Reduction of Regulatory T Cells by Mogamulizumab, a Defucosylated

Published OnlineFirst November 5, 2014; DOI: 10.1158/1078-0432.CCR-14-0830

Cancer Therapy: Clinical Clinical Cancer Research Reduction of Regulatory T Cells by Mogamulizumab, a Defucosylated Anti-CC Chemokine Receptor 4 Antibody, in Patients with Aggressive/Refractory Mycosis Fungoides and Sezary Syndrome Xiao Ni1, Jeffrey L. Jorgensen2, Meghali Goswami1, Pramoda Challagundla2, William K. Decker3, Youn H. Kim4, and Madeleine A. Duvic1

Abstract

Purpose: The CC chemokine receptor 4 (CCR4) is expressed on achieved a response in blood had high baseline CCR4 expres- malignant T cells in cutaneous T-cell lymphoma (CTCL) as well as sion on malignant T cells. Tregs in blood were 58.6% to 100% on regulatory T cells (Treg). When mogamulizumab, a defuco- positive for CCR4 at baseline and showed decreased numbers þ sylated monoclonal antibody, binds to CCR4, it induces anti- and CCR4 expression after treatment. CD8 T cells in blood þ body-dependent cellular cytotoxicity against CCR4 malignant T were 3.2% to 23.2% positive for CCR4 at baseline and showed cells. The goal of this study was to determine the effect of limited reduction of CCR4 expression with increased percen- þ þ mogamulizumab on CCR4 Tregs in patients with CTCL. tages of CD8 T cells after treatment. Of 14 patients tested for Experimental Design: Peripheral blood of 24 patients with NK cells in blood, 10 showed increased percentages after CTCL participating in a phase I/II trial was analyzed for CCR4 treatment. Four of 6 patients tested showed increased NK cell þ expression on different T-cell subsets by ﬂow cytometry, before cytotoxicity. Sixteen of 18 patients who had CCR4 lympho- and after one course of mogamulizumab. The number and cytes in baseline lesions showed decreased numbers after function of natural killer (NK) cells were also analyzed. Lesional treatment. þ biopsies were examined for CCR4, Foxp3, and CD16 expression Conclusions: Mogamulizumab reduces levels of CCR4 þ by immunohistochemistry. malignant T cells and also CCR4 Tregsinpatientswith Results: Malignant T cells in peripheral blood were 20.8%– CTCL, which may in turn improve immune proﬁles. Clin 100% positive for CCR4 at baseline. Fourteen patients who Cancer Res; 21(2); 1–12. 2014 AACR.

Introduction malignant T cells in the peripheral blood or with a clonal T-cell receptor gene rearrangement (1). Advanced stage or trans- Cutaneous T-cell lymphomas (CTCL) are characterized by formed mycosis fungoides and also Sezary syndrome are con- malignant clonal proliferation of skin-homing T cells. Mycosis sidered to be aggressive rather than indolent. Patients with fungoides and its leukemic form, Sezary syndrome are the most aggressive mycosis fungoides/Sezary syndrome have depressed common of CTCL variants. Mycosis fungoides often begins as cellular immunity, poor prognosis, and are often refractory to indolent skin patches or plaques, but may disseminate to treatment (2, 3). There is an unmet need for more effective lymph nodes, liver, spleen, lung, and blood at advanced stages. targeted therapies for aggressive/refractory mycosis fungoides/ Sezary syndrome is characterized by erythroderma covering at Sezary syndrome that are less immunosuppressive and can lease of 80% of the body surface area with at least 1,000/mL induce durable complete remissions. The CC chemokine receptor type 4 (CCR4) is a seven-trans- 1Department of Dermatology, The University of Texas MD Anderson membrane, G-protein–coupled receptor which is speciﬁcfor 2 Cancer Center, Houston,Texas. Department of Hematopathology,The the CC chemokines, macrophage-derived chemokine (MDC or University of Texas MD Anderson Cancer Center, Houston, Texas. 3Department of Stem Cell Transplantation, The University of Texas CCL22), and thymus and activation-regulated chemokine MD Anderson Cancer Center, Houston,Texas. 4Department of Derma- (TARC or CCL17). CCR4 is expressed on activated Th2 cells, tology, Stanford Cancer Center, Stanford, California and is critical for T-cell skin-homing (4, 5). CCR4 is also highly Note: Supplementary data for this article are available at Clinical Cancer expressed on malignant T cells in mycosis fungoides skin Research Online (http://clincancerres.aacrjournals.org/). lesions and on circulating malignant T cells in patients with Corresponding Authors: Madeleine Duvic, Department of Dermatology, The Sezary syndrome, making it an ideal molecule for targeted University of Texas MD Anderson Cancer Center, Unit 1452, 1515 Holcombe Blvd. therapy (5, 6). A defucosylated anti-CCR4 monoclonal anti- Houston, TX 77030. Phone: 713-745-4615; Fax: 713-745-3597; E-mail: body, KW-0671 or mogamulizumab, was developed to spe- þ [email protected]; and Xiao Ni, [email protected] ciﬁcally target CCR4 malignant T cells in T-cell lymphomas doi: 10.1158/1078-0432.CCR-14-0830 (7,8).BecauseofaknockoutoftheFUT8gene,thebackboneof 2014 American Association for Cancer Research. this antibody lacks fucose leading to increased antibody-

www.aacrjournals.org OF1

Ni et al.

mouse model (7). Therefore, we hypothesized that mogamu- Translational Relevance þ lizumab can not only reduce CCR4 malignant T cells but also þ The CC chemokine receptor 4 (CCR4) is expressed on deplete CCR4 Tregsinpatientswithmycosisfungoides/Sezary malignant T cells in cutaneous T-cell lymphoma (CTCL) as syndrome, and subsequently restore NK cell function. Twenty- well as on regulatory T cells (Treg). Mogamulizumab (KW- four patients with aggressive/refractory mycosis fungoides/ 0761), a defucosylated monoclonal antibody, can induce Sezary syndrome, participating in the phase I/II clinical trial þ antibody-dependent cellular cytotoxicity against CCR4 of mogamulizumab, were studied. Flow cytometry was used to malignant T cells when it binds to CCR4. Our translational analyze the expression of CCR4 on different T-cell subsets study accompanying a phase I/II trial found that mogamuli- before and after one treatment course. Foxp3 and CCR4 mRNA þ zumab not only eliminates CCR4 malignant T cells locally levels were assessed. The number and function of NK cells were and systemically following one course of therapy, but also also analyzed. Lesional biopsies were examined for CCR4, þ reduces CCR4 Tregsthat may be instrumentalfor restoring NK Foxp3, and CD16 expression. cell antitumor function in patients with CTCL. Thus, besides broad applications for treating patients with CTCL, mogamu- Materials and Methods lizumab therapy may have value in treating many other tumors Patients and study design with immunosuppressive mechanisms involving Tregs. This translational study was approved by MD Anderson Cancer Center and Stanford Institutional Review Boards and conducted according to the Declaration of Helsinki in parallel with the phase I/II clinical trial of mogamulizumab (Clinical- dependent cytotoxicity (ADCC) activity. This antibody binds to Trials.gov identifier: NCT00888927; M.A. Duvic; submitted for the N-terminus of CCR4 and has no neutralizing activity of publication). All patients signed a written consent. Patients ligand and no complement-dependent cytotoxicity. Strong were eligible for the phase I/II study if they had a histologi- ADCC against primary malignant T cells in vitro and antitumor cally/cytologically confirmed diagnosis of CTCL, and failed at activity was observed in an mycosis fungoides/Sezary syndrome least one prior systemic therapy. Fresh peripheral blood and mouse model (7, 8). lesional skin biopsies were collected at baseline and after the Regulatory T cells (Treg) are a subset of immunosuppressive T first treatment course. In the first treatment course, patients cells that also have high CCR4 expression (9). Tregs play a key role received weekly infusions of 0.1 or 0.3 or 1.0 mg/kg of in maintaining self-tolerance and modulating adaptive immune mogamulizumab for 4 weeks, followed by a 2-week observa- responses; they inhibit an excessive immune response during tion period. An early clinical response to mogamulizumab on inflammation by suppressing effector T cells. Increased numbers day 29 was assessed by one of two attending dermatologists of Tregs are correlated with poor prognosis in patients with (M.A. Duvic and Y.H. Kim). Blood, skin, lymph node, and tumors, suggesting that Tregs may weaken antitumor immune overall response were graded as complete response (CR), responses and foster immune privilege (10, 11). In Hodgkin þ þ partial response (PR), stable disease (SD), or progressive lymphoma, CCR4 Foxp3 Tregs created an environment in disease (PD) as previously described (21–23). which Hodgkin lymphoma cells could escape from host immunity (12). Therefore, depleting Tregs or inhibiting the function of Tregs in patients with cancer has been recently attempted to boost Blood and tissue specimens antitumor immunity (13). Monoclonal antibodies specific for cell Peripheral blood mononuclear cells (PBMC) were separated surface molecules predominantly expressed by Tregs are under and aliquoted for flow cytometry, total RNA extraction, and NK development (13). assays. Peripheral blood samples from normal donors (ND) were The role of Tregs in mycosis fungoides/Sezary syndrome is still obtained from the Department of Transfusion Medicine at our controversial because Sezary cells share many features with Tregs. institution. Skin lesion biopsies were fixed with 4% paraformal- Malignant T cells in patients with Sezary syndrome are anergic and dehyde and stored in 80% ethanol until paraffin embedding. Five immunosuppressive, they secrete IL10 and TGFb, and they may micrometer sections were consecutively cut for hematoxylin and also express Foxp3 (14, 15). However, these circulating malignant eosin (H&E) and immunohistochemical staining. T cells differ from classic Tregs in several aspects. Sezary cells have little to no expression of CD26 or/and CD7 (16), usually Flow cytometry analysis for different T-cell subsets, NK cells, have normal to low CD25 expression (17), and have normal IL7 and CCR4 expression receptor (CD127) expression. In contrast, classic Tregs have high In this study, we defined malignant T cells in peripheral blood þ þ CD25 expression and dim or absent CD127 (18). Therefore, it as CD3 CD4 CD26 and/or CD7 by flow cytometry (16). þ is still possible to distinguish classic Tregs with CD3 They additionally often showed alteration in levels of CD3 þ þ CD4 CD25high CD127dim/ phenotypes from malignant T cells and/or CD4 expression. Fresh peripheral blood and/or PBMCs in patients with mycosis fungoides/Sezary syndrome. at baseline and after one course of mogamulizumab were þ Mogamulizumab was approved in Japan for adult T-cell analyzed for malignant T cells, CCR4 malignant T cells, þ þ þ þ þ þ lymphoma (ATL; refs. 19, 20). In the United States, a phase CD3 CD4 CD25 CD127dim/ Tregs, CCR4 Tregs, CD3 CD8 þ þ þ þ þ I/II multicenter clinical trial was conducted in patients T cells, CCR4 CD3 CD8 T cells, and CD3 CD16 CD56 NK with CTCL with overall response rates of 37% for all patients cells by multicolor flow cytometry analysis using BD FACSCanto (21–24). A recent study reports that KM-2760, another defu- II cytometers (BD Biosciences). Cell surface expression of CCR4 þ cosylated anti-CCR4 antibody, was able to reduce the number (CD194) on malignant T cells, Tregs, and CD8 T cells was þ of tumor-infiltrating Foxp3 Tregs and increase the number of assessed by a commercial available antibody, 1G1, from BD þ tumor-infiltrating CD56 NKcellsinaHodgkinlymphoma Biosciences, Pharmingen. The percentage and absolute number

OF2 Clin Cancer Res; 21(2) January 15, 2015 Clinical Cancer Research

Effect of KW-0761 on Tregs and NK Cells in CTCL

of each cell population were calculated at baseline and after Results treatment. The early blood response is seen after one course of treatment with mogamulizumab Quantitative real-time PCR for Foxp3 and CCR4 mRNA Demographics and baseline laboratory data for 24 patients expression with CTCL in this translational study are summarized in Table 1. Total RNA was extracted from PBMCs with RNeasy Mini Kit Eleven patients with mycosis fungoides and 13 Sezary syndrome (Qiagen). First strand cDNA was synthesized from 400 ng of total participating in the Phase I/II clinical trial were enrolled. Seven RNA with an oligo (dT) 12–18 primer using SuperScript III reverse patients participated in phase I, and 17 were participants of phase transcriptase (Life Technologies Inc.). Preformulated TaqMan pri- II. Twenty patients (including 17 in phase II) received four mers and probes for Foxp3 and CCR4 (Hs99999919-m1) and doses of mogamulizumab at 1.0 mg/kg. Two patients received GAPDH were used. Quantitative PCR was performed and the 0.1 mg/kg, and two received 0.3 mg/kg in phase I. After complet- relative fold changes were calculated as described previously ing one course of treatment, 14 of 19 patients (73.7%) with initial (15, 25). blood involvement had responses in blood, including 10 of 13 Sezary syndrome patients (76.9%) and 4 of 6 in mycosis fun- Immunohistochemical staining for CCR4, Foxp3, and CD16 in goides patients (66.7%). A response in skin was seen in only 5 of CTCL skin lesions 24 (20.8%), including 2 of 13 Sezary syndrome patients (15.4%), Mycosis fungoides/Sezary syndrome tissue sections were and 3 of 11 mycosis fungoides patients (27.3%). Responses in stained with H&E for pathologic diagnosis and confirmation blood were more common than responses in skin (P < 0.01), (26). The expression of CCR4 was assessed by immunohisto- suggesting that an early blood response after one course of chemical staining with DAKO EnVisionþ System (DAB; Dako mogamulizumab occurs in most patients with CTCL, especially North America, Inc.) using the murine parent monoclonal anti- patients with Sezary syndrome who have high blood burden. body, KM2160 (2.0 mg/mL, Kyowa Hakko Kirin Pharma, Inc.). þ KM2160 binds to the same epitope as KW-761. HH cell mouse Decreased circulating CCR4 malignant T cells are seen after xenograft sections served as a positive control and the SR cell one course of mogamulizumab mouse xenograft sections and IgG1 served as negative controls. The percentages of circulating malignant T cells (out of total The expressions of Foxp3 and CD16 were assessed using anti- lymphocytes) at baseline for all patients are shown in Table 1. All Foxp3 monoclonal antibody, 236A/E7 (1.0 mg/mL, 1:10, Abcam 13 Sezary syndrome patients and 6 of 11 mycosis fungoides Inc.) and prediluted anti-CD16 antibody (2H7; Abcam Inc.) with patients had detectable circulating malignant T cells (Fig. 1A). the same DAKO EnVisionþ System. The percentages of lympho- The mean percentages and absolute numbers of malignant T cells þ cytic infiltrates positive for CCR4 and Foxp3 as well as CD16 cells in patients with CTCL were 39.4% and 4154.4/mL. As expected, in skin lesions were recorded. patients with Sezary syndrome had higher numbers of circulating malignant T cells than patients with mycosis fungoides who had limited blood involvement (Table 1). Chromium release assay Further flow cytometry analysis for CCR4 expression on malig- Target K562 cells (ATCC) were radiolabeled with 200 pCi/mL nant T cells was performed with 1G1 anti-CCR4 antibody. Before Na [51Cr] (Amersham) at 1 l06/mL for 60 minutes. Labeled treatment, the average percentage of malignant T cells positive for K562 cells were suspended at 5 l04 cells/mL, and dispensed in CCR4 was 83.7% (Fig. 1B; Table 1). Interestingly, although similar 100 mL aliquots into wells of a V-bottom 96-well plate (Costar). blood malignant T cells at baseline, 14 patients who achieved PBMC effector cells were suspended at concentrations of 2.5 responses in blood had higher baseline CCR4 expression on 105,1 106, and 2.5 l06 cells/mL. Then, 100 mL aliquots of malignant T cells (93.5% 9.0%) than 5 nonresponders effector cells were added to wells containing target cells to reach (56.2% 42.2%; P < 0.01, Fig. 1C). effector (E) to target (T) ratios of 5:1, 20:1, and 50:1. Maximum After one course of mogamulizumab treatment, malignant T- release was determined by incubating target cells with BRIJ-35 cell numbers were decreased in 16 of 19 patients, and 14 patients detergent (Sigma). Spontaneous release was determined by incu- reached 50% reduction for a response in blood (Table 1; Fig. bating target cells in medium only. After 4 hours of incubation at 1A). The average decrease was from 39.4% to 19.7% (Table 2, P < 37 C, supernatants were harvested and 51Cr release was measured 0.01). Twelve of 13 Sezary syndrome patients had an average in a gamma counter. Percentages of specific release were calcu- reduction from 64.3% to 32.8% (P < 0.01), and 4 of 6 patients lated as (experimental release spontaneous release)/(maximum with mycosis fungoides with 9.9% to 4.3% (P ¼ 0.12). release spontaneous release) 100. All samples were analyzed As shown in Fig. 1B, the decrease in malignant T cells appears to in quadruplicate. þ correspond to elimination of CCR4 malignant T cells. The þ Statistical analysis average fraction of CCR4 cells in 19 patients decreased from For flow cytometry analysis, the percentage of lymphocytes and 83.7% to 25.2% of malignant T cells (P < 0.01, Table 2). All 13 absolute numbers of each cell population were obtained for all patients with Sezary syndrome showed decreases in CCR4 expres- patients at baseline and after one course of treatment. For real- sion on residual lymphoma cells, with an average reduction from time PCR analysis, fold changes were obtained for all samples. The 93.2% to 34.4% (P < 0.01). Five of 6 patients with mycosis means and SDs were then calculated for all groups. Statistical fungoides also showed a decrease in CCR4 expression, with an significance was determined by c2 tests, Student t test, and paired t average reduction from 63.5% to 5.2% (P < 0.05). test as appropriate. The correlation between two parameters was Next, we used real-time PCR to assess the expression of CCR4 analyzed using the Pearson correlation test. Differences between mRNA in PBMCs of 19 patients. Patients with CTCL had much groups were considered significant if P < 0.05. higher CCR4 mRNA levels (23.39 33.50 fold) compared with

www.aacrjournals.org Clin Cancer Res; 21(2) January 15, 2015 OF3

Ni et al.

Figure 1. Circulating malignant T cells, CCR4þ malignant T cells, and CCR4 mRNA in aggressive/refractory mycosis fungoides/Sezary syndrome patients before and after mogamulizumab. Multicolor flow cytometry analysis was done to detect CD3þCD4þCD26 and/or CD7 malignant T cells in the blood before and after one course of mogamulizumab. A, representative flow plots of lymphocytes from patient #24 before and after mogamulizumab treatment are shown. The percentages of malignant T cells before and after treatment are also shown beneath. B, CCR4 expression on malignant T cells was further analyzed with 1G1 anti-CCR4 þ þ antibody. Representative flow plots of CD3 CD4 CD26 cells from patient #24 before and after mogamulizumab are shown. The CCR4-positive cell fractions þ are circled in each plot. The percentages of CCR4 malignant T cells before and after treatment are also shown beneath. C, the percentages of malignant T cells (% of lymphocytes) as well as malignant T cells positive for CCR4 (% of malignant T cells) at baseline in blood responders (R, n ¼ 14) versus nonresponders (NR, þ n ¼ 5) are shown. NS, not significant; , P < 0.01. D, the correlation between CCR4 mRNA levels (fold changes) and absolute numbers of CCR4 malignant T cells (/mL) by 1G1 anti-CCR4 antibody is shown; , dots for patient #22; Pearson correlation, r ¼ 0.741, P < 0.001. The levels of CCR4 mRNA in PBMCs of 19 patients at baseline and 14 patients after one course of mogamulizumab are shown.

healthy donors (2.08 0.64-fold, P < 0.05). As expected, higher The results conﬁrm that the majority of circulating malignant T CCR4 mRNA levels were seen in patients with Sezary syndrome cells in patients with CTCL, especially those with Sezary syn- than in patients with mycosis fungoides. The levels of CCR4 drome, were CCR4 positive and sensitive to mogamulizumab mRNA were correlated with the absolute numbers of circulating treatment. Higher CCR4 expression on malignant T cells was þ CCR4 malignant T cells in these patients (Fig. 1D, r ¼ 0.731; P < associated with an early blood response.

0.001). The highest CCR4 mRNA level (134.0-fold) was found in þ Sezary syndrome patient #22 who also had the highest numbers Circulating CCR4 Tregs also decrease after one course of of malignant T cells with 99.5% positive for CCR4 expression. Of mogamulizumab 5 mycosis fungoides patients who had no detectable malignant T To determine whether mogamulizumab also reduces the num- cells, 4 showed low levels of CCR4 mRNA and one patient (#9) bers of Tregs, we simultaneously monitored the numbers of had a moderate level of CCR4 mRNA which could be attributed to circulating Tregs during therapy. As shown in Table 1, 18 of 24 þ þ þ þ dim/ increased CCR4 Tregs. After treatment, all 14 patients showed patients (75%) had detectable CD3 CD4 CD25 CD127 decreased CCR4 mRNA levels (Fig. 1D) with an average fold circulating Tregs at baseline (Fig. 2A). Five Sezary syndrome and change from 23.39- to 1.73-fold (P < 0.05, Table 2), supporting one mycosis fungoides patients had no detectable Tregs. The þ our ﬁnding of reduced CCR4 malignant T cells following mean percentage and absolute numbers of Tregs (of total lym- mogamulizumab. phocytes) were 1.2% and 39.2/mL for all patients with CTCL.

OF4 Clin Cancer Res; 21(2) January 15, 2015 Clinical Cancer Research

Table 1. The demographics, malignant T cells, Tregs, CD8þ T cells, NK cells, and responses to mogamulizumab in patients with CTCL Malignant CCR4þ on Clinical response Published OnlineFirstNovember5,2014;DOI:10.1158/1078-0432.CCR-14-0830 T cells in malignant Tregs in CCR4þ CD8þ T cells CCR4þ on NK cells in CCR4þ þ

clincancerres.aacrjournals.org Primary Patient Dose/ peripheral T cells peripheral on Tregs in peripheral CD8 T cells peripheral cells in skin Lymph diagnosis ID Stage Age, y Gender course Phase blood (%)a (% CCR4) blood (%)a (% CCR4) blood (%)a (% CCR4) blood (%)a lesions (%)a Blood Skin node Overall Mycosis 1 IIB 75 M 0.3 mg/kg/1 I 0 — 2.3 90.5 14.3 20.5 —— —PD SD PD fungoides 2 IVA 81 M 1.0 mg/kg/1 I 34.4 95.7 1.5 100 16.7 4.6 16.6 90 PR SD SD PR (n ¼ 11) 3 IIB 73 M 1.0 mg/kg/1 II 0 — 0 — 18.5 22.2 16.4 0 — SD — SD 4 IIB 66 M 1.0 mg/kg/1 II 0 — 1.6 94.4 11.5 6.2 36.4 1 — PR — PR 5 IIIA 48 F 1.0 mg/kg/1 II 0 — 2.1 93.8 27.9 18.3 14.9 2.5 — PR — PR 6 IIB 76 M 1.0 mg/kg/1 II 16.2 3.3 0.7 90.2 39.6 6.5 24.7 90 SD SD — SD 7 IB 79 M 1.0 mg/kg/1 II 21.2 94.9 2.3 96 13.4 17.4 32.9 60 PR SD SD SD 8 IVA 68 M 0.1 mg/kg/1 I 5.6 99.5 0.8 100 11.3 18.7 — 75 PR PR SD PR 9 IIB 66 M 1.0 mg/kg/1 II 0 — 2 78.2 15.9 8.9 1.4 85 — SD SD SD 10 IIB 53 M 1.0 mg/kg/1 II 13.7 20.8 1.5 58.6 22 9.2 — 45 SD SD — SD Cancer Research. 11 IVB 67 M 1.0 mg/kg/3 II 18.0 64.1 3.8 86.1 10.5 13.8 — 0PRSDSDSD

on September 28, 2021. © 2014American Association for 68 10/1 1/1/9 3/8 4/6 3/11 0/5 4/11 Mean SD (48–81) (M/F) (0.1/0.3/1.0) (I/II) 9.9 11.6 63.1 42.3 1.7 1.0 88.8 12.5 18.3 8.3 13.3 6.1 20.5 11.9 44.8 38.2 66.7% 27.3% — 36.4% Sezary 12 IVB 64 F 1.0 mg/kg/1 I 66.4 91.5 1.2 100 15.7 3.9 — 7.5 CR SD PD PD syndrome 13 IIIB 84 F 1.0 mg/kg/1 II 84.7 99.5 0 — 4.3 10.1 6.7 — SD SD PD PD (n ¼ 13) 14 IVA 62 F 1.0 mg/kg/1 II 95 95.1 0 — 2.9 20 4.8 0 PR SD SD SD 15 IVB 79 F 1.0 mg/kg/1 II 73.8 97.4 1.5 97.8 3.5 22 6.0 80 PR SD SD SD 16 IVA 69 M 0.1 mg/kg/1 I 30.4 95.9 0.1 100 17.7 12.6 — 35 CR PR PR PR 17 IVA 65 F 0.3 mg/kg/1 I 7.7 100.0 2.5 59.3 23.6 3.2 — 60 CR SD — SD 18 IVA 47 M 1.0 mg/kg/1 I 58.7 88.3 2.9 97.9 9.5 16.7 7.3 0 SD SD SD SD — — 19 IVA 61 M 1.0 mg/kg/1 II 55.2 96.5 0 6.2 7.2 90 CR SD SD SD CTCL in Cells NK and Tregs on KW-0761 of Effect 20 IVA 63 M 1.0 mg/kg/1 II 40.9 69.1 1 91.7 8.7 5 — 95 SD SD PD SD 21 IVA 76 M 1.0 mg/kg/1 II 82.3 91.3 0.4 95.7 15.9 23.2 1.3 85 PR SD SD SD 22 IVA 62 M 1.0 mg/kg/1 II 94.5 99.5 0 — 0.8 8.9 1.3 100 CR PR PR PR 23 IVA 72 F 1.0 mg/kg/1 II 60.1 96.5 0 — 8.8 ——65 PR SD SD SD lnCne e;2()Jnay1,2015 15, January 21(2) Res; Cancer Clin 24 IVB 86 M 1.0 mg/kg/1 II 86.5 90.5 0.4 97.1 3.4 16.8 12.4 70 PR SD SD SD Mean SD 65 7/6 1/1/11 4/9 64.3 26.2 93.2 8.1 0.8 1.0 87.1 20.5 9.3 6.7 12.5 6.9 5.7 2.7 57.3 35.8 10/13 2/13 2/12 2/13 P (47–86) (M/F) (0.1/0.3/1.0) (I/II) <0.01 <0.05 <0.05 NS <0.01 NS <0.01 NS 76.9% 15.4% 16.7% 15.4% Sum n ¼ 24 67.5 17/7 2/2/20 7/17 39.4 34.4 83.7 27.2 1.2 1.1 88.0 16.3 13.4 8.7 12.9 6.5 13.1 11.4 51.6 37.4 14/19 5/24 2/17 6/24 (47–86) (M/F) (0.1/0.3/1.0) (I/II) 73.7% 20.8% 11.8% 25.0% NOTE: Italics indicate statistical signiﬁcance. Abbreviations: M, male; F, female; NS, not signiﬁcant. aCell numbers as a percentage of total lymphocytes. OF5 Published OnlineFirst November 5, 2014; DOI: 10.1158/1078-0432.CCR-14-0830

Ni et al.

Table 2. Effects of mogamulizumab on lymphoma cells, Tregs, CD8þ T cells, CCR4 expression, Foxp3 expression, in peripheral blood and skin lesions of patients with CTCL Before treatment After treatment P Patients with CTCL (n ¼ 24) (Mean SD) (Mean SD) (t test) Lymphoma cells in peripheral blood Percentage of total lymphocytes (%) 39.4 34.4 19.7 26.2 <0.01 Absolute # (/mL) 4,154.4 7,646.3 1,089.6 3,294.5 <0.05 CCR4 expression on lymphoma cells Percentage (1G1, % of total lymphoma cells) 83.7 27.2 25.2 29.2 <0.01 Absolute # (1G1, /mL) 3,934.8 7,411.9 652.6 2,185.9 <0.05 Tregs in peripheral blood Percentage of total lymphocytes (%) 1.2 1.1 0.3 0.5 <0.01 Absolute # (/mL) 39.2 46.8 5.6 8.7 <0.01 CCR4 expression on Tregs Percentage (1G1, % of Tregs) 88.0 16.3 30.9 32.0 <0.01 Absolute # (1G1, /mL) 36.5 46.3 2.7 6.8 <0.01 CD8þ T cells in peripheral blood Percentage of total lymphocytes (%) 13.4 8.7 23.1 9.9 <0.01 Absolute # (/mL) 334.5 246.7 363.5 314.8 NS CCR4 expression on CD8þ T cells Percentage (1G1, % of Tregs) 12.9 6.5 8.6 6.4 NS Absolute # (1G1, /mL) 45.3 46.6 36.9 60.3 NS Foxp3 mRNA in PBMCs (fold) 0.57 0.91 0.07 0.08 <0.05 CCR4 mRNA in PBMCs (fold) 23.39 33.50 1.73 2.35 <0.05 NK cells in peripheral blood (n ¼ 14) Percentage of total lymphocytes (%) 13.1 11.4 21.4 10.5 0.05 Absolute # (/mL) 380.9 153.6 362.8 175.4 NS Killing (%; E:T ¼ 50:1; n ¼ 6) 4.9 3.8 8.7 8.5 NS þ CCR4 lymphocytes in skin lesions (%, n ¼ 18) 51.6 37.4 23.9 28.4 <0.01 Abbreviation: NS, not signiﬁcant.

þ þ Patients with mycosis fungoides had higher percentages of Tregs CD4 CD25 CD127dim/ circulating Tregs in our patients (Fig. than patients with Sezary syndrome. Further flow cytometry 2C; r ¼ 0.804, P < 0.001). For example, Sezary syndrome patient analysis using 1G1 anti-CCR4 antibody found that CCR4 was #18, who had the highest number of Tregs (198.9/mL) among positive on 58.6%–100% of Tregs, with a mean of 88.0% (Fig. patients, also exhibited the highest Foxp3 mRNA level (4.06- 2B; Table 1). Fourteen of 18 patients had more than 90% of Tregs fold). Meanwhile, 5 patients with undetectable Tregs showed only positive for CCR4. low levels of Foxp3 mRNA. After treatment, 14 of 14 patients with After treatment, there was a clear trend towards decreased Tregs follow-up blood samples all showed decreased Foxp3 mRNA in all 18 patients who had detectable Tregs at baseline, regardless levels (Fig. 2D; Table 2). of their clinical response. The decrease was very significant (Table These results indicate that the majority of Tregs in the blood of 2). Tregs in patients with mycosis fungoides was decreased from mycosis fungoides/Sezary syndrome were CCR4 positive and 1.7% or 26.7/mL to 0.4% or 5.1/mL(P < 0.01) and in patients with were very sensitive to mogamulizumab treatment. Sezary syndrome from 0.8% or 54.8/mL to 0.3% or 6.1/mL(P < þ 0.05). The percentages of CD8 T cells increase after one course of All 18 patients had decreased CCR4 expression on Tregs as well mogamulizumab þ þ as decreased absolute CCR4 Tregs after treatment, regardless of The percentages of CD8 T cells of total lymphocytes in þ þ clinical response. The decrease in CCR4 Tregs was very significant peripheral blood and the expression of CCR4 on CD8 T cells þ (Table 2). The percentages of CCR4 cells and absolute counts in for all patients at baseline are shown in Table 1. As expected, there þ patients with mycosis fungoides decreased from 88.8% and 23.5/ were lower percentages of CD8 T cells in Sezary syndrome mL to 16.8% and 0.8/mL, and in patients with Sezary syndrome patients (9.3% 6.7%) than in mycosis fungoides patients from 87.1% and 52.9/mL to 46.5% and 4.6/mL. (18.3% 8.3%, P < 0.01). CCR4 was positive on 3.2% to þ We also assessed the mRNA expression of Foxp3, a transcrip- 23.2% of CD8 T cells, with an average of 12.9% 6.5% for all tion factor important for the development of Tregs. We were able patients which were much lower than CCR4 on malignant T cells to detect Foxp3 mRNA in PBMCs of all 19 patients tested at (83.7% 27.2%; P < 0.001) and Tregs (88.0% 16.3%; P < baseline. Higher expression of Foxp3 mRNA was present in 0.001). patients with CTCL (0.57 0.91 fold) compared with healthy After one course of mogamulizumab, CCR4 expression was þ donors (0.09 0.04 fold, P < 0.05). Foxp3 mRNA levels were reduced on CD8 T cells in patients with mycosis fungoides, with slightly higher in patients with Sezary syndrome (0.79 1.20 an average from 13.3% 6.1% to 7.0% 5.2% (P < 0.05) but not fold) than in patients with mycosis fungoides (0.33 0.30 fold, P seen in Sezary syndrome patients (12.5% 6.8% to 10.0% þ ¼ 0.27). This may be due to the higher absolute numbers of Tregs 6.9%, P > 0.05). Interestingly, the percentages of CD8 T cells were in patients with Sezary syndrome (54.8/mL) compared with increased in 23 of 24 patients, an average increase from 13.4% mycosis fungoides patients (26.7/mL) and Foxp3 expression by 8.7% to 23.1% 9.9% in all patients (Table 2, P < 0.01). The malignant T cells (15). The levels of Foxp3 mRNA in PBMCs were increases were seen in both Sezary syndrome patients (20.8 þ well correlated with the absolute numbers of CD3 10.9%, P < 0.01) and mycosis fungoides patients (25.9 7.5% P <

OF6 Clin Cancer Res; 21(2) January 15, 2015 Clinical Cancer Research

Effect of KW-0761 on Tregs and NK Cells in CTCL

Figure 2. Circulating Tregs, CCR4þ Tregs, and Foxp3 mRNA in aggressive/refractory mycosis fungoides/Sezary syndrome patients before and after mogamulizumab. þ þ þ A, multicolor flow cytometry analysis was done to detect CD3 CD4 CD25 CD127dim/ Tregs in the blood before and after one course of mogamulizumab. Representative flow plots from patient #6 before and after mogamulizumab are shown. The percentages (% of total lymphocytes) and absolute numbers (/mL) of Tregs from 18 patients before and after treatment are shown. B, CCR4 expression on Tregs was further measured with 1G1 anti-CCR4 antibody before and after mogamulizumab. Representative flow plots from patient #6 before and after mogamulizumab are shown. The CCR4-positive cell fractions are circled in þ þ each plot. The percentages of CCD4 cells (% of total Tregs) and absolute numbers (/mL) of CCR4 Tregs from 18 patients before and after treatment are þ þ þ shown. , P < 0.01. C, the correlation between Foxp3 mRNA levels (fold changes) and absolute numbers of CD3 CD4 CD25 CD127dim/ Tregs (/mL) is shown; , dots for patient#18; Pearson correlation, r ¼ 0.804, P < 0.001. D, the levels of Foxp3 mRNA in PBMCs of 19 patients at baseline and in 14 patients after one course of mogamulizumab are shown.

þ þ 0.01). The absolute numbers of CD8 T cells remained pretty even CCR4 lymphocytes in CTCL skin lesions decrease after one in Sezary syndrome patients (baseline: 325.7 219.5/mL; post: course of mogamulizumab 383.5 374.2/mL, P ¼ 0.55) and mycosis fungoides patients Using the murine parent anti-CCR4 antibody, KM2160, we þ (baseline: 344.9 275.1/mL; post: 339.9 223.1/mL, P ¼ 0.95). assessed CCR4 inﬁltrating lymphocytes in skin lesions from 22 þ The results suggest that the expression of CCR4 on CD8 T cells patients with CTCL by immunohistochemistry. Eighteen of 22 þ was relatively low compared with CCR4 on malignant T cells and baseline lesions had CCR4 inﬁltrating lymphocytes, ranging þ Tregs, and that CD8 T cells are less sensitive to mogamulizumab from 1% to 100% (Fig. 3A–C). Four lesions (2 mycosis fungoides þ treatment. and 2 Sezary syndrome) had no detectable CCR4 cells. Average

www.aacrjournals.org Clin Cancer Res; 21(2) January 15, 2015 OF7

Ni et al.

Figure 3. CCR4 and Foxp3 expression by immunohistochemical staining (IHC) in mycosis fungoides/Sezary syndrome skin lesions before and after mogamulizumab. A, mycosis fungoides/Sezary syndrome skin specimens before and after treatments were stained with hematoxylin and eosin (H&E) for pathologic diagnosis and confirmation, patient #15; magnification, 200. B, by IHC, CCR4 protein was examined using the murine parent anti-CCR4 antibody (KM2160, 2.0 mL/mL) in skin þ lesions before and after treatment, from patient #15; magnification, 200. C, CCR4 -infiltrating lymphocytes (% of total infiltrating lymphocytes) in skin lesions from 18 patients before and after treatment are presented, paired t test; , P < 0.01. D, by IHC, Foxp3 protein was examined using antibody (236A/E7, 1.0 mg/mL, 1:10) in skin lesion before and after treatment, from patient #19; magnification, 200.

þ þ percentages of CCR4 cells in Sezary syndrome lesions (57.3%) (Fig. 3D), with decreased CCR4 cells as well. Patient #24 had 2% þ and mycosis fungoides lesions (44.8%, P ¼ 0.46, Table 1) were Foxp3 cells in both baseline and posttreatment lesions while þ þ þ similar. There was no difference in CCR4 cell numbers in the CCR4 cells were also unchanged. There were various CD16 cells baseline lesions between patients who had responses in skin (n ¼ in baseline lesions which showed little change after treatment 5, 42.7% 35.9%) and patients without a response in skin (n ¼ (data not included). 17, 54.3% 36.4%, P ¼ 0.57). Interestingly, after just one course of treatment, all 5 patients who showed clinical partial responses NK cell numbers and function are restored after treatment with þ in skin had a decrease in CCR4 cells. In addition, another 11 mogamulizumab þ patients with stable disease in skin also showed decreased CCR4 To assess the subsequent effect of mogamulizumab on NK cells, þ þ cells to various degrees, with a decrease in the mean from 51.6% to we examined the numbers of CD3 CD56 CD16 NK cells in the 23.9% (16/18, P < 0.01, Fig. 3C; Table 2). Only one treated patient blood and their cytotoxicity before and after treatment. Flow þ (#24) retained a high percentage of CCR4 cells in the lesion, and cytometry analysis showed that the percentages of NK cells in showed no clinical improvement in skin lesions. A posttreatment 14 patients with CTCL tested (13.1 11.4%, Fig. 4A; Tables 1 lesional biopsy was not available in another patient. Four patients and 2) was comparable with those in healthy donors (n ¼ 5, 14.6 with skin lesions negative for CCR4 at baseline remained negative 12.7%). However, much lower percentages and absolute num- and none of them showed skin improvement. bers of NK cells were present in Sezary syndrome patients (n ¼ 7, þ Foxp3 -inﬁltrating lymphocytes were found in only 10 of 22 5.7 2.7% or 357.6 124.3/mL) than in mycosis fungoides baseline skin lesions. Eight lesions showed single or scattered patients (n ¼ 7, 20.5 11.9%, P < 0.01 or 404.2 178.0/mL). þ Foxp3 cells, and only 2 lesions (#19 and #24) had countable After treatment, 10 of 14 patients demonstrated increased NK þ þ numbers of Foxp3 cells. Patient #19 had 10% Foxp3 cells in the cell percentages with an average of 13.1% at baseline versus 21.4% baseline lesion that dropped to 5% positivity following treatment after treatment (P ¼ 0.05), but with similar absolute numbers

OF8 Clin Cancer Res; 21(2) January 15, 2015 Clinical Cancer Research

Effect of KW-0761 on Tregs and NK Cells in CTCL

Figure 4. Circulating NK cells and their function in aggressive/refractory mycosis fungoides/Sezary syndrome patients before and after mogamulizumab. Multicolor flow cytometry analysis was used to detect CD3CD16þCD56þ NK cells in peripheral blood before and after mogamulizumab. A, representative flow plots from patient #6 before and after one course of mogamulizumab are shown. B, the percentages (% of total lymphocytes) and the absolute numbers (/mL) of þ þ CD3 CD16 CD56 NK cells in 14 paired patients before and after treatment are shown. C, NK cell–specific killing (%) by a 51Cr release assay before (-^-) and after treatment (-&-) for patients #4, #7, #13, #24, #14, and #6 at E:T ratios of 5:1, 20:1, and 50:1 are shown.

(380.9 versus 362.8/mL, P ¼ 0.74; Fig. 4B; Table 2). NK cells were The results suggest that NK cell percentages and function may increased in 3 of 7 mycosis fungoides patients, but not in the be restored after anti-CCR4 treatment, especially in patients with remaining 4 mycosis fungoides patients (average from 20.5% or Sezary syndrome, and could be attributable to the removal of 404.2/mL to 21.5% or 402.5/mL, P ¼ 0.99). In contrast, increases in malignant T cells and/or Tregs. NK cell percentages were seen in all 7 tested Sezary syndrome P patients, with an average increase from 5.7% to 21.3% ( < 0.01). Discussion All 7 Sezary syndrome patients also had decreased levels of both malignant T cells and Tregs, with 5 of 7 patients achieving We studied 24 aggressive/refractory mycosis fungoides/Sezary responses in blood. syndrome patients participating in a phase I/II clinical trial of We also assessed NK cell cytotoxicity by a 51Cr release assay. mogamulizumab and found that majority of circulating malig- Among 6 patients assessed, 4 patients demonstrated an increase in nant T cells in these patients were positive for CCR4, and NK cell killing at different E:T ratios (Fig. 4C). Three of these 4 decreased after only one 4-week course of mogamulizumab þ patients exhibited responses in blood and one mycosis fungoides infusion. Indeed, a high proportion of CCR4 malignant T cells patient, without blood involvement, achieved a partial response in the blood are associated with the early blood response. Cir- in skin and overall assessment. The average specific killing at the culating Tregs were also positive for CCR4, and were decreased 50:1 ratio was up from 4.9% at baseline to 8.7% after treatment following therapy, confirming the effect of mogamulizumab on but significant (Table 2; P ¼ 0.27). NK cell killing in one patient Tregs in vivo. The expression levels of Foxp3 and CCR4 mRNA in was similar before and after treatment (#14). Another patient PBMCs of these patients were consistent with the observed num- þ whose NK cell numbers remained unchanged exhibited a decrease bers of Tregs and malignant T cells. Meanwhile, CCR4 -infiltrat- in NK cell killing (#6). Blood and skin improvement were not ing lymphocytes in skin lesions were also reduced after 4 weeks of present after one course of treatment in these 2 patients. treatment. A posttreatment increase in NK cell percentages in

www.aacrjournals.org Clin Cancer Res; 21(2) January 15, 2015 OF9

Ni et al.

Sezary syndrome patients and NK cell cytotoxicity in some where CCR4 ligands, including CCL17 (TARC) and CCL22 patients suggests that potential restoration of NK cell function (MDC) are increased (11, 38). High levels of CCL17 and CCL22 þ may also be achieved with 4 weeks of therapy. Thus, mogamu- were found in CTCL lesions, and could attract CCR4 malignant T þ þ lizumab treatment not only eliminates CCR4 malignant T cells cells as well as CCR4 Tregs into the skin (6). A recent mouse study þ locally and systemically, but also reduces CCR4 Tregs that may suggests that high levels of CCL17 and CCL22 were found in be instrumental for restoring NK cell antitumor function in inflammatory lesions where there was a heavily skewed Th2-type patients with mycosis fungoides/Sezary syndrome. To our knowl- cytokine response (39). Advanced CTCL, especially Sezary syn- edge, studies monitoring the effects of mogamulizumab on drome, are malignancies with skewed Th2-type cytokine profiles. different T-cell subsets and NK cells in patients with mycosis Therefore, anti-CCR4 treatment in CTCL could not only remove þ fungoides/Sezary syndrome in vivo have not been reported. CCR4 malignant T cells but also deplete immunosuppressive þ þ Multiple strategies have been attempted to target CCR4 in CCR4 Tregs. In this study, circulating CCR4 Tregs as well as CTCL. Han and colleagues described an anti-CCR4 gene transfer Foxp3 mRNA levels in PBMCs were reduced in all patients studied þ system in a xenograft mouse model. This approach showed in vivo after treatment. In CTCL skin lesions, CCR4 -infiltrating lympho- þ antitumor activity in mice model and a reduction of CCR4 cytes were decreased in most skin lesions after treatment. A few þ malignant T cells and human cell lines in vitro (27, 28), but Foxp3 cells were seen in half of lesions. We postulate that Tregs þ human trials have not been reported. Mogamulizumab or KW- comprise a portion of the CCR4 cells in CTCL lesions. A more 0761, a humanized antibody given by intravenous infusion, has sensitive assay for detecting Tregs in skin lesions could be used in shown promising results in clinical trials for refractory ATLL future studies. and PLCL (19, 20) and for mycosis fungoides/Sezary syndrome Human Tregs are divided into three subsets: effector Tregs (19–21, 29). (eTreg), na€ve Tregs, and nonclassic Tregs (35, 40, 41). Effector This translational study ran in parallel with the phase I/II trial, Tregs are differentiated and highly suppressive, and they are þ þ but clinical responses were assessed after only one course or four characterized by a Foxp3hi CD45RA CD25hi phenotype. Na€ve þ þ þ weekly doses of treatment. A high response rate was seen in blood Tregs are Foxp3lo CD45RA CD25lo and can differentiate into þ in 14 of 19 or 73.7% of patients studied, compared with one-fifth eTregs after antigenic stimulation. Nonclassic Tregs are Foxp3lo þ of patients with responses in skin (5/24, 20.8%). The difference CD45RA CD25lo and lack suppressive functional activity but do between the responses in blood and in skin may be attributed to secrete proinflammatory cytokines (35, 40, 41). Sugiyama and the intravenous administration of mogamulizumab and to the colleagues found that CCR4 was predominantly expressed on þ high numbers of CCR4 cells in blood of Sezary syndrome eTregs but not on na€ve Tregs in peripheral blood of healthy patients (average of 61.8% in Sezary syndrome and 10.5% in donors and patients with melanoma (35). They also found that þ mycosis fungoides). CCR4 cells were decreased in 16 of 18 skin anti-CCR4 antibody treatment resulted in selective reduction of lesions. A longer treatment time may be required to eliminate eTregs, while preserving na€ve Tregs. Our study confirms that þ þ CCR4 cells from skin lesions. In the completed phase I/II clinical CCR4 Tregs in mycosis fungoides/Sezary syndrome patients trial, skin responses were seen in more than 40% of patients account for about 90% of Tregs in peripheral blood and they are continuing on biweekly treatment schedules (M.A. Duvic; sub- very sensitive to mogamulizumab. Therefore, mogamulizumab þ mitted for publication). treatment may lead to selective elimination of CCR4 eTregs in Among multiple cell surface makers on malignant T cells of patients with CTCL. CTCL, CCR4 is a superior target in comparison with other markers Numerous strategies have or are being investigated to eliminate like CD3, CD4, or CD52 for therapeutic targeting. CD3 is or inhibit Tregs in patients with cancer. These include anti-CD25 expressed on all T cells, CD4 is present on all helper T cells, and antibody and denileukin diftitox, an IL2–diptheria fusion protein CD52 is expressed not only on all mature T lymphocytes, but also (42). However, elimination of Tregs by these approaches can elicit on B lymphocytes, NK cells, and dendritic cells (30). Targeting harmful autoimmunity. If anti-CCR4 treatment reduces only these markers may deplete all T cells, normal helper T cells, or eTregs during treatment, such severe autoimmunity may be other innate immune cells, raising the risk of infection and other avoided because na€ve Tregs are still available to differentiate adverse events (31, 32). In contrast, CCR4 is highly expressed on into eTregs. Of note, none of mycosis fungoides/Sezary syndrome skin-homing T cells which give rise to malignant T-cell clones in patients treated with mogamulizumab in this study experienced a þ CTCL (6). Moreover, CCR4 is rarely expressed on CD8 T cells, B severe immune-related adverse event, except mild skin rash (M.A. cells, NK cells, monocytes, and/or macrophages. Although CCR4 Duvic; submitted for publication). Therefore, anti-CCR4 therapy may be expressed on platelets and important for platelet activa- maybe a unique cancer immunotherapy by depleting eTregs tion (33, 34), we found a small effect of mogamulizumab on without eliciting serious autoimmunity. þ platelets in our study (Supplementary Fig. S1). Thus, CCR4 is a We found that there are low numbers of CD8 T cells and NK highly specific target on the surface of CTCL malignant T cells, and cells in Sezary syndrome patients compared with mycosis fun- targeting CCR4 may protect innate immune cells from destruction goides patients, which may contribute to more severe immune þ (35). In addition, CCR4 malignant T cells are not only found suppression in patients with Sezary syndrome. Interestingly, þ mycosis fungoides and Sezary syndrome, but also in other types of CD8 T cells and NK cell numbers rebounded in patients with CTCL such as primary cutaneous anaplastic large cell lymphoma, Sezary syndrome after treatment and NK cell function was and some peripheral T-cell lymphomas not otherwise specified increased in some patients. Removal of immunosuppressive Tregs (6, 36, 37). Hence, broad clinical applications for anti-CCR4 and/or malignant T cells by mogamulizumab may be instrumen- treatment are expected. tal for restoration of NK cell numbers and function in patients High expression of CCR4 on Tregs is reported to be related to with mycosis fungoides/Sezary syndrome. The improvement in þ tumor escape from immune surveillance. The expression of CCR4 the immune profiles of CD8 T cells and NK cells was correlated on the surface of Tregs allows them migrate into the tumor site with clinical responses in blood in patients with Sezary syndrome.

OF10 Clin Cancer Res; 21(2) January 15, 2015 Clinical Cancer Research

Effect of KW-0761 on Tregs and NK Cells in CTCL

Future studies with longer follow-up are warranted. We see a Kirin, and Millenium. No potential conflicts of interest were disclosed by the rationale for use of mogamulizumab to deplete Tregs in vaccina- other authors. tion trials in patients with CTCL (25, 43, 44). Similar to other monoclonal antibodies (mAb) used in Authors' Contributions cancer immunotherapy, mogamulizumab works by ADCC. Conception and design: X. Ni, M. Goswami, M.A. Duvic Development of methodology: X. Ni, J.L. Jorgensen, M. Goswami, W.K. Decker Therefore, any strategy which stimulates NK cell function and Acquisition of data (provided animals, acquired and managed patients, enhance ADCC can be synergistic with mogamulizumab ther- provided facilities, etc.): X. Ni, M. Goswami, P. Challagundla, W.K. Decker, apy. A study reported that an agonistic anti-CD137 (4-1BB) Y.H. Kim, M.A. Duvic mAb boosted NK cell function when combined with tumor- Analysis and interpretation of data (e.g., statistical analysis, biostatistics, specific mAbs in xenotransplant models of lymphoma and computational analysis): X. Ni, J.L. Jorgensen, M. Goswami, P. Challagundla, Y. breast cancer (45–47). Another study reported that ADCC H. Kim, M.A. Duvic Writing, review, and/or revision of the manuscript: X. Ni, J.L. Jorgensen, W.K. induced by KW2760, another defucosylated anti-CCR4 anti- Decker, Y.H. Kim, M.A. Duvic body, is greatly enhanced by the immunomodulatory cytokines Administrative, technical, or material support (i.e., reporting or organizing IL12, IFNa-2b, and IFNg (7). Future studies combining moga- data, constructing databases): X. Ni, M.A. Duvic mulizumab with other reagents and ADCC enhancers can test Study supervision: X. Ni, M.A. Duvic this potential for synergy. In summary, our study demonstrates that mogamulizumab Acknowledgments þ can concurrently reduce circulating CCR4 malignant T cells The authors thank all patients who participated in this study. The authors þ and CCR4 Tregs in patients with aggressive/refractory mycosis also thank Nehal Mohammed (Ph.D.) and Takeshi Takahashi (Ph.D.) for fungoides/Sezary syndrome in vivo.Thesefindings will be valuable discussion during the study, Timothy Langridge for technical help, and Humphrey (MD) and Karen Dwyer for reviewing this article. further confirmed during the ongoing phase III randomized multicenter clinical trial. We speculate that mogamulizumab Grant Support therapy may not only have broad applications for treating The work was supported in part by an investigator-initiated research grant patients with CTCL, but also may have value in treating many from Kyowa Hakko Kirin Co., Ltd. (LS2009-00025650JW; to X. Ni and M.A. other tumors with immunosuppressive mechanisms involving Duvic). Tregs. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked Disclosure of Potential Conflicts of Interest advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate J.L. Jorgensen reports receiving a commercial research grant from Kyowa this fact. Hakko Kirin. M.A. Duvic reports receiving commercial research grants from Eisai, Galderma, Kyowa Hakko Kirin, Millennium, and Seattle Genetics and Received April 7, 2014; revised August 20, 2014; accepted October 22, 2014; is a consultant/advisory board member for Celgene, Eisai, Kyowa Hakko published OnlineFirst November 5, 2014.

References 1. Olsen E, Vonderheid E, Pimpinelli N, Willemze R, Kim Y, Knobler R, et al. 9. Iellem A, Mariani M, Lang R, Recalde H, Panina-Bordignon P, Sinigaglia F, Revisions to the staging and classification of mycosis fungoides and Sezary et al. Unique chemotactic response profile and specific expression of syndrome: a proposal of the International Society for Cutaneous Lympho- chemokine receptors CCR4 and CCR8 by CD4(þ)CD25(þ) regulatory T mas (ISCL) and the cutaneous lymphoma task force of the European cells. J Exp Med 2001;194:847–53. Organization of Research and Treatment of Cancer (EORTC). Blood 10. Sakaguchi S. Naturally arising CD4þ regulatory t cells for immunologic 2007;110:1713–22. self-tolerance and negative control of immune responses. Annu Rev 2. Talpur R, Singh L, Daulat S, Liu P, Seyfer S, Trynosky T, et al. Long-term Immunol 2004;22:531–62. outcomes of 1,263 patients with mycosis fungoides and Sezary syndrome 11. Curiel TJ, Coukos G, Zou L, Alvarez X, Cheng P, Mottram P, et al. Specific from 1982 to 2009. Clin Cancer Res 2012;18:5051–60. recruitment of regulatory T cells in ovarian carcinoma fosters immune 3. Vidulich KA, Talpur R, Bassett RL, Duvic M. Overall survival in erythro- privilege and predicts reduced survival. Nat Med 2004;10:942–9. dermic cutaneous T-cell lymphoma: an analysis of prognostic factors in a 12. Ishida T, Ishii T, Inagaki A, Yano H, Komatsu H, Iida S, et al. Specific cohort of patients with erythrodermic cutaneous T-cell lymphoma. Int J recruitment of CC chemokine receptor 4-positive regulatory T cells in Dermatol 2009;48:243–52. Hodgkin lymphoma fosters immune privilege. Cancer Res 2006;66: 4. Imai T, Nagira M, Takagi S, Kakizaki M, Nishimura M, Wang J, et al. 5716–22. Selective recruitment of CCR4-bearing Th2 cells toward antigen-presenting 13. Nishikawa H, Sakaguchi S. Regulatory T cells in tumor immunity. Int J cells by the CC chemokines thymus and activation-regulated chemokine Cancer 2010;127:759–67. and macrophage-derived chemokine. Int Immunol 1999;11:81–8. 14. Berger CL, Tigelaar R, Cohen J, Mariwalla K, Trinh J, Wang N, et al. 5. Wu XS, Lonsdorf AS, Hwang ST. Cutaneous T-cell lymphoma: roles for Cutaneous T-cell lymphoma: malignant proliferation of T-regulatory cells. chemokines and chemokine receptors. J Invest Dermatol 2009;129: Blood 2005;105:1640–7. 1115–9. 15. Shiue LH, Ni X, Prieto VG, Jorgensen JL, Curry JL, Goswami M, et al. A case 6. Ferenczi K, Fuhlbrigge RC, Pinkus JL, Pinkus GS, Kupper TS. Increased of invisible leukemic cutaneous T cell lymphoma with a regulatory T cell CCR4 expression in cutaneous T cell lymphoma. J Invest Dermatol clone. Int J Dermatol 2013;52:1111–4. 2002;119:1405–10. 16. Jones D, Dang NH, Duvic M, Washington LT, Huh YO. Absence of CD26 7. Yano H, Ishida T, Inagaki A, Ishii T, Ding J, Kusumoto S, et al. Defucosylated expression is a useful marker for diagnosis of T-cell lymphoma in periph- anti CC chemokine receptor 4 monoclonal antibody combined with eral blood. Am J Clin Pathol 2001;115:885–92. immunomodulatory cytokines: a novel immunotherapy for aggressive/ 17. Heid JB, Schmidt A, Oberle N, Goerdt S, Krammer PH, Suri-Payer E, et al. refractory Mycosis fungoides and Sezary syndrome. Clin Cancer Res FOXP3(þ)CD25() tumor cells with regulatory function in sezary syn- 2007;13:6494–500. drome. J Invest Dermatol 2009;129:2875–85. 8. Ishii T. [Development of an anti-CCR4 antibody, mogamulizumab, the 18. Banham AH. Cell-surface IL-7 receptor expression facilitates the puri- first approved antibody made by POTELLIGENT((R)) technology]. Nihon fication of FOXP3(þ) regulatory T cells. Trends Immunol 2006;27: Yakurigaku Zasshi 2013;142:167–71. 541–4.

www.aacrjournals.org Clin Cancer Res; 21(2) January 15, 2015 OF11

Ni et al.

19. Yamamoto K, Utsunomiya A, Tobinai K, Tsukasaki K, Uike N, Uozumi 32. Papalas JA, Sebastian S, Wang E. Molecular evidence of a genotypically K, et al. Phase I study of KW-0761, a defucosylated humanized anti- novel large T-cell lymphoma after anti-CD4 therapy for refractory mycosis CCR4 antibody, in relapsed patients with adult T-cell leukemia-lym- fungoides. Leuk Lymphoma 2011;52:905–7. phoma and peripheral T-cell lymphoma. J Clin Oncol 2010;28: 33. Abi-Younes S, Si-Tahar M, Luster AD. The CC chemokines MDC and TARC 1591–8. induce platelet activation via CCR4. Thromb Res 2001;101:279–89. 20. Ishida T, Joh T, Uike N, Yamamoto K, Utsunomiya A, Yoshida S, et al. 34. Clemetson KJ, Clemetson JM, Proudfoot AE, Power CA, Baggiolini M, Wells Defucosylated anti-CCR4 monoclonal antibody (KW-0761) for relapsed TN. Functional expression of CCR1, CCR3, CCR4, and CXCR4 chemokine adult T-cell leukemia-lymphoma: a multicenter phase II study. J Clin Oncol receptors on human platelets. Blood 2000;96:4046–54. 2012;30:837–42. 35. Sugiyama D, Nishikawa H, Maeda Y, Nishioka M, Tanemura A, Katayama I, 21. Duvic M, Pinter-Brown L, Foss FM, Sokol L, Jorgensen J, Spitalny GL, et al. et al. Anti-CCR4 mAb selectively depletes effector-type FoxP3þCD4þ Results of a phase 1/2 study for KW-0761, a monoclonal antibody directed regulatory T cells, evoking antitumor immune responses in humans. Proc against CC chemokine receptor type 4 (CCR4), in CTCL patients. Blood Natl Acad Sci U S A 2013;14:14. (ASH Annual Meeting Abstract) 2010;116:962. 36. Sokolowska-Wojdylo M, Wenzel J, Gaffal E, Lenz J, Speuser P, Erdmann S, 22. Kim YH, Pinter-Brown L, Foss F, Sokol L, Jorgensen JL, Spitalny GL, et al. et al. Circulating clonal CLA(þ) and CD4(þ) T cells in Sezary syndrome CO26. Results of a phase 1/2 study for KW-0761, a monoclonal antibody express the skin-homing chemokine receptors CCR4 and CCR10 as well as directed against CC chemokine receptor type 4 (CCR4), in CTCL patients. the lymph node-homing chemokine receptor CCR7. Br J Dermatol Melanoma Res 2011;21:e15–e6. 2005;152:258–64. 23. Duvic M, Pinter-Brown L, Foss F, Sokol L, Jorgensen J, Ni X, et al. Corre- 37. Jones D, O'Hara C, Kraus MD, Perez-Atayde AR, Shahsafaei A, Wu L, et al. lation of target molecule expression and overall response in refractory Expression pattern of T-cell-associated chemokine receptors and their cutaneous T-cell lymphoma patients dosed with mogamulizumab (KW- chemokines correlates with specific subtypes of T-cell non-Hodgkin lym- 0761), a monoclonal antibody directed against CC chemokine receptor phoma. Blood 2000;96:685–90. type 4 (CCR4). Blood 2012;120:3697. 38. Olkhanud PB, Baatar D, Bodogai M, Hakim F, Gress R, Anderson RL, et al. 24. Ni X, Jorgensen JL, Goswami M, Challagundla P, Decker WK, Kim YH, et al. Breast cancer lung metastasis requires expression of chemokine receptor Effects of anti-CCR4 antibody (KW-0761) on regulatory T cells and natural CCR4 and regulatory T cells. Cancer Res 2009;69:5996–6004. killer cells in patients with cutaneous T-cell lymphoma. Blood 2011; 39. Jakubzick C, Wen H, Matsukawa A, Keller M, Kunkel SL, Hogaboam CM. 118:204–5. Role of CCR4 ligands, CCL17 and CCL22, during Schistosoma mansoni 25. Ni X, Richmond HM, Liao XM, Decker WK, Shiue LH, Shpall EJ, et al. egg-induced pulmonary granuloma formation in mice. Am J Pathol Induction of T-cell responses against cutaneous T-cell lymphomas ex vivo 2004;165:1211–21. by autologous dendritic cells transfected with amplified tumor mRNA. 40. Sakaguchi S, Miyara M, Costantino CM, Hafler DA. FOXP3þ regulatory T J Invest Dermatol 2008;128:2631–9. cells in the human immune system. Nat Rev Immunol 2010;10:490–500. 26. Goswami M, Duvic M, Dougherty A, Ni X. Increased Twist expression in 41. Miyara M, Sakaguchi S. Human FoxP3(þ)CD4(þ) regulatory T cells: their advanced stage of mycosis fungoides and Sezary syndrome. J Cutan Pathol knowns and unknowns. Immunol Cell Biol 2011;89:346–51. 2012;39:500–7. 42. Golovina TN, Vonderheide RH. Regulatory T cells: overcoming suppres- 27. Han T, Abdel-Motal UM, Chang DK, Sui J, Muvaffak A, Campbell J, et al. sion of T-cell immunity. Cancer Journal 2010;16:342–7. Human anti-CCR4 minibody gene transfer for the treatment of cutaneous 43. Kim YH, Gratzinger D, Harrison C, Brody JD, Czerwinski DK, Ai WZ, et al. T-cell lymphoma. PLoS ONE 2012;7:4. In situ vaccination against mycosis fungoides by intratumoral injection of a 28. Chang DK, Sui J, Geng S, Muvaffak A, Bai M, Fuhlbrigge RC, et al. TLR9 agonist combined with radiation: a phase 1/2 study. Blood Humanization of an anti-CCR4 antibody that kills cutaneous T-cell lym- 2012;119:355–63. phoma cells and abrogates suppression by T-regulatory cells. Mol Cancer 44. Pere H, Montier Y, Bayry J, Quintin-Colonna F, Merillon N, Dransart E, Ther 2012;11:2451–61. et al. A CCR4 antagonist combined with vaccines induces antigen-specific 29. Ogura M, Ishida T, Hatake K, Taniwaki M, Ando K, Tobinai K, et al. CD8þ T cells and tumor immunity against self antigens. Blood 2011;118: Multicenter phase II study of mogamulizumab (KW-0761), a defucosy- 4853–62. lated anti-cc chemokine receptor 4 antibody, in patients with relapsed 45. Houot R, Kohrt H, Levy R. Boosting antibody-dependant cellular cytotox- peripheral T-cell lymphoma and cutaneous T-cell lymphoma. J Clin Oncol icity against tumor cells with a CD137 stimulatory antibody. Oncoimmu- 2014;32:1157–63. nology 2012;1:957–8. 30. Rao SP, Sancho J, Campos-Rivera J, Boutin PM, Severy PB, Weeden T, et al. 46. Kohrt HE, Houot R, Goldstein MJ, Weiskopf K, Alizadeh AA, Brody J, et al. Human peripheral blood mononuclear cells exhibit heterogeneous CD52 CD137 stimulation enhances the antilymphoma activity of anti-CD20 expression levels and show differential sensitivity to alemtuzumab medi- antibodies. Blood 2011;117:2423–32. ated cytolysis. PLoS ONE 2012;7:25. 47. Kohrt HE, Houot R, Weiskopf K, Goldstein MJ, Scheeren F, Czerwinski D, 31. Frankel AE, Zuckero SL, Mankin AA, Grable M, Mitchell K, Lee YJ, et al. Anti- et al. Stimulation of natural killer cells with a CD137-specific antibody CD3 recombinant diphtheria immunotoxin therapy of cutaneous T cell enhances trastuzumab efficacy in xenotransplant models of breast cancer. lymphoma. Curr Drug Targets 2009;10:104–9. J Clin Invest 2012;122:1066–75.

OF12 Clin Cancer Res; 21(2) January 15, 2015 Clinical Cancer Research

Reduction of Regulatory T Cells by Mogamulizumab, a Defucosylated Anti-CC Chemokine Receptor 4 Antibody, in Patients with Aggressive/Refractory Mycosis Fungoides and Sézary Syndrome

Xiao Ni, Jeffrey L. Jorgensen, Meghali Goswami, et al.

Clin Cancer Res Published OnlineFirst November 5, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-14-0830

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2015/01/04/1078-0432.CCR-14-0830. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.