Dear Editor,

The CD30 antigen is a member of the tumour necrosis factor receptor superfamily, expressed on activated T cells and B cells. In lymphoma, CD30 is expressed on the neoplastic cells in Hodgkin lymphoma, anaplastic large cell lymphoma, and some cases of mature T cell lymphoma and B cell lymphoma. Therapeutic targeting of CD30 by the anti-CD30 antibody conjugate brentuximab vedotin has been shown to be highly effective in Hodgkin lymphoma [1], anaplastic large cell lymphoma [2], and selected cases of mature T cell lymphoma [3] and cutaneous T cell lymphoma [4, 5]. Interestingly, although expression of CD30 is apparently necessary for brentuximab vedotin to be effective, the response does not appear to be directly proportional to the actual expression level.

A 17-year-old woman presented with abdominal distension. Computed tomography of the abdomen showed a mass around the common bile duct. A biopsy showed large areas of necrosis, with sheets of large lymphoma cells exhibiting angiocentricity (Fig. 1a). Neoplastic cells were positive for CD2, cytoplasmic CD3, CD8, and the cytotoxic marker TIA1, but negative for CD4 and CD56. Some cells were positive for T cell receptor alpha/beta. CD30 was expressed in about 30 % of cells (Fig. 1b). In situ hybridization showed the presence of EBV-encoded RNA (EBER) in the neoplastic cells. Marrow trephine biopsy showed EBER-positive atypical lymphoid cells, indicating lymphoma involvement. Positron emission tomography-computed tomography (PET-CT) showed hypermetabolic lesions in the lungs and supraclavicular lymph node. Plasma EBV DNA was assayed by quantitative polymerase chain reaction (Q-PCR) [6] and was elevated to 2.4 × 103 IU/mL (reference range: undetectable). A diagnosis of NK/T cell lymphoma was made according to World Health Organization classification criteria.

Fig. 1
figure 1

Complete remission of refractory NK/T cell lymphoma to brentuximab vedotin and bendamustine. a Angiocentricity in the lymphoma, with neoplastic cells infiltrating the blood vessel wall (haematoxylin eosin, original magnification ×400). b CD30 was positive in about 30 % of neoplastic cells, some of which were arranged in an angiocentric pattern (immunoperoxidase, original magnification ×400). c Positron emission tomography at relapse, showing multiple hypermetabolic foci (arrows) at nodal and extranodal sites, as well as the spleen. d Positron emission tomography after three courses of brentuximab vedotin and bendamustine, showing complete resolution of all hypermetabolic foci

Full size image

She was treated with a protocol comprising gemcitabine (750 mg/m2/day, days 1 and 15), cisplatin (25 mg/m2/day, days 1–3), ifosfamide (1200 mg/m2/day, with MESNA, 1200 mg/m2/day, days 2–4), etoposide (100 mg/m2/day, days 2–4), dexamethasone (40 mg/day, days 2–4), and l-asparaginase (6000 units/m2/day, days 8, 10, 12, 14, 16, and 20). A complete remission (CR) radiologically (interim PET/CT showing no lesions) and molecularly (undetectable plasma EBV DNA) was achieved after two courses. However, following two additional cycles of chemotherapy, the lymphoma relapsed, with new hypermetabolic lymphadenopathy above and below the diaphragm shown on PET/CT (Fig. 1c), and resurgence of plasma EBV DNA to 8.4 × 104 IU/mL.

Salvage treatment with brentuximab vedotin (1.8 mg/kg, day 1) and bendamustine (90 mg/m2, days 1, 2) every 3 weeks was administered. She achieved a metabolic CR on PET/CT after three cycles of treatment (Fig. 1d), although plasma EBV DNA was still detectable at 2 × 102 IU/mL. Therapy-related toxicity was mainly haematological (grade III neutropenia and grade II thrombocytopenia), and with supportive treatment, no dose adjustments were necessary. On confirmation of CR, she underwent consolidation with a haploidentical haematopoietic stem cell transplantation (HSCT) from her father, as a matched donor was not available. Fludarabine, cyclophosphamide, and total body irradiation were used as conditioning, and cyclophosphamide, tacrolimus, and mycophenolate mofetil were used as prophylaxis for graft-versus-host disease. A PET/CT performed 1 month post-HSCT showed continued metabolic CR. Serial Q-PCR showed undetectable plasma EBV DNA, indicating molecular CR. At the latest follow-up 5 months post-HSCT, she has remained asymptomatic.

CD30 expression has been reported to occur in about 40 % of NK/T cell lymphoma (with 5 % positivity as cutoff) [7], suggesting that brentuximab vedotin may be useful. The efficacy of single agent brentuximab vedotin is variable in mature T cell lymphomas, being better in cutaneous T cell lymphomas (overall response rate, ORR 70–73 %, CR 35 %) [4, 5] than in other mature T cell lymphomas (ORR 41 %, CR 23 %) [3]. In an isolated report of one case, brentuximab vedotin achieved merely a transient response in a patient with refractory NK/T cell lymphoma, which was complicated by dose-limiting dyspnoea [8].

Refractory NK/T cell lymphoma has a dismal outcome. In this patient, we adopted a combination approach with brentuximab vedotin and bendamustine for several reasons. Brentuximab vedotin as a single agent induces low CR rates in non-cutaneous T cell lymphomas, suggesting that combination approaches may be needed for better responses. Furthermore, the safety of brentuximab vedotin when combined with anthracycline-containing regimens has previously been demonstrated [9]. Bendamustine is an attractive partner, as it has activity in about 50 % of patients with relapsed/refractory T cell lymphoma [10]. Our case achieved CR with combined brentuximab vedotin and bendamustine without serious toxicity. Hence, in NK/T cell lymphomas with CD30 expression, brentuximab vedotin and bendamustine may be an active combination that should be further evaluated.