HIV has been associated with an increased risk of malignancy,1 both AIDS defining2 and non-AIDS defining. Initiation of highly active antiretroviral therapy in the developed world has resulted in a significant decrease in AIDS-defining neoplasms such as Kaposi sarcoma (KS) and certain lymphomas, particularly primary central nervous system lymphoma and high grade B-cell non-Hodgkin lymphoma (NHL).1,3-5 NHL is generally not considered a major cause of AIDS morbidity in Africa. Explanations for this may include under-recognition due to limited access to diagnosis and care in poorly equipped medical facilities or due to competing mortality from other causes among HIV-infected people. Differences in lymphoma prevalence and presentation may well be expected in a region characterized by the huge genetic diversity of its populations, a high burden of HIV with diversity of circulating HIV subtypes and a high prevalence of communicable diseases contributing to malignancies such as Epstein-Barr virus6,7 and KS-associated herpesvirus.8-11
Generally, poor diagnostic capability and weak or nonexistent national cancer registries have resulted in limited descriptions of lymphomas emanating from African countries. South Africa has retained pathology skills related to conducting morphological assessment, flow cytometry immunophenotyping, and molecular characterization of malignancies using conventional cytogenetics, fluorescence-in-situ hybridization, and polymerase chain reaction (PCR). Despite this, the collection of data on AIDS-related malignancies is difficult because HIV is not a notifiable disease, HIV test data are unlinked to cancer registry data, and data analysis is compounded by a historical reluctance to capture HIV status on pathology reports.
This article makes an important contribution to the characterization of these lymphomas and allows establishment of a baseline that will enable ongoing monitoring of trends in lymphoproliferative disorders as the HIV epidemic matures and access to antiretroviral therapy increases.
Sample Population and Data Collection
A retrospective analysis was conducted of cases referred to the pathology departments at the Charlotte Maxeke Johannesburg Hospital, a major tertiary referral centre in Johannesburg, South Africa. All newly diagnosed lymphoproliferative disorders during the period of review (January 2004-December 2006) were included in this study. The antenatal clinic HIV prevalence was approximately 30% in the 2004-2006 period of review. A national household survey conducted in 2005 revealed an overall HIV prevalence of 15.8% [95% confidence interval (CI): 13.0-19.1] in the 15- to 49-year age group with much lower values above 60 years of age (<5%).12 Age, sex, and geographical differences in the prevalence of HIV are important to consider when analyzing these data. Data were collected by searching the laboratory information system at the site hosted by the National Health Laboratory Service. Limited demographic data were accessed including age, sex, referring health facility, and CD4 count when available at diagnosis. Ethics consent was obtained from the ethics committee of the University of the Witwatersrand (ethics number: M0-61110).
Classification of Tumors
Tumors were classified according to the World Health Organization classification of hematopoietic and lymphoid tumors of 2001.13
Determination of HIV Status
The HIV status was established by a systematic review of patients' results on the laboratory information system [enzyme-linked immunosorbent assay (ELISA)-based HIV serology] or established by HIV DNA PCR targeting the gag region of the genome following direct DNA extraction from stored paraffin embedded tissue blocks or slides. For the PCR assay, DNA from available tissue was extracted using the High Pure PCR Template Preparation Kit (Roche, Mannheim, Germany). PCR amplification and detection of the target HIV DNA was performed using the Roche Amplicor HIV-1 DNA test version 1.5,14 with modifications to the manufacturer's instructions (Roche Molecular Systems, Branchburg, NJ). The assay has an internal control to control for PCR inhibition.
During the period under review, 1897 new lymphoproliferative disorders were diagnosed and are summarized in Table 1. As expected, lymphomas were largely of B-cell origin15 and for this reason were summarized in more detail. The most commonly diagnosed lymphomas included diffuse large B-cell lymphoma (DLBCL), B-cell non-Hodgkin lymphoma (BNHL)/B-cell lymphoproliferative disorder not otherwise specified, small lymphocytic lymphoma/chronic lymphocytic leukemia (SLL/CLL), Hodgkin lymphoma, plasma cell dyscrasias, precursor B-cell acute lymphoblastic leukemia of the pre-B/common immunophenotype (Pre-B/common ALL), and Burkitt lymphoma/leukemia (BL). Rarer B-cell lymphoma entities not listed in Table 1 include B-cell acute lymphoblastic leukemia (1.3%; n = 25), hairy cell leukemia (0.07%; n = 15), lymphoplasmacytic lymphoma (0.02%; n = 4), and mucosa-associated lymphoid tissue lymphoma (1.2%; n = 24). T-cell lymphomas comprised only 3% (n = 74) of all lymphomas diagnosed and were placed in the following categories: anaplastic large T-cell lymphoma (1.4%; n = 27), angioimmunoblastic T-cell lymphoma (0.1%, n = 2), peripheral T-cell lymphoma: (1%; n = 19) and T-cell NHL (1.3%; n = 26) and will not be discussed further.
The overall prevalence of HIV infection across a subset of patients where HIV results were available (n = 709) was 37% (Table 1) and varied dramatically between the different lymphoma categories. HIV ELISA serology results were available for 505 (71%) of these individuals at time of diagnosis and HIV PCR tests were conducted for the remaining 194 cases. HIV prevalence ranged from a total absence in diseases such as follicular or mantle cell lymphoma to a low prevalence in diseases like SLL/CLL and pre-B/common ALL (4%-5%) to a very high prevalence (greater than 80%) in DLBCL and BL (Table 1). Patients diagnosed with primary effusion lymphoma (n = 5) were all HIV positive.
DLBCL was the most frequently diagnosed lymphoma in our patient population, with an overall prevalence of 21%. The mean age at first diagnosis was 43.9 years (median 40 years; range, 2-87 years). Of the 401 newly diagnosed cases of DLBCL, HIV status could only be ascertained in 169 (42%). Within this group, 135 (80%) of patients were HIV positive (Table 1). In the HIV-positive patients with DLBCL with available CD4 counts (n = 78), the mean CD4 count at diagnosis of DLBCL was 158 cells/mm3 (median 104 cells/mm3, range 9-1200 cells/mm3). A large group of B lymphoproliferative disorders comprising 17% (n = 330) of the total group could not be further characterized due to submission of inadequate biopsy material or only cells from fine needle aspiration for flow cytometry. HIV status could be determined in 105 of these cases and 59% were HIV positive. Fifty individuals in the HIV-positive group had CD4 counts at diagnosis with a mean of 143 cells/mm3 (median 73 cells/mm3; range 5-664 cells/mm3).
The plasmablastic variant of DLBCL, a reportedly rare HIV-associated neoplasm, comprised 24% (n = 97) of all DLBCL cases and 5% of all lymphomas. Of those with HIV results, 94% (47/50) were HIV positive and the mean age at presentation was 38 years (median 36; range, 6-78 years). The oral cavity was involved in 33% (n = 32) of cases with the remainder including the gastrointestinal tract (11%), anorectal region (13%), skin (5%), bone marrow (4%), genitalia (9%), and other sites (25%). In addition to clinical features, common immunophenotypic features included a lack of definite B-lineage expression (CD20 and CD22 negative, possible dim CD19 expression) and the presence of plasma cell markers (namely, bright CD38 and CD138 expression).
Burkitt/atypical Burkitt lymphoma accounted for 6% of all newly diagnosed lymphomas and were strongly associated with HIV infection. In fact, 86% of all BL cases were HIV positive. In adults in the 20- to 40-year age range, the prevalence of HIV infection was 100% (refer Fig. 1). Outside this age range, HIV-negative status possibly reflects the presence of migration from an endemic region and sporadic cases of BL, well described before the era of HIV infection. Within the HIV-positive group, a CD4 count was obtained for 46 cases and revealed a mean of 138 cells/mm3 (range 14-960 cells/mm3). All cases of BL screened for the t(8:14) translocation were positive, however, 6 cases showed additional atypical abnormalities not usually described such as aneuploidy, a low S phase fraction and lack of light chain restriction using conventional flow cytometric techniques.
In the study period, there were 137 new cases of Hodgkin lymphoma. Forty-six percent (24/52) of Hodgkin lymphomas tested were HIV positive (Table 1). The nodular sclerosis subtype was the commonest subtype (39%) followed by the mixed cellularity subtype (26%). Primary effusion lymphomas characterized by their clinical presentation [peritoneal (n = 4), pleural (n = 1)], the absence of B- and T-cell immunophenotypic markers together with a monoclonal immunoglobulin heavy chain gene rearrangement, was a rare diagnosis (0.3%).
It was estimated that 25,000 new NHL cases were diagnosed in sub-Saharan Africa and a further 5000 cases in North Africa in 2002.16 This study represents the first detailed description of lymphoproliferative disorders in Johannesburg, coinciding with the introduction of combination antiretroviral treatment. These data support the association between HIV-1 and certain NHL categories seen in other African studies,17,18 and the reports that highlight an increased incidence of these tumors from both Harare (Zimbabwe)19 and Kampala (Uganda).20 This analysis follows a recent hospital-based case-control study in Johannesburg describing the spectrum of HIV-associated malignancies in the pre-highly active antiretroviral therapy era (1995-2004). This report revealed an increased risk for KS [odds ratio (OR) = 47.1, 95% CI = 31.9-69.8), NHL (OR = 5.9, 95% CI = 4.3-8.1), cancer of the cervix (OR = 1.6, 95% CI = 1.3-2.0), Hodgkin lymphoma (95% CI = 1.0-2.7), cancers of the anogenital region (OR = 2.2; 95% CI = 1.4-3.3), and squamous cell carcinoma of the conjunctiva (OR = 2.6, 95% CI = 1.4-4.9).21 The 6-fold increased risk in HIV infection for developing NHL was consistent with other African sites, but substantially lower than in developed countries16,17 and the reasons for this remain largely unexplained.
In the period of review, DLBCL was the most commonly diagnosed histological subtype in the public sector in Johannesburg. Despite this, the overall prevalence of 21% is probably an underestimate, with a significant number of cases likely to be represented in the B-cell lymphoproliferative group that could not be classified further due to the absence of adequate lymph node biopsy material. Patients with DLBCL in this analysis were younger than counterparts from developed nations, a finding consistent with previous reports from Central Africa.22 Eighty percent of patients with DLBCL tested for HIV were indeed HIV positive, representing an increase over that described for similar categories in the earlier case-control study.21 The expanding spectrum of anatomical sites for the plasmablastic variant of DLBCL has been previously reported23 and recently acknowledged in the revised World Health Organization classification of lymphomas.24 In contrast to previous work from western populations,25 patients were of equal gender and had greater involvement of sites other than the oral cavity. Although it should be noted that difficulties can be experienced in distinguishing immunoblastic from plasmablastic variants of DLBCL.
Burkitt lymphoma was strongly associated with HIV in the young adult population in Johannesburg. Burkitt lymphoma is not endemic in South Africa and cases of jaw tumors affecting young children generally reflects population migration from east, central, and west Africa, where this disease may account for 70% of all childhood cancers.26 A survey in Kenyan adults also showed a 3-fold increase in BL (66% were HIV positive) between 1992 and 1996.27 A finding that is confirmed in HIV-infected individuals in Uganda.20,27 Atypical clinical and laboratory findings may make it difficult to distinguish between Burkitt/atypical Burkitt lymphoma and DLBCL. Although recent gene expression profiling has revealed a unique molecular signature for BL, a group of cases remained that were intermediate between DLBCL and BL.28,29
The high prevalence of HIV in Hodgkin Lymphoma supports the Stein study which revealed an increased risk for the malignancy in HIV (95% CI = 1.0-2.7).21 Hodgkin lymphoma of the nodular sclerosis type was the commonest subtype which is in contrast to other studies in Africa, where mixed cellularity is most prevalent. Further studies are needed to exclude the possibility that this picture is merely a reflection of the background HIV prevalence in the young adult population rather than an association of this lymphoma subtype with HIV. Primary effusion lymphoma was rarely diagnosed in Johannesburg despite a high prevalence of HIV and KS-associated herpesvirus in the population. This may reflect under diagnosis due to lack of awareness or poor access to appropriate health care facilities.
The limitations of the study include its retrospective nature, the failure to determine HIV status on all biopsy samples submitted (hence some of the wide confidence intervals noted), or that HIV status may be underestimated in PCR-based analyses. This study further underscores the importance of establishing sustainable South African and regional cancer registries and highlights the need for linking these to HIV-testing databases. The strengthening of regional cancer registries is essential to monitor the potential rise of non-AIDS-defining malignancies that may emerge as a result of the HIV epidemic.
1. Engels EA, Biggar RJ, Hall HI, et al. Cancer risk in people infected with human immunodeficiency virus in the United States. Int J Cancer
2. CDC. Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. Morb Mortal Wkly Rep
3. Cinti SK, Gandhi T, Riddell J. Non-AIDS-defining cancers: should antiretroviral therapy be initiated earlier? AIDS Read
. 2008;18:18-20, 6-32.
4. Palmieri C, Treibel T, Large O, et al. AIDS-related non-Hodgkin's lymphoma in the first decade of highly active antiretroviral therapy. Q J Med
5. Besson C, Goubar A, Gabarre J, et al. Changes in AIDS-related lymphoma since the era of highly active antiretroviral therapy. Blood
6. Queiroga E, Gualco G, Chioato L, et al. Viral studies in Burkitt lymphoma: association with Epstein-Barr virus but not HHV-8. Am J Clin Pathol
7. Carbonea A, Gloghinia A, Dottib G. EBV-associated lymphoproliferative disorders: classification and treatment. The Oncologist
8. Laurent C, Meggetto F, Brousset P. Human herpesvirus 8 infections in patients with immunodeficiencies. Hum Pathol
9. Adjei AA, Armah HB, Boamah I, et al. Seroprevalence of HHV-8, CMV, and EBV among the general population in Ghana, West Africa. BMC Infect Dis
10. Sarmati L. HHV-8 infection in African children. Herpes
11. Engels EA, Mbulaiteye SM, Othieno E, et al. Kaposi sarcoma-associated herpesvirus in non-Hodgkin lymphoma and reactive lymphadenopathy in Uganda.. Hum Pathol
12. Shisana O, Rehle T, Simbayi L, et al. South African national HIV prevalence, HIV incidence, behaviour and communications survey. Cape Town, South Africa: Human Sciences Research Council; 2005.
13. Jaffe ES, Harris NL, Stein H, et al. World Health Organization Classification of Tumours. Pathology and genetics of tumours of haematopoietic and lymphoid tissues. Lyon, France: IARC Press; 2001.
14. Sherman GG, Cooper PA, Coovadia AH, et al. Polymerase chain reaction for diagnosis of human immunodeficiency virus infection in infants in low resource settings. Pediatr Infect Dis J
15. Levine AM, Seneviratne L, Espina BM, et al. Evolving characteristics of AIDS-related lymphoma. Blood
16. Parkin D, Sitas F, Chirenje M, et al. Part I: cancer in Indigenous Africans-burden, distribution, and trends. Lancet Oncol
17. Mbulaiteye SM, Katabira ET, Wabinga H, et al. Spectrum of cancers among HIV-infected persons in Africa: the Uganda AIDS-Cancer Registry Match Study. Int J Cancer
18. Newton R, Ziegler J, Beral V, et al. A case-control study of human immunodeficiency virus infection and cancer in adults and children residing in Kampala, Uganda. Int J Cancer
19. Chokunonga E, Levy LM, Bassett MT, et al. Aids and cancer in Africa: the evolving epidemic in Zimbabwe. Aids
20. Parkin DM, Wabinga H, Nambooze S, et al. AIDS-related cancers in Africa: maturation of the epidemic in Uganda. Aids
21. Stein L, Urban MI, O'Connell D, et al. The spectrum of human immunodeficiency virus-associated cancers in a South African black population: results from a case-control study, 1995-2004. Int J Cancer
22. Cool C, Bitter M. The malignant lymphomas in Kenya; morphology, immunophenotype and frequency of EBV in 73 cases. Hum Pathol
23. Chetty R, Hlatswayo N, Muc R, et al. Plasmablastic lymphoma in HIV+
patients: an expanding spectrum. Histopathology
24. Swerdlow SH, Campo E, Harris NL, et al. WHO classification of tumours of haematopoietic and lymphoid tissues. Lyon, France: IARC; 2008.
25. Riedel DJ, Gonzalez-Cuyar LF, Zhao XF, et al. Plasmablastic lymphoma of the oral cavity: a rapidly progressive lymphoma associated with HIV infection. Lancet Infect Dis
26. Stiller CA, Parkin DM. International variations in the incidence of childhood lymphomas. Paediatr Perinat Epidemiol
27. Otieno MW, Remick SC, Whalen C. Adult Burkitt's lymphoma in patients with and without human immunodeficiency virus infection in Kenya. Int J Cancer
28. Boerma EG, Siebert R, Kluin PM, et al. Translocations involving 8q24 in Burkitt lymphoma and other malignant lymphomas: a historical review of cytogenetics in the light of todays knowledge. Leukaemia
29. Salaverria I, Zettl A, Beà S, et al. Chromosomal alterations detected by comparative genomic hybridization in subgroups of gene expression-defined Burkitt's lymphoma. Haematologica