Mogamulizumab, approved by the FDA for relapsed or refractory mycosis fungoides and Sézary syndrome, improves progression-free survival compared with vorinostat in the largest trial to date in cutaneous T-cell lymphoma, with particular efficacy in leukemic disease, but carries a risk of immune-mediated toxicities with concomitant depletion of regulatory T cells.
See related article by Kasamon et al., p. 7275
In this issue of Clinical Cancer Research, Kasamon and colleagues report on the findings of the MAVORIC trial that led to FDA approval of mogamulizumab for relapsed or refractory mycosis fungoides and Sézary syndrome and provide regulatory insights on the FDA deliberation (1). The MAVORIC trial, the largest randomized controlled trial in cutaneous T-cell lymphoma (CTCL) to date, exemplifies the latest standard set by the FDA in trial design for this disease (2). It is notable for its rigorous assessment of response at more frequent intervals, utilization of a global composite response score, use of a comparator arm, and selection of progression-free survival as the primary endpoint. However, the worse than expected performance of the comparator drug, vorinostat, potential early censoring in the mogamulizumab arm, and data quality problems with the patient-reported outcomes are potential caveats of the study. The modest overall response rate (ORR) in CTCL of 28% reflects the difficulty of showing efficacy in all anatomic compartments captured in the composite response score. This challenge may be overcome if mogamulizumab can be safely combined with other therapies. Immune-related adverse events (irAE) were reported, and the severity, nature of autoimmune events, and long-term impact are unknown and require further study. Finally, mogamulizumab performed better in Sézary syndrome than mycosis fungoides. Biologics that function through antibody-dependent cellular cytotoxicity (ADCC) may have greater efficacy in Sézary syndrome and may be less efficacious in treating non-recirculating (mycosis fungoides) lymphoma cells in skin.
Mogamulizumab, a defucosylated IgG1κ antibody, binds to C-C motif chemokine receptor 4 (CCR4; Fig. 1; ref. 2). Defucosylation of the Fc portion of mogamulizumab specifically enhances ADCC (2). Mogamulizumab does not inhibit CCR4, nor does it induce complement-dependent cytotoxicity (3). CCR4, important for lymphocyte-specific chemotaxis to the skin, is predominately expressed in activated skin tropic and regulatory T cells (Treg; ref. 4). Not surprisingly, T cells in CTCL skin and blood largely express CCR4 (2). MAVORIC provides the largest assessment of CCR4 expression in cutaneous lesions of mycosis fungoides and Sézary syndrome. The receptor was detected in skin biopsies of all 290 evaluable patients (median CCR4 expression = 80%), but skin CCR4 levels did not correlate with response (2). In fact, a significant portion of nonresponders had expression of CCR4 in greater than 80% of lymphocytes in the skin (1). While in the phase I/II trial, CCR4 expression in peripheral blood lymphoma cells did not correlate with response in mycosis fungoides or Sézary syndrome, patients with higher CCR4 expression in circulating malignant T cells had faster clearance in the blood and residual leukemic cells had a lower percentage of CCR4 expression (3, 5). CCR4 expression in leukemic disease may be a more relevant biomarker for response in the blood, which is obfuscated when combined in the composite score.
MAVORIC included patients with mycosis fungoides and Sézary syndrome, the two most common subtypes of CTCL and revealed better responses in Sézary syndrome (ORR 37%) compared with mycosis fungoides (ORR 21%). While included in the same staging paradigm for historical reasons, they are now considered distinct diseases based on clinical presentation, genomic features, and molecular T-cell markers (6, 7). Mycosis fungoides, characterized by fixed, well-defined patches or plaques in early stages, expresses surface markers characteristic of non-recirculating skin resident memory T cells (TRM; refs. 6, 8). Sézary syndrome, characterized by leukemic disease and erythroderma, exhibits markers of central memory T cells (TCM), a recirculating T-cell population (Fig. 1; refs. 6, 8, 9). The different migratory properties of malignant T cells in mycosis fungoides and Sézary syndrome may explain the superior effect of mogamulizumab in leukemic disease, as is the case for alemtuzumab, an anti-CD52 antibody that induces ADCC used in CTCL (8, 9). Following treatment with alemtuzumab, patients with CTCL had depletion of benign and malignant TCM from skin but not TRM, the nonmigratory population (8), suggesting that ADCC is not as effective in skin. Interestingly, there is a striking difference in efficacy of mogamulizumab in the blood (ORR 68%) compared with the skin (ORR 42%; ref. 2). Furthermore, as with alemtuzumab, the time to response in the skin is delayed compared with the blood (2). Taken together, this suggests that mogamulizumab's efficacy in the skin is dependent on malignant T-cell recirculation from the skin to blood. Ineffective ADCC in the skin may also explain the lack of correlation between drug response and cutaneous CCR4 expression. As noted previously, this finding stands in contrast to the correlation between blood clearance and circulating malignant T cell CCR4 expression.
The ability of mogamulizumab to deplete Tregs may serve as another mechanism of drug action independent of lymphoma cell CCR4 expression (3). Most circulating Tregs express CCR4 and many express the skin homing marker CLA (Fig. 1; ref. 4). Because mogamulizumab depletes circulating Tregs, patients with active autoimmune diseases were excluded from MAVORIC. While Tregs are generally associated with poor prognosis in cancer, their role in CTCL is less well understood, as it is possible that Tregs may function not only to dampen the antitumor immune response but also to suppress malignant T cells. For example, elevated levels of tumor-infiltrating Tregs in skin lesions of early stage CTCL have been associated with improved survival (10).
Not surprisingly, given depletion of Tregs by mogamulizumab, irAEs have become evident (Fig. 1). As with immune checkpoint inhibitors (ICI), dermatologic toxicities are the most common irAE (25% of patients), which is particularly challenging in these patients with preexisting skin disease. Up to 18% of patients discontinued therapy due to rash or drug eruption. Other reported irAEs included myositis, myocarditis, polymyositis, hepatitis, pneumonitis, hypothyroidism, and a variant of Guillain-Barré syndrome. In contrast to ICI toxicities, and perhaps related to differential effects in different organ compartments, colitis was notably absent. To date, post market cases of photosensitivity and Steven's Johnson Syndrome/Toxic epidermal necrolysis have been reported with mogamulizumab in patients with T-cell lymphoma. The conceptual framework for management of irAE remains the same as with ICI with a notable exception. While controversial in the management of ICI-induced irAEs in those with solid tumors, corticosteroids are used extensively in the management of CTCL, and we feel high dose corticosteroids should be implemented early and used liberally for life-threatening or severe mogamulizumab toxicity. Importantly, as with ICI-induced irAEs, it appears that these toxicities may persist long after the drug has stopped; awareness of and counselling on delayed and persistent toxicities is critical (11).
The persistence of drug, the attendant Treg depletion, and the potential for delayed irAEs has led to significant concern about severe and refractory graft-versus-host disease (GVHD) after hematopoietic stem cell transplantation (HSCT). HSCT prior to 50 days from last dose carries a considerable risk of severe and potentially fatal GVH that exceeds the mortality from mycosis fungoides/Sézary syndrome (12). While mogamulizumab can be removed via plasma exchange, the recovery of Tregs may not be rapid enough to prevent irAEs, including GVHD. Thus, the greater the interval from mogamulizumab cessation without relapse to transplant, the better. This makes transplant planning and coordination logistically critical, but challenging.
Overall, mogamulizumab is a welcome addition to the therapeutic armamentarium, and is a particularly attractive therapy for patients with leukemic disease. The majority of approved therapies used in CTCL are limited by poor response rates, short duration of response, dose-limiting toxicities, or lack of efficacy across disease compartments. Durable responses are elusive, save for allogeneic HSCT, which carries with it significant morbidity and mortality. There remains an unmet need for many patients with advanced CTCL: a therapy that is relatively well-tolerated and prevents disease progression. Mogamulizumab may help bridge this gap.
Disclosure of Potential Conflicts of Interest
C. Larocca is an employee/paid consultant for Kyowa Kirin. N.R. LeBoeuf is an employee/paid consultant for Bayer and Seattle Genetics. No potential conflicts of interest were disclosed by the other author.
T.S. Kupper is supported by the NIH (R01 CA210372, R01 AI127654, and R01 AR065807).