The Immunopathogenesis and Immunotherapy of Cutaneous T Cell Lymphoma: Part II, Current and Future Approaches

Abstract

In the past few decades, immunotherapy has emerged as an effective therapeutic option for patients with cutaneous T cell lymphoma (CTCL). CTCL is characterized by progressive impairment of multiple arms of the immune system. Immunotherapy targets these deficits to stimulate a more robust anti-tumor response, thereby both clearing the malignant T cells and repairing the immune dysfunction. By potentiating rather than suppressing the immune system, immunotherapy can result in longer treatment responses than alternatives such as chemotherapy. In recent years, advances in our understanding of the pathogenesis of CTCL have led to the development of several new agents with promising efficacy profiles. In part II of this review, we describe the current immunotherapeutic options for treatment of CTCL, with a focus on how they interact with the immune system and their treatment outcomes in case studies and clinical trials. In particular, we will discuss established CTCL immunotherapies such as interferons, photopheresis, and retinoids; emerging therapies such as interleukin-12 and TLR-agonists; and new approaches to targeting tumor antigens and checkpoint molecules such as mogamulizumab, anti-PD1, anti-CD47, and CAR-T therapy. We also describe the principles of multimodality immunotherapy and the use of TSEBT in such regimens.

Established Therapies

IFN-alpha

Interferons (IFN) are among the first immune potentiating agents successfully used to treat CTCL. Both IFN-alpha and IFN-gamma are cytokines produced as part of the innate immune response, and their multifold immunomodulatory effects have been harnessed to combat a variety of malignancies, including CTCL.

IFN-alpha, produced by plasmacytoid dendritic cells (DCs), targets several of the immune deficits present in CTCL, helping to bolster an antitumor response. Specifically, IFN-alpha has been found to stimulate CD8+ T cells and NK cells, thereby activating antitumor cytotoxicity.¹ Moreover, IFN-alpha may upregulate MHC class I expression on malignant lymphocytes.² IFN-alpha can also blunt the excess TH2 production of IL-4 and IL-5 by malignant T cells, helping to restore the host TH1/TH2 balance.^3–5 While these effects have proven useful in inducing disease regression and control, the development of resistance may occur.⁶ Proposed mechanisms of resistance include the production of neutralizing antibodies, decreased expression of IFN receptors, and mutations and deletions of STAT1.^7–11 Importantly, IFN-alpha should be used with care in CD8+ CTCL, and not used at all for gamma/delta T cell lymphomas or for the initial phase of treating panniculitic T cell lymphomas, as these conditions are typically associated with cytotoxic T cells that release proinflammatory cytokines.

While there are few randomized controlled trials using IFN-alpha for CTCL, both retrospective and prospective studies have described impressive responses to this agent. Importantly, there are several forms of IFN-alpha: recombinant IFN-alpha-2a and alpha-2b, as well as pegylated IFN-alpha-2a and alpha-2b. Most studies have evaluated recombinant IFN-alpha, but both recombinant and pegylated IFN-alpha may be used to treat CTCL and anecdotal reports suggest higher response rates with pegylated forms. One prospective analysis by Papa et al. observed that 80% of early stage patients and 70% of patients with Sézary syndrome responded to IFN-alpha.¹² A trial by Olsen et al. similarly discovered that 59% of patients of all stages responded to the drug after 10 weeks of treatment.¹³ This same study found a dose-responsive curve indicating a possible advantage of higher dosage strategies when patients are resistant to lower doses.¹³ IFN-alpha has also demonstrated value in multimodality regimens. In particular, IFN-alpha has been shown to augment clinical response rates when added to PUVA. ⁶Smaller scale studies also indicate possible efficacy of IFN-alpha when combined with oral retinoids, photopheresis, total skin electron beam therapy (TSEBT), and nbUVB.^4,6

IFN-alpha is generally tolerable at low doses, but various adverse effects may occur. Shortly after initiation of treatment, patients may develop flu-like symptoms that typically diminish and usually resolve within a few weeks. Potential chronic complications include fatigue, weight loss, depression, and impaired cognitive function.¹⁴ Myelosuppression may also occur, which usually does not require discontinuation if neutrophils remain above 1000/μl. Less common adverse effects include hypothyroidism, peripheral neuropathy, visual impairment, cardiac toxicity with depressed ejection fraction, and worsening of autoimmune disorders.^15,16

IFN-gamma

IFN-gamma, produced mainly by CD8+ T cells and NK cells, enhances cell-mediated cytotoxicity, antigen presentation, and TH1 immunity. IFN-gamma, like IFN-alpha, activates CD8+ T cells and NK cells. Importantly, IFN-gamma can prime and activate DCs and macrophages, thus pairing well with proapoptotic agents by enhancing processing of released tumor antigens. As the prototypical TH1 cytokine, IFN-gamma also counteracts the global TH2 bias that commonly suppresses cytotoxic immunity in CTCL.¹⁷ Recently, a prospective study found that 11 of 15 MF patients had a partial response to IFN-gamma treatment.¹⁸ IFN-gamma has also been described as useful in combination with other treatments, including ECP and bexarotene.¹⁸ However, IFN-gamma is less well-studied than IFN-alpha and more large-scale studies are needed to evaluate IFN-gamma for treatment of CTCL.

The side-effects of IFN-gamma and IFN-alpha are similar, but the former is less likely to cause cognitive decline, depression, and neuropathy, making it a more tolerable drug for older individuals.

Photopheresis

Following a successful clinical trial for advanced CTCL, photopheresis was FDA approved for the treatment of advanced refractory CTCL in 1987.¹⁹ This treatment involves leukapheresis every 2–4 weeks with separation of approximately 2% of the circulating leukocytes, which are exposed to 8-methoxypsoralen and ultraviolet A irradiation.¹⁹ This process leads to apoptosis and possibly necrosis of malignant lymphocytes, which are then reinfused into the patient.²⁰ As monotherapy, a partial response rate (> 50% skin improvement) of 33% and a near-complete response rate (> 90% skin improvement) of 41% have been reported.²¹ The most responsive population includes those with stages IIIB-IVA disease, early in their course and with adequate numbers of circulating DCs and CD8+ T cells. These patients are more likely to have the capacity to generate an effective immune response to the antigens released by dying malignant cells.

Importantly, concomitant therapy with immune potentiating agents has demonstrated useful synergy with photopheresis. IFN-alpha, due to its ability to activate NK and CD8+ T cells, is often used with photopheresis. Moreover, IFN-gamma has the ability both to prime antigen presenting cells, which are critical for the afferent response, and to stimulate NK and CD8+ T cells. Wysocka and colleagues demonstrated that subcutaneous GM-CSF injections can increase the numbers of circulating DCs, which are characteristically diminished in SS.²² In mice, studies have demonstrated that administering an excessively large number of apoptotic cells may suppress DC function, further supporting the use of DC-potentiating cytokines with photopheresis.²³ Additional therapies that can be useful to pair with photopheresis include oral retinoids, PUVA therapy, and low dose (12 Gy) total skin electron beam therapy.

Photopheresis is very well-tolerated and few adverse effects have been reported.^24,25 Among erythrodermic patients, it is recommended that peripheral vascular access be obtained and that external Hickman catheters be avoided as there has been a high frequency of S. aureus contamination.

Retinoids

Retinoids are another standard therapeutic option that at least partially target the immune dysregulation of CTCL. Retinoids act through retinoid X receptors (RXRs) and retinoic acid receptors (RARs). When bound, RXRs and RARs can form heterodimers with other nuclear receptors to act as transcription factors, resulting in multiple beneficial immunomodulatory effects.

Bexarotene is a synthetic rexinoid that selectively binds RXRs, and it has demonstrated usefulness and safety in all stages of CTCL. Though the full effects of bexarotene in CTCL have not been elucidated, it notably induces apoptosis of malignant T cells.²⁶ Moreover, bexarotene inhibits production of TH2 cytokines that are upregulated in CTCL.²⁶ Bexarotene also inhibits skin trafficking of malignant T cells by reducing CCR4 expression and chemotaxis.^26,27 Resistance has been noted in many patients, possibly through loss of RXR expression on malignant T cells.²⁸ Use of bexarotene in combination with other agents may decrease the frequency of resistance and optimize treatment outcomes. The most frequent side-effects of bexarotene are hyperlipidemia and hypothalamic hypothyroidism, which occur in the majority of patients.^29,30 Thus, the monitoring of thyroid hormone replacement should be made by utilizing the serum free T4 as serum levels of TSH are depressed by bexarotene.

RAR retinoids, particularly isotretinoin, may provide added benefit in all stages of CTCL. There are several settings in which they are especially useful, including the treatment of advanced folliculotropic mycosis fungoides and suppression of new squamous cell cancers among patients with this additional issue.³¹ In contrast, there is no evidence that bexarotene can slow squamous cell skin cancer growth.

New and Emerging Therapies

IL-12

Interleukin-12 (IL-12) is a 70 kD heterodimeric protein produced by many cell types within the immune system. Notably, like IFN-alpha and IFN-gamma, it is considered a product of the innate immune response, but it represents a critical link between innate and adaptive T cell immunity. Monocytes and, particularly, DCs, are the most significant producers of IL-12. This cytokine can potently activate NK cells and CD8+ T cells with the subsequent production of IFN-gamma. IFN-gamma then enforces a TH1 immune response, which is critical for the development of a robust anti-tumor immune response.

Studies of immune dysregulation in CTCL found that the majority of advanced stage patients, particularly those with SS, exhibited profound defects in TH1 immunity associated with reduced numbers of circulating DCs and markedly diminished production of IFN-gamma and IL-12.^22,32 As there is a bias towards TH2 immunity with increased production of IL-4, IL-13 and often IL-5, it was notable that IL-12 exhibited the ability to suppress TH2 cytokine production from SS patients’ circulating mononuclear cells.^3,32 These observations served as the impetus for the initiation of several successful clinical trials with recombinant IL-12.^33,34,35,36 As IL-12 is presently not in clinical development for CTCL, details of the use of this potentially valuable and clinically effective cytokine are beyond the scope of this review.

TLR-agonists

IFN-alpha, IFN-gamma and IL-12 are all products of innate immune activation and have each demonstrated significant clinical efficacy for CTCL. Therefore, another logical immunotherapeutic approach for CTCL is direct triggering of innate immunity via Toll-like receptor (TLR) agonists. Stimulation of TLR-7 or 9 on plasmacytoid DCs results in the production of IFN-alpha while activation of TLR-8 on myeloid DCs results in the release of IL-12, TNF-alpha and other immune potentiating cytokines, including IFN-gamma. It was hypothesized by Suchin et al. (2002) nearly 20 years ago that the application to active CTCL skin lesions of imiquimod cream, a TLR-7 agonist, should prove beneficial through the induction of the local release of IFN-alpha.³⁷ Indeed, the initial patient experienced a complete clinical response.³⁷ Subsequent studies have demonstrated mixed results, although some series have produced high response rates. One series of 20 early-stage patients treated with imiquimod had an overall response rate of 80%, and 45% had a complete response.³⁸

A confounding issue with imiquimod is low bioavailability. Persistent, long term use typically produces greater efficacy. Furthermore, increasing the amount applied to a larger surface area can be quite helpful, although this is limited by the drug being dispensed in 0.25 g packages. Quite importantly, the simultaneous use of topical steroids at the same sites of application of imiquimod can eliminate the beneficial effects as the functions of DCs are inhibited by corticosteroids.

Because myeloid DCs, which express TLR-8 but not TLR-7, populate the skin in greater numbers than TLR-7-expressing plasmacyotoid DCs, there is rationale for employing a TLR-8 agonist.³⁹ In that regard, a phase I trial of resiquimod gel, a TLR-7/8 agonist, for early stage disease conducted by Rook and colleagues (2015) resulted in high response rates among 12 treated patients.³⁹ All but one patient experienced significant improvement in skin lesions. Two patients experienced long term complete remission. Using high-throughput T cell receptor sequencing analysis of skin DNA samples from serial skin biopsies of treated lesions, the malignant T cell clone appeared to be eradicated from 4 of 10 patient samples and significantly reduced in another 5 patient samples. Moreover, a common finding during 8 weeks of topical therapy was intense infiltration of treated lesions with CD8+ T cells that were expressing granzyme and IFN-gamma. In addition, non-treated distant lesions quite commonly regressed, clearly indicating that resiquimod has the potential for an abscopal effect.

Resiquimod also has the ability to activate circulating cells. Many patients treated in the phase I trial exhibited activation of both circulating DCs and NK cells. The ability to produce systemic immune activation with a topical agent that can induce the release of both IFN-alpha and IL-12 as well as produce an abscopal effect is likely to be a valuable addition to the profile of immunotherapy drugs available for CTCL.

Limited trials of systemically administered TLR-agonists have also been undertaken for CTCL. Cytosine-guanine oligodeoxynucleotides (CpG-ODN), which bind to TLR-9, and thereby activate plasmacytoid DCs, were administered subcutaneously in increasing doses in a phase I trial for advanced refractory CTCL.⁴⁰ A 32% response rate was observed. Considering that the maximal tolerated dose was not reached, these initial results appear to be quite promising.

Common side effects of imiquimod include skin irritation, inflammation, and itchiness, although one case report describes the development of flu-like symptoms.^41,42 Resiquimod is similarly well-tolerated and has been found in clinical trials to generally cause mild adverse effects such as skin irritation.³⁹

Targeting of Tumor Antigens and Checkpoint Molecules

At present, the only unique tumor marker identified on Sézary cells appears to be the specific T cell receptor of the malignant T cell characterized by a unique T cell receptor gene rearrangement. Because these vary greatly from patient to patient, targeting the T cell receptor is considered to be a highly complex approach. However, there are certain cell surface molecules which tend to be highly expressed on the malignant T cell population, including CCR4, PD-1, and CD47, which could provide for a common targeting approach.

Mogamulizumab

Mogamulizumab is an anti-CCR4 antibody that is FDA approved for treatment of relapsed or refractory CTCL. By targeting CCR4, which is overexpressed in the tumor cells of many patients with CTCL, mogamulizumab mediates elimination of malignant T cells by antibody-dependent cell-mediated cytotoxicity (ADCC). Moreover, mogamulizumab has the capacity to remove CCR4+ Tregs, which can counteract Treg-mediated immune suppression.⁴³ This therapy appears to be most efficacious in the setting of leukemic CTCL prior to the development of bulky lymph nodes or skin tumors.⁴⁴ More recent anecdotal reports have described using mogamulizumab in a multimodality approach with interferon, which is known to enhance ADCC, as well as with photopheresis.

The most common adverse effects of mogamulizumab in the pivotal clinical trial were infusion reactions, drug rashes, diarrhea, and fatigue.⁴⁴ More severe adverse events such as pyrexia and cellulitis occurred in 41% of patients. There were two treatment-related deaths due to sepsis and polymyositis.

Anti-PD1

Programmed cell death protein 1 (PD-1) is expressed by activated T cells, on which it mediates T cell exhaustion and immune tolerance. It is often highly expressed on Sézary cells.⁴⁵ Binding of PD ligand 1 (PD-L1) to PD-1 on T cells limits T cell proliferation and function and may produce T cell “anergy.” A phase II trial using pembrolizumab in 24 patients with advanced refractory CTCL had an overall response rate of 38%, including two complete responses.⁴⁶ Responses were long-lasting with a median duration not reached at trial conclusion. Nivolumab has also demonstrated efficacy in early studies and case reports, including a phase I trial with an overall response rate of 15% among 13 MF patients as well as a case report of a favorable response in a patient with refractory SS.^47,48 Currently, studies are underway to better evaluate the safety and efficacy of combining PD-1 inhibitors with other therapies such as IFN-gamma. The side effects of anti-PD1 treatment in CTCL are largely similar to those documented in treatment of other malignancies.⁴⁶

Anti-CD47

CD47 is a glycoprotein expressed on many normal cells that inhibits phagocytosis by macrophages via interaction with SIRP-alpha on the macrophage. On Sézary cells, the expression of CD47 is significantly increased. In recent studies by Akilov and colleagues, a decoy receptor called TTI-621 (Signal regulatory protein αFc; SIRPαFc) has been administered intravenously to patients with Sézary syndrome, resulting in a marked decline of circulating Sézary cells.⁴⁹ Observed side effects of anti-CD47 therapy have been minor, including fatigue, chills, decreased appetite, headache, and pruritus.⁵⁰ Because TTI-621 mediates a prominent component of its effect via ADCC, agents that enhance this arm of cell-mediated immunity, such as IFN-alpha, can prominently increase in vitro destruction of the malignant T cells.⁵¹

Chimeric Antigen Receptor T cells

The success of chimeric antigen receptor (CAR) T cells in B cell malignancies has raised expectations for these therapies in other cancers, including CTCL.^52,53 A CAR is a synthetic surface receptor that combines extracellular antibody fragments for target cell recognition with intracellular T cell activation and co-stimulation domains.⁵⁴ Patient T cells are derived from leukapheresis products, transduced with the CAR gene, and re-infused into the patient where they selectively lyse antigen-positive tumor cells.

For CTCL, one CAR-T strategy is targeting universal T cell antigens.^55,56 Initially, the expression of CARs specific for the pan-T antigens CD3- and CD7 led to the fratricide of CAR-T cells, precluding their ex vivo manufacturing.^57,58 However, deleting these antigens from CAR-T cells can eliminate fratricide and allow creation of a functional product.^57,58 Interestingly, a CAR targeting CD5 did not lead to CAR-T fratricide.⁵⁹ CD5- and CD7-directed CAR-T cells are now being evaluated in clinical trials for T cell malignancies (NCT03081910, NCT03690011).

More tumor-restricted antigens targeted with CAR-T have included CCR4, CD4, and CD30. A CAR targeting CCR4 eradicated xenografted CTCL and ATL tumors in mice.⁶⁰ The targeting of CD4 led to fratricide of CD4+ CAR-T cells, but the CD8+ CAR-T cells remained and may have therapeutic potential.⁶¹ CD30-directed CAR-T has produced impressive responses in clinical trials for Hodgkin lymphoma (HL), and this strategy is promising for CD30+ CTCL.^55,62 The targeting of the clonal malignant TCR is another intriguing strategy. One approach is to use CARs specific for the TCR beta chain constant regions TCRB1 or TRCB2. As T cells express only one of these two genes, this may target tumor cells while preserving the benign T cell repertoire.⁵⁶ This approach is being evaluated in a Phase I/II clinical trial for T-Non-Hodgkin lymphomas (NCT03590574).

Multimodality Immunotherapy:

At our own and other cutaneous lymphoma programs, multimodality immunotherapy regimens have become the standard of care for advanced CTCL. In the multimodality approach, we usually “layer on” additional therapies rather than replacing a single therapy that does not work. For example, we often initially treat patients with stage IVA SS with a combination of IFN-alpha, photopheresis, and skin-directed treatments. If a poor response occurs during the first four months, we typically add bexarotene and sometimes IFN-gamma as well. Many of our patients experience significant clinical improvement with this regimen. If they do not, we may add mogamulizumab or low-dose (12 Gy) TSEBT. Through this approach, we have recently learned that addition of low-dose TSEBT to multimodality immunotherapy may lead to long-term remissions for some patients with refractory SS.⁶³ Only after failure of this immunostimulatory approach do we consider histone deacetylase inhibitors, pralatrexate, brentuximab, alemtuzumab, and other accepted therapies that blunt the immune response. In the setting of rapidly progressive disease or in preparation for stem cell transplantation, chemotherapeutics are considered an acceptable approach.

Conclusion

In part II, we reviewed the major immune augmenting therapies available for CTCL. Among established therapies, interferons, photopheresis, and retinoids have reliably demonstrated efficacy in the treatment of CTCL. New immunotherapeutic therapies, such as IL-12, TLR-agonists, monoclonal antibodies, and CAR-T, have displayed promise in clinical trials or are in development. Unlike immunosuppressive alternatives, these immune-potentiating modalities do not raise the risk of serious infection or hinder the ability to induce long-term remissions. It is recommended that immunotherapies be employed in the initial management of advanced CTCL and in successive combination with each other for optimal outcomes.

Table 1:

Immunotherapy options for CTCL – summary of mechanisms.

Therapy	Mechanism
Established Therapies
IFN-alpha	Enhances cell-mediated cytotoxicity Decreases TH2 production by tumor cells Inhibits malignant T-cell proliferation
IFN-gamma	Enhances cell-mediated cytotoxicity Decreases TH2 production by tumor cells Enhances dendritic cell function Inhibits malignant T-cell growth
Photopheresis	Apoptosis of tumor cells
Retinoids (RXRs)	Apoptosis of tumor cells Decreases TH2 production by tumor cells Inhibits CCR4 expression & skin trafficking
New & Emerging Therapies
IL-12	Enhances cell-mediated cytotoxicity Activates IFN-gamma production Decreases TH2 production by tumor cells Enhances dendritic cell function
TLR-agonists	TLR-7/9: activates IFN-alpha production TLR-8: activates IL-12, TNF-alpha, IFN-gamma production
Targeting Tumor Antigens & Checkpoint Molecules
Mogamulizumab	Antibody targeting of tumor cells via CCR4 Depletes CCR4+ Tregs
Anti-PD1	Immune checkpoint inhibition
Anti-CD47	Antibody targeting of tumor cells via CD47 Enhances phagocytosis by macrophages
CAR-T	T-cell targeting of tumor cells

IFN, interferon. RXRs, retinoid X receptors. IL-12, interleukin-12. TLR, toll-like receptor. PD1, programmed cell death protein 1. CAR-T, chimeric antigen receptor T cells.