Absence of Human T-Lymphotropic Virus Type I in Japanese Patients With Cutaneous T-Cell Lymphoma

CUTANEOUS T-cell lymphoma (CTCL) is a disease entity characterized by a primary sporadic T-cell proliferation in the skin, which includes mycosis fungoides (MF ), Sezary syndrome (SS), and other peripheral T-cell lymphomas primarily occurred in the skin.1 Clinical appearance of CTCL varies from indurated patches to multiple tumors of the skin but limitation to the skin is common.2 The disease progression is indolent and dissemination of the tumor cells to the internal organ means a terminal event.2,3 Morphologically, the tumor cells are large but usually uniform in size and their nuclei show cerebriform, expressing CD4⁺/helper phenotype.3 In addition to these characteristics detection of the DNA rearrangement of the T-cell receptor without the monoclonal integration of human T-lymphotropic virus type 1 (HTLV-1) genome is essential in the diagnosis. Patients with CTCL are seronegative for HTLV-1 except CTCL patients in HTLV-1 carrier,3a and CTCL is considered to be sporadic.2 3 

HTLV-1 is a type of retrovirus first detected in human tumors,4 which causes adult T-cell leukemia/lymphoma (ATLL), a neoplasm of the mature T-lymphocyte.5 Patients with ATLL are seropositive for HTLV-1 and often develop a rapid dissemination to the internal organs from the early-stages.6 Skin manifestation of ATLL constitutes papules or multiple skin nodules, but sometimes show indurated erythema or erythroderma as seen in CTCL. The tumor cells, expressing CD4⁺/suppressor phenotype, are heterogenous in size and their nuclei have irregular contours showing flower-like arrangements.5 The prognosis of ATLL is very poor.6 

Involvement of the skin characterizes both disorders, making it difficult to differentiate between them. Furthermore, several authors have recently shown the HTLV-1 genome in the genomic DNA from patients with CTCL and proposed a causal relation between HTLV-1 and CTCL,7-12 which causes more confusions in the diagnosis as well as the meaning of HTLV-1 in these two disorders. However, in these studies there were some problems in the diagnosis of CTCL as well as in the interpretation of the laboratory data as follows: (1) The diagnoses of CTCL was not always differentiated from chronic or smouldering ATLL,12-14 (2) HTLV-1 carriers (seropositive for HTLV-1) were included in some ploymerase chain reaction (PCR) studies,12 (3) Conditions (stringency) of PCR varied,11 (4) Samples employed in these studies varied including fresh lesional skin, peripheral blood mononuclear cells (PBMC) and cultured cell lines,7-12 (5) Non-T–cell lines established from patients with CTCL were used to detect HTLV-1.7,9,15 16 

In an endemic area of ATLL like Japan diagnoses of CTCL are very critical because the treatment as well as the prognosis of ATLL are quite different from those of CTCL. Thus, in this study, we examined 128 DNA specimens of fresh lesional skin, PBMC, and T-lymphocyte clones derived from 50 Japanese patients with CTCL, who showed typical clinical and histopathological features and had no serum HTLV-1 antibodies to clarify the relationship between HTLV-1 and CTCL in Japan.

Selection of patients.We have selected 50 patients with CTCL, in which clinical features, pathological findings and DNA analyses were all typical for CTCL and serum anti-HTLV-1 antibodies were negative in both particle agglutination and enzyme-linked immunosorbent assay (ELISA) methods. The patients included 28 men and 22 women and the ages varied from 24 to 88 (mean; 55.9). The diagnoses of the patients included erythema stage of MF (n = 9), plaque stage of MF (n = 9), tumor stage of MF (n = 7), SS (n = 4), primary cutaneous anaplastic large cell lymphoma of T-cell type expressing Ki-1 (CD30) antigen (Ki-1⁺CTCL) (n = 4), CTCL other than MF, SS, or Ki-1⁺CTCL (n = 17). A case of smouldering ATLL in which monoclonal integrations of HTLV-1 genome were proved by Southern blot analysis and two cases of serum anti-HTLV-1 antibody positive CTCL in which monoclonal integrations of HTLV-1 genome were not detected (CTCLs in HTLV-1 carrier),3a were employed as positive controls.

Cell culture.The fresh skin specimens from the patients with CTCL were incubated in RPMI medium 1640 containing 20% fetal bovine serum and interleukin-2 after an initial 72-hour stimulation with phytohemagglutinin. When the cultured cells once constantly grew in the medium, phenotypic analyses were done by using flow cytometry with anti-CD4 monoclonal antibody. If the cultured cells express CD4 on their surface, genomic DNAs were derived for the subsequent PCR.

DNA preparation.Genomic DNAs were derived from the fresh lesional skin and PBMC of the patients with CTCL. Briefly, after phenol chloroform extraction the DNA was precipitated with ethanol and redissolved in sterile H₂O. DNA concentration was estimated by absorption at 260 nm.

PCR amplification.Amplification of the DNA was performed in a total volume of 50 μL in a reaction mixture containing 225 μmol/L each of deoxyadenosine triphosphate, deoxycytidine triphosphate, deoxyguanosine triphosphate, and deoxythymidine triphosphate, 50 to 125 pmol each of primer, 50 mmol/L KCl, 2.5 mmol/L MgCl₂ , and 10 mmol/L tris-HCl (pH 8.3), and 2 U of Taq polymerase (Takara, Shiga, Japan). Solutions were covered with mineral oil to prevent condensation. Fifty cycles of denaturation 2 minutes at 94°C, primer anealing for 1 minute at 60°C, and chain elongation for 30 seconds at 72°C were performed, using reagents from Takara Gene Amp reagents kit and the Perkin Elmer Thermal Cycler (Perkin Elmer Cetus, Norwalk, CT). The primers used for the amplification of HTLV-1 are shown in Table 1.

T-cell cultures were established in 4 cases of erythema stage MF, 5 of plaque stage MF, 5 of tumor stage MF, 3 of SS, 4 of Ki-1⁺ CTCL, and 13 of CTCL other than MF, SS, or Ki-1⁺ CTCL. Totally, 128 DNA samples were taken from 50 lesional skin, 44 PBMC and 34 cultured T-lymphocyte clones. The results are summarized in Table 2. No samples revealed positive for HTLV-1 genome. Two CTCLs in HTLV-1 carrier revealed positive with all four sets of primers and a smouldering ATLL revealed strongly positive with all sets of primers examined (Fig 1).

In Japan, where has the highest prevalence of HTLV-1 infection in the world, antibodies against HTLV-1 have been detected in 1.2 million people (most of them are carriers of HTLV-1), which are estimated at approximately 1.0% of all Japanese population, and newly more than 700 cases of ATLL are diagnosed each year.6 Thus, the presence of serum HTLV-1 antibodies is not diagnostic for ATLL and it is essential to confirm the monoclonal integration of the HTLV-1 DNA by using Southern blot analysis17 or inverse PCR,¹⁸ a newly developed method to recognize the monoclonality of HTLV-1 under the detection level of Southern blotting. In practice, such borderline cases as a smouldering ATLL showing indolent MF-like course (our no. 51), HTLV-1 negative ATLL,19 and CTCLs accompanied by serum anti-HTLV-1 antibodies (our nos. 52 and 53)3a are present.

In lymphoproliferative disorders except ATLL the seronegativity for HTLV-1 is common. However, more sensitive method such as ELISA has recently enabled to show positivity for HTLV-1 up to 30% in various leukemia/lymphoma patients including MF.20,21 In certain regions of the United States HTLV-1 seroprevalence has been particularly high in prostitutes, homosexuals, and intravenous drug users.22,23 These tendencies are completely different from our backgrounds (patients who have serum anti-HTLV-1 antibodies are exceptional in our series), and they also suggest the higher detection rates of HTLV-1 genome in their PCR studies. As shown in Fig 1, it is reasonable that the PCR for HTLV-1 revealed positive in HTLV-1 carriers (but the positive bands were weaker than those in smouldering ATLL), which should not be included into these HTLV-1 detection studies. Moreover, it is no doubt that patients having monoclonal integrations of HTLV-1 should be diagnosed as ATLL.17 

In addition to these confusions in the diagnoses of ATLL and in the interpretation of serum anti-HTLV-1 antibodies, there are several problems in the positive HTLV-1 results with PCR in CTCL. First, technical errors including contamination can easily occur especially in PCR studies. Second, many positive bands can be detected if the stringency of PCR is low (data not shown). These results indicate the presence of endogenous HTLV-1–like sequences within human genomic DNA or the detection of other retroviral sequences similar to HTLV-1 (HTLV-5, etc). Moreover, the false positive results for HTLV-1 can be detected unless the complete sequencings for the PCR products are done.11 Third, deleted forms of HTLV-1 genome may exist in the genomic DNA of CTCL patients.8 If the HTLV-1 genome was inserted into the host DNA as a deleted form, it is reasonable that the defective HTLV-1 genome neither induce the host's antibody production nor the replication of the ATL cells. Lastly, in the positive studies7-12 most of the HTLV-1 DNA were proved not in the fresh specimens but in the cultured cells including non-T–cell lines.8,16 During the cultivating procedures not only clonal expansions but also modifications of the culturing cells can occur, thus the DNA derived from those cell lines does not exactly reflect the conditions in the original CTCL. Furthermore, HTLV-1 sequences have been detected even in squamous cell carcinoma of the skin,24 suggesting that an opportunistic or a coincidental infection of HTLV-1 should be considered.

We conclude that CTCL which does not include HTLV-1 is present although the pathogenesis of CTCL may be different by respective areas or races.