HIV and Malignancies S. De Wit St Pierre Hospital Brussels

HIV and cancer AIDS-defining malignancies: Kaposi’s sarcoma Non Hodgkin lymphoma 1985 Cervical cancer 1993 Non AIDS-defining malignancies (NADM) is increasing Linked with virus HPV (Anal), HBV and HCV (Liver), EBV (HL) Linked with previous immunodeficiency HHV8 EBV HPV

Background Before introduction of HAART, ADCs common, including Kaposi’s sarcoma, NHL, and invasive cervical carcinoma Rate of ADCs significantly increased from early to late pre-HAART era and then significantly decreased following introduction of HAART Rates of nADCs stable during pre-HAART eras and then significantly increased following introduction of HAART Crum-Cianflone N, et al. AIDS. 2009;23:41-50.

SIR = Standardised Incidence Ratio Nb cases of cancer in the HIV population Expected nb of cases in the general population, calculated with local cancer registry incidence =

Cancer Incidence in AIDS Patients Cancer type No. cases SIR 95% CI AIDS-defining cancers Kaposi sarcoma 3136 5321 5137 - 5511 Non-Hodgkin lymphoma 3345 32 31 - 33 Cervical cancer 101 5.6 5.5 - 6.8 Non-AIDS-defining cancers Anal cancer 219 27 24 - 31 Liver cancer 86 3.7 3.0 - 4.6 Lung cancer 531 3.0 2.8 - 3.3 Hodgkin lymphoma 184 9.1 7.7 - 11 All non-AIDS related cancers 2155 1.7 1.5 - 1.8 Study of cancer risk in AIDS patients from 1980-2006 (N=372,364) Predominantly male (79%), non-hispanic black (42%), MSM (42%) Median age of 36 years at the onset of AIDS Aims: To address the question what impact the decline on AIDS-related mortality had on cancer rates. Overall 3 questions were asked: 1) Cancer risk during year 3-5 after AIDS onset relative to the general population. 2) 5. year cumulative incidence of AIDSdefining cancer (ADC) and non-AIDS-defining cancer (NADC). 3) Fraction of deaths among people with AIDS attributable to cancer Methods: Records of n=372.364 people with AIDS (1980-2006) were linked to corresponding cancer registry records. The observation for outcomes (cancer or death) began at month 4 after AIDS among those who were alive and cancer-free. Conclusions: Elevated risks for ADCs and some NADCs relative to the general population. Particularly cancers of the anus, liver (but cave no adjustment for smoking which is found more frequent in HIV-serpositive cohorts), lung and Hodgkin lyphoma are found more often in the HIV-positive population. SIR=Standardized Incidence Ratios Simard E, et al. 17th CROI; San Francisco, CA; February 16-19, 2010. Abst. 27.

Cancer Mortality in AIDS Patients Population attributable risk among people with AIDS in the US Cumulative Incidence (%) Cumulative incidence of ADCs has declined, consistent with improvements in immune status. Cumulative incidence of some NADCs increasing in the HAART era particularly cancers of the anus, liver, lung and hodgkins lymphomaancers of the anus, liver, lung and Hodgkins lymphoma. Simard E, et al. 17th CROI; San Francisco, CA; February 16-19, 2010. Abst. 27.

Increased rates of nADCs. Why ? Increasing survival of patients with HIV might be associated with an increase of traditional cancer Aging of the HIV population Life style Long-term toxicity of ART ?

HIV associated cancers Possible explanations: Confounding by shared lifestyle cancer risk factors A direct effect of HIV, likely through an effect of immune deficiency Importance: If immune deficiency is responsible, then reversing immune deficiency might decrease cancer risk

Pathogenesis of NADC Some are virally-induced cancers, but not all HIV-tat may transactivate cellular genes or proto-oncogenes, inhibit tumor suppressor genes Microsatellite alterations (MA) due to genetic instability in HIV (e.g 6 fold higher number of MA in HIV lung CA over non-HIV)1 Increase susceptibility to effects of carcinogens (tobacco) Population differences based on genetics and exposure to carcinogens Decreased immune surveillance 1Wistuba, AIDS 1999;13:415-26

HIV & Cancers Role of immune deficiency ? Cancer rate should also be increased in other immunosuppressive disorders

Infection-related cancers Grulich et al. Lancet, 2007, 370, 59–

Grulich et al. Lancet, 2007, 370, 59–

Common epithelial cancers Grulich et al. Lancet, 2007, 370, 59–

Cancers in HIV and transplant patients The range of cancers occurring at increased rates is strikingly similar in the two groups Mostly those known or suspected to be caused by infective agents Impact of immunodeficiency on these cancers

CD4 and risk of liver cancer Clifford and Franceschi, Future Oncology 2009

Current CD4 count and death from cancer D:A:D study group AIDS 2008, 22:2143–

Characteristics of cancer immune control CD4 cell count CTL function NK Immune memory Central/effector memory Level of immune activation: PD-1, IL-10, Treg Immune system on pre-cancerous lesions

Cancer Incidence in AIDS Patients Cancer type No. cases SIR 95% CI AIDS-defining cancers Kaposi sarcoma 3136 5321 5137 - 5511 Non-Hodgkin lymphoma 3345 32 31 - 33 Cervical cancer 101 5.6 5.5 - 6.8 Non-AIDS-defining cancers Anal cancer 219 27 24 - 31 Liver cancer 86 3.7 3.0 - 4.6 Lung cancer 531 3.0 2.8 - 3.3 Hodgkin lymphoma 184 9.1 7.7 - 11 All non-AIDS related cancers 2155 1.7 1.5 - 1.8 Aims: To address the question what impact the decline on AIDS-related mortality had on cancer rates. Overall 3 questions were asked: 1) Cancer risk during year 3-5 after AIDS onset relative to the general population. 2) 5. year cumulative incidence of AIDSdefining cancer (ADC) and non-AIDS-defining cancer (NADC). 3) Fraction of deaths among people with AIDS attributable to cancer Methods: Records of n=372.364 people with AIDS (1980-2006) were linked to corresponding cancer registry records. The observation for outcomes (cancer or death) began at month 4 after AIDS among those who were alive and cancer-free. Conclusions: Elevated risks for ADCs and some NADCs relative to the general population. Particularly cancers of the anus, liver (but cave no adjustment for smoking which is found more frequent in HIV-serpositive cohorts), lung and Hodgkin lyphoma are found more often in the HIV-positive population. SIR=Standardized Incidence Ratios Simard E, et al. 17th CROI; San Francisco, CA; February 16-19, 2010. Abst. 27.

Anogenital Cancers Invasive cervical carcinoma Anal cancer1 Considered an AIDS-defining condition Anal cancer1 Not AIDS defining but very common HPV involvement1-2 Both derive from premalignant dysplastic lesions due to HPV Most oncogenic strains: 16, 18, 31, 33, 35, 45 Repeated infections and infection with multiple HPV strains increase the risk of developing neoplasia 1Phelps RM, et al. Int J Cancer. 2001;94:753-757. 2Martin F, et al. Sex Transm Infect. 2001;77:327-331.

HPV-induced cancer Cervix Vulva Vagina Anal Oro-pharyngal Penis 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 69, 82

The Natural History of HPV Infection and Cervical Cancer HIV- HIV+ Persistent HPV 5-10% 20-40% Cervical cancer x 3-11 Vulva & vagina cancer x 4-10 The Natural History of HPV Infection and Cervical Cancer. The peak prevalence of transient infections with carcinogenic types of HPV (blue line) occurs among women during their teens and 20s, after the initiation of sexual activity. The peak prevalence of cervical precancerous conditions occurs approximately 10 years later (green line) and the peak prevalence of invasive cancers at 40 to 50 years of age (red line). (The peaks of the curves are not drawn to scale.) The conventional model of cervical-cancer prevention is based on repeated rounds of cytologic examination, including Papanicolaou smears, and colposcopy (small blue arrows). Alternative strategies include HPV vaccination of adolescents (large beige arrow), one or two rounds of HPV screening at the peak ages of treatable precancerous conditions and early cancer (large reddish-brown arrows), or both. Schiffman, M. et al. N Engl J Med 2005;353:2101-2104

Infection with oncogenic HPV in HIV women Prevalence is higher :20-40% (vs.5-10%) Multiple genotypes: 40 % (vs. 12% ) New infection? Reactivation of latent infection Linked with younger age, lower CD4 and higher HIV VL Paleksky. J Natl Cancer Inst 1999 Strickler. Journal of the National Cancer 2005

Saint-Pierre Cohort N=592 Prevalence of HR-HPV infection according to both age and CD4 cell strata (count/µL). (p=0.03, logistic regression) D Konopnicki, Y Manigart, C Gilles, de Marchin J*, M Delforge, F Feoli*, P Barlow, S De Wit and N Clumeck. ECCMID 2011

Cancer screening – EACS Problem Patients Procedure Evidence of benefits Screening interval Additional Comments Breast cancer Women 50–70 yrs Mammography ↓breast cancer mortality 1–3 years Cervical cancer Sexually active women Papanicolau test, HPV DNA test ↓cervical cancer mortality Target age group should include at least the age range 30 to 59 years. Longer screening interval if prior screening tests repeatedly negative Colorectal cancer Persons 50–75 yrs Fecal Occult Blood test ↓colorectal cancer mortality Benefit is marginal EACS guidelines 2011. Available at http://www.europeanaidsclinicalsociety.org/guidelinespdf/EACS-EuroGuidelines_FullVersion.pdf . Accessed March 2011.

Screening in developing countries Screen-and-treat approach Randomised , n=6555 with 956 HIV-positive women in South Africa, 35-65 years first screen. Excluded macroscopic lesions (6%) 3 arm HPV test and cryotherapy Visual inspection+ acetowhite detection and cryotherapy Control : delayed at 6 months Women had colposcopy and biopsy at Month 6 (all), 12, 24 and 36 (subset) HIV pos HIV neg ≥CIN2 at M36 15% 5% p=.0006 Screen HPV RR M36 0.2 (0.06-0.07) 0.3 (0.02-.005) ps for both Screen VIA 0.51 (0.29-0.89) p=ns Kuhn and al. AIDS 2010

Anal Cancer Invasive cancer SIR 6-8 (in USA, St-Pierre Cohort) Piketty AIDS 2008 132 cases of invasive anal cancer among 86322 HIV-patients Median survival 5 years Recent Pre Early Median CD4 188 227 288 Death due to AC 50% 40% 68.8%

Anal Cancer Incidence Incidence and risk of invasive anal cancer Higher in HIV-infected vs age- and gender-matched general population (P < .001) 60/100,000 PYs (95% CI, 40-89) vs 0.52/100,000 PYs (95% CI, 0.27-0.78) Nonsignificant difference in pre-HAART and post-HAART era for HIV-positive individuals (P > .05) 35/100,000 PYs (95% CI, 15-72) vs 92/100,000 PYs (95% CI, 52-149) Higher relative risk of anal cancer vs general population in post-HAART era Pre-HAART era, 67 Post-HAART era, 176 http://clinicaloptions.com/hiv Bower M, et al. J Acquir Immune Defic Syndr. 2004;37:1563-1565.

Anal Cytology Screening for AIN in HIV-positives Screening Pap Normal ASCUS LSIL HSIL Repeat in 12 months Anoscopy with biopsy No lesion seen LSIL HSIL Treat or follow Treat Chin-Hong PV et al. J Infect Dis. 2004;90:2070-2076.

In summary HPV-induced cancers are not reduced after cART introduction Screening should be improved for cervical cancer and for anal cancer Preventive vaccination against HPV should be more extensively studied and applied in HIV patients

Cancer Incidence in AIDS Patients Cancer type No. cases SIR 95% CI AIDS-defining cancers Kaposi sarcoma 3136 5321 5137 - 5511 Non-Hodgkin lymphoma 3345 32 31 - 33 Cervical cancer 101 5.6 5.5 - 6.8 Non-AIDS-defining cancers Anal cancer 219 27 24 - 31 Liver cancer 86 3.7 3.0 - 4.6 Lung cancer 531 3.0 2.8 - 3.3 Hodgkin lymphoma 184 9.1 7.7 - 11 All non-AIDS related cancers 2155 1.7 1.5 - 1.8 Aims: To address the question what impact the decline on AIDS-related mortality had on cancer rates. Overall 3 questions were asked: 1) Cancer risk during year 3-5 after AIDS onset relative to the general population. 2) 5. year cumulative incidence of AIDSdefining cancer (ADC) and non-AIDS-defining cancer (NADC). 3) Fraction of deaths among people with AIDS attributable to cancer Methods: Records of n=372.364 people with AIDS (1980-2006) were linked to corresponding cancer registry records. The observation for outcomes (cancer or death) began at month 4 after AIDS among those who were alive and cancer-free. Conclusions: Elevated risks for ADCs and some NADCs relative to the general population. Particularly cancers of the anus, liver (but cave no adjustment for smoking which is found more frequent in HIV-serpositive cohorts), lung and Hodgkin lyphoma are found more often in the HIV-positive population. SIR=Standardized Incidence Ratios Simard E, et al. 17th CROI; San Francisco, CA; February 16-19, 2010. Abst. 27.

Hodgkin’s Disease Hodgkin’s disease: RR: 5 to 30 Association with HIV-infection Hodgkin’s disease: RR: 5 to 30 Non-Hodgkin’s disease: RR: 24 to 165 Patients with HIV present with: B symptoms (70% to 96%), worse histology, higher-stage tumor (74% to 92% are III or IV), bone marrow involvement (40% to 50%), pancytopenia Good response to MOPP/ABV Complete response: 74.5% 2-year disease-free survival: 62% Early better results with Stanford V and BEACOPP Gerard L, et al. AIDS. 2003;17:81-87.

Risk of Hodgkin lymphoma by CD4 count Clifford and Franceschi, 2009

Cancer Incidence in AIDS Patients Cancer type No. cases SIR 95% CI AIDS-defining cancers Kaposi sarcoma 3136 5321 5137 - 5511 Non-Hodgkin lymphoma 3345 32 31 - 33 Cervical cancer 101 5.6 5.5 - 6.8 Non-AIDS-defining cancers Anal cancer 219 27 24 - 31 Liver cancer 86 3.7 3.0 - 4.6 Lung cancer 531 3.0 2.8 - 3.3 Hodgkin lymphoma 184 9.1 7.7 - 11 All non-AIDS related cancers 2155 1.7 1.5 - 1.8 Aims: To address the question what impact the decline on AIDS-related mortality had on cancer rates. Overall 3 questions were asked: 1) Cancer risk during year 3-5 after AIDS onset relative to the general population. 2) 5. year cumulative incidence of AIDSdefining cancer (ADC) and non-AIDS-defining cancer (NADC). 3) Fraction of deaths among people with AIDS attributable to cancer Methods: Records of n=372.364 people with AIDS (1980-2006) were linked to corresponding cancer registry records. The observation for outcomes (cancer or death) began at month 4 after AIDS among those who were alive and cancer-free. Conclusions: Elevated risks for ADCs and some NADCs relative to the general population. Particularly cancers of the anus, liver (but cave no adjustment for smoking which is found more frequent in HIV-serpositive cohorts), lung and Hodgkin lyphoma are found more often in the HIV-positive population. SIR=Standardized Incidence Ratios Simard E, et al. 17th CROI; San Francisco, CA; February 16-19, 2010. Abst. 27.

LUNG CANCER

Excess of risk of lung cancer in HIV Pre-HAART epidemiological studies Author n HIV Study Pre-HAART SIR* Post-HAART SIR* 8.93 yes 1 R 8640 Bower 2 P 77,025 Herida no 3.8 31,616 Grulich 6.5 26,181 Parker 4 302,834 Frish 2.4 60,421 Dal Maso Reviewed in Lavolé, Lung Cancer 2005. *SIR is defined by the number of LC observed in the HIV-population/number of LC expected in the general population matched for age

Excess of risk of lung cancer in HIV Bias due to difference of smoking habits in HIV ? % of smokers 57 risk factors for cardiovascular disease age 35 to 44 years old HIV patients, n=274 (APROCO cohort) non HIV-persons, n=1038 (WHO-MONICA project) HIV Non HIV 33 Savès, CID 2003

Excess of risk of lung cancer in HIV Bias due to difference of smoking habits in HIV expected number of LC in the general population if 100 % of the persons were smokers 40 SIR = 6.5 40 SIR = 2.5 30 30 LC observed in HIV Number of LC Number of LC 20 20 LC expected in HIV 10 10 unknown % of smokers 100 % of smokers Parker, Chest 1998

Excess of risk of lung cancer in HIV Hypothesies for causal factors… increased frequency of smoking in HIV population, but intensity and duration not different HIV status seems probable, but the mechanisms remain unknown : degree of immune deficiency duration of immune deficiency oncogenic role of HIV per se other oncogenic virus role of HAART Cadranel, Respiration 1999; Bower, AIDS 2004

Excess of risk, which mechanisms Smoking + HIV + ID + HAART… Normal Dysplasia Hyperplasia Metaplasia Carcinoma 3p LOH, microsatellite alterations 9p21 LOH telomerase upregulation, MYC over expression 8p21-23 LOH neoangiogenesis, loss of FHIT, P53 mutations, aneuploidy, methylation 5q21 APC-MCC LOH, K-ras 12 mutation Increase of genomic instability ? Wistuba, JAMA 1997

Lung Cancer Most frequent NADC in HAART era Incidence 2-4 fold higher than general population SIRS between 2 and 3 and stable over time Diagnosed at younger age with advanced disease and primarily in smokers Adenocarcinoma is most frequent sub-type No clear screening strategy No argument to treat differently than non-HIV infected patients Confounded by role of tobacco consumption as higher in HIV infected individuals

Cancer Incidence in AIDS Patients Cancer type No. cases SIR 95% CI AIDS-defining cancers Kaposi sarcoma 3136 5321 5137 - 5511 Non-Hodgkin lymphoma 3345 32 31 - 33 Cervical cancer 101 5.6 5.5 - 6.8 Non-AIDS-defining cancers Anal cancer 219 27 24 - 31 Liver cancer 86 3.7 3.0 - 4.6 Lung cancer 531 3.0 2.8 - 3.3 Hodgkin lymphoma 184 9.1 7.7 - 11 All non-AIDS related cancers 2155 1.7 1.5 - 1.8 Aims: To address the question what impact the decline on AIDS-related mortality had on cancer rates. Overall 3 questions were asked: 1) Cancer risk during year 3-5 after AIDS onset relative to the general population. 2) 5. year cumulative incidence of AIDSdefining cancer (ADC) and non-AIDS-defining cancer (NADC). 3) Fraction of deaths among people with AIDS attributable to cancer Methods: Records of n=372.364 people with AIDS (1980-2006) were linked to corresponding cancer registry records. The observation for outcomes (cancer or death) began at month 4 after AIDS among those who were alive and cancer-free. Conclusions: Elevated risks for ADCs and some NADCs relative to the general population. Particularly cancers of the anus, liver (but cave no adjustment for smoking which is found more frequent in HIV-serpositive cohorts), lung and Hodgkin lyphoma are found more often in the HIV-positive population. SIR=Standardized Incidence Ratios Simard E, et al. 17th CROI; San Francisco, CA; February 16-19, 2010. Abst. 27.

Hepatocellular Carcinoma Linked to coinfection by hepatitis B and C viruses No clear difference between HAART users and non-users Estimated to be 7 times more frequent than in general population Optimal treatment similar to general population

Breast cancer No higher incidence in HIV-positive women There might even be a lower incidence: Significant decrease was recorded in Tanzania following HIV epidemics. Amir. J Natl Med Assoc 2000 Significant decrease in relative risk (observed cases/expected cases based incidence in general population ). Frisch. JAMA 2001

Why breast cancer could be less frequent in HIV women? Reduced incidence is also found in other immunosuppressed patients Steward. Lancet 1995 Suggesting that physiological immune response is a facilitating factor in breast carcinogenesis

Why breast cancer could be less frequent in HIV women? Hormone production is reduced in HIV: oestradiol or testosterone Body composition change with HAART (waist gain)…and the USA obesity epidemics

Why breast cancer could be less frequent in HIV women? CXCR4-tropic HIV is protective against breast cancer because In vitro: this receptor is highly expressed by tumor cells and CXCR4 HIV induces tumor cells apoptosis Endo M. Curr HIV Res 2008 In vivo : decreased incidence of breast cancer when compared to CCR5 HIV-infected patients Hessol N . PloS ONE Dec 2010. vol 5;12:e14349. Ritonavir has been studied in preclinical trials for its activity against breast cancer growth

Cancer screening – EACS Problem Patients Procedure Evidence of benefits Screening interval Additional Comments Breast cancer Women 50–70 yrs Mammography ↓breast cancer mortality 1–3 years Cervical cancer Sexually active women Papanicolau test, HPV DNA test ↓cervical cancer mortality Target age group should include at least the age range 30 to 59 years. Longer screening interval if prior screening tests repeatedly negative Colorectal cancer Persons 50–75 yrs Fecal Occult Blood test ↓colorectal cancer mortality Benefit is marginal EACS guidelines 2011. Available at http://www.europeanaidsclinicalsociety.org/guidelinespdf/EACS-EuroGuidelines_FullVersion.pdf . Accessed March 2011.

Other Malignancies Non-melanomatous skin cancer Conjunctival cancer Sarcoma Melanoma Germ cell tumors Other hematopoietic neoplasms including myeloma and leukemia Many present with advanced disease at diagnosis

HAART and chemotherapy Many patients will receive HAART and chemotherapy concurrently with high likelihood of drug interactions Protease inhibitors and non-nucleoside reverse transcriptase inhibitors are substrates and potent inhibitors or inducers of cytochrome P450 system (CYP) Many anti-neoplastic drugs also metabolized by CYP system leading to either drug accumulation and possible toxicity or decreased efficacy Paclitaxel and docetaxel Vinca alkaloids

Summary Since introduction of HAART, NHL and KS incidence has decreased Incidence of other cancers has increased related to other risk factors – immunosuppression, viral coinfections, smoking Prevention via risk factor control and screening Optimization of antiretroviral treatment Treatment strategies similar for non-HIV infected individuals