Persons infected with HIV (PHIV), especially injecting drug users (IDU), have a greatly increased incidence of hepatitis B virus (HBV) and hepatitis C virus (HCV) infection compared with the general population, and hence are at a greater risk of hepatocellular carcinoma (HCC) [1,2]. In an era of highly active antiretroviral therapy (HAART), liver disease is becoming a leading cause of death among PHIV [3,4], of which HCC appears to account for an increasingly large fraction . Nevertheless, though studies on HBV/HCV natural history have shown that HIV-related immune suppression worsens the risk of cirrhosis and liver-related death [6–11], they have failed to identify a direct effect of HIV-related immunodeficiency on HCC risk [2,10,12].
In addition to potential indirect effects on HCC risk through improvements in immune reconstitution and survival, HAART is known to have some direct hepatotoxic effects, especially among PHIV chronically infected with HBV or HCV .
We thus undertook a case–control study of HCC nested within the Swiss HIV Cohort Study (SHCS), specifically designed to characterize the effect of immune-suppression and HAART on the development of HCC among PHIV, by carefully accounting for important confounding factors related to shared transmission of HBV, HCV and HIV.
Patients and methods
The SHCS is an ongoing study that has been enrolling PHIV since 1984 from seven large hospitals in Switzerland (www.shcs.ch), including 85 821 person years of follow-up until May 2007. Women contribute 25 976 person years and representation of HIV-transmission categories are balanced between men having sex with men (MSM, 28 557 person years), IDU (29 213 person years) and heterosexual/other (28 049 person years). Detailed information on disease, CD4+ cell count and HIV-related treatments are collected at enrolment, and at each 6-month follow-up visit. From 1998 onwards, results for markers of HCV and HBV at enrolment are also included, and HCV testing is repeated in participants with HCV-negative serology every 2 years thereafter.
In total, 36 patients with primary liver cancers were identified through an update of record linkage with eight Swiss Cantonal Cancer Registries , and three more were identified in the records from the Bern SHCS centre (a canton with no overlapping cancer registry). Of these 39 cases, 12 patients with liver lymphoma and one small cell carcinoma were excluded, leaving 26 HCC patients for the present analysis. Histological or cytological confirmation of HCC was available for 22 (85%) patients, whereas in the remaining four patients, diagnosis was based on imaging techniques or exclusion of lymphoma through lymph-node biopsies or both.
For each patient with HCC, 10 controls were matched at random from eligible SHCS participants. Eligible controls had at least the same length of follow-up as the matched case, and had a CD4+ cell count result available within 1 year prior to the reference date (occurring after a similar length of follow-up in the SHCS as the HCC diagnosis for the corresponding case). This value is hereafter referred to as ‘latest CD4+ cell count’. Matching criteria were: SHCS centre; sex; HIV-transmission category (IDU, MSM, heterosexual/other); age group at enrolment (in 10-year groups); and calendar period at enrolment (1985–1989, 1990–1992, 1993–1995, 1996–1998, and 1999–2001). For four HCC patients, only six, eight, eight and nine controls, respectively, could be matched, leaving 251 controls for the present study (Table 1).
HAART use was defined as prescription of at least three antiretroviral drugs, including a protease inhibitor or a non-nucleoside reverse transcriptase inhibitor, or three nucleoside reverse transcriptase inhibitors including abacavir.
The present study was approved by the local ethical committees of the clinics collaborating with the SHCS and of the International Agency for Research on Cancer. Written informed consent was obtained from all SHCS participants.
Serology results for HBV and HCV markers closest, but prior, to HCC diagnosis, or the corresponding reference date for controls, were considered from available measurements recorded in the SHCS database. Valid results for hepatitis B surface antigen (HBsAg), antibodies against HCV (anti-HCV) or both were missing for four (15.4%) HCC patients and 34 (13.5%) controls. Missing serology results for HCC patients, but not controls, were obtained from stored serum aliquots using enzyme-linked immunosorbent assay (ELISA; AxSYM HBsAg version 2.0 and AxSYM anti-HCV, Abbott Diagnostic Division, Weisbaden, Germany) at the Institute for Medical Microbiology, University of Basel, Basel, Switzerland. CD4+ cell count and total lymphocyte count were measured by flow cytometry and were retrieved from available measurements recorded in the SHCS database. CD4+ cell percentage was calculated as the proportion of CD4+ cells among all lymphocytes.
Logistic regression, conditioned on matching variables, was used to calculate odds ratios (ORs) and corresponding 95% confidence intervals (CI). CD4+ cell counts were divided into categories of 100 cells/μl, CD4+ cell percentages into categories of 10% or lymphocytes into categories of 500 cells/μl, and were fitted into the model as continuous variables.
Table 1 shows the distribution of 26 HCC patients and controls by selective matching and HCC-specific variables. The majority of HCC patients, and hence matched controls, were men (92.3%) between 30 and 39 years of age (61.5% of HCC cases). The HIV-transmission categories of IDU (53.9%) and MSM (34.6%) accounted for most of the HCC patients. The large majority of HCC patients had been enrolled in the SHCS prior to 1999 (92.3%), and had been followed up for more than 5 years prior to HCC diagnosis (73.1%). Three-quarters of HCC patients were diagnosed after 1999. HCC cases occurred at a mean age of 46.4 years. Only one of the 26 HCC patients, a liver transplant recipient, had survived for more than 2 years after HCC diagnosis.
All HCC patients were positive for HBsAg, anti-HCV or both. Ten HCC patients were positive for HBsAg only, 11 for anti-HCV only and five for both markers. Results for HCV RNA were not available systematically for SHCS participants, but all nine anti-HCV-positive HCC patients and 86% of 77 anti-HCV-positive controls with an available HCV RNA result were also HCV RNA-positive (data not shown). Only one (3.8%) of the 14 anti-HCV-positive HCC patients and 10 (8.6%) of the 116 anti-HCV-positive controls had been treated with peg-interferon/ribavirin (data not shown).
The associations of various measures of immunodeficiency with HCC risk are shown in Table 2. CD4+ cell count at enrolment in the SHCS was not related to HCC risk (OR per 100 cells/μl decrease = 0.96, 95% CI 0.84–1.10), nor was nadir CD4+ cell count (OR per 100 cells/μl decrease = 1.03, 95% CI 0.76–1.41) (data not shown). Latest CD4+ cell count, however, was related to HCC risk (OR per 100 cells/μl decrease = 1.33, 95% CI 1.06–1.68) (Table 2). A significant association was also observed between a history of AIDS diagnosis and HCC risk (OR = 2.40, 95% CI 1.06–5.44) (Table 2). Among patients who developed AIDS before cancer diagnosis/reference date, average CD4+ cell counts at AIDS diagnosis were similar among HCC patients (56% <200 cells/μl) and controls (61% <200 cells/μl) (data not shown).
Latest total lymphocyte count was not significantly different between HCC patients and controls (OR per 500 cells/μl decrease = 0.98, 95% CI 0.89–1.08) (Table 2). Hence, the substitution of latest CD4+ cell count with CD4+ cell percentage revealed a similar significant association with HCC risk (OR per 10% decrease = 1.65, 95% CI 1.01–2.71). HAART use (OR for ever versus never = 0.59, 95% CI 0.18–1.91) and HIV viral load (OR for <100 versus ≥10 000 copies/ml = 1.65, 95% CI 0.54–5.04) were not significantly related to HCC risk.
There was a strong relationship between HIV-transmission category and hepatitis virus type (Table 3). HCC patients who were IDU included 10 positive for anti-HCV only, one for HBsAg only and three positive for both; HCC in MSM/heterosexual/other included one positive for anti-HCV only, nine for HBsAg only and two positive for both. This dichotomy was also seen among controls, for whom anti-HCV prevalence was much higher among IDU (90.9%) than MSM/heterosexual/other (5.4%). Both HBsAg (OR = ∞, 95% CI 29.2–∞ and 5.53, 95% CI 1.15–26.5, respectively) and anti-HCV (OR = 7.34, 95% CI 1.14–47.4 and OR = 1.49, 95% CI 0.16–13.9, respectively) were more strongly associated with HCC among MSM/heterosexual/other than IDU.
When HCC patients and controls were stratified according to HIV-transmission category, the effect of latest CD4+ cell count on HCC risk was more evident for MSM/heterosexual/other (OR per 100 cells/μl decrease = 1.68, 95% CI 1.15–2.46) than IDU (OR = 1.13, 95% CI 0.86–1.48) (Table 3), though there was no statistically significant heterogeneity. This was also observed for CD4+ cell percentage (OR per 10% decrease = 2.19, 95% CI 1.05–4.56 among MSM/heterosexual/other versus 1.26, 95% CI 0.63–2.49 among IDU) (data not shown).
Our present nested case–control study showed, for the first time, a specific association between HCC risk and low CD4+ cell count in the year preceding HCC diagnosis. These findings, therefore, complement previous reports that declining CD4+ cell counts increase overall liver-related deaths, predominantly due to liver failure, among PHIV [3,6,11].
The relationship between immunodeficiency and HCC among PHIV has been difficult to establish. First, the excess risk of HCC in PHIV was slow to emerge, and became apparent only as PHIV on HAART started living long enough for the most severe sequelae of hepatitis viruses to manifest themselves [14–16]. Then, once established, the excess risk could be simply attributed to high prevalence of coinfection with HCV and HBV, as suggested by the largest single study of cancer risk in people with HIV/AIDS which found no relationship between CD4+ cell count at AIDS diagnosis and HCC . However, also in our present analysis, more historical measures of CD4+ cell count, including that at SHCS enrolment or nadir, were insensitive markers of HCC risk. Furthermore, only about half of HCC patients and about one-quarter of controls in our present study had a history of AIDS, highlighting the limitations in assessing HCC risk among AIDS patients only [12,17].
Previous case-series of HCC in Italy/Spain and the United States have shown that in PHIV, the tumour is more commonly associated with hepatitis viruses than in the corresponding general population [18,19] and is caused primarily by HCV [2,18,19]. In Switzerland, however, HIV-positive HCC patients were equally distributed between HBV and HCV. Differing proportions of HIV-transmission categories in the different studies likely explain this discrepancy. Indeed, among HCC patients in the present study, there was a strong dichotomy between HBV-positive HCC arising in MSM and HCV-positive HCC in IDU.
Matching for transmission category hampered us from evaluating the importance of hepatitis virus infections as independent risk factors for HCC among PHIV. Stratified analyses revealed, however, that the association between hepatitis virus infection and HCC risk tended to be weaker among IDU than MSM/heterosexual/other, and appeared stronger for HBsAg than anti-HCV in both groups. The association between lower CD4+ cell counts and HCC risk was also particularly strong for MSM/heterosexual/other and hence, due to the strong dichotomy mentioned above, for HBV-related HCC. Although this is the first report of a direct effect of HIV-related immune suppression on HBV-related HCC, it adds to the continuum of evidence that HIV worsens the clinical outcome of HBV infection by hastening progression to cirrhosis  and liver-related death [11,21], particularly at low CD4+ cell counts . The hypothesis that immune suppression might actually reduce HCC risk due to less immune destruction of HBV-infected hepatocytes  appears, therefore, unlikely.
Conversely, there was no strong evidence of an association between immunodeficiency and HCC arising in IDU, who were, in the vast majority, infected with HCV. A similar lack of excess HCC risk was also seen among HIV/HCV coinfected United States veterans, who were also predominantly IDU, compared with those infected with HCV alone [2,10]. Nevertheless, previous studies have shown that progression to cirrhosis or liver disease [8–10] and liver-related death [6,8] is increased in HCV-positive persons coinfected with HIV, particularly those with low CD4+ cell counts [8,23,24]. In the present study, IDU-matched controls tended to be more immunodeficient than their MSM/heterosexual/other counterparts. Thus, the effect of immunodeficiency on HCV-related HCC may have been masked by a worse immune status or high competing HCV-related or other mortality in IDU in general. The risk of HCC in anti-HCV-positive patients is not expected to have been heavily influenced either by treatment with peg-interferon/ribavirin, which was relatively rare in this population, nor by screening for HCC, which is not routine in the SHCS.
Cirrhosis often precedes HCC and has been associated with low CD4+ cell counts even in the absence of HIV infection, secondary to portal hypertension and splenic sequestration of lymphocytes . Thus, to exclude any possible reverse causality, we showed that total lymphocyte counts did not differ greatly between HCC patients and controls in our study and that significant findings were reproduced even when CD4+ cells were calculated as a percentage of all lymphocytes.
HAART is known to have some direct hepatotoxic effects which are worsened by coinfection with HBV or HCV [13,18], raising the possibility that HAART per se might hasten the progression to cirrhosis and hence HCC. Our findings did not support an overall increase in HCC risk in HAART users, and no HCC cases occurred in HAART users in the absence of HCV or HBV infection. HAART use has been reported to reduce liver-related mortality in HCV-coinfected  or HBV-coinfected persons , though this effect was not seen in a much larger cohort analysis .
The SHCS has many strengths, including the size, duration and regularity of follow-up and comprehensiveness of clinical and laboratory information. Approximately half of PHIV in Switzerland have been enrolled in the SHCS, and both sexes and different risk categories are well represented. Additional strengths were the supplementation of cancer diagnoses through linkage with cancer registries , meaning a more comprehensive follow-up for HCC, and the availability of histological or cytological confirmation for a majority of HCC patients, which allowed exclusion of metastatic hepatic lesions as well as lymphomas with liver involvement. The use of a nested case–control approach allowed careful matching for many important correlates of HCC risk and immune status, with complete HBV and HCV status in HCC patients. The principal weakness of the study is the relative small number of HCC patients that have accrued in the SHCS, which limited the possibility to draw conclusions on the strongly correlated effects of HIV-transmission category and hepatitis viral type.
As PHIV live longer and are exposed to the effects of mild and prolonged immune deficiency, it can be expected that hepatitis viruses will increasingly manifest their oncogenic potential. Thus, the present findings are also relevant for the millions of PHIV worldwide who live in HBV endemic areas in sub-Saharan Africa and Asia as their survival also improves on HAART.
This study was performed within the framework of the Swiss HIV Cohort Study, supported by the Swiss National Science Foundation (Grant 33CSCO-108787) and was funded by a grant from the OncoSuisse (ICP OCS 01355-03-2003) and the Istituto Superiore di Sanità, (grant 20 G.3).
The funding agencies had no role in the design of the study; the collection, analysis and interpretation of the data; the decision to submit for publication; or in the writing of the manuscript.
The authors thank Dr Peter Erb at the Institute for Medical Microbiology, University of Basel, for help with HBV/HCV serology testing and Dr Gernot Jundt (Basel), Dr Andrea Bordoni (Ticino) and Dr Silvia Ess (St Gallen and Appenzell) for data linkage exercise and collection of additional information on cancer cases from their respective cancer registries.
The authors of this manuscript have no conflict of interest to declare.
G.C. and S.F. conceived the study and drafted the manuscript. M.R. was responsible for management of the clinical databases of the Swiss HIV Cohort Study (SHCS) and liaison with the SHCS centres and Scientific Board. J.P. and L.D.M. were responsible for the linkage of the Swiss HIV Cohort with the Cantonal Cancer Registries as well as data managerial and statistical issues. I.S. performed the HCV and HBV serology testing. N.M.P.H., C.B., and F.L. are the representatives of the individual cantonal cancer registries that were responsible for the data linkage exercise and collection of additional information on cancer cases. A.R. and B.L. are experts on the clinical and epidemiological aspects, respectively, of HCV and HBV infection in the Swiss HIV Cohort, and gave important scientific input to the conception, the analysis and interpretation of the study. All authors read and gave feedback to the final version of the manuscript.
1. Grulich AE, Tvan Leeuwen M, Falster MO, Vajdic CM. Incidence of cancers in people with HIV/AIDS compared with immunosupressed transplant recipients: a meta-analysis. Lancet 2007; 370:59–67.
2. McGinnis KA, Fultz SL, Skanderson M, Conigliaro J, Bryant K, Justice AC. Hepatocellular carcinoma and non-Hodgkin's lymphoma: the roles of HIV, hepatitis C infection, and alcohol abuse. J Clin Oncol 2006; 24:5005–5009.
3. Weber R, Sabin CA, Friis-Moller N, Reiss P, El-Sadr WM, Kirk O, et al
. Liver-related deaths in persons infected with the human immunodeficiency virus: the D:A:D study. Arch Intern Med 2006; 166:1632–1641.
4. Louie JK, Hsu LC, Osmond DH, Katz MH, Schwarcz SK. Trends in causes of death among persons with acquired immunodeficiency syndrome in the era of highly active antiretroviral therapy, San Francisco, 1994–1998. J Infect Dis 2002; 186:1023–1027.
5. Rosenthal E, Pialoux G, Bernard N, Pradier C, Rey D, Bentata M, et al
. Liver-related mortality in human-immunodeficiency-virus-infected patients between 1995 and 2003 in the French GERMIVIC Joint Study Group Network (MORTAVIC 2003 Study). J Viral Hepat 2007; 14:183–188.
6. Darby SC, Ewart DW, Giangrande PL, Spooner RJ, Rizza CR, Dusheiko GM, et al
. Mortality from liver cancer and liver disease in haemophilic men and boys in UK given blood products contaminated with hepatitis C. UK Haemophilia Centre Directors' Organisation. Lancet 1997; 350:1425–1431.
7. Benhamou Y, Bochet M, Thibault V, Di Martino V, Caumes E, Bricaire F, et al
. Long-term incidence of hepatitis B virus resistance to lamivudine in human immunodeficiency virus-infected patients. Hepatology 1999; 30:1302–1306.
8. Di Martino V, Rufat P, Boyer N, Renard P, Degos F, Martinot-Peignoux M, et al
. The influence of human immunodeficiency virus coinfection on chronic hepatitis C in injection drug users: a long-term retrospective cohort study. Hepatology 2001; 34:1193–1199.
9. Graham CS, Baden LR, Yu E, Mrus JM, Carnie J, Heeren T, et al
. Influence of human immunodeficiency virus infection on the course of hepatitis C virus infection: a meta-analysis. Clin Infect Dis 2001; 33:562–569.
10. Kramer JR, Giordano TP, Souchek J, Richardson P, Hwang LY, El-Serag HB. The effect of HIV coinfection on the risk of cirrhosis and hepatocellular carcinoma in U.S. veterans with hepatitis C. Am J Gastroenterol 2005; 100:56–63.
11. Thio CL, Seaberg EC, Skolasky R Jr, Phair J, Visscher B, Munoz A, et al
. HIV-1, hepatitis B virus, and risk of liver-related mortality in the Multicenter Cohort Study (MACS). Lancet 2002; 360:1921–1926.
12. Frisch M, Biggar RJ, Engels EA, Goedert JJ. Association of cancer with AIDS-related immunosuppression in adults. JAMA 2001; 285:1736–1745.
13. Sulkowski MS, Thomas DL, Chaisson RE, Moore RD. Hepatotoxicity associated with antiretroviral therapy in adults infected with human immunodeficiency virus and the role of hepatitis C or B virus infection. JAMA 2000; 283:74–80.
14. Clifford GM, Polesel J, Rickenbach M, Dal Maso L, Keiser O, Kofler A, et al
. Cancer risk in the Swiss HIV cohort study: associations with immunodeficiency, smoking, and highly active antiretroviral therapy. J Natl Cancer Inst 2005; 97:425–432.
15. Newnham A, Harris J, Evans HS, Evans BG, Moller H. The risk of cancer in HIV-infected people in southeast England: a cohort study. Br J Cancer 2005; 92:194–200.
16. Galceran J, Marcos-Gragera R, Soler M, Romaguera A, Ameijide A, Izquierdo A, et al
. Cancer incidence in AIDS patients in Catalonia, Spain. Eur J Cancer 2007; 43:1085–1091.
17. Dal Maso L, Franceschi S, Polesel J, Braga C, Piselli P, Crocetti E, et al
. Risk of cancer in persons with AIDS in Italy, 1985–1998. Br J Cancer 2003; 89:94–100.
18. Puoti M, Bruno R, Soriano V, Donato F, Gaeta GB, Quinzan GP, et al
. Hepatocellular carcinoma in HIV-infected patients: epidemiological features, clinical presentation and outcome. AIDS 2004; 18:2285–2293.
19. Brau N, Fox RK, Xiao P, Marks K, Naqvi Z, Taylor LE, et al
. Presentation and outcome of hepatocellular carcinoma in HIV-infected patients: a U.S.-Canadian multicenter study. J Hepatol 2007; 47:527–537.
20. Colin JF, Cazals-Hatem D, Loriot MA, Martinot-Peignoux M, Pham BN, Auperin A, et al
. Influence of human immunodeficiency virus infection on chronic hepatitis B in homosexual men. Hepatology 1999; 29:1306–1310.
21. Puoti M, Spinetti A, Ghezzi A, Donato F, Zaltron S, Putzolu V, et al
. Mortality for liver disease in patients with HIV infection: a cohort study. J Acquir Immune Defic Syndr 2000; 24:211–217.
22. Feitelson MA. Hepatitis B virus in hepatocarcinogenesis. J Cell Physiol 1999; 181:188–202.
23. Rockstroh JK, Spengler U, Sudhop T, Ewig S, Theisen A, Hammerstein U, et al
. Immunosuppression may lead to progression of hepatitis C virus-associated liver disease in hemophiliacs coinfected with HIV. Am J Gastroenterol 1996; 91:2563–2568.
24. Martin-Carbonero L, Benhamou Y, Puoti M, Berenguer J, Mallolas J, Quereda C, et al
. Incidence and predictors of severe liver fibrosis in human immunodeficiency virus-infected patients with chronic hepatitis C: a European collaborative study. Clin Infect Dis 2004; 38:128–133.
25. McGovern BH, Golan Y, Lopez M, Pratt D, Lawton A, Moore G, et al
. The impact of cirrhosis on CD4+ T cell counts in HIV-seronegative patients. Clin Infect Dis 2007; 44:431–437.
26. Puoti M, Cozzi-Lepri A, Arici C, Moller NF, Lundgren JD, Ledergerber B, et al
. Impact of lamivudine on the risk of liver-related death in 2041 HBsAg- and HIV-positive individuals: results from an inter-cohort analysis. Antivir Ther 2006; 11:567–574.
27. Qurishi N, Kreuzberg C, Luchters G, Effenberger W, Kupfer B, Sauerbruch T, et al
. Effect of antiretroviral therapy on liver-related mortality in patients with HIV and hepatitis C virus coinfection. Lancet 2003; 362:1708–1713.