Ruxolitinib Shows Noteworthy Responses in Patients With Rare Leukemias

With the absence of a standard of care for chronic neutrophilic leukemia and atypical chronic myeloid leukemia, 2 rare BCR-ABL1-negative myeloid neoplasms, ruxolitinib may be a viable treatment option after a phase II study showed potential for response in this patient population.

With the absence of a standard of care for chronic neutrophilic leukemia (CNL) and atypical chronic myeloid leukemia (aCML), 2 rareBCR-ABL1-negative myeloid neoplasms, ruxolitinib (Jakafi) may be a viable treatment option after a phase II study (NCT02092324) showed potential for response in this patient population.1

The efficacy analyses from a study of ruxolitinib in 44 patients with CNL (n = 21) and aCML (n = 23) showed that 33% of the first 15 patients in stage 1 of the trial responded to treatment, which allowed investigators to move to stage 2 of the Simon’s design. Thirty-two percent of the first 25 patients responded, with all responses observed being partial responses. The objective response rate (ORR) seen by diagnosis was 58% among patients in the CNL group and 8% of the patients in the aCML group (P= .011). For patients with CSF3Rmutations, the ORR was 54%, whereas patients withCSF3R—wild type disease had an ORR of 8% (P= .030).

Study participants had a median overall survival (OS) of 18.8 months (95% CI, 15.3-25.2). The median OS was 15.6 months for nonresponders and 23.1 months for the patients who responded to treatment. Patients with CML versus those with aCML had similar survival probability until around the 9-month mark. The Kaplan-Meier curves for survival split at that point and those with CNL were seen to have a greater likelihood for survival (P= .2). Patients with an International Prognostic Scoring System (IPSS) score ≤2 had significant improvements in OS versus those with an IPSS score >2 (P= .01).”

Investigators tracked changes in disease from baseline to the end of treatment cycle 6 out of a possible 24 cycles, and changes were broken down by disease type and the presence ofCSF3Rmutations. In the overall population of study participants, white blood counts (WBC), absolute neutrophil counts (ANC), spleen volume, and hemoglobin values were reduced after treatment with ruxolitinib. There was also a decrease in the absolute median change, and the median change of symptoms related to myeloproliferative neoplasms (MPNS), according to the MPN Symptom Assessment Form Total Symptom Score (MPN-SAF TSS). Platelet count, spleen reduction, International Prostate System Score IPSS increased.

The 19 patients with CNL had higher percentages of decreased disease characteristics in the study than patients with aCML, with the exception of hemoglobin levels. The WBC absolute median change with CNL was —33.5 × 109/L (—195.9 to 2.6) compared with –21.2 × 109/L (—63.7 to 47.4) in patients with aCML (P= .053). The change in ANC was —29.8 × 109/L (—189.1 to 3.6) versus –10.7 × 109/L (—51.6 to 27.5), respectively (P= .043). The respective changes for MPN-SAF TSS were —4.5 (–37.0 to 7.0) compared with –3.0(—25.0 to 8.0;P= .514). Spleen volume decreased in the CNL group by —249.7 cm3(—1,5553.8 to 259.9) versus -219.7 cm3(—1078.9-1061.4) in patients with aCML (P= .941). Hemoglobin levels were —0.4 g/dL (2.2) and –0.6 g/dL (1.8), respectively (P= .848).

The incidence of some disease characteristic was significantly higher in patients with CNL in contrast to aCML, and in some instances, changes showed a decrease in patients with aCML while increasing in the CNL. Platelet counts according to the absolute median change measurement were 64.0 × 109/L (—214.0 to 203.0) versus –8.0 × 109/L (—84.0 to 260.0), respectively (P= .096). The number of patients without spleen reduction ≥35% was 4 in the CNL group (28.6%) and 4 of those with aCML (57.1%). The number of patients who had spleen reduction ≥35% was 10 in the CNL group (71.4%) and 3 in the aCML group (42.9%). The changes in IPSS were not significant in either group.

In the efficacy analysis of patients withCSF3R—wild type andCSF3R-mutant myeloid neoplasms, all changes in disease characteristics decreased after 6 cycles of ruxolitinib treatment except platelet counts and spleen reduction. Platelets increased by 11.0 × 109/L (—84.0 to 135.0) in patients withCSF3R—wild type compared with 56.0 × 109/L (—214.0 to 260.0) in patients withCSF3R-mutant disease (P= .216). The number of study participants withCSF3R—wild type disease who did not experience spleen reduction ≥35% was 5 (71.4%) versus 3 (21.4%) in patients withCSF3Rmutations. However, 2 patients withCSF3R—wild type disease did have a spleen reduction ≥35% (28.6%), as well as 11 individuals withCSF3R-mutant disease (78.6%;P= .056).

The decreases observed in WBCs were —8.5 × 109/L (—63.0 to 47.4) versus –38.3 × 109/L (—195.9 to 14.5), in theCSF3R—wild type versus theCSF3R-mutant group (P= .028). For ANC, the changes were —6.5 × 109/L (—51.6 to 27.5) compared with ­–33.0 × 109/L (—189.1 to 13.8), respectively (P= 0.22). The hemoglobin levels observed were —1.0 g/dL (1.8) versus –0.1 g/dL (2.1), respectively (P= .292). The final increase in change in disease characteristics after cycle 6 was with the MPN-SAF TSS, which was —2.5 (–25.0 to 7.0) in patients with CSF3R—wild type disease versus –4.0 (–37.0 to 8.0) in patients withCSF3R-mutant disease (P= .599). There was also a slight difference in values of MPN-SAF TSS ≥20% in these 2 groups. In subjects who were CSF3Rwild type, the value observed was —10.0 (–25.0 to 0.00) compared with –11.0 (–37.0 to 3.0;P= .462).

Ruxolitinib was considered to be tolerable in the overall population of patients in the study. Hematologic adverse events (AEs) were seen more commonly than other types of AEs. The most common AEs overall were anemia (n = 15), platelet count decrease (n = 6), leukocytosis (n =5), fatigue (n = 4), lung infection ( n = 4), and pneumonitis ( n = 4).

A total of 15 patients discontinued treatment due to disease progression with or without blastic transformation. Another 11 patients discontinued due to provider preference which was found to be primarily due to lack of clinical benefit. Thirty-one patients died since the study began, with the most common cause being disease progression.

The open-label, single-arm, multicenter, investigator-initiated phase II study of ruxolitinib’s efficacy for patients with CNL or aCML withCSF3Rmutations followed a Simon’s 2-stage minimax design. The primary end point of the study was the rate of hematologic response. Secondarily, the study also assessed the incidence of AEs and other safety and efficacy determinants.

According to the study’s statistical analysis protocol, which required the ORR for ruxolitinib to be ≥30%, ruxolitinib is a promising agent for treatment of patients with CNL or CML withCSF3Rmutations.


Dao KT, Gotlib J, Deininger MM, et al. Efficacy of ruxolitinib in patients with chronic neutrophilic leukemia and atypical chronic myeloid leukemia. [Published online December 27, 2019].J Clin Oncol. doi: 10.1200/JCO.19.00895.