Lack of serologic association of human herpesvirus-8 (KSHV) in patients with monoclonal gammopathy of undetermined significance with and without progression to multiple myeloma

After Rettig et al1 presented evidence that human herpesvirus (HHV)-8 DNA is present in the bone marrow dendritic cells of patients with multiple myeloma (MM), other reports quickly followed confirming2-10 or refuting the association of HHV-8 with MM, using polymerase chain reaction (PCR) techniques with various HHV-8 primers.11-16 In addition, a few reports also showed that MM patients lack antibody to HHV-8.11-14,16-20 Gao et al,21however, did find HHV-8 IgG antibody in MM patients by latent nuclear antigen (LNA) immunoblot in 81% of the cases. The HHV-8 antibody data in this report were not convincing enough to establish a serologic relation of MM with HHV-8.14 MM patients often have antibodies to either γ-herpesvirus (Epstein-Barr virus [EBV]) or β-herpesviruses (cytomegalovirus, HHV-6, HHV-7), which supports the hypothesis that MM patients have no defect in their immunologic syste ms.11-13,16,17 19 

One fourth of patients with monoclonal gammopathy of undetermined significance (MGUS) develop MM or related serious B-cell disorders. Because Rettig et al1 suggested that HHV-8 (also called KSHV [Kaposi sarcoma–associated herpesvirus]) may be required for the transformation from MGUS to MM and for perpetuation of the growth of malignant plasma cells, we tested sera from 362 MGUS patients and 110 MM patients. To make sure our serologic assays were able to detect accurately IgG antibody to HHV-8, we used a whole-virus enzyme-linked immunosorbent assay (ELISA), latent antinuclear antigen immunofluorescence assay (LANA IFA), and an IFA to lytic antigens. To see whether MGUS patients were immunocompetent, we also tested study sera for the presence of EBV virus capsid antigen (VCA) antibody.

MGUS sera were collected and frozen at −20°C by Drs M. B. Rettig, J. R. Berenson, and R. A. Kyle. Samples of MM, Kaposi sarcoma/human immunodeficiency virus-1 (KS/HIV), and classic KS were obtained frozen from Drs M. Kaplan, A. Friedman-Kien, D. Viza, and P. Gill. Donor sera were from a well-characterized panel used at Advanced Biotechnologies Inc (Columbia, MD). All sera were screened for antibodies to lytic proteins by the whole-virus lysate ELISA kit (Advanced Biotechnologies Inc).22 Positive sera were confirmed by IFA for lytic HHV-8 antigens.22 D. Bourboulia did the testing for IgG antibody to LANA by the IFA method of Simpson et al23 in the laboratory of Dr C. Boshoff in London. The EBV-VCA ELISA was done using native gp125 as a source of antigen.22 All samples were coded before testing.

Table 1 shows the seroprevalence of antibodies to HHV-8 lytic proteins for MGUS, MM, KS, and healthy controls. Twenty-six of 362 MGUS sera (7.1%), 6 of 110 MM sera (5.4%), 20 of 20 KS sera (100%), and 6 of 102 sera of healthy blood donors (5.9%) were positive by ELISA for IgG antibodies to HHV-8 antigen. For confirmation, the positive samples were retested by IFA, and all tested positive except for 2 of the MGUS sera and 2 of the donor sera (Table 2). These results show that the prevalence of antibody to lytic HHV-8 proteins in MGUS and MM patients is similar to that observed in the healthy donor population (Table 1). To ensure that we were not missing patients who had antibody only to latent proteins, we also used the LANA IFA to test randomly picked MGUS sera (95), MM sera (65), KS sera (20), and donor sera (50). For the samples tested, 3 MGUS (3.2%), 3 MM (4.6%), 18 KS (90%), and 2 healthy donors (4.0%) were LANA positive. The prevalence of antibody to the LANA in MGUS and MM is even lower than that of antibody to lytic proteins (Table 1). Analysis of HHV-8 antibody status in MGUS patients with and without progression to MM indicated no change in seroprevalence (Table 3). This argues against any role for HHV-8 in the transformation or pathogenicity of MGUS to MM.

These data are consistent with other antibody studies done largely on MM populations showing a lack of epidemiologic evidence for HHV-8 involvement in MM.11,12,16,17,19,20 Chauhan et al5 reported the analysis of 53 MM sera and found no antibody to latent or lytic proteins of HHV-8; they did, however, find HHV-8 DNA by PCR in the MM tissues. Whitby et al17 also failed to detect HHV-8 antibody in 4 MGUS patients who subsequently developed overt MM. Two other MGUS sera in their study were antibody positive but the patients had no evidence of MM after 36 and 40 months, respectively. The only study reporting the presence of HHV-8 antibodies in MM patients is by Gao et al.21 Using an immunoblot method instead of IFA,16,19 they reported the detection of antibodies to ORF-64 and LNA antigens.21 With the use of IFA, one can observe the difference in staining patterns between nonspecific and specific reactivity, but with the immunoblot method, it is possible to score only positive and negative results. If any nonspecific reactivity survived the blocking step, it would appear on the immunoblot as a positive result. This difference in technique may account for the discrepancy in results with other researchers. Therefore, on the strength of our data and that of the other researchers mentioned earlier, we conclude that it is extremely doubtful that HHV-8 has a role in the progression to MM.

To establish that MGUS patients are competent to make antibody to herpesviruses, we also tested the same 95 MGUS sera for IgG antibody to EBV-VCA using a native EBV (gp125 protein) ELISA. Ninety-two of 95 MGUS sera tested (96.8%) were positive for EBV-VCA antibody, demonstrating that MGUS patients were immunocompetent and responded to EBV infection and other human herpesviruses, as shown previously in MM patients.12,17 19 Both EBV and HHV-8 are classified as human γ-herpesviruses.

The work done by Rettig et al1 was based largely on PCR HHV-8 DNA data, as opposed to the present report, which is a study of serologic antibodies. It is interesting to note that in studies done to compare the frequency of PCR data with that of serology,24,25 the percentage of HIV-1⁺/KS⁺ patients positive for HHV-8 DNA in peripheral blood mononuclear cells by PCR is much lower than the percentage found positive for HHV-8 IgG antibodies by IFA for lytic proteins. It is therefore curious that Rettig et al1 could find cells positive for HHV-8 DNA by PCR, but could not detect HHV-8 antibodies in these patients. Berenson and Vescio2attributed the lack of serologic responses to HHV-8 in MM to interpatient differences as well as to consistent changes in ORF-65 sequencing derived from MM as compared to primary effusion B-cell lymphoma and KS patients. They argued that because ORF-65 is responsible for a major part of the serologic response to HHV-8, deletion of a base pair likely results in a change in the protein product, and this may explain the lack of or low level of serologic response to HHV-8 in MM patients.2 Although such deletions may cause a change in the function of the protein, it is unlikely that they would cause much change in antigenic response.

In light of this mounting molecular evidence based on HHV-8 data and the fact that these same patients lack HHV-8 antibody, MM patients may carry another herpesvirus yet to be identified. This theoretical herpesvirus could share common sequences with HHV-8 but be immunologically distinct from HHV-8.22,26 In much the same way, HHV-8 and EBV share homologous DNA sequences,26 but lack serologic cross reactivity.22 

We thank Ms Ruth Weinroth, Ms Alice Long, and Ms Kelly Lehr for their assistance in the preparation of this report.