Risk of Virus-Related Cancers and Lung Cancer Higher With HIV in U.S.
Overall cancer risk was 69% higher with HIV than in the general U.S. population in 1996-2012, according to an analysis involving 448,258 HIV-positive people. HIV conferred a higher risk of AIDS-defining cancers, most non-AIDS virus-related cancers and a few non-AIDS virus-unrelated cancers, including lung cancer.
Higher cancer risk with HIV reflects impaired immune control of cancer-causing viral infections, as well as lifestyle factors such as smoking and alcohol use. National Cancer Institute (NCI) researchers who conducted the new study note that wider and earlier use of antiretroviral therapy (ART) might lower the risk of some cancers but may not completely reverse the impact of early immunosuppression and ongoing immune dysfunction and inflammation. Because few recent population-based studies address relative cancer risk with HIV, these investigators analyzed HIV-related cancer incidence over time by matching population-based HIV and cancer registries.
The study examined data from the HIV/AIDS Cancer Match (HACM) Study, which links data on U.S. people with HIV to cancer registries. This analysis focused on HIV-positive people in HACM during the period 1996-2012 in nine jurisdictions: Colorado, Connecticut, Georgia, Maryland, Michigan, New Jersey, New York, Puerto Rico and Texas. The researchers excluded the first three months of follow-up in each participant to avoid analyzing prevalent cancers that may have prompted HIV testing and reporting. Then, they determined how many people with HIV had incident cancer by linking them to corresponding cancer registries.
The NCI team divided cancers into AIDS-defining cancers (Kaposi sarcoma, non-Hodgkin lymphoma, cervical cancer), non-AIDS virus-related cancers and non-AIDS virus-unrelated cancers. To compare cancer incidence in people with HIV and the U.S. general population, the researchers calculated standardized incidence ratio (SIR), which equals observed incident cancers in people with HIV divided by the expected number in that group. They estimated expected numbers of incident cancers by applying general population cancer incidence to person-time in the HIV group based on sex, age, race/ethnicity, calendar year and registry. The researchers used Poisson regression to test SIR differences by HIV status over time.
The 448,258 study participants were predominantly male (71%), and 30% were 50 years old or older. Almost half of the population was black (47%) and about one-quarter Hispanic (28%) and one-quarter white (25%). Most people, 85%, had fewer than 10 years of follow-up between HIV diagnosis and cancer onset, though that proportion fell over time from more than 99% in 1996-1999 to 76% in 2009-2012.
The NCI investigators counted 21,294 new cancers in people with HIV to yield an overall SIR of 1.69, meaning HIV-positive people had a 69% higher cancer risk than the general U.S. population (P < .0001 for this and all following SIRs). HIV infection conferred a 14 times higher risk of AIDS-defining cancers (SIR 13.97) and a 5.39-fold higher risk of non-AIDS virus-related cancers, including anal cancer (SIR 19.06), vulvar cancer (SIR 9.35), Hodgkin lymphoma (SIR 7.70), penile cancer (SIR 5.33), vaginal cancer (SIR 3.55), liver cancer (SIR 3.21) and human papillomavirus-related oral cancer (SIR 1.64).
Considered together, non-AIDS virus-unrelated cancers proved less frequent with than without HIV (SIR 0.92), with two exceptions: Cancer of the lung (SIR 1.97) and larynx (SIR 2.11) were both about twice as likely in people with HIV than in the general U.S. population. In contrast, breast cancer (SIR 0.63) and prostate cancer (SIR 0.48) were both significantly less frequent with HIV.
The most frequent individual cancers diagnosed were non-Hodgkin lymphoma, lung cancer, Kaposi sarcoma, anal cancer, prostate cancer, liver cancer and Hodgkin lymphoma.
Compared with HIV-positive people who never had AIDS, those with AIDS had a significantly higher incidence of all cancers (adjusted SIR ratio [aSIRr] 1.83), grouped AIDS cancers (aSIRr 3.15), grouped non-AIDS cancers (aSIRr 1.45), grouped non-AIDS virus-related cancers (aSIRr 2.21) and grouped non-AIDS virus-unrelated cancers (aSIRr 1.25) (P < .0001 for all comparisons).
Over the 1996-2012 study period, the SIR comparing cancer incidence in people with HIV versus the general population did not increase for any cancer. SIRs dropped significantly for grouped AIDS-defining cancers; Kaposi sarcoma; each type of AIDS-defining non-Hodgkin lymphoma; grouped non-AIDS virus-related cancers; cancers of the anus, liver and lung; and non-AIDS-defining non-Hodgkin lymphoma. Nevertheless, SIRs remained significantly higher with HIV in the latest study period (2009-2012).
The authors suggest that declining SIRs over time -- indicating a narrowing difference in cancer incidence in HIV-positive people relative to the general population -- "presumably" partly reflect wider access to stronger antiretroviral regimens and earlier ART use over time. But, they add that "the elevated risk for many cancers, especially after AIDS onset, highlights the continuing contribution of immunosuppression to cancer risk in this population."
The NCI team stresses that lung cancer ranked second in incidence -- behind an AIDS cancer, non-Hodgkin lymphoma -- in people with HIV. High smoking prevalence in HIV populations only partly explains high lung cancer incidence, they argue, because higher lung cancer incidence in people with AIDS (versus HIV alone) and falling SIRs over time as ART access improved "both support a contribution from immunosuppression."