYes≥dailyIDUInnoLipa3a (299)NA (719)601 (3a)Possible
 260151MYIDU/Yes<dailyIDU/Core3a (88)3a (865)1113 (1aα/3a/1aβ)Confirmed
    Sexual  SexualCore1aα (88)3a (865)111  
        Core3a (865)1aβ (1059)194  
        Core1aα (88)1aβ (1059)971  
Untreated            
 20630FNIDUYes≥dailyIDUE1/HVR1NA (222)2b (301)111 (2b)Possible
 63536MYIDU/Yes≥dailyIDU/InnoLipa3a (173)NA (495)462 (3a/2a)Confirmed
    Sexual  Sexual5′UTRNA (173)2a (495)462 (3a/2a) 
Superinfection
Treated            
 11017MNIDUYes≥dailyIDU5′UTR2b (67)1a (123)83 (2b/1a/3a)Confirmed
        E1/HVR-11a (123)3a (855)105  
        Core1a (123)3a (855)105  
 201025FNIDUYes<dailyIDUE1/HVR-13a (225)1b (309)122 (3a/1b)Confirmed
        Core3a (225)1b (253)4  
 230431MNIDUYes<dailyIDUE1/HVR-11a (165)2a (291)182 (1a/2a)Confirmed
        Core1a (165)2a (291)18  
Untreated            
 60218FNIDUYes≥dailyIDUE1/HVR-13aα (471)3aβ (853)553 (3aα/3aβ/3aδ)Confirmed
        E1/HVR-13aβ (853)3aδ (1163)44  
        E1/HVR-13aα (471)3aδ (1163)99  
 140324MNIDUYes≥dailyIDUE1/HVR-13a (394)1b (527)192 (3a/1b)Confirmed
        Core3a (394)1b (527)19  
 200229MNIDUYesNoneIDUE1/HVR-13aα (148)3aβ (177)42 (3aα/3aβ)Confirmed

Incidence of HCV reinfection was 4.7 cases per 100 PY (95% CI: 1.9, 11.2) following treatment-induced virological suppression and 12.3 per 100 PY (95% CI: 5.1, 29.6) following SVR. Ongoing IDU during follow-up was reported in 38% of participants with treatment-induced virological suppression (33 of 88). HCV reinfection in this subgroup was 7.3 per 100 PY (95% CI: 2.3, 22.6).

HCV Reinfection Among Untreated Participants with Recent Infection.

Of 52 untreated participants in the ATAHC study, 58% (30 of 52) had spontaneous virological suppression and HCV RNA recurrence was observed in 10% (3 of 30). HCV RNA sequencing identified viral relapse in one participant with a homologous viral strain at both timepoints and two participants with reinfection (one confirmed by sequencing and one possible). Patient ID635 spontaneously resolved an HCV 3a infection and following four negative HCV RNA tests was subsequently reinfected with an HCV 2a strain 28 weeks after primary clearance. The one participant with possible reinfection (ID206) could not be sequenced at the timepoint of primary infection due to low-level HCV RNA (<10 IU/mL).

Incidence of HCV reinfection was 6.1 per 100 PY (95% CI: 1.5, 24.6) following spontaneous virological suppression. Incidence of HCV reinfection stratified by HCV treatment status, HIV infection status or IDU during follow-up is shown in Table 3.

Table 3. Incidence of Reinfection Among Participants with Treatment-Induced (n=88) or Spontaneous (n=30) Virological Suppression During Recent HCV Infection (n=118)
Incidence TypeCasesNumber at RiskPerson-Years of Follow-UpIncidence per 100 Person Years95% CI95% CI
  • *

    Numbers do not equal 118 due to missing data on injecting during follow-up in seven participants.

Overall      
 Confirmed/possible reinfection71181405.02.410.5
 Confirmed reinfection51181403.61.58.6
 Confirmed persistent reinfection31181402.10.76.7
Untreated      
 Confirmed/possible reinfection230336.11.524.6
 Confirmed reinfection130333.10.421.8
 Confirmed persistent reinfection030330.0--
Treated      
 Confirmed/possible reinfection5881074.71.911.2
 Confirmed reinfection4881073.71.49.9
 Confirmed persistent reinfection3881072.80.98.7
HIV uninfected      
 Confirmed/possible reinfection479974.11.510.9
 Confirmed reinfection279972.10.58.2
 Confirmed persistent reinfection179971.00.17.3
HIV infected      
 Confirmed/possible reinfection339427.12.322.0
 Confirmed reinfection339427.12.322.0
 Confirmed persistent reinfection239424.71.218.9
No injecting during follow-up      
 Confirmed/possible reinfection266*812.50.69.9
 Confirmed reinfection166*811.21.78.8
 Confirmed persistent reinfection166*811.21.78.8
Injecting during follow-up      
 Confirmed/possible reinfection545*568.93.721.4
 Confirmed reinfection445*567.12.719.0
 Confirmed persistent reinfection245*563.60.914.2

Natural History of HCV Reinfection.

The individual HCV RNA and ALT profiles of the seven participants with reinfection (five treated and two untreated) are shown in Fig. 2. ALT and changes were common following reinfection (Fig. 2, Table 4). Five of seven participants (71%) demonstrated peak ALT levels greater than 1.5 times the ULN around the time of HCV reinfection (319, 1,875, 1,135, 361, and 894 U/L, Table 4) compared with one of seven at enrollment (14%, 120 U/L, Table 4). Reinfection remained persistent in three of seven cases (43%). Among those with HCV reinfection, two participants demonstrated evidence of HCV mixed infection during follow-up (ID810 and ID2601, Fig. 2).

Table 4. ALT levels, HCV RNA, and Virological Outcomes Among Participants with HCV Reinfection (n=7) and Superinfection (n=6) During Recent HCV Infection
IDEstimated Duration of Primary Infection (Weeks)ALT at Enrollment (U/L)Peak ALT Following Reinfection/ Superinfection (U/L)HCV RNA at Enrollment (log10 IU/ml)Peak HCV RNA at Reinfection/ Superinfection (log10 IU/ml)Outcome
Reinfection cases
Treated      
 63752443194.35.9Persistence
 810265818756.55.0Persistence
 110125120534.94.6Clearance
 260196011354.33.6Persistence
 23032823143.43.2Clearance
Untreated      
 2063211361<1.04.9Clearance
 63525108942.83.7Clearance
Superinfection cases
Treated      
 110102159834.75.7Persistence
 2010231141815.46.9Clearance
 230424844283.92.8Clearance
Untreated      
 60262762243.04.9Persistence
 140356130785.05.8Persistence
 200221453553.34.7Clearance

HCV Superinfection Among Treated and Untreated Participants with Recent Infection.

Among treated participants with HCV RNA persistence (lack of virological suppression through week 12) HCV RNA sequencing detected 16% (3 of 19) with superinfection (Fig. 3). Among untreated participants with HCV RNA persistence, 17% (3 of 18) were superinfected. The demographic and detailed demographic characteristics of participants with superinfection are shown in Tables 1 and 2.

Figure 3.

Overview of study population at risk of HCV superinfection.

Natural History of HCV Superinfection.

The individual HCV RNA, ALT profiles, and virological outcomes of these six participants with superinfection are shown in Fig. 4 and Table 4. ALT and HCV RNA changes were common following superinfection. Five of six participants (83%) demonstrated peak ALT levels greater than 1.5 times the ULN around the time of superinfection (983, 181, 428, 224, and 355 U/L) as compared with three of six at enrollment (50%; 215, 114, and 130 U/L; Table 2). Further, five of six participants (83%) demonstrated greater peak HCV RNA levels at the time of superinfection as compared with enrollment (ID110, 5.7 versus 4.7; ID2010, 6.9 versus 5.4; ID602, 4.9 versus 3.0; ID1403, 5.8 versus 5.0; and ID2002, 4.7 versus 3.3 log10 IU/mL; Table 4). Among those with HCV superinfection, two participants demonstrated evidence of HCV mixed infection during follow-up (ID110 and ID2010, Fig. 4).

Figure 4.

Natural history of HCV superinfection in (A) treated (n = 3), and (B) untreated (n = 3) with recent HCV infection and persistence. Circles indicate HCV RNA (light gray shading indicates persistent viremia) and squares indicate ALT levels. Subtypes are color-coded circles as shown in the key (sizes of the circle correlate to log of viral load [IU/mL] of each HCV population at each timepoint). Treatment with PEG-IFN and/or RBV is indicated by the dark gray box. HCV genotype is indicated with either an arrow (indicating genotype at a single timepoint) or a line with hashes indicating that all samples between that specific time period were of that same genotype. Greek superscripts indicate infection with a genetically different HCV subtype.

Risk Factors for HCV Reinfection/Superinfection.

The demographic characteristics of the seven participants with reinfection and six participants with superinfection are shown in Tables 1 and 2. In four of seven participants with reinfection (ID637, ID1101, ID2303, and ID206), IDU was the attributed mode of acquisition for both initial infection and reinfection. In the three other participants (ID635, ID810, and ID2601, all HIV-infected males), both initial infection and reinfection were attributed to high-risk homosexual contact, although two of these cases (ID635 and ID2601) also reported IDU (frequency less than weekly and denied any injecting equipment sharing) at the time of both exposures. In all six participants with superinfection, IDU was the attributed mode of acquisition for both initial infection and reinfection.

Factors associated with reinfection/superinfection were evaluated by comparing participants with reinfection/superinfection (n = 13) to those demonstrating virological suppression without reinfection and those with viral persistence without superinfection (n = 137; Supporting Table 1). In unadjusted analyses in the overall population, reinfection/superinfection occurred more often in those with poor social functioning, those with IDU in previous 6 months at enrollment, and those with IDU during follow-up. In adjusted analysis, reinfection/superinfection occurred more often in participants with poorer social functioning at enrollment (score ≤14 versus score >14 adjusted OR [AOR] 6.05, 95% CI: 1.19, 30.87, P = 0.030) and those with IDU during follow-up (AOR 4.35, 95% CI: 1.09, 17.45, P = 0.038).

We performed further analyses to assess specific injecting behaviors during follow-up associated with reinfection/superinfection among those with a history of IDU (n = 112, Supporting Table 2). In unadjusted analyses in those with a history of IDU, reinfection/superinfection occurred more often in those with poor social functioning, ≥daily peak IDU frequency during follow-up, syringe borrowing during follow-up, and methamphetamine/cocaine injecting during follow-up. In adjusted analysis, reinfection/superinfection occurred more often in participants with poorer social functioning at enrollment (score ≤14 versus score >14, AOR 5.85, 95% CI: 1.11, 30.92, P = 0.038) and in those with methamphetamine injecting during follow-up (AOR 7.29, 95% CI: 1.86, 28.54, P = 0.004).

Discussion

In this prospective study we characterized the virological, clinical, and epidemiological features of reinfection/superinfection in treated and untreated participants with recent HCV, consisting predominantly of IDUs. Among those with reinfection and superinfection, elevations in HCV RNA and ALT levels around the time of secondary infection were common, providing potentially useful clinical markers for reexposure. Reinfection/superinfection was associated with poor social functioning at enrollment and ongoing IDU (but not injecting at enrollment). Among injectors, the greatest odds of reinfection/superinfection occurred among those with poor social functioning at enrollment and ongoing methamphetamine IDU. The outcomes of infection following reinfection and superinfection were heterogeneous, with four of seven participants with reinfection demonstrating spontaneous clearance of reinfection (including two following treatment). This study provides a greater understanding of the natural history and factors associated with reinfection/superinfection, particularly in the setting of HCV treatment among IDUs.

Elevations in ALT occurred in almost three-quarters of participants following reinfection/superinfection. The observed increases in ALT following reinfection are consistent with data from chimpanzees18-22 and a small series of case studies in humans.11, 23 Similarly, the observed increases in ALT following superinfection is consistent with available data from a small number of superinfection case studies in chimpanzees24 and humans.25, 26 It is possible that given the frequency of testing, elevations in ALT in some participants may have been missed. However, it is interesting that those with reinfection and superinfection were similar with respect to the observed increases in ALT following reexposure. Increases in HCV RNA levels occurred in greater than three-quarters of participants around the time of HCV superinfection. These data have important clinical implications, as regular monitoring of ALT and HCV RNA elevations among participants at high risk may provide a clinically useful tool for detecting reinfection and superinfection.

Reinfection and superinfection were independently associated with poor social functioning at enrollment and ongoing IDU (but not IDU at the time of study enrollment). Among those with a history of IDU, the greatest risk of reinfection/superinfection infection was among those with poor social functioning at enrollment and ongoing methamphetamine IDU. Given the small number of cases of reinfection and superinfection reported in studies performed to date,16, 27-34 there are no previous studies that have identified factors independently associated with reinfection and/or superinfection. Although IDU during follow-up was associated with reinfection/superinfection, a history of IDU and recent IDU (past 30 days and past 6 months) at study enrollment was not. Thus, within IDU populations it may be difficult to predict on the basis of injecting characteristics at enrollment who will subsequently become reinfected/superinfected with HCV. As such, it is important to ensure that adequate harm reduction measures are in place for those who relapse back to ongoing IDU.

Among treated participants, the rate of reinfection was 4.7 cases per 100 PY in those with virologic suppression and 12.3 cases per 100 PY in those with SVR. Our results are slightly higher than previously reported retrospective studies of reinfection following successful treatment of chronic HCV infection among IDUs.8-12 In Germany, among 18 IDUs followed for a mean of 2.8 years after successful treatment for HCV infection (50% relapsed to IDU following treatment), 0-2 cases of reinfection were observed (reinfection <4.1 cases per 100 PY).8 Similar results were reported from 27 IDUs followed for 5.4 years following SVR in Norway.9 The incidence of reinfection was 0.8 cases per 100 PY, 2.5 cases per 100 PY among the 9 of 27 (33%) who returned to IDU during follow-up.9 In Canada, two reinfections were observed among 35 IDUs with SVR and the rates of reinfection were 3.2 per 100 PY (95% CI: 0.4, 11.5) overall and 5.3 per 100 PY (95% CI: 0.6, 19.0) among those reporting injecting following SVR.11 The observed rate does not support withholding therapy for IDU populations, where the greatest burden of HCV-related disease is concentrated. However, given that reinfection can occur, education around harm reduction and the risks of reinfection following successful treatment should be incorporated into the plan of care for those with a history of IDU.

There are a number of limitations to this study. Among participants with superinfection, it is possible that the viral strain at the time of superinfection was present at enrollment. Although the subtype-specific nRT-PCR assays were sensitive (86 IU/mL), we cannot exclude the possibility of the presence of a virus previously present in low copy number. Further, in those with HCV persistence we cannot exclude the possibility of rapid clearance with subsequent reinfection between testing intervals. This may have been more possible in untreated participants, given that participants were tested for HCV RNA more frequently (monthly) during treatment. We also cannot exclude the possibility that some reinfections with rapid clearance occurred during periods when sampling of HCV RNA was only every 12 weeks, potentially also underestimating the rate of reinfection. That being said, this would not have had a major impact on the detection of HCV reinfections that remain persistent, which is clinically more important. It is possible that some of those with HCV persistence following reinfection were not followed for a sufficient time to allow for spontaneous clearance. If those with HCV reinfection and persistence subsequently cleared infection given longer follow-up, this would have underestimated the proportion with clearance of HCV reinfection. However, the two with clearance of reinfection had two negative HCV RNA tests ≥60 days following clearance (well-accepted definition of spontaneous clearance), so we are confident of clearance in these participants.

The study findings have important implications for the clinical management of HCV in IDUs and for HCV vaccine development. Regular HCV RNA and/or ALT testing could be used as useful clinical tools for the detection of some cases of HCV reinfection/superinfection. Although this study was performed among those with recent HCV, it is possible that sharp ALT rises sometimes observed in chronic HCV may be attributed to superinfection. Although many clinicians may not have access to the sequencing technologies employed here, a standard HCV genotype following suspected reinfection or superinfection (based on ALT or HCV RNA elevations) may identify some individuals with reexposure. Those at greatest risk of HCV reinfection/superinfection include IDUs, particularly those with poor social functioning and ongoing injecting (particularly methamphetamine). Social functioning at baseline may provide some indication of those most likely to become reinfected/superinfected. Harm reduction and education around the risks of HCV reinfection and superinfection should be incorporated into the clinical management of all IDUs initiating treatment for HCV. However, IDUs should not be denied care on the basis of concerns of HCV reinfection. These results also have implications for HCV vaccine design. The observation that both participants with spontaneous HCV clearance and two of five with treatment-induced clearance demonstrated spontaneous clearance following HCV reinfection is important. It suggests that the induction of partial protective immunity may be possible, providing hope for a vaccine that enhances spontaneous HCV clearance.

Acknowledgements

Contributions: G.J.D., G.V.M., M.H., and J.M.K. designed the original ATAHC study and wrote the protocol. J.G., S.T.P., G.V.M., K.P., J.M.K., G.J.D., and P.A.W. designed the HCV reinfection and superinfection study. S.T.P. and R.A.B. performed all laboratory work. J.G. drafted the primary statistical analysis plan, which was reviewed by K.P., S.T.P., G.J.D., G.V.M., and P.A.W. The primary statistical analysis was conducted by J.G. and K.P. All authors reviewed data analysis. J.G., S.T.P., G.D., and P.A.W. wrote the first draft; all authors contributed to and approved the final article.

Ancillary