ABSTRACT
Ruxolitinib is first‐line therapy for intermediate/high‐risk myelofibrosis (MF), but ∼50% of patients discontinue within 1 year due to loss of efficacy or intolerance. Four JAK inhibitors (gecacitinib, fedratinib, pacritinib, momelotinib) are approved for ruxolitinib‐pretreated MF, with no head‐to‐head trials comparing their efficacy and safety. This study used matching‐adjusted indirect comparison (MAIC) to compare these four agents, aiming to provide evidence‐based insights for treatment decision‐making in ruxolitinib‐resistant or intolerant MF. Individual patient data (IPD) of gecacitinib (100 mg BID; ZGJAK006/ZGJAK017, n = 78) and published data of comparators (fedratinib: JAKARTA‐2/FREEDOM2; pacritinib: PAC203; momelotinib: SIMPLIFY‐2/MOMENTUM) were analyzed. Eight baseline characteristics were matched. Efficacy outcomes (week‐24 SVR35, TSS50, transfusion independence [TI]) were reported as odds ratios (ORs); safety as risk differences (RDs). Gecacitinib showed superior SVR35 versus fedratinib (JAKARTA‐2: OR = 3.96, 95% CI = 1.37–11.39, P = 0.0108), pacritinib (PAC203: OR = 6.10, 95% CI = 1.54–24.23, P = 0.0101), and momelotinib (SIMPLIFY‐2: OR = 8.65, 95% CI = 1.86–40.31, P = 0.0060), and superior TSS50 versus pacritinib (OR = 7.62 95% CI = 1.84–31.51, P = 0.0050) and momelotinib (MOMENTUM: OR = 7.52, 95% CI = 1.63–34.61, P = 0.0096). Numerically, gecacitinib had better TI. It also had significantly lower incidences of diarrhea, nausea, and AE‐related treatment discontinuation. Hematologic AE profiles of gecacitinib varied by comparator cohort. Gecacitinib showed favorable efficacy and tolerability signals versus several comparators, suggesting it may be a valuable second‐line option.
Keywords: gecacitinib, JAK inhibitor, matching‐adjusted indirect comparison (MAIC), myelofibrosis, second‐line treatment
1. Introduction
Primary myelofibrosis (PMF), post‐polycythemia vera myelofibrosis (PPV‐MF), and post‐essential thrombocythemia myelofibrosis (PET‐MF) are collectively classified as BCR::ABL1‐negative myeloproliferative neoplasm‐associated myelofibrosis (MF). Clinical manifestations of MF include hematologic abnormalities (e.g., elevated white blood cell count, anemia, thrombocytopenia), splenomegaly, extramedullary hematopoiesis, and portal hypertension, with ∼20% of patients progressing to acute myeloid leukemia within 10 years of diagnosis [1].
Ruxolitinib, the first approved Janus kinase (JAK) inhibitor for intermediate‐ and high‐risk MF, effectively mitigates splenomegaly and constitutional symptoms [2, 3]. However, 40%–70% of patients discontinue ruxolitinib within 1 year due to loss of efficacy or intolerance [4]. Following discontinuation, symptoms and splenomegaly rapidly revert to pretreatment levels, and salvage therapy for ruxolitinib‐failed MF yields suboptimal response rates, with a median overall survival (OS) of 6–24 months [4, 5]. Ruxolitinib resistance drives disease progression, while drug‐induced anemia and intolerance‐related thrombocytopenia further compromise patient outcomes—highlighting an urgent unmet clinical need for second‐line agents with improved efficacy and favorable hematologic safety profiles.
Currently, no universal consensus defines “ruxolitinib failure,” which typically encompasses primary resistance, loss of initial response, intolerance, or progressive disease during treatment. According to the NCCN Guidelines, three JAK inhibitors are approved for ruxolitinib‐pretreated MF: fedratinib, pacritinib, and momelotinib [6, 7]. Fedratinib, a JAK2‐selective inhibitor, prioritizes splenomegaly reduction, as supported by data from the single‐arm JAKARTA‐2 [8] and two‐arm FREEDOM2 studies [9]. Pacritinib, with minimal JAK1 activity, exhibits mild myelosuppression, making it suitable for MF patients with concurrent cytopenias; it was evaluated in the single‐arm PAC203 study [10] and two‐arm PERSIST‐2 (48% ruxolitinib‐pretreated patients) [11] trials. Momelotinib, a dual JAK1/JAK2 and ACVR1 inhibitor, improves anemia by suppressing ACVR1‐mediated hepcidin production, tested in SIMPLIFY‐2 and MOMENTUM [12, 13].
In China, gecacitinib—a novel JAK inhibitor approved in 2025—was authorized for ruxolitinib‐relapsed/refractory(R/R) or intolerant MF based on two single‐arm trials (ZGJAK006 and ZGJAK017) [14, 15, 16, 17, 18]. Numerical differences in key efficacy outcomes exist between gecacitinib and the FDA approved JAK inhibitors: at week 24, the proportion of patients achieving ≥ 35% spleen volume reduction (SVR35) and ≥ 50% total symptom score reduction (TSS50) varies across trials [8, 9, 10, 12, 13]. Safety profiles also differ: fedratinib is associated with gastrointestinal toxicities (nausea, diarrhea) and requires monitoring for Wernicke's encephalopathy [8, 9, 19]; momelotinib's prominent adverse events (AEs) include diarrhea and peripheral neuropathy [20]; pacritinib was linked to diarrhea and nausea in PERSIST‐1 and PERSIST‐2, but PAC203 reported significantly lower incidences of these AEs, attributed to revised inclusion criteria and dose adjustment strategies [10, 11, 21]; Numerically, gecacitinib generally causes mild AEs, with lower peripheral neuropathy incidence than momelotinib and reduced diarrhea/nausea versus fedratinib.
Notably, no head‐to‐head trials have directly compared these four JAK inhibitors, limiting evidence‐based decision‐making for ruxolitinib‐pretreated MF. To address this gap, we employed a matching‐adjusted indirect comparison (MAIC) to adjust for baseline disparities among trials, systematically evaluating the efficacy and safety of gecacitinib versus fedratinib, pacritinib, and momelotinib.
2. Methods
2.1. Data Sources
Data were extracted from completed second‐line clinical trials of four JAK inhibitors for ruxolitinib‐pretreated MF, with priority given to peer‐reviewed publications. Week‐24 efficacy and safety outcomes were included from the following trials (Table 1). Individual patient data (IPD) for gecacitinib (100 mg twice daily [BID]) were pooled from two Phase 2 trials: ZGJAK006 (ruxolitinib‐intolerant MF; NCT04217993) [14] and ZGJAK017 (ruxolitinib‐relapsed/refractory [RR] MF; NCT04851535) [15]. Aggregate data for comparators were derived from published reports: fedratinib (JAKARTA‐2 [8], NCT01523171; FREEDOM2 [9], NCT03952039), momelotinib (SIMPLIFY‐2 [12], NCT02101268; MOMENTUM [13], NCT04173494), and pacritinib (PAC203 [10], NCT04884191).
TABLE 1.
Characteristics of Included Studies for Matching‐adjusted indirect comparison (MAIC) in Ruxolitinib‐pretreated Myelofibrosis.
| Investigated drug | Trial name | NCT number | Phase | Primary comparison/design | Patient population |
|---|---|---|---|---|---|
| Gecacitinib | ZGJAK006 | NCT04217993 | Phase 2 | Single‐arm; | Ruxolitinib‐intolerant |
| Gecacitinib | ZGJAK017 | NCT04851535 | Phase 2 | Single‐arm | Ruxolitinib‐relapsed/refractory |
| Fedratinib | JAKARTA‐2 | NCT01523171 | Phase 2 | Single‐arm; | Ruxolitinib‐resistant or ‐intolerant |
| Fedratinib | FREEDOM2 | NCT03952039 | Phase 3b | Fedratinib versus Best available therapy (77.6% ruxolitinib) | Previously treated with ruxolitinib |
| Pacritinib | PAC203 | NCT04884191 | Phase 3 | Pacritinib versus Physician's choice | Ruxolitinib intolerant (3%) or failure (76%), 50% met criteria for both |
| Momelotinib | SIMPLIFY‐2 | NCT02101268 | Phase 3 | Momelotinib versus Best available therapy (89% ruxolitinib) | Ruxolitinib suboptimal responses or hematologic intolerance |
| Momelotinib | MOMENTUM | NCT04173494 | Phase 3 | Momelotinib versus Danazol | JAK inhibitor‐pretreated anemic MF,100% with ruxolitinib and 5% with fedratinib |
2.2. MAIC Analysis
This matching‐adjusted indirect comparison (MAIC) was conducted in adherence to the Declaration of Helsinki, Good Clinical Practice (GCP) guidelines established by the International Conference on Harmonization (ICH), and received prior approval from the relevant local institutional review boards (IRBs) or independent ethics committees (IECs). Eight baseline characteristics were selected based on established MF trial literature [22, 23, 24, 25]: Platelet count, HGB level, baseline TSS, transfusion independence (TI) status, and duration of prior JAK inhibitor therapy were regarded as potential effect modifiers, and DIPSS risk category, MF subtype, age were considered prognostic factors.
Across trials, SVR35 was defined using consistent MRI/CT criteria. Although TSS questionnaire versions varied, cross‐sectional validity ensured inter‐study comparability [26, 27], the version used in each trial is listed in Supporting Information S1. Mean hemoglobin (Hb) levels at baseline and week 24 were extracted from published line graphs using WebPlotDigitize version 4.6 (https://automeris.io/WebPlotDigitizer/) [28]. In our study, Hb response and TI were defined following each trial's predefined criteria and one‐to‐one comparison with gecacitinib, respectively. All safety data were confined to the 24‐week treatment period, and AEs were graded per Common Terminology Criteria for Adverse Events (CTCAE) v.4.0 to v5.0.
To assess baseline balance and the preservation of statistical information, the effective sample size (ESS) post‐weighting was calculated using the equation (Σwj) [2]/Σwj [2], with wj denoting the individual‐specific weighting factor. For comparators reporting only median values for continuous variables, variables were dichotomized using the median as the cutoff to enable proportional matching.
Pooled IPD for gecacitinib were reweighted using a MAIC approach to align eight baseline characteristics with those reported for each comparator trial. Efficacy outcomes (SVR35, TSS50, TI) were analyzed using weighted logistic regression models with a robust sandwich variance estimator, reported as odds ratios (ORs) with 95% confidence intervals (CIs) and p‐values; an OR > 1 indicated superior efficacy of gecacitinib. Safety outcomes (anemia, thrombocytopenia, neutropenia, diarrhea, nausea, peripheral neuropathy, herpes zoster, AE‐related treatment discontinuation, grade 3/4 AEs) were reported as risk differences (RDs) with 95% CIs and p‐value; an RD < 0 indicated a lower risk of the AE with gecacitinib.
Standard errors for ORs and RDs were calculated using a robust sandwich estimator to account for potential residual confounding. Statistical significance was defined as a two‐sided p‐value < 0.05. All analyses were performed using R software version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria).
3. Results
3.1. Baseline Characteristic Matching
IPD for gecacitinib were pooled from ZGJAK006 and ZGJAK017 (n = 78). Comparator cohorts included fedratinib (FREEDOM2, n = 134; JAKARTA‐2 stringent criteria cohort, n = 79), momelotinib (SIMPLIFY‐2, n = 104; MOMENTUM, n = 130), and pacritinib (PAC203 200 mg twice daily [BID], n = 54).
Due to the heterogeneity of baseline characteristics in ruxolitinib‐pretreated MF patients and the inclusion of eight matching variables, the ESS post‐weighting was relatively modest (Table 2). ESS and corresponding population proportions (in descending order) were: JAKARTA‐2 fedratinib (ESS = 20.43, 26.19%), PAC203 pacritinib (ESS = 17.73, 22.73%), FREEDOM2 fedratinib (ESS = 12.17, 15.61%), SIMPLIFY‐2 momelotinib (ESS = 11.63, 14.91%), and MOMENTUM momelotinib (ESS = 8.46, 10.84%). After weighting, baseline characteristics were well‐balanced between gecacitinib and each comparator. Summary statistics before/after weighting (Table 2) and versus comparators (Supporting Information S1) are provided.
TABLE 2.
Baseline characteristics of gecacitinib before and after adjustment.
| Before adjustment (n = 78) | After adjustment to JAKARTA‐2 (ESS = 20.43) | After adjustment to FREEDOM2 (ESS = 12.17) | After adjustment to PAC203 (ESS = 17.73) | After adjustment to SIMPLIFY‐2 (ESS = 11.63) | After adjustment to MOMENTUM (ESS = 8.46) | |
|---|---|---|---|---|---|---|
| Age, years | ||||||
| Mean (SD) | 57.9 (10.95) | 63.2 (9.89) | 64.9 (8.56) | 63.8 (9.52) | 66.4 (6.73) | 65.7 (7.59) |
| Median | 59.5 | 65.0 | 70.0 | 68.0 | 65.0 | 70.0 |
| Sex | ||||||
| Male | 59.0 | 43.8 | 35.0 | 62.4 | 72.0 | 59.4 |
| Eemale | 41.0 | 56.2 | 65.0 | 37.6 | 28.0 | 40.6 |
| Disease subtype | ||||||
| PMF | 83.3 | 60.0 | 56.0 | 68.5 | 62.0 | 60.0 |
| Post‐ET‐MF | 9.0 | 17.0 | 19.0 | 13.0 | 21.0 | 19.0 |
| Post‐PV‐MF | 7.7 | 23.0 | 25.0 | 18.5 | 17.0 | 21.0 |
| DIPSS risk status | ||||||
| Intermediate‐1 | 29.5 | 14.0 | 2.0 | 22.2 | 22.0 | 7.0 |
| Intermediate‐2 | 62.8 | 52.0 | 76.0 | 51.9 | 60 | 55.0 |
| High | 7.7 | 34.0 | 22.0 | 25.9 | 18.0 | 38.0 |
| MPN‐SAF TSS | ||||||
| Mean (SD) | 17.2 (14.46) | 20.0 (18.60) | 24.9 (16.93) | 19.0 (15.42) | 18.5 (13.29) | 22.0 (13.89) |
| Median | 12.50 | 12.0 | 25.0 | 17.0 | 16.0 | 25.0 |
| HGB, g/L | ||||||
| Mean (SD) | 88.4(21.75) | 87.0 (19.01) | 82.6 (13.84) | 82.0 (19.81) | 94.0 (17.76) | 76.3 (17.24) |
| Median | 87.0 | 88.0 | 86.0 | 79.0 | 93.0 | 79.0 |
| PLT, 109/L | ||||||
| Mean (SD) | 203.8 (153.07) | 224.5 (216.08) | 155.3 (87.01) | 303.0 (325.80) | 170.8 (156.88) | 151.7 (56.30) |
| Median | 143.0 | 145.0 | 123.0 | 136.0 | 136.0 | 126.0 |
| TI* rate, % | 55.1 | 63.5 | 52.8 | 37.0 | 31.0 | 49.4 |
| Duration of prior JAKi therapy, days | ||||||
| Mean (SD) | 599.8 (517.76) | 350.0 (256.07) | 609.4 (369.79) | 584.4 (411.40) | 614.3 (370.38) | 969.5 (672.01) |
| Median | 394.0 | 265.0 | 654.0 | 501.0 | 501.0 | 891.0 |
| Slpeen volume, cm3 | ||||||
| Mean (SD) | 1757.9 (1106.6) | 1437.5 (645.2) | 1370.4 (742.7) | 1361.7 (696.6) | 1580.4 (778.4) | 1676.8 (999.0) |
| Median | 1515.4 | 1253.8 | 1185.5 | 1124.9 | 1567.3 | 1187.0 |
Note: Data are presented as median (no range reported), mean (SD), or percentage (%).
Abbreviations: ESS, effective sample size; DIPSS, Dynamic International Prognostic Scoring System; TSS, total symptom score; HGB, hemoglobin level; PLT, platelet count; TI, transfusion independence. *The definitions of transfusion independence used for matching pacritinib and momelotinib were not identical.
3.2. Efficacy Analysis
3.2.1. SVR35
At week 24, gecacitinib's weighted SVR35 rate was higher than most comparators, except MOMENTUM momelotinib (19.8% vs. Twenty‐three percent; OR = 0.94, 95% CI: 0.17 to 5.12, P = 0.9431). Statistically significant superiority was observed for gecacitinib versus JAKARTA‐2 fedratinib (OR = 3.96, 95% CI: 1.37 to 11.39, P = 0.0108), PAC203 pacritinib (OR = 6.10, 95% CI: 1.54 to 24.23, P = 0.0101) and SIMPLIFY‐2 momelotinib (OR = 8.65, 95% CI: 1.86 to 40.31, P = 0.0060), and (Figure 1). The most prominent numerical difference in SVR35 rates was between gecacitinib and JAKARTA‐2 fedratinib (60.5% vs. 30.0%). Detailed weighted and unweighted SVR35 rates are reported in Supporting Information S1.
FIGURE 1.
Comparative analysis of SVR35 at Week 24 among gecacitinib, fedratinib, pacritinib and momelotinib in second‐line therapy. ORs of > 1 indicate outcomes favoring gecacitinib; OR, Odds ratio.
3.2.2. TSS50
Gecacitinib's weighted TSS50 rate exceeded all comparators, with significant advantages versus PAC203 pacritinib (OR = 7.62, 95% CI: 1.84 to 31.51, P = 0.0050) and MOMENTUM momelotinib (OR = 7.52, 95% CI: 1.63 to 34.61, P = 0.0096), Figure 2. The highest weighted TSS50 rate for gecacitinib was observed versus MOMENTUM momelotinib (71.1% vs. 25.0%). Supporting Information S1 provides detailed TSS50 rate comparisons.
FIGURE 2.
Comparative analysis of TSS50 at Week 24 among gecacitinib, fedratinib, pacritinib and momelotinib in second‐line therapy. ORs of > 1 indicate outcomes favoring gecacitinib; OR, Odds ratio.
3.2.3. HGB Improvement
Patients were matched per each trial's predefined Hb‐response criteria. Per PAC203 pacritinib criteria—≥ 10 or ≥ 20 g/L Hb increase from baseline (Hb ≤ 100 g/L) without transfusions in the 8 weeks before week 24—gecacitinib achieved a higher response rate (≥ 10 g/L: 26.1% vs. 9.5%; OR = 3.36, 95% CI: 0.82–13.83, P = 0.093; ≥ 20 g/L: 16.5% vs. 4.8%; OR = 3.95, 95% CI: 0.71–22.00, P = 0.12). Per MOMENTUM momelotinib criteria—transfusion‐independent patients with a sustained ≥ 20 g/L Hb increment over any 12 consecutive weeks within 24 weeks—gecacitinib showed comparable response rates (39.0% vs. 41.2%; OR = 0.93, 95% CI: 0.14–6.15, P = 0.94).
3.2.4. TI Improvement
Using a consistent TI definition (no red blood cell [RBC] transfusion in the 12 weeks pre‐treatment; HGB ≥ 80 g/L, per SIMPLIFY‐2 and MOMENTUM), gecacitinib's week‐24 TI rate was numerically lower than momelotinib (SIMPLIFY‐2: OR = 0.52, 95% CI: 0.12–2.23, P = 0.38) but higher than momelotinib (MOMENTUM: OR = 2.63, 95% CI: 0.52–13.24, P = 0.24). Gecacitinib showed numerical TI advantage over MOMENTUM momelotinib (53.9% vs. 31.0%). Notably, TSS, hemoglobin improvement, and TI either required harmonization or had residual heterogeneity; we therefore suggest considering these variables as exploratory.
3.3. Safety Analysis
Gecacitinib demonstrated favorable safety profiles, with significantly lower incidences of any‐grade nausea and AE‐related treatment discontinuation across all comparator cohorts (Table 3).
TABLE 3.
Safety outcomes of gecacitinib, fedratinib, pacritinib and momelotinib in second‐line therapy myelofibrosis.
| Gecacitinib versus JAKARTA‐2 fedratinib stringent criteria cohort | Gecacitinib versus FREEDOM2 fedratinib | Gecacitinib versus PAC203 pacritinib 200 mg BID | Gecacitinib versus SIMPLIFY‐2 momelotinib | Gecacitinib versus MOMENTUM momelotinib | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Weighted RD (95% CI) | p‐value | Weighted RD (95% CI) | p‐value | Weighted RD (95% CI) | p‐value | Weighted RD (95% CI) | p‐value | Weighted RD (95% CI) | p‐value | |
| Anemia | ||||||||||
| Grade 3/4 | −0.17 (−0.43, 0.09) | 0.2117 | 0.13 (−0.20, 0.46) | 0.4491 | 0.20 (−0.06, 0.47) | 0.1339 | 0.49 (0.17, 0.80) | 0.0027 | −0.21 (−0.59, 0.17) | 0.2773 |
| Any | −0.16 (−0.42, 0.10) | 0.2255 | NE | NE | 0.28 (−0.02, 0.57) | 0.0669 | 0.49 (0.18, 0.81) | 0.0023 | −0.55 (−0.93, −0.16) | 0.0055 |
| Thrombocytopenia | ||||||||||
| Grade 3/4 | −0.05 (−0.19, 0.09) | 0.4747 | 0.06 (−0.27, 0.39) | 0.7282 | −0.07 (−0.31, 0.16) | 0.5258 | 0.28 (−0.05, 0.60) | 0.0930 | 0.01 (−0.32, 0.34) | 0.9547 |
| Any | 0.17 (−0.09, 0.43) | 0.1961 | NE | NE | 0.24 (−0.01, 0.49) | 0.0639 | 0.37 (0.03, 0.70) | 0.0310 | −0.29 (−0.69, 0.11) | 0.1483 |
| Neutropenia | ||||||||||
| Grade 3/4 | NE | NE | NE | NE | 0.14 (−0.04, 0.33) | 0.1309 | NE | NE | 0.00 (−0.22, 0.23) | 0.9768 |
| Any | NE | NE | NE | NE | 0.25 (0.03, 0.47) | 0.0238 | NE | NE | −0.09 (−0.36, 0.18) | 0.5076 |
| Diarrhea | ||||||||||
| Any | −0.46 (−0.69, −0.22) | 0.0001 | −0.20 (−0.43, 0.02) | 0.0796 | −0.17 (−0.35, 0.01) | 0.0700 | −0.23 (−0.42, −0.05) | 0.0112 | 0.04 (−0.27, 0.35) | 0.8215 |
| Nausea | ||||||||||
| Any | −0.53 (−0.64, −0.42) | < 0.0001 | NE | NE | −0.26 (−0.38, −0.13) | < 0.0001 | −0.19 (−0.26, −0.11) | < 0.0001 | −0.16 (−0.23, −0.10) | < 0.0001 |
| Peripheral neuropathy | ||||||||||
| Grade 3/4 | NE | NE | NE | NE | NE | NE | −0.01 (−0.03, 0.01) | 0.3196 | NE | NE |
| Any | NE | NE | NE | NE | NE | NE | −0.10 (−0.16, −0.04) | 0.0008 | −0.04 (−0.07, −0.01) | 0.0237 |
| TEAEs (discontinuation) | −0.20 (−0.29, −0.11) | < 0.0001 | −0.10 (−0.15, −0.04) | 0.0002 | −0.17 (−0.28, −0.07) | 0.0014 | −0.13 (−0.20, −0.07) | < 0.0001 | −0.18 (−0.24, −0.11) | < 0.0001 |
Note: RDs of < 0 indicate outcomes favoring gecacitinib.
Abbreviations: NE, Not Estimable; RD, risk difference.
3.3.1. Versus JAKARTA‐2 Fedratinib
Gecacitinib had a lower incidence of AE‐related treatment discontinuation and significantly reduced risks of any‐grade diarrhea (RD = −0.46, 95% CI: −0.69–‐0.22, p = 0.0001) and nausea (RD = −0.53, 95% CI: −0.64–‐0.42, p < 0.0001).
3.3.2. Versus PAC203 Pacritinib
Gecacitinib showed a distinct safety advantage, with fewer AE‐related treatment discontinuations and a significantly lower risk of any‐grade nausea (RD = −0.26, 95% CI: −0.38–‐0.13, p < 0.0001). Numerically higher rates of other hematologic AEs were noted with gecacitinib, but the clinically meaningful reduction in severe thrombocytopenia was notable.
3.3.3. Versus Momelotinib
Safety profiles differed by trial cohort.
Peripheral neuropathy: Significantly lower any‐grade incidence versus SIMPLIFY‐2 (RD = −0.10, 95% CI: −0.16–‐0.04, P = 0.0008) and MOMENTUM (RD = −0.04, 95% CI: −0.07–‐0.01, P = 0.24) momelotinib;
Anemia: Higher any‐grade/grade 3–4 incidence versus SIMPLIFY‐2 momelotinib (any‐grade: RD = 0.49, 95% CI: 0.17–0.80, P = 0.0027; grade 3–4: RD = 0.49, 95% CI: 0.18–0.81, P = 0.0023) but lower versus MOMENTUM momelotinib (any‐grade: RD = −0.21, 95% CI: −0.59–0.17, P = 0.28; grade 3–4: RD = −0.55, 95% CI: −0.93–‐0.16, P = 0.0055);
Thrombocytopenia: Significantly higher any‐grade incidence versus SIMPLIFY‐2 momelotinib (RD = 0.37, 95% CI: 0.03–0.70, P = 0.031) but numerically lower versus MOMENTUM momelotinib (RD = −0.29, 95% CI: −0.69–0.11, P = 0.15);
Diarrhea: Significantly lower any‐grade incidence versus SIMPLIFY‐2 momelotinib (RD = −0.23, 95% CI: −0.42−0.05, P = 0.011) but numerically higher versus MOMENTUM momelotinib (RD = 0.04, 95% CI: −0.27–0.35, P = 0.82).
4. Discussion
This study utilized MAIC to systematically evaluate the efficacy and safety of four JAK inhibitors (gecacitinib, fedratinib, pacritinib, momelotinib) in ruxolitinib‐pretreated MF, addressing the unmet need for head‐to‐head trial data. Key findings indicate that gecacitinib was associated with numerically higher SVR35 and TSS50 rates versus most comparators, with a favorable safety profile.
Ruxolitinib failure lacks a universal definition; however, overlapping inclusion criteria across trials support the validity of indirect comparisons [7, 8, 9, 12, 13]. Notably, the MOMENTUM cohort was restricted to ruxolitinib‐intolerant, anemic MF patients, whereas gecacitinib's pooled population included both ruxolitinib‐intolerant and relapsed/refractory (RR) subgroups—with RR patients historically exhibiting poorer responses to second‐line JAK inhibitors (24‐week SVR35: 43.2% in intolerant vs. 32.4% in RR patients) [14, 15]. This underscores gecacitinib's robust efficacy, even in a more refractory patient subset. Despite this potential confounding, the agent still demonstrated significant SVR35/TSS50 advantages versus SIMPLIFY‐2 momelotinib, PAC203 pacritinib, and JAKARTA‐2 fedratinib.
Anemia and transfusion dependence are major drivers of morbidity in MF, with momelotinib's ACVR1‐inhibitory activity conferring unique anemia benefits [13, 29, 30]. Our analysis showed no significant TI differences between gecacitinib and momelotinib, though gecacitinib numerically outperformed MOMENTUM momelotinib (53.9% vs. 31.0%); notably, HGB improvement rates were virtually identical between the two agents (39.0% vs. 41.2%). The lack of statistical significance for transfusion independence may reflect small ESS or enrichment for anemic patients in MOMENTUM, highlighting the need for subgroup analyses in future studies. While comparisons between pacritinib and gecacitinib did not reach statistical significance, gecacitinib consistently demonstrated markedly higher HGB improvement rates. Importantly, the ORs for HGB improvement observed with gecacitinib versus pacritinib (≥ 10 g/L: OR 3.36; ≥ 20 g/L: OR 3.95) are strikingly similar to momelotinib versus pacritinib (10 g/L: OR 4.89; ≥ 20 g/L: OR 3.40) [25], despite the latter analysis including mixed first‐/second‐line populations.
In terms of safety, gecacitinib's consistent advantage in reducing peripheral neuropathy versus momelotinib, a well‐documented AE of momelotinib [31], is clinically relevant. Notably, gecacitinib's lower gastrointestinal AE rates (diarrhea, nausea) versus fedratinib and SIMPLIFY‐2 momelotinib were observed without prophylactic interventions—contrasting with fedratinib trials, which implemented thiamine monitoring and antiemetic/antidiarrheal prophylaxis [8, 9]. This suggests inherent tolerability of gecacitinib's JAK inhibitory profile, reducing supportive care burdens. Versus PAC203 pacritinib, gecacitinib’ was associated with lower grade 3/4 thrombocytopenia (RD = −0.18, P = 0.026), indicating favorable tolerability in patients with thrombocytopenia.
Second‐line MF therapy requires balancing efficacy, safety, and treatment persistence. Gecacitinib's dual advantages in spleen/symptom control and reduced treatment discontinuation align with unmet needs in this population. Gecacitinib may be considered in ruxolitinib‐exposed patients with high spleen/symptom burden or intolerance to other JAK inhibitors. Anemia‐related findings were mixed and not statistically significant. Additionally, as the first Chinese‐approved JAK inhibitor for ruxolitinib‐pretreated MF, gecacitinib addresses regional access gaps, providing a locally available option for Asian patients.
This study has several limitations. First, despite balancing eight baseline characteristics, we could not match on other potential confounders—such as molecular mutations, prior JAK inhibitor dose, ruxolitinib failure, and ruxolitinib intolerance rate—due to the limitations of aggregate data from comparator trials. Second, variable definitions of ruxolitinib failure across trials may limit cross‐cohort comparability. Third, the small effective sample sizes after weighting, especially for comparisons with momelotinib, mean that the risk differences may be unstable, these results require confirmation in head‐to‐head trials. Fourth, the short follow‐up duration (24 weeks) precluded analysis of long‐term outcomes, which are critical for evaluating MF therapy evaluation.
5. Conclusion
This study suggests that gecacitinib may be a valuable second‐line therapeutic option for ruxolitinib‐pretreated MF patients, potentially offering a balanced combination of efficacy and tolerability. Future long‐term studies and head‐to‐head trials are warranted to confirm these observations and to inform personalized treatment algorithms.
Author Contributions
Y.Z. and J.J. conceived and designed the study. Y.Z., H.Z., Q.K.Z., Q.C.L., S.J.G., Z.J.X., M.H.D., L.M.M., W.W., H.B.D., G.S.H., H.M.J., J.M.L., H.P.Y., H.L.Z., C.K.C., Y.Q.C., X.D., M.H., X.L., J.S.W., N.X., and S.S.S. served as study investigators, collected clinical data, and aided in data analysis and interpretation. Y.Z. and J.J. prepared the initial draft of the manuscript. All authors conducted critical reviews of the manuscript, provided substantial revisions, and approved the final version for submission.
Funding
This research was funded by the Zhejiang Provincial Natural Science Foundation of China (Grant Nos. LMS25H080004; recipient: Y.Z.) and 2022 Ningbo Yongjiang Talent Program (Grant No. 2022B‐018‐G; recipient: J.J.).
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Supporting Information S1
Acknowledgments
We acknowledge financial support from Suzhou Zelgen Biopharmaceuticals Co. Ltd. and the Zhejiang Provincial Natural Science Foundation of China (Grant No. LMS25H080004) and 2022 Ningbo Yongjiang Talent Program (Grant No. 2022B‐018‐G). We express our gratitude to all enrolled patients and their families for their valuable participation, as well as the core research teams for their commitment to the conduct and oversight of the clinical trials.
Data Availability Statement
De‐identified trial data will be made available to the scientific community following proper ethical review and approval. All requests for data access should be submitted to the corresponding author and the trial's dedicated clinical trials unit.
References
- 1. Garmezy B., Schaefer J. K., Mercer J., and Talpaz M., “A Provider's Guide to Primary Myelofibrosis: Pathophysiology, Diagnosis, and Management,” Blood Reviews 45 (2021): 100691, 10.1016/j.blre.2020.100691. [DOI] [PubMed] [Google Scholar]
- 2. Harrison C., Kiladjian J.‐J., Al‐Ali H. K., et al., “JAK Inhibition With Ruxolitinib Versus Best Available Therapy for Myelofibrosis,” New England Journal of Medicine 366, no. 9 (2012): 787–798, 10.1056/nejmoa1110556. [DOI] [PubMed] [Google Scholar]
- 3. Verstovsek S., Mesa R. A., Gotlib J., et al., “A Double‐Blind, Placebo‐Controlled Trial of Ruxolitinib for Myelofibrosis,” New England Journal of Medicine 366, no. 9 (2012): 799–807, 10.1056/nejmoa1110557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Harrison C. N., Schaap N., and Mesa R. A., “Management of Myelofibrosis After Ruxolitinib Failure,” Annals of Hematology 99, no. 6 (2020): 1177–1191, 10.1007/s00277-020-04002-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kuykendall A. T., Shah S., Talati C., et al. “Between a Rux and a Hard Place: Evaluating Salvage Treatment and Outcomes in Myelofibrosis After Ruxolitinib Discontinuation,” Annals of Hematology 97, no. 3 (2018): 435–441, 10.1007/s00277-017-3194-4. [DOI] [PubMed] [Google Scholar]
- 6. National Comprehensive Cancer Network . “Myeloproliferative Neoplasms (Version 2.2025),”2025, [cited July 8, 2025], https://www.nccn.org/professionals/physician_gls/pdf/mpn.pdf.
- 7. Tefferi A., “Primary Myelofibrosis: 2023 Update on Diagnosis, risk‐stratification, and Management,” American Journal of Hematology 98, no. 5 (2023): 801–821, 10.1002/ajh.26857. [DOI] [PubMed] [Google Scholar]
- 8. Harrison C. N., Schaap N., Vannucchi A. M., et al., “Janus Kinase‐2 Inhibitor Fedratinib in Patients With Myelofibrosis Previously Treated With Ruxolitinib (JAKARTA‐2): A Single‐Arm, Open‐Label, Non‐Randomised, Phase 2, Multicentre Study,” Lancet Haematology 4, no. 7 (2017): e317–e324, 10.1016/s2352-3026(17)30088-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Harrison C. N., Mesa R., Talpaz M., et al., “Efficacy and Safety of Fedratinib in Patients With Myelofibrosis Previously Treated With Ruxolitinib (FREEDOM2): Results From a Multicentre, Open‐Label, Randomised, Controlled, Phase 3 Trial,” Lancet Haematology 11, no. 10 (2024): e729–e740, 10.1016/s2352-3026(24)00212-6. [DOI] [PubMed] [Google Scholar]
- 10. Gerds A. T., Savona M. R., Scott B. L., et al., “Determining the Recommended Dose of Pacritinib: Results From the PAC203 Dose‐Finding Trial in Advanced Myelofibrosis,” Blood Advances 4, no. 22 (2020): 5825–5835, 10.1182/bloodadvances.2020003314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Mascarenhas J., Hoffman R., Talpaz M., et al., “Pacritinib Vs Best Available Therapy, Including Ruxolitinib, in Patients With Myelofibrosis: A Randomized Clinical Trial,” JAMA Oncology 4, no. 5 (2018): 652–659, 10.1001/jamaoncol.2017.5818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Harrison C. N., Vannucchi A. M., Platzbecker U., et al., “Momelotinib Versus Best Available Therapy in Patients With Myelofibrosis Previously Treated With Ruxolitinib (SIMPLIFY 2): A Randomised, Open‐Label, Phase 3 Trial,” Lancet Haematology 5, no. 2 (2018): e73–e81, 10.1016/s2352-3026(17)30237-5. [DOI] [PubMed] [Google Scholar]
- 13. Verstovsek S., Gerds A. T., Vannucchi A. M., et al. “Momelotinib Versus Danazol in Symptomatic Patients With Anaemia and Myelofibrosis (MOMENTUM): Results From an International, Double‐Blind, Randomised, Controlled, Phase 3 Study,” Lancet 401, no. 10373 (2023): 269–280, 10.1016/s0140-6736(22)02036-0. [DOI] [PubMed] [Google Scholar]
- 14. Zhang Y., Zhou H., Duan M., et al., “Safety and Efficacy of Jaktinib (A Novel JAK Inhibitor) in Patients With Myelofibrosis Who Are Intolerant to Ruxolitinib: A Single‐Arm, Open‐Label, Phase 2, Multicenter Study,” American Journal of Hematology 98, no. 10 (2023): 1588–1597, 10.1002/ajh.27033. [DOI] [PubMed] [Google Scholar]
- 15. Zhang Y., Zhang Q., Liu Q., et al., “Safety and Efficacy of Jaktinib (A Novel JAK Inhibitor) in Patients With Myelofibrosis Who Are Relapsed or Refractory to Ruxolitinib: A Single‐Arm, Open‐Label, Phase 2, Multicenter Study,” American Journal of Hematology 98, no. 10 (2023): 1579–1587, 10.1002/ajh.27031. [DOI] [PubMed] [Google Scholar]
- 16. Administration N. M. P., “NMPA Approves Gecacitinib Hydrochloride Tablets in China,”2025, : [cited May 29, 2025], https://www.nmpa.gov.cn/zhuanti/cxylqx/cxypxx/20250529152155157.html.
- 17. Tefferi A., Gangat N., and Pardanani A., “Jaktinib (JAK1/2 Inhibitor): A Momelotinib Derivative With Similar Activity and Optimized Dosing Schedule,” American Journal of Hematology (2022). [DOI] [PubMed] [Google Scholar]
- 18. Tefferi A., “Jaktinib and Momelotinib for the Treatment of Myelofibrosis‐Birds of a Feather?,” American Journal of Hematology (2023). [DOI] [PubMed] [Google Scholar]
- 19. Pardanani A., Harrison C., Cortes J. E., et al., “Safety and Efficacy of Fedratinib in Patients With Primary or Secondary Myelofibrosis: A Randomized Clinical Trial,” JAMA Oncology 1, no. 5 (2015): 643–651, 10.1001/jamaoncol.2015.1590. [DOI] [PubMed] [Google Scholar]
- 20. Verstovsek S., Mesa R., Gupta V., et al., “Momelotinib Long‐Term Safety and Survival in Myelofibrosis: Integrated Analysis of Phase 3 Randomized Controlled Trials,” Blood Advances 7, no. 14 (2023): 3582–3591, 10.1182/bloodadvances.2022009311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Mesa R. A., Vannucchi A. M., Mead A., et al., “Pacritinib Versus Best Available Therapy for the Treatment of Myelofibrosis Irrespective of Baseline Cytopenias (PERSIST‐1): An International, Randomised, Phase 3 Trial,” Lancet Haematology 4, no. 5 (2017): e225–e236, 10.1016/s2352-3026(17)30027-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Abdelmagid M., Palmer J. M., Al‐Kali A., et al., “Predictors of Response to Pacritinib Therapy and Survival Among 60 Mayo Clinic Patients With Myelofibrosis Treated Outside of Clinical Trials,” Haematologica 110, no. 9 (2025): 2224–2229, 10.3324/haematol.2025.287576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Zhang Y., Zhou H., Xiao Z., et al., “Exploratory Analysis of Gecacitinib in Ruxolitinib‐Intolerant Myelofibrosis Patients: Impact of Baseline Anemia and Prior Ruxolitinib Treatment Duration On Response To Gecacitinib,” HemaSphere 9, no. 1 (2025): e70152, 10.1002/hem3.70152. [DOI] [Google Scholar]
- 24. Masarova L., Verstovsek S., Palandri F., et al., “Indirect Treatment Comparisons of Momelotinib Vs Pacritinib Safety and Anemia Outcomes in Patients With Myelofibrosis,” Future Oncology 21, no. 16 (2025): 2067–2075, 10.1080/14796694.2025.2511562. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Masarova L., Verstovsek S., Palandri F., et al., “Indirect Treatment Comparison of Momelotinib Vs Fedratinib Safety in Patients With Myelofibrosis,” Future Oncology 21, no. 16 (2025): 2077–2087, 10.1080/14796694.2025.2511564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Dueck A. C., Gwaltney C., Chen W.‐H., et al., “Quantitative Testing of the Myelofibrosis Symptom Assessment Form Version 4.0, A Harmonized Patient‐Reported Outcome Measure for Collecting Key Secondary Endpoint Data in Myelofibrosis Clinical Trials,” Blood 130 (2017): 2168, 10.1182/blood.V130.Suppl_1.2168.2168. [DOI] [Google Scholar]
- 27. Gupta V., Mascarenhas J., Kremyanskaya M., et al., “Matching‐Adjusted Indirect Comparison of the Pelabresib‐Ruxolitinib Combination Vs Jaki Monotherapy in Myelofibrosis,” Blood Advances 7, no. 18 (2023): 5421–5432, 10.1182/bloodadvances.2023010628. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Rohatgi A. 2024, [cited 7 July 2025], https://automeris.io/WebPlotDigitizer/.
- 29. Duminuco A., Chifotides H. T., Giallongo S., Giallongo C., Tibullo D., and Palumbo G. A., “ACVR1: A Novel Therapeutic Target to Treat Anemia in Myelofibrosis,” Cancers 16, no. 1 (2023): 154, 10.3390/cancers16010154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Tefferi A., Pardanani A., and Gangat N., “Momelotinib (JAK1/JAK2/ACVR1 Inhibitor): Mechanism of Action, Clinical Trial Reports, and Therapeutic Prospects Beyond Myelofibrosis,” Haematologica (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Abdelrahman R. A., Begna K. H., Al‐Kali A., et al., “Momelotinib Treatment‐Emergent Neuropathy: Prevalence, Risk Factors and Outcome in 100 Patients With Myelofibrosis,” British Journal of Haematology 169, no. 1 (2015): 77–80, 10.1111/bjh.13262. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information S1
Data Availability Statement
De‐identified trial data will be made available to the scientific community following proper ethical review and approval. All requests for data access should be submitted to the corresponding author and the trial's dedicated clinical trials unit.