Ruxolitinib or Dasatinib in Combination with Chemotherapy for Patients with Relapsed/Refractory Philadelphia (Ph)-like Acute Lymphoblastic Leukemia: A Phase I-II Trial

Background: Ph-like acute lymphoblastic leukemia (Ph-like ALL) is a high-risk subtype of ALL in both children and adults (Roberts, NEJM 2014, JCO 2017; Jain, Blood 2017). Ph-like ALL is characterized by a multitude of genetic aberrations, leading to activation of the JAK-STAT signaling pathway or the ABL kinase pathway, which may be amenable to targeted therapies.

Methods: We designed an investigator-initiated phase I/II trial using ruxolitinib (JAK inhibitor) or dasatinib (ABL1 kinase inhibitor), in combination with chemotherapy for patients (pts) with relapsed/refractory (R/R) Ph-like ALL aged ≥10 years (yrs) old (NCT02420717). Pts with BCR-ABL1 fusion were excluded. Pts with R/R B-ALL were screened for genomic aberrations seen in Ph-like ALL. Testing methods included flow cytometry for CRLF2 ; NGS panel, covering all exons of JAK2 ; FISH panel (CRLF2, ABL1, ABL2, CSF1R, JAK2, EPOR, PDGFRB); Ph-like ALL fusion panel (Child Lab, Columbus, OH). Once the genomic aberration was identified, pts were either enrolled on ruxolitinib arm (CRLF2 overexpression, JAK mutation/fusion, EPOR fusion, SH2B3 deletion, IL7R mutation), or dasatinib arm (ABL1, ABL2, PDGFRB fusions) (Figure 1). The primary objective (Phase I) is to determine the safety and MTD of ruxolitinib in combination with chemotherapy. The secondary objectives include efficacy. Three dose levels of ruxolitinib are planned (level 1: 15mg BID, level 2: 20mg BID, level 3: 25mg BID). The dose of dasatinib is 100mg once daily. Ruxolitinib (or dasatinib) was initiated on cycle 1 day 1 and continued daily. Hyper-CVAD chemotherapy was added after cycle 1, or earlier if evidence of disease progression. Phase II will commence after the MTD is established. We report here interim results of the phase I part of the trial.

Results: 9pts have been treated. Median age 24 yrs (range 18-62). Median no. of prior therapies 3 (2-10). 3 pts had a prior allo-SCT, and 1 pt had failed CD19 CART. 8 pts were treated on the ruxolitinib arm (15mg BID, n=5; 20mg BID, n=3), and 1 pt on the dasatinib arm. Genomic aberrations in the ruxolitinib arm included 7 pts with CRLF2 overexpression by flow-cytometry (confirmed by CRLF2 FISH), and 1 pt with a novel JAK2 fusion (HMBOX1-JAK2). 3 of the 7 pts with CRLF2 rearrangement had a JAK2 mutation (R683S, T875N, I682F). Dose levels 1 and 2 have been completed without any DLT; the trial is now enrolling at dose level 3. Two pt in dose level 1 did not receive chemotherapy, and the cohort was expanded to 5 pts to allow for DLT assessment. One pt with CRLF2+/JAK2 R683S treated at the dose level 2 achieved a CRp. One pt with CRLF2+/JAK2 I682F (dose level 1) had a transient reduction in peripheral blood (PB) blasts during the ruxolitinib monotherapy. The sole pt on the dasatinib arm had NUP214-ABL1 fusion. This pt had reduction in PB blasts with dasatinib, but did not respond in the bone marrow (BM).

CyTOF analysis was performed on pretreatment and post-TKI (Day 21) BM or PB samples in 5 pts (4 pts ruxolitinib arm; 1 pt dasatinib arm). TKI therapy reduced pTyr by >70% in CD10+/CD19+ ALL cells in 3 of the 4 pts, and to a lesser degree the downstream p4EBP1 in 4 pts (33-91% inhibition). One pt with wtJAK2/CRLF2+ had no modulation of pTyr, pStat3 or pp38; another pt with same genotype had no change in pStat3 or pp38 despite decrease in pTyr, and both pts progressed during ruxolitinib therapy. In two pts with most significant reduction in pTyr there was concomitant reduction in pStat3 in CRLF2+ ALL and of pp38 and p4EBP1, and corresponding decrease in BM blasts (NUP214-ABL1 rearranged ALL, 58%→8%; and JAK2 T875N CRLF2+ ALL, 90%→58%, Figure 2). Notably, expression of anti-apoptotic protein BCL-2 remained unchanged or was induced in residual ALL cells in all except one pt.

Conclusions: The combination of ruxolitinib and chemotherapy is safe, with no DLT seen with combinations of ruxolitinib at two initial dose levels or dasatinib with hyper-CVAD. Preliminary clinical activity in R/R heavily pre-treated ALL pts is limited at the dose levels studied. CyTOF analyses showed lack of modulation of signaling in pts with wtJAK2/CRLF2 accompanied by ALL progression; partial down regulation of JAK-STAT signaling in mutJAK2/CRLF2+ by ruxolitinib and in ABL1-rearranged ALL by dasatinib, associated with decrease in % BM blasts. Notably, we observed compensatory induction of anti-apoptotic BCL-2 protein, potentially conferring resistance to TKIs. The study continues to enroll at a higher dose of ruxolitinib.

View largeDownload slide

View largeDownload slide

Close modal