Systematic Review and Meta-Analysis of Spleen Volume Reduction and Overall Survival in Myelofibrosis Patients Treated with JAK-Inhibitors

Background:

Myelofibrosis (MF) is a myeloproliferative neoplasm marked by the constitutive activation of the JAK/STAT pathway with clinical manifestations including splenomegaly and a shortened survival. Since the first successful trial of ruxolitinib (RUX), JAK-inhibitors have become the cornerstone of MF treatment. Following RUX, 3 additional JAK-inhibitors have been approved, extending the benefits of this treatment to patients (pts) with cytopenias or those intolerant/refractory to RUX. The primary endpoint of most JAK inhibitor trials in MF is spleen volume reduction (SVR). However, ultimately overall survival (OS) is the outcome of paramount importance. Surrogate measures are urgently needed to identify promising therapeutic approaches before survival outcomes are available. Therefore, we performed a systemic review and meta-analysis to clarify the role of SVR as a surrogate endpoint for OS in patients receiving JAK-inhibitors.

Methods:

A literature search was conducted on 01/31/2024 across the following databases: EMBASE, MEDLINE, CINAHL, Cochrane CENTRAL. The search terms included keywords and controlled vocabulary for myelofibrosis and JAK-inhibitors (ruxolitinib, fedratinib, momelotinib, and pacritinib). Two independent authors screened the articles for inclusion criteria of: prospective or retrospective studies of MF pts treated with a JAK-inhibitor; studies comparing SVR among pts treated with the same JAK-inhibitor and its association with OS; availability of full-text manuscripts in English. Hazard ratios (HR) were extracted along with confidence intervals (CI) and p values when available. Heterogeneity was assessed using Paule-Mandel estimator T² and Higgins' I² statistic, both with corresponding CIs. Random-effects meta-analysis was conducted. Publication bias was visualized using funnel plot and tested using Begg and Mazumdar's rank correlation test and Egger's linearity test. The study evaluated SVR% response cutoffs, JAK-inhibitor types, SVR measurement methods, and whether HRs were adjusted. Random-effects meta-regression was conducted on these variables, followed by subgroup analyses.

Results:

The search identified 5576 studies, with 1685 duplicates removed. We reviewed 3891 titles and abstracts, of which 3835 were excluded. Full-text evaluation was conducted on 56 studies, with 48 studies excluded, resulting in 8 studies included in our analysis. This included prospective data from COMFORT-1/2, SIMPLIFY-1/2, PERSIST-2, and 5 retrospective studies.

The publications ranged from 2012 to 2023. Five studies included pts on RUX; 1 with momelotinib, 1 with pacritinib, and 1 with multiple JAK-inhibitors. The SVR% cutoff ranged from 10% to 50%. Five studies used imaging as a measure to confirm spleen response while 3 used palpation. Five studies included adjusted HR, while 3 reported unadjusted HR.

In our primary analysis, MF patients on JAK-inhibitors who reached a specified SVR% cutoff had a lower risk of death (HR 0.45, 95% CI: 0.30-0.69, p < 0.001) compared to who did not reach the SVR cutoff.

Data analysis revealed high heterogeneity T²=0.21 (95% CI: 0.01-1.37); I² = 60% (95% CI: 12%-82%). Visual examination of the funnel plot suggested an impression of symmetry. This was supported by Begg and Mazumdar's test (p = 0.40) and Egger's test (p = 0.22), both of which provided evidence against the presence of publication bias.

Higher spleen response cutoffs in meta-regression were associated with a significantly lower HR (improved survival), with no impact from JAK-inhibitor type, method of SVR measure, or analysis adjusted observed.

Subgroup analysis comprised studies with SVR% response cutoffs of greater than 50% (n=2), 35% (n=4), and (25% or 30%) (n=2), yielding HR of 0.20 (95% CI: 0.11-0.38), 0.62 (95% CI: 0.35-1.12), and 0.51 (95% CI: 0.36-0.72), respectively.

Conclusions:

In this systematic review and meta-analysis, we found that JAK-inhibitor-treated MF patients achieving an SVR% endpoint had a 55% lower mortality risk (p<0.001) compared to MF patient on the same JAK-inhibitor who did not reach the SVR% endpoint. Further research should identify the optimal SVR% cutoff, additional survival markers, and baseline predictors of JAK-inhibitor response. Our findings suggest SVR as a potential surrogate endpoint, but further validation is needed.

Mascarenhas:AbbVie: Consultancy, Research Funding; PharmaEssentia: Consultancy, Research Funding; Disc: Consultancy; Keros: Consultancy; Kartos: Consultancy, Research Funding; Geron: Consultancy, Research Funding; Celgene: Consultancy, Other: Travel Support, Speakers Bureau; Novartis: Consultancy, Other: Travel Support , Research Funding, Speakers Bureau; GSK: Consultancy; Ajax: Research Funding; Blueprint Medicines: Consultancy; Sumitomo: Consultancy; Karyopharm: Consultancy; Pfizer: Research Funding; Merck: Consultancy; MorphoSys: Consultancy; Icahn School of Medicine at Mount Sinai: Current Employment; Bristol Myers Squibb: Research Funding; Roche: Consultancy; NS Pharma: Research Funding; CTI BioPharma/SOBI: Consultancy, Research Funding; Incyte Corporation: Consultancy, Speakers Bureau; Ariad: Speakers Bureau; Astellas: Research Funding. Tremblay:Sobi: Consultancy, Research Funding; Sumitomo: Research Funding; Cogent Biosciences: Consultancy, Research Funding; Gilead: Research Funding; Novartis: Consultancy; Abbvie: Consultancy; Pharmaessentia: Consultancy; Sierra Oncology: Consultancy; GSK: Consultancy.