TRANSFORM-1 was an international, randomized, double-blind phase III trial of navitoclax, a BCL-2/BCL-XL inhibitor, plus ruxolitinib vs ruxolitinib plus placebo in 252 patients who were JAK inhibitor naive and had intermediate 2‒ or higher-risk MF by Dynamic International Prognostic Scoring System (DIPSS) Plus and measurable splenomegaly. In addition, patients were required to have evidence of symptoms related to MF (NCT04472598).²

Patients were randomly assigned to receive a starting dose of navitoclax at either 100 mg or 200 mg daily depending on platelet count. Patients with a platelet count >150 x 10⁹/L started at 200 mg daily; all other patients started at 100 mg daily and were escalated to 200 mg after 7 days if platelets stayed above 75,000. Ruxolitinib was administered at a dose of either 15 mg or 20 mg twice daily depending on platelet count. Patients with platelets >200 x 10⁹/L received 20 mg, and those with platelets of 100-200 x 10⁹/L received 15 mg.

The primary endpoint of the study was SVR35 at Week 24, which was assessed for superiority using MRI or CT imaging by International Working Group (IWG) criteria. The secondary endpoints included change in TSS from baseline to Week 24, anemia response using IWG criteria, and safety. The trial had a median follow-up of 14.9 months.

TRANSFORM-1 follows in the footsteps of the phase II REFINE study, which demonstrated efficacy with the addition of navitoclax to ruxolitinib in patients with MF and progressive disease or suboptimal response to single-agent ruxolitinib and is based on the biological rationale that BCL-2 and BCL-XL are upregulated in patients who are receiving JAK inhibitor therapy.³

Baseline characteristics are displayed here, with a few key points to be made. Median spleen volume was essentially matched in both arms at approximately 1400-1600 cm³, with some dramatic ranges.² Putting this into context, in the phase III COMFORT trials of ruxolitinib vs best available therapy or placebo, the baseline spleen volumes of patients were approximately 2300-2600 cm³.^4,5 Median TSS in TRANSFORM-1 was 21 in the navitoclax arm and 24 in the placebo arm, so these were symptomatic patients; only a small percentage of patients were transfusion dependent.

The majority of patients in both arms were intermediate-2 by DIPSS Plus, with a smaller percentage who were higher or intermediate-1 risk. Patients with high-risk mutations comprised approximately one half of the patient population in both arms.

Fewer than 20% of patients discontinued treatment for any reason. The most common reason for discontinuation of navitoclax was AEs (14%).²

In the primary endpoint analysis, there was a clear superiority of navitoclax plus ruxolitinib vs placebo plus ruxolitinib in rate of SVR35 at Week 24 (63.2% vs 31.5%; a difference of 31% in favor of the combination; P <.0001).² Putting this finding into some historical context, an SVR35 of 31% with ruxolitinib alone is really very much akin to what was observed with the COMFORT studies in aggregate, so that does give us a measure of confidence in these findings.^4,5 As such, the TRANSFORM-1 study met its primary endpoint, with a significantly higher rate of SVR35 at Week 24 with navitoclax plus ruxolitinib vs placebo plus ruxolitinib.² Spleen responders had higher average doses of navitoclax and ruxolitinib vs nonresponders, so dosing was a factor.

In terms of SVR35 rate at any time during the study, there was again a statistical superiority with the combination vs placebo plus ruxolitinib. Median time to SVR35 was 12 weeks in both arms, and the percentage of patients who lost their SVR35 was numerically higher with ruxolitinib alone.

Of note, there was no statistical difference in TSS between the 2 arms from baseline to Week 24.²

In terms of safety, grade ≥3 AEs were observed in 85% of patients who received combination therapy and 70% of patients who received placebo plus ruxolitinib.²

The most common reason for dose reductions/interruptions was thrombocytopenia, and no patients were observed for clinically significant bleeding, suggesting that the dose reductions/interruptions for thrombocytopenia were due to a numerical reduction in platelet counts.

In terms of AEs, more patients who received combination therapy had grade 3 thrombocytopenia compared with patients who received placebo plus ruxolitinib.² Instances of neutropenia were significantly increased with combination therapy compared with placebo plus ruxolitinib. Differences in other events were much smaller. 

The most common AEs with combination therapy were thrombocytopenia, anemia, neutropenia, and diarrhea, and the most common serious AEs were COVID-19 pneumonia and pneumonia; this study was ongoing during the height of the pandemic.

Data from MANIFEST-2 also were presented at ASH 2023. The MANIFEST-2 study was a randomized, double-blind, placebo-controlled phase III trial of pelabresib (a novel oral BET inhibitor) plus ruxolitinib vs placebo plus ruxolitinib for patients with splenomegaly and symptomatic, JAK inhibitor‒naive, intermediate-1 or greater‒risk MF.⁶ Data from the phase II MANIFEST study showed that adding pelabresib to ruxolitinib resulted in spleen size reductions and anemia improvement for many patients with JAK inhibitor‒naive MF, providing clinical support for MANIFEST-2.⁷

In MANIFEST-2, patients on one arm received pelabresib 125 mg daily for 14 days in combination with ruxolitinib at either a 10-mg or 15-mg starting dose, with a mandatory increase of 5 mg after cycle 1 if the patients met platelet count criteria. Patients on the second arm were given placebo for 14 days with continuous ruxolitinib at the same dosage as arm 1. The primary endpoint of this study was SVR35 at Week 24, and key secondary endpoints were change in TSS and safety.

In contrast to TRANSFORM-1, the majority of patients on this trial had intermediate-1‒risk MF (59.8% and 58.8% on the combination vs placebo plus ruxolitinib arms).⁶

Median baseline hemoglobin was relatively preserved at approximately 11 g/dL in both arms, and median platelet counts were both >280,000. A very small proportion of patients required red blood cell transfusions, and median baseline spleen volume was similar to that observed in patients on TRANSFORM-1 at approximately 1300 cm³. Median TSS was 26.6 in the experimental arm and 24.7 in the placebo arm, indicating highly symptomatic MF.

SVR35 was achieved by 65.9% of patients receiving the combination vs 35.2% of patients receiving placebo plus ruxolitinib, a statistically significant difference in favor of the combination by approximately 30% (P <.001).⁶

The SVR35 response at Week 24 was consistently higher with combination therapy across all key subgroups, including DIPSS score, type of MF, baseline spleen volume, platelet count, and hemoglobin level.

The mean change in TSS from baseline to Week 24 numerically favored the combination vs placebo plus ruxolitinib (-15.99 vs -14.05). Regarding TSS50, there was a numerically greater decrease in the proportion of patients achieving this endpoint with combination therapy vs placebo plus ruxolitinib, but this did not reach statistical significance.

One of the analyses assessed the proportion of patients who achieved both an SVR35 and TSS50. This was achieved by 40% of patients receiving combination therapy vs 18% of patients receiving placebo plus ruxolitinib.⁶

Additional outcomes also were reported. A mean increase in hemoglobin response of 1.5 g/dL was observed in approximately 9% of patients who received the combination vs 5% of patients who received placebo plus ruxolitinib.⁶

The number of patients needing red blood cell transfusion at screening was higher in the combination therapy arm, but at Week 24, approximately 30% of patients who received combination therapy required a red blood cell transfusion vs 41% of patients who received placebo plus ruxolitinib.

Finally, looking at reticulin fibrosis changes, more patients on the combination therapy arm achieved a reduction of ≥1 grade level in bone marrow fibrosis vs patients on the control arm. Unchanged reticulin fibrosis rates were approximately the same between arms, and worsening by ≥1 grade level was seen in a higher proportion of patients who received placebo plus ruxolitinib. There was also a greater reduction in inflammatory cytokines with combination therapy vs placebo plus ruxolitinib, so some key biological changes were occurring with pelabresib therapy.

There were numerically more grade ≥3 AEs with placebo plus ruxolitinib vs pelabresib plus ruxolitinib.⁶ Discontinuation rates were slightly higher with the combination, as were dose interruptions. However, the overall safety profile with the combination generally was consistent with that observed with prior ruxolitinib trials.

Grade 3 anemia and thrombocytopenia events were seen more frequently with the placebo arm vs the combination arm.⁶ By contrast, there was a slightly higher rate of diarrhea and dysgeusia with combination therapy; these effects were dose dependent with pelabresib and were attenuated by reducing the dose. Only 1 patient discontinued because of dysgeusia, and essentially all the other AEs were similar across the board.⁶

The last MF trial presented at ASH 2023 that we will discuss is a study of bomedemstat (an LSD1 inhibitor) plus ruxolitinib for patients with MF.⁸ This was an open-label phase II study with 2 cohorts. Patients on cohort A were refractory to or relapsed on ruxolitinib. On that cohort, patients were treated with bomedemstat plus ruxolitinib, the ruxolitinib dose being the prior stable dose that the patient was receiving—essentially just adding bomedemstat to a fixed dose of ruxolitinib.

Cohort B included patients with JAK inhibitor‒naive MF with measurable disease burden, defined as MF-related symptoms with a >10 score on the Myelofibrosis Symptom Assessment Form (MFSAF) or ≥2 symptoms with scores of 3 and a spleen that was palpable 5 cm below the left costal margin. The dosing of bomedemstat here was the same as in cohort A, but ruxolitinib was used at a fixed dose of 10 mg twice daily.

The primary endpoint of the study was safety. Secondary endpoints were spleen length reduction, SVR35 at Week 24, and symptom burden score by MFSAF.

At baseline, median spleen size by palpation was 8 cm below the left costal margin, and median spleen volume was 922 cm³. These were smaller spleen volumes than those observed in the prior trials we have discussed.⁸

The majority of patients experienced grade 1/2 AEs, with 17% experiencing grade 3/4 events. Grade 3/4 thrombocytopenia affected approximately 17% of patients, and grade 3/4 anemia was reported in approximately 20% of patients.⁸

Regarding nonhematologic AEs, grade 3/4 infections and heart failure were observed but occurred uncommonly. Grade 1/2 diarrhea, infections, liver function abnormalities, and renal function abnormalities were observed.

Bomedemstat dose reductions/interruptions did not necessarily increase over time and were seen throughout cycles 1-6.

In total, 20% of patients who previously had been exposed to ruxolitinib and 33% of patients who were ruxolitinib naive experienced TSS50 by Week 24.⁸ SVR35 also was higher among patients who were ruxolitinib naive. The number of patients available for analysis here was relatively small.

Improved or stable hemoglobin was achieved by 56% of patients with prior ruxolitinib and 33% of patients who were ruxolitinib naive. Changes in the JAK2 V617F allele burden were stable or slightly reduced in 79% of patients at Week 24 and were reduced by 50% in 16% of patients.

We will start with a phase IIb trial of the LSD1 inhibitor bomedemstat in patients with ET.⁹ This was a single-arm study that enrolled 73 patients with ET who required cytoreduction and had an inadequate response or were intolerant to 1 or more standard therapies. Patients were required to have platelet counts >450,000, hemoglobin >100 g/L, and no splenectomy. Participants were treated with bomedemstat titrated to their platelet count and followed.

The primary endpoints were safety, tolerability, and response (platelets ≤400 x 10⁹/L with no new thromboembolic events). Exploratory endpoints included hematologic effects, durable platelet reduction for >12 weeks, durable white blood cell count reduction, patient symptom reduction, bleeding and clotting events, mutant allele reduction, and transformation to acute myeloid leukemia.

At baseline, the mean white blood cell count was <10 x 10⁹/L, and mean hemoglobin was >100 g/L; mean platelets in this population were 808,000, which is perhaps the most important parameter.⁹ There was a history of bleeding in 33% of patients and history of thrombotic events in 18% of patients.

Most patients had been previously treated with hydroxyurea, with anagrelide being the next most common previous therapy.

At Week 24, 77% of patients had responded and maintained a platelet count <400,000 with no new thromboembolic events.⁹

Thrombotic events occurred in a small number of patients (3%). The number of hemorrhagic events was higher, occurring in 21% of patients with various event types.⁹

There was no change in symptoms from baseline at Week 24 according to the Myeloproliferative Neoplasm Symptom Assessment Form (MPN-SAF), but when using another symptom assessment tool, the Patient Global Impression of Change (PGIC), there was a response in approximately 72% of patients at Week 24.⁹ The driver mutation variant allele frequency had decreased in 85% of patients and increased in 15% of patients.

In total, 47% of patients experienced grade 3/4 AEs, 37% experienced serious AEs, 40% experienced AEs that led to temporary treatment interruption, and 15% experienced AEs that led to treatment discontinuation. The most common grade 3/4 event was thrombocytopenia. Regarding any-grade events, dysgeusia was the most common AE, followed by arthralgia, constipation, and thrombocytopenia.

Ruxolitinib is approved for patients with PV who are intolerant or resistant to hydroxyurea based on the RESPONSE trial.¹⁰ Then, the RESPONSE 2 trial assessed ruxolitinib in patients with PV who were intolerant or resistant to hydroxyurea who did not have splenomegaly.¹¹ Most recently, the MAJIC trial assessed ruxolitinib in patients who were intolerant or resistant to hydroxyurea, demonstrating the first event-free survival advantage with ruxolitinib vs best available therapy. This included thrombotic events and transformation events, so the data that emerged from this trial were quite important.¹²

RuxoBEAT was a phase IIb study of ruxolitinib vs best available therapy as first-line treatment in high-risk PV.¹³ This study enrolled adult patients with previously untreated high-risk PV or ET with an indication for cytoreductive therapy due to progressive myeloproliferation. The plan was to enroll a cohort of 190 patients, but the prespecified interim analysis was to be conducted once enrollment reached 78 participants.

The primary endpoint of the study was the rate of complete hematologic response—normalization of blood counts—at 6 months.¹⁴ Secondary endpoints included overall response, which was the rate of complete plus partial response; spleen reduction; and patient-reported outcomes.¹⁴

Participants were randomly assigned to receive ruxolitinib 10 mg twice daily or best available therapy. Crossover was allowed to the ruxolitinib arm for intolerance or resistance to best available therapy after 6 months.¹³

There are a few key baseline characteristics to point out. First, a significant number of patients had thrombocytosis and leukocytosis. Also, median hematocrits were 45.6% with ruxolitinib and 43.8% with best available therapy. Furthermore, splenomegaly was seen in approximately one half of patients. The median duration of treatment was similar between treatment arms.¹³

Complete response was achieved in a small proportion of patients on both arms with no significant difference.¹³ The rate of complete response plus partial response was achieved in a numerically higher proportion of patients who received ruxolitinib vs best available therapy, but this did not reach statistical significance. 

There was no difference between arms in white blood cell count or hematocrit percentage at 6 months, but there was a significant trend toward a moderately higher platelet count among patients receiving ruxolitinib vs best available therapy. The number of patients with phlebotomy requirements was comparable between arms. Spleen size reduction, fatigue scores, and pruritus scores all favored ruxolitinib vs best available therapy.

From baseline to Month 6, there was a statistically significant decrease in platelet count, white blood cell count, and median hematocrit percentage with both ruxolitinib and best available therapy.¹³ Regarding spleen size, there was a reduction with ruxolitinib but not with best available therapy.

There was a clear difference between treatment arms in symptom burden. According to patient-reported outcomes, early satiety, pruritus, and quality of life were improved with ruxolitinib; this was not observed with best available therapy.

In terms of safety, more any-grade gastrointestinal events were reported with best available therapy, but this was the only statistical difference between the 2 treatment arms.¹³

Similarly, with grade ≥3 AEs, a greater proportion of patients receiving best available therapy experienced gastrointestinal events vs ruxolitinib.¹³