Influence of Hepatitis C Virus Co-Infection and Hepatitis C Virus Treatment on Risk of Chronic Kidney Disease in HIV-Positive Persons

Amanda Mocroft; Lene Ryom; Cristiana Oprea; Qiuju Li; Andri Rauch; Christoph Boesecke; Vilma Uzdaviniene; Dalibor Sedlacek; Josep M. Llibre; Karine Lacombe; Lars N. Nielsen; Eric Florence; Inka Aho; Nikoloz Chkhartishvili; János Szlavik; Gordana Dragovic; Clifford Leen; Helen Sambatakou; Therese Staub; Montse Laguno; Hila Elinav; Janez Tomažič; Lars Peters


AIDS. 2020;34(10):1485-1495. 

In This Article


Of 22 826 persons enrolled in EuroSIDA, 6806 were excluded because of unknown HCV status, insufficient follow-up or with CKD before baseline. An additional 1266 persons were excluded with unknown HCV-RNA status for those who were anti-HCV positive, or with missing baseline CD4+ cell counts and viral load. Compared to the 14 754 included, the 1266 excluded were less likely to be MSM, were less likely to be from Central, or West Europe and more likely to be from Central East, Eastern Europe or Argentina compared with southern Europe. They were also less likely to have suppressed HIV viral load and more likely to have a prior AIDS diagnosis (all P < 0.05).

Table 1 shows the characteristics of the 14 754 included persons, stratified by baseline HCV strata. The five HCV strata were quite heterogenous, and there were many significant differences across the groups (see footnote to Table 1). As would be expected, the proportion of injecting drug uses (IDUs) was lowest in those anti-HCV-negative, the proportion with prior ESLD (only three persons had a prior diagnosis of hepatorenal syndrome) was highest in those HCV-RNA-positive after treatment and the burden from F3/F4 liver fibrosis was the highest in both those cured and HCV-RNA-positive after treatment, as was the proportion who had received tenofovir disoproxil fumarate (TDF) at baseline. The median age was 43 years [interquartile range (IQR) 37--51], baseline CD4+ cell count was 470 cells/μl (IQR 318–669) and CD4+ nadir 174 cells/μl (IQR 70–281). One thousand, seven hundred and sixty-four persons had been previously treated for HCV; the majority of these (1467; 83.2%) had been treated with interferon with ribavirin. At baseline, 181 had received a DAA with interferon, and 275 had received DAAs without interferon.

The analysis included 280 022 eGFRs with a median of 16 (IQR 8–28) per person and 2.4 (IQR 1.9–3.0) per year of follow-up. The number of measures per person per year were similar across the five HCV strata, ranging from 2.2/year (IQR 1.7–3.0) in spontaneous clearers to 2.4 per year in those anti-HCV-negative, those cured and those HCV-RNA-positive after treatment. The median eGFR at baseline was 99 ml/min per 1.73 m2 (IQR 85--110). Four thousand, four hundred and twenty (30%) were at low risk of CKD using the D:A:D risk score, 5089 (34.5%) were at medium risk and 5243 (35.5%) were at high risk, with significant differences between HCV strata. At baseline, 2842 of those anti-HCV-negative were at high risk (30.6%), increasing to 545 of those cured (59.1%) and 425 in those HCV-RNA-positive after treatment (50.5%).

The Incidence of Chronic Kidney Disease in Hepatitis C Virus Strata

During 115 335 PYFU; a median 7 (IQR 3.7–12.4) per person, 1130 (7.7%) developed CKD; the crude incidence rate per 1000 person-years of follow-up was 9.8 [95% confidence interval (CI) 9.2–10.4]. Table 2 shows the crude incidence rate in each of the HCV strata. The incidence rate was lowest in those HCV-RNA-positive following treatment; incidence rate 7.7/1000 PYFU (95% CI 5.2–10.1) and highest in those cured; 12.9/1000 PYFU (95% CI 10.4–15.3). Figure 1 shows the multivariate incidence rate ratios of CKD compared with those cured. After adjustment (model 3, adjusting separately for the components of the D:A:D CKD risk score, liver fibrosis stage at baseline and use of integrase inhibitors) those anti-HCV negative [adjusted incidence rate ratio (aIRR) 0.50; 95% CI 0.39–0.63] and spontaneous clearers (aIRR 0.67; 95% CI 0.47–0.97) had significantly lower rates of CKD compared with those cured. Those chronically infected (aIRR 0.85; 95% CI 0.65–1.12) and HCV-RNA-positive after treatment (aIRR 0.71; 95% CI 0.49–1.04) had nonsignificant reduced rates of CKD compared with those cured.

Figure 1.

Multivariate incidence rate ratios of chronic kidney disease. All factors are included at baseline with the exception of HCV group. *Model additionally adjusted for starting integrase inhibitors as a time-updated variable. HCV, hepatitis C virus.

The proportion of follow-up time with eGFR greater than 90 ml/min per 1.73 m2 was 62.5%, and was highest in those with chronic infections (69.2%) and lowest in those cured (55%). Of 1128 who developed CKD, 926 (82.1%) had at least two further eGFRs and 3 months follow-up. Of these 926, 442 (47.7%) had a reversal of CKD during subsequent follow-up. By 12 months after CKD, 17.2% were estimated to have reversed CKD (95% CI 14.7–19.7) from Kaplan--Meier estimates, with no differences between the HCV strata at development of CKD (P = 0.56). The proportion who reversed CKD was lowest overall for those cured (23/72, 31.9%) and highest for those chronically infected (53/102, 52.0%), but this was not statistically significant (P = 0.083). The median eGFR at CKD was 53.4 (IQR 47.2–57.0 ml/min per 1.73 m2) and was lowest in those chronically infected (median 50.4, IQR 44.2–56.3 ml/min per 1.73 m2), and highest in those anti-HCV-negative (median 53.6, IQR 48.2–57.0 ml/min per 1.73 m2).

Sensitivity Analyses

The results from a wide range of sensitivity analyses showed similar results. Of note, an analysis excluding those with F3/F4 or unknown liver fibrosis at baseline included 442 events during 52 085 PYFU (incidence of CKD 8.5/1000 PYFU; 95% CI 7.7–9.3) and showed similar results; albeit with wider confidence intervals. In this analysis, those anti-HCV-negative had significantly reduced rates of CKD (aIRR 0.65; 95% CI 0.47–0.89) compared with those cured, with no significant differences between other groups (left hand side; Figure 2).

Figure 2.

Univariate and multivariate* incidence rate ratios of chronic kidney disease: sensitivity analyses. *Adjusted for eGFR, use of nephrotoxic antiretroviral, AIDS, hypertension, diabetes, baseline CD4+, age, liver fibrosis stage and baseline date, all at baseline and starting integrase inhibitors as a time-updated variable. eGFR, estimated glomerular filtration rate.

Our results were also consistent when we investigated separately HCV treatments including interferon or DAAs in those treated and cured or HCV-RNA-positive after treatment, with limited power in the latter analysis. There were 1068 events during 111 228 PYFU when DAA treatments were excluded from those cured or HCV-RNA-positive after treatment with an overall incidence rate of 9.6 (9.0–10.3), and the results are shown in the middle panel of Figure 2. Similarly, when only including DAA treatments in those cured or HCV-RNA-positive after treatment, there were 1036 events during 105 291 PYFU, and the results are shown on the right hand side of Figure 2. In this analysis, those anti-HCV-negative had significantly lower rates of CKD and those with spontaneous clearance had marginally lower rates of CKD compared with those cured.

Having a more stringent definition for CKD of a confirmed 25% decline in eGFR to less than 60 ml/min per 1.73 m[2] resulted in a lower incidence of CKD (1001 events during 116 369 PYFU, rate 8.6/1000 PYFU; 95% CI 8.1–9.1), but also showed a lower incidence of CKD in those anti-HCV-negative), consistent with our main findings.

Characteristics of Hepatitis C Virus-treated Persons at Chronic Kidney Disease or Last Visit

Our final analysis focused further on those treated for HCV. Characteristics of persons at CKD or last visit for those not developing CKD are shown in Table 3. Of note, there was a much higher proportion of persons with ESLD in those cured who developed CKD, likely reflecting targeted treatment to those with most advanced liver disease when DAAs first became available. As would be expected, those cured had a much higher proportion of people who had received DAA treatment compared with those HCV-RNA-positive after treatment, regardless of whether they developed CKD or not.