There are several distinct clinical and epidemiologic forms of Kaposi's sarcoma (KS*), including an iatrogenic form occurring in patients who have received organ allografts (1). The incidence of KS among renal transplant recipients in Western countries is many times the expected incidence in the general population(2). Nonetheless, the overall incidence of KS among renal transplant recipients is relatively low; about 0.4% of renal transplant recipients develop this tumor, and KS represents only 5% of cancer in transplant recipients (3). However, in Saudi Arabia, KS occurs in 4.1% of renal transplant recipients, about 10-fold higher than in Western countries, and accounts for 70% of malignancies in these patients(4, 5). The DNA of a newly identified human gammaherpesvirus known as Kaposi's sarcoma-associated herpesvirus (KSHV) or human herpesvirus eight (HHV8) has been identified in all forms of the disease(6). However, important questions about the association of the new virus with KS remain unanswered. One question is “How specific is the association of KSHV infection with the clinical expression of KS?” A second related question is, “How widespread is infection with the new virus in different populations at high risk for development of KS?” Seroepidemiologic studies are a classical way to address such questions. The determination of antibodies to HHV8 in Saudi Arabian patients with and without KS represented a unique epidemiologic opportunity to evaluate the role of the new virus in renal transplant recipients.
In previous serologic studies from our laboratory, a three-to fourfold higher prevalence of antibodies to HHV8 had been detected in the sera of patients with human immunodeficiency virus (HIV)-associated KS than in HIV-infected patients without KS (7). These studies, initially employing a cell-associated HHV8 antigen, p40, were later confirmed using a recombinant HHV8 small viral capsid antigen (8). Similar results were obtained when other serologic tests measuring antibodies to nuclear antigens and recombinant capsid antigens were used in patients with HIV-related and classical KS (9, 10). Thus serologic studies provided additional evidence for an etiologic link of the newly discovered virus with KS.
In HIV-infected patients, KS is thought to be a sexually transmitted disease. Serologic studies in homosexual and bisexual HIV-infected patients could not exclude the possibility that transmission of HHV8 occurred in parallel with transmission of a yet-unidentified etiologic agent of KS. Therefore, we sought serologic evidence for KSHV in another high risk population, namely renal transplant recipients, in whom sexual transmission of the virus is considered to be less likely. This group has not been previously studied for seroreactivity to HHV8. As controls, we studied Saudi Arabian patients with chronic renal failure, other malignancies, and healthy individuals.
Subjects with posttransplant KS and control subjects were selected from the outpatient clinics of the King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia. All transplant recipients had received a standard immunosuppressive regimen, consisting of prednisone plus cyclosporine and azothiaprine. Clinical and demographic data were recorded on standardized forms that were linked to the serum samples by numerical code.
All serologic assays were performed and interpreted while the serum samples remained coded. The sera were heated at 56°C for 30 min before testing. Sera were screened at a dilution of 1:100. Antibodies to butyrate-induced HHV8 p40 were measured by an immunoblot assay described previously(7). Untreated BC-1 cells served as a negative control.
Antibodies were also measured to a recombinant capsid antigen protein designated sVCA expressed in Escherichia coli. sVCA was isolated from a cDNA library obtained 48 hr after butyrate induction of BC-1 cells(8). sVCA represents a 170-amino acid KSHV capsid protein which is the homologue of Epstein-Barr virus (EBV) BFRF3 and Herpes saimiri virus open reading frame 65 (10). Extensive studies have shown that KSHV sVCA is not immunologically cross-reactive with EBV BFRF3. Specifically, neither polyclonal human antibodies nor monospecific rabbit antibodies to EBV BFRF3 react with KSHV sVCA. Conversely, monospecific rabbit antibodies to KSHV sVCA do not recognize EBV BFRF3 (8). sVCA was cloned in pET30b (Novagen); pET30b without an insert served as a negative control. The expression plasmids were transformed into E coli strain BL21 (DE3)pLysS. The sVCA antigen was purified on a nickel column according to the manufacturer's directions. Comparable fractions of the vector control extract served as a negative control.
A serum sample was considered to have antibody to HHV8 lytic cycle antigens if there was antibody to either p40 or sVCA (Fig. 1). Twenty-one of the 76 sera examined contained antibody to both antigens, 52 sera lacked antibody to both antigens; only 3 sera (4%) were discordant in the two tests. Two samples contained antibody to p40 and lacked antibody to sVCA; one sample contained antibody to sVCA and lacked antibody to p40.
Rates of KSHV seropositivity were compared using chisquare and Fisher's exact test; odds ratios (ORs) were calculated with Cornfield 95% confidence intervals (CIs). Associations between seroreactivity and other variables were compared using logistic regression. Differences in age distribution were calculated using Student's t test. All calculations were performed using EpiInfo and SAS.
Serum was collected from 76 individuals. The study population included all of fourteen identified transplant recipients with KS (group 1) as well as the following control groups: renal transplant recipients without KS (group 2, n=18), patients with chronic renal failure (group 3, n=14), oncology patients with normal renal function (group 4, n=15), and healthy volunteers (group 5, n=15). Three renal transplant recipients in group 2 without KS had other neoplasms: one cancer of the thyroid, one cancer of the rectum, and one non-Hodgkin's lymphoma. The median time from transplantation to development of KS was 13 months (range, 4-36 months).
Demographic characteristics and seroprevalence rates in cases and controls are described in Table 1. Seroreactivity to HHV8 antigens was significantly more prevalent in KS cases compared to transplant patients without KS (group 2; OR=33.80, CI=2.96-904, P<0.001) and compared to the aggregate control group (groups 2-5; OR=60.27, CI=6.6-1393,P<0.000001). Seroreactivity was more common in male subjects than in females (41% vs. 22% seropositive) but this difference was not statistically significant (OR=0.43, CI=0.14-1.32, P=0.096).
Among the control groups, overall seroprevalence was 18%. Seroreactivity was more common among patients with renal failure (28% seropositive in groups 2 and 3 combined) compared to other controls (6.7% seropositive in groups 3 and 4 combined), but in a logistic regression model, including the variables of age, renal failure, and transplant status, only age was independently associated with KSHV seropositivity. The small sample size prevents further conclusions about the role of factors such as blood transfusion or immunosuppression which may be associated with renal failure, in seroprevalence to HHV8.
By demonstrating that Saudi Arabian patients with post-transplant KS are nearly universally seroreactive to KSHV antigens, this study strengthens the etiologic link between KSHV infection and clinical KS. About 90% of American HIV-1-infected patients with KS are also seroreactive with recombinant sVCA(8, 10). Thus, in the two clinical forms of KS that are associated with overt immunosuppression, there is extremely high concordance between antibodies to KSHV and the disease.
The prevalence of antibodies to KSHV in these two groups of high risk patients without KS is also remarkably similar. Among American HIV-1-infected patients without KS, 20% were seropositive to sVCA (8). In this study, 28% of Saudi patients with renal failure were seroreactive with the same antigen. In earlier studies of HIV-1-infected patients, it has been proposed that the seroepidemiology of KSHV conforms to that of a sexually transmitted disease (9). The demonstration of a similar seroprevalence in Saudi patients with renal disease but without KS makes it possible that other routes of transmission, including transfusion of blood or blood products, infection from the transplanted organ, or opportunistic infection secondary to immunosuppression of chronic renal disease, may be equally important.
Reliable estimates of the prevalence of antibodies to HHV8 in the general population of Saudi Arabia await further large scale studies. However, the low seroprevalence in control groups 4 and 5 (7%) is similar to that reported for healthy blood donors in the US (5%) using a similar recombinant sVCA preparation (10).
What accounts for the high incidence of KS among Saudi Arabian transplant recipients? There is nothing unusual about the immunosuppressive regimen that is used to maintain the renal allografts. All patients receive corticosteroids; the patients also receive cyclosporine and azothiaprine. These are standard regimens. Previous studies have explored the possible role of genetic and geographic factors. In initial studies, it appeared that KS was associated with the HLA A2 genotype, but this association was not confirmed in additional patients (5). HLA DR5 is important in classical and epidemic forms of KS but not in Saudi Arabia. The increased incidence of KS in Saudi Arabian transplant recipients might be related to the source of the donated kidneys, to variations in locally circulating strains of HHV8, or to immunologic or cellular factors that influence host-virus interactions.
In summary, seroreactivity was strongly associated with clinical KS. Seroreactivity was more common among patients with renal failure than in other control groups, but this association may be confounded by age or other variables.
Acknowledgments. The authors thank Daad Abu-rub and Dr. Adnan Ezzat for assistance in obtaining blood samples and Josephine Shepard for assistance with biostatistics.
Abbreviations: CI, confidence interval; EBV, Epstein-Barr virus; HHV8, human herpesvirus eight; HIV, human immunodeficiency virus; KS, Kaposi's sarcoma; KSHV, Kaposi's sarcoma-associated herpesvirus; OR, odds ratio.
1. A-Sulaiman MH, A-Khader AA. Kaposi's sarcoma in renal transplant recipients. Transplant Sci 1994; 4: 46.
2. Penn I. Kaposi's sarcoma in immunosuppressed patients. J Clin Lab Immunol 1983; 12: 1.
3. Penn I. Kaposi's sarcoma in organ transplant recipients. Transplantation 197; 27: 8.
4. Qunibi W, Akhtar M, Sheth K, et al. Kaposi's sarcoma: the most common tumor after renal transplantation in Saudi Arabia. Am J Med 1988; 84: 225.
5. Qunibi WY, Barri Y, Alfurayh O, et al. Kaposi's sarcoma in renal transplant recipients: a report on 26 cases from a single institution. Transplant Proc 1993; 25: 1402.
6. Chang Y, Cesarman E, Pessin MS, et al. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi's sarcoma. Science 1994; 265: 1865.
7. Miller G, Rigsby M, Heston L, et al. Antibodies to butyrate-indicuble antigens of Kaposi's sarcoma-associated herpesvirus in patients with HIV-1 infection. N Engl J Med 1996; 334: 1292.
8. Lin S-F, Sun R, Heston L, et al. Identification, expression and immunogenicity of Kaposi's sarcoma-associated herpesvirus-encoded small viral capsid antigen. J Virol 1997; 71: 3069.
9. Kedes D, Operskalski E, Busch M, et al. The seroepidemiology of human herpesvirus 8 [Kaposi's sarcoma associated herpesvirus]: distribution of infection in KS risk groups and evidence of sexual transmission. Nat Med 1996; 2: 918.
10. Simpson GR, Schulz TF, Whitby D, et al. Prevalence of Kaposi's sarcoma associated herpesvirus infection measured by antibodies to recombinant capsid protein and latent immunofluorescence antigen. Lancet 1996; 348: 1133.