Abstract
Background
Human immunodeficiency virus (HIV)–infected patients are at increased risk of liver-related mortality. The effect of occult cirrhosis (OcC), defined as preclinical compensated cirrhosis without any clinical findings, on liver-related events is unknown.
Methods
HIV-infected patients from 2 Canadian cohorts underwent transient elastography (TE) examination and were classified as (1) OcC (TE ≥13 kPa with no sign of cirrhosis, including absence of thrombocytopenia and signs of advanced liver disease on ultrasound or gastroscopy); (2) overt cirrhosis (OvC) (TE ≥13 kPa with signs of cirrhosis); or (3) noncirrhotic patients (TE <13 kPa). Incidence and risk factors of liver-related events were investigated through Kaplan-Meier and Cox regression analyses, respectively. We estimated monitoring rates according to screening guidelines for hepatocellular carcinoma (HCC) by OcC and OvC status.
Results
A total of 1092 HIV-infected patients (51% coinfected with hepatitis C virus) were included. Prevalence of OcC and OvC at baseline was 2.7% and 10.7%, respectively. During a median follow-up of 1.8 (interquartile range, 1.5–2.8) years, the incidence of liver-related events in noncirrhosis, OcC, and OvC was 3.4 (95% confidence interval [CI], 1.2–7.3), 34.0 (95% CI, 6.0–104.0), and 37.0 (95% CI, 17.0–69.1) per 1000 person-years, respectively. Baseline OcC (adjusted hazard ratio [aHR], 7.1 [95% CI, 1.3–38.0]) and OvC (aHR, 8.5 [95% CI, 2.8–26.0]) were independently associated with liver-related events. Monitoring rates for HCC were lower in patients with OcC (24%) compared to those with OvC (40%).
Conclusions
HIV-infected patients with OcC have a high incidence of liver-related events. Greater surveillance and earlier recognition with appropriate screening strategies are necessary for improved outcomes.
Keywords: HIV, occult cirrhosis, transient elastography, liver-related events, HCC surveillance
Occult cirrhosis affects 2.7% of HIV-infected patients, representing 20% of all cirrhosis cases. HIV-infected patients with occult cirrhosis have the same risk of liver-related events as those with clinically overt cirrhosis. HIV-infected patients with occult cirrhosis should receive greater surveillance for hepatocellular carcinoma.
Liver disease is the leading cause of non-AIDS-related deaths in people living with human immunodeficiency virus (HIV) [1]. HIV-infected individuals have multiple risk factors for liver injury, including coinfections with hepatitis C (HCV) and hepatitis B (HBV) viruses, metabolic conditions triggering nonalcoholic fatty liver disease (NAFLD), excessive alcohol intake, and antiretroviral therapy (ART) inducing hepatotoxicity [2, 3]. This hypothetical multihit process can then lead to cirrhosis, hepatocellular carcinoma (HCC), and complications associated with end-stage liver disease (ESLD) such as ascites, hepatic encephalopathy, and variceal bleeding [4]. Up to 63% of patients with liver disease are diagnosed with liver cirrhosis only at the first episode of hepatic decompensation [5]. Therefore, identifying patients with underlying liver cirrhosis is critical as surveillance for HCC and esophageal varices can be promptly initiated [6].
A major obstacle to the diagnosis of cirrhosis at the preclinical stage, also known as occult cirrhosis (OcC), is the lack of any clinical, laboratory, and imaging findings. Patients with OcC do not have thrombocytopenia and have no features consistent with cirrhosis on imaging. Nonetheless, OcC is identified in up to 12% of patients referred for transient elastography (TE) examination and accounts for up to 37% of all cirrhosis cases [7, 8]. HIV-negative patients with OcC are at risk of developing ESLD events, but they receive suboptimal surveillance for HCC when compared to patients with clinically overt cirrhosis (OvC) [7–9].
Liver biopsy is the gold standard for the diagnosis of liver cirrhosis, but it is invasive and prone to sampling errors leading to the misdiagnosis of cirrhosis [10–12]. As such, it is unpractical as a screening tool or for the follow-up of HIV-infected patients, where the prevalence of liver disease is high [13]. The measurement of liver stiffness (LSM) by TE is a validated noninvasive method to diagnose liver cirrhosis, with a reported area under the curve of 0.94 [14]. TE is superior to simple fibrosis biomarkers, including the aspartate aminotransferase-to-platelet ratio index (APRI) and the Fibrosis-4 Index for Liver Fibrosis score (FIB-4), in HIV-infected patients [15–17].
Thus far, the prevalence of OcC in HIV-infected patients and its impact on incident liver-related events has not been reported. We used data from 2 large clinical Canadian cohorts to fulfill the following aims: (1) determine the prevalence and associated factors of OcC diagnosed by TE; (2) evaluate the incidence of liver-related events in patients with OcC; and (3) investigate monitoring rates, according to guidelines, for screening of HCC in patients with OcC.
PATIENTS AND METHODS
Study Design and Population
We conducted a retrospective analysis of the Canadian Coinfection Cohort (CCC) and the LIVEr disease in HIV (LIVEHIV) studies [18, 19]. The CCC is a prospective cohort of patients coinfected with HIV and HCV enrolled at 18 centers across Canada where they are followed up every 6 months since 2003. Centers participating in the FibroScan substudy assessed LSM by TE examination every 6 months. The LIVEHIV Cohort is a prospective screening program for NAFLD and liver fibrosis established in 2013 at McGill University Health Centre (MUHC), Montreal, Canada. All patients underwent screening for liver disease with yearly TE examination.
Ethics
All participants provided informed written consent. The research ethics board of the Research Institute of MUHC approved the study (codes 14-182-BMD and 2006-1875), which was conducted according to the Declaration of Helsinki.
Eligibility
All consecutive HIV-infected adults aged ≥18 years were eligible. Patients in the CCC who were enrolled in centers not participating in the FibroScan substudy were excluded. Patients with failure of TE examination or unreliable LSM were also excluded.
Clinical and Biological Parameters
Data collected included demographic information, HIV and medication history, body mass index (BMI), history of type 2 diabetes mellitus [20], liver biochemistries, and hematological and virological parameters. Alcohol intake was measured by the Alcohol Use Disorders Identification Test (AUDIT-C) questionnaire, with a score ≥4 for men and ≥3 for women considered as hazardous alcohol intake [21]. The following fibrosis biomarkers were computed: APRI (with a cutoff value ≥2 suggesting liver cirrhosis) and FIB-4 (with a cutoff value ≥3.25 suggesting advanced liver fibrosis) [17, 22]. Results of abdominal imaging and gastroscopies were recorded.
TE Examination
The TE examination was performed in patients fasting for at least 3 hours. The standard M probe was used in all patients. The XL probe was used in cases of failure of TE with the M probe or if the BMI was >30 kg/m2. The following criteria were applied to define the result of TE as reliable: at least 10 validated LSM and an interquartile range (IQR) <30% of the median LSM [23]. Hepatic steatosis measurement by controlled attenuation parameter was available in the LIVEHIV Cohort. Any grade hepatic steatosis (>5% of hepatocytes) was defined as controlled attenuation parameter ≥288 decibels per meter [24].
Definition of Study Groups
Patients were divided into 4 mutually exclusive subgroups. Patients with OcC had LSM ≥13 kPa (10.1 kPa with the XL probe [25]) and no evidence of advanced liver disease or portal hypertension—namely, absence of thrombocytopenia (<140 × 109/L), cirrhosis on abdominal imaging (liver nodularity, enlarged caudate lobe, splenomegaly ≥14 cm), esophageal varices or hypertensive portal gastropathy on gastroscopy, and ascites. Patients with OvC had LSM ≥13 kPa with evidence of advanced liver disease or portal hypertension, as listed above. Patients classified as noncirrhotic had LSM <13 kPa and no evidence of advanced liver disease or portal hypertension. Finally, patients classified as noncirrhotic portal hypertension had LSM <13 kPa and evidence of advanced liver disease or portal hypertension. The 13-kPa cutoff has been chosen based on a meta-analysis [14].
Outcome Measures
Liver-related events were prospectively collected by means of a dedicated outcome measures form. Liver-related events included occurrence of de novo ascites, variceal bleeding or banding, spontaneous bacterial peritonitis, hepatorenal syndrome, hepatic encephalopathy, HCC, liver transplantation, and death from liver-related causes. Screening interval for HCC was defined as receiving abdominal imaging at least every 12 months during the study period [26, 27].
Follow-up
Patients were followed from their first TE examination (baseline, or time zero) until they had a liver-related event, died, withdrew consent, were lost to follow-up (no visits for more than 1.5 years), or until administrative censoring (31 March 2017 for the CCC and 14 January 2017 for the LIVEHIV Cohort). Non-liver-related deaths were censored. The decision to initiate surveillance for HCC was left to the treating physician.
Statistical Analysis
Cross-sectional Component
The period prevalence was reported as the ratio of the total number of prevalent and incident cases to the total number of patients included in the study. The associations between various risk factors at inclusion and prevalent or incident OcC—among patients at risk for OcC—were assessed using a cross-sectional approach with unadjusted and adjusted logistic regression models. We reported results as adjusted odds ratios. Due to the cross-sectional nature of any such measured associations, we made no causal claims from these analyses.
Longitudinal Component
Kaplan-Meier plots were used to show the cumulative incidences over time, as stratified by study group and coinfection status, using the log-rank test to test for differences between groups. The association between various risk factors and the incidence of de novo liver-related events were assessed with unadjusted and adjusted Cox proportional hazards models. Results were reported as adjusted hazard ratios with 95% confidence intervals (CIs). Finally, as an exploratory analysis, the proportion of patients—among those at risk for HCC—with adequate cirrhosis surveillance were reported descriptively and stratified by study group. The same analysis was stratified by source cohort and by HCV coinfection status. A 2-sided level of significance of 5% was used for all statistical inferences.
Covariates
All adjusted regression models included covariates that were determined a priori to be clinically important; this list was further restricted due to limited event counts. The following baseline covariates were chosen: age, sex, aboriginal ethnicity, hazardous alcohol intake in past year, BMI, diabetes in past year, lipid-lowering therapy in the past year, nadir CD4 cell count, detectable HIV RNA in the past year, HCV RNA positive, HIV duration, and ART duration. Where multiple models were estimated for the same outcome, the model with the best goodness-to-fit measure, based on the Akaike information criterion, was chosen.
RESULTS
After applying the inclusion and exclusion criteria, 1092 HIV-infected individuals were included, of whom 499 were HIV monoinfected, 556 were HIV/HCV coinfected, and 37 were HIV/HBV coinfected (Figure 1). Failed or unreliable LSM comprised 7.0% of patients from the CCC and 5.8% of patients from the LIVEHIV Cohort.
Figure 1.
Flowchart displaying the selection of study participants, stratified by study cohort. 1Administrative censoring was applied on 31 March 2017 for the CCC and on 14 January 2017 for the LIVE HIV Cohort. 2Liver stiffness measurements by transient elastography were considered reliable if the ratio of the interquartile range over the median of the 10 measures was no more than 30%. Abbreviations: CCC, Canadian Coinfection Cohort; LIVEHIV, LIVEr disease in HIV study.
Prevalence, Incidence, and Factors Associated With Occult Cirrhosis
OcC and OvC represented 2.7% and 10.7% of our study population, respectively. OcC patients represented 19.9% of all patients with cirrhosis. The main characteristics by study group status at baseline are presented in Table 1. Figure 2 and Supplementary Table 1 report the distribution of study groups and the main characteristics by HCV and HBV coinfection status, respectively. The prevalence of OcC increased during the study period with 14 new cases, for an overall pooled period prevalence of 3.9%. Furthermore, over a median follow-up time of 1.8 (IQR, 1.5–2.8) years, the incidence rate of OcC was 10.0 (95% CI, 5.7–16.3) per 1000 person-years (PY). HIV-monoinfected patients showed similar incidence of OcC compared with HCV-coinfected patients. Conversely, those coinfected with HCV were more likely to develop OvC during follow-up compared to monoinfected patients (Figure 3A and 3B). Logistic regression identified HIV duration as an independent factor associated with the presence of OcC as compared to noncirrhotic patients (adjusted odds ratio, 1.44 [95% CI, 1.18–1.77]; Table 2). We also observed a tendency for higher BMI to be a factor associated with OcC.
Table 1.
Patient Characteristics at Baseline by Study Group Status (N = 1092)
| Characteristic | Noncirrhotic (n = 795) | Occult Cirrhosis (n = 29) |
Overt Cirrhosis (n = 117) |
Noncirrhotic Portal Hypertension (n = 151) | P Value |
|---|---|---|---|---|---|
| Follow-up time, y | 1.53 (1.44–2.60) | 2.62 (1.50–2.96) | 2.48 (1.55–3.09) | 2.36 (1.50–2.96) | <.001 |
| Age, y | 48.7 (40.7–54.8) | 50.5 (45.2–54.8) | 52.6 (48.1–57.5) | 51.5 (45.6–56.6) | <.001 |
| Female sex | 206 (26) | 8 (28) | 25 (21) | 37 (25) | .742 |
| MSM | 258 (32) | 14 (48) | 51 (44) | 54 (36) | .04 |
| Ethnicity | |||||
| White | 339 (43) | 15 (52) | 88 (75) | 97 (64) | <.001 |
| Black | 166 (21) | 3 (10) | 9 (8) | 15 (10) | <.001 |
| Latino | 58 (7) | 1 (3) | 0 | 6 (4) | .01 |
| Asian | 20 (3) | 3 (10) | 3 (3) | 3 (2) | .073 |
| Indigenous | 125 (16) | 5 (17) | 12 (10) | 17 (11) | .251 |
| Unknown | 93 (12) | 3 (10) | 5 (4) | 15 (10) | .111 |
| BMI, kg/m2 | 25.1 (22.7–28.4) | 26.8 (24.1–30.7) | 25.2 (22.1–27.5) | 24.3 (21.0–26.8) | .005 |
| Obesity (BMI >30 kg/m2) | 106 (13) | 6 (21) | 15 (13) | 12 (8) | .17 |
| Hazardous alcohol intake | 104 (13) | 4 (14) | 23 (20) | 33 (22) | .02 |
| Active tobacco smoker | 266 (33) | 11 (38) | 62 (53) | 80 (53) | <.001 |
| Active IDU | 197 (25) | 5 (17) | 44 (38) | 64 (42) | <.001 |
| HIV duration, y | 11 (6–19) | 19 (14–26) | 19 (12–24) | 16 (9–22) | <.001 |
| CD4 nadir, cells/μL | 209 (99–350) | 121 (28–290) | 153 (89–277) | 180 (105–302) | .039 |
| CD4 cell count, cells/μL | 514 (340–720) | 526 (416–673) | 406 (272–606) | 536 (367–719) | .004 |
| HIV viral load, units, log10 | 1.59 (1.59–1.63) | 1.59 (1.59–1.60) | 1.59 (1.59–1.59) | 1.59 (1.59–1.59) | .434 |
| Undetectable HIV viral load (≤50 copies/mL) | 590 (74) | 22 (76) | 90 (77) | 118 (78) | .949 |
| Duration of ART, y | 7.2 (3.2–12.9) | 11.4 (4.6–15.8) | 11.1 (5.9–17.2) | 9.0 (4.8–16.2) | <.001 |
| Current ART used | 655 (82) | 20 (69) | 103 (88) | 134 (89) | .019 |
| NRTI | 617 (78) | 19 (66) | 97 (83) | 123 (81) | .147 |
| NNRTI | 212 (27) | 6 (21) | 30 (26) | 31 (21) | .410 |
| PI | 329 (41) | 10 (34) | 49 (42) | 71 (47) | .501 |
| Integrase inhibitors | 237 (30) | 8 (28) | 44 (38) | 59 (39) | .066 |
| Other | 59 (7) | 1 (3) | 7 (6) | 11 (7) | .818 |
| Exposure to didanosine, y, mean ± SD | 0.26 ± 1.58 | 0.32 ± 1.17 | 0.26 ± 1.55 | 0.37 ± 1.49 | .648 |
| Exposure to stavudine, y, mean ± SD | 0.56 ± 1.95 | 0.29 ± 0.95 | 0.23 ± 1.09 | 0.69 ± 2.26 | .266 |
| HCV coinfection | 337 (42) | 18 (62) | 99 (85) | 102 (68) | <.001 |
| HBsAg positive | 16 (2) | 1 (3) | 6 (5) | 14 (9) | <.001 |
| Hypertension | 125 (16) | 9 (31) | 25 (21) | 24 (16) | .082 |
| Diabetes | 145 (18) | 7 (24) | 27 (23) | 31 (21) | .526 |
| INR | 1.00 (0.93–1.03) | 1.02 (1.00–1.13) | 1.09 (1.00–1.15) | 1.00 (0.98–1.07) | <.001 |
| Platelets, 109/L | 212 (182–259) | 208 (171–241) | 130 (98.8–189) | 133 (113–180) | <.001 |
| Total cholesterol, mmol/L | 4.5 (3.8–5.2) | 3.8 (3.5–4.5) | 3.9 (3.1–4.4) | 4.1 (3.6–4.9) | <.001 |
| LDL-C, mmol/L | 2.6 (2.0–3.2) | 1.7 (1.2–2.7) | 2.0 (1.5–2.5) | 2.3 (1.8–2.8) | <.001 |
| HDL-C, mmol/L | 1.2 (1.0–1.4) | 0.9 (0.7–1.2) | 1.1 (0.9–1.3) | 1.1 (0.9–1.3) | <.001 |
| Triglycerides, mmol/L | 1.3 (0.9–2.0) | 1.7 (1.1–2.5) | 1.3 (0.9–1.7) | 1.3 (0.9–2.1) | .300 |
| Creatinine, µmol/L | 80 (68–93) | 78.0 (66.0–104) | 82.5 (68.8–94.8) | 77 (68–89) | .794 |
| Insulin, pmol/L | 75 (47–137) | 148 (112–262) | 92 (60–190) | 66 (42–145) | <.001 |
| ALT, IU/L | 29 (20–47) | 41 (29–84) | 45 (29–79) | 33 (20–52) | <.001 |
| AST, IU/L | 28 (21–39) | 46 (29–55.2) | 51 (32–80) | 32 (23–48) | <.001 |
| Total bilirubin, µmol/L | 10 (7–15) | 12 (9.35–18.5) | 14 (9–20.1) | 11 (8–17) | < .001 |
| Albumin, g/L | 41 (39–44) | 41 (37–44) | 40.5 (36.2–43) | 42 (39–45) | .007 |
| APRI ≥2 | 13 (2) | 4 (14) | 33 (28) | 10 (7) | < .001 |
| FIB-4 ≥3.25 | 16 (2) | 4 (14) | 47 (40) | 27 (18) | < .001 |
| CAP, dB/m | 228 (192–259) | 248 (212–292) | 208 (172–237) | 218 (174–259) | .006 |
| CAP ≥288 | 82/558 (15) | 6/22 (27) | 5/56 (9) | 16/97 (16) | .217 |
Continuous variables are expressed as median (interquartile range) and categorical variables are expressed as frequencies (%), unless otherwise specified. For the 3-way comparison of patients with no cirrhosis, occult cirrhosis, or overt cirrhosis, P values were computed using Fisher exact test for dichotomous variables, χ2 test for categorical variables, and a Kruskal-Wallis test for continuous variables, and are considered significant when <.05. CAP was available in 733 of 1092 patients.
Abbreviations; ALT, alanine aminotransferase; APRI, aspartate aminotransferase-to-platelet ratio index; ART, antiretroviral therapy; AST, aspartate aminotransferase; BMI, body mass index; CAP, controlled attenuation parameter; FIB-4, Fibrosis-4 Index for Liver Fibrosis score; HBsAg, hepatitis B surface antigen; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HDL-C, high-density lipoprotein cholesterol; IDU, injection drug use; INR, international normalized ratio; LDL-C, low-density lipoprotein cholesterol; MSM, men who have sex with men; NNRTI, nonnucleoside reverse transcriptase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; SD, standard deviation.
Figure 2.
Distribution of study groups according to hepatitis C virus and hepatitis B virus coinfection status. Abbreviations: HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; OcC, occult cirrhosis; OvC, overt cirrhosis.
Figure 3.
Survival curves. A, Incidence of occult cirrhosis by hepatitis C virus (HCV) coinfection status. B, Incidence of overt cirrhosis by HCV coinfection status. C, Probability of liver-related events by study group category. D, Probability of liver-related events by HCV coinfection category. Log-rank P values are shown. Abbreviation: HIV, human immunodeficiency virus.
Table 2.
Factors Associated With Occult Cirrhosis by Logistic Regression Analysis
| Variable | OR (95% CI) |
Model 1 aOR (95% CI) |
Model 2 aOR (95% CI) |
|---|---|---|---|
| Age (per 10 years) | 1.36 (.99–1.88) | 1.45 (1.03–2.06) | … |
| Female sex (yes vs no) | 0.97 (.42–2.03) | 1.07 (.46–2.28) | 1.19 (.51–2.57) |
| Indigenous ethnicity (yes vs no) | 0.51 (.12–1.46) | … | … |
| Hazardous alcohol intake in past year (yes vs no) | 1.04 (.35–2.51) | … | … |
| BMI (per 5 units) | 1.19 (.9–1.5) | 1.20 (.92–1.52) | 1.25 (.95–1.58) |
| Diabetes in past year (yes vs no) | 1.67 (.73–3.52) | … | … |
| Lipid-lowering therapy in past year (yes vs no) | 1.12 (.41–2.57) | … | … |
| Nadir CD4 count (per 100 cells) | 0.94 (.77–1.12) | … | … |
| Detectable HIV RNA in past year (yes vs no) | 0.69 (.2–1.79) | … | … |
| HCV coinfection (yes vs no) | 1.21 (.61–2.36) | 1.46 (.72–2.94) | 1.30 (.65–2.56) |
| HIV duration (per 5 years) | 1.39 (1.15–1.69) | … | 1.44 (1.18–1.77) |
| ART duration (per 5 years) | 1.14 (.88–1.47) | … | … |
| AIC | … | 296.55 | 288.33 |
Occult cirrhosis patients (n = 29) are compared to noncirrhotic patients (n = 795).
Abbreviations: AIC, Akaike information criterion; aOR, adjusted odds ratio; ART, antiretroviral therapy; BMI, body mass index; CI, confidence interval; HCV, hepatitis C virus; HIV, human immunodeficiency virus; OR, odds ratio.
Incidence and Risk Factors of Liver-related Events
The overall incidence of liver-related events was 8.6 (95% CI, 4.9–14.0) per 1000 PY, mainly driven by the occurrence of de novo ascites and HCC (Table 3). The incidence rate of liver-related events was similar between patients with OcC and OvC (Table 3; Figure 3C). Conversely, noncirrhotic patients had a very low incidence of liver-related events, whereas patients with noncirrhotic portal hypertension had an intermediate incidence. HCV coinfection showed a tendency to be associated with an increased incidence of liver-related events (Figure 3D). The presence of OcC and OvC at baseline, as well as longer HIV duration, were independent risk factors for liver-related events, with adjusted hazard ratios of 7.12 (95% CI, 1.33–37.99), 8.5 (95% CI, 2.83–25.53), and 1.37 (95% CI, 1.09–1.73), respectively (Table 4).
Table 3.
Incidence Rate of Liver-related Events During Follow-up
| Event |
Noncirrhotic (n = 795) |
Occult Cirrhosis (n = 29) |
Overt Cirrhosis (n = 117) |
Noncirrhotic Portal Hypertension (n = 151) | ||||
|---|---|---|---|---|---|---|---|---|
| Cases | Incidence rate (95% CI) |
Cases | Incidence rate (95% CI) |
Cases | Incidence rate (95% CI) |
Cases | Incidence rate (95% CI) |
|
| Any liver event | 5 | 3.4 (1.2–7.3) | 2 | 34.0 (6.0–104.0) | 8 | 37.0 (17.0–69.1) | 4 | 12.7 (3.94–29.5) |
| Ascites | 1 | 0.68 (.04–2.97) | 2 | 33.34 (5.54–102.89) | 5 | 21.48 (7.7–46.16) | 3 | 9.51 (2.37–24.66) |
| Varices with bleeding or banding | 2 | 1.35 (.22–4.16) | 0 | … | 0 | … | 1 | 3.08 (.18–13.55) |
| Spontaneous bacterial peritonitis | 1 | 0.67 (.04–2.96) | 0 | … | 1 | 3.83 (.22–16.85) | 0 | … |
| Hepatic encephalopathy | 0 | … | 0 | … | 3 | 11.56 (2.87–29.97) | 0 | … |
| Hepatorenal syndrome | 0 | … | 0 | … | 1 | 3.71 (.21–16.32) | 0 | … |
| HCC | 1 | 0.67 (.04–2.96) | 1 | 15.49 (.88–68.13) | 3 | 11.61 (2.89–30.08) | 1 | 3.08 (.18–13.55) |
| ESLD-related death | 0 | … | 1 | 15.33 (.87–67.44) | 1 | 3.71 (.21–16.31) | 0 | … |
Incidence rates are presented as 1000 person-years.
Abbreviations: CI, confidence interval; ESLD, end-stage liver disease; HCC, hepatocellular carcinoma.
Table 4.
Risk Factors Associated With Any Liver Event by Cox Proportional Hazards Regression Analysis
| Variable | HR (95% CI) | aHR (95% CI) |
|---|---|---|
| Occult cirrhosis at baseline (yes vs no) | 4.45 (.96–20.62) | 7.12 (1.33–37.99) |
| Overt cirrhosis at baseline (yes vs no) | 8.49 (3.03–23.75) | 8.5 (2.83–25.53) |
| Age (per year) | 1.04 (1.01–1.07) | … |
| Female sex (yes vs no) | 0.79 (.22–2.78) | … |
| Hazardous alcohol intake in the past year (yes vs no) |
2.97 (1.05–8.39) | … |
| BMI ≥30 kg/m2 in past year (yes vs no) | 0.95 (.22–4.17) | … |
| Diabetes in past year (yes vs no) | 1.92 (.61–5.98) | … |
| Lipid-lowering therapy in past year (yes vs no) | 0.46 (.06–3.47) | … |
| Nadir CD4 count (per 100 cells) | 0.79 (.59–1.06) | … |
| Detectable HIV RNA in past year (yes vs no) | 1.71 (.54–5.38) | 1.55 (.51–4.67) |
| HCV coinfection (yes vs no) | 1.83 (.65–5.18) | … |
| HIV duration (per 5 years) | 1.55 (1.23–1.96) | 1.37 (1.09–1.73) |
| ART duration (per 5 years) | 1.63 (1.14–2.32) | … |
Abbreviations: aHR, adjusted hazard ratio; ART, antiretroviral therapy; BMI, body mass index; CI, confidence interval; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HR, hazard ratio.
Surveillance for HCC
Patients with OcC received 0.70 (95% CI, .51–.92) abdominal imaging visits per PY, compared to 0.87 (95% CI, .76–.98) for those with OvC. Furthermore, the monitoring rate of HCC according to screening guidelines [26, 27] was 24% (95% CI, 10%–44%) and 40% (95% CI, 30%–49%) in those with OcC and OvC, respectively (P = .06; Supplementary Table 2). In patients with at least 1 HCC screening imaging, the annual rate of imaging was 0.97 (95% CI, .72–1.28) and 1.33 (95% CI, 1.16–1.51) per patient in those with OcC and OvC, respectively. One patient (HIV/HBV coinfected) with OcC developed HCC during the follow-up, accounting for an incidence of 15 per 1000 PY. At diagnosis, there were 2 lesions consistent with HCC, with the largest one measuring 6 cm and the smaller one 3 cm. This tumor was beyond any transplant criteria, including Milan criteria, total tumor volume, up-to-7 criteria, and University of California, San Francisco criteria [28]. The patient was offered transarterial chemoembolization and Sorafenib but declined. He died 7.8 months after the diagnosis. In the OvC group, the 3 patients who developed HCC were HIV/HCV coinfected and were diagnosed at an earlier stage (single lesion ranging from 1.7 to 3.2 cm). They were offered transarterial chemoembolization (n = 1) or liver transplant (n = 2) and are still alive.
DISCUSSION
In this well-characterized clinical cohort, we showed that OcC is a frequent occurrence, representing nearly 20% of all patients with liver cirrhosis in HIV-infected individuals. We also showed that liver-related events had similar incidence rates in HIV-infected patients with OcC and OvC, rendering it essential for the HIV clinician to be aware of this clinical entity. Finally, patients with OcC do not receive screening at recommended intervals for detection of important ESLD complications, specifically HCC, which may result in a late diagnosis.
Until recently, the description of a liver as “cirrhotic” was sufficient to define patients’ prognosis. Garcia-Tsao et al suggested that cirrhosis is rather a dynamic condition encompassing a more complex clinicopathological spectrum [29]. Given the lack of any clinical sign, patients at the preclinical stage of compensated cirrhosis represent a diagnostic challenge for clinicians, and often remain undiagnosed. The diagnosis of OcC based on LSM and the absence of clinical signs has previously been shown, in different patient populations, to be an indicator of poor prognosis [7, 9]. Individuals living with HIV are known to be at increased risk of liver cirrhosis and ESLD [1]. Our study is the first to report on the prevalence of OcC and its impact on liver-related events in HIV-infected patients.
In our large combined cohort of 1092 HIV-infected patients, OcC was found in 2.7% of cases, a prevalence similar to that reported in other populations [7, 9]. Although seemingly comprising only a small proportion of patients at baseline TE, the prevalence of OcC increased to 3.9% during the follow-up period, with an annual incidence rate of 10 per 1000 PY. These changes in LSM may herald a poor prognosis, as reported in other patient populations [30–32].
The only independent factor associated with OcC at baseline vs noncirrhotic subgroup was longer duration of HIV infection, which is a proxy for exposure to multiple hepatotoxic hits [18]. Interestingly, HCV coinfection was not associated with an increased prevalence or incidence of OcC, but rather with the development of OvC. This could be explained by either a faster progression of HCV-coinfected patients from earlier stage of liver disease to OvC [33] or because they are more frequently monitored with TE, and thus diagnosed more often. Higher BMI had a trend in being an independent risk factor for OcC, thus confirming the emerging concern around NAFLD in HIV-infected patients [18, 34]. Accordingly, hepatic steatosis diagnosed by controlled attenuation parameter had the highest frequency in the OcC group (27%). Taken together, our findings suggest that clinicians should suspect OcC in patients with longer HIV infection and higher BMI even in the absence of any sign of liver cirrhosis. Moreover, screening for liver cirrhosis should not be limited to HIV patients coinfected with HCV, as multiple risk factors may still be driving progressive liver fibrosis and we would miss important opportunities to prevent liver decompensation.
Patients with OcC were more likely to develop a liver-related event when compared to patients without cirrhosis, with a 10 times higher incidence rate. On the other hand, the incidence of a liver-related event was similar between those with OcC and OvC. Our findings are similar to a previous study by Merchante and colleagues where patients with cirrhosis had an incidence of hepatic decompensation of 46 per 1000 PY [32]. Similarly, Macías et al reported that patients with LSM <14.6 kPa have a lower risk of decompensation [35]. However, these studies did not differentiate between patients with and without clinical evidence of liver cirrhosis. Furthermore, they only included HIV/HCV-coinfected cases, so information about dynamics and evolution of OcC in HIV-monoinfected patients was missing. In our study, the diagnosis of OcC at baseline was independently associated with the development of liver-related events. These findings underline that HIV-infected patients with OcC, whether or not HCV coinfected, constitute a group of patients beyond the traditional subset of patients with OvC, who are at high-risk of liver-related events.
Despite a high incidence of liver-related events, our cohort study shows that HIV-infected patients with OcC appear to be less frequently monitored with abdominal ultrasounds when compared to those with OvC. Guidelines advocate regular screening ultrasounds for HCC in patients with a known diagnosis of cirrhosis and in HBV/HIV-coinfected patients. Adherence to HCC screening guidelines remains low in HIV-infected patients, with reported rates between 20% and 35% [36, 37]. Underrecognition of cirrhosis at its preclinical stage, lack of awareness of liver-related complications, and potential patient-related factors may explain this low adherence [36]. Importantly, HIV-infected patients with HCC have a shorter expected survival when compared to HIV-negative patients [38, 39]. Our findings suggest that HIV-infected patients should be screened for OcC with noninvasive diagnostic tools, such as TE, and that those with OcC should be considered for HCC surveillance.
Our study has several strengths. To our knowledge, it is the first study characterizing risk factors and dynamics of liver cirrhosis in its preclinical stage in HIV-infected patients. It relies on 2 prospectively maintained large cohorts of consecutive individuals living with HIV. We employed an accurate and validated noninvasive tool to diagnose liver cirrhosis [13, 40]. We also acknowledge several limitations of our study. First, although we included a large number of patients, the follow-up period was relatively short. We observed a limited number of clinical outcomes and thus our estimates of incidence rates are imprecise. However, our findings showed that OcC patients behave as those with OvC in cases of HIV infection. Second, because a cross-sectional approach was used to determine factors associated with OcC, no inferential conclusions can be derived. However, these associated factors could help identify who might be prioritized for screening. Third, we did not have data about the indication for ultrasound, which could have been performed for reasons other than HCC screening. Fourth, there is a potential for competing risk from death from other causes. Fifth, liver biopsy was not available in our study. However, we validated our findings longitudinally by demonstrating that the OcC group had worse clinical outcomes than the noncirrhotic group. Using liver biopsy in our cohort of >1000 patients would not have been feasible.
In conclusion, OcC accounts for 1 in 5 cases of cirrhosis in HIV-infected patients. These patients, whether HCV coinfected or not, constitute a high-risk group prone to liver-related events. We advocate that HIV-infected patients with OcC should receive surveillance for HCC to the same degree as patients with OvC. Screening for cirrhosis with point-of-care noninvasive diagnostic tools, such as TE, should be considered in HIV-infected patients, especially in case of long duration of HIV infection and higher BMI.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Author contributions. A. B. contributed to study design, data, interpretation of the data, and first draft of the manuscript. R. N. contributed to study design, statistical analysis, and interpretation of data. T. P., A. S. H., P. W., and M. D. contributed to data and interpretation of data. P. G. and M. B. K. contributed to conception, study design, data, and interpretation of the data. G. S. contributed to conception, study design, data, and interpretation of the data, and first draft of the manuscript. All authors approved the final version of the article.
Disclaimer. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. G. S. is supported by a research salary from the Department of Medicine of McGill University.
Financial support. The Canadian Coinfection Cohort was funded by the Canadian Institutes of Health Research (HOP-90182 to M. B. K.). ViiV and Merck provided a grant to establish the diagnostic center for hepatic fibrosis and steatosis at the McGill University Health Centre, which is in use for the LIVEr disease in HIV Cohort.
Potential conflicts of interest. P. G. has acted as consultant for Merck and Gilead. M. B. K. has acted as a consultant for ViiV, Gilead, Janssen, and Merck, and has received research funding from Réseau sida et maladies infectieuses du Fonds de la recherche santé du Québec, the National Institutes of Health, Merck, Gilead, and ViiV Healthcare. M. D. has served as an advisory board member for Merck, Janssen, and Gilead. P. W. has acted as a consultant for Bristol-Myers Squibb (BMS), Gilead, Merck, and Novartis. G. S. has received speaker’s fees from Merck, BMS, Gilead, AbbVie, and ViiV; has served as an advisory board member for Merck, BMS, and Novartis and has received research funding from Merck, ViiV, and Echosens. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1. Smith CJ, Ryom L, Weber R, et al. Trends in underlying causes of death in people with HIV from 1999 to 2011 (D:A:D): a multicohort collaboration. Lancet 2014; 384:241–8. [DOI] [PubMed] [Google Scholar]
- 2. Pereyra F, Jia X, McLaren PJ, et al. ; International HIV Controllers Study The major genetic determinants of HIV-1 control affect HLA class I peptide presentation. Science 2010; 330:1551–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Rockstroh JK, Mohr R, Behrens G, Spengler U. Liver fibrosis in HIV: which role does HIV itself, long-term drug toxicities and metabolic changes play? Curr Opin HIV AIDS 2014; 9:365–70. [DOI] [PubMed] [Google Scholar]
- 4. Shah AG, Lydecker A, Murray K, Tetri BN, Contos MJ, Sanyal AJ; Nash Clinical Research Network Comparison of noninvasive markers of fibrosis in patients with nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol 2009; 7:1104–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. D’Amico G, Morabito A, Pagliaro L, Marubini E. Survival and prognostic indicators in compensated and decompensated cirrhosis. Dig Dis Sci 1986; 31:468–75. [DOI] [PubMed] [Google Scholar]
- 6. de Franchis R; Baveno VI Faculty Expanding consensus in portal hypertension: Report of the Baveno VI Consensus Workshop: stratifying risk and individualizing care for portal hypertension. J Hepatol 2015; 63:743–52. [DOI] [PubMed] [Google Scholar]
- 7. Chen T, Wong R, Wong P, et al. Occult cirrhosis diagnosed by transient elastography is a frequent and under-monitored clinical entity. Liver Int 2015; 35:2285–93. [DOI] [PubMed] [Google Scholar]
- 8. Bertot LC, Jeffrey GP, Wallace M, et al. Nonalcoholic fatty liver disease-related cirrhosis is commonly unrecognized and associated with hepatocellular carcinoma. Hepatol Commun 2017; 1:53–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Kim MN, Kim SU, Kim BK, et al. Increased risk of hepatocellular carcinoma in chronic hepatitis B patients with transient elastography-defined subclinical cirrhosis. Hepatology 2015; 61:1851–9. [DOI] [PubMed] [Google Scholar]
- 10. Rockey DC, Caldwell SH, Goodman ZD, Nelson RC, Smith AD; American Association for the Study of Liver Diseases Liver biopsy. Hepatology 2009; 49:1017–44. [DOI] [PubMed] [Google Scholar]
- 11. Colloredo G, Guido M, Sonzogni A, Leandro G. Impact of liver biopsy size on histological evaluation of chronic viral hepatitis: the smaller the sample, the milder the disease. J Hepatol 2003; 39:239–44. [DOI] [PubMed] [Google Scholar]
- 12. Wong JB, Koff RS. Watchful waiting with periodic liver biopsy versus immediate empirical therapy for histologically mild chronic hepatitis C. A cost-effectiveness analysis. Ann Intern Med 2000; 133:665–75. [DOI] [PubMed] [Google Scholar]
- 13. Benmassaoud A, Ghali P, Cox J, et al. Screening for nonalcoholic steatohepatitis by using cytokeratin 18 and transient elastography in HIV mono-infection. PLoS One 2018; 13:e0191985. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Friedrich-Rust M, Ong MF, Martens S, et al. Performance of transient elastography for the staging of liver fibrosis: a meta-analysis. Gastroenterology 2008; 134:960–74. [DOI] [PubMed] [Google Scholar]
- 15. Wai CT, Greenson JK, Fontana RJ, et al. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology 2003; 38:518–26. [DOI] [PubMed] [Google Scholar]
- 16. de Lédinghen V, Douvin C, Kettaneh A, et al. Diagnosis of hepatic fibrosis and cirrhosis by transient elastography in HIV/hepatitis C virus-coinfected patients. J Acquir Immune Defic Syndr 2006; 41:175–9. [DOI] [PubMed] [Google Scholar]
- 17. Sterling RK, Lissen E, Clumeck N, et al. ; APRICOT Clinical Investigators Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology 2006; 43:1317–25. [DOI] [PubMed] [Google Scholar]
- 18. Pembroke T, Deschenes M, Lebouché B, et al. Hepatic steatosis progresses faster in HIV mono-infected than HIV/HCV co-infected patients and is associated with liver fibrosis. J Hepatol 2017; 67:801–8. [DOI] [PubMed] [Google Scholar]
- 19. Klein MB, Saeed S, Yang H, et al. Cohort profile: the Canadian HIV-hepatitis C co-infection cohort study. Int J Epidemiol 2010; 39:1162–9. [DOI] [PubMed] [Google Scholar]
- 20. IDF Clinical Guidelines Task Force. Global guideline for type 2 diabetes: recommendations for standard, comprehensive, and minimal care. Diabet Med 2006; 23:579–93. [DOI] [PubMed] [Google Scholar]
- 21. Reinert DF, Allen JP. The Alcohol Use Disorders Identification Test (AUDIT): a review of recent research. Alcohol Clin Exp Res 2002; 26:272–9. [PubMed] [Google Scholar]
- 22. Loko MA, Castera L, Dabis F, et al. ; Groupe d’Epidémiologie Clinique du SIDA en Aquitaine (GECSA) Validation and comparison of simple noninvasive indexes for predicting liver fibrosis in HIV-HCV-coinfected patients: ANRS CO3 Aquitaine cohort. Am J Gastroenterol 2008; 103:1973–80. [DOI] [PubMed] [Google Scholar]
- 23. Castera L. Noninvasive methods to assess liver disease in patients with hepatitis B or C. Gastroenterol 2012; 142:1293–302.e4. [DOI] [PubMed] [Google Scholar]
- 24. Caussy C, Alquiraish MH, Nguyen P, et al. Optimal threshold of controlled attenuation parameter with MRI-PDFF as the gold standard for the detection of hepatic steatosis. Hepatology 2018; 67:1348–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Myers RP, Pomier-Layrargues G, Kirsch R, et al. Feasibility and diagnostic performance of the FibroScan XL probe for liver stiffness measurement in overweight and obese patients. Hepatology 2012; 55:199–208. [DOI] [PubMed] [Google Scholar]
- 26. Heimbach JK, Kulik LM, Finn RS, et al. AASLD guidelines for the treatment of hepatocellular carcinoma. Hepatology 2018; 67:358–80. [DOI] [PubMed] [Google Scholar]
- 27. European Association for the Study of the Liver; European Organisation for Research and Treatment of Cancer. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol 2012; 56:908–43. [DOI] [PubMed] [Google Scholar]
- 28. Menon KV, Hakeem AR, Heaton ND. Review article: liver transplantation for hepatocellular carcinoma—a critical appraisal of the current worldwide listing criteria. Aliment Pharmacol Ther 2014; 40:893–902. [DOI] [PubMed] [Google Scholar]
- 29. Garcia-Tsao G, Friedman S, Iredale J, Pinzani M. Now there are many (stages) where before there was one: in search of a pathophysiological classification of cirrhosis. Hepatology 2010; 51:1445–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Macías J, Camacho A, Von Wichmann MA, et al. Liver stiffness measurement versus liver biopsy to predict survival and decompensations of cirrhosis among HIV/hepatitis C virus-coinfected patients. AIDS 2013; 27:2541–9. [DOI] [PubMed] [Google Scholar]
- 31. Vergniol J, Boursier J, Coutzac C, et al. Evolution of noninvasive tests of liver fibrosis is associated with prognosis in patients with chronic hepatitis C. Hepatology 2014; 60:65–76. [DOI] [PubMed] [Google Scholar]
- 32. Merchante N, Téllez F, Rivero-Juárez A, et al. ; Grupo Andaluz para el Estudio de las Hepatitis Víricas (HEPAVIR) de la Sociedad Andaluza de Enfermedades Infecciosas (SAEI) Progression of liver stiffness predicts clinical events in HIV/HCV-coinfected patients with compensated cirrhosis. BMC Infect Dis 2015; 15:557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Klein MB, Althoff KN, Jing Y, et al. Risk of end-stage liver disease in HIV-viral hepatitis coinfected persons in North America from the early to modern antiretroviral therapy eras. Clin Infect Dis 2016; 63:1160–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Lemoine M, Lacombe K, Bastard JP, et al. Metabolic syndrome and obesity are the cornerstones of liver fibrosis in HIV-monoinfected patients. AIDS 2017; 31:1955–64. [DOI] [PubMed] [Google Scholar]
- 35. Macías J, Mancebo M, Márquez M, et al. Low risk of liver decompensation among human immunodeficiency virus/hepatitis C virus-coinfected patients with mild fibrosis in the short term. Hepatology 2015; 61:1503–11. [DOI] [PubMed] [Google Scholar]
- 36. Beauchamp E, Rollet K, Walmsley S, Wong DK, Cooper C, Klein MB; Canadian Co-infection Cohort Study (CTN222) Missed opportunities for hepatocellular carcinoma screening in an HIV/hepatitis C virus-coinfected cohort. Clin Infect Dis 2013; 57:1339–42. [DOI] [PubMed] [Google Scholar]
- 37. Hearn B, Chasan R, Bichoupan K, et al. Low adherence of HIV providers to practice guidelines for hepatocellular carcinoma screening in HIV/hepatitis B coinfection. Clin Infect Dis 2015; 61:1742–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Gelu-Simeon M, Sobesky R, Haïm-Boukobza S, et al. Do the epidemiology, physiological mechanisms and characteristics of hepatocellular carcinoma in HIV-infected patients justify specific screening policies? AIDS 2014; 28:1379–91. [DOI] [PubMed] [Google Scholar]
- 39. Berretta M, Garlassi E, Cacopardo B, et al. Hepatocellular carcinoma in HIV-infected patients: check early, treat hard. Oncologist 2011; 16:1258–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Morse CG, McLaughlin M, Matthews L, et al. Nonalcoholic steatohepatitis and hepatic fibrosis in HIV-1-monoinfected adults with elevated aminotransferase levels on antiretroviral therapy. Clin Infect Dis 2015; 60:1569–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.