The efficacy and safety of the JAK 1/2 inhibitor ruxolitinib in TCL

The pursuit of molecular targets in T-cell lymphomas (TCL) is vital, given their heterogeneous nature and lack of effective treatment in both the frontline and relapsed/refractory settings. Upregulation of JAK/STAT signalling from various JAK-activating mutations and via the tumor microenvironment provides a logical avenue for therapeutic intervention, with the aim of reducing tumor cell proliferation. The potential for this was demonstrated using a JAK1/2 inhibitor, ruxolitinib, which in multiple preclinical studies reduced JAK/STAT signalling and tumor growth in TCL cell lines.¹

The next step is to observe whether inhibition of JAK/STAT signalling translates into a clinical benefit. The results of a phase II trial (NCT02974647) were recently published by Moskowitz et al.¹ in Blood, which investigated the efficacy and safety of ruxolitinib in patients with relapsed/refractory peripheral T-cell lymphoma (PTCL) or mycosis fungoides (MF). We summarize key results below.

Study design

In total, 53 patients were enrolled into one of three cohorts based on baseline JAK/STAT activity:

• Cohort 1: Patients whose tumors harbor known activating mutations in JAK1, JAK2, JAK3, STAT3, or STAT5B (allele frequency ≥0.1)
• Cohort 2: Patients whose tumors lack mutations in these factors but have functional evidence of JAK/STAT activation defined as ≥ 30% phospho-STAT3 (pSTAT3) staining in tumor cells by immunohistochemistry (IHC)
• Cohort 3: Patients who did not meet criteria for Cohort 1 or 2, or who lacked adequate tissue for sequencing or assessment of pSTAT3 by IHC

All treatment cohorts were given twice daily ruxolitinib at 20 mg orally on 28-day cycles until disease progression, unacceptable toxicity, or at the discretion of the treating physician.

Eligibility criteria

• Aged ≥ 18 years
• Diagnosis of PTCL, peripheral NK cell lymphoma or cutaneous T-cell lymphoma (Stage 1B and greater)
• At least one prior line of therapy unless diagnosed with T-cell prolymphocytic leukemia (T-PLL)
• Adequate organ function

Results

Patient characteristics for the overall cohort (N = 53) are summarized in Table 1.

Table 1. Patient characteristics*

In Cohort 1:

• Mutations included STAT5B (n = 8), STAT3 (n = 7), JAK1 (n = 2), JAK2 (n = 1), and JAK3 (n = 6)
• Two patients (one with a STAT3 mutation and one with STAT3 and JAK1 mutations) were positive by IHC for pSTAT3 in tumor cells; 12 were negative for pSTAT3 expression, and seven did not have adequate tissue available for testing

In Cohort 2:

• The median percent of pSTAT3 positive tumors cells (by IHC) was 65%
• Nine patients had adequate tissue for next-generation sequencing (NGS), and none showed JAK/STAT mutations

In Cohort 3:

• Six patients had incomplete JAK/STAT activation data

Efficacy

The overall response rate (ORR) in 52 evaluable patients was 25% and the clinical benefit rate (CBR) was 35%, with a complete response (CR) rate of 6% and partial response (PR) rate of 19%; 10% had stable disease lasting at least 6 months. The CBR for Cohorts 1, 2, and 3 were 48%, 36%, and 18% respectively (p = 0.076), and the ORR was 33%, 29%, and 12% (p = 0.2). The median time to best response was 6.3 months (5.5–7.2 months). The median PFS was 2.8 months (95% CI, 1.8–4.5 months) and median OS was 26.2 months (95% CI, 11.5–not reached).

Given the heterogenous nature of TCL within each cohort, the authors investigated efficacy according to subtype, with survival outcomes summarized in Table 2.

Table 2. Survival outcomes stratified by TCL subtype*

JAK/STAT activation was most commonly observed in T-cell prolymphocytic leukemia

(T-PLL) and T-cell large granular lymphocyte leukemia (T-LGL). Ruxolitinib response was promising for T-PLL, with a CBR of 50% across patients harboring JAK/STAT mutations (n = 7) and one patient with only pSTAT3 expression. The same was observed in T-LGL (n = 5), with all four patients who responded remaining progression free for >1 year. However, clinical benefit was also observed irrespective of evidence of JAK/STAT involvement, with two patients from Cohort 3 responding to ruxolitinib.

Among patients with PTCL (n = 45), the CBR in Cohorts 1, 2, and 3 were 53%, 45%, and 13%, respectively, and the ORR was 37%, 36%, and 7%. Of the most common PTCL entities (PTCL-NOS, AITL, and ALCL; n = 24), the CBR was significantly higher in patients from Cohort 1 (n = 5) and Cohort 2 (n = 9) compared with Cohort 3 (n = 9) (p = 0.048), indicating increased sensitivity to JAK inhibition with baseline JAK/STAT involvement. Two patients, one with PTCL-NOS and another with AITL achieved a response >1 year.

Cutaneous MF was associated with a low ORR and CBR, and most of these patients carried JAK/STAT mutations (2/8) or pSTAT3 overexpression (3/8), indicating that the presence of JAK/STAT activation is not a certain predictor of response to inhibition. One patient did, however, remain progression free for >1 year.

Safety

Adverse events (AEs) observed with ruxolitinib were consistent with its known side-effect profile. The most common Grade ≥ 3 and serious treatment-related AEs are summarized in Table 3.

Table 3. Grade ≥3 AEs and serious treatment-related AEs*

IHC and multispectral immunofluorescent staining (MIF)

In 12 out of 14 evaluable cases with baseline JAK/STAT mutations, IHC revealed a lack of pSTAT3 staining while 9 out of 11 cases with pSTAT3 staining lacked JAK/STAT mutations. This indicates the presence of other factors that influence JAK/STAT signalling, perhaps tumor microenvironment signals.

MIF staining was performed in nine patients: three with CBR and six with no CBR. There was no association found between the fraction of TCL cells expressing pSTAT3/5 and response to ruxolitinib. However, in five out of six patients with no CBR, phosphorylated S6 ribosomal protein (pS6) expression in TCL cells was ≥25% at baseline compared with zero out of three patients with CBR (p = 0.05). These data demonstrate the potential of pS6 as a biomarker for predicting response to ruxolitinib. Additionally, pS6 markers are associated with PI3K/mTOR signalling, indicating that this pathway may potentially produce resistance to ruxolitinib.

Conclusion

Overall, this phase II trial provides evidence for the clinical relevance of JAK/STAT activation in TCL. Clinical activity for the single agent oral ruxolitinib was demonstrated across various subtypes of PTCL (n = 45) and with low toxicity. Clinical benefit was observed regardless of the presence of baseline JAK/STAT mutations. Given the possible association of PI3K/MTOR signalling for resistance to ruxolitinib in patients with no or less durable responses, the next study will investigate a combination of ruxolitinib with the PI3K inhibitor duvelisib.

Limitations of this study included the omission of external factors such as cytokine receptors and negative regulators to define JAK/STAT activation. Also, only pSTAT3 staining by IHC was included, and the cutoff point to define JAK/STAT activation (30%) was perhaps too high, as evident by a high responder to ruxolitinib showing 20% expression. Finally, a significant limitation highlighted by the authors was the limited tissue available for genetic, IHC, and MIF assessment.