Big 3 Lung Cancer Risk Factors With HIV: Smoking, Low CD4s, Lung Infection

July 3, 2018


All the lung cancer risk factors that apply to the general population also apply to people with HIV infection. Smoking is by far the top lung cancer risk factor for people with and without HIV. In HIV populations smoking may account for 94% of lung cancers. A simulation model predicts that HIV-positive men and women who quit smoking at age 40 dramatically cut their risk of dying from lung cancer through age 80. Two other factors inflate lung cancer risk in people with HIV -- compromised immunity indicated by various CD4-cell measures and prior lung disease. Although low CD4 count did not predict lung cancer in models adjusted for most major risk factors, a low CD4/CD8 ratio did remain an independent predictor in a study of US veterans. Prior bacterial pneumonia also predicted lung cancer in the fully adjusted model. Prior Pneumocystis pneumonia or chronic obstructive pulmonary disease also boosts lung cancer risk in some analyses.

Smoking. It so dominates the lung cancer risk list that other factors struggle for a small slice of the spotlight. That sounds like hyperbole, but NA-ACCORD researchers recently produced compelling data to back it up. If no one with HIV ever smoked, they figured 94% of all lung cancers would be prevented in people with HIV.1

That startling statistic comes from an analysis of 52,411 North American adults treated for HIV in whom 2306 cancers developed in about 4 years.1 Splitting the population into ever-smokers and never-smokers, NA-ACCORD researchers used Cox proportional hazards models adjusted for demographic and clinical factors to determine that ever smoking boosted the risk of all cancers 33% (hazard ratio [HR] 1.33, 95% confidence interval [CI] 1.18 to 1.49), more than doubled the risk of smoking-related cancers (HR 2.31, 95% CI 1.80 to 2.98), and upped the risk of lung cancer almost 18-fold (HR 17.8, 95% CI 5.60 to 56.63).

Adjusted population-attributable fraction analysis2 showed that never smoking would prevent 19% of all cancers, 50% of smoking-related cancers, and a whopping 94% of lung cancers.1 If one accepts the estimate that about 33,700 HIV-positive people in the United States have lung cancer, never smoking would slice prevalence to 2022.

Because available data dictated that NA-ACCORD researchers compare ever-smokers with never-smoker, they could not estimate lung cancer risk in HIV-positive people who quit smoking. But other researchers have scrutinized how quitting affects lung cancer mortality. For the general population, quitting for 10 years halves the risk of dying from lung cancer.3

Related: Lung Cancer Risk and Screening With HIV -- and an Intriguing New Treatment Opportunity

Boston collaborators used a simulation model to make a similar analysis in HIV-positive people in care in the United States.4 Among 40-year-old men who continued to smoke, estimated cumulative mortality from lung cancer by age 80 stood at 28.9% in heavy smokers, 23.0% in moderate smokers, and 18.8% in light smokers. Among men who quit smoking at age 40, cumulative lung cancer mortality by age 80 plunged to 7.9% in heavy smokers, 6.1% in moderate smokers, and 4.3% in light smokers. For 40-year-old women who kept smoking, cumulative lung cancer mortality at age 80 measured 27.8% in heavy smokers, 20.9% in moderate smokers, and 16.6% in light smokers. Among women who quit at age 40, those death rates swooned to 7.5%, 5.2%, and 3.7%. Men who never smoked had a cumulative lung cancer mortality of 1.6% at age 80, while women who never smoked had a cumulative 1.2% lung cancer death rate at age 80. If smokers among an estimated 644,240 HIV-positive US residents between ages 20 and 64 do not quit smoking, the authors project that 59,900 (9.3%) will die from lung cancer.

Smoking tops American Cancer Society and CDC lists of lung cancer risk factors in the general population, followed by several other factors that all apply to people with HIV (Table 1). Researchers believe smoking can trigger lung cancer in two ways, (1) by suppressing pulmonary immunologic defenses, and (2) by torqueing up systemic immune activation.1 Two other portentous lung cancer risk factors specific to people with HIV also work by upsetting the immune system -- low CD4 count and previous lung disease.

Table 1. Lung Cancer Risk Factors in the General Population
  • Smoking
  • Second-hand smoke
  • Older age
  • Personal or family history of lung cancer
  • Radiation therapy to chest
  • Air pollution
  • Exposure to radon
  • Arsenic in drinking water
  • Exposure to workplace substances including diesel exhaust, silica, and chromium
  • Beta carotene supplements in smokers

Sources: American Cancer Society. Lung cancer risk factors; CDC. Lung cancer. What are the risk factors for lung cancer?

Impact of Low CD4 Count on Lung Cancer Risk

Studies of nonlung cancers in people with HIV suggest that prolonged periods at a low CD4 count heighten cancer risk. But similar lung cancer studies in HIV populations yield diverse results impossible to summarize in a neat declarative sentence.


National Cancer Institute (NCI) researchers offered an early analysis of immunosuppression and lung cancer risk in early antiretroviral days (1978 to 1996).5 Matching 302,834 adults with AIDS to cancer registry data, they found that lung cancer met all 3 criteria they had set for potential association with immunosuppression: (1) elevated overall relative risk (RR) from 60 months before to 27 months after AIDS diagnosis; (2) elevated RR in the 4- to 27-month post-AIDS period; and (3) increasing trend in RR from before to after AIDS onset.

Three large longitudinal cohort studies in the United States and France link poor immune control -- measured various ways -- to incident lung cancer.6-8 The US study involves 21,666 HIV-positive Veterans Aging Cohort Study (VACS) members tracked for at least 3 years in a period spanning January 1998 through December 2012.6 During a median follow-up of 7.4 years, lung cancer developed in 277 veterans (1.3%).

Cox regression models adjusted for age, race/ethnicity, smoking, HCV infection, alcohol or drug use, and history of chronic obstructive pulmonary disease (COPD) or occupational lung disease identified three immune predictors of incident lung cancers in 12-month time-lagged analyses: low CD4 count, low CD4/CD8 ratio, and high viral load (HIV RNA) (Table 2).6 These analyses were time-lagged for 12 months to account for the possibility that falling CD4 counts could reflect undiagnosed lung cancer. More cumulative bacterial pneumonia episodes also predicted incidentlung cancer, as discussed in the following section. In an additional analysis that adjusted for all these factors, low CD4/CD8 ratio and bacterial pneumonia remained independent predictors of lung cancer.

Table 2. Independent Predictors of Lung Cancer in U.S. Veterans, 1998-2012*
VariableHazard Ratio (95% Confidence Interval)
CD4 count 100 to 199 (vs over 500)2.3 (1.6 to 3.4)
CD4 count 200 to 500 (vs over 500)1.2 (1.2 to 2.1)
CD8 count over 1000 (vs under 600)1.5 (0.9 to 2.1)
CD8 count 600 to 1000 (vs under 600)1.4 (1.0 to 2.3)
CD4/CD8 ratio below 0.4 (vs above 1.0)2.6 (1.6 to 4.1)
CD4/CD8 ratio 0.4 to 1.0 (vs above 1.0)1.9 (1.2 to 3.0)
Viral load above 500 (vs under 500)1.4 (1.1 to 1.9)
Bacterial pneumonia, 2 of more episodes (vs 0)1.8 (1.0 to 2.5)
Bacterial pneumonia, 1 episode (vs 0)1.7 (1.2 to 2.4)
* 12-month lagged analysis. Source: Sigel K, et al. 2017.6

Keith Sigel, principal investigator of the VACS study,6 suggested two interesting implications of linking a low CD4/CD8 ratio to higher lung cancer risk: (1) low CD4/CD8 ratio may indicate "a persistent immune disturbance that contributes to inflammation or abnormal cancer surveillance" and (2) the CD4/CD8 ratio may be a useful lung cancer risk factor even in people with an apparently normal CD4 count.9

A study of 52,278 members of the French Hospital Database on HIV through a median follow-up of 4.9 years in 1998-2006 tested 78 different models to estimate the impact of current CD4 count on lung cancer risk.7 Compared with a current count above 500 cells/mm3, a CD4 tally between 350 and 499 cells/mm3 inflated lung cancer risk from 2.2 to 8.5 times, depending on the model (rate ratio [RR] 2.2, 95% CI 1.3 to 3.6, to 8.5, 95% CI 4.3 to 16.7, P < 0.0001).

More recently some of these same French investigators compared incidence of four non-AIDS cancers in 84,504 people with HIV and the general population by calculating standardized incidence ratios (SIRs) in three periods: early combination ART (1997-2000), intermediate ART (2001-2004), and late ART (2005-2009).8 Through an average 6.8 years of follow-up for each person with HIV, lung cancer incidence remained higher with HIV across the three study periods, though the difference diminished with time:

  • 1997-2000 SIR 4.7 (95% CI 4.1 to 5.5)
  • 2001-2004 SIR 3.8 (95% CI 3.3 to 4.3)
  • 2005-2009 SIR 2.8 (95% CI 2.5 to 3.1)

Among people who reached a CD4 count above 500 cells/mm3 for at least 2 years on ART, lung cancer incidence became nearly equivalent in people with and without HIV (SIR 0.9, 95% CI 0.6 to 1.3) but not in people with two other non-AIDS cancers, Hodgkin lymphoma (SIR 9.4, 95% CI 7.9 to 16.8) and liver cancer (SIR 2.4, 95% CI 1.4 to 4.1).8 Together these findings suggest that gaining CD4 cells with stronger antiretroviral regimens trims lung cancer risk. Besides reduced pulmonary inflammation with immune recovery, the authors note that decreased incidence of recurrent pneumonia probably contributed to falling lung cancer incidence (2.4 pneumonia cases per 1000 person-years in 1997 to 0.9 per 1000 in 2009)

A comparison of 24,768 HIV-positive people and 257,600 HIV-negatives in California's Kaiser Permanente healthcare system during 1996-2011 also marks low CD4 count as a lung cancer predictor, but not after adjustment for pneumonia.10

The Kaiser team offers three layers of adjustment for lung cancer risk: (1) demographics (age, sex, race/ethnicity, calendar period), (2) cancer risk factors (smoking, drug/alcohol abuse, overweight/obesity), and (3) any history of pneumonia (Pneumocystis pneumonia or at least two episodes of bacterial or other pneumonia).10 Lung cancer incidence measured 66 per 100,000 person-years in people with HIV and 33 per 100,000 in the HIV-negative group. Poisson regression models linked HIV to higher lung cancer risk after adjustment for demographics (RR 1.9, 95% CI 1.5 to 2.4) and demographics plus cancer risk factors (RR 1.4, 95% CI 1.1 to 1.7) but not after additional adjustment for pneumonia (RR 1.1, 95% CI 0.9 to 1.5).

In a CD4-stratified analysis with 12-month time lagging, as in the VACS study,6 after adjustment for demographics a CD4 count below 200 cells/mm3 almost tripled lung cancer risk compared with HIV-negative controls (RR 2.7, 95% CI 1.5 to 4.8) and a CD4 count between 200 and 499 cells/mm3 almost doubled the risk (RR 1.7, 95% CI 1.2 to 2.5).10 But neither low CD4 bracket boosted lung cancer risk after further adjustment for cancer risk factors and pneumonia. As in the French study,8 however, people with a recent CD4 count above 500 cells/mm3 did not run a higher lung cancer risk in unadjusted or adjusted models.

The Kaiser team concludes that excess lung cancer risk in these people with HIV can be traced to differences in demographics, cancer risk factors like smoking, and higher pneumonia risk -- but not to low CD4 counts.10 A Swiss HIV Cohort Study (SHCS) analysis pinpointed only smoking as a lung-factor risk factor in a case-control comparison of 68 cohort members with lung cancer and 337 HIV-positive matched controls (Table 3).11 Current versus never smoking inflated odds of lung cancer more than 14-fold (odds ratio 14.4, 95% CI 3.36 to 62.1). But recent CD4 count, earlier CD4 count, use of combination antiretroviral therapy, and a history of AIDS did not predict lung cancer.

Table 3. Four Studies With Varying Findings on How CD4 Tallies Affect Lung Cancer Risk
 Low CD4 or CD4/CD8 Raises RiskLittle or No CD4 Impact on Risk
n21,666 HIV+; 277 lung ca+/21,389 lung ca-84,504 HIV+ (48,914 1997-2000; 56,846 2001-2004; 67,688 2005-2009)24,768 HIV+; 257,600 HIV-68/337 HIV+; 68 lung ca+, 337 lung ca-a
Years analyzed1998-20121997-20091996-20111985-2010
Age (y)50 lung ca+/45 lung ca-35.5 1997-2000; 38.3 2001-2004; 40.9 2005-20094035% 25-44, 38% 45-54, 27% 55+b
Male/female (%)99/1~70/3090/1080/20
Comparison groupHIV+ VACS members without lung caGeneral population of FranceHIV- Kaiser patientsHIV+ SHCS members without lung ca
Race46% white, 48% black, 4% Hispanic~13% AfricanHIV+/HIV-: 56%/44% white, 18%/10% black, 21%/27% Hispanic; 4%/16% AsianNR, total SHCS 78% white, 14% black, 8% otherc
Smoking73% current, 15% former22% current (unknown for 52%)45% HIV+, 31% HIVever smokers96% lung ca+/73% lung ca-d
Adjustment variablesAge, race, smoking, HCV, alcohol/drug use, COPD, occupational lung diseaseAge, sexAge, sex, race, calendar era, smoking, drug/alcohol use, overweight/obesity, recent CD4, pneumonia historyAge, sex, HIV route, smoking status, year, SHCS center
CD4 outcomeCD4, CD8, CD4/CD8 ratioAt least 500 CD4s after 2 y ARTRecent CD4Nadir, recent, or distant CD4

a Matched by age, sex, HIV transmission route, calendar period, and SHCS center to HIV-positive cohort members without lung cancer.
b Age at lung cancer diagnosis.
dSmoking status known for about 75% of participants.
ART, antiretroviral therapy; ca, cancer; COPD, chronic obstructive pulmonary disease; Kaiser, Kaiser Permanente healthcare system; NR, not reported; SHCS, Swiss HIV Cohort Study; VACS, Veterans Aging Cohort Study.

Why do the Kaiser10 and SHCS11 results appear to differ from the VACS6 and French8 analyses? Comparison of the four study populations, methods, and risk variables suggests the results actually differ less than they first seem to (Table 3).

Alone among the four studies, the SHCS found no hint of a CD4-count impact on lung cancer risk in any adjusted analysis.11 This study also differs most from the other three, with a case-control design involving (1) only 68 lung cancer patients with HIV and 337 HIV-positive controls with no cancer, (2) a largely white study population, (3) 96% smoking prevalence in the lung cancer group, and (4) an observation period stretching 10 years back into the pre-combination ART era (Table 3).

The other three studies all involve 21,000 or more people with HIV and hundreds of lung cancer cases (Table 3).6,8,10 In these three studies follow-up began with the dawn of combination ART and reached 2009, 2010, or 2011. All three studies found tenuous ties between low CD4 count and lung cancer risk. In the VACS6 and Kaiser10 studies, significant ties disappeared in fully adjusted models including pneumonia history. In the VACS study a fully adjusted model linked lower (worse) CD4/CD8 ratio to lung-cancer risk;6 none of the other studies evaluated this variable. In the French8 and Kaiser10 studies, people who reached a CD4 count above 500 cells/mm3 had a lung cancer risk equivalent to the general population, a finding suggesting that maintaining or regaining a robust CD4 tally helps shield people from lung cancer.

Faulty immune function has other possibly baneful effects related to lung cancer -- younger age at lung cancer diagnosis and shorter survival with lung cancer. NA-ACCORD researchers compared median age at cancer diagnosis in 88,018 North Americans treated for HIV and the general population represented by the Surveillance, Epidemiology and End Results (SEER) Program.12 An analysis weighted to yield groups equivalent in age, race, and calendar period linked a recent CD4 count below 200 cells/mm3 to a 4-year younger age at lung cancer diagnosis (51 versus 55 years) when compared with either a recent CD4 count of 200 to 499 cells/mm3 or a recent count of 500 cells/mm3 or more.

In a 1984-2011 study of 2549 women in the Women's Interagency HIV Study (WIHS) and 4274 men in the Multicenter AIDS Cohort Study (MACS), all participants had a history of smoking.13 In an analysis limited to 42 lung cancer patients with HIV, a Cox model adjusted for ART use, AIDS diagnosis, prior AIDS pneumonia, viral load, and CD4 count identified only one predictor of shorter survival -- a nadir CD4 count below 200 cells/mm3.

Lung Cancer Risk Posed By Previous Pulmonary Disease

Several studies tie prior episodes of pneumonia or other lung disease to heightened lung cancer risk -- and people with HIV run a higher risk of pneumonia.14 In the 21,666-veteran analysis described in the preceding section, 1 or more episodes of bacterial pneumonia nearly doubled the risk of incident lung cancer in a 12-month time-lagged analysis (Table 1).6 Lung cancer risk was even higher in analyses lagged 24 months, 36 months, or 48 months. And that association remained significant in the final adjusted model (HR 1.3, 95% CI 1.1 to 2.3 for 1 versus 0 episodes). A previous VACS analysis of 37,294 veterans with HIV and 75,750 without HIV figured that bacterial pneumonia upped the risk of lung cancer 50% (incidence rate ratio [IRR] 1.5, 95% CI 1.1 to 2.0), while COPD nearly doubled the risk (IRR 1.9, 95% CI 1.5 to 2.3).15

A National Cancer Institute (NCI) study involved 322,675 people with HIV in 11 US regions in 1997-2002.16 Linking AIDS patients to cancer registries allowed the NCI team to estimate risk of incident lung cancer in analyses adjusted for age, race, sex, HIV acquisition route, CD4 count, and AIDS diagnosis year. Recurrent pneumonia (2 or more episodes in 1 year) boosted lung cancer risk more than 60% in the overall study group (HR 1.63, 95% CI 1.08 to 2.46) for 10 years after AIDS began and doubled the risk in people younger than 50 (HR 1.99, 95% CI 1.26 to 3.16). These risks with recurrent pneumonia remained elevated, but nonsignificantly, after indirect adjustment for smoking. The analysis found no link between Pneumocystis pneumonia or tuberculosis and incident lung cancer.

The combined MACS and WIHS analysis discussed above involved 4274 men (1860 with HIV) and 2549 women (1875 with HIV), all of them current or prior smokers.13 In a multiple regression model prior Pneumocystis pneumonia or recurrent bacterial pneumonia more than tripled lung cancer risk in the combined MACS and WIHS groups for all study years (1984-2011) (IRR 3.56, 95% CI 1.67 to 7.61) and for the combination ART era (1995-2011) (IRR 3.51, 95% CI 1.61 to 7.67). These associations held true when the researchers lagged the pneumonia diagnosis up to 5 years.

In the already-noted Kaiser study involving 24,768 people with HIV and 257,600 HIV-negative Kaiser patients, HIV infection and lower CD4 counts boosted lung-cancer risk in some multivariable models, but not after adjustment for pneumonia (Pneumocystis pneumonia or at least two episodes of bacterial or other pneumonia).10 This statistical impact of pneumonia underlines its strength as a lung-cancer predictor.

Researchers who conducted the VACS analysis propose that pneumonia-induced inflammatory injury could heighten lung cancer risk and "dysfunctional immune activation could lead to more deleterious inflammatory responses to bacterial pneumonia in patients with HIV than in those without HIV."6 Putting all the risk pieces together, they suggest clinicians can identify HIV-positive people with a high lung cancer risk by looking for the 5 risk factors listed in Table 4.

Table 4. Five Key Factors to Identify HIV-Positive People With High Lung-Cancer Risk
  • Smoking
  • Older age
  • Low CD4/CD8 ratio
  • Bacterial pneumonia history
  • COPD history

Source: Sigel K, et al.6


  1. Altekruse SF, Shiels MS, Modur SP, et al. Cancer burden attributable to cigarette smoking among HIV-infected people in North America. AIDS. 2018;32:513-521.
  2. World Health Organization. Health statistics and information systems. Metrics: Population-attributable fraction (PAF).
  3. American Lung Association. Benefits of quitting.
  4. Reddy KP, Kong CY, Hyle EP, et al. Lung cancer mortality associated with smoking and smoking cessation among people living with HIV in the United States. JAMA Intern Med. 2017;177:1613-1621.
  5. Frisch M, Biggar RJ, Engels EA, Goedert JJ; AIDS-Cancer Match Registry Study Group. Association of cancer with AIDS-related immunosuppression in adults. JAMA. 2001;285:1736-1745.
  6. Sigel K, Wisnivesky J, Crothers K, et al. Immunological and infectious risk factors for lung cancer in US veterans with HIV: a longitudinal cohort study. Lancet HIV. 2017;4:e67-e73.
  7. Guiguet M, Boué F, Cadranel J, et al. Effect of immunodeficiency, HIV viral load, and antiretroviral therapy on the risk of individual malignancies (FHDH-ANRS CO4): a prospective cohort study. Lancet Oncol. 2009;10:1152-1159.
  8. Hleyhel M, Hleyhel M, Bouvier AM, et al. Risk of non-AIDS-defining cancers among HIV-1-infected individuals in France between 1997 and 2009: results from a French cohort. AIDS. 2014;28:2109-2118.
  9. Sigel K. The changing future of HIV: non-AIDS cancers and lung cancer. Slide presentation. Personal communication. 2018.
  10. Marcus JL, Leyden WA, Chao CR, et al. Immunodeficiency, AIDS-related pneumonia, and risk of lung cancer among HIV-infected individuals. AIDS. 2017;31:989-993.
  11. Clifford GM, Lise M, Franceschi S, et al. Lung cancer in the Swiss HIV Cohort Study: role of smoking, immunodeficiency and pulmonary infection. Br J Cancer. 2012;106:447-452.
  12. Shiels MS, Althoff KN, Pfeiffer RM, et al. HIV infection, immunosuppression, and age at diagnosis of non-AIDSdefining cancers. Clin Infect Dis. 2017;64:468-475.
  13. Hessol NA, Martínez-Maza O, Levine AM, et al. Lung cancer incidence and survival among HIV-infected and uninfected women and men. AIDS. 2015;29:1183-1193.
  14. Benard A, Mercie P, Alioum A, et al. Bacterial pneumonia among HIV-infected patients: decreased risk after tobacco smoking cessation. ANRS CO3 Aquitaine Cohort, 2000-2007. PloS One. 2010;5:e8896.
  15. Sigel K, Wisnivesky J, Gordon K, et al. HIV as an independent risk factor for incident lung cancer. AIDS. 2012;26:1017-1025.
  16. Shebl FM, Engels EA, Goedert JJ, Chaturvedi AK. Pulmonary infections and risk of lung cancer among persons with AIDS. J Acquir Immune Defic Syndr. 2010;55:375-379.

[Note from TheBodyPRO: This article was originally published by The Center for AIDS Information & Advocacy in June 2018. We have cross-posted it with their permission.]

Related Stories

Lung Cancer Risk and Screening With HIV -- and an Intriguing New Treatment Opportunity

This article was provided by The Center for AIDS Information & Advocacy. It is a part of the publication Research Initiative/Treatment Action!. Visit CFA's website to find out more about their activities and publications.

No comments have been made.

Add Your Comment:
(Please note: Your name and comment will be public, and may even show up in
Internet search results. Be careful when providing personal information! Before
adding your comment, please read's Comment Policy.)

Your Name:

Your Location:

(ex: San Francisco, CA)

Your Comment:

Characters remaining:


The content on this page is free of advertiser influence and was produced by our editorial team. See our content and advertising policies.