Effect of anti-CCR4 monoclonal antibody (mogamulizumab) on adult T-cell leukemia–lymphoma: Cutaneous adverse reactions may predict the prognosis

Authors


Correspondence: Kentaro Yonekura, M.D., Ph.D., Department of Dermatology, Imamura Bun-in Hospital, 11-23 Kamoikeshinmachi, Kagoshima 890-0064, Japan. Email: ke.yonekura@jiaikai.jp

Abstract

Adult T-cell leukemia–lymphoma (ATL) is one of the most malignant lymphomas with poor prognosis. ATL cells express CC chemokine receptor 4 (CCR4) and mogamulizumab, a monoclonal antibody against CCR4 that exhibits very strong cytotoxicity for ATL cells via antibody-dependent cellular cytotoxicity. Although its effect is dramatic in ATL, serious adverse reactions such as Stevens–Johnson syndrome have been reported. However, these eruptions can appear as therapeutic signs of mogamulizumab. We evaluated the effectiveness of mogamulizumab in five acute-type ATL patients. Peripheral blood (PB) and lymph nodes (LN) were affected in three and four patients, respectively. In PB, complete response (CR) was obtained in all three patients and partial response (PR) was recorded in LN of one patient. In skin lesions, four of five patients manifested CR; in two, the lesions worsened after the start of mogamulizumab treatment and subsequently improved. In these lesions, CD4+825+ ATL cells were replaced by CD3+8+ cytotoxic T cells. Cutaneous adverse reactions (CAR) developed in two patients with CR; they did not show a relapse of ATL over the course of 9 months. Our findings suggest that mogamulizumab should be continued and surface marker evaluation should be performed even in patients whose skin lesions show aggravation, and that CAR may be a marker for a favorable prognosis.

Introduction

Chemokines are a group of small polypeptides that play fundamental roles in the immune system by regulating the cell trafficking of leukocytes and lymphocytes. CC chemokine receptor 4 (CCR4) is a receptor for CC chemokines including macrophage inflammatory protein-1, regulated and normal T-cell expressed and secreted, thymus and activation regulated chemokine and monocyte chemotactic protein-1; it is expressed on the surface of regulatory T cells (Treg) cells.[1, 2] CCR4 is often expressed on the tumor cells of T-cell lymphomas such as mycosis fungoides, peripheral T-cell lymphoma, not otherwise specified, and adult T-cell leukemia–lymphoma (ATL).[3] ATL is one of the most malignant T-cell neoplasms; its prognosis is poor and combinations of various chemotherapeutic agents have yielded only limited success.[4]

The novel defucosylated humanized monoclonal antibody against CCR4, mogamulizumab, has been shown to be highly cytotoxic for ATL cells via antibody-dependent cellular cytotoxicity (ADCC).[5] In a phase II study, it exhibited strong effects on ATL. The development by many patients of cutaneous adverse reactions (CAR) was suggested to indicate an association between CAR and the prognosis.[6] Herein, we report the results of our clinical trial designed to assess the effectiveness of mogamulizumab in patients with acute-type ATL paying particular attention to CAR, and present detailed histological findings.

Methods

Patients

This study was approved by the human research ethics committee of our hospital and prior written informed consent was obtained from all participants. Patients of at least 20 years of age with CCR4-positive acute type ATL with cutaneous manifestations that relapsed after at least one prior chemotherapy were eligible. We enrolled five patients, three men and two women, who started treatment with mogamulizumab at our hospital between May 2012 and March 2013. Their age ranged 57–73 years (median, 63). Their diagnosis, treatments, laboratory studies, therapeutic results and CAR were obtained from their clinical records. Peripheral blood (PB) and lymph nodes (LN) were affected in three and four patients, respectively and the central nervous system (CNS) was involved in one patient. The patients' clinical profiles are shown in Table 1.

Table 1. Characteristics of patients with acute-type adult T-cell leukemia–lymphoma
Case no.Age/sexPrevious treatmentATL lesions at the start of mogamulizumab treatment
  1. ATL, adult T-cell leukemia–lymphoma; CDE-11, CBDCA, dexamethasone, ETP, CPT-11; CHOP, cyclophosphamide, doxorubicin, vincristine, prednisolone; CNS, central nervous system; ETP, etoposide; mLSG-15, modified LSG-15 protocol (VCAP [vincristine, cyclophosphamide, doxorubicin, prednisolone], AMP [vincristine, cyclophosphamide, doxorubicin, prednisolone], VECP [vindesine, etoposide, carboplatin, prednisolone]);[4] LN, lymph node; mEPOCH, etoposide, prednisolone, vincristine, carboplatin, doxorubicin; PB, peripheral blood; THP-COP, pirarubicin, cyclophosphamide, vincristine, prednisolone.

162/MCHOP, mLSG-15, CDE-11PB, skin, LN
273/FTHP-COP, ETPPB, skin, LN
357/MmLSG-15, CDE-11, mEPOCHSkin, LN
463/FCHOP, mLSG-15PB, skin, CNS
564/MCHOPSkin, LN

Mogamulizumab treatment

Mogamulizumab (1 mg/kg) was administrated by i.v. infusion once a week for 8 consecutive weeks. All patients received antihistamine and acetaminophen p.o. before each infusion.

Assessment of clinical symptoms

The response of the cutaneous lesions was evaluated by physician global assessment, taking into consideration erythema, induration, papules and nodules. Responses to mogamulizumab were recorded as complete response (CR), partial response (PR), stable disease (SD) or progressive disease (PD) according to the criteria of the Japan Clinical Oncology Group. CR was defined as the disappearance of all clinical and radiographic evidence of disease and the normalization of lactate dehydrogenase (LDH) for at least 4 weeks. As human T-lymphotropic virus type-1 carriers frequently harbor a small percentage of abnormal lymphocytes, CR was recorded when less than 5% of such cells were present if the absolute lymphocyte count was less than 4 × 109/L. PR was defined as a reduction in measurable disease indices by at least 50% with a more than 75% reduction in the absolute abnormal lymphocyte count for at least 4 weeks without the development of new lesions or disease progression. We also required that LDH was decreased to less than 1.5 of the normal upper limit. PD was defined by a 25% or more increase in measurable disease indices or the appearance of new lesions during treatment. Lastly, SD was defined by a response between PR and PD.[7] Objective responses were assessed independently on photographs by three dermatologists after the 4th and 8th infusion session.

Histopathological and immunohistochemical findings

Biopsy specimens were obtained from ATL skin lesions before and after mogamulizumab treatment and in the transiently aggravated phase. Lesions that newly developed after treatment were also biopsied. ATL cells were defined by positive reactions to CD4 and CD25. CD3-, CD8- and CD56 positivity was used to define pan-T cells, cytotoxic T lymphocytes and natural killer cells. The presence of the cytotoxic molecules granzyme B, perforin and TIA-1 was also recorded. Immunohistochemical studies were performed using the VECTASTAIN Elite ABC standard kit (Vector Laboratories, Burlingame, CA, USA).

Results

Response of clinical symptoms to mogamulizumab

As shown in Tables 2 and 3, CR of their skin lesions was obtained in four of five patients. In two patients (cases 2 and 3) the lesions worsened after the start of mogamulizumab treatment and subsequently improved. All three patients with PB lesions (cases 1, 2, 4) experienced CR and one patient with LN lesions (case 2) manifested PR. On the other hand, the CNS symptoms in case 4 failed to respond.

Table 2. Effect and cutaneous adverse reactions elicited by mogamulizumab
Case no.Treatment cyclesBest response after treatmentInfusion reactionCutaneous adverse reactionEarly ATL relapse
PBSkinLN
  1. CR, complete response; LN, lymph node; NA, not assessable; PB, peripheral blood; PD, progressive disease.; PR, partial response; SD, stable disease.

14CRCRSD++
28CRCRPR++
37CRSDNone+
48CRCR++
54SDPDNoneNA
Table 3. Clinical course during treatment with mogamulizumabThumbnail image of

While ATL cells in the PB disappeared soon after the start of mogamulizumab treatment, the response of cutaneous lesions was relatively slow. These lesions disappeared in 3–4 weeks in two patients (cases 1 and 4). In another two patients (cases 2 and 3), these lesions manifested aggravation at 4–5 weeks after the start of treatment, subsequently subsided, and disappeared at 7–8 weeks with the continuous administration of mogamulizumab (Figs 1, 2).

Figure 1.

Case 2. (a) Before treatment with mogamulizumab. There were many cutaneous nodules on the patient's thigh. (b) After three cycles of treatment. The cutaneous lesions appear to have worsened clinically. (c) After eight cycles of treatment. The cutaneous lesions disappeared with the continuous administration of mogamulizumab.

Figure 2.

Case 3. (a) Several cutaneous tumors were seen on the patient's chest before treatment with mogamulizumab. (b) They appear to have enlarged after the second cycle of treatment. (c) While the lesions subsided and disappeared after the seventh cycle, (d) erythroderma with many papules developed at this time. As this was thought to be attributable to mogamulizumab treatment was discontinued.

Adverse events

Three patients developed an infusion reaction, two had CAR, one patient (case 3) presented with long-lasting erythroderma (Fig. 2d), and one (case 4) with generalized exudative erythema. In case 3, the skin rash appeared 1 week after the seventh infusion of mogamulizumab and prolonged for 8 months despite the treatment of 0.5 mg/kg oral prednisolone. The skin erythema of case 4 appeared 3 weeks after the fifth infusion and gradually subsided in a few weeks by the topical steroid treatment.

Relapse of ATL after mogamulizumab therapy

Two patients (cases 1 and 2) suffered ATL relapse, developed new skin lesions, and manifested a few abnormal cells in PB at 3 and 4 weeks after the discontinuation of mogamulizumab, respectively. Two patients (cases 3 and 4) experienced no ATL relapse in the skin or PB for nine and more than 10 months, respectively. Two patients (cases 1 and 2) without CAR suffered a relapse of ATL, while another two (cases 3 and 4) with CAR had no relapse in the course of more than 9 months. One patient (case 5) failed to respond to mogamulizumab and had no CAR.

Histopathological examinations

We subjected samples of skin lesions that showed CR after transient worsening (case 2) to histopathological study. Samples obtained before treatment showed dense infiltration of CD4+825+ atypical lymphocytes in the upper to middle dermis (Fig. 3a). In the early aggravation phase, the number of these atypical infiltrates decreased and prominent edema was seen in the dermis. Immunohistochemical studies showed a decrease in the number of CD4+25+ ATL cells and the appearance of many CD3+8+ cells positive for granzyme B, perforin and TIA-1 (Fig. 3b). We observed no CD56+ cells (data not shown).

Figure 3.

Case 2. (a) A skin biopsy specimen from the adult T-cell leukemia–lymphoma skin lesion before treatment with mogamulizumab showed dense infiltration of atypical lymphocytes in the upper to middle dermis. Immunohistochemically, these lymphocytes were positive for CD4 and CD25, but negative for CD8. (b) A skin biopsy specimen from a clinically worsened lesion after three cycles of treatment revealed decreased numbers of atypical lymphocytes and prominent edema in the dermis. Immunohistochemically, the number of CD4+25+ cells was lower. Many CD3+8+ lymphocytes positive for granzyme B, perforin and TIA-1 appeared (original magnifications ×200).

A biopsy specimen taken from an area of CAR in case 3 showed spongiosis in the epidermis, liquefactive degeneration of basal cells and heavy lymphocyte infiltration in the upper dermis. These cells were CD3+8+ lymphocytes positive for granzyme B, perforin and TIA-1. CD4+25+ cells were sparse and CD56+ cells were absent (Fig. 4).

Figure 4.

Case 3. A skin biopsy specimen obtained after the seventh cycle showed liquefactive degeneration of basal cells, spongiosis in the epidermis and heavy lymphocyte infiltration in the dermis. The lymphocytes were positive for CD3, CD8, granzyme B, perforin and TIA-1. There were few residual CD4+25+ adult T-cell leukemia–lymphoma cells (original magnifications ×200).

Discussion

Mogamulizumab, a newly developed defucosylated humanized monoclonal antibody against CCR4, exerts potent antitumor effects on ATL cells through ADCC.[5] Because approximately 90% of ATL cells express CCR4 on their surface,[3] mogamulizumab is expected to be highly useful for the treatment of ATL.

A multicenter single-arm phase II trial that included 28 patients showed that mogamulizumab exerted clinically meaningful antitumor activity in patients with relapsed ATL and that its toxicity profile was acceptable, findings suggestive of an association between CAR and the prognosis.[6] In 63% of patients, CAR appeared after the 4th or subsequent infusion. Of the 14 patients who had grade 2 or higher skin rashes, 13 manifested a clinical response; eight patients experienced CR. On the other hand, no objective responses were observed in 12 patients with no or grade 1 skin rashes. In addition, all of the patients who had grade 2 or higher skin rashes received five or more infusion of mogamulizumab, therefore, association between skin rashes and the number of mogamulizumab infusion was suggested.

Based on these earlier observations, we paid particular attention to CAR in our series. Mogamulizumab was remarkably effective against ATL cells in skin lesions and PB but less so with respect to LN or CNS. The clinical course of two of our patients (cases 2 and 3) in response to mogamulizumab appears to be of significance. At the start of mogamulizumab treatment, their ATL skin lesions became worse for 3 or 4 weeks. Histopathological evaluation revealed that this was attributable to dermal edema and the infiltration by CD3+8+ cytotoxic T lymphocytes that replaced CD4+825+ ATL cells. We continued mogamulizumab and the lesions had almost completely disappeared after the 7th or 8th infusion. Histopathologically, the CD3+8+ lymphocytes expressed various cytotoxic molecules such as granzyme B, perforin and TIA-1 indicative of enhanced antitumor immunity. We posit that these cytotoxic cells are promoted by the reduction of CD4+25high Treg cells because CCR4 is expressed on the surface of Treg as well as ATL cells. With this in mind, we suggest that mogamulizumab exhibits its clinical effectiveness not only via ADCC but also by inducing cytotoxic T lymphocytes and that its administration should be continued when histological study shows that CD3+8+ lymphocytes are increased in lesions that are exacerbated after the start of mogamulizumab treatment.

Two of our patients who experienced CR without CAR suffered relapse a few weeks after treatment. On the other hand, during more than 9 months of follow up, neither patient who experienced CR despite CAR relapsed. Our findings coincide with an earlier phase II study[6] and suggest that CAR are a favorable sign for a good prognosis. On the other hand, some patients treated with mogamulizumab developed severe Stevens–Johnson syndrome.[8] However, the differentiation between prognostically favorable CAR and severe systemic drug eruptions such as Stevens–Johnson syndrome is very difficult. The pathophysiological mechanism(s) underlying CAR remain to be elucidated but in the meantime, in the clinical settings, skin biopsies must be evaluated in patients treated with mogamulizumab.

Conflict of interest

A. U. received honoraria from Kyowa Hakko Kirin for his works.

Ancillary