P991: IMPAIRMENT OF TCR SIGNALLING IN PATIENTS WITH MYELOFIBROSIS TREATED WITH RUXOLITINIB

Abstract

Background: Ruxolitinib (rux) is a JAK1/JAK2 inhibitor with varied immune effects including on T cell, NK cell and dendritic cell function. Therapeutic effects are largely attributed to marked reduction in pro-inflammatory cytokines. Aims: We hypothesised that impairment of T cell receptor (TCR) signalling through off-target binding of Src kinases, including Lck, would be an additional mechanism of immune dysfunction. Methods: We performed phosphoflow cytometry in Tregs, T effectors and NK cells to assess the effect on signalling downstream from the TCR and activating NK cell receptors. 11-colour flow cytometry was performed after cells were activated with H2O2 for 15 minutes, due to its activity as a potent phosphatase inhibitor and cells were analysed for phosphorylation of ZAP70 as a surrogate of TCR signalling and STAT5. A gating strategy of CD4+/CD25+/FOXP3+/CD127lo cells was used for identification of Tregs. Results: Phosphoflow analysis was performed in 7 patients with a diagnosis of MF managed with rux and compared with 7 healthy controls (HC). 4 (57%) of patients were male and mean age was 61 (range 44-76). Patients were taking a median rux dosage of 30mg daily (10-50). Three patients had a DIPPS+ score of Int-2, 2 Int-1 and 2 low-risk. Results of phosphoflow analysis are expressed as relative fluorescence intensity (RFI) defined as MFI in H2O2 stimulated sample/MFI in unstimulated sample. As expected, patients on rux had significantly reduced phosphorylation of pSTAT5 when compared with HC in all cell subsets evaluated. The mean RFI of pSTAT5 in CD4+ and CD8+ cells in HC was 21.9 and 31.1 respectively, compared with 6 and 8.5 in rux patients (p=0.001/p=0.002). In CD56+ cells this was 46.2 vs 7.7 (p=0.005) and in Tregs this was 21.9 vs 6.1 (p=0.019). Of note, a difference in phosphorylation was also observed for pZAP70 indicative of inhibition of TCR proximal signalling. For pZAP70 mean RFI in CD4+ cells was 8.8 in controls compared with 2.2 in rux patients (p=0.05). In CD8+ cells mean RFI was 11.4 in controls and 2.4 in rux patients (p=0.04). In CD56+ cells mean RFI was 17.8 vs 2.4 (p=0.02), whilst in Tregs, significance was not reached with mean RFI of 6.3 vs 2.2 (p=0.13). Univariate analysis of rux patients showed no effect of age, comorbidities, DIPPS+ score or rux dosage on TCR or STAT5 signalling. Findings were compared with those observed in 4 patients with CML and deep molecular response on nilotinib, in view of the highly specific ABL1 binding properties of this TKI. As expected, patients on nilotinib had significantly higher RFI for pSTAT5 compared with rux patients, at 23, 32.8, 21.2 and 46.5 in CD4+, CD8+, Treg and NK cells respectively (p=0.002/p=0.002/p=0.006/p=0.003). Interestingly, patients on nilotinib also had significantly higher RFI for pZAP70 in CD4+, CD8+ and NK cells at 6.7, 7.4 and 8.1 compared with rux patients (p=0.035/p=0.025/p=0.037). Image:Summary/Conclusion: Studies have shown that rux inhibits T cell proliferation and reduces CD4+ cell cytokine secretion. Previous in vitro analysis of rux on CD4+ cells failed to identify an effect on TCR signalling after CD3/CD28 stimulation. In our ex-vivo analysis, using H2O2 stimulation, we observe TCR signalling inhibition in CD4+, CD8+ and NK cells, although reduced compared with effects on STAT5. It has been shown that rux inhibits Lck with an IC50 of 3.6uM. Whilst this is significantly less than the inhibitory effect on JAK kinases, there is ~34% sequency homology between Src and JAK family kinases in the kinase domain, suggesting off target kinase inhibition is implicated in the impairment of TCR signalling.