Inhibition Of The JAK/STAT Signaling Pathway As a Novel Therapy In CD8+ T-Large Granular Lymphocyte Leukemia

Mainstream immunosuppressive therapies in patients with CD8+ T cell large granular lymphocyte leukemia (T-LGLL) include cyclosporine (CsA), methotrexate (MTX), and cyclophosphamide (CTX), used to alleviate often severe cytopenias, in particular neutropenia and reticulocytopenic anemia. Retrospective analysis of the primary response rate of 169 patients treated at our facility was CsA 56%, MTX 37%, and CTX 45% with an overall combined response rate of CsA 59%, MTX 43%, and CTX 45%, illustrating that there are a significant proportion of patients who remain refractory. A number of salvage therapies have been applied, including alemtuzumab and anti-thymocyte globulin. Due to the frequent association of T-LGLL with rheumatoid arthritis (RA), additional salvage therapies include rheumatologic drugs such as abatacept or infliximab. During our practice we have encountered 5 cases with T-LGLL associated severe cytopenia who remained refractory to standard and salvage therapies. One of these patients was positive for a STAT3 mutation. To treat coexisting RA these patients ultimately received tofacitinib citrate, a Janus-associated kinase (JAK) inhibitor. Activation of STAT3 due to IL-6 signaling or STAT3 mutation is becoming recognized as a hallmark of T-LGLL, and we wondered what effect upstream inhibition of the JAK/STAT pathway would have on associated cytopenia. 4/5 (80%) patients with cytopenia demonstrated hematologic improvement. Of these, one patient with both anemia and neutropenia achieved a complete response, with normalization of hemoglobin levels and restoration of normal neutrophil counts. Three patients showed partial responses and all patients reported improvement of systemic symptoms including fatigue and arthralgias.

These clinical observations, coupled with the recent discovery of STAT3 mutations in T-LGLL, prompted us to conduct in vitro studies examining the effect of JAK/STAT inhibition on a cohort of 10 CD8+ T-LGLL patients, 4 of which had the STAT3 mutation detected by ARMS-PCR, and none had received tofacitinib citrate. For inhibitors, we chose STA-21, a novel STAT3-selective low molecular weight compound which acts by preventing STAT3 dimerization and DNA binding; ruxolitinib, a JAK1/2 selective inhibitor; and tofacitinib citrate, a JAK3 selective inhibitor. Cells were cultured in presence of inhibitor or vehicle control at 1uM and 10uM concentration and apoptosis of CD8+ cells was measured by flow cytometry using Annexin V and propidium iodide. The percentage of cell death was calculated by subtracting vehicle control background from inhibitor results. While all three inhibitors demonstrated in vitro activity, at 10uM STA-21 was far more efficient at inducing apoptosis than either tofacitinib citrate (44% vs. 5%, p<.0001, paired t-test) or ruxolitinib (44% vs. 8%, p=.0008). Only STA-21 exhibited significant dose dependent responses at concentrations tested. Subgroup analysis of patients with and without the STAT3 mutation clearly demonstrated increased efficacy of JAK/STAT inhibition in patients with the STAT3 mutation (Table 1). To see if inhibitor results were T-LGLL specific, we tested 6 healthy controls, but there was no statistical difference in CD8+ apoptosis between patients and controls in response to inhibitor treatment.

Our results suggest that direct inhibition of STAT3 offers the most potential benefit for rational therapy of T-LGLL patients, especially those with STAT3 mutations. JAK2 inhibitors may constitute alternative treatment effective in both STAT3 mutant and wild-type cases. Furthermore, clinical responses indicate that in T-LGLL patients with RA who have failed to respond to first line therapies, tofacitinib citrate may offer a potential alternative. Prospective clinical trials are necessary to confirm this preliminary data.