Why This Study Matters
Myelofibrosis (MF) is a rare myeloproliferative neoplasm with substantial morbidity and mortality. Ruxolitinib, a JAK1/2 inhibitor, is the standard first-line treatment for most patients, but the authors note that 50% to 70% of patients discontinue within 3 to 5 years because of loss of response or intolerance. Navitoclax is an oral inhibitor of the antiapoptotic proteins BCL-XL, BCL-2 and BCL-W; preclinical work showed greater activity when combined with ruxolitinib in JAK2 V617F–mutated cells.
Study Design
TRANSFORM-1 is a randomized, double-blind, placebo-controlled, multicenter phase 3 study in JAK inhibitor–naïve adults. All patients received ruxolitinib at the labeled dose and were randomized 1:1 to navitoclax (200 mg/d, or 100 mg/d escalated to 200 mg/d depending on baseline platelet count) or placebo.
252 patients enrolled; 1:1 randomization; 125 allocated to NAV+RUX and 127 to PBO+RUX; median follow-up 20.3 months (range 0.0-35.3). Number of sites and countries: not specified in source.
Patient Population
Eligible adults had intermediate-2 or high-risk MF by DIPSS Plus, measurable splenomegaly, and ECOG performance status of ≤2. More than 80% had intermediate-2 risk and almost 50% had high-molecular-risk (HMR) mutations. Median hemoglobin was below 10 g/dL in both arms (9.9 g/dL NAV+RUX; 9.8 g/dL PBO+RUX).
Primary Endpoint Results
The primary endpoint was a ≥35% spleen volume reduction at week 24 (SVR35W24) by central imaging. It was achieved by 63.2% (n = 79) in the NAV+RUX arm and 31.5% (n = 40) in the PBO+RUX arm (P < .0001). SVR35 at any time was reached by 76.8% (n = 96) vs 44.1% (n = 56) (nominal P < .0001).
In JAKi-naïve patients with MF, NAV+RUX led to a twofold improvement in SVR35W24 vs PBO+RUX.
— Pemmaraju et al., Blood 2026 — Conclusions
Because the TSS secondary endpoint was not met, formal testing of remaining secondary endpoints was not performed; subsequent P values are nominal.
Secondary and Exploratory Outcomes
| Outcome | NAV+RUX | PBO+RUX | Statistic |
|---|---|---|---|
| Mean TSS change, week 24 | −10.2 | −11.6 | P = .2852 |
| TSS reduction ≥10 points (any time) | 37.6% (95% CI, 29.1-46.7) | 43.3% (95% CI, 34.6-52.4) | Exploratory |
| Median overall survival | NE (95% CI, 41.4 to NE) | 48.5 mo (95% CI, 48.5 to NE) | HR 1.25 (0.75-2.08); P = 0.3867 |
| Median PFS (exploratory) | 31.9 mo (25.5-35.7) | 30.2 mo (22.9-31.7) | HR 0.755 (0.496-1.147); P = 0.1822 |
| Anemia response | 28.8% (18.3-41.3), n = 66 | 30.6% (20.2-42.5), n = 72 | Nominal P = .8524 |
| Bone marrow fibrosis grade reduction | 57% | 49% | Descriptive |
| ≥20% driver-gene VAF reduction | 58.5% | 45.5% | Exploratory |
| ≥50% driver-gene VAF reduction | 24.6% | 23.1% | Exploratory |
Subgroup Analyses
In a prespecified exploratory analysis, the authors report that across nearly all subgroups SVR35W24 favored NAV+RUX; subgroup-level numbers are shown in a figure and are not reproduced in the text. Statistics stated explicitly for the HMR subgroup are below.
| Subgroup analysis | Result | Nature |
|---|---|---|
| HMR: median SVR35 duration | NAV+RUX not estimable (95% CI, 27.6 to NE); PBO+RUX 19.4 months (95% CI, 13.6 to NE); HR 0.594 (95% CI, 0.246-1.437) | Post hoc; the authors state durations between arms were not different |
| HMR: progression or death | HR 0.450 (95% CI, 0.260-0.779) | Post hoc |
| SVR35 achievers at week 24: progression or death | HR 0.542 (95% CI, 0.284-1.034) | Post hoc |
Safety Profile
Any-grade adverse events occurred in 100% of NAV+RUX patients (n = 124) and 97.6% of PBO+RUX patients (n = 125); grade 3/4 events occurred in 90.3% and 79.2%. Cytopenias were more frequent and began earlier with the combination.
| Adverse event | NAV+RUX any / grade 3-4 | PBO+RUX any / grade 3-4 |
|---|---|---|
| Thrombocytopenia | 91.9% / 54.0% | 54.4% / 19.2% |
| Anemia | 67.7% / 49.2% | 53.6% / 44.0% |
| Neutropenia | 46.0% / 40.3% | 12.8% / 8.8% |
| Diarrhea | 41.9% / 5.7% | 16.8% / 0% |
| Serious TEAEs | 33.8% (42) | 44.8% (56) |
| TEAEs leading to NAV/PBO interruption | 78.2% | 43.2% |
| TEAEs leading to NAV/PBO dose reduction | 83.9% | 37.6% |
| Deaths | 35 (28.2%) | 28 (22.4%) |
Median relative ruxolitinib dose intensity was 49.4% in the combination arm and 73.6% in the placebo arm. None of the dose reductions or interruptions for NAV/PBO were due to bleeding. Of the 35 deaths in the NAV+RUX arm, 10 (8.1%) occurred within 30 days of the last navitoclax dose; later causes of death were not reported.
Interpretation and Broader Context
The authors conclude that cytopenias were generally manageable and reversible with dose adjustments, and that the lack of symptom benefit may reflect a ceiling effect of ruxolitinib, lower ruxolitinib dosing in the combination arm, and a relatively low baseline TSS. They also argue that SVR and TSS may be insufficient endpoints for combination studies. They note that the sponsor decided to discontinue the navitoclax clinical development program, and that all active patients have discontinued the study. The authors state that the efficacy measures favoring NAV+RUX did not extend PFS or translate into improved OS in the overall population within the current follow-up, while noting that some subgroups may derive particular benefit.
Limitations
Limitations noted in the source: the TSS secondary endpoint was not met, so formal testing of the remaining secondary endpoints was not performed and their P values are nominal. The HMR and SVR35-achiever analyses are post hoc, and subgroup-level SVR35W24 results are shown only in a figure. Progression-free survival and the driver-gene analyses are exploratory. The authors suggest the absence of symptom benefit may reflect a ceiling effect of ruxolitinib, lower ruxolitinib dose intensity in the combination arm (49.4% versus 73.6%), and a relatively low baseline TSS. Overall survival and PFS did not favor the combination within the current follow-up, later causes of death were not reported, and the sponsor discontinued the navitoclax program, with all active patients having discontinued the study. Number of sites and countries is not specified in the source.