SVR With Pacritinib Is Linked With OS Benefit in Myelofibrosis With Thrombocytopenia

Image Credit: © alionaprof
– stock.adobe.com

The achievement of spleen volume response (SVR) of at least 10% (SVR10) at week 12 with pacritinib (Vonjo) was positively associated with overall survival (OS) in patients with myelofibrosis and thrombocytopenia, although more stringent SVR thresholds were not significantly associated with OS, according to findings from a post hoc landmark analysis of the phase 3 PERSIST-2 trial (NCT02055781).1

The investigators assessed OS among SVR responders vs nonresponders using 4 different SVR thresholds: at least 35% (SVR35), at least 20% (SVR20), SVR10, and greater than 0% (SVR0). The data, which were published in the European Journal of Haematology, showed that the greatest separation of OS curves between responders and nonresponders was observed among patients who received pacritinib (n = 89) at 200 mg twice daily and achieved SVR10 vs those who did not respond at that threshold (HR, 0.00; 95% CI, 0.00-0.14; P < .01). No deaths were observed among the responders, whereas 5 nonresponders died.

“We found that a significant association was not observed with more stringent SVR or total symptom score [TSS] thresholds, likely because the number of responders was lower, and thus a larger number of observations would be needed to detect statistical significance at such thresholds,” lead study author Helen Ajufo, MD, MS, of the Department of Medicine, Leukemia Service, at Memorial Sloan Kettering Cancer Center, in New York, New York, and coinvestigators explained in the paper.

SVR0 and SVR20 also conferred improved OS, although these more stringent response thresholds were less prognostic of the pacritinib survival benefit.

In contrast, in the best available therapy (BAT) arm (n = 84), which included patients who received ruxolitinib (Jakafi; n = 39), no SVR threshold at week 12 was associated with improved OS. Only 33% of patients in this arm achieved SVR10, 82% of whom had received ruxolitinib (78% at ≤ 10 mg twice daily; 43% at ≤ 5 mg twice daily). The incidence of death was comparable between patients treated with BAT who achieved SVR10 (11%) and those who did not respond at that threshold (14%; HR, 0.63; 95% CI, 0.17-2.37; P = .49).

Notably, among patients with myelofibrosis and platelet counts at or below 100 x 109/L, SVR with pacritinib, but not BAT, was associated with a significant OS benefit.

PERSIST-2 Background and Design

The study authors noted that although retrospective analyses have shown an association between SVR and an OS benefit in patients with myelofibrosis and platelet counts greater than 100 x 109/L, this association had not previously been demonstrated in patients with myelofibrosis and platelet counts lower than 100 x 109/L, which denotes thrombocytopenia.

The randomized, controlled, multicenter PERSIST-2 trial evaluated the efficacy and safety of pacritinib vs BAT in patients with myelofibrosis and thrombocytopenia. A total of 311 patients were randomly assigned 1:1:1 to receive pacritinib at 400 mg once daily (n = 104), pacritinib at 200 mg twice daily (n = 107), or BAT (n = 100). Notably, the 400-mg once daily dose is no longer in development.

Previously, PERSIST-2 showed that pacritinib improved SVR vs BAT (including ruxolitinib) among patients with myelofibrosis and platelet counts at or below 100 x 109/L. Patients who received pacritinib at 200 mg twice daily (the FDA-approved dose) achieved a significantly higher rate of SVR35 vs those who received BAT, at 22% vs 3% (P = .001).2 Treatment with pacritinib at the FDA-approved dose was also associated with a trend toward improved OS vs treatment with BAT (HR, 0.68; 95% CI, 0.30-1.53). However, this analysis was limited by crossover from the BAT arm to the pacritinib arm in 51% of patients from the BAT arm at or after week 24, as well as the placement of a full clinical hold on the pacritinib clinical development program during study enrollment, which led to incomplete trial accrual and shortened survival follow-up.1,2

The present landmark survival analysis included patients from PERSIST-2 who were alive and still on the study at the start of the 12-week SVR assessment window and received pacritinib at 200 mg twice daily or BAT.1 In total, 83% of patients from the pacritinib 200 mg arm and 84% of patients from the BAT arm were included in this analysis. The median follow-up from the landmark analysis to death or last follow-up was 246 days in the pacritinib arm and 226 days in the BAT arm.

Additional Response Data

Most patients in the pacritinib arm of this analysis achieved SVR10 (73%). Many baseline characteristics were similar among those responders vs nonresponders, including median age (66 years [IQR, 58-69] vs 67 years [IQR, 60.5-70.5]), high-risk status per the Dynamic International Prognostic Scoring System (18% vs 46%), median platelet counts (58 x 109/L [IQR, 39-90] vs 67 x 109/L [IQR, 30-95]), median hemoglobin levels (9.7 g/dL [IQR, 8.4-11.1] vs 9.3 g/dL [IQR, 7.7-11.2]), and median spleen volumes (2573 cm3 [IQR, 1665-3419] vs 2094.5 cm3 [IQR, 1521-3040]). The median relative dose intensity of pacritinib through week 12 was 100% among both SVR10 responders and nonresponders.

The hazard ratio for death when stratified by SVR10 was not affected in either the pacritinib or BAT arms after adjusting for baseline age, high-risk disease status, platelet count, or spleen volume. Investigators saw similar SVR10 findings among patients who had baseline platelet counts less than 50 x 109/L.

When the landmark OS analysis was repeated without censoring at the clinical hold, SVR10 with pacritinib was still associated with a survival benefit (HR, 0.22; 95% CI, 0.08-0.65; P < .01), but responses diminished when treatment was abruptly discontinued for all patients irrespective of response. The investigators noted similar trends with the SVR20 threshold. These findings indicate that the OS benefit associated with spleen response to ruxolitinib depends on pacritinib continuation.

“Our observation that the OS curves converge after withdrawal of pacritinib in the context of the clinical hold also supports that the survival benefit seen is attributable to the drug itself and argues against any influence of baseline disease and patient characteristics,” the authors noted. “This is a novel and clinically relevant finding that has not been demonstrated for other JAK inhibitors previously, which used continuous dosing of JAK inhibitors.”

Notably, a spleen length reduction of at least 20% was associated with a subsequent OS benefit (HR, 0.14; 95% CI, 0.02-1.26; P = .04).

Relationship Between TSS and OS

Among all the tested response thresholds based on a reduction in TSS at week 12, investigators observed the greatest separation of the OS curves between pacritinib responders and nonresponders with a TSS of at least 10%. Among the 59 responders, 1 patient died; 4 deaths occurred among the 30 nonresponders (HR, 0.12; 95% CI, 0.01-1.04; P = .02). Statistically significant OS curve separations with pacritinib were not observed at higher TSS response thresholds. In the BAT arm, TSS response was not prognostic for OS at any threshold.

Investigators also conducted a landmark analysis of OS based on a 20% or greater reduction in physical function scores (sum of “tiredness” and “inactivity”), cytokine-related symptom scores (sum of “itching,” “night sweats,” and “bone pain”), and spleen-related symptom scores (sum of “abdominal discomfort,” “early satiety,” and “left rib pain”). Only cytokine-related symptom response was significantly associated with OS; 0 deaths occurred among the 35 responders vs 2 among the 15 nonresponders (HR, 0.00; 95% CI, 0.0-0.71; P = .03). However, deaths were also less frequent among physical function responders and spleen-related symptom responders vs nonresponders.

Relationship Between Transfusion Response and OS

A total of 36 patients in the pacritinib arm required red blood cell transfusions at baseline. Among these, 7 became transfusion independent over at least 12 weeks through the time of the landmark analysis. No transfusion-independent responders died; 2 deaths occurred among the transfusion-independent nonresponders, although the association between transfusion-independent nonresponse and death was not statistically significant (P = .45). In the BAT arm, 2 of 38 responders became transfusion independent, which was too small of a population to evaluate associations with OS.

“Although SVR35 is the most commonly used threshold for spleen response and has been accepted by the FDA as an endpoint for drug approval, it is unclear whether this is the optimal SVR cut-off as a surrogate for survival. In fact, our analysis demonstrated that the separation of the survival curves for patients treated with pacritinib was greater for lower SVR thresholds (≥ 10% and ≥ 20%) again supporting that OS benefits can be achieved already with a more modest SVR and are mediated by the drug-specific effect,” the authors concluded.

References

1. Ajufo H, Bewersdorf JP, Harrison C, et al. Pacritinib response is associated with overall survival in myelofibrosis: PERSIST-2 landmark analysis of survival. Eur J Haematol. 2025;114(2):238-247. doi:10.1111/ejh.14321
2. Mascarenhas J, Hoffman R, Talpaz M, et al. Pacritinib vs best available therapy, including ruxolitinib, in patients with myelofibrosis: a randomized clinical trial. JAMA Oncol. 2018;4(5):652-659. doi:10.1001/jamaoncol.2017.5818