Abstract
Background:
Chronic hepatitis B (HBV) is the predominant cause of hepatocellular carcinoma (HCC) worldwide. Although HBV coinfection is common in HIV, the determinants of HCC in HIV/HBV coinfection are poorly characterized. We examined the predictors of HCC in a multi-cohort study of HIV/HBV-coinfected individuals.
Methods & Results:
We included HIV/HBV-coinfected persons within 22 cohorts of the North American AIDS Cohort Collaboration on Research and Design (1995–2016). First occurrence of HCC was verified by medical record review and/or cancer registry. We used multivariable Cox regression to determine adjusted hazard (aHRs [95% confidence intervals]) of factors assessed at cohort entry (age, sex, race, body mass index), ever during observation (heavy alcohol use, hepatitis C), or time-updated (HIV RNA, CD4+ percentage, diabetes mellitus, HBV DNA).
Among 8,354 HIV/HBV-coinfected individuals (median age, 43 years; 93% male; 52.4% non-white), 115 HCC cases were diagnosed over 65,392 person-years (incidence rate, 1.8 [95% CI, 1.5–2.1] events/1,000 person-years). Risk factors for HCC included age 40–49 years (aHR, 1.97 [1.22–3.17]), age ≥50 years (aHR, 2.55 [1.49–4.35]), hepatitis C coinfection (aHR, 1.61 [1.07–2.40]), and heavy alcohol use (aHR, 1.52 [1.04–2.23]), while time-updated HIV RNA >500 copies/mL (aHR, 0.90 [0.56–1.43]) and time-updated CD4+ percentage <14% (aHR, 1.03 [0.56–1.90]) were not. The risk of HCC was increased with time-updated HBV DNA >200 IU/mL (aHR, 2.22 [1.42–3.47]) and was higher with each 1.0 log10 IU/mL increase in time-updated HBV DNA (aHR, 1.18 [1.05–1.34]). HBV suppression with HBV-active antiretroviral therapy (ART) for ≥1 year significantly reduced HCC risk (aHR, 0.42 [0.24–0.73]).
Conclusion:
HIV/HBV-coinfected individuals on ART with detectable HBV viremia remain at risk for HCC. To gain maximal benefit from ART for HCC prevention, sustained HBV suppression is necessary.
Introduction
Liver cancer is the sixth most common cancer and third leading cause of cancer-related mortality worldwide (1). Chronic hepatitis B virus (HBV) infection, both via inflammation and virally mediated pro-oncogenic mechanisms, is the most common cause of hepatocellular carcinoma (HCC) (2). Coinfection with chronic HBV is common among people living with HIV, with an estimated HBV prevalence of 5–15% (3). As the population of HIV-infected persons has aged during the antiretroviral therapy (ART) era, HCC has emerged as a leading cause of non-AIDS-defining cancer and cancer-attributable death (4).
Despite the high prevalence of HBV coinfection, no population-based studies have examined the incidence rates and determinants of HCC exclusively among HIV/HBV-coinfected individuals. Existing knowledge of risk factors for HBV-associated HCC originates from native Asian or Caucasian cohorts without HIV infection (5, 6). Consequently, it remains unclear if higher levels of HIV RNA, prolonged HIV viremia, and greater HIV-related immunosuppression contribute to development of HCC in HBV coinfection. It is also unknown if the biological gradient of risk for HCC with increasing HBV DNA levels, which has been observed among HBV-monoinfected persons, is present in HIV coinfection (5).
In this multi-cohort study using data from the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD), we examined the incidence of HBV-associated HCC and evaluated the determinants of this malignancy, including HIV-related factors (HIV viremia, immunosuppression), HBV viremia, and traditional HCC risk factors, to help guide preventive measures and early identification of this cancer.
Methods
Study design and data source
We performed a longitudinal cohort study of HIV/HBV-coinfected persons enrolled in 22 US and Canadian cohorts of the NA-ACCORD from 1995–2016. The NA-ACCORD is the largest consortium of interval and clinic-based HIV cohorts in the region (7). At regular intervals, NA-ACCORD cohorts transfer demographic, diagnostic, medication, socio-behavioral, laboratory, and vital status information to the Data Management Core (University of Washington), which performs quality control for completeness and accuracy and harmonizes data across cohorts that are sent to the Epidemiology/Biostatistics Core (Johns Hopkins University) where analytic-ready summary files are created. NA-ACCORD research has been approved by the Institutional Review Boards of each cohort. Informed consent was waived given the de-identified nature of these data. This study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the University of Pennsylvania and University of Washington Institutional Review Boards.
Study patients
The study population included HIV-infected individuals who had: 1) age ≥18 years, 2) active HBV coinfection (defined by at least one of the following: positive HBV surface antigen, positive HBV e antigen [HBeAg], or detectable HBV DNA) between January 1, 1995 and December 31, 2016, and 3) HIV RNA and CD4+ cell measurement during this period. To minimize the likelihood of including patients with acute HBV infection, we excluded patients who had an alanine aminotransferase or aspartate aminotransferase >1,000 U/L within ±30 days of their first HBV surface antigen, HBeAg, or HBV DNA and no subsequent positive HBV laboratory test ≥6 months after the initial result.
To ensure prevalent HCC diagnoses were not misclassified as incident events, we defined the start of follow-up as 180 days after the latest of the date of NA-ACCORD enrollment, start of the HCC observation window for the patient’s cohort (during which HCC diagnoses were ascertained), or assessment of HIV RNA or CD4+ cell measurements. We excluded patients who had HCC diagnosed prior to start of follow-up. Follow-up continued until first occurrence of: HCC, death, cohort-specific end date for reporting HCC diagnoses, date lost to follow-up (defined by NA-ACCORD as the earlier of the last HIV RNA or CD4+ cell measurement plus 540 days), or December 31, 2016.
Main study outcome
The primary outcome was incident HCC diagnosis. Within each cohort, HCC diagnoses were adjudicated based on cancer registry linkage or through comprehensive review of medical records by trained medical record abstracters under the supervision of a physician using a web-based standardized abstraction protocol to verify the diagnosis (i.e., histopathologic diagnosis, supportive radiographic imaging, or clinician-confirmed diagnosis), as previously described (8).
Covariates
We examined age, sex, race/ethnicity, body mass index, HIV transmission risk factors, diabetes mellitus, alcohol use, chronic hepatitis C virus (HCV) infection, and use of ART. Diabetes was assessed throughout observation and defined by: 1) hemoglobin A1c ≥6.5%, 2) prescription of specific anti-diabetic medications (e.g. insulin), or 3) diabetes diagnosis plus prescription of diabetes-related medications (9). Heavy alcohol use was defined as ever having had while under observation: 1) International Classification of Diseases, Ninth Revision hospital or outpatient diagnosis of alcohol dependence/abuse, or 2) ≥3 drinks/day or ≥7 drinks/week for females; ≥4 drinks/day or ≥14 drinks/week for males on the self-reported Alcohol Use Disorders Identification Test-Consumption questionnaire (10). Chronic HCV coinfection was defined by detectable HCV RNA or available HCV genotype recorded at any time during observation. ART was defined as use of three antiretrovirals from at least two classes or a triple nucleoside/nucleotide reverse transcriptase inhibitor regimen (previously accepted as ART). We determined exposure to ART, HBV-active antiretrovirals (tenofovir disoproxil fumarate [TDF], lamivudine, emtricitabine), and entecavir throughout observation.
We examined all HIV RNA levels, CD4+ cell percentages, HBeAg, and quantitative and qualitative HBV DNA results before and during follow-up. We evaluated CD4+ cell percentage instead of absolute CD4+ cell count because absolute CD4+ count may decrease during cirrhosis due to portal hypertension-induced splenic sequestration (11). HIV RNA level and CD4+ cell percentages were lagged by 180 days (approximate mean doubling time of <5 cm HCCs (12)) to reduce the possibility that HCC influenced these variables (i.e., reverse causality). Due to changes in the sensitivity of HIV RNA assays over time, detectable HIV was defined as >500 copies/mL. Hepatitis delta coinfection status was not available within the NA-ACCORD data.
Statistical analysis
We first determined unadjusted incidence rates of HCC (events/1,000 person-years). We used multivariable Cox regression to determine adjusted hazard ratios (aHRs [95% confidence intervals]) of HCC for risk factors of interest. HIV-related factors included detectable HIV RNA (>500 copies/mL) and CD4+ cell percentage, both time-updated. We examined traditional HCC risk factors including older age (≥ 40 years), male sex, race/ethnicity, obese body mass index (≥30 kg/m2), time-updated diabetes mellitus, heavy alcohol use, and chronic HCV coinfection (2). Baseline values of time-updated variables (i.e., HIV RNA, CD4+ cell percentage, and diabetes status) were carried forward until the date that a new value was recorded. The aHRs of HCC for time-updated variables therefore reflect associations between current values of that variable and HCC. Analyses additionally adjusted for the year that follow-up started. ART use was not included because of its collinearity with HIV RNA and CD4+ percentage. To evaluate fully the effects of HIV viremia on HCC, we also examined alternative definitions of HIV viremia as a: 1) categorical variable (≤500, 501–10,000, >10,000 copies/mL), 2) continuous variable per 1.0 log10 copies/mL increase, and 3) cumulative value with increasing consecutive months of detectable HIV (compared to those with undetectable HIV), as previously described (13). We performed two secondary analyses. First, we determined aHRs of HCC accounting for death as a competing risk (14). Second, to assess if risk factors for HCC varied by presence of advanced hepatic fibrosis, we stratified our analysis by platelet count at start of follow-up (<150,000/μL; ≥150,000/μL), since platelet count <150,000/μL is associated with advanced hepatic fibrosis by liver biopsy (Ishak stage 4–6) among chronic HBV patients and is a marker of cirrhosis-induced portal hypertension (15). We chose not to adjust for platelet count in primary analyses because cirrhosis is in the causal pathway to HCC and controlling for advanced hepatic fibrosis status could potentially adjust away associations between risk factors of interest and HCC and might attenuate our ability to detect important clinical factors.
Next, to examine the effects of HBV viremia on HCC risk, we first restricted the sample to those who had quantitative HBV DNA measured prior to the end of follow-up. We constructed separate Cox models to examine aHRs of HCC associated with the following time-updated categories of HBV DNA: 1) >200 international units [IU]/mL, 2) ≤200; 201–2,000; >2,000 IU/mL, and 3) ≤200; 201–200,000; >200,000 IU/mL. We also determined the aHR of HCC with higher time-updated quantitative HBV DNA level (per 1.0 log10 IU/mL increase).
We then included those who had quantitative or qualitative HBV DNA assessed. We used Cox regression to evaluate the risk of HCC with time-updated detectable HBV (compared to those with undetectable HBV). We evaluated aHRs of HCC associated with increasing consecutive months of detectable HBV (compared to those with undetectable HBV). Given the potential impact of both HIV and HBV on development of HCC, we used Cox regression to evaluate the risk of HCC as a composite variable of time-updated detectable HIV and HBV with four categories: 1) detectable HIV and HBV, 2) undetectable HIV with detectable HBV, 3) detectable HIV with undetectable HBV, and 4) undetectable HIV and HBV. We then determined the aHRs of HCC for above groups 1–3 compared to those with undetectable HIV and HBV.
Among individuals who received HBV-active ART (i.e., TDF, lamivudine, emtricitabine) or entecavir and who had quantitative or qualitative HBV DNA assessed, we determined if the risk of HCC was reduced with increasing consecutive months of: 1) undetectable HBV, and 2) undetectable HIV and HBV. Consecutive months of undetectable viremia were counted until detectable HIV or HBV was observed. If undetectable viremia was again achieved, the consecutive months suppressed was restarted at one month.
Analyses evaluating HBV viremia were adjusted for time-updated HIV RNA, time-updated CD4+ percentage, and other HCC risk factors. Analyses were repeated accounting for death as a competing risk and stratified by platelet count (<150,000/μL; ≥150,000/μL) at start of follow-up. In the subgroup of persons who had HBeAg assessed, we performed an exploratory analysis examining the association between positive HBeAg status and HCC. Further, to begin to explore the effect of advanced hepatic fibrosis on risk of HCC, we conducted an additional exploratory analysis evaluating the effect of baseline platelet count <150,000/μL on HCC risk.
We assessed proportionality of hazards with log-log plots and Schoenfeld residuals. We implemented multiple imputation using chained equations to address the potential bias of missing risk factor data, by means of ten imputations using all variables in Table 1 (16). Results across the ten datasets were combined to arrive at confidence intervals that accounted for within- and across-dataset variances. Data were analyzed using SAS 9.4 (SAS Institute Inc., Cary, NC).
Table 1.
Baseline Characteristics of HIV/hepatitis B virus-coinfected persons in the North American AIDS Cohort Collaboration on Research and Design (1995–2016).
| Characteristic | (n=8,354) |
|---|---|
| Age (n, %) | |
| Median (years, IQR) | 42.9 (36.2–49.2) |
| <40 years | 3,224 (38.6%) |
| 40–49 years | 3,204 (38.4%) |
| ≥50 years | 1,926 (23.1%) |
| Male sex (n, %) | 7,775 (93.1%) |
| Race (n, %) | |
| White | 3,973 (47.6%) |
| Black or African American | 3,421 (41.0%) |
| Asian/Pacific Islander | 126 (1.4%) |
| Multiracial, Other, Unknown | 834 (10.0%) |
| Hispanic (n, %) | 665 (8.4%) |
| Body mass index (n, %) | |
| Median (IQR) | 24.4 (22.0–27.3) |
| Underweight (<18.50 kg/m2) | 273 (3.6%) |
| Normal (18.50–24.99 kg/m2) | 4,041 (52.6%) |
| Overweight (25.00–29.99 kg/m2) | 2,435 (31.7%) |
| Obesity (≥30.00 kg/m2) | 929 (12.1%) |
| Missing | 676 (8.1%) |
| Diabetes mellitus (n, %) | 523 (6.3%) |
| Heavy alcohol use (n, %) | |
| Ever | 2,946 (35.3%) |
| Never | 5,207 (62.3%) |
| No data on alcohol use available | 201 (2.4%) |
| HIV transmission risk factors (n, %) | |
| Men who have sex with men | 3,939 (47.2%) |
| History of injection drug use | 1,806 (21.6%) |
| Receipt of blood transfusion, etc. | 31 (0.4%) |
| Heterosexual contact | 842 (10.1%) |
| Other | 151 (1.8%) |
| Unknown | 2,020 (24.2%) |
| Chronic hepatitis C virus infection (n, %) | 1,803 (21.6%) |
| HIV RNA (n, %) | |
| Median (log10 copies/mL, IQR) | 2.6 (1.7–4.1) |
| ≤500 copies/mL | 4,567 (54.7%) |
| >500 copies/mL | 3,787 (45.3%) |
| Absolute CD4+ cell count (n, %) | |
| Median (cells/mm3, IQR) | 354.0 (184.0–557.0) |
| ≥500 cells/mm3 | 2,626 (31.4%) |
| 200–499 cells/mm3 | 3,450 (41.3%) |
| <200 cells/mm3 | 2,278 (27.3%) |
| CD4+ cell percentage (n, %) | |
| Median (%, IQR) | 21.0 (12.2–30.0) |
| ≥28% | 2,539 (30.4%) |
| 14–27.99% | 3,454 (41.3%) |
| <14% | 2,361 (28.3%) |
| Platelet count (n, %) | |
| <150,000/μL | 1,584 (19.0%) |
| ≥150,000/μL | 6,285 (75.2%) |
| Not assessed at start of follow-up | 485 (5.8%) |
| HBV DNA | |
| Median HBV DNA (log10 IU/mL, IQR) | 2.0 (1.3–5.0) |
| Quantitative HBV DNA (initial assessment during observation; n, %) | 3,054 (36.6%) |
| ≤200 IU/mL | 1,688 (55.3%) |
| 201–2,000 IU/mL | 215 (7.0%) |
| >2,000 IU/mL | 1,151 (37.7%) |
| Assessed only for qualitative HBV DNA during observation (n, %) | 2,307 (27.6%) |
| Never assessed for quantitative or qualitative HBV DNA during observation (n, %) | 2,993 (35.8%) |
| HBV e antigen (n, %) | |
| Negative | 1,748 (20.9%) |
| Positive | 2,174 (26.0%) |
| Never tested before or during follow-up | 4,432 (53.1%) |
| On ART at start of follow-up (n, %) | 6,702 (80.2%) |
| Anti-HBV ART regimen at start of follow-up (n, %) * | |
| Lamivudine or emtricitabine alone | 3,997 (47.8%) |
| Tenofovir disoproxil fumarate alone | 142 (1.7%) |
| Tenofovir disoproxil fumarate + (lamivudine or emtricitabine) | 2,230 (26.7%) |
| On ART, but no anti-HBV antiretroviral | 333 (4.0%) |
| Not on ART | 1,652 (19.8%) |
| Year of start of follow-up (n, %) | |
| 1995–2000 | 2,490 (29.8%) |
| 2001–2006 | 3,398 (40.7%) |
| 2007–2016 | 2,466 (29.5%) |
Abbreviations: ART=antiretroviral therapy; HBV=hepatitis B virus; HIV=human immunodeficiency virus; IQR=interquartile range; RNA=ribonucleic acid
Results represent initial ART regimen.
Age was measured as year of baseline - year of birth.
Sex, race/ethnicity, and history of injection drug use were collected at enrollment into the NA-ACCORD.
History of heavy alcohol use was defined as ever having reported while under observation in the NA-ACCORD: 1) inpatient or outpatient diagnosis of alcohol dependence/abuse, or 2) ≥3 drinks/day or ≥7 drinks/week for females; ≥4 drinks/day or ≥14 drinks/week for males on the self-reported Alcohol Use Disorders Identification Test-Consumption questionnaire.
HIV transmission risk factors were not mutually exclusive.
Diabetes mellitus was defined by: 1) hemoglobin A1c ≥6.5%, 2) prescription of certain anti-diabetic medications, or 3) diabetes diagnosis plus prescription of certain anti-diabetic medications.
Chronic hepatitis C virus infection was defined by detectable HCV RNA or available HCV genotype recorded at any time during observation.
ART was measured as a combination of 3 antiretroviral agents from at least two classes or a triple nucleoside/nucleotide reverse transcriptase inhibitor regimen.
Results
Patient characteristics
Among 130,594 persons with HIV in the 22 NA-ACCORD cohorts from 1995–2016, 10,661 (8.2%) were HBV-coinfected. After exclusions, 8,354 HIV/HBV-coinfected individuals remained in the sample (Figure 1). The median age was 43 years, and participants were predominantly male (93.1%) and non-white (52.4%; Table 1). Heavy alcohol use was reported in 35.3%. A total of 1,806 (21.6%) had chronic HCV coinfection. At start of follow-up, the median HIV RNA was 2.6 log10 copies/mL, and the median CD4+ cell count 354 cells/mm3 (median percentage, 21.0%). At any time before or during follow-up, 3,054 (36.6%) had a quantitative HBV DNA assessed (median measures/patient, 2 [interquartile range (IQR), 1–5]), 2,795 (33.5%) had a qualitative HBV DNA assessed (median measures/patient, 3 [IQR, 1–6]), and 3,922 (47.0%) had HBeAg tested. Among those who had HBeAg tested, 2,174 (55.4%) had a positive result.
Figure 1.
Selection of HIV/hepatitis B virus-coinfected persons within the North American AIDS Cohort Collaboration on Research and Design (1995–2016).
Abbreviations: HBV, hepatitis B virus; HCC, hepatocellular carcinoma; NA-ACCORD, North American AIDS Cohort Collaboration on Research and Design; RNA, ribonucleic acid
* Includes 22 contributing cohorts: ALIVE, CWRU/PCRD, FENWAY, HOMER, HOPS, JHHCC, KPMAS, KPNC, MACS, MCHS-II, MONT, OHTN, RRC, SAC, SCOPE, UAB, UCHCC, UCSD, UW, VAND, VACS, WIHS
† HBV co-infection determined by an ever positive for HBV surface antigen, HBV e antigen, or HBV DNA.
‡ Presumed acute HBV infection defined as alanine aminotransferase or aspartate aminotransferase >1,000 U/L within ±30 days of the first positive HBV laboratory test (HBV surface antigen, HBV e antigen, or HBV DNA) with no subsequent positive HBV laboratory test ≥6 months after the initial result.
At start of follow-up, 47.8% were on ART with lamivudine or emtricitabine as the only HBV-active antiretroviral, 26.7% received HBV-active ART with TDF plus either emtricitabine or lamivudine, 1.7% received HBV-active ART with TDF alone, and 4.0% were on ART without an HBV-active antiretroviral (Table 1). A total of 1,652 (19.8%) were not on ART at start of follow-up; of these, 1,277 (77.8%) started ART during follow-up. Among 1,985 (23.8%) patients not on an HBV-active antiretroviral at study entry, 1,509 (76.0%) received HBV-active ART during follow-up (1,143 TDF-based; 360 with lamivudine or emtricitabine alone). A total of 129 patients received entecavir during follow-up (of whom 118 [91.5%] were also on ART).
Among the 8,354 individuals, 115 (1.4%) developed HCC during 65,392 person-years of follow-up (incidence rate, 1.8 [95% CI, 1.5–2.1] events/1,000 person-years). The median duration of follow-up was 6.9 (IQR, 2.8–13.0) years.
Risk factors for HCC
The risk of HCC was not increased with time-updated HIV RNA >500 copies/mL (aHR, 0.90 [0.56–1.43]; Table 2). Analyses evaluating alternative definitions of HIV viremia showed that the risk of HCC was not increased with higher time-updated HIV RNA (aHR, 1.03 [0.86–1.24] per 1.0 log10 copies/mL increase), higher category of time-updated HIV RNA (501–10,000 copies/mL: aHR, 0.59 [0.28–1.24]; >10,000 copies/mL: aHR, 1.16 [0.68–1.99] versus ≤500 copies/mL), or greater consecutive months with detectable HIV (<12 months: aHR, 0.64 [0.30–1.38]; ≥12 months: aHR, 1.07 [0.64–1.79] versus those with undetectable HIV). HCC risk was also not increased with CD4+ percentage <14% (aHR, 1.03 [0.56–1.90] versus >28%). Several traditional risk factors for HCC were associated with increased HCC risk (Table 2), specifically age 40–49 years (aHR, 1.97 [1.22–3.17]), age ≥50 years (aHR, 2.55 [1.49–4.35]), heavy alcohol use (aHR, 1.52 [1.04–2.23]), and chronic HCV (aHR, 1.61 [1.07–2.40]). The risk of HCC was increased with diabetes, but the result did not meet statistical significance (aHR, 1.79 [0.95–3.38]).
Table 2.
Factors associated with hepatocellular carcinoma among HIV/hepatitis B virus-coinfected persons in the North American AIDS Cohort Collaboration on Research and Design (1995–2016; n=8,354; 115 incident hepatocellular carcinoma events identified).
| Characteristic | No. Exposed* | No. Events | Person-Time | Incidence Rate (95% CI), Events/1,000 Person-Years | Unadjusted HR (95% CI) | Adjusted HR† (95% CI) |
|---|---|---|---|---|---|---|
| Age | ||||||
| <40 years | 3,224 | 25 | 26,141 | 1.0 (0.6–1.4) | Reference | Reference |
| 40–49 years | 3,204 | 56 | 26,221 | 2.1 (1.6–2.8) | 2.21 (1.38–3.54) | 1.97 (1.22–3.17) |
| ≥50 years | 1,926 | 34 | 13,029 | 2.6 (1.8–3.6) | 2.82 (1.68–4.73) | 2.55 (1.49–4.35) |
| Sex | ||||||
| Female | 579 | 3 | 4,364 | 0.7 (0.1–2.0) | Reference | Reference |
| Male | 7,775 | 112 | 61,028 | 1.8 (1.5–2.2) | 2.61 (0.83–8.21) | 1.92 (0.60–6.14) |
| Race | ||||||
| Non-white | 4,381 | 51 | 33,540 | 1.5 (1.1–2.0) | Reference | Reference |
| White | 3,973 | 64 | 31,852 | 2.0 (1.5–2.6) | 1.30 (0.90–1.88) | 1.38 (0.94–2.03) |
| Body mass index | ||||||
| Not obese (<30 kg/m2) | 6,749 | 95 | 54,954 | 1.7 (1.4–2.1) | Reference | Reference |
| Obese (≥30 kg/m2) | 929 | 12 | 7,218 | 1.7 (0.9–2.9) | 0.98 (0.54–1.79) | 1.00 (0.55–1.83) |
| Time-updated diabetes mellitus ‡ | ||||||
| No | 7,903 | 104 | 62,138 | 1.7 (1.4–2.0) | Reference | Reference |
| Yes | 449 | 11 | 3,253 | 3.4 (1.7–6.1) | 2.08 (1.12–3.87) | 1.79 (0.95–3.38) |
| Heavy alcohol use § | ||||||
| No | 5,207 | 51 | 36,804 | 1.4 (1.0–1.8) | Reference | Reference |
| Yes | 2,946 | 63 | 27,103 | 2.3 (1.8–3.0) | 1.62 (1.12–2.35) | 1.52 (1.04–2.23) |
| Chronic hepatitis C virus infection || | ||||||
| No | 6,551 | 75 | 50,504 | 1.5 (1.2–1.9) | Reference | Reference |
| Yes | 1,803 | 40 | 14,888 | 2.7 (1.9–3.7) | 1.79 (1.22–2.63) | 1.61 (1.07–2.40) |
| Time-updated HIV RNA | ||||||
| ≤500 copies/mL | 7,436 | 87 | 45,829 | 1.9 (1.5–2.3) | Reference | Reference |
| >500 copies/mL | 6,719 | 28 | 19,563 | 1.4 (1.0–2.1) | 0.89 (0.57–1.37) | 0.90 (0.56–1.43) |
| Time-updated CD4+ cell percentage | ||||||
| >28% | 4,961 | 40 | 25,873 | 1.5 (1.1–2.1) | Reference | Reference |
| 14–28% | 6,219 | 59 | 27,469 | 2.1 (1.6–2.8) | 1.47 (0.98–2.21) | 1.47 (0.97–2.21) |
| <14% | 3,973 | 16 | 12,049 | 1.3 (0.8–2.2) | 0.98 (0.55–1.76) | 1.03 (0.56–1.90) |
| Year at start of follow-up (per year) | .. | .. | .. | .. | 0.96 (0.91–1.02) | 0.95 (0.89–1.01) |
Abbreviations: CI=confidence interval; HBV=hepatitis B virus; HIV=human immunodeficiency virus; HR=hazard ratio; RNA=ribonucleic acid
For time-updated variables, a given patient may be included within more than one category.
Hazard ratios adjusted for all other risk factors listed in this table.
Diabetes mellitus was defined by: 1) hemoglobin A1c ≥6.5%, 2) prescription of anti-diabetic medication, or 3) record of a diabetes diagnosis plus the prescription of diabetes-related medication prior to start of follow-up.
History of heavy alcohol use defined as ever having reported while under observation in the NA-ACCORD: 1) inpatient or outpatient diagnosis of alcohol dependence/abuse, or 2) ≥3 drinks/day or ≥7 drinks/week for females; ≥4 drinks/day or ≥14 drinks/week for males on the self-reported Alcohol Use Disorders Identification Test-Consumption questionnaire.
Chronic hepatitis C virus infection was defined by detectable HCV RNA or available HCV genotype recorded at any time during observation.
Among persons who had a quantitative HBV DNA assessed during follow-up (n=3,054; Table 3), the risk of HCC increased with higher time-updated HBV DNA (aHR, 1.18 [1.05–1.34] per 1.0 log10 IU/mL increase). HBV DNA >200 IU/mL was associated with a 2.7-fold increased risk of HCC (aHR, 2.70 [1.23–5.93]). The risk was especially elevated at HBV DNA >200,000 IU/mL (aHR, 4.34 [1.72–10.94]).
Table 3.
Risk of incident hepatocellular carcinoma associated with different categories of time-updated hepatitis B virus (HBV) DNA among HIV/HBV-coinfected persons who had quantitative HBV DNA measured prior to follow-up in the North American AIDS Cohort Collaboration on Research and Design (1995–2016; n=3,054; 30 incident hepatocellular carcinoma events identified).
| Characteristic | No. Exposed* | No. Events | Person-Time | Incidence Rate (95% CI), Events/1,000 Person-Years | Unadjusted HR (95% CI) | Adjusted HR† (95% CI) |
|---|---|---|---|---|---|---|
| Time-updated HBV DNA, 200 IU/mL cut-off | ||||||
| ≤200 IU/mL | 2,355 | 17 | 15,844.9 | 1.1 (0.6–1.7) | Reference | Reference |
| >200 IU/mL | 1,468 | 13 | 5,031.4 | 2.6 (1.4–4.4) | 2.60 (1.25–5.41) | 2.70 (1.23–5.93) |
| Time-updated HBV DNA, 2,000 IU/mL cut-off | ||||||
| ≤200 IU/mL | 2,355 | 17 | 15,844.9 | 1.1 (0.6–1.7) | Reference | Reference |
| 201–2,000 IU/mL | 596 | 2 | 973.1 | 2.1 (0.2–7.4) | 2.11 (0.48–9.17) | 2.20 (0.50–9.59) |
| >2,000 IU/mL | 1,232 | 11 | 4,058.3 | 2.7 (1.4–4.8) | 2.72 (1.26–5.86) | 2.85 (1.24–6.57) |
| Time-updated HBV DNA, 200,000 IU/mL cut-off | ||||||
| ≤200 IU/mL | 2,355 | 17 | 15,844.9 | 1.1 (0.6–1.7) | Reference | Reference |
| 201–200,000 IU/mL | 1,122 | 5 | 2,924.0 | 1.7 (0.6–4.0) | 1.71 (0.63–4.67) | 1.77 (0.63–4.94) |
| >200,000 IU/mL | 774 | 8 | 2,107.4 | 3.8 (1.6–7.5) | 3.87 (1.66–9.06) | 4.34 (1.72–10.94) |
Abbreviations: CI=confidence interval; HBV=hepatitis B virus; HIV=human immunodeficiency virus; HR=hazard ratio; RNA=ribonucleic acid
Since HBV DNA is a time-updated variable, a given patient may be included within more than one category.
Hazard ratios adjusted for age and year at start of follow-up.
Among those who had quantitative or qualitative HBV DNA assessed during follow-up (n=5,316; Table 4), the risk of HCC was increased with ≥1 year of detectable HBV (aHR, 2.15 [1.34–3.42]) but not with <1 year (aHR, 2.02 [0.95–4.32) compared to those with consistently undetectable HBV (test for trend p=0.0012). In the analysis examining combined HIV and HBV viremia, the highest risk of HCC was observed among coinfected patients with both detectable HIV and HBV (aHR, 2.21 [1.17–4.18]; Table 4) compared to those with both undetectable HIV and HBV. Patients with undetectable HIV but detectable HBV also had significantly elevated HCC risk (aHR, 1.77 [1.07–2.92]), but the risk of HCC was not increased with detectable HIV but undetectable HBV (aHR, 0.27 [0.06–1.14]).
Table 4.
Risk of hepatocellular carcinoma associated with time-updated hepatitis B virus (HBV) DNA level and time-updated detectable HIV and HBV status among HIV/HBV-coinfected persons who had quantitative or qualitative HBV DNA assessed in the North American AIDS Cohort Collaboration on Research and Design (1995–2016; n=5,316; 87 incident hepatocellular carcinoma events identified).
| Characteristic | No. Exposed* | No. Events | Person-Time | Incidence Rate (95% CI), Events/1,000 Person-Years | Unadjusted HR (95% CI) | Adjusted HR† (95% CI) |
|---|---|---|---|---|---|---|
| Time-Updated HBV DNA | ||||||
| Undetectable | 3,656 | 44 | 22,692 | 1.9 (1.4–2.6) | Reference | Reference |
| Detectable | 3,364 | 43 | 12,850 | 3.3 (2.4–4.5) | 1.87 (1.22–2.85) | 2.22 (1.42–3.47) |
| Time-Updated Detectable HIV and HBV Status ‡ | ||||||
| Undetectable HIV and HBV | 3,494 | 42 | 19,164 | 2.2 (1.6–3.0) | Reference | Reference |
| Detectable HIV, undetectable HBV | 1,881 | 2 | 3,529 | 0.6 (0.07–2.0) | 0.29 (0.07–1.21) | 0.27 (0.06–1.14) |
| Undetectable HIV, detectable HBV | 2,835 | 27 | 8,510 | 3.2 (2.1–4.6) | 1.55 (0.95–2.52) | 1.77 (1.07–2.92) |
| Detectable HIV and HBV | 2,480 | 16 | 4,340 | 3.7 (2.1–6.0) | 1.93 (1.07–3.49) | 2.21 (1.17–4.18) |
Abbreviations: CI=confidence interval; HBV=hepatitis B virus; HIV=human immunodeficiency virus; HR=hazard ratio
Since HIV RNA and HBV DNA are time-updated variables, a given patient may be included within more than one category.
Model evaluating time-updated HBV DNA adjusted for age, sex, race/ethnicity, diabetes, time-updated HIV RNA, time-updated CD4+ cell percentage, heavy alcohol use, and year at start of follow-up. Model evaluating time-updated detectable HIV and HBV status adjusted for age, sex, race/ethnicity, diabetes, time-updated CD4+ cell percentage, heavy alcohol use, and year at start of follow-up.
Detectable HIV=HIV RNA >500 copies/mL; Detectable HBV=HBV DNA identified on quantitative or qualitative assay; Undetectable HIV=HIV RNA ≤500 copies/mL; Undetectable HBV=no HBV DNA identified on quantitative or qualitative assay.
For persons who received HBV-active ART and who had quantitative or qualitative HBV DNA assessed (n=4,891; Table 5), increasing consecutive months of undetectable HBV was associated with a lower risk of HCC compared to those with detectable HBV (Table 5; test for trend p=0.0024). This pattern was also seen when evaluating consecutive months of combined HIV and HBV suppression. The risk of HCC was significantly reduced with ≥1 year of suppression of both viruses (S1 Table).
Table 5.
Risk of hepatocellular carcinoma associated with different categories of increasing consecutive months with undetectable (HBV) DNA and, separately, both undetectable HIV RNA and HBV DNA, among HIV/HBV-coinfected persons who had quantitative or qualitative HBV DNA assessed and received HBV-active antiretroviral therapy in the North American AIDS Cohort Collaboration on Research and Design (1995–2016; n=4,891; 78 incident hepatocellular carcinoma events identified).
| Characteristic* | No. Exposed† | No. Events | Person-Time | Incidence Rate (95% CI), Events/1,000 Person-Years | Unadjusted HR (95% CI) | Adjusted HR‡ (95% CI) |
|---|---|---|---|---|---|---|
| Duration of Undetectable HBV § | ||||||
| Detectable HBV | 3,226 | 38 | 11,529 | 3.3 (2.3–4.5) | Reference | Reference |
| Undetectable HBV for <1 year | 3,263 | 13 | 3,110 | 4.2 (2.2–7.1) | 1.41 (0.74–2.69) | 1.12 (0.55–2.28) |
| Undetectable HBV for ≥1 year | 2,843 | 27 | 15,481 | 1.7 (1.1–2.5) | 0.50 (0.30–0.82) | 0.42 (0.24–0.73) |
| Duration of Undetectable HBV || | ||||||
| Detectable HBV | 3,226 | 38 | 11,529 | 3.3 (2.3–4.5) | Reference | Reference |
| Undetectable HBV for <1 year | 3,263 | 13 | 3,110 | 4.2 (2.2–7.1) | 1.42 (0.75–2.71) | 1.14 (0.56–2.31) |
| Undetectable HBV for 1–4 years | 2,843 | 12 | 7,107 | 1.7 (0.9–2.9) | 0.56 (0.29–1.08) | 0.55 (0.28–1.07) |
| Undetectable HBV for ≥4 years | 1,773 | 15 | 8,374 | 1.8 (1.0–3.0) | 0.45 (0.25–0.84) | 0.34 (0.17–0.67) |
| Duration of Undetectable HIV and HBV ¶ | ||||||
| Detectable HIV and HBV | 4,181 | 40 | 13,923 | 2.9 (2.1–3.9) | Reference | Reference |
| Undetectable HIV and HBV for <1 year | 3,145 | 15 | 4,068 | 3.7 (2.1–6.1) | 1.34 (0.74–2.44) | 1.25 (0.68–2.29) |
| Undetectable HIV and HBV for ≥1 year | 2,609 | 23 | 11,996 | 1.9 (1.2–2.9) | 0.62 (0.37–1.05) | 0.51 (0.29–0.91) |
| Duration of Undetectable HIV and HBV # | ||||||
| Detectable HIV and HBV | 4,181 | 40 | 13,923 | 2.9 (2.1–3.9) | Reference | Reference |
| Undetectable HIV and HBV for <1 year | 3,145 | 15 | 4,068 | 3.7 (2.1–6.1) | 1.34 (0.74–2.45) | 1.24 (0.68–2.29) |
| Undetectable HIV and HBV for 1–4 years | 2,609 | 13 | 6,597 | 2.0 (1.0–3.4) | 0.70 (0.37–1.32) | 0.64 (0.34–1.23) |
| Undetectable HIV and HBV for ≥4 years | 1,396 | 10 | 5,399 | 1.9 (0.9–3.4) | 0.54 (0.26–1.10) | 0.38 (0.17–0.83) |
Abbreviations: CI=confidence interval; HBV=hepatitis B virus; HIV=human immunodeficiency virus; HR=hazard ratio
Detectable HIV=HIV RNA >500 copies/mL; Detectable HBV=HBV DNA identified on quantitative or qualitative assay; Undetectable HIV=HIV RNA ≤500 copies/mL; Undetectable HBV=no HBV DNA identified on quantitative or qualitative assay.
Since HIV RNA and HBV DNA are time-updated variables, a given patient may be included within more than one category.
Models adjusted for age, sex, race/ethnicity, diabetes, time-updated HIV RNA, time-updated CD4+ cell percentage, heavy alcohol use, and year at start of follow-up.
Test for trend: p=0.0024
Test for trend: p=0.0010
Test for trend: p=0.0314
Test for trend: p=0.0132
Similar findings were observed in competing risk analyses (S2–S4 Tables). Among persons with platelet counts ≥150,000/μL at start of follow-up, age ≥40 years, HCV coinfection, and heavy alcohol use remained risk factors for HCC, while time-updated HIV RNA >500 copies/mL and CD4+ percentage <14% were not (S5 Table). Within this subgroup, associations with HCC were slightly stronger for time-updated detectable HBV DNA and detectable HIV/HBV (S6 Table) as well as increasing consecutive years with both undetectable HIV and HBV (S7 Table).
Finally, in our exploratory analysis examining the association between positive HBeAg status and HCC among persons who had HBeAg assessed, we observed no increased risk of HCC with HBeAg-positive status after adjustment for HIV-related and traditional HCC risk factors (aHR, 1.47 [0.92–2.35]). In a separate exploratory analysis, baseline platelet count <150,000/μL was strongly associated with an increased risk of HCC (aHR, 4.05 [2.62–6.26]), after adjusting for detectable HBV, detectable HIV, CD4+ percentage, and traditional HCC risk factors (S8 Table). When accounting for competing risks, the results of these analyses were nearly identical.
Discussion
In this large sample of HIV/HBV-coinfected persons from North America, we found that all detectable measures of HBV viremia, including time-updated detectable HBV >200 IU/mL, higher current HBV DNA, and greater consecutive months of detectable HBV, were associated with increased risk of HCC. There was a biological gradient of risk of HCC with higher HBV DNA, especially at levels >200,000 IU/mL. Notably, neither CD4+ percentage nor any definition of HIV viremia were associated with increased HCC risk. In analyses evaluating both detectable HIV and HBV, the risk of HCC was highest when HBV viremia was detectable. Finally, sustained HBV suppression with HBV-active ART for ≥1 year was associated with a 58% reduction in HCC risk.
These findings complement and extend existing knowledge of HBV natural history, and do so specifically for HIV-coinfected persons. The REVEAL-HBV study from Taiwan provided important initial data on the role of HBV replication on HCC development in the pre-HBV antiviral therapy era (5). This study demonstrated a strong relationship between baseline HBV DNA level and risk of HCC among persons with chronic HBV monoinfection. A threshold for increased HCC risk was observed at 2,000 IU/mL, with incrementally greater risk for HCC with higher levels of HBV DNA. Additionally, persistent HBV viremia during follow-up, particularly at higher HBV DNA levels and for longer duration, was found to be a risk factor for incident HCC and later for recurrent HCC (17, 18).
Our study is the first to show that HBV viremia is associated with HCC risk in a large, racially diverse cohort of adults outside of Asia. The epidemiology of HBV infection is distinctly different in Asia, where most HBV infections are acquired perinatally followed by years of immune tolerance. In contrast, horizontal HBV transmission via sexual or parenteral routes is more common in Western countries. Despite this difference in acquisition and natural history, HBV viremia remained adversely associated with HCC in this non-Asian, HIV-infected population, with a threshold of risk that may begin as low as 200 IU/mL.
Our findings also provide some insight into the association between HBV viremia and excess mortality among people living with HIV and HBV that has been observed in a variety of cohort studies from Africa and Europe (19–21). In these studies, the probability of death increased at HBV DNA levels >2,000 IU/mL, and in one study, was incrementally higher with every 1.0 log10 IU/mL increase of HBV DNA (19). HCC, a lethal cancer with high case fatality, may be more challenging to diagnose and treat in the setting of HIV (22, 23), and may have been a contributing factor to the increased mortality observed in these studies.
Notably, in contrast to REVEAL-HBV, our sample of HIV/HBV-coinfected individuals was a mostly treated cohort, with 76% on HBV-active ART at start of follow-up and 76% of those not on HBV-active ART at baseline initiating this therapy during follow-up. This observation highlights that antiviral therapy reduces, but does not eliminate, the risk of HCC. The overall incidence of HCC in our study was comparable to estimates reported among HBV-monoinfected patients without cirrhosis receiving TDF (24–26). Interestingly, HBV viremia was a stronger predictor of HCC than HIV viremia in our cohort, which underscores the necessity of HBV-specific control with HBV-active ART to reduce HCC risk. Just as suboptimal adherence on HBV antiviral therapy and HBV viremia have been shown to have detrimental effects on development of HCC (27), so can persistent HBV viremia during HBV-active ART. These data highlight the importance of ongoing HBV DNA monitoring and optimization of HBV-active ART to achieve HBV DNA suppression, through maximal adherence and use of tenofovir-based ART.
As a corollary to our findings on HBV viremia, we found that sustained HBV suppression had a protective association with HCC risk. The potential for antiviral therapy to prevent liver cancer was first noted in a randomized, controlled trial of lamivudine in patients with chronic HBV and advanced hepatic fibrosis, which reported a 51% reduction in HCC risk with lamivudine (28). Multiple observational studies in HBV monoinfection have since demonstrated that prolonged use of the more potent and durable antiviral nucleos(t)ide analogues tenofovir or entecavir is associated with a reduction in risk of HCC by as much as 70% (6, 25, 26). Our findings suggest that HIV/HBV-coinfected patients are most likely to derive protective benefit from HBV-active ART when they have maintained HBV suppression beyond a year. We observed even lower hazard ratios with longer durations of HBV suppression.
Older age, heavy alcohol use, and chronic HCV coinfection – all known risk factors for HCC in HBV-monoinfected cohorts (29) – also significantly increased the risk of HCC in our cohort of HIV/HBV-coinfected persons. Heavy alcohol use and HCV coinfection were prevalent in our cohort and may continue to influence the natural history of HCC despite HBV suppression. Our findings suggest that avoiding excessive alcohol consumption and initiating antiviral therapy for chronic HCV infection might help to reduce the risk of HCC among HIV/HBV-coinfected persons.
Our study has several limitations. First, clinical factors were prospectively collected as part of routine clinical care, not per standardized protocol, which explains why HBV DNA monitoring was inconsistently performed. Second, we did not have data on hepatitis delta coinfection, HBV genotype, or non-alcoholic fatty liver disease, which have all been shown to be associated with increased risk of HCC (30, 31). Third, we used thrombocytopenia as a surrogate for advanced hepatic fibrosis because cirrhosis diagnoses were not available and since aspartate aminotransferase-to-platelet ratio index and Fibrosis-4 Index for Hepatic Fibrosis, which are commonly used non-invasive measures of hepatic fibrosis, perform poorly in determining cirrhosis among HIV/HBV-coinfected patients (32). Platelet count <150,000/μL is associated with advanced hepatic fibrosis as determined by liver biopsy among chronic HBV-infected patients and is a marker of portal hypertension (15). Fourth, chronic HCV infection was determined by detectable HCV RNA or genotype. HCV antibody-positive persons who were never assessed for HCV RNA or genotype might have been misclassified as uninfected. Finally, our cohorts represent the North American demographic of HIV/HBV-coinfected individuals. Our results may not be generalizable to other geographic regions.
A major strength of this study is examination of validated cancer diagnoses within a large multi-cohort population of HIV/HBV-coinfected persons who were followed over an extended length of time. Our evaluation of individual-level clinical data, as well as time-updated measures of HBV DNA, HIV RNA, and CD4+ percentage, enabled a comprehensive and granular examination of HCC risk factors. Use of lagged HIV RNA and CD4+ percentage minimized potential for reverse causality. Care was also taken to account for competing risk of death in our analyses.
Our findings highlight the burden of HCC on HIV/HBV-coinfected individuals in the ART era. In the absence of a cure for chronic HBV infection, prevention of HBV with immunization, early identification of HBV coinfection, and prompt initiation of HBV-active ART remain essential to preventing HBV-associated HCC. To gain maximal benefit from ART for HCC prevention, sustained and ideally uninterrupted suppression of HBV may be necessary over years. Further work is needed to determine how best to intervene with risk factor modification, antiviral therapy, and HCC screening to reduce the impact of this major cancer among HIV/HBV-coinfected persons.
Supplementary Material
Acknowledgements
We are grateful for the participants and representatives of the NA-ACCORD Collaborating Cohorts:
AIDS Clinical Trials Group Longitudinal Linked Randomized Trials: Constance A. Benson and Ronald J. Bosch
AIDS Link to the IntraVenous Experience: Gregory D. Kirk
Emory-Grady HIV Clinical Cohort: Vincent Marconi and Jonathan Colasanti
Fenway Health HIV Cohort: Kenneth H. Mayer and Chris Grasso
HAART Observational Medical Evaluation and Research: Robert S. Hogg, Viviane Lima, P. Richard Harrigan, Julio SG Montaner, Benita Yip, Julia Zhu, Kate Salters, and Karyn Gabler
HIV Outpatient Study: Kate Buchacz and Jun Li
HIV Research Network: Kelly A. Gebo and Richard D. Moore
Johns Hopkins HIV Clinical Cohort: Richard D. Moore
John T. Carey Special Immunology Unit Patient Care and Research Database, Case Western Reserve University: Jeffrey Jacobson
Kaiser Permanente Mid-Atlantic States: Michael A. Horberg
Kaiser Permanente Northern California: Michael J. Silverberg
Longitudinal Study of Ocular Complications of AIDS: Jennifer E. Thorne
MACS/WIHS Combined Cohort Study: Todd Brown, Phyllis Tien, and Gypsyamber D’Souza
Maple Leaf Medical Clinic: Graham Smith, Mona Loutfy, and Meenakshi Gupta
The McGill University Health Centre, Chronic Viral Illness Service Cohort: Marina B. Klein
Multicenter Hemophilia Cohort Study-II: Charles Rabkin
Ontario HIV Treatment Network Cohort Study: Abigail Kroch, Ann Burchell, Adrian Betts, and Joanne Lindsay
Parkland/UT Southwestern Cohort: Ank Nijhawan
Retrovirus Research Center, Universidad Central del Caribe, Bayamon Puerto Rico: Robert F. Hunter-Mellado and Angel M. Mayor
Southern Alberta Clinic Cohort: M. John Gill
Study of the Consequences of the Protease Inhibitor Era: Jeffrey N. Martin
Study to Understand the Natural History of HIV/AIDS in the Era of Effective Therapy: Jun Li and John T. Brooks
University of Alabama at Birmingham 1917 Clinic Cohort: Michael S. Saag, Michael J. Mugavero, and James Willig
University of California at San Diego: Laura Bamford and Maile Karris
University of North Carolina at Chapel Hill HIV Clinic Cohort: Joseph J. Eron and Sonia Napravnik
University of Washington HIV Cohort: Mari M. Kitahata and Heidi M. Crane
Vanderbilt Comprehensive Care Clinic HIV Cohort: Timothy R. Sterling, David Haas, Peter Rebeiro, and Megan Turner
Veterans Aging Cohort Study: Lesley Park and Amy Justice
NA-ACCORD Study Administration:
Executive Committee: Richard D. Moore, Keri N. Althoff, Stephen J. Gange, Mari M. Kitahata, Jennifer S. Lee, Michael S. Saag, Michael A. Horberg, Marina B. Klein, Rosemary G. McKaig, and Aimee M. Freeman
Administrative Core: Richard D. Moore, Keri N. Althoff, and Aimee M. Freeman
Data Management Core: Mari M. Kitahata, Stephen E. Van Rompaey, Heidi M. Crane, Liz Morton, Justin McReynolds, and William B. Lober
Epidemiology and Biostatistics Core: Stephen J. Gange, Jennifer S. Lee, Brenna Hogan, Bin You, Elizabeth Humes, Lucas Gerace, Cameron Stewart, and Sally Coburn
Financial Support:
The content is solely the responsibility of the authors and does not necessarily represent the views of the National Institutes of Health. This study was funded by a research grant from the National Institute of Allergy and Infectious Diseases (R21-AI124868). This work was supported by National Institutes of Health grants U01AI069918, F31AI124794, F31DA037788, G12MD007583, K01AI093197, K01AI131895, K23EY013707, K24AI065298, K24AI118591, K24DA000432, KL2TR000421, N01CP01004, N02CP055504, N02CP91027, P30AI027757, P30AI027763, P30AI027767, P30AI036219, P30AI050409, P30AI050410, P30AI094189, P30AI110527, P30MH62246, R01AA016893, R01DA011602, R01DA012568, R01 AG053100, R24AI067039, U01AA013566, U01AA020790, U01AA020793, U01AI038855, U01AI038858, U01AI068634, U01AI068636, U01AI069432, U01AI069434, U01DA03629, U01DA036935, U10EY008057, U10EY008052, U10EY008067, U01HL146192, U01HL146193, U01HL146194, U01HL146201, U01HL146202, U01HL146203, U01HL146204, U01HL146205, U01HL146208, U01HL146240, U01HL146241, U01HL146242, U01HL146245, U01HL146333, U24AA020794,U54MD007587, UL1RR024131, UL1TR000004, UL1TR000083, Z01CP010214 and Z01CP010176; contracts CDC-200-2006-18797 and CDC-200-2015-63931 from the Centers for Disease Control and Prevention, USA; contract 90047713 from the Agency for Healthcare Research and Quality, USA; contract 90051652 from the Health Resources and Services Administration, USA; grants CBR-86906, CBR-94036, HCP-97105 and TGF-96118 from the Canadian Institutes of Health Research, Canada; Ontario Ministry of Health and Long Term Care; and the Government of Alberta, Canada. Additional support was provided by the National Institute Of Allergy And Infectious Diseases (NIAID), National Cancer Institute (NCI), National Heart, Lung, and Blood Institute (NHLBI), Eunice Kennedy Shriver National Institute Of Child Health & Human Development (NICHD), National Human Genome Research Institute (NHGRI), National Institute for Mental Health (NIMH) and National Institute on Drug Abuse (NIDA), National Institute On Aging (NIA), National Institute Of Dental & Craniofacial Research (NIDCR), National Institute Of Neurological Disorders And Stroke (NINDS), National Institute Of Nursing Research (NINR), National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institute on Deafness and Other Communication Disorders (NIDCD), and National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). These data were collected by cancer registries participating in the National Program of Cancer Registries (NPCR) of the Centers for Disease Control and Prevention (CDC).
Abbreviations:
- aHR
adjusted hazard ratio
- ART
antiretroviral therapy
- HBeAg
hepatitis B e antigen
- HBV
hepatitis B virus
- HCC
hepatocellular carcinoma
- HCV
hepatitis C virus
- IQR
interquartile range
- IU
international units
- NA-ACCORD
North American AIDS Cohort Collaboration on Research and Design
- TDF
tenofovir disoproxil fumarate
References
- 1.International Agency on Research for Cancer WHO. Global Cancer Observatory: All Cancers Fact Sheet. In; 2018.
- 2.El-Serag HB. Epidemiology of viral hepatitis and hepatocellular carcinoma. Gastroenterology 2012;142:1264–1273.e1261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Platt L, French CE, McGowan CR, Sabin K, Gower E, Trickey A, McDonald B, et al. Prevalence and burden of HBV co-infection among people living with HIV: A global systematic review and meta-analysis. J Viral Hepat 2020;27:294–315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Robbins HA, Shiels MS, Pfeiffer RM, Engels EA. Epidemiologic contributions to recent cancer trends among HIV-infected people in the United States. AIDS 2014;28:881–890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chen C-J, Yang H-I, Su J, Jen C-L, You S-L, Lu S-N, Huang G-T, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 2006;295:65–73. [DOI] [PubMed] [Google Scholar]
- 6.Papatheodoridis GV, Idilman R, Dalekos GN, Buti M, Chi H, van Boemmel F, Calleja JL, et al. The risk of hepatocellular carcinoma decreases after the first 5 years of entecavir or tenofovir in Caucasians with chronic hepatitis B. Hepatology 2017;66:1444–1453. [DOI] [PubMed] [Google Scholar]
- 7.Gange SJ, Kitahata MM, Saag MS, Bangsberg DR, Bosch RJ, Brooks JT, Calzavara L, et al. Cohort profile: the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD). Int J Epidemiol 2007;36:294–301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Silverberg MJ, Lau B, Achenbach CJ, Jing Y, Althoff KN, D’Souza G, Engels EA, et al. Cumulative incidence of cancer among persons with HIV in North America: A cohort study. Ann Intern Med 2015;163:507–518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Crane HM, Kadane JB, Crane PK, Kitahata MM. Diabetes case identification methods applied to electronic medical record systems: their use in HIV-infected patients. Curr HIV Res 2006;4:97–106. [DOI] [PubMed] [Google Scholar]
- 10.Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA. The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identification Test. Arch Intern Med 1998;158:1789–1795. [DOI] [PubMed] [Google Scholar]
- 11.McGovern BH, Golan Y, Lopez M, Pratt D, Lawton A, Moore G, Epstein M, et al. The impact of cirrhosis on CD4+ T cell counts in HIV-seronegative patients. Clin Infect Dis 2007;44:431–437. [DOI] [PubMed] [Google Scholar]
- 12.Barbara L, Benzi G, Gaiani S, Fusconi F, Zironi G, Siringo S, Rigamonti A, et al. Natural history of small untreated hepatocellular carcinoma in cirrhosis: a multivariate analysis of prognostic factors of tumor growth rate and patient survival. Hepatology 1992;16:132–137. [DOI] [PubMed] [Google Scholar]
- 13.Torgersen J, Kallan MJ, Carbonari DM, Park LS, Mehta RL, D’Addeo K, Tate JP, et al. HIV RNA, CD4+ percentage, and risk of hepatocellular carcinoma by cirrhosis status. J Natl Cancer Inst 2020;112:747–755. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Fine J, Gray RJ. A proportional hazards model for the subdistribution of competing risk. J Am Stat Assoc 1999;94:496–509. [Google Scholar]
- 15.Lu SN, Wang JH, Liu SL, Hung CH, Chen CH, Tung HD, Chen TM, et al. Thrombocytopenia as a surrogate for cirrhosis and a marker for the identification of patients at high-risk for hepatocellular carcinoma. Cancer 2006;107:2212–2222. [DOI] [PubMed] [Google Scholar]
- 16.Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, Wood AM, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ 2009;338:b2393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Chen C-F, Lee W-C, Yang H-I, Chang H-C, Jen C-L, Iloeje UH, Su J, et al. Changes in serum levels of HBV DNA and alanine aminotransferase determine risk for hepatocellular carcinoma. Gastroenterology 2011;141:1240–1248, 1248.e1241–1242. [DOI] [PubMed] [Google Scholar]
- 18.Wang W, Tian SL, Wang H, Shao CC, Wang YZ, Li YL. Association of hepatitis B virus DNA level and follow-up interval with hepatocellular carcinoma recurrence. JAMA Netw Open 2020;3:e203707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Nikolopoulos GK, Paraskevis D, Psichogiou M, Hatzakis A. HBV-DNA levels predict overall mortality in HIV/HBV coinfected individuals. J Med Virol 2016;88:466–473. [DOI] [PubMed] [Google Scholar]
- 20.Velen K, Charalambous S, Innes C, Churchyard GJ, Hoffmann CJ. Chronic hepatitis B increases mortality and complexity among HIV-coinfected patients in South Africa: a cohort study. HIV Med 2016;17:702–707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kouame GM, Boyd A, Moh R, Badje A, Gabillard D, Ouattara E, Ntakpe JB, et al. Higher mortality despite early antiretroviral therapy in human immunodeficiency virus and hepatitis B virus (HBV)-coinfected patients with high HBV replication. Clin Infect Dis 2018;66:112–120. [DOI] [PubMed] [Google Scholar]
- 22.Pinato DJ, Dalla Pria A, Sharma R, Bower M. Hepatocellular carcinoma: an evolving challenge in viral hepatitis and HIV coinfection. AIDS 2017;31:603–611. [DOI] [PubMed] [Google Scholar]
- 23.Rimassa L, Personeni N, Czauderna C, Foerster F, Galle P. Systemic treatment of HCC in special populations. J Hepatol 2020. [DOI] [PubMed] [Google Scholar]
- 24.Buti M, Fung S, Gane E, Afdhal NH, Flisiak R, Gurel S, Flaherty JF, et al. Long-term clinical outcomes in cirrhotic chronic hepatitis B patients treated with tenofovir disoproxil fumarate for up to 5 years. Hepatol Int 2015;9:243–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kim WR, Loomba R, Berg T, Aguilar Schall RE, Yee LJ, Dinh PV, Flaherty JF, et al. Impact of long-term tenofovir disoproxil fumarate on incidence of hepatocellular carcinoma in patients with chronic hepatitis B. Cancer 2015;121:3631–3638. [DOI] [PubMed] [Google Scholar]
- 26.Nguyen MH, Yang HI, Le A, Henry L, Nguyen N, Lee MH, Zhang J, et al. Reduced incidence of hepatocellular carcinoma in cirrhotic and noncirrhotic patients with chronic hepatitis B treated with tenofovir-A propensity score-matched study. J Infect Dis 2019;219:10–18. [DOI] [PubMed] [Google Scholar]
- 27.Shin JW, Jung SW, Lee SB, Lee BU, Park BR, Park EJ, Park NH. Medication nonadherence increases hepatocellular carcinoma, cirrhotic complications, and mortality in chronic hepatitis B patients treated with entecavir. Am J Gastroenterol 2018;113:998–1008. [DOI] [PubMed] [Google Scholar]
- 28.Liaw YF, Sung JJ, Chow WC, Farrell G, Lee CZ, Yuen H, Tanwandee T, et al. Lamivudine for patients with chronic hepatitis B and advanced liver disease. N Engl J Med 2004;351:1521–1531. [DOI] [PubMed] [Google Scholar]
- 29.Raffetti E, Fattovich G, Donato F. Incidence of hepatocellular carcinoma in untreated subjects with chronic hepatitis B: a systematic review and meta-analysis. Liver Int 2016;36:1239–1251. [DOI] [PubMed] [Google Scholar]
- 30.Qu LS, Zhou GX. Significance of viral status on occurrence of hepatitis B-related hepatocellular carcinoma. World J Gastroenterol 2014;20:5999–6005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Streba LA, Vere CC, Rogoveanu I, Streba CT. Nonalcoholic fatty liver disease, metabolic risk factors, and hepatocellular carcinoma: an open question. World J Gastroenterol 2015;21:4103–4110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Sterling RK, King WC, Wahed AS, Kleiner DE, Khalili M, Sulkowski M, Chung RT, et al. Evaluating noninvasive markers to identify advanced fibrosis by liver biopsy in HBV/HIV co-infected adults. Hepatology 2020;71:411–421. [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.