Infection With Kaposi’s Sarcoma–Associated Herpesvirus Among Families of Patients With Classic Kaposi’s Sarcoma FREE

Kaposi’s sarcoma–associated herpesvirus (KSHV), also called human herpesvirus 8, is an essential etiologic agent of all clinical and epidemiological variants of Kaposi’s sarcoma (KS), including the classic, endemic (African), iatrogenic, and AIDS-associated (epidemic) forms.^{1 - 2} Generally, the prevalence of KSHV varies between geographic regions from an overall infection rate of 2% to 7% in Western Europe and North America, to 10% to 20% in Mediterranean countries, and up to about 100% in certain sub-Saharan African countries.³ Correspondingly, KS presents an uneven geographic distribution. AIDS-associated KS appears in a broad range of geographic areas and predominantly develops in homosexual men, who demonstrate a relatively high overall infection rate.

Among homosexual men, KSHV infection is associated with a high lifetime number of sexual partners, other sexually transmitted diseases, and sexual contact with other men from communities where AIDS-associated KS is common.^{4 - 5} Thus, sexual transmission of KSHV appears to play an important role among homosexual men. The transmission routes of KSHV among heterosexuals are not yet fully understood. In areas endemic to KSHV infection, such as sub-Saharan Africa, horizontal transmission is acquired in childhood and the prevalence of KSHV steadily increases with age in the absence of sexual exposure.^{5 - 7} This supports an alternate nonsexual horizontal transmission route. Mother-to-child transmission, during pregnancy or at delivery, is also likely to occur in these geographic areas, but the low KSHV seroprevalence in children younger than 5 years argues against this route as a major mode of transmission.^{8 - 11} In addition, several studies in African countries demonstrated an association between KSHV seropositivity, multiple sex partners, and sexually transmitted diseases.^{12 - 13} In North America and Western Europe, infection is rare in childhood and the acquisition of KSHV infection occurs mainly after puberty.¹⁴ However, sexual activity has not been specifically associated with an increased risk of KSHV infection among heterosexuals.^{4 - 5} Likewise, in the developing world, most children become infected with Epstein-Barr virus within the first decade of life, whereas in the more developed western world, up to half of children remain seronegative at the end of their first decade of life and become infected in adolescence or young adulthood.¹⁵

In Italy, KSHV DNA has been found in the peripheral blood mononuclear cells of up to 50% of individuals with AIDS-associated KS^{16 - 17} and in 8% to 10% of healthy subjects.^{17 - 18} Infectious viruses and genomic KSHV DNA were found in the saliva of patients with KS,^{19 - 20} and viral DNA was detected in semen and prostatic fluid of patients with KS.^{21 - 22} Accordingly, KSHV could potentially be transmitted through blood contact, sexual contact, saliva, and transplanted organs.

Classic KS (CKS) is found mostly among elderly people of Mediterranean or Jewish origin with no apparent immunodeficiency.²³ Between 1960 and 1998, the overall age-standardized incidence rates (±SD) of CKS in the Israeli Jewish population were 20.7 ± 9.3 per million among men and 7.5 ± 3.4 per million in women.²⁴ These rates represent one of the highest incidence rates of CKS in the developed world. Seroprevalence surveys^{25 - 26} found 6.5% to 9.9% of the Israeli Jewish population to be infected with KSHV. In view of the intermediate infection rates with KSHV and the high living standards, Israel presents a unique opportunity to investigate the epidemiology of CKS and to evaluate the transmission modes of KSHV.

The objective of the present study was to assess the intrafamilial patterns of KSHV transmission to obtain insight into the transmission modes of KSHV among heterosexuals. We have focused on first-degree family members of patients with CKS and compared this population with a randomly selected hospital-based control population. In what we believe to be the largest CKS family study conducted thus far, we present evidence of intrafamilial transmission of KSHV in families of Israeli CKS patients.

Enrollment of CKS patients to the study was based on Israel Cancer Registry information and used a systematic sampling approach of Jewish patients from 3 dermatological departments in Israel (Rambam Medical Center in Haifa, HaEmek Medical Center in Afula, and Rabin Medical Center in Petah-Tikva) whose diagnosis was ascertained. Patients with AIDS-associated KS were excluded from the study, and selected patients had at least a 2-year history of disease. In total, 200 CKS patients were sampled; 64 (48 men and 16 women) were located and agreed to participate in the study. Concurrently, 146 of their first-degree family members—spouses, offspring, and siblings, all of them living in Israel—agreed to participate in the study. Because the CKS patients were rather old, no children were included in the family member group. For the control population, 189 subjects were consecutively recruited from the emergency and orthopedic departments. Written informed consent was obtained from all study participants, and ethical approval was obtained from the Helsinki committee of the Israeli Ministry of Health.

Blood samples were collected during 1999-2002, and serum was separated and stored at −20°C until testing. All study participants underwent a face-to-face interview. Data were recorded on standardized forms that were linked to the serum samples by numeric code. Statistical analyses were performed using SPSS statistical software (version 10.0, SPSS Inc, Chicago, Ill). The null hypothesis for the final stage of the analysis was that there would be no association between the serostatus of KSHV and the category of familial relationship to the CKS patients. A 2-sided P<.05 was considered statistically significant and sufficient to refute the null hypothesis.

All serologic assays were performed and interpreted while the serum samples remained coded. We adopted a unique flowchart approach for defining KSHV seropositivity (Figure). Each serum sample was first tested for KSHV antibodies using the 2 widely used immunofluorescence assays (IFAs): latent IFA with BCP-1 cells using a serum dilution of 1:100²⁷ and lytic IFA at a dilution of 1:40 (Biotrin International, Dublin, Ireland). Results were independently scored by at least 2 investigators. If the results were concordant, a final positive or negative serologic diagnosis was made. Because lytic IFA can produce a false-positive diagnosis, we used a confirmatory Western blot assay to screen for antibodies to the small capsid-related protein encoded by KSHV open reading frame 65 (ORF65)²⁸ at a dilution of 1:100 for samples negative by latent IFA but positive by lytic IFA. Seropositivity established by both lytic IFA and ORF65 Western blot defined serum samples as positive for KSHV. If this test was not confirmatory, all 3 assays were repeated. Serum samples that were negative by lytic IFA but positive by latent IFA were also tested by the KSHV-negative Ramos cell line to exclude nonspecific nuclear reactivity. Serum samples that tested negative by this assay and positive by latent IFA were scored as positive.

Sixty-four CKS patients were studied, including 48 men and 16 women. The mean ± SD age of the CKS patients was 71.8 ± 11.0 years, and is comparable (P = .16) to that of Jewish Israeli CKS patients, which was 69.2 ± 14.8 years during the period between 1970 and 2000.²⁴ Our serologic evaluation strategy indicated that 62 (96.9%) of the patients were seropositive. Of all 64 patients, 62 (96.9%) were seropositive by latent IFA, while 61 (95.3%) were seropositive by lytic IFA. One 70-year-old woman with CKS presented a seropositive response by lytic IFA and seronegative responses by latent IFA and the ORF65 Western blot assay; based on our serostatus evaluation strategy, this patient was designated as being seronegative. In addition, a 53-year-old woman with CKS, who was also diagnosed as having pulmonary KS, was found to be seronegative by all assays and therefore was designated as being seronegative as well. The concordance rate between latent and lytic IFAs among CKS patients was 96.9% (62 of 64).

We next tested the seroprevalence of KSHV among 146 family members (1-7 family members for each of 60 patient with CKS) and 189 hospital controls. Following the completion of the serologic evaluation, we excluded from the statistical analysis 3 control subjects and 1 family member who reported homosexual or bisexual sexual preferences, as well as 2 family members with indeterminate KSHV serostatus. Characteristics of the 2 groups are shown in Table 1. Because the characteristics of the 2 groups differed, we repeated all statistical analyses separately for each group.

The overall seroreactivity to KSHV in family members of CKS patients was 39.2% (56 of 143 family members). A significantly lower overall seroreactivity to KSHV was found among the hospital controls (11.3% [21 of 186 hospital controls], P<.001). The estimated KSHV prevalence rate in our control population was similar to the adjusted rate reported previously in Israel.²⁶ The presence of antibodies to KSHV was strongly associated with clinical CKS, whether we were comparing the CKS patients with their family members (P<.01) or with the hospital controls (P<.01).

Similar frequencies of KSHV seropositivity were observed among men and women in the family members and the hospital controls (Table 1). The KSHV seropositivity peaked in family members 65 years and older and in the 50- to 64-year age group in the hospital controls (Table 1). Among family members, the mean ± SD age of those who were seropositive was 53.1 ± 17.0 years, and that of those who were seronegative was 46.8 ± 15.2 years. Among the hospital controls, the mean age of the seropositive persons was 59.9 ± 15.5 years, and that of those who were seronegative was 53.6 ± 19.2 years. Overall, seropositive individuals among family members of CKS patients and hospital controls tended to be older than seronegative individuals. This trend, however, was not statistically significant for hospital controls (Table 1).

The first-degree family members of CKS patients included spouses, offspring, and siblings (Table 2). Having a family member with CKS was a strong predictor of KSHV seropositivity (odds ratio, 6.44; 95% confidence interval, 3.36-12.32; P<.001). Among the relatives of CKS patients, seroprevalence rates varied between 56% in siblings, 49% in spouses, and 34% in offspring. However, these differences were not statistically significant (Table 2). Closer analysis of the families of CKS patients demonstrated clustering of serologic status, especially among families in which the mother had CKS or was KSHV-seropositive (Table 3). Overall, seroprevalence for KSHV in families in which the father had CKS and the mother was KSHV-seropositive was relatively higher compared with families in which the father had CKS and the mother was KSHV-seronegative (P = .07) (Table 3). This observation is partially in line with data presented by Angeloni and colleagues,²⁹ who described full concordance between the KSHV serostatus of mothers and their offspring.

We identified a relatively high seroprevalence of antibodies to KSHV among family members of CKS patients. Because the mean ± SD age of the hospital controls in our study (54.3 ± 18.8 years) was significantly higher than that of the family members (49.2 ± 16.1 years) (P = .01), and because the seroprevalence of KSHV increases with age, this difference further amplifies the inequality of KSHV seroprevalence between the 2 groups studied. Overall, we found that family members were 6.44-fold (95% confidence interval, 3.36-12.32) more likely to be infected with KSHV than were control subjects. Although it is unknown whether familial clustering is due to direct intrafamilial transmission or due to shared risk factors for infection, the latter seems unlikely because of the high seroprevalence of KSHV among family members of CKS patients (39.2% vs 11.3% in hospital controls). The seroprevalence of KSHV was higher among spouses than offspring, suggesting possible sexual transmission of KSHV, but this difference was not statistically significant (Table 2). Likewise, a study from Sardinia²⁹ of 20 CKS patients and 36 of their first-degree family members demonstrated clustering of seropositive individuals within families of CKS patients, with no significant difference between seroprevalence in spouses compared with seroprevalence in other family members. The correlation of KSHV serostatus between mother and child (Table 3), although only marginally significant, supports an increased risk of mother-to-child KSHV transmission compared with father-to-child transmission. This correlation between the serostatus of the mother and that of her offspring could reflect the greater role of the mother in intimate care and does not necessarily represent transmission during pregnancy, delivery, or breastfeeding. Significant correlations in KSHV serostatus between mothers and offspring, between siblings, and between spouses were reported in Israel and in Tanzania in families not affected with KS.^{26 ,30} In a study⁹ of people of African origin living in French Guyana, similar correlations were noted between the serostatus of children and that of their mothers and between siblings, but not between the serostatus of children and that of their fathers nor between spouses.

Salivary shedding of KSHV may provide a possible mode of intrafamilial transmission, because exposure to saliva may occur while playing, sharing eating implements, and kissing during childhood and between adults. Sexual and vertical transmission among family members of CKS patients may also take place. Still, the similarity in the risk of KSHV infection among spouses, offspring, and siblings supports similar modes of virus infection among first-degree family members with different relationships (spouse, sibling, parent or child). Therefore, our study suggests a predominantly nonsexual, person-to-person spread among family members. The increased correlation between seropositivity of the mother with that of the offspring further supports this notion and could suggest that mothering responsibilities result in increased exposure to maternal transmission. The relatively limited transmission of KSHV from KSHV-seropositive fathers to their offspring is similar to the intrafamilial transmission patterns of hepatitis B virus infection.^{31 - 32}

We found no significant sex difference in the prevalence of KSHV antibodies among family members or hospital controls. This suggests that some factor other than an increased exposure to virus infection must account for the observed male-female ratio of 2.8:1 in the incidence of CKS in Israel.²⁴ A similar trend has been reported worldwide.²³

In conclusion, our data underscore the issue of intrafamilial spread of KSHV infection among family members of CKS patients. Our findings may also apply, although to a lesser extent, to KSHV-seropositive individuals who have not developed the disease. Exposure to the saliva of infected individuals, especially in the familial setting, may be a major mode of horizontal KSHV spread. Nevertheless, assuming that KSHV, like Epstein-Barr virus, reactivates periodically in infected individuals and is shed in the saliva, and given that infection with KSHV could occur years before disease onset, it is reasonable to conclude that, although we have found an increased risk for virus infection among family members of CKS patients, the transmission of KSHV is inefficient. This may also explain the nonubiquitous nature of KSHV infection in most populations.

Correspondence: Ronit Sarid, PhD, Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel 52900 (saridr@mail.biu.ac.il).

Accepted for Publication: March 10, 2005.

Author Contributions: Drs Guttman-Yassky, Kra-Oz, Dubnov, Barchana, Bergman, and Sarid had full access to all data and take full responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Guttman-Yassky, Kra-Oz, Friedman-Birnbaum, Linn, Silbermann, Barchana, Bergman, and Sarid. Acquisition of data: Guttman-Yassky, Kra-Oz, Zaltzman, Roth, Schwartz, Rozenman, David, Barchana, Bergman, and Sarid. Analysis and interpretation of data: Guttman-Yassky, Kra-Oz, Dubnov, Segal, Linn, Silbermann, Barchana, Bergman, and Sarid. Drafting of the manuscript: Guttman-Yassky, Dubnov, Bergman, and Sarid. Critical revision of the manuscript for important intellectual content: Friedman-Birnbaum, Rozenman, David, and Silbermann. Statistical analysis: Dubnov, Segal, Linn, and Barchana. Obtained funding: Guttman-Yassky, Silbermann, Bergman, and Sarid. Administrative, technical, and material support: Zaltzman, Roth, Schwartz, Rozenman, David, and Barchana.

Financial Disclosure: None.

Funding/Support: This study was supported in part by the Israel Cancer Association, Givatayim, Israel, and the Middle East Cancer Consortium, Haifa, Israel.

Role of the Sponsors: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; or in the preparation, review, or approval of the manuscript.

Previous Presentations: This study was presented in part at the Fifth International Workshop on Kaposi’s Sarcoma–Associated Herpesvirus (KSHV/HHV-8) and Related Agents; August, 9, 2002; Kloster Irsee, Germany; and the 62nd Meeting of the American Academy of Dermatology; February 9, 2004; Washington, DC.

Acknowledgment: We thank Patrick Moore, MD, MPH, and Scott Laney, MPH, of the University of Pittsburgh, Pittsburgh, Pa, for critically reading the manuscript and providing comments.