FDA Approval Summary: Mogamulizumab-kpkc for Mycosis Fungoides and Sézary Syndrome

The prognosis with mycosis fungoides or Sézary syndrome, the main types of cutaneous T-cell lymphoma (CTCL), varies widely and is related to clinical stage (1, 2). Although patients with early-stage mycosis fungoides have a favorable prognosis, mycosis fungoides and Sézary syndrome are usually incurable and can cause significant morbidity as well as shortened survival. Mycosis fungoides primary presents in the skin, with potential involvement of lymph nodes, blood, and viscera. The skin lesions may be localized or widespread, manifesting as patches and plaques, tumors, or erythroderma. Large-cell transformation occurs in approximately 10% of cases and is associated with an estimated median overall survival of 2 years (2). Patients with Sézary syndrome, a closely related, rare, and aggressive leukemic variant of mycosis fungoides, have an estimated median overall survival of 3 years (2). In both diseases, pruritus, skin exfoliation, erosion, and superinfection from constant scratching are common, as are opportunistic infections.

Whereas limited-stage disease is treated with skin-directed therapies or immunomodulatory agents, management of refractory early-stage and advanced-stage disease may also require systemic chemotherapy (1, 3, 4). However, mycosis fungoides and Sézary syndrome tend to be resistant to chemotherapy. Approximately 30%–40% of patients respond to a variety of biologic agents, although deep remissions are rare (Table 1).

In August 2018, the FDA granted regular approval of mogamulizumab-kpkc (POTELIGEO; Kyowa Kirin, Inc.) for the treatment of adult patients with relapsed or refractory mycosis fungoides or Sézary syndrome after at least one prior systemic therapy. We summarize the FDA clinical review and regulatory insights on this licensing application.

Mogamulizumab-kpkc (hereafter, mogamulizumab) is a humanized, IgG1 kappa mAb that is directed against CC chemokine receptor type 4 (CCR4). CCR4 is a G protein–coupled receptor for CC chemokines that is implicated in lymphocyte trafficking to the skin and sites of skin inflammation (5, 6). CCR4 expression by nonmalignant cells is primarily restricted to regulatory T cells and type 2 T helper (Th2) cells (5, 7–9). CCR4 is also overexpressed by some T-cell malignancies, including the majority of cases of mycosis fungoides and Sézary syndrome (10–13). In mycosis fungoides, a higher frequency of CCR4 expression has been reported with advancing stage (14). Binding of mogamulizumab to CCR4-expressing cells may potentiate antibody-dependent cellular cytotoxicity (ADCC) and depletion of the target cells (15). Because its Fc region is defucosylated, mogamulizumab may elicit more potent ADCC than traditional glycosylated antibodies (10).

With a terminal half-life of 17 days, steady-state concentrations of mogamulizumab are reached after 8 doses (12 weeks). Clinically significant changes in pharmacokinetics were not observed in patients with renal impairment or mild or moderate hepatic impairment. The pharmacokinetic effect of severe hepatic impairment is not defined (16).

In patients with mycosis fungoides or Sézary syndrome, the recommended dose of mogamulizumab is 1 mg/kg infused intravenously over at least 60 minutes on days 1, 8, 15, and 22 of the first 28-day cycle, then on days 1 and 15 of each subsequent cycle until disease progression or unacceptable toxicity (16). Premedication with diphenhydramine and an antipyretic is advised for the first infusion to mitigate infusion-related reactions (16).

The basis of approval is a multicenter, open-label, randomized phase III trial (Study 0761-010; NCT01728805) in adults with relapsed or refractory, stage IB-IV mycosis fungoides or Sézary syndrome after at least one prior systemic therapy (17). The trial randomized 372 patients with a 1:1 ratio to either mogamulizumab at the approved dose-schedule or vorinostat 400 mg once daily. Treatment continued until progressive disease (PD) or unacceptable toxicity, with cross-over permitted from vorinostat to mogamulizumab.

The trial enrolled patients regardless of tumor CCR4 expression status and excluded patients with active histologic transformation, prior allogeneic hematopoietic stem cell transplantation (HSCT), and active autoimmune disease. Requirements included an Eastern Cooperative Oncology Group performance status <2, absolute neutrophil count ≥1,500/μL (≥1,000/μL with bone marrow involvement), platelet count ≥100,000/μL (≥75,000/μL with bone marrow involvement), adequate renal function, and hepatic transaminases ≤2.5 times the upper limit of normal in the absence of liver involvement.

The primary endpoint was progression-free survival (PFS) per investigator, measured from randomization until death or documented progressive disease (PD). The trial was designed to detect a 50% improvement in median PFS from 5.6 months with vorinostat to 8.4 months. Secondary endpoints included overall objective response rate (ORR) per INV based on the global composite response score (18), efficacy per independent review committee (IRC), and patient-reported outcomes. All reported P values are two-sided. The response criteria combine measures from each disease compartment (skin, blood, lymph nodes, and viscera; ref. 18). Responses required confirmation at two successive disease assessments performed ≥4 weeks apart, which included the modified Severity Weighted Assessment Tool (mSWAT), skin photographs, central flow cytometry, and computed tomography.

Table 2 summarizes the characteristics of the intention-to-treat efficacy population. In the mogamulizumab arm (n = 186), 56% of patients had mycosis fungoides and 44% had Sézary syndrome, with a similar distribution in the vorinostat arm (n = 186). The patients enrolled had a median of 3 prior systemic therapies, with 81% having at least 2 prior systemic therapies and 40% having at least 4. At study baseline, 38% of all patients had stage IB-II disease (11% stage IB), 10% stage III, and 52% stage IV. Two-thirds of patients had blood and/or nodal involvement, whereas visceral involvement was rare.

CCR4 expression status, as determined by skin biopsy IHC, was available in 290 patients (88%), of whom all had CCR4 expression in ≥1% of lymphocytes and 97% had CCR4 expression in ≥10% of lymphocytes. The median CCR4 percentage was 80% (range 1–100).

During randomized treatment, the median duration of exposure was two times longer in the mogamulizumab arm (5.6 months) than the vorinostat arm (2.8 months), with 48% and 22% of patients, respectively, having at least 6 months of exposure (Table 2). After a median of 2.4 months, 73% of the vorinostat arm crossed over to mogamulizumab because of inefficacy (80%) or side effects (20%).

Investigator-assessed PFS was statistically significantly longer in the mogamulizumab arm, with an estimated median PFS of 7.6 months compared with 3.1 months in the vorinostat arm (Table 3; Fig. 1). The corresponding HR was 0.53 (95% CI: 0.41–0.69; P < 0.001 based on a stratified log-rank test). The results of a blinded independent review were supportive. The estimated median PFS per IRC was 6.7 months (95% CI: 5.6–9.4) with mogamulizumab and 3.8 months (95% CI: 3.0–4.7) with vorinostat (HR = 0.64; 95% CI: 0.49–0.84).

In these analyses, patients with clinical progression that did not meet formal PD criteria were censored at the time of clinical progression. A sensitivity analysis of investigator-assessed PFS, treating clinical progression as an event, was consistent with the primary analysis (HR = 0.61; 95% CI: 0.47–0.78).

Table 3 summarizes response rates overall and by disease compartment. The investigator-assessed ORR [confirmed complete response (CR) or partial response (PR)], a key secondary efficacy endpoint, was statistically significantly higher in the mogamulizumab arm (28%) than the vorinostat arm (5%; P <0.001). The estimated median duration of response was 13.9 months and 9.0 months, respectively. Most responses to mogamulizumab were partial, with 25% of patients achieving a best overall response of PR and 2% achieving CR. Similarly, by IRC assessment, the ORR was 23% (95% CI: 17–30) with mogamulizumab and 4% (95% CI: 2–8).

A higher ORR with mogamulizumab than vorinostat was observed in both the mycosis fungoides and Sézary syndrome subgroups (Table 3). The activity of mogamulizumab appeared greater in the blood and skin than in lymph nodes. There were too few patients with visceral involvement to inform activity in that compartment (Table 3). Responses to mogamulizumab occurred across a wide range of CCR4 expression levels (Supplementary Fig. S1).

The U.S. prescribing information for mogamulizumab includes warnings and precautions for dermatologic toxicity (one of the most common toxicities of mogamulizumab), infusion reactions, infections, autoimmune complications, and complications of allogeneic HSCT, including severe and refractory graft-versus-host disease (GvHD).

After randomized treatment with mogamulizumab (n = 184), adverse reactions (ARs) reported in at least 20% of patients included rash, infusion-related reaction, fatigue, diarrhea, musculoskeletal pain, and upper respiratory tract infection. Other common ARs (incidence at least 10%) included skin infection, pyrexia, nausea, edema, thrombocytopenia, headache, constipation, mucositis, anemia, cough, and hypertension. ARs led to treatment discontinuation in 18% of patients, with the leading cause being rash or drug eruption followed by infection. Serious ARs occurred in 36% of patients, most often from infection (16% of all patients), including pneumonia, lower respiratory tract infection, sepsis, and skin infection.

Randomized treatment with mogamulizumab carried a safety profile similar to crossover treatment after vorinostat. On pooled analysis of the 319 mogamulizumab recipients in this study, 2.2% had fatal ARs within 90 days of the last mogamulizumab dose. Infusion-related reactions occurred in 35% of patients, in most cases occurring during or shortly after the first infusion, and were severe in 8% of the cases. Drug eruption, with a variable onset (median 15 weeks) and appearance, occurred in 25% of patients, with 18% of the cases being severe. Use of systemic immunosuppressants for immune-mediated ARs was reported in 1.9% of patients. Immune-mediated or possibly immune-mediated ARs after mogamulizumab included myositis, myocarditis, polymyositis, hepatitis, pneumonitis, hypothyroidism, and a variant of Guillain–Barré syndrome.

In patients with relapsed or refractory mycosis fungoides or Sézary syndrome after systemic therapy, this randomized phase III clinical trial established the efficacy of mogamulizumab relative to vorinostat, a widely used salvage regimen. Efficacy was based on investigator-assessed PFS and confirmed ORR using global composite response criteria, with supportive data from a blinded, independent review. On the basis of these results, mogamulizumab received Breakthrough Therapy Designation for its approved indication.

Study 0761-010 is one of the largest randomized, actively controlled trials reported in CTCL. Table 1 provides an overview of approved systemic therapies for CTCL. With the exception of the mogamulizumab and brentuximab vedotin, these drugs received regular approval on the basis of single-arm studies, with ORR determined primarily or solely in the skin compartment. More recently, the FDA has recommended a randomized clinical trial with a PFS endpoint as the initial registration approach in mycosis fungoides and Sézary syndrome. The more recent approvals have also been tailored to the specific subtypes of CTCL that were represented in the trials, with efficacy determinations by an IRC as well as by investigators In clinical trials, response assessment in the skin can pose multiple challenges, such as the subjective component of the assessment, inter-reader variability, adequacy of response documentation (including photographs), susceptibility to bias, and distinguishing dermatologic toxicity from the underlying disease. In light of these issues and the open-label design of Study 0761-010, the randomized trial design and IRC determinations of efficacy were important in supporting the regular approval of mogamulizumab.

From an efficacy standpoint, the poor performance of vorinostat is notable. Although cross-study comparisons are limited, the confirmed ORR of 5% contrasts sharply with the 30% ORR reported previously in single-arm trials (Table 1; refs. 19, 20). Compliance issues seem unlikely to explain the low activity of vorinostat, as the reported mean relative dose intensity of vorinostat was 89%. In addition to differences in patient and disease characteristics, a potential explanation is the use of more stringent response criteria in Study 0761-010, which required assessment of all disease compartments and persistence of response for at least 4 weeks. Inadequate exposure to vorinostat is also possible in some cases, given the generally short exposure to vorinostat prior to crossover to mogamulizumab (median 2.4 months; Table 2). Whereas the median time to response to vorinostat is reportedly 8 weeks (range 4 to 24 weeks) in the skin compartment (19), approximately 40% of patients on the vorinostat arm completed less than 2 months of treatment, raising the question of premature crossover.

A limitation of this study is the amount of early censoring for PFS and duration of response. In the mogamulizumab arm, of the 41% of patients censored for investigator-assessed PFS, 47% were censored less than 6 months after randomization. Because of the relatively short follow-up, the estimate of median PFS may be unstable. The censoring was driven by clinical progression that did not meeting formal PD criteria, which in turn raised concern for premature, informative censoring. However, as discussed in the results, sensitivity analyses for PFS were consistent with the main analysis. Meaningful conclusions could not be drawn about overall survival because of the cross-over design and need for longer-term follow-up.

Assessments of patient-reported outcomes were key secondary or exploratory endpoints in this study. Comprehensive assessments of patient-reported outcomes are needed in trials of CTCL, particularly because of the substantial impact of disease-related symptoms on quality of life. Although some metrics of quality of life improved in the mogamulizumab arm, the assessments were hampered by missing data, the choice of instrument (types of questionnaires and scales), and the open-label study design. Therefore, the patient-reported outcome data were not sufficient to support a labeling claim for efficacy.

In considering the indication statement for mogamulizumab, the FDA team deliberated whether to restrict the indication to CCR4-expressing cases of mycosis fungoides and Sézary syndrome. Although CCR4 is expressed in almost all cases of adult T-cell leukemia, its expression in mycosis fungoides and Sézary syndrome, while frequent, appears less consistent in the literature (10, 11, 14). Several published studies have also described less frequent CCR4 expression in earlier stages of mycosis fungoides (14). Cross-study comparisons of CCR4 expression levels are limited by the variable CCR4 detection methods, by patient heterogeneity, and by sample size. However, Study 0761-010 provides one of the largest sets of data available on CCR4 expression patterns in mycosis fungoides and Sézary syndrome. In this study, all 290 evaluable patients had detectable CCR4 expression by skin biopsy. Efficacy data in low-CCR4 expressing mycosis fungoides and Sézary syndrome are limited. However, the minimum CCR4 expression level necessary for the activity of mogamulizumab is not determined, and the trial enrolled patients regardless of CCR4 expression status. FDA therefore granted an indication that is not restricted to CCR4-expressing disease. Of note, the protocol-specified threshold for “positive” CCR4 expression was defined as ≥10% of lymphocytes and was selected to reduce false-positive results. However, this threshold is not based on any predictive correlation associated with a <10% versus ≥10% expression level.

The safety profile of mogamulizumab is acceptable for the approved population, with an overall favorable benefit/risk balance. However, particular vigilance is needed for dermatologic toxicity, infusion-related reactions, serious infections, and immune-mediated complications. The dermatologic toxicity of mogamulizumab necessitates monitoring throughout the treatment course. It has a variable onset, with 25% of cases occurring after 7 months, does not have characteristic pattern, and may be challenging to differentiate from disease progression in the skin.

Immune-mediated complications after mogamulizumab, including but not limited to the allogeneic HSCT setting, are biologically plausible because mogamulizumab can reduce levels of CCR4-expressing T-regulatory cells (9, 11). Because of this theoretical concern, the FDA review team identified the safety signal of autoimmune complications after mogamulizumab. The incidence of such adverse reactions may be underreported. Importantly, the safety of allogeneic HSCT after receipt of mogamulizumab is not defined, nor is the safety of giving mogamulizumab after allogeneic HSCT. There is a consistent signal in the literature that pretransplantation mogamulizumab may increase risk of posttransplant complications, including severe GVHD, steroid-refractory GVHD, and possibly nonrelapse death (21–23). Among recipients of pretransplantation mogamulizumab, a shorter interval between the last mogamulizumab dose and the transplant has been associated with a higher risk of transplant complications (21, 23). Thus, as a postmarketing safety requirement, FDA has requested that the sponsor conduct a study to further characterize the safety of allogeneic HSCT after mogamulizumab.

On the basis of statistically significant improvements in PFS and confirmed ORR compared with vorinostat, the approval of mogamulizumab-kpkc provides an alternative treatment option for patients with relapsed or refractory mycosis fungoides and Sézary syndrome despite at least one prior systemic therapy.

No potential conflicts of interest were disclosed.

The Editor handling the peer review and decision-making process for this article has no relevant employment associations to disclose.

Conception and design: Y.L. Kasamon, L. Nie, R. Pazdur

Development of methodology: R. Pazdur

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y.L. Kasamon, R. Pazdur

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y.L. Kasamon, H. Chen, R.A. de Claro, L. Nie, J. Ye, G.M. Blumenthal, A.T. Farrell, R. Pazdur

Writing, review, and/or revision of the manuscript: Y.L. Kasamon, H. Chen, R.A. de Claro, J. Ye, G.M. Blumenthal, A.T. Farrell

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): G.M. Blumenthal, R. Pazdur