Abstract

Key Summary Points

Introduction

Anemia represents a substantial medical need in myelofibrosis; approximately one-third of patients have hemoglobin (Hb) of < 100 g/L at diagnosis, and nearly all become anemic over time [1, 2]. Many patients receive red blood cell (RBC) transfusions to manage anemia, but transfusion dependency is associated with diminished quality of life and is a negative prognostic factor for survival [1]. While Janus kinase (JAK) inhibitors are the backbone of myelofibrosis treatment due to their efficacy in managing symptoms and splenomegaly, some, such as ruxolitinib and fedratinib, do not alleviate, and may exacerbate, anemia [3–7]. In the absence of therapies that directly address the mechanisms of anemia in myelofibrosis—including increased pro-inflammatory signaling, RBC sequestration, and dysregulated iron metabolism [1]—anemia supportive therapies such as erythropoiesis-stimulating agents (ESAs; e.g., epoetin alfa), androgens (e.g., danazol), corticosteroids, and immunomodulatory drugs have been traditionally added in combination with continued ruxolitinib or fedratinib to manage new or worsening anemia [1, 3]. However, these adjunct treatments offer limited efficacy, durability, and tolerability, while dose reduction of the concomitant JAK inhibitor to mitigate myelosuppressive effects may compromise clinical benefit [3].

Momelotinib, a JAK1/JAK2/activin A receptor type 1 (ACVR1) inhibitor approved specifically for the treatment of patients with myelofibrosis who have anemia [7], has demonstrated clinically meaningful and durable improvements in anemia, splenomegaly, and symptoms in both JAK inhibitor-naive and -experienced patients with myelofibrosis across 3 phase 3 trials [8–10]. Thus, both JAK inhibitor-naive patients who already present with disease-related anemia and JAK inhibitor-experienced patients who develop new or worsening anemia on ruxolitinib or fedratinib may benefit from momelotinib. While clinical guidelines include momelotinib for patients with myelofibrosis and anemia, traditional anemia supportive therapies continue to be considered as well, particularly for patients who also have splenomegaly and constitutional symptoms that are well controlled on an alternative first-line JAK inhibitor [11].

In the phase 3 SIMPLIFY-2 trial, momelotinib was evaluated versus best available therapy (BAT) in JAK inhibitor-experienced patients who had experienced hematologic toxicity while receiving ruxolitinib [9]. While the primary endpoint of superiority in the rate of spleen volume reduction ≥ 35% from baseline at week 24 with momelotinib versus BAT was not met (6.7% vs. 5.8%; P = 0.90), momelotinib was associated with nominally significant improvements in the rates of total symptom score (TSS) reduction ≥ 50% (26.2% vs. 5.9%; nominal P = 0.0006) and RBC transfusion independence at week 24 (43.3% vs. 21.2%; nominal P = 0.0012) [9]. Notably, treatment washout from prior ruxolitinib was not permitted, and 88.5% of patients in the BAT arm remained on ruxolitinib, including some who continued to receive therapeutic doses [9]. Thus, this trial provides an opportunity to evaluate the anemia-related benefits of switching to momelotinib versus continuing ruxolitinib and managing anemia with supportive therapies and/or RBC transfusions. To that end, we performed a post hoc descriptive analysis of SIMPLIFY-2 focused on patient subgroups for whom anemia management is a key consideration: (1) patients with moderate-to-severe anemia, defined by baseline Hb of < 100 g/L, and (2) patients who were not transfusion independent (TI) at baseline.

Methods

Results

Anemia-Related Efficacy

In both subgroups of interest, week 24 transfusion independence rates were higher with momelotinib versus BAT/ruxolitinib. Among patients with baseline Hb of < 100 g/L, 22 (33.3%) were TI at week 24 with momelotinib per the prespecified terminal 12-week definition; 25 (37.9%) were TI by week 24 per the rolling 12-week definition (Fig. 1). In contrast, only 5 (12.8%) and 7 (17.9%) in this subgroup were TI per the week 24 terminal and rolling definitions, respectively, with BAT/ruxolitinib.

Open in a separate window

Fig. 1

Week 24 transfusion independence in the baseline Hb < 100 g/L and non-TI subgroups. Transfusion independence at week 24 (terminal 12-week definition; defined as no RBC transfusions and no Hb of < 80 g/L in the last 12 weeks before week 24) or by week 24 (rolling 12-week definition; defined as no RBC transfusions and no Hb of < 80 g/L during any 12-week period through week 24). BAT best available therapy, Hb hemoglobin, RBC red blood cell, TI transfusion independent

Similarly, 25 patients (34.7%) in the non-TI subgroup were TI with momelotinib per the terminal 12-week definition and 29 (40.3%) were TI per the rolling 12-week definition versus only 1 (3.0%) and 3 (9.1%) with BAT/ruxolitinib, respectively (Fig. 1). Transfusion independence rates at week 24 based on baseline erythropoietin (EPO) level were also assessed within these subgroups of interest; there was no influence of baseline EPO levels on transfusion independence rates with momelotinib, and response rates were consistently higher with momelotinib versus BAT/ruxolitinib (Supplementary Material).

Similar to the overall ITT population [9], mean Hb levels in these subgroups improved rapidly with momelotinib and remained higher over time than in those treated with BAT/ruxolitinib, including additional improvement in the BAT/ruxolitinib groups after crossover to momelotinib at week 24 (Fig. 2). These Hb improvements in the momelotinib versus BAT/ruxolitinib groups were apparent despite median transfusion rates through week 24 being comparable or lower with momelotinib [units/month (range), Hb < 100 g/L: 1.6 (0–8.2) vs. 1.4 (0–7.6); non-TI: 1.2 (0–8.7) vs. 1.8 (0–7.6)].

Open in a separate window

Fig. 2

Mean Hb levels over time in the baseline Hb < 100 g/L (A) and non-TI (B) subgroups. Results through week 84 are shown for illustrative purposes, although the study continued beyond this time point. BAT best available therapy, BL baseline, Hb hemoglobin, TI transfusion independent

Spleen, Symptom, and Dual/Triple Responses

Although these subgroups of interest were selected primarily to evaluate the anemia-related benefits of momelotinib versus BAT/ruxolitinib, rates of splenic response [Hb < 100 g/L: 6 (9.1%) vs. 2 (5.1%); non-TI: 7 (9.7%) vs. 1 (3.0%)] and symptom response [Hb < 100 g/L: 21 (32.3%) vs. 1 (2.6%); non-TI: 21 (29.2%) vs. 0] at week 24 were also higher with momelotinib (Fig. 3).

Open in a separate window

Fig. 3

SVR35 and TSS50 at week 24 in the baseline Hb < 100 g/L and non-TI subgroups. For TSS50, response rates are based on the number of patients evaluable for TSS at week 24. BAT best available therapy, Hb hemoglobin, SVR35 spleen volume reduction ≥ 35%, TI transfusion independent, TSS50 total symptom score reduction ≥ 50%. ^aIn the Hb < 100 g/L subgroup, n = 65 for the momelotinib arm and n = 38 for the BAT/ruxolitinib arm

Many responders with momelotinib achieved 2 or all 3 endpoints (splenic, symptom, and transfusion independence response), including 15 of 33 responders (45.5%) in the Hb < 100 g/L subgroup and 16 of 36 responders (44.4%) in the non-TI subgroup; there were no dual or triple responses in the BAT/ruxolitinib arm in either subgroup. In the Hb < 100 g/L subgroup, 3 of 6 splenic responders (50.0%) and 10 of 21 symptom responders (47.6%) with momelotinib were also TI at week 24. Similarly, in the non-TI subgroup, 4 of 7 splenic responders (57.1%) and 10 of 21 symptom responders (47.6%) with momelotinib were also TI at week 24 (Supplementary Material).

Discussion

These exploratory analyses of the phase 3 SIMPLIFY-2 trial focused on JAK inhibitor-experienced patient subgroups for whom anemia management is typically a key consideration: patients with Hb of < 100 g/L and patients who were non-TI. In both populations, anemia-related benefits and continued spleen and symptom control were greater with momelotinib than with BAT/ruxolitinib, suggesting that it may be advantageous to switch to momelotinib versus continuing ruxolitinib and managing anemia with supportive care. Because there was no washout of prior ruxolitinib in SIMPLIFY-2, these results complement a recent analysis of the phase 3 SIMPLIFY-1 trial illustrating that patients can be immediately transitioned from ruxolitinib to momelotinib at full dose with no safety concerns, derive rapid anemia benefits, and maintain splenic and symptom control [12].

The subgroups in these analyses comprise similar but not completely overlapping patient populations. Hb of < 100 g/L broadly constitutes moderate-to-severe anemia, and while supportive therapies are typically considered at this threshold, the introduction of RBC transfusions may be reserved for more severely anemic individuals [3, 13–17]. Thus, it is possible to encounter patients with Hb of < 100 g/L who require anemia-directed therapy but are nevertheless TI (14 patients in the present analysis). Conversely, as transfusions may themselves increase Hb levels, there may be patients with Hb of ≥ 100 g/L who are nevertheless TD or TR (as is the case for a separate group of 14 patients in the present analysis). While anemia-related benefits are of particular importance in the management of both of these subgroups, the goals of treatment may vary (e.g., Hb improvement versus transfusion burden reduction). Momelotinib was associated with comprehensive anemia benefits, including higher mean Hb levels over time and increased rates of week 24 transfusion independence, versus BAT/ruxolitinib in both subgroups.

Anemia supportive therapies do not address the underlying mechanism of anemia in myelofibrosis and are unable to halt its progression; however, adding these supportive therapies to ongoing treatment with a JAK inhibitor such as ruxolitinib may be perceived as a preferred approach, particularly in patients who are otherwise experiencing symptom and spleen control [1, 3, 18]. In fact, our analyses suggest that this approach provides not only limited anemia benefit but also lower symptom and spleen control compared with switching to momelotinib. Consistent with the ITT analysis, in which the primary superiority endpoint was not met, splenic response rates were low across treatment arms and subgroups, likely attributable to the lack of washout from prior ruxolitinib [9]. However, splenic response rates with momelotinib in these subgroups were higher than in the ITT population (ITT, 6.7%; subgroups, 9.1% and 9.7%), while those with BAT/ruxolitinib were lower (ITT, 5.8%; subgroups, 5.1% and 3.0%). A similar trend was observed with symptom responses; notably, no patients in the non-TI subgroup, and only 1 in the Hb < 100 g/L subgroup, treated with BAT/ruxolitinib had a symptom response, while approximately 30% did with momelotinib. These poor splenic and symptom response rates in the BAT/ruxolitinib arm may be the result of suboptimal ruxolitinib dosing, as most patients in these subgroups received ≤ 20 mg daily. These observations are consistent with recent real-world analyses of ruxolitinib, which found lower overall doses, splenic and symptom response rates, and survival in patients with cytopenias, primarily anemia [19]. In contrast, most patients in the present analysis received full daily doses of momelotinib, which not only provides spleen and symptom benefits through inhibition of JAK1/JAK2 but directly addresses anemia in myelofibrosis through inhibition of ACVR1, leading to increased serum iron availability and erythropoiesis [3, 20]. Thus, the decision to continue first-line treatment or switch may be informed in part by the dose of the initial JAK inhibitor that can be maintained and the corresponding depth and durability of spleen and symptom responses.

The prespecified terminal 12-week definition of transfusion independence in SIMPLIFY-2 [9] ensures the durability of the rate reported at week 24, as it requires patients to have remained transfusion-free for at least the previous 12 weeks. In contrast, the rolling 12-week definition [21] is less stringent, as patients may achieve transfusion independence over any 12-week period through week 24; even if this independence is later lost, they will still be counted as responders at week 24. The fact that week 24 transfusion independence rates with momelotinib were similar by the terminal and rolling 12-week definitions (e.g., 40.3% vs. 34.7% in the non-TI subgroup) provides further evidence of durability, as it indicates that few patients lost response with momelotinib once achieved; in contrast, few patients met the strict terminal 12-week criteria for transfusion independence in the BAT/ruxolitinib arm.

The most common anemia supportive therapy, other than transfusions, in the BAT/ruxolitinib arm of SIMPLIFY-2 was ESAs (5 patients). Current guidelines recommend ESAs primarily for patients with serum EPO levels of < 500 IU/L [11], with responses more common in those with levels of < 125 IU/L [16]. In the subgroups evaluated in the present analysis, more patients in the BAT/ruxolitinib arm versus the momelotinib arm had baseline EPO levels of < 500 IU/L, including approximately 40–50% versus 30% with levels of < 100 IU/L. However, transfusion independence responses with momelotinib were observed regardless of baseline EPO level, suggesting that this criterion should not impact consideration of momelotinib in patients who require an anemia benefit. Notably, no patients in either SIMPLIFY-2 subgroup of interest achieved transfusion independence at week 24 with ESAs.

The primary limitation of the present analyses is their post hoc and descriptive nature; SIMPLIFY-2 was not powered to prospectively evaluate momelotinib versus BAT/ruxolitinib in these patient subgroups; thus, statistical significance could not be determined. The crossover design of SIMPLIFY-2 also represents a limitation, as this precludes comparative analyses of endpoints such as duration of transfusion independence beyond week 24.

Conclusion

Collectively, these data suggest that, in patients with moderate-to-severe anemia and/or in need of RBC transfusions, outcomes—notably anemia benefits, including week 24 transfusion independence rates, onset and duration of transfusion independence, median transfusion rates through week 24, and mean Hb levels over time—are improved by switching to momelotinib rather than continuing ruxolitinib and using ESAs or other supportive therapies to manage anemia.

Acknowledgements

Author Contributions

Claire N. Harrison, Alessandro M. Vannucchi, Christian Recher, Francesco Passamonti, Aaron T. Gerds, Juan Carlos Hernandez-Boluda, Abdulraheem Yacoub, Shireen Sirhan, and Uwe Platzbecker were involved in data collection. Catherine Ellis, Bharat Patel, and Bryan Strouse were involved in planning and/or performance of the analysis. All authors had access to the data and contributed to its interpretation, participated in writing and reviewing the manuscript, approved the final version, and agreed to be accountable for the accuracy and integrity of the data.

Funding

These analyses of the SIMPLIFY-2 trial were supported by GSK. The journal’s Rapid Service and Open Access Fees were also funded by GSK.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Information on GSK’s data sharing commitments and requesting access to anonymized individual participant data and associated study documents can be found at https://www.gsk-studyregister.com/en/.

Declarations

Footnotes

References