Adult T-Cell Leukemia-Lymphoma Presenting Concurrently with Myelopathy Open Access

Human T-cell leukemia virus type 1 (HTLV-1) is an oncogenic retrovirus that is endemic in southern Japan, the Caribbean, South America, parts of Africa, and Iran [1]. Approximately ten to twenty million individuals are infected worldwide. Approximately 5% of infected individuals will develop adult T-cell leukemia/lymphoma (ATLL), and 0.25–4% will develop HTLV-associated myelopathy/tropical spastic paresis (HAM/TSP). Although most viral gene expression is silenced when ATLL develops, the HTLV-1 helix basic zipper (HBZ) gene is continuously expressed at all stages of infection [2]. Here, we describe a novel case of a patient presenting concurrently with CD30+ ATLL and HAM/TSP, who responded to brentuximab-vedotin chemotherapy regimen. The diagnosis was confirmed with a novel HBZ RNAscope assay.

A 51-year-old African-American woman, born in the Midwest with no relevant travel history, with a past medical history of diet-controlled diabetes mellitus type 2, alcohol abuse, pancreatitis, anemia, osteoarthritis, gastric ulcers, and gastroesophageal reflux disease, presented with a 1-month history of progressive and profound lower extremity weakness and mild numbness. She reported inability to walk and severe back pain, bowel and bladder incontinence, 20 pound weight loss, drenching night sweats, fevers, chills, chest pain, and dyspnea. Laboratory studies were remarkable for a lactate dehydrogenase value of 507 U/L (ULN 250 U/L), as well as normal thyroid studies, vitamin B12 level, and negative infectious studies including syphilis and HIV, except positive results for antibodies to HTLV-1 (Table 1). Cerebrospinal fluid (CSF) showed 79 mg/dL protein and 42 nucleated cell (94% mature lymphocytes) but negative cytology and negative paraneoplastic antibody studies. HTLV-1 proviral load (PVL) in the CSF was 0.09 copies/mononuclear cell (Table 2). HTLV-1 antibodies were also detected in the CSF (Table 1).

Chest computerized tomography revealed a 9.4 × 4.2-cm right hilar mass extending from the mediastinum to the right and left paratracheal and subcarinal regions, obliterating the right upper and middle lobe bronchi, which was fluorodeoxyglucose (FDG) positive on positron emission tomography (PET) (Fig. 1a). Hypermetabolic supraclavicular and mesenteric nodes were also noted. An MRI of cervical, thoracic, and lumbar spine showed a faint area of T2/STIR hyper-intense signal within the cervical cord at C2-C4, with additional possible areas of involvement at T1-T3, concerning for a demyelinating condition. The lower extremity paralysis and bowel and bladder incontinence were ascribed to HTLV-1-associated myelopathy (HAM).

Bronchoscopic biopsy of the paratracheal lymph node showed a diffuse infiltrate of atypical medium to large CD3+CD4+CD30+ T-cells, negative for CD1, CD10, anaplastic lymphoma kinase-1 (ALK-1), and Epstein-Barr virus encoded RNA-1 and near complete loss of CD5 and CD7, with rare nonneoplastic CD20 cells (Fig. 2a–e). The atypical T cells had irregular nuclear contours and high nuclear-to-cytoplasmic ratio, prominent nucleoli, increased apoptosis, mitotic figures, and foci of necrosis, resulting in the diagnosis of Lugano stage IIIB ATLL, lymphoma subtype [3].

Her initial course was complicated by deep venous thrombosis, pulmonary embolus, and cardiogenic shock, requiring an intravenous filter and anticoagulation. Electromyography showed severe loss of compound muscle action potentials and absent sensory nerve action potentials in the legs. She was treated with six cycles of brentuximab-vedotin, cyclophosphamide, doxorubicin, and prednisone (BV-CHP), complicated by febrile neutropenia in the first cycle, Escherichia coli urosepsis during the fifth cycle, and pneumonia during the sixth cycle. Following chemotherapy, she continued to have urinary tract infections and oral candidiasis and lower limb paralysis. Repeat PET scan showed significant improvement in mediastinal lymphadenopathy with no residual FDG avidity but new multifocal lesions in the cerebrum and cerebellum (Fig. 1b). However, a brain MRI showed no lesions and no leptomeningeal enhancement.

The diagnosis was confirmed by RNAscope showing foci of HBZ RNA in the patient’s lymphoma cells (Fig. 2f). Negative control results were obtained with a nonspecific RNAscope probe (E. coli dapB) (Fig. 2g), and positive control results were obtained with a RNAscope probe (human PPIB) directed against a constitutively expressed cellular RNA (Fig. 2h).

Immunodeficient NSG-Kit mice were inoculated with a short-term culture of peripheral blood mononuclear cells (PBMCs) from the patient, which had a PVL of 0.76 copies/mL (Table 2). After 8 and 9 weeks, respectively, the mice experienced weight loss and dehydration and were sacrificed. Necropsy revealed massive hepatosplenomegaly, tumor infiltration into the lungs, and enlarged brachial and inguinal nodes. Both mice had elevated leukocyte counts (10.3 K/mm³ and 15.2 K/mm³). Blood smears revealed abnormal large and small lymphocytes with convoluted nuclei (Fig. 3a–h). FACS analysis revealed human CD4+CD45+ cells in the spleen, lung, liver, bone marrow, and peripheral blood, with a mixture of single and double CD4- and CD8-positive cells (Fig. 3i–k). Sequence analyses of PDX lymphoma tissue showed identical polymorphisms to those present in the patient (i.e., lymphoma was patient-derived) and oligoclonal integration of the HTLV-1 provirus (Fig. 3l) with oligoclonal indices of 0.41 and 0.43. On a scale of 0–1, 0 represents polyclonal integration, and 1.0 represents monoclonal integration. The recurrent CARD11 R423W substitution described by Kataoka et al. [4] was present in PDX-derived cells (2.4% and 4.1%) but only found in 0.08% of reads in the patient’s PBMC-derived DNA.

Consistent with the clinical course of lymphoma, the HTLV-1 PVL in the peripheral blood decreased after treatment from 0.18 to 0.06 (Table 2). However, after activation and expansion in culture ex vivo in the presence of anti-CD3/CD28 and IL-2, the PVL increased significantly to 0.76 and increased even more dramatically in the splenocytes of the PDX mice (4.49 and 10.34).

The patient was seropositive for HTLV-1 and presented concurrently with ATLL and HAM/TSP. Although she did not have spasticity, which is typically associated with HAM/TSP, she did have the characteristic progressive weakness of bilateral lower extremities and urinary disturbances [5]. CSF analyses for malignancy, paraproteins, and biomarkers of multiple sclerosis were negative in this patient. The patient’s PET scan after completion of chemotherapy showed new multifocal lesions in the cerebrum and cerebellum, which are more consistent with HAM/TSP than cerebral lymphoma, in light of the negative brain MRI findings. Other less likely considerations for her paraplegia were nutritional deficiency, a paraneoplastic syndrome, myelitis from another virus, and amyotrophic lateral sclerosis, but laboratory studies were negative for these disorders.

This case is notable because HAM/TSP rarely presents concurrently with ATLL, most likely due to the fact that each disorder develops in only 0.25–5% of individuals at some time in their lifetime. The current therapy for ATLL with glucocorticoid and cytotoxic agents may have therapeutic activity for HAM/TSP, but our patient did not show improvement in her progressive myelopathy after chemotherapy treatment. In contrast, a therapeutic antibody against CCR4, mogamulizumab, has reported activities in both ATLL and HAM/TSP [6]. However, this agent was not available at the time of presentation of our patient.

Previous case reports described instances of ATLL developing after HAM/TSP diagnosis, with very few showing HAM/TSP after or concurrent with ATLL presentation [7]. Takeda et al. [7] described a 5-year prospective cohort study of 527 HAM/TSP patients and described ATLL prevalence and incidence in this cohort of 3.0% and 3.8 cases per 1,000-person-years of follow-up, respectively. Genomic analysis revealed somatic mutations of HTLV-1-infected cells in 90% of these individuals, and 11% of them had dominant clones and developed ATLL during the observation period [4, 7]. Similar observations were made in a prospective cohort of HTLV-1-infected patients, including individuals with HAM/TSP, who subsequently developed ATLL [8].

The patient described in the current report received a regimen of BV-CHP. Brentuximab-vedotin is an antibody-drug conjugate directed against CD30. She was given BV-CHP rather than cyclophosphamide, doxorubicin, vincristine, and prednisone based on the results from the ECHELON-2 clinical trial [9]. This was a randomized, placebo-controlled trial comparing BV-CHP with standard-of-care cyclophosphamide, doxorubicin, vincristine, and prednisone in CD30+ mature peripheral T-cell lymphoma in 452 eligible subjects, resulting in median progression-free survival of 48 versus 21 months, febrile neutropenia in 18% versus 15%, and peripheral neuropathy in 52% versus 55%, respectively. The multicenter trial only had four cases of CD30+ ATLL out of 226 patients receiving BV-CHP, so there is limited evidence of the therapy in this context. Two ongoing studies in refractory T-cell lymphomas include ATLL patients [10].

The patient’s posttreatment PET scan showed complete resolution of mediastinal, supraclavicular, and mesenteric lymphadenopathy with no residual FDG avidity, indicating successful anti-lymphoma therapy. Due to the patient’s resistance to hospitalization, she was not a candidate for allogeneic hematopoietic cell transplantation. Zidovudine and interferon alpha are often used concurrently with chemotherapy [11]. However, since it is unclear whether or not this improves response rates or overall survival and likely increases the toxicity of the regimen, it was not considered for this patient. Although zidovudine and interferon alpha are active in the acute subtype of ATLL, there is very little activity in the lymphoma subtype [12].

The PDX mouse studies provide support for the diagnosis of ATLL (Fig. 3). The finding of “chronic lymphocytic leukemia (CLL)-type” ATLL has been described [13], with small cells, no nucleoli, and a clumped chromatin pattern. CLL-type ATLL is thought to be a disorder of immature lymphocytes with a less aggressive course and a chronic disease. The patient did not have characteristic ATLL “flower” cells in the peripheral blood nor did she have leukemia. However, lymphoma cells circulating in the peripheral blood of this patient, expanded in culture and in NSG mouse xenografts, revealed small cells with highly lobulated nuclei similar to CLL-type ATLL, and carried highly elevated HTLV-1 PVLs. The PVL is increased with severity in HAM/TSP and is high in patients with rapid progression [14]. A PVL of 0.18 copies/PBMC at clinical presentation is consistent with a highly aggressive HAM/TSP.

The diagnosis of ATLL is determined by both clinical and laboratory features. HTLV-1 serology is a required test. To detect monoclonal integration of the HTLV-1 provirus genome, Southern blot analysis or inverse PCR may be done. However, these methods are time-consuming. ATLL is often misdiagnosed as various other mature T-cell malignancies, including mycosis fungoides and Sezary syndrome, cerebriform variants of T-cell prolymphocytic leukemia, peripheral T-cell lymphoma, unspecified (PTCL-U), angioimmunoblastic T-cell lymphoma, or Hodgkin’s lymphoma. PTCL-U and ATLL are usually difficult to differentiate due to similar morphology, and thus, the presence of HTLV in the cells is essential to differentiate the two disorders.

The HTLV-1 provirus genome consists of the gag, pol, and env genes. The TAX protein is an important viral factor in the pathogenesis of ATLL, but recent studies have reported that TAX expression is undetectable in about 60% of ATLL cases. Additionally, the frequency of transcription of the tax gene is very low and could only be detected by RT-PCR, not Western blot analysis. Although TAX expression decreases, another important HTLV-1 protein, HBZ, is expressed throughout the course of disease. A previous study showed that HBZ transcript detection by in situ hybridization of formalin-fixed, paraffin-embedded patient biopsy sections is an accurate, reliable way to diagnose ATLL by confirming the presence of the virus in the malignant cells [15]. HBZ RNA expression was detected by RNAscope in tumor cells. Although this method does not detect HTLV clonality, it does not require cells to be cultured and thus provides an easy, useful histological diagnostic test for ATLL, especially in sero-indeterminate patients or in cases with atypical pathological or clinical features. Takatori et al. [15] described the identification of HTLV-1-infected cells in 17 of 18 ATLL cases. When combined with the use of HTLV-1 PCR assays, they found 100% sensitivity in distinguishing ATLL from non-ATLL cases in HTLV-1 carriers [15].

This case report and data from a PDX model and HBZ RNAscope of the patient’s samples demonstrate methods to improve diagnosis and outcomes by making clinicians and pathologists more aware of ATLL. Moreover, response to brentuximab-vedotin in this CD30+ ATLL mediastinal tumor highlights a new approach to therapy. The co-occurrence of HAM/TSP represents an interesting and unusual feature in this patient.

Written informed consent was obtained from the patient for publication of this case report and any accompanying images. This study protocol was reviewed and approved by the Institutional Review Board of Washington University School of Medicine, approval number 201104376.

The authors have no conflicts of interest to declare.

This study was funded by Public Health Service grants from the National Cancer Institute. Funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Sneha Poondru and Lee Ratner wrote the manuscript, and Ancy Joseph, John C. Harding, Hemalatha Sundaramoorthi, Neha Mehta-Shah, Patrick Green, Anjun Hassan, and Daniel A. Rauch performed experiments or provided patient management. Sneha Poondru, Ancy Joseph, John C. Harding, Hemalatha Sundaramoorth, Neha Mehta-Shah, Patrick Green, Anjum Hassan, Daniel A. Rauch, and Lee Ratner approved the manuscript.

This study includes no supplementary material. All data that support the findings of this study are included in this article. Further inquiries can be directed to the corresponding author.

Sneha Poondru and Ancy Joseph contributed equally to this work.