Lack of confirmation of an association between HTLV-I infection and myelodysplastic syndrome.

Abstract

Karlic et al (1997) recently reported the presence of antibodies to HTLV-I, as detected by enzyme and immunofluorescence assays, in the sera from 11/65 patients with myelodysplastic syndrome (MDS) in a non-endemic region of central Europe. The presence of specific HTLV-I pol and tax sequences was also documented by polymerase chain reaction (PCR) in the whole cell lysates from uncultured peripheral blood obtained from the same seropositive patients. HTLV-I infection was documented in all the subtypes of the French–American–British (FAB) Cooperative Group classification (refractory anaemia, RA; RA with excess of blasts, RAEB; RAEB in transformation, RAEBt; chronic myelomonocytic leukaemia, CMML) with the exception of RA with ringed sideroblasts (RARS). HTLV-I infection was found to be strictly associated with the occurrence of del(5q), which was detected in 5/8 HTLV-I-positive MDS cases, but in only 1/38 HTLV-I-negative cases. More recently, the same authors reported the detection of HTLV-I infection in 2/13 cases of acute myeloid leukaemia (AML) which had evolved from MDS, and documented the presence of del(5q) and monosomy 5 in the two cases, respectively (Karlic et al, 1998). We investigated the frequency of HTLV-I infection in a series of 39 MDS cases from another non-endemic area (Northern Italy) (Table I). DNA was extracted with phenol/chloroform methods from the whole cell lysates obtained from peripheral blood and bone marrow cells of all patients, and from long-term bone marrow cultures in eight cases. DNA was tested by PCR using primers for the HTLV-I pol and tax genes, and the amplified products were hybridized with a specific oligonucleotide probe, using the same protocols described by Bazarbachi et al (1997). The sensitivity of the PCR assay was assessed by successive 10-fold dilutions of the positive control DNA (HUT-102 and MT-2 cell lines) in an HTLV-I-negative DNA (MOLT-4 cell line), showing that amplification was obtained at up to one HTLV-I-positive cell in 1 × 105 cells. None of the samples tested revealed positivity for HTLV-I tax and pol sequences (Table I), not even in four patients carrying del(5q). The involvement of HTLV-1 infection in human diseases in non-endemic areas is not clear. The presence of deleted forms of HTLV-I genomes in HTLV-I seronegative patients with mycosis fungoides and Sezary syndrome, initially reported in a few studies, has not been confirmed (Bazarbachi et al, 1997). The presence of antibodies to the specific BA21 epitope of the envelope protein of HTLV-I/II in patients with large granular lymphocyte leukaemia is due to a cross-reactivity with a protein with homology to viral BA21, in the absence of prototypical HTLV-I/II infection (Loughran et al, 1997). The reported detection of HTLV-I antigens, reverse transcriptase (RT) activity and HTLV-I-like retroviral particles in the pathologic samples from AML patients (Xu et al, 1996) needs confirmation, since it is well known that, for example, RT activity itself may arise from mycoplasma infection and cellular DNA polymerase activity and human endogenous retroviruses may have cross-reactivity to HTLV-I antigens and have RT activity. Relevant to this, Boyd et al (1997) initially reported the detection of retrovirus-like particles and RT activity in platelet lysates from patients with myeloproliferative diseases (essential thrombocythaemia and polycythaemia vera) but further discovered that these particles package RNA molecules that encode human endogenous retroviral (HERV-K)-related pol genes. In conclusion, much caution is needed in establishing an association between exogenous retroviruses and human disease. Our study provides evidence against a role for HTLV-I infection in MDS, although the presence of a retrovirus closely related to HTLV-1, or an as yet unidentified unrelated virus, in MDS remains a possibility. This work was funded by a grant from Associazione Italiana per la Ricerca sul Cancro (A.I.R.C.), Milan, Italy. We thank Professor T. F. Schulz, Department of Medical Microbiology, University of Liverpool, U.K., for helpful discussion.