Graft‐Versus‐Tumour Effect Following Allogeneic Haematopoietic Stem Cell Transplantation in a Paediatric Patient With Mogamulizumab‐Treated Primary Cutaneous Peripheral T‐Cell Lymphoma

ABSTRACT

Introduction

Primary cutaneous peripheral T‐cell lymphoma, not otherwise specified (pcPTCL‐NOS), is exceptionally rare in children, and evidence guiding diagnosis and treatment is limited.

Case Presentation

A 10‐year‐old boy with pcPTCL‐NOS refractory to chemotherapy received mogamulizumab prior to allogeneic haematopoietic stem cell transplantation (allo‐HSCT). Despite a 60‐day washout period, he developed prolonged steroid‐refractory acute graft‐versus‐host disease requiring ruxolitinib and extracorporeal photopheresis. Concurrently, transient skin lesions developed at previous lymphoma sites and regressed spontaneously, with biopsy supporting the interpretation of visible graft‐versus‐lymphoma.

Conclusion

This case underscores the need to optimise allo‐HSCT timing after mogamulizumab and preserve graft‐versus‐lymphoma effect while limiting graft‐versus‐host disease.

Trial Registration: The authors have confirmed clinical trial registration is not needed for this submission.

1. Introduction

Peripheral T‐cell lymphomas are a rare and heterogeneous group of non‐Hodgkin lymphomas. Primary cutaneous peripheral T‐cell lymphoma, not otherwise specified (pcPTCL‐NOS), is an exceptionally rare subtype in children and lacks defining clinical and histopathological features. In contrast to other paediatric cutaneous T‐cell lymphomas (CTCLs) such as mycosis fungoides and lymphomatoid papulosis, which typically follow an indolent course, pcPTCL‐NOS often presents with advanced disease, aggressive behaviour and poor prognosis, necessitating intensive systemic therapy [1]. Diagnosis and management are challenged by rarity, clinical heterogeneity and limited understanding of disease biology, with responses to conventional chemotherapy frequently being short‐lived [2].

Achieving complete remission followed by consolidative allogeneic haematopoietic stem cell transplantation (allo‐HSCT) remains the only potentially curative treatment for advanced pcPTCL‐NOS, reflecting a graft‐versus‐lymphoma (GvL) effect mediated by donor T cells [3]. However, this benefit is counterbalanced by a substantial risk of treatment‐related toxicity, particularly acute graft‐versus‐host disease (aGvHD) [4].

In children, data on optimal treatment strategies and long‐term outcomes are extremely limited. Management is largely extrapolated from adult CTCL, where allo‐HSCT is infrequently pursued due to toxicity, or from protocols used in other paediatric lymphomas despite limited disease‐specific evidence [1, 3]. Consequently, treatment often involves multiple therapy lines before allo‐HSCT. Mogamulizumab, a CCR4‐targeting humanized IgG1 antibody approved for relapsed or refractory adult CTCL, improves response rates, although complete responses remain low [5].

Patients with CTCL appear particularly prone to aGvHD, with reported 100‐day incidences of Grade II–IV aGvHD of 50%–70%, compared with 30%–40% in other haematologic malignancies [6].

Here, we present a rare paediatric case of pcPTCL‐NOS in leukaemic phase illustrating the diagnostic challenges and the dual nature of donor T‐cell alloreactivity.

2. Case Presentation

A previously healthy 4‐year‐old boy developed multiple persistent cutaneous nodules on the upper extremities and face, later involving both ear helices (Figure 1A). At 6 years of age, a skin biopsy was interpreted as Rosai–Dorfman disease [7]. Baseline blood tests, imaging and abdominal ultrasound showed no systemic involvement, and skin‐directed therapy was continued.

FIGURE 1.

(A) At presentation, before diagnostic re‐evaluation. Progressive nodular skin lesions on cheek and ear helix. (B) Post‐TSEBT and prior to allo‐HSCT all skin lesions were in complete remission, here left ear.(C) Day +6 post‐HSCT. Flare‐up of lesion on the ear. (D) Day +38 post‐HSCT: Macular lesions on the back, clinically inconsistent with GvHD and predominantly localised to sites of prior lymphoma involvement. Biopsy demonstrated low‐level persistence of the diagnostic malignant clone by PCR (E) Day +63. Second GvHD episode involving > 80% of body surface area, primarily the upper body. The picture is taken 7 days after reintroducing steroid treatment and the day of initiation of ruxolitinib. Allo‐HSCT, allogeneic haematopoietic stem cell transplantation; GvHD, graft‐versus‐host disease; GvL, graft‐versus‐lymphoma; TSEBT, total skin electron beam therapy.

Over the next 18 months, the lesions slowly progressed despite treatment, accompanied by recurrent urticaria, severe pruritus and chronic erythema. At Age 9, accelerated growth of facial lesions prompted repeated biopsy, demonstrating a clonal T‐cell population with an aberrant immunophenotype (cyCD3⁺, surface CD3⁻, CD4⁺, variable CD7) in tumour tissue and peripheral blood. Molecular studies identified clonal TCRβ and TCRγ rearrangements and a balanced t(1;7)(q23;q34) translocation involving TCRB1, also detectable in blood. Whole‐genome sequencing identified no pathogenic somatic sequence variants or fusion genes. Germline analysis revealed no cancer predisposition, genodermatoses or immunodeficiency variants. Minimal residual disease (MRD) at diagnosis was 5.7% in peripheral blood and 0.39% in bone marrow by flow cytometry, with negative cerebrospinal fluid. PET‐CT showed FDG‐avid lesions in the cheek and ear helices without nodal or visceral involvement. Following multidisciplinary review, including external expert pathology consultation, the overall clinicopathological findings were considered most consistent with a mature primary CTCL despite the unusual t(1;7)(q23;q34), reflecting the prolonged cutaneous course with subsequent leukaemisation. The diagnosis was revised to pcPTCL‐NOS, Stage IVA (T3N0M0B2) according to the WHO–EORTC classification.

Given leukaemisation, first‐line treatment followed the ALLTogether‐1A high‐risk protocol but resulted in persistent MRD (blood 3.6%, marrow 1.0%, Figure 2A) [8]. Second‐line therapy following the ALCL‐99 protocol induced a transient MRD reduction (blood 0.98%) before relapse (blood 4.9%) [9]. Subsequent mogamulizumab (1 mg/kg) markedly reduced skin lesion and MRD (blood 0.02%, marrow 0.003%). However, after six cycles, new cutaneous lesions emerged with persistence of the original malignant T‐cell clone. Consolidation with total skin electron beam therapy (1 Gy × 10 fractions) resulted in complete cutaneous remission (Figure 1B), while MRD in blood continued to decline, indicating compartment‐specific disease activity. The last mogamulizumab dose was administered 60 days before allo‐HSCT.

FIGURE 2.

Disease course timeline. (A) Pre‐HSCT course showing therapeutic strategies and corresponding MRD values in peripheral blood (blue) and bone marrow (red). MRD values are shown by flow cytometry; however, there was good concordance with PCR‐based MRD, supporting the association between the aberrant immunophenotype and TCR clonality. (B) Post‐HSCT timeline highlighting the clinical course of acute GvHD and GvL effect on the skin and indicating immunosuppresive treatment used. Numbers indicate post‐HSCT days. A2G, ALLTogether protocol; EBT, electron beam therapy; ECP, extracorporal photophorese; GI, gastrointestinal; GvHD, graft‐versus‐host disease; GvL, graft‐versus‐lymphoma; HR‐in, high risk induction; HSCT, haematopoietic stem cell transplantation; MRD, minimal residual disease.

Allo‐HSCT was performed using an HLA‐identical sibling donor after a myeloablative conditioning with total body irradiation (2 Gy × 6) and etoposide (60 mg/kg). GvHD prophylaxis included cyclosporine, switched to tacrolimus on Day +20 to ensure compliance. Neutrophil engraftment occurred on Day +14.

The post‐HSCT course was complicated by two episodes of aGvHD. The first began on Day +10 reaching Stage 3 skin aGvHD, overall Grade II, and resolved with methylprednisolone. The second episode started on Day +52 with extensive cutaneous and mild gastrointestinal involvement, progressing despite increased calcineurin inhibition and showing poor response to steroids (Figure 1E). Additional treatment with ruxolitinib and extracorporeal photopheresis (ECP) was initiated, with gradual disease control, although recurrent flare‐ups persisted until Day +270 and, hence, very gradual tapering of immunosuppressants and a total of 31 series of ECP (Figure 2B).

In addition to the aGvHD manifestations, clinical findings interpreted as GvL activity were observed in the skin. From Day +6, prior lymphoma lesions briefly reappeared before resolving by Day +9. On Day +38, during tapering of steroids after the first aGvHD episode, new macular lesions appeared on the back and left cheek, overlapping prior lymphoma sites (Figure 1C,D). Clinically, these lesions were inconsistent with aGvHD, and biopsy demonstrated low‐level persistence of the malignant clone by PCR. MRD in blood was 0.001% at Day +42 and detectable below the quantitative range at Day +57, with all subsequent tests being MRD negative. Donor chimerism was 99.6% at Day +42 and > 99.95% by Day +77. These findings were considered consistent with a GvL reaction, although mogamulizumab‐associated rash cannot be excluded. The lesions regressed spontaneously by Day +48, preceding the second aGvHD episode (Figure 2B).

At Day +230 post‐HSCT, regulatory T‐cell (Treg) counts were reduced (0.006 × 10⁹/L), proportionate to CD4⁺ T‐cell lymphopenia (CD4⁺ 0.048 × 10⁹/L). At last follow‐up, 1‐year post‐HSCT, the patient had no active GvHD and was off all immunosuppressive therapy, including ruxolitinib. MRD remained negative with complete donor chimerism. The immune reconstitution remained delayed with T‐cell counts (CD3⁺ 0.43 × 10⁹/L; CD4⁺ 0.21 × 10⁹/L; CD8⁺ 0.17 × 10⁹/L), whereas Treg counts had normalised (0.031 × 10⁹/L).

3. Discussion

This case highlights key challenges in paediatric primary cutaneous peripheral T‐cell lymphoma, including delayed diagnosis, refractoriness to conventional therapy and severe aGvHD following allo‐HSCT. Notably, the case provides rare clinical evidence of concurrent GvL activity in the skin.

The patient's initial presentation at 4 years of age was misinterpreted as Rosai–Dorfman disease based on a skin biopsy [7], and the lack of pathognomonic clinical and histological features likely contributed to the prolonged diagnostic delay until age 9. At that time, the disease was already advanced with blood and bone marrow involvement, underscoring the challenges of timely diagnosis in paediatric CTCL, particularly in pcPTCL‐NOS.

Therapeutically, the disease was refractory to two conventional chemotherapy regimens, consistent with the limited chemosensitivity typically observed in pcPTCL‐NOS [10]. Allo‐HSCT remains the only potentially curative approach for advanced pcPTCL‐NOS. However, experience in the paediatric population is extremely limited, and current recommendations regarding transplantation timing, particularly following mogamulizumab, are based on retrospective adult studies [6, 11]. Although reduced‐intensity conditioning is generally preferred in adult CTCL, the optimal conditioning intensity in paediatric pcPTCL‐NOS is unknown. Given the patient's young age, excellent performance status, aggressive refractory disease and blood and bone marrow involvement, myeloablative conditioning was selected; however, its potential contribution to the subsequent severe aGvHD cannot be excluded.

Several factors may explain the increased GvHD risk in CTCL. First, the skin represents both the primary disease site and the main GvHD target organ, complicating the distinction between relapse and alloreactivity and potentially amplifying tissue damage. Second, the chronic immune activation and a pro‐inflammatory microenvironment of CTCL may prime donor lymphocytes toward exaggerated alloreactivity. Third, treatment‐related factors, particularly mogamulizumab exposure, play a crucial role. As an anti‐CCR4 antibody, mogamulizumab not only targets malignant T cells but also depletes Tregs, which normally suppress alloreactive responses [12]. Consequently, pre‐transplant mogamulizumab exposure has been strongly associated with increased incidence and severity of aGvHD, especially when administered close to HSCT [4, 11]. In our case, transplantation was performed 60 days after the final mogamulizumab dose, in accordance with recommendations of a minimum washout period of 50 days [13].

Nevertheless, the patient developed steroid‐refractory aGvHD requiring multiple lines of immunosuppressive therapy, including ruxolitinib and ECP. At the same time, the patient demonstrated a GvL effect in the skin. The overlap with prior lymphoma lesions and their spontaneous regression support donor‐derived immune activity against malignant T cells. Importantly, this visible GvL activity prompted a cautious approach during the second episode of aGvHD, slightly delaying systemic corticosteroids to preserve anti‐lymphoma immunity. This clinical dilemma—balancing the benefits of GvL with the risks of uncontrolled aGvHD—represents a central therapeutic challenge in allo‐HSCT for CTCL.

The delayed immune reconstitution likely reflected prior mogamulizumab exposure and intensive steroid treatment. Tacrolimus was withdrawn as soon as clinically feasible in attempt to support immune reconstitution.

To our knowledge, no previous reports have described GvL activity so directly visible in the skin. Franke et al. previously demonstrated a haematologically confirmed graft‐versus‐Sézary effect in a 54‐year‐old patient undergoing non‐myeloablative allo‐HSCT [14].

4. Conclusion

Our case highlights the delicate balance between therapeutic benefit and treatment‐related toxicity in allo‐HSCT for CTCL. Incorporating emerging biomarkers such as ST2 and Reg3α into early risk stratification may help tailor prophylaxis and interventions [15]. Future studies should define the optimal timing of allo‐HSCT after mogamulizumab and identify strategies that preserve GvL activity while limiting GvHD.

Author Contributions

P.P., L.H., S.M., H.V.M. and L.D.S. contributed to diagnosis and pre‐HSCT treatment planning. M.I., H.H.U., T.M., M.E.S., S.M., H.V.M., L.D.S and L.S. contributed to HSCT management and treatment of treatment‐related complications. P.P., M.H., H.H.U., T.M., M.I. and M.E.S. were involved in clinical patient care. M.E.S. drafted the manuscript. T.M., H.H.U., H.V.M., L.D.S., L.H., L.S., M.E.S., M.I. and P.P. contributed essential clinical data and its interpretation. All authors critically revised the manuscript for important intellectual content and approved the final version.

Funding

The authors have nothing to report.

Ethics Statement

According to national regulations, ethical approval was not required for this case report. The study was conducted in accordance with the principles of the Declaration of Helsinki.

Consent

Written informed consent was obtained from the patient's parents for publication of this case report and accompanying images.

Conflicts of Interest

The authors declare no conflicts of interest.

Sørum M. E., Masmas T., Uhlving H., et al. “Graft‐Versus‐Tumour Effect Following Allogeneic Haematopoietic Stem Cell Transplantation in a Paediatric Patient With Mogamulizumab‐Treated Primary Cutaneous Peripheral T‐Cell Lymphoma.” eJHaem 7, no. 5 (2026): e70403. 10.1002/jha2.70403

Data Availability Statement

The data are available upon request.