Absence of HTLV-Related Sequences in Skin Lesions and Peripheral Blood of Cutaneous T-Cell Lymphomas

TO THE EDITOR The involvement of retroviruses and more specifically of viruses belonging to the human T-cell lymphotropic virus (HTLV) family in cutaneous T-cell lymphomas’ (CTCL) pathomechanisms remains a fiercely debated issue. Indeed, this hypothesis has remained attractive owing to the presence of a number of similarities between subsets of HTLV-1-associated adult T-cell leukemia and erythrodermic forms of CTCL. Overall, studies conducted in Europe have been generally negative (Bazarbachi et al., 1993, 1997) with the notable exception of a report from Italy, which claimed the isolation of a new retrovirus distantly related to HTLV-1 (then designated HTLV-V) from a continuous cell line derived from a patient with CD4þ Tac CTCL/leukemia as well as in other patients with CTCL (Manzari et al., 1987). This alleged breakthrough has not been confirmed by subsequent studies, including further investigations conducted by the same team. Another, more recent, report identified HTLV-1 tax-like sequences in blood and saliva from Russian CTCL patients, and established that these sequences were indeed expressed up to the protein level (Morozov et al., 2005). In addition, a significant proportion of USA-based investigations seemed to find out precise clues regarding the presence of retroviral agents close to HTLV-1 in skin lesions and/or peripheral blood in CTCL, and some authors have considered it reasonable to conclude that mycosis fungoides/ Sezary syndrome was an HTLV-associated disease (Hall et al., 1991; Pancake et al., 1995; Khan et al., 1996) even though other studies originating from the same geographical area have remained negative, as have been most of European studies (Wood et al., 1997). The discrepancy in the results and concepts developed in these reports is puzzling, and the fact that these different studies have been conducted with heterogeneous tools makes it difficult to apply strict comparisons. Furthermore, virtually all tools used in these studies, of which most of them have been conducted more than 10 years ago, were either poorly specific (serological tests or identification of viral particles, for instance), or were instead specifically targeting HTLV-1 and notably the tax sequence, thereby significantly reducing the chances to uncover even slightly divergent retroviruses. To overcome this difficulty, we used a recently described powerful semi-nested DNA amplification method (Kim et al., 2006), allowing the amplification of a sequence from the HTLV envelope gene located in the envelope receptor-binding domain (Kim et al., 2004). This PCR amplification of the highly variable envelope region, as Abbreviations: CTCL, cutaneous T-cell lymphoma; HTLV, human T-cell lymphotropic virus; PTLV, primate T-cell lymphotropic virus