As the HIV-infected population ages, non-AIDS-defining cancers have become a major cause of death, and in recent years the incidence of non-AIDS-defining cancers has surpassed that of AIDS-defining cancers [1–3]. The mechanisms behind the increased risk of non-AIDS-defining cancers in HIV patients compared to the background population have not been fully elucidated. Smoking and other lifestyle-associated risk factors are more prevalent among HIV-infected than in uninfected individuals [4–6]. A number of studies have demonstrated higher levels of markers of inflammation and immune activation in HIV-infected individuals and shown associations with all-cause mortality . It has been hypothesized that HIV infection causes accelerated aging, but the relative contributions of immunodeficiency versus lifestyle-related factors, such as smoking, substance abuse and sexual behaviour to morbidity and mortality, among HIV-infected individuals are controversial [8,9]. We have previously shown that smoking is associated with a larger loss of live years than HIV itself among HIV patients in Denmark .
The objective of the present study was to estimate the population-attributable fractions (PAFs) of cancer that are associated with smoking, immune deficiency and with being HIV-infected. We therefore compared the incidence rates of cancer in the HIV-infected population with those of a matched cohort from the background population. Analyses were stratified on smoking status and nadir CD4+/duration of CD4+ below 200 cells/μl, and cancers were categorized as smoking-related, related to viral infections or ‘other’.
The healthcare system in Denmark is funded through taxes. All pathology specimens and cases of cancer are registered with dates and diagnoses in nationwide registries.
HIV care takes place at eight highly specialized centres in the country. In 2010, an estimated 70% of individuals diagnosed with HIV had HIV-RNA below 500 copies/ml .
The Danish civil registration system
A national register was established in 1967 which contains data on migration and viral status of all Danish citizens .
The Danish cancer registry
It was established in 1943 and since March 1987 reporting of data on all malignant diseases has been mandatory . Approximately 90% of cases in the registry are verified by histological examination. There is an extensive data validation programme which ensures the quality and completeness of the registry.
The Danish HIV cohort study
It is a population-based nationwide cohort study, described in details elsewhere , of all HIV-infected individuals who receive care at Danish HIV centres after 1 January 1995. Individuals are consecutively enrolled. We included all HIV-1-infected individuals who were 16 years or above at HIV diagnosis, alive and received care in Danish HIV centres 1 year after HIV diagnosis in the period between 1 January 1995 and 31 December 2011, and had available data on smoking status.
The Copenhagen general population study
This is a prospective study of a cohort of individuals randomly selected from greater Copenhagen [14–16]. The study participants were interviewed about lifestyle and health-related factors. From this study, we identified a population control cohort which was frequency-matched on sex and age (5-year intervals). We aimed to include four population controls for each HIV-infected individual, but because of few participants aged below 35 years in the Copenhagen General Population Study, only two population controls were included for each HIV-infected individual aged below 35 years. All analyses were adjusted for age.
The date of study inclusion was the date 1 year after HIV diagnosis or the date of first available data on smoking status, whichever came last. Individuals diagnosed with cancer prior to this date were excluded. Study participants were followed until the date of study outcome, emigration, death or 31 December 2011, whichever occurred first.
Exposures and outcomes
We analysed the following exposures: smoking (current or previous versus never), HIV (HIV-positive versus negative status) and two measures of immune deficiency: nadir CD4+ cell count below 200 cells/μl, as well as CD4+ below 200 cells/μl for at least 2 years before start of follow-up.
The study outcome was cancer (excluding non-melanoma skin cancer) as registered in the National Danish Cancer Registry. Cancer diagnoses were categorized as smoking-related, virological or other.
Individuals smoking any type of tobacco were categorized as smokers. Individuals were categorized as ever-smokers or never-smokers on the basis of information on smoking status at the time of enrolment, and did not change category during the observation period.
Nadir CD4+ cell count was the lowest count measured from date of HIV diagnosis until 90 days after study inclusion. Cumulative duration of immune suppression was assessed from date of first CD4+ cell count until study inclusion [median 4.5 years, inter-quartile range (IQR) 1.2–8.0] and was calculated as the sum of time intervals from the date of CD4+ cell count below 200 until date of a CD4+ cell count at least 200 cells/μl. If the first measured CD4+ cell count was less than 200 cells/μl, it was assumed that the CD4+ cell count had declined at a rate of 50 cells/μl/year .
Cancers were coded according to the International Classifications of Diseases, 10th version (ICD-10). Cancers, in which data from WHO indicate that the proportion attributed to smoking is above 40%, were classified as smoking-related . Cancers categorized as smoking-related were: lung cancer (C34), head and neck cancer (C0–C14 and C30–C32), oesophagus cancer (C15) and bladder cancer (C67). Cancers classified as virological were: lymphomas (C81–C86), Kaposi sarcoma (C46), liver cell carcinoma (C22.0), anal cancer (C21), cervical cancer (C53), vulva cancer (C51) and penile cancer (C60). All other cancers were categorized as ‘other’. We did not include non-melanoma skin cancers (C44).
Incidence rates were calculated as the numbers of first diagnoses of cancers per 10 000 person-years. Excess incidence rates were calculated using the formula: Excess incidence rate = incidence rateexposed − incidence rateunexposed. Adjusted incidence rate ratios (IRRs) of cancer according to HIV, smoking and immune status were estimated using Poisson regression analyses. In analyses including only HIV patients, we adjusted for the following variables: sex, age (time updated in 2-year intervals), route of transmission (MSM, heterosexual, injection drug use or other), duration of ART (time updated in 1-year intervals) and year of study inclusion. In analyses including population controls, we included sex, age, smoking status and year of study inclusion in the model. In comparisons of HIV-infected individuals versus population controls, the interaction term for smoking and HIV was insignificant in all analyses, whereas the interaction term was excluded from final analyses.
Population-attributable fractions were calculated using the formula: PAF = Pe(RRe − 1)/[1 + Pe(RRe − 1)], where Pe is the proportion of population with exposure and RRe is the relative risk of cancer among exposed compared to unexposed, adjusted for the variables mentioned above.
The study was approved by the Danish Data Protection Agency. Ethics approval and individual consent are not required by Danish legislation governing this type of study on the HIV-infected individuals; however, studies on population controls were approved by a Danish ethical committee (#H-KF-01–144/01) and all population controls provided written informed consent.
SPSS statistical software, Version 15.0 (Norusis; SPSS Inc., Chicago, Illinois, USA) and Stata, Version 8.0 (Stata Corporation, College Station, Texas, USA) were used for data analysis.
We included 3503 HIV patients and 12 979 matched population controls who were followed for a total of 18 679 and 55 957 person-years, respectively. At time of study inclusion, 2685 (77%) HIV patients were on ART, median 3.3 years (IQR 1.0–6.6) and in 92% of the observation time the HIV patients were on ART. The median CD4+ cell count at study inclusion was 450 (IQR 310–630). The proportion of ever-smokers was larger among HIV patients compared to population controls (67 versus 53%) and the proportion of current smokers was larger among ever-smoking HIV patients compared to ever-smoking population controls (76 versus 39%). The proportions of men, people of Danish origin and injection drug users were higher among ever-smoking compared to never-smoking HIV patients (Table 1). Among population controls, the median age was higher for ever-smokers compared to never-smokers.
Risk of cancer among HIV patients versus population controls
The incidence rates of cancers, by category and location among HIV patients and population controls, are summarized in Table 2. The IRR of all non-skin cancers was 2.0 [95% confidence interval (CI) 1.6–2.5] among HIV patients compared to population controls, with virological cancers accounting for the highest excess as well as the highest relative risk (excess incidence rate 33.2 per 10 000 person-years, 95% CI 24.4–42.0; IRR 11.5, 95% CI 6.5–20.5). The absolute and relative risks of smoking-related cancers were also increased among HIV patients compared to population controls (excess incidence rate 12.3 per 10 000 person-years, 95% CI 5.6–19.0; IRR 2.8, 95% CI 1.6–4.9), whereas the risk of other cancers did not differ (excess incidence rate −7.0 per 10 000 person-years, 95% CI −16.1–2.1; IRR 1.0, 95% CI 0.7–1.3).
Among non-smokers, the increase in risk of cancer associated with HIV infection was confined to virological cancers (Table 3).
Risk of cancer associated with smoking
Smoking was not associated with risk of virological or ‘other’ cancers among HIV patients (Table 3). For smoking-related cancers, the absolute and relative risks associated with smoking were higher among HIV patients than population controls. There was no significant interaction between smoking and HIV. Only one of 15 head and neck cancers and none of 17 lung cancers occurred among never-smoking HIV patients, whereas three of 17 lung cancers and four of 17 head and neck cancers occurred among never-smoking population controls. The risk of non-virological cancers did not differ among never-smoking HIV patients compared to never-smoking population controls.
Risk of cancer associated with current and previous smoking
The IRRs of cancer associated with current versus never-smoking were 1.74 (95% CI 1.18–2.57) among HIV patients and 1.92 (95% CI 1.38–2.65) among controls. The IRRs of previous versus never-smoking were 1.09 (95% CI 0.66–1.81) among HIV patients and 1.44 (95% CI 1.06–1.97) among controls. The IRRs of smoking-related cancer associated with current smoking were 21.35 (95% CI 2.88–158.5) among HIV patients and 4.12 (95% CI 1.74–9.78) among controls. The IRR of previous smoking were 8.28 (95% CI 0.99–69.1) among HIV patients and 1.20 (95% CI 0.46–3.14) among controls. Neither current nor previous smoking was associated with increased risk of virological or other cancers among HIV patients. Tests for interaction between HIV and smoking were insignificant in all analyses.
Risk of cancer associated with immune deficiency
Nadir CD4+ below 200 cells/μl was associated with increased risk of smoking-related cancer (Table 4). The risk associated with nadir CD4+ below 200 cells/μl was increased for lung cancer (IRR 3.54, 95% CI 1.00–12.59), but not for head and neck cancer (IRR 1.49, 95% CI 0.45–5.00). There was no association between duration of CD4+ below 200 cells/μl and risk of lung cancer (IRR 1.07 per year, 95% CI 0.87–1.32). No individual with nadir CD4+ at least 200 cells/μl developed smoking-related cancer. For other cancer types, the risk associated with smoking did not differ between individuals with versus without low CD4+ cell counts.
Population-attributable fractions associated with smoking, immune deficiency and with being HIV-infected
The relative contributions of the three categories of cancers in the HIV-infected population versus the control cohort, and the proportions associated with smoking are illustrated in Fig. 1. Smoking-related and virological cancers accounted for 23 and 43% of cancers in the HIV-infected population, respectively, whereas virological cancers were rare among population controls. The fractions of all cancers in the HIV-infected population attributable to smoking and to being HIV-infected were 27 and 49%, respectively (Table 3). For cancer types that are considered strongly related to smoking, the proportion attributable to smoking was 91% and the fraction of virological cancers attributable to being HIV-infected was 91%. Other cancers were not associated with immune deficiency or with being HIV-infected, and the fraction associated with smoking was marginal (Tables 3 and 4).
In this nationwide, population-based cohort study, we found that HIV patients, with long-term engagement in care, had increased risk of cancers that are considered strongly related to smoking or viral infections. In contrast, they did not have increased risk of other cancers compared to the background population. We found no association between immune deficiency and non-virological cancers that are not strongly related to smoking.
Cancers related to smoking and viral infections accounted for 23 and 43% of all cancers, respectively, and for these cancers, the PAFs associated with smoking and with being HIV-infected, respectively, were approximately 90%. Immune deficiency had no impact on risk of non-virological cancers in absence of smoking in this population, where 75% of the HIV patients were on antiretroviral therapy (ART) at time of study inclusion and 92% of observation time was during ART. If observation time prior to ART was excluded, the estimates altered only marginally and conclusions did not change (data not shown).
Strengths of our study include the nationwide, population-based design with a matched control group from the background population with information on smoking status. We obtained data on study outcomes for both patients and population controls from the National Cancer Registry which is regularly validated and where approximately 90% of cases are verified by histology; thus the risk of bias due to misclassifications is limited. In Denmark, it is mandatory to report all malignancies to this registry. We therefore believe that ascertainment of study outcomes was high.
The study is limited by the rather low number of study outcomes, which prevented us from assessing specific cancer types or differences by sex or age groups. We categorized poorly differentiated tumours as ‘other’; some of these could be lymphoma and we may thereby have misclassified some virological cancers as ‘other’. All population controls were of Danish origin and resident in greater Copenhagen, which could potentially introduce a bias. However, analyses restricted to HIV patients of Danish origin or those followed in HIV centres in Copenhagen yielded similar results (data not shown). Individuals who have volunteered to participate in the Copenhagen General Population study, which involves outpatient visits and interviews on lifestyle, may represent a healthier subset of the background population.
We did not have data on duration of smoking, and data on the number of cigarettes smoked per day were only available for a sub-group of HIV patients. We were therefore not able to determine if the increased risk associated with smoking among HIV patients compared to population controls was caused by a higher proportion of heavy smokers in the smoking HIV-infected population or an interaction between adverse effects of tobacco and HIV. Previous studies have found very high rates of heavy smokers among HIV patients . In our cohort, 20% of those with available data on amount of smoking smoked more than 20 cigarettes per day. The distribution of current and previous smokers differed between HIV patients and population controls categorized as ‘ever-smokers’, with a higher proportion of current smokers among HIV patients. We believe this is likely the most probable explanation for higher risk associated with smoking among HIV patients compared to population controls. We analysed the risk associated with ever-smoking rather than current smoking because most cancers develop over several years. We did not have data on time of smoking cessation. However, continuous smoking and number of pack-years may not explain the complete difference in risk of smoking-related cancer among HIV patients compared to population controls. We found that the risk of lung cancer was higher among HIV patients with low versus high nadir CD4+ cell count, although there was no association with duration of immune deficiency. Individuals with low nadir CD4+ cell count were older, but analyses were adjusted for age, and it is possible that immune deficiency may exacerbate the effect of smoking on risk of lung cancer in the HIV-infected population.
We categorized head and neck cancers as smoking-related. Some head and neck cancers are related to human papilloma virus (HPV) , and there are data indicating that the risk associated with HPV is exacerbated by smoking . For cervical and anal cancers, which are strongly associated with HPV, there is an association between low CD4+ cell count (for anal cancer: cumulative time with low CD4+ cell count) and increased risk of cancer . We did not have data on the proportions of head and neck cancers that were associated with HPV. When we categorized tonsil cancers, which have the strongest association with HPV, as virological the results did not change markedly (data not shown). We found no association between immune deficiency and risk of head and neck cancer.
Our results are similar to findings from previous studies showing that HIV patients have a 2–3-fold increased risk of cancer compared to the background population [23–25]. The impact of impaired immune surveillance and accelerated aging on risk of non-AIDS cancers has been vividly debated [8,26,27]. In the present study we found that the increased risk of non-AIDS cancer was largely confined to cancers associated with smoking and viral infections, for example, anal cancer, liver cell carcinoma and Hodgkin's lymphoma, of which the two former are associated with sexual behaviour and injection drug use [28,29]. In contrast, the risk of cancers that are not considered strongly related to smoking or viral infections did not differ between the HIV-infected and the background population, and the impact of immune deficiency was limited. Other studies have reached the opposite conclusion regarding risk of non-AIDS cancers among HIV patients compared to the background population . The majority of HIV-infected individuals in the present study were on ART, had a high CD4+ cell count at baseline, and did not represent a socially deprived part of the population, which may explain the discrepancy between conclusions of these studies.
In untreated HIV-infected populations and among individuals recently enrolled in care immune deficiency is likely to have a much larger impact on the overall incidence of cancer . Previous studies have shown that rates of AIDS-defining cancers have decreased markedly, whereas rates of non-AIDS cancers have remained constant or increased parallel with the increasing age of the HIV-infected population from the pre-combination ART to the early and late-combination ART periods [31,32]. Patients who were diagnosed with cancer within the first year of HIV diagnosis were not included in the present study. It is well known that a low CD4+ cell count at the time of HIV diagnosis is associated with significantly increased risk of AIDS-defining malignancy within the first year  and our estimate of the proportion of virological cancers is therefore likely to be considerably lower than among individuals recently diagnosed with HIV.
We found that for cancers overall, the PAF associated with being HIV-infected was approximately 50%. The PAF was estimated by comparing the HIV-infected cohort to a cohort from the Copenhagen General Population Study, which was matched for sex and age. However, individuals in these cohorts differed in many ways in addition to HIV status. Thus, both biological and behavioural factors may contribute to the observed associations between being HIV-infected and risk of cancer. In addition to immune deficiency and immune activation, other factors such as co-infections, substance abuse, heavy smoking, sexual behaviour and so on are likely to contribute to the increased risk of cancers related to smoking and viral infections.
Smoking may be associated with socio-economic status, alcohol intake and poor diet , which could confound the observed associations with smoking.
Under the assumption that most cancers develop over several months or years, we used nadir CD4+ cell count and duration of CD4+ below 200 cells/μl at study inclusion as a measures of immune deficiency rather than latest CD4+ cell count. We have previously shown that the risk of cancer is substantially increased the first 6 months following a drop in CD4+ cell count among virally suppressed HIV patients, indicating that CD4+ decline can be a marker of undiagnosed cancer . This could confound analyses of associations between immune deficiency and cancer if latest CD4+ cell count is used as a marker of immune deficiency.
The date of study inclusion was at least 1 year after HIV diagnosis and the majority of HIV patients in the study was on ART at study inclusion and had high CD4+ cell counts; thus results can only be generalized to HIV-infected populations engaged in care. Associations between smoking, HIV infection and cancer may be different in HIV-infected populations with a different distribution regarding age, sex, access to care, lifestyle and socio-economic status.
We conclude that in a HIV-infected population with long-term engagement in care, the risk of cancer is increased approximately two-fold compared to the background population. The risk is increased for smoking-related and virological cancers, whereas the risk of other cancers does not differ between HIV-infected and uninfected individuals, and does not seem to be associated with immune deficiency. Among non-smokers, the increase in risk of cancer associated with HIV infection is confined to cancers related to viral infections.
We thank the staff of our clinical departments for their continuous support and enthusiasm.
All of the authors contributed to the conception and design of the study and analyses and interpretation of data. The manuscript was drafted by M.H., J.G. and N.O., and was critically reviewed and approved by all authors.
No funding sources were involved in study design, data collection, analysis, report writing, or the decision to submit the paper.
Conflicts of interest
N.O. has received research funding from Bristol-Myers Squibb, Merck Sharp & Dohme, GlaxoSmithKline, Abbott, Boehringer Ingelheim, and Gilead. C.P. has received research funding from Abbott, Roche, Bristol-Myers Squibb, Merck Sharp & Dohme, GlaxoSmithKline, Swedish Orphan, Jansen Pharma/Tibotec and Boehringer Ingelheim. J.G. has received research funding from Abbott, Roche, Bristol-Myers Squibb, Merck Sharp & Dohme, ViiV, Swedish Orphan and Gilead.
The remaining authors report no conflicts of interest.
1. Palella FJ Jr, Baker RK, Moorman AC, Chmiel JS, Wood KC, Brooks JT, et al. Mortality in the highly active antiretroviral therapy era: changing causes of death and disease in the HIV outpatient study
. J Acquir Immune Defic Syndr
2. Helleberg M, Kronborg G, Larsen CS, Pedersen G, Pedersen C, Gerstoft J, Obel N. Causes of death among Danish HIV patients compared with population controls in the period 1995–2008
3. Simard EP, Pfeiffer RM, Engels EA. Mortality due to cancer among people with AIDS: a novel approach using registry-linkage data and population attributable risk methods
4. Armah KA, McGinnis K, Baker J, Gibert C, Butt F AA, Bryant KJ, et al. HIV status, burden of comorbid disease, and biomarkers of inflammation, altered coagulation, and monocyte activation
. Clin Infect Dis
5. Helleberg M, Afzal S, Kronborg G, Larsen CS, Pedersen G, Pedersen C, et al. Mortality attributable to smoking among HIV-1-infected individuals: a nationwide, population-based cohort study
. Clin Infect Dis
6. Galvan FH, Bing EG, Fleishman JA, London AS, Caetano R, Burnam MA, et al. The prevalence of alcohol consumption and heavy drinking among people with HIV in the United States: results from the HIV Cost and Services Utilization Study
. J Stud Alcohol
7. Deeks SG, Lewin SR, Havlir DV. The end of AIDS: HIV infection as a chronic disease
8. Cutrell J, Bedimo R. Non-AIDS-defining Cancers Among HIV-Infected Patients
. Curr HIV /AIDS Rep
9. Shcherba M, Shuter J, Haigentz M Jr. Current questions in HIV-associated lung cancer
. Curr Opin Oncol
10. Helleberg M, Haggblom A, Sonnerborg A, Obel N. HIV care in the Swedish-Danish HIV Cohort 1995–2010, closing the gaps
. PLoS One
11. CPR-kontoret. Det Centrale Personregister. www.cpr.dk
[Accessed 4 July 2013]
12. Gjerstorff ML. The Danish Cancer Registry
. Scand J Public Health
13. Obel N, Engsig FN, Rasmussen LD, Larsen MV, Omland LH, Sorensen HT. Cohort profile: the Danish HIV cohort study
. Int J Epidemiol
14. Zacho J, Tybjaerg-Hansen A, Jensen JS, Grande P, Sillesen H, Nordestgaard BG. Genetically elevated C-reactive protein and ischemic vascular disease
. N Engl J Med
15. Langsted A, Nordestgaard BG. Nonfasting lipids, lipoproteins, and apolipoproteins in individuals with and without diabetes: 58 434 individuals from the Copenhagen General Population Study
. Clin Chem
16. Nordestgaard BG, Palmer TM, Benn M, Zacho J, Tybjaerg-Hansen A, Davey Smith G, et al. The effect of elevated body mass index on ischemic heart disease risk: causal estimates from a Mendelian randomisation approach
. PLoS Med
17. Helleberg M, Kronborg G, Larsen CS, Pedersen G, Pedersen C, Obel N, Gerstoft J. No change in viral set point or CD4 cell decline among antiretroviral treatment-naive, HIV-1-infected individuals enrolled in the Danish HIV Cohort Study in 1995–2010
. HIV Med
18. Danaei G, Vander HS, Lopez AD, Murray CJ, Ezzati M. Causes of cancer in the world: comparative risk assessment of nine behavioural and environmental risk factors
19. Browning KK, Wewers ME, Ferketich AK, Diaz P. Tobacco use and cessation in HIV-infected individuals
. Clin Chest Med
20. Gillison ML, Koch WM, Capone RB, Spafford M, Westra WH, Wu L, et al. Evidence for a causal association between human papillomavirus and a subset of head and neck cancers
. J Natl Cancer Inst
21. Sinha P, Logan HL, Mendenhall WM. Human papillomavirus, smoking, and head and neck cancer
. Am J Otolaryngol
22. Guiguet M, Boue F, Cadranel J, Lang JM, Rosenthal E, Costagliola D. Effect of immunodeficiency, HIV viral load, and antiretroviral therapy on the risk of individual malignancies (FHDH-ANRS CO4): a prospective cohort study
. Lancet Oncol
23. Franceschi S, Lise M, Clifford GM, Rickenbach M, Levi F, Maspoli M, et al. Changing patterns of cancer incidence in the early- and late-HAART periods: the Swiss HIV Cohort Study
. Br J Cancer
24. Albini L, Calabresi A, Gotti D, Ferraresi A, Festa A, Donati F, et al. Burden of non-AIDS-defining and non-virus-related cancers among HIV-infected patients in the combined antiretroviral therapy era
. AIDS Res Hum Retroviruses
25. Calabresi A, Ferraresi A, Festa A, Scarcella C, Donato F, Vassallo F, et al. Incidence of AIDS-defining cancers and virus-related and nonvirus-related non-AIDS-defining cancers among HIV-infected patients compared with the general population in a large health district of northern Italy, 1999–2009
. HIV Med
26. Frisch M, Biggar RJ, Engels EA, Goedert JJ. Association of cancer with AIDS-related immunosuppression in adults
27. Dubrow R, Silverberg MJ, Park LS, Crothers K, Justice AC. HIV infection, aging, and immune function: implications for cancer risk and prevention
. Curr Opin Oncol
28. MacDonald DC, Nelson M, Bower M, Powles T. Hepatocellular carcinoma, human immunodeficiency virus and viral hepatitis in the HAART era
. World J Gastroenterol
29. Piketty C, Selinger-Leneman H, Bouvier AM, Belot A, Mary-Krause M, Duvivier C, et al. Incidence of HIV-related anal cancer remains increased despite long-term combined antiretroviral treatment: results from the french hospital database on HIV
. J Clin Oncol
30. Silverberg MJ, Chao C, Leyden WA, Xu L, Horberg MA, Klein D, et al. HIV infection, immunodeficiency, viral replication, and the risk of cancer
. Cancer Epidemiol Biomarkers Prev
31. Shiels MS, Pfeiffer RM, Gail MH, Hall HI, Li J, Chaturvedi AK, et al. Cancer burden in the HIV-infected population in the United States
. J Natl Cancer Inst
32. Powles T, Robinson D, Stebbing J, Shamash J, Nelson M, Gazzard B, et al. Highly active antiretroviral therapy and the incidence of non-AIDS-defining cancers in people with HIV infection
. J Clin Oncol
33. Yanik EL, Napravnik S, Cole SR, Achenbach CJ, Gopal S, Olshan A, et al. Incidence and timing of cancer in HIV-infected individuals following initiation of combination antiretroviral therapy
. Clin Infect Dis
34. Robinson WT, Brown MC, Moody-Thomas S. Smoking and experiences with tobacco cessation among men who have sex with men: New Orleans, 2011
. AIDS Behav
35. Helleberg M, Kronborg G, Larsen CS, Pedersen G, Pedersen C, Obel N, Gerstoft J. CD4 decline is associated with increased risk of cardiovascular disease, cancer, and death in virally suppressed patients with HIV
. Clin Infect Dis