Increase of non-AIDS defining cancers in HOPS cohort

Simon Collins, HIV i-Base

An oral presentation of an analysis from Patel and colleagues looked at the age-, race-, smoking-, and gender-adjusted relative rates of five cancers that are not traditionally associated with AIDS (lung, head/neck, Hodgkin’s disease [HD], anorectal [ARC], melanoma) in 7,900 patients treated at two large Chicago HIV clinics with those observed in the 20 million County and 92 million State cancer registry patients. A second group of around 4,050 HIV-patients from the HOPS (HIV Out-Patient Study) cohort was compared with 334 million patients from the general population. The study period covered 1992 to 2002. [1]

The incidence of the five non-AIDS malignancies was much higher in the HOPS population than in the general population.

Chicago Chicago HOPS HOPS
Adj. RR 95% CI Adj. R 95% CI
Lung 3.63 2.18 – 6.05 2.13 1.06 – 4.27
HD 77.43 19.37 – 309.55 4.58 7.48 – 13.72
Anorectal 5.03 4.76 – 5.33 10.13 7.48 – 13.72
Melanoma 4.10 9.39 – 152.70 2.99 1.71 – 5.22
Head/neck 9.96 2.49 – 39.79

In a discussion session, Joel Palefsky provided an overview of incidence and treatment of HPV (human papilloma virus)-associated anal cancer, in reference to two other studies presented at the conference. [2]The study noted that while the incidence of KS and cervical cancer have reduced in the HAART era, these five malignancies appear to be increasing. The incidence of other common cancers such as breast, colon and prostate cancer were not significantly increased in either the HIV-positive or general population groups.

Palefsky also highlighted previous research from his group that showed that the risk for HIV-positive gay men of developing HPV-associated anal cancer is around 35-fold higher than for men in the general population and twice as high as HIV-negative gay men. Although the absolute risks for anal cancer are reported as 35/100,000 incidence in HIV-positive men he noted that this was comparable to rates of cervical cancer prior to the introduction of effective screening and treatment programmes. These programmes have successfully been able to reduce rates of cervical cancer down to 8/100,000. Early detection and treatment of anal cancer significantly improves prognosis.

Patrick Sullivan and colleagues from the US Centres for Disease Control and Prevention (CDC) in Atlanta looked at incidence and risk factors for anorectal cancer from 1990-2002 in a cohort of more than 58,000 HIV-positive patients and compared this to the NCI general population dataset. [3]

During more than 231,000 patient years of follow-up, 150 patients were documented as having anorectal cancer. Among HIV-positive cases, 92% were male, 44% were aged 35 to 44 years, and 79% were exposed to HIV through male-male sex. The age-standardised anorectal cancer rate was 84.9 cases per 100,000 person-years (95% CI:  52.4 to 117.5); the age- and sex-standardised rate was 66.1 per 100,000 person-years (CI:  38.0 to 94.2). The general population rate was 21.8 per 100,000 person-years (CI:  21.6 to 22.0).

adjusted odds ratio 95% CI
HIV exposure MSM vs MSW 5.6 1.8 to 17.3
Age Ž 45 yrs vs <35 yrs 2.5 1.6 to 3.9
Age 35 – 44 yrs vs <35 yrs 1.7 1.6 to 3.9
Clinical AIDS diagnosis 1.5 1.1 to 2.2
CD4 count <200 vs >500 4.0 1.9 to 8.4
CD4 counts 200-499 vs >500 2.4 1.1 to 5.1

The study concluded that these data suggest that clinicians must be vigilant for ARC during periodic examinations of HIV-infected persons, especially men with a history of male-male sex, older persons, and those at more advanced stages of disease.

Diamond and colleagues from the University of California performed a match between the AIDS and cancer registries for San Diego County from 1988-2000.

They identified 39 cases of anal squamous cell carcinoma. All were men and 38 (97%) were men who have sex with men (MSM). The median age was 42 years (range 25 to 59); 28 (72%) were Caucasian, two (5%) were Black, seven (18%) were Latino, and two (5%) were of unknown race/ethnicity.

The median CD4 count was 120 cells/mm3 (range 2 to 551). Among the 36 patients diagnosed with HIV prior to or simultaneously with their anal cancer diagnoses, the median duration of known HIV infection was 78 months (range 0 to 175). No cases of anal cancer were diagnosed before 1992.The median duration of HIV infection in the pre-HAART era (1992 to 1995) was 29 months while post-HAART (1996 to 2000) it was 84 months (p = 0.01).

Eight cases (21%) were diagnosed pre-HAART, while 31 (79%) were diagnosed post-HAART. The number of cases increased from 2.8/1000 AIDS cases in 1992 to 24.7/1000 in 2000 despite a declining incidence of AIDS (r = 0.83, p = 0.005). However, the study reported 3/8 (38%) and 8/31 (26%) in pre- and post-HAART period were in-situ which many clinicians would not count as a registered cancer and would remove rather than treat.

Twenty-eight patients received surgical treatment, one received radiation therapy, one received chemotherapy, 16 (41%) received both, and 21 received neither. At most recent follow-up, 20 (51%) were alive. Among the 19 deceased, six died of HIV/AIDS, six died of anal cancer, and seven died of other/unknown causes.

The study concluded that increased incidence could be related to increased screening for anal cancer or increased longevity with the use of HAART but that the fewer in-situ tumors in the post-HAART era argue against a screening phenomenon. The longer duration of HIV infection post-HAART suggests that HAART increases the time at risk for the development of anal cancer.


Screening availability is general by specialist referral in HIV clinics in the UK. However, cost effectiveness should be higher than for cervical screening as the population at risk is easily defined.

Although the natural history can lead to regression, this is only in younger men and rarely in men over 30 years old.


  1. Patel P, Novak RM, Tong T et al. Incidence of non-AIDS-defining malignancies in the HIV Out-Patient Study. 11th CROI 2004, Oral abstract 81.
  2. Palefsky J. Discussant: HPV infection and anal cancer. 11th CROI 2004, Session 107.
  3. Sullivan PS, Juhasz M, Brooks et al. Incidence and associations of anorectal cancers in persons with HIV infection. 11th CROI 2004, Abstract 777.
  4. Diamond C, Taylor TH, Culver HA. Dramatic increase in anal cancer diagnoses in the era of Highly Active Antiretroviral Therapy. 11th CROI 2004, Abstract 778.

Links to other websites are current at date of posting but not maintained.