Recurrence of an erythrodermic mycosis fungoides as a biologically indolent anaplastic large T‐cell lymphoma initially misdiagnosed as being of donor origin

A 50-year-old woman with a 2-year history of gluteal patches and plaques that did not improve with topical corticosteroids developed an erythrodermic rash involving more than 75% of her body surface (Figure 1A). A biopsy showed a dense lymphoid infiltrate in the upper dermis comprising intermediate-sized atypical lymphocytes with epidermotropism (Figure 1B) expressing CD3/CD4/CD30 (50% of them; Figure 1C). High-throughput sequencing studies of T-cell receptor (TCR) detected the clonal TCRGAMMA Vg3-Jg1/2 sequence (62% reads) and the V11-2/J1-4 TCRBETA sequence (26.5% reads). A diagnosis of cutaneous T-cell lymphoma (CTCL), mycosis fungoides (MF)-type, was made. Subsequent next-generation sequencing (NGS) studies with a customized panel detected four mutations at CSNK1A1-Ser27Phe (VAF = 44%, VUS), CCDN3-Leu84_His85ins47 (VAF = 7%, likely pathogenic), CDKN1B-Asn3Asp (VAF = 6%, VUS) and TRRAP-Ala1911Ser (VAF = 5%, VUS) genes. Enlarged axillary and inguinal lymph nodes were identified. Flow cytometry of peripheral blood revealed 5868 lymphoid cells/μL, CD3+/CD4++/CD7−/CD26− compatible with MF, T2bNxM0B2, and IVA2 stage (ISCLC and EORTC). The patient started treatment with methotrexate/UVB-phototherapy, which was ineffective. Treatment was changed to brentuximab therapy, without complete response. Eight months afterwards, allogeneic hematopoietic stem cell transplantation (HSCT) from her brother was performed. Four weeks after the HSCT, a diagnosis of acute cutaneous graft-versus-host disease Grade II was established (Figure 1D–F). She received high-dose systemic corticosteroids, which resolved the cutaneous lesions within a month. Twenty-six months after the HSCT, the patient developed a nodular cutaneous lesion on her right breast (Figure 1G). A dense non-epidermotropic lymphocytic infiltrate in the upper mid-dermis with no epidermotropism was found (Figure 1H). Neoplastic cells were small to medium-sized with hyperchromatic nuclei, a high nucleus: cytoplasm ratio, and no remarkable nucleolus, and there was an intralymphatic tumor (Figure 1I). All neoplastic cells were CD30/CD3/TCRBETA+. CD4+ tumoral lymphocytes were in the center of the infiltrate, while reactive CD8+ lymphocytes predominated at the tumor periphery (Figure 1J). PD1, TCRGAMMA, and in situ hybridization for the Epstein–Barr virus were negative. These findings prompted a diagnosis of posttransplant pcALCL. According to the new WHO Classification,1 lymphoid proliferations in the immunosuppressive setting should be classified as in non-immunodeficient patients adding its relation to viruses and highlighting the underlying cause of immunodeficiency. The patient underwent an excision of the cutaneous nodule without further systemic treatment, remaining disease-free after 13 months. FISH studies of X/Y chromosomes (two-colored probe) in the breast nodule showed numerous cells with red and green signals, corresponding to male/donor origin (XY karyotype cells), intermingled with other cells carrying double red signals of female/patient origin (Figure 1K,L). Recipient chimerism studies confirmed the 98% donor profile of the bone marrow cells analyzed by karyotype (XY cells). These findings suggested a new lymphoma of donor origin. Nevertheless, molecular studies identified the same clonal TCRGAMMA (Vg3-Jg1/2, AF = 24%) sequence as in the initial biopsy and a common somatic mutation in the CSNK1A1-Ser27Phe gene (36% reads). A new clonal TCRGAMMA sequence was identified (Vg10-Jg1/2, AF = 5%) and TCRBETA analysis identified two new clonal sequences at Vb6-1/Jb2-7 (AF = 22%) and Vb5-8/Jb2-3 (AF = 11%). These findings are consistent with previous studies that suggest that MF malignant transformation take place during early T-cell development and further alterations could harbor at different locations or stages of the disease. In line with this hypothesis, different MF lesions may share the same TCRGAMMA rearrangement and acquire different clonotypes. Moreover, the presence of the CSNK1A1-Ser27Phe somatic mutation with a VAF of 44% and 36% in both lesions before and after HSCT suggests it was a germline mutation. This data also supports a common clonal origin of both lesions acquiring further divergent evolution.2, 3 Presence of both X/X and X/Y cells, especially these last ones in CD8+ cells of the pcALCL lesion, suggest a robust T-cell response from the donor. We present a patient who suffered a recurrence of her erythrodermic MF as a biologically indolent pcALCL initially misdiagnosed as being of donor origin. Donor-origin T-cell lymphomas occur under two scenarios with different clinical features: when the transplant is from a T-cell lymphoma patient, in which the tumor is often fatal, or from a safe donor, which tends to be indolent. First reported is the case of two sisters, the donor and the recipient of HSCT for a previous systemic anaplastic large cell lymphoma (sALCL), who developed identical subcutaneous panniculitis-like T-cell lymphoma 3 years after the procedure, and the finding of identical T-cell clones in both lymphomas, different from those of sALCL implied transmission during HSCT.4 Other reports5, 6 highlighted posttransplant CTCL in recipients whose donors had a previous history of MF. Identical TCR clonal patterns detected in the recipient and donor were decisive in diagnosing them as posttransplant CTCL from the donor, suggesting malignant stem cell transfer within the allograft, and clinical outcomes were fatal.2 Conversely, other two cases of posttransplant CTCL following HSCT in a recipient previously treated for mantle cell lymphoma and a folliculotropic posttransplant MF included XY FISH studies that supported a conclusion of donor origin.7, 8 No further molecular studies were undertaken. None of the previously reported posttransplant CTCLs were aggressive, since one was a Stage IA and the other, Stage IVA that was responsive to brentuximab. Similar to these previous XY FISH reports,7, 8 the indolent behavior of the lesion, the discovery of many XY karyotype neoplastic cells and the 98% donor chimeric profile prompted us initially to consider the new lymphoma to be of donor origin, despite the donor being healthy. Nevertheless, pre/post-HSCT lymphomas shared a dominant TCRGAMMA clonal sequence at Vg3-Jg1/2 and the CSNK1A1 gene mutation. These molecular findings, the mixture of CD4+ and CD8+ lymphocytes present in the biopsy specimen, and the mixture of XY/XX karyotype cells finally led us to conclude that the posttransplant CTCL was a recurrence of her lymphoma, with an exuberant CD8+ antitumoral response of donor origin, rather than a new lymphoma of donor origin. Although HSCT is the only curative treatment available to date, nearly half of MF patients relapse and die.9, 10 Our clinical presentation and biological behavior of the pre/posttransplant neoplasms clearly differ: the former is an erythrodermic MF with systemic involvement, and the latter is a single localized lesion that disappeared after surgically excision. Clinical presentation of the latter as a large solitary nodule, consisting of a diffuse infiltrate with marked intralymphatic infiltration of CD30+ neoplastic cells, supported our diagnosis of pcALCL. Different clonally related clinical presentations of the same disease as MF, pcALCL, or LyP have been previously reported.11 MF cases with concomitant pcALCL lesions seem to share TCR clonotypes, while different MF lesions at different time points share a common clonotype but acquire additional ones as far as the disease progresses or presents at different locations.2, 3, 11 The prognosis for patients with pcALCL is very favorable, with some developing spontaneous remission. Whether the specific biology of these lesions was responsible for the nodule remaining localized or if it was the effect of the antitumoral response of donor-derived cells remains to be determined. We emphasize that not all posttransplant CTCLs from healthy donors are necessarily of donor origin since relapses of the original neoplasm, with the subsequent appearance of new clonotypes and a huge antitumoral response of donor origin, may occur. In this context, we suggest that in addition to XY FISH studies, TCR/NGS studies may be of interest. The authors declare no conflict of interest. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.