In our study, ruxolitinib showed acceptable efficacy for relapsed/refractory HL patients. However, only one patient achieved CR, and the duration of response was relatively short. This outcome was similar to that of a phase II study of ruxolitinib treatment for relapsed/refractory HL patients. In the study, best overall response rate was 18.8% (6/32 patients), and the median duration of response was 7.7 months. The survival analysis reported the median PFS of 3.5 months (95% CI: 1.9 - 4.6), and the median OS of 27.1 months (95% CI: 14.4-27.1) similar to that of our study [13]. Notably, our study showed ruxolitinib was not effective in PMBCL patients although the number was relatively small. These outcomes were not consistent with previous preclinical studies.
Actually, the JAK2 selective inhibitor, fedratinib decreased the cellular proliferation of HL and PMBCL cell lines, and reduced tumor growth in murine xenograft models of cHL and MLBCL with 9p24.1/JAK2 amplification [14]. The anti-tumor effect of ruxolitinib was also demonstrated in both HL and PMBL cells and xenograft models [15]. Although this discrepancy of previous pre-clinical studies with the results of our study could not be clearly explained, the influence of JAK2 signaling on the growth of tumor cells in human might be different from in vitro and in vivo models, and this might be related with different outcome of ruxolitinib in HL and PMBCL of this pilot study.
To find a biomarker for predicting the response to ruxolitinib, we evaluated the presence of JAK2 amplification in patients enrolled onto our study. Although the number of patients in our study was small, our FISH analysis showed a possible association of JAK2 amplification with response to ruxolitinib in HL, because three of four HL patients shown to have a JAK2 mutation responded to treatment. In a previous phase II study of ruxolitinib analyzing 12 patients for JAK2 amplification, specific JAK2 amplification was found in only one patient who achieved PR [13]. However, our PMBCL cases with JAK2 amplification did not respond, suggesting that PMBCL might be less dependent than HL on JAK2 activation. Furthermore, it is not certain whether our dosage was appropriate for inhibiting JAK2 in HL tumor cells because we used the dosage recommended for myeloproliferative neoplasms. A phase I/II study of ruxolitinib in acute myeloid leukemia reported the tolerability of ruxolitinib at doses up to 200 mg twice daily [16]. Given the tolerable toxicity profiles in our study, dosage escalation might improve the efficacy profile of ruxolitinib in HL and PMBCL patients.
Conclusions
In conclusion, this pilot study suggested that ruxolitinib might have single-agent activity against HL at the current dosage, especially in case of patients with JAK2 amplification. However, ruxolitinib might not be effective against PMBCL regardless of JAK2 amplification. Considering the biological rationale for the use of JAK2 inhibitor as a treatment of HL, further study should be warranted to explore the optimal usage of JAK2 inhibitor such as combined approach of JAK2 inhibitor with brentuximab vedotin, nivolumab or pembrolizumab.