Clinical Significance of Emergent Leukocytosis in Patients with Myelofibrosis Receiving JAK Inhibitor Therapy

Ruxolitinib is the first available JAK inhibitor (JAKi) therapy for amelioration of constitutional symptoms or splenomegaly in myelofibrosis (MF). Duration of response varies, as patients discontinue ruxolitinib due to disease progression or drug toxicity. Ruxolitinib failure is a poorly defined clinical entity, but may include suboptimal or lost spleen response, worsening cytopenias, accelerated or blast phase (AP/BP) progression, or non-hematologic side effects. It is not clear what other clinical parameters should be considered indicative of JAKi failure.

Leukocytosis is a known prognostic factor in MF and is included in various prognostic models. We have observed that some patients on stable doses of JAKi therapy develop progressive leukocytosis in the absence of other signs of MF progression. The significance of this event is not known, and it is not clear whether the onset of leukocytosis should prompt changes in clinical management.

To assess the clinical significance of emergent leukocytosis, we evaluated leukocyte counts in our database of MF patients receiving Ruxolitnib or Momelotinib as first-line JAKi therapy. We defined emergent leukocytosis as any of:

1. New onset of WBC ≥25 x 10⁹/L in patients with WBC ≤12.5 x 10⁹/L at JAKi start.
2. Doubling of WBC from the nadir value in patients with WBC >12.5 x 10⁹/L at JAKi start and nadir WBC >12.5 x 10⁹/L.
3. WBC ≥25 x 10⁹/L in patients with WBC > 12.5 x 10⁹/L at JAKi start after attaining a nadir WBC ≤12.5 x 10⁹/L.

Leukocytosis had to be sustained over consecutive blood counts at least one month apart and had to occur in the absence of infection, steroid therapy, AP/BP transformation, splenectomy or JAKi dose reduction. Exclusion criteria included concurrent hematologic diagnoses, and splenectomy or AP/BP MF preceding JAKi initiation.

Of 290 patients with MF receiving JAKi therapy, 217 met study criteria. Of these 217 patients, 27 developed emergent leukocytosis while receiving JAKi. The cumulative incidence of leukocytosis was 4%, 10% and 15% at 1, 3, and 5 years from the start of JAKi therapy, respectively. Transformation to AP/BP, splenectomy, bone marrow transplant, or death from any cause were considered as competing risks in the calculation of cumulative incidence. In a multivariate analysis, clinical parameters associated with emergent leukocytosis included presence of baseline anemia (HR 4.94 [95% CI, 1.13-21.53]; p = 0.03) or leukocytosis (HR 5.01 [95% CI, 1.44-17.41]; p = 0.01) at JAKi start and female gender (HR 0.21 for male [0.08-0.06]; p=0.002). Baseline leukocytosis as WBC ≥25 x 10⁹/L, and anemia was as a hemoglobin <100g/L.

In patients with available targeted sequencing performed prior to JAKi therapy (n=141, 65%), TET2 mutations were associated with increased risk of leukocytosis (HR 5.48 [95% CI, 1.73-17.35]; p=0.004), but JAK2, CALR, MPL or ASXL1 mutations were not.

Among 27 patients who experienced emergent leukocytosis, 21 were deceased while 6 are in ongoing follow-up. Median overall survival calculated from the date of leukocytosis was 14 months [95% CI, 8.6-27.9]. Overall, 7 patients transformed to AP/BP, 7 died from progressive MF, 4 underwent transplant, one underwent splenectomy, 6 died of miscellaneous causes and two remain on first-line JAKi without any complicating events.

Among 27 patients experiencing leukocytosis, paired samples for molecular analysis at JAKi initiation and at the time of leucocytosis were available in 21 (78%). Results from targeted sequencing of a panel of 49 myeloid genes were available from 4 patients with paired samples (Table 1). New mutations at leukocytosis were seen in NRAS and CEBPA in the first patient, KRAS and SH2B3 in the second, and CBL in the third. The fourth patient acquired no new mutations but demonstrated marked expansion of a dominant clone. The remaining 17 paired samples have been submitted for sequencing with results pending.

In summary, MF patients with baseline anemia, leukocytosis, or female gender are at higher risk for emergent leucocytosis while on JAKi therapy. The median overall survival of 14 months in patients who experience emergent leukocytosis is comparable to that observed in patients following ruxolitinib failure. Our data also suggest that onset of leukocytosis may be indicative of underlying clonal evolution, which will need to be confirmed in a larger number of patients. These findings suggest consideration of emergent leukocytosis as a criterion for JAKi failure in MF.