Tukei, Vincent J.a; Kekitiinwa, Adeodataa; Beasley, R. Palmerb
Kaposi's sarcoma, non-Hodgkin's lymphoma (NHL), cancer of the cervix, and a host of other rare cancers have consistently been shown to occur more frequently in HIV-infected than in uninfected individuals [1–3]. In sub-Saharan Africa, the prevalence of these HIV-associated malignancies has followed the pattern of the HIV epidemic [2,4,5].
Among African children, the incidence of these malignancies has been changing over the years [6,7]. In Uganda, childhood Kaposi's sarcoma, known to be very rare before the HIV epidemic, is now very common [8–10].
Most studies on HIV-associated malignancies conducted among Ugandan patients have largely assessed prevalence of HIV among adult patients with malignancies. Studies on HIV-infected children have remained few and information on the prevalence of malignancies among HIV-infected children enrolling for care is lacking, yet this information is vital for early diagnosis and treatment of these children. In addition, response to antiretroviral therapy (ART) and survival of HIV-infected Ugandan children that present with these malignancies is still unknown.
The objective of this study was to determine the prevalence of HIV-associated malignancies among children enrolling for care at an HIV/AIDS clinic in Kampala, Uganda; to determine the survival of these children; and to determine associations between these malignancies and other relevant variables such as age, sex, and immune status of the children. In this article, we report on a series of children that were diagnosed with HIV-associated malignancies at the time of enrolment for care at Baylor-Uganda clinic in Kampala, Uganda.
We conducted a retrospective case series that involved review of charts of HIV-infected patients who enrolled for care and treatment at the Baylor-Uganda clinic between 1 January 2004 and 31 December 2008. The Baylor-Uganda clinic is an HIV outpatient clinic caring for HIV-infected children aged 6 weeks to 18 years. The clinic, which is donor funded, is located within Mulago National Referral Hospital in Kampala, Uganda and is a major pediatric HIV service center in Uganda. One-third of patients seen at the clinic are referred from wards within Mulago Hospital, another third come as self-referrals, and the remainder are from other HIV treatment centers within Kampala city. Baylor-Uganda clinic provides free clinical care to over 10 000 of the estimated 150 000 HIV-infected children in Uganda.
During the study period, the clinic routine involved clinical and laboratory diagnosis of HIV, opportunistic infections, and malignancies. New patients seeking care and treatment at the clinic were tested for HIV infection and those found to be infected were examined and investigated for opportunistic infections and malignancies. All malignancies were systematically diagnosed by trained physicians equipped with flowcharts and lists of expected symptoms and signs and appropriate laboratory investigations to help in diagnosis. For all patients suspected of having a malignancy, tissue biopsy for histopathologic examination was carried out within the first 4 weeks of enrolling for care at the clinic. A positive histopathology report confirmed the diagnosis; negative reports in the presence of clinical evidence suggestive of malignancies were reviewed by another pathologist in order to establish accurate diagnosis.
All patients with confirmed malignancies were started on ART and cancer-specific chemotherapy. Patients without malignancies were assessed for ART eligibility based on WHO treatment guidelines [11,12]. A combination of three antiretroviral drugs that contained two nucleoside reverse transcriptase inhibitors (NRTIs) with the third drug being either a protease inhibitor or a non-NRTI (NNRTI) was the standard of care.
Cancer chemotherapy, which was prescribed by pediatric oncologists, included the use of vincristine and bleomycin for patients with Kaposi's sarcoma, and cyclophosphamide, vincristine, doxorubicin, and prednisone for patients with NHL.
Patient data, recorded by physicians who saw the patients, was double entered into a Microsoft Access database by trained data clerks. We selected patient charts for data abstraction based on an inclusion criterion that required study participants to be registered as patients at the clinic; be 6 weeks to 18 years of age at the time of enrolment for care; and be HIV positive and ART naive. All malignancies diagnosed within the first 4 weeks of enrolment for care were considered prevalent cases. Malignancies diagnosed after 4 weeks were considered incident cases and were excluded from the analysis.
Using a structured questionnaire, information on relevant variables such as age, sex, date of registration at the clinic, diagnoses, and baseline CD4 cell counts was abstracted. Data on type of malignancy, ART given, change in CD4 cell count, death (if any), and loss to follow-up were also abstracted.
Descriptive statistical analysis of proportions, means and SDs, and assessment of survival by Kaplan–Meier method was conducted using STATA 10.1 (StataCorp., College Station, Texas, USA). Cox proportional hazards models for univariate and multivariable analysis were developed to identify baseline factors associated with death during follow-up. All variables in the univariate analysis that were associated with death at a P value less than 0.2 together with other clinically relevant variables were included in the final multivariable model. Differences between the baseline and follow-up CD4 cell percentages were assessed using the Wilcoxon signed-rank test.
Written permission to conduct the study was obtained from Makerere University Faculty of Medicine Research and Ethics Committee (Uganda); Uganda National Council of Science and Technology (Uganda); Baylor College of Medicine Institutional Review Board (USA); and The University of Texas School of Public Health Research Office and IRB (USA). In addition, a waiver of patient consent was granted by Makerere University Faculty of Medicine Research and Ethics Committee.
Of the 6530 patients aged 6 weeks to 18 years who presented to the clinic during the study period, 109 patients (1.67%) had malignancies. One of the 109 patients was excluded because the record was incomplete both in the patient's chart and in the clinic's electronic database. Data on 108 patients (1.65%) were abstracted. The mean age of patients at baseline was 8.8 years (SD = 4.28), and 63 (57.8%) were male. Kaposi's sarcoma and NHL were the only malignancies seen during the study period and no patient had both malignancies (Table 1). Close to half the patients were already orphaned at the time of enrolment.
All but one of the 108 patients with malignancies continued to be HIV positive on follow-up specimens. The one patient who was found to be HIV negative on follow-up was a 2-year-old boy with Kaposi's sarcoma, who had earlier tested positive for HIV, but repeat tests after several months of follow-up showed him to be HIV-negative, so he was discharged from the clinic.
Of the 108 patients, 33 died and 21 were lost to follow-up (LTF). Eleven of those who died and 14 of those who were LTF did not start ART. Eighty-three patients started ART, 39 were on a protease inhibitor-based regimen, and 44 were on an NNRTI-based regimen. Upon starting treatment, CD4 cell percentage increased from a baseline median of 6% [interquartile range (IQR) 0–24%], to 14% (IQR 5–33%) at 6 months (P < 0.001), and to 15.8% (IQR 3–33%) at 12 months of ART (P = 0.032 for the 6–12 month increase).
Eighty patients (75.5%) received chemotherapy specific to their cancer; the remainder died, were LTF, or did not require chemotherapy as judged by the clinician.
Among patients with Kaposi's sarcoma, 32% had lesions only on lymph nodes, 26% had lesions involving the skin, 7% had lesions in the oral mucosa, 5% had lesions involving visceral organs, and 28% had disseminated disease involving two or more anatomic sites. All patients with visceral lesions died during follow-up and no patients with only mucosal lesions died. The majority of NHL patients were boys less than 12 years of age and all cases presented as Burkitt's lymphoma.
Patients with low CD4 cell counts (<15%) were more likely to have Kaposi's sarcoma compared with those with higher CD4 cell percentages (P = 0.032).
The probabilities of surviving during follow-up were 0.98 [95% confidence interval (CI) 0.92–0.99, n = 97], 0.91 (95% CI 0.83–0.95, n = 87), 0.78 (95% CI 0.68–0.85, n = 70), 0.73 (95% CI 0.63–0.81, n = 59), and 0.69 (95% CI 0.58–0.77, n = 44) at 1, 3, 6, 12, and 24 months of care, respectively. Deaths were extremely rare after 2 years of follow-up. Only three patients (two Kaposi's sarcoma and one NHL) died after their second year of follow-up.
In multivariable Cox regression analysis, the risk of death was not related to sex (hazard ratio 0.8, 95% CI 0.31–2.19); age category (≤12 years, hazard ratio 0.7, 95% CI 0.24–2.11); baseline CD4 cell percentage (hazard ratio 0.9, 95% CI 0.88–1.02); ART regimen (NNRTI versus protease inhibitor, hazard ratio 1.2, 95% CI 0.46–2.97); or type of malignancy (Kaposi's sarcoma, hazard ratio 0.7, 95% CI 0.08–5.58).
In this study, the proportion of children presenting with malignancies was 1.7%. Despite significant advances in diagnosis and treatment of HIV over the last two decades, HIV-associated malignancies remain an important cause of morbidity and mortality for children in this Ugandan hospital setting.
We found Kaposi's sarcoma the most prevalent malignancy occurring at an almost equal frequency between sexes, whereas NHL was predominantly seen among young boys. This sex distribution agrees with data from other studies conducted among children in the region [13–15].
Comparative epidemiological data on similar patients without HIV/AIDS is not available; however, incidence rates computed by Parkin et al. based on data from the Kampala Cancer Registry clearly indicate that these cancers were rarely seen in children before the outbreak of HIV/AIDS. The 2-year-old HIV-negative child that was excluded is an example of a rare case of Kaposi's sarcoma in an HIV-uninfected child.
Leiomyosarcoma was not seen nor did we see squamous cell carcinoma of the conjunctiva. We did not expect to find cancer of the cervix in this population that consisted of children and adolescents; however, this malignancy is not uncommon among adult Ugandan women .
The immune recovery seen in this series of patients compares favorably with that seen in similar patients without malignancies . Although most patients had low CD4 cell counts and were on chemotherapeutic drugs, they were able to mount robust immune recovery after starting ART. Having a low CD4 cell count did not increase the probability of death.
The contribution of these malignancies to child morbidity and mortality is often overshadowed by the effect of more prevalent opportunistic infections such as tuberculosis and pneumonias; however, in the Baylor-Uganda setting in which patient crude mortality rates are 5–10%, the need to reduce mortality directly resulting from these malignancies is critical. This descriptive study is a significant first step in our quest to understand malignancies in Ugandan children with HIV and how best to diagnose and treat them.
Our study may be limited by the descriptive nature of the study design, its use of retrospective routine data, lack of a comparison group of children without HIV, and lack of data on related viral infections such as Epstein–Barr virus, and human herpes virus 8. Nevertheless, we believe the results reported here are informative to clinicians seeking to understand the epidemiology and survival of children burdened with HIV and malignancies in the resource-constrained settings of Africa.
This study was supported in part by Grant D43 TW01036 from the Fogarty International Center of the National Institutes of Health.
The authors thank Dr Mark Kline, Program director for Baylor International Pediatric AIDS Initiative (BIPAI); Meg Ferris, Senior Vice President BIPAI; and all BIPAI (Houston) staff for facilitating this work.
The authors thank the entire Center for International Training and Research (CITAR) staff at the University of Texas (Houston) for their help and constructive criticism of the research proposal and manuscript review process and also thank Dr Alice Asiimwe (Baylor-Uganda) for her help during the IRB approval process.
This study was performed in partial fulfillment of the MPH thesis of V.J.T. at the School of Public Health of the University of Texas at Houston, USA.
Conflicts of interest
There are no conflicts of interest.
1. Longo DL, Steis RG, Lane HC, Lotze MT, Rosenberg SA, Preble O, et al. Malignancies in the AIDS patient: natural history, treatment strategies, and preliminary results. Ann N Y Acad Sci 1984; 437:421–430.
2. Parkin DM, Wabinga H, Nambooze S, Wabwire-Mangen F. AIDS-related cancers in Africa: maturation of the epidemic in Uganda. AIDS 1999; 13:2563–2570.
3. Biggar RJ, Frisch M, Goedert JJ. Risk of cancer in children with AIDS. AIDS-Cancer Match Registry Study Group. JAMA 2000; 284:205–209.
4. Chokunonga E, Levy LM, Bassett MT, Borok MZ, Mauchaza BG, Chirenje MZ, et al. AIDS and cancer in Africa: the evolving epidemic in Zimbabwe. AIDS 1999; 13:2583–2588.
5. Mosam A, Aboobaker J, Shaik F. Kaposi's sarcoma in sub-Saharan Africa: a current perspective. Curr Opin Infect Dis 2010; 23:119–123.
6. Chintu C, Athale UH, Patil PS. Childhood cancers in Zambia before and after the HIV epidemic. Arch Dis Child 1995; 73:100–104.discussion 4–5.
7. Mutalima N, Molyneux EM, Johnston WT, Jaffe HW, Kamiza S, Borgstein E, et al. Impact of infection with human immunodeficiency virus-1 (HIV) on the risk of cancer among children in Malawi: preliminary findings. Infect Agent Cancer 2010; 5:5.
8. Olweny CL, Kaddumukasa A, Atine I, Owor R, Magrath I, Ziegler JL. Childhood Kaposi's sarcoma: clinical features and therapy. Br J Cancer 1976; 33:555–560.
9. Ziegler JL, Katongole-Mbidde E. Kaposi's sarcoma in childhood: an analysis of 100 cases from Uganda and relationship to HIV infection. Int J Cancer 1996; 65:200–203.
10. Mbulaiteye SM, Katabira ET, Wabinga H, Parkin DM, Virgo P, Ochai R, et al. Spectrum of cancers among HIV-infected persons in Africa: the Uganda AIDS-Cancer Registry Match Study. Int J Cancer 2006; 118:985–990.
13. Stefan DC, Stones DK, Wainwright L, Newton R. Kaposi sarcoma in South African children. Pediatr Blood Cancer 2011; 56:392–396.
14. Gantt S, Kakuru A, Wald A, Walusansa V, Corey L, Casper C, et al. Clinical presentation and outcome of epidemic Kaposi sarcoma in Ugandan children. Pediatr Blood Cancer 2010; 54:670–674.
15. Orem J, Maganda A, Mbidde EK, Weiderpass E. Clinical characteristics and outcome of children with Burkitt lymphoma in Uganda according to HIV infection. Pediatr Blood Cancer 2009; 52:455–458.
16. Parkin DM, Nambooze S, Wabwire-Mangen F, Wabinga HR. Changing cancer incidence in Kampala, Uganda. Int J Cancer 2010; 126:1187–1195.
17. Kekitiinwa A, Lee KJ, Walker AS, Maganda A, Doerholt K, Kitaka SB, et al. Differences in factors associated with initial growth, CD4, and viral load responses to ART in HIV-infected children in Kampala, Uganda, and the United Kingdom/Ireland. J Acquir Immune Defic Syndr 2008; 49:384–392.
© 2011 Lippincott Williams & Wilkins, Inc.