JAK/STAT: a pathway through the maze of PTCL?

In this issue of Blood, Moskowitz et al¹ report the result of treatment with ruxolitinib for patients with relapsed/refractory peripheral T-cell lymphoma (PTCL), demonstrating that the JAK/STAT pathway is clinically relevant in T-cell lymphoma.

In the hematology-oncology community, few would argue that some of the biggest therapeutic disappointments in lymphoid malignancies are seen in non–mycosis fungoides PTCL. The reasons for our slow advancement in this area are many, including the rarity of these disorders, their extreme heterogeneity, and their geographic diversity. In the 2017 World Health Organization Classification, there are least 29 different subtypes of PTCL; each a relative challenge to pathologically categorize and each a unique entity.² 

We are, however, making progress through this maze of PTCL, moving away from combination chemotherapy regimens that have been borrowed from the B-cell lymphomas, like cyclophosphamide, doxorubicin, vincristine, prednisone and ifosfamide, carboplatin, etoposide, toward lineage-specific chemotherapy, such as pralatrexate. We are now using in the front line a drug that targets an antigen, CD30, found to a variable degree on PTCL, with improvement in therapeutic outcome. We use therapies that rely on T-cell lymphoma’s unique vulnerability to epigenetic disruption, such as the histone deacetylase inhibitors and hypomethylating agents. In doing so, we are made aware that the treatment of PTCL certainly needs to be lineage specific and will likely become more subtype specific. There is not a “one-size-fits-all” for PTCL. Only some subtypes of PTCL have CD30 overexpression and/or recurrent mutations in genes that govern the epigenome. Certainly, making our way through this maze will not and has not been simple and will require learning more about the malignant T cell and its vulnerabilities.

Signaling through JAK1 and/or JAK2 within the tumor or its microenvironment has been commonly reported in PTCL.³ Given the high rate of refractory and relapsed disease, and the need for more efficacious therapy, Moskowitz et al conducted a biomarker-driven study of ruxolitinib for the treatment of patients with relapsed/refractory PTCL in this multicenter, investigator-initiated, phase 2 trial (#NCT02974647). Ruxolitinib, an oral agent, is a JAK1/2 inhibitor. Fifty-two evaluable patients with a variety of PTCLs were treated with ruxolitinib 20 mg orally twice daily until either progressive disease or unacceptable toxicity. Assessments for response were performed after 56 and 140 days of therapy and then every 84 days. The primary endpoint of the study was assessment of clinical benefit rate (CBR), defined as complete and partial responses, or stable disease that lasted at least 6 months.

Patients were assigned and evaluated in 3 cohorts: cohort 1 included those with known activating JAK and/or STAT mutations; cohort 2 included those with functional activation of JAK/STAT by demonstration of ≥30% pSTAT3 3 expression in tumor cells by immunohistochemistry; and cohort 3, which those with neither genetic or phenotypic evidence of JAK/STAT activation or insufficient tissue for this assessment. Ruxolitinib activity was found to be greater in those patients with PTCL with JAK/STAT activating mutations or active signaling, with cohort 1 having a CBR of 53%; cohort 2 having a CRB of 45%; and cohort 3 having a CRB of 13%. The median time to best response was 6.3 months. Adverse events were consistent with known side effects of this drug and primarily involved cytopenias.

Although an overall CBR of 35% is likely commensurate (although many may agree it is also commiserate) with other approved agents, 8 exceptional responders were identified with responses >1 year. These included 4 of the 5 patients with T-large granular lymphocytosis (T-LGL), the CBR of which was 80%. Of the 8 patients with T-prolymphocytic leukemia (T-PLL), CBR was 50%, all in patients with evidence of JAK/STAT activation. In contrast, only 1 of 7 patients with mycosis fungoides experienced a response to this treatment.

From this study, we learned that an already available oral JAK inhibitor was active against various subtypes of PTCL with acceptable toxicity most often when JAK/STAT pathway activation could be demonstrated. Although lack of JAK/STAT activation appears to predict a reduced response to ruxolitinib in PTCL, the presence of activation was not a guarantee of response, and the lack of activation did not assure there would be no response. This was clearly demonstrated in some cases of T-LGL.

The responses seen in the small number of patients with T-PLL and T-LGL are particularly encouraging. Here, we have 2 T-lymphocytic entities characterized by circulation of malignant cells, cytopenias, and splenomegaly. T-PLL is an aggressive disorder with very limited treatment options. T-LGL is an indolent disorder with treatment options that are successful in approximately half of cases. Of interest, in T-PLL, up to 76% of cases have mutations in JAK1, 2, or 3 or STAT3 or 5B; in LGL, up to 40% of cases have mutations. Although 60% of the patients with LGL are symptomatic, STAT3 mutations have been associated with increased symptoms and shorter time to treatment failure.^4-7 

We look forward to studies that would specifically address these 2 neglected subtypes of PTCL with either expansion of single-agent investigation and/or combination with other pathway inhibitors, such as bcl-2 inhibitors or Pi3 kinase inhibitors. We also anticipate further characterization of the role that JAK/STAT inhibition may play in the treatment of the larger group of patients diagnosed with non–mycosis fungoides PTCL who need improved outcomes.