How to switch therapies in patients with myelofibrosis

Myelofibrosis (MF) is a heterogeneous disease with limited treatment options, which may cause treatment challenges when patients no longer respond to one therapy. Donal P. McLornan and Claire N. Harrison recently published a paper providing guidance on changing therapy choice in myelofibrosis in Blood. This article provides a summary of the current guidance for treating patients with MF who require therapy switching from a JAK inhibitor to another JAK inhibitor or to a different treatment, or who are considered for allogeneic hematopoietic stem cell transplantation (allo-HSCT).¹

The characterization of the JAK2 driver mutation and the description of JAK signaling cascade in myeloproliferative neoplasms (MPN) led to therapeutic advances, such as the identification of the JAK1/JAK2 inhibitor, ruxolitinib, which was approved by both the U.S. Food & Drug Administration (FDA) and European Medicines Agency for the treatment of MF. Recently, another JAK inhibitor, fedratinib, was FDA approved for MF, both in the frontline and second-line setting for patients with intermediate-2 and high-risk MF. As newer therapies are being explored, questions arise of where they fit into the treatment landscape, with the focus on agents second-in-line to ruxolitinib treatment to overcome challenges such as ruxolitinib resistance or intolerance.

Challenges with JAK inhibitor treatment

There are concerns of systemic deterioration and proinflammation that result from a substantial reduction in dosage or a complete stop of ruxolitinib. However, it is unknown whether these withdrawal reactions occur with fedratinib treatment as well. As both are viable treatment options, questions are likely to arise, such as when the best time to switch between the two treatments is. Also, regimen schedules should translate from clinical trials to clinical practice effectively, for example patients in clinical practice are more likely to switch between the two drugs without a washing-out period from the first drug. Overall, patients with MF have a poor prognosis when discontinuing ruxolitinib, and it is unclear if this also happens when frontline fedratinib fails. Moreover, the optimal time for patients to receive allo-SCT with JAK inhibitors is debatable, as some practices switch patients from frontline JAK inhibitor treatment directly to allo-SCT, whereas others gradually reduce treatments before allo-SCT. Another unaddressed question is whether a change of treatment to the other JAK inhibitor agent during the lead-in time to allo-SCT could improve patient outcome.  

Considerations when switching therapies

1. Frontline therapy

When administering frontline treatment, physicians should provide sufficient time for patients to respond to treatment and monitor if dose optimization has occurred, whilst balancing dosage against side effects.   

2. Frontline failure

Recognition of ruxolitinib or fedratinib failure or intolerance and assessment of disease dynamic to monitor for clonal evolution is important. Physicians should also consider if the patient is suitable for a switch of treatment, e.g. from ruxolitinib to fedratinib or vice versa, or allo-HSCT, or enrolment into a clinical trial.

3. Confounders

Monitoring that encapsulates patient symptoms, such as the MPN Symptom Assessment Form or MPN10, and accurate spleen size measurements should be mandated, and possible confounding factors that affect JAK inhibitors, such as such as depression, drug–drug interactions, and guidelines for managing anemia or thrombocytopenia should be regularly reviewed.

4.  Switch agents

Recommendations encourage that patients should be weaned off JAK inhibitors before starting a second-line agent or overlap two JAK inhibitor treatment, and possibly using, steroid cover. Sufficient time for adequate exposure and optimal dosing needs to be ensured before assessing response to second-line agent.

Discontinuation of JAK inhibitors

A study by Palandri et al., examining the outcome of patients with MF after ruxolitinib treatment on 442 patients, at a median follow-up of 30.5 months, reported 20% of patients died while receiving ruxolitinib and 48% had discontinued. The median survival of patients who had discontinued ruxolitinib (n = 171) was 22.6 months. Patients who received another JAK inhibitor or investigational agent after discontinuation because of intolerance or resistance had a better survival rates compared with patients who received subsequent therapies with drugs such as danazol or hydroxycarbamide, indicating the importance of appropriate treatment sequence. Similar observations were made in the 5-year analyses of COMFORT-1 and COMFORT-2 studies, which showed 72% and 73% of patients randomized to ruxolitinib had discontinued, respectively.

A phase I/II study examining 107 patients with MF found 86 patients had discontinued ruxolitinib after a median follow-up of 79 months and median survival after discontinuation was 14 months. This study found patients with clonal evolution or thrombocytopenia (platelets < 100 × 10⁹/L) at discontinuation had a poor prognosis.

Defining refractoriness or intolerance to JAK inhibitors

Definitions of JAK inhibitor failure fluctuates across clinical trials, therefore establishing practical criteria, that factors in dose optimization, is crucial to enhance patient-individualized response.

Definitions are based on JAKARTA-2 reanalyses

• Loss of response or refractoriness to ruxolitinib
  • ≥ 3 months administration of ruxolitinib with an initial response followed by either regrowth of the spleen or an insignificant response (defined as < 10% reduction of spleen volume or < 30% reduction in spleen size from baseline)
• Intolerance was defined as ≥ 28 days ruxolitinib therapy that led to the need of transfusion (≥ 2 units per month for two months) or the development of Grade ≥ 3 thrombocytopenia, anemia, hematoma, and/or hemorrhage during ruxolitinib treatment

Potential toxicities of JAK inhibitors  

Comparisons across different clinical trials regarding the management of treatment-emergent adverse events remain difficult as treatment strategies differ. The balance between acceptable low platelet count and sufficient clinical response is also yet to be defined as both ruxolitinib and fedratinib are associated with hematological toxicities, particularly anemia and thrombocytopenia. The two landmark JAKARTA studies analyses indicate that fedratinib may confer greater suitability to patients with baseline platelet counts of < 100 × 10⁹/L and splenic volume reduction were similar in patients with platelets < or > 100 × 10⁹/L. Also, discontinuation or dose alteration because of Grade 3–4 thrombocytopenia were uncommon.

It is yet to be elucidated if the enhanced JAK2 specificity of fedratinib will impact rates of non-melanoma skin cancers and infectious complications seen with ruxolitinib treatment. However, it is currently known that fedratinib is associated with low-grade gastrointestinal toxicity and has a black-box warning for thiamine deficiency and encephalopathy; these factors may influence physician choice of frontline treatment and the timing of therapeutic switch.  

Potential considerations from exemplary clinical cases

1. For a patient who lost response but develops inflammatory response when weaned off ruxolitinib, potential treatment considerations include overlapping ruxolitinib and fedratinib while ruxolitinib is being weaned off. Alternatively, the use of a steroid during ruxolitinib wean before fedratinib treatment is established
2. For a patient who lost response under stable low dosing of ruxolitinib, physicians should consider adjusting dose increment of ruxolitinib before switching to fedratinib
3. In a patient who lost response and has been gradually discontinued of ruxolitinib, potential options are to start fedratinib treatment, to enrol into a clinical trial, or to perform allo-HSCT
4. A patient who develops blastic phase disease while under fedratinib therapy has a dismal prognosis. Here, options include best supportive care, low-intensity chemotherapy, or hypomethylating agents
5. For a patient with optimal spleen response under fedratinib therapy who is considered for allo-HSCT, no formal recommendation on the management of fedratinib prior to transplantation can be given as the effects on engraftment and posttransplantation immune reconstitution are currently unknown

Conclusion

Although the approval of the two JAK inhibitors in MF is fortunate, there is a need to establish practice guidelines and summaries of product characteristics. Recommendations should be able to translate into clinical practice and be personalized to the patient’s clinical status, line of therapy, and comorbidities.