Kaposi's sarcoma-associated herpesvirus is not only associated with Kaposi's sarcoma but with tumours such as multicentric Castleman's disease and body cavity based lymphomas ( Ganem, 1997). The incidence of infection of KSHV in lymphomas is very low, but recent data suggest a strong association between the virus and multiple myeloma ( Rettig et al, 1997 ; Said et al, 1997 ; Brousset et al, 1997 ). In these cases, KSHV might infect stromal (dendritic) cells of the bone marrow ( Rettig et al, 1997 ), but this result has not been reproduced by others ( Cull et al, 1998 ). Because of the failure to amplify DNA sequences of KSHV from marrow aspirates or mononuclear cells, the investigation of bone marrow core biopsies with PCR and in situ hybridization has been proposed ( Rettig et al, 1997 ; Said et al, 1997 ; Brousset et al, 1997 ). There are many similarities in the pathophysiology of multiple myeloma and Waldenstrom's macroglobulinaemia (WM) such as plasma cell differentiation and immunoglobulin production. It has been postulated that, in multiple myeloma, KSHV could play a pathogenetic role through the expression of its human IL-6 homologue (vIL6) ( Rettig et al, 1997 ; Said et al, 1997 ). A similar scenario may be proposed for the aetiology of Waldenstrom's macroglobulinaemia. To verify this hypothesis, we investigated DNA samples from 20 acetone-fixed and paraffin-embedded bone marrow biopsies with Waldenstrom's macroglobulinaemia, for the detection of Kaposi's sarcoma-associated herpesvirus (KSHV).
Kaposi's sarcoma-associated herpesvirus (KSHV) is suspected to play a role in the aetiology of multiple myeloma. Because of similarities in the pathophysiology of multiple myeloma and Waldenstrom's macroglobulinaemia (WM), we investigated DNA samples from 20 bone marrow biopsies with WM for the detection of KSHV by PCR (KS330/ORF26). We performed two rounds of amplification and found that only 1/20 of the DNA samples obtained from biopsies had a detectable KSHV sequence. The positive patient was also infected by the human immunodeficiency virus (HIV). Our data provide evidence that KSHV cannot be implicated in the pathogenesis of WM.
MATERIAL AND METHODS
Bone marrow biopsies of 20 patients with WM were retrieved from our files at the Pathology Laboratory, Purpan Hospital, Toulouse. All bone marrow biopsies had been acetone-fixed and paraffin-embedded according to the ModAMeX protocol ( Delsol et al, 1989 ) This method has been proved to be useful for DNA preservation. DNA was extracted from the biopsies after deparaffinization with standard protocols ( Galoin et al, 1996 ).
For the detection of KSHV we used the protocol described by Chang et al (1994 ). The primers (KS1, KS2) and the internal probe were identical and corresponded to the KS330233 sequence. We changed the protocol by performing two rounds of 30 cycles of PCR amplification. After the first round, the amplification products were diluted at 1:100 before reamplification. The amplification products were visualized on 3% agarose gel after ethidium bromide staining under UV illumination and subsequently transferred onto a nylon membrane for hybridization with the 32P-labelled internal probe as described ( Chang et al, 1994 ). The positive control consisted of a case of Kaposi's sarcoma processed by the same method. In all cases the integrity of the DNA was checked by amplifying a 258 bp fragment corresponding to the c-raf-1 gene ( Galoin et al, 1996 ).
KSHV detection in bone marrow biopsies
KSHV DNA sequences were present in only one out of the 20 biopsies (5%) included in this study (Fig 1). The patient infected with KSHV was also HIV-positive. The PCR result was obtained after two rounds (30 × 2 cycles) of amplification. After the first round the positive case gave a clear band, at 233 bp after ethidium bromide staining. After the two rounds of PCR the positive band was of strong intensity and was specifically hybridized with the internal probe. The negative cases were interpretable because the DNA extracted was of good quality. A single round of amplification yielded a clear 258 bp band with primers corresponding to the c-raf-1 gene. None of the other biopsies were positive for KSHV.
The possible association of KSHV with multiple myeloma represents significant information concerning the epidemiology and pathophysiology of this tumour. The founding paper by Rettig et al (1997 ) pointed to a close link between the virus and malignant plasma cell proliferation. In their report Rettig et al (1997 ) found that the virus infected the bone marrow dendritic cells. However, recent studies have provided conflicting results by demonstrating a negative serology for KSHV ( MacKenzie et al, 1997 ), together with the lack of confirmation of its presence in cultured dendritic cells of patients with multiple myeloma ( Cull et al, 1998 ).
Since we have been able to detect KSHV DNA sequences in bone marrow core biopsies of patients with multiple myeloma ( Brousset et al, 1997 ), we sought to determine the incidence of KSHV infection in patients with WM by PCR amplification of DNA samples obtained from bone marrow core biopsies. We found that 1/20 of the DNA samples obtained from biopsies had detectable KSHV sequences (KS330/ORF26). The positive patient was also infected by the human immunodeficiency virus (HIV) whereas the others were not (0/19). None of the control cases (17 bone marrow biopsies with follicular lymphoma, three with myelodysplastic syndromes, seven with reactive processes) had KSHV sequences. In our previous report ( Brousset et al, 1997 ) we found that KSHV was present in 90% of bone marrow biopsies from patients with multiple myeloma. Our method was found to be sensitive and the viral detection was achieved after two rounds of amplification. However, in some cases a positive amplification could be visualized on gel after only one round of 30 cycles. Despite these highly sensitive conditions for the PCR-assay we failed to detect KSHV in HIV-negative cases of WM. The data presented here provide evidence that KSHV cannot be implicated in the pathogenesis of WM at least in HIV-negative patients. However, the finding of the virus in the bone marrow of an AIDS patient may simply reflect the high incidence of KSHV infection in this setting.
This work was supported by the Délégation à la Recherche Clinique (DRC) and the Institut Universitaire de France.
We thank Professor S. M. Chittal, University of St Johns, Newfoundland, Canada, for critical reading of the manuscript.