Predicting spontaneous clearance of acute hepatitis C virus in a large cohort of HIV-1-infected men

Abstract

Objective: An epidemic of acute hepatitis C virus (HCV) infection in HIV-positive men-who-have-sex-with-men (MSM) is emerging in Europe, Australia and the USA. The aim of this study was to characterise the natural history of primary HCV in this setting and to assess host and viral factors which predict spontaneous clearance.

Methods: This prospective longitudinal cohort study was carried out in 112 HIV-positive patients who were followed in a single centre (the St Mary's Acute HCV Cohort). Plasma and peripheral blood mononuclear cells (PBMCs) were obtained at monthly intervals for 3 months and at 3-monthly intervals thereafter for a median of 45 months (IQR = 29-69 months). The primary end point was spontaneous clearance of HCV. Cox regression was used to assess the impact of clinical and virological variables on outcome, including liver function, CD4 count, rate of HCV RNA decline, T cell response and clonal sequence evolution within the HCV E2 envelope gene.

Results: 15% of patients cleared HCV spontaneously, while 85% progressed towards chronicity. The latter group included a significant proportion of 'fluctuating' progressors (37.5%), in whom a fall followed by a rise (>1 log₁₀) in viraemia was observed. This was associated with superinfection with new HCV strains and partially effective T cell responses. Spontaneous clearance was strongly associated with a 2.2 log₁₀ viral load drop within 100 days of infection (HR = 1.78; p < 0.0001), elevated bilirubin (≥ 40 μmol/l; HR = 5.04; p = 0.006), elevated alanine aminotransferase (ALT; ≥ 1000 IU/ml; HR = 2.62; p = 0.048) and baseline CD4 count ≥ 650 × 10⁶/l (HR = 2.66; p = 0.045), and only occurred in patients with genotype 1 infection. Evolution to spontaneous clearance occurred in patients with low viral diversity in the presence of an early multispecific T cell response.

Conclusions: Spontaneous clearance of acute HCV in HIV-positive men can be predicted by a rapid decline in viral load, high CD4 count, elevated bilirubin and ALT, and is associated with low viral diversity and strong T cell responses.

Figure 1

Patient cohort diagnosis, treatment and follow-up. A total of 112 HIV-positive patients with acute hepatitis C virus (HCV) were recruited to the the St Mary's Acute HCV Cohort study of whom 17 spontaneously cleared HCV. Treatment with pegylated interferon α (IFNα) and ribavirin was given for 48 weeks to 62 patients and withheld or delayed in a further 26. The sustained virological response rate was 66%. Ab, antibody; LFTs, liver function tests; RT-PCR, reverse transcription–PCR.

Figure 2

Clinical measures and outcome in (A) the whole cohort and (B) representative individual patients. Longitudinal hepatitis C virus (HCV) viral load measurements taken prior to treatment revealed three patterns of infection: spontaneous clearance (SC; 15%, red) and two patterns of progression (85%); peak and plateau (PV; 47.5%, black) or fluctuating viraemia (FV; 37.5%, blue) defined as a viral load peak followed by a decline (>1 log₁₀) and subsequent rise in viraemia (>1 log₁₀). Viral loads (VLs) obtained from spontaneous clearers (red), plateau progressors (black) and fluctuating progressors (blue) groups were pooled and plotted using a linear fit model from the time of peak viraemia to 300 days (insert). VL data were right-censored at the time of treatment for HCV. Grey shading indicates 95% CIs.

Figure 3

Patterns of acute HCV based on phylogenetic analysis in subjects 1–10. (A) Spontaneous clearance and plateau progression, (B) fluctuating viraemia associated with superinfection and (C) reactivation. A circular neighbour-joined phylogenetic tree analysis (Kimura 2-parameter model) of 305 HVR-1 sequences derived from subjects 1–10 at 2–4 time points is shown in order to illustrate patterns observed in different patient groups. Similar trees were created for 50 patients. Clones derived from each patient are given a different colour and those from two representative patients are shown in further detail (filled circles represent time point 1, clear circles time point 2, filled triangles time point 3 and clear triangles time point 4). (A) A gradual increase in genetic diversity can be seen in the progressor patient (black), while clones derived from the spontaneous clearer (red) reveal a narrower repertoire. Bootstrap values greater than 70% are shown. (B) A patient with evidence of superinfection is highlighted. Sequences from time point 4 (clear triangles, red brackets) consist of a dominant superinfecting strain (one clone related to the previous lineage is also present). This patient had a fluctuating viral load, with a negative HCV PCR (limit of detection=12 IU/ml) between time points 3 and 4. (C) A case of HCV reactivation is shown. The patient had a sustained virological response (PCR negative 6 months post-treatment) following 48 weeks of IFNα and ribavirin. One year later, he developed a rapid decrease in CD4 count (nadir 106x10/l) and was hospitalised with Pneumocystis jeroveci pneumonia. During this time, his ALT became elevated and HCV RNA was once again detected in his blood. He denied any risk factors for reacquisition of HCV. Branches derived from patient HVR-1 sequences (n=56) are coloured in filled green (baseline sample) and open green (3 years later). These samples derive from the same lineage.

Figure 4

T cell responses and genetic variation within the E2 hypervariable region-1 (HVR-1) over time. Cumulative genetic variation (corrected Hamming distance and dN/dS ratio (non-synonymous substitutions per non-synonymous site divided by synonymous substitutions per synonymous site)) and T cell responses (interferon γ (IFNγ) spot-forming units (SFU)/10⁶ cells) were assessed in each group of patients at two time points (sample 1, range 28–189 days; and sample 2, range 115–246 days from the last negative PCR test).

Figure 5

T cell responses across the hepatitis C virus (HCV) genome at baseline and 3–6 months. (A) Total T cell responses and (B) T cell responses across the HCV genome. T cells responses to pooled peptides spanning the entire HCV genome were assessed by ELISpot in 40 patients with genotype 1a infection at two time points. Significant results (p<0.05) are highlighted with asterisks (red indicates significant difference between SC and both FV and PV patients; green indicates significant difference between SC and PV patients). These responses were measured in total (A) and to 10 overlapping peptide pools (B) from across the entire HCV genome.