Real-world prevalence and outcome of elevated right ventricular systolic pressure in myelofibrosis Authors: Alexis Behne Sharma, Elicia Wang, Sruthi Selvakumar, Himara Koelmeyer, Sanath Shetty, Kinley Buckley, Qianxing Mo, Teagen Smith, Austin Chao, Carson Balen, Gowtam Mannam, Jacob Yudiono, Kyle Lien, Sami Khatib, Varshith Paduchuri, Najla Al Ali, Jinming Song, Taiga Nishihori, Jeffrey E Lancet, Mohammed Alomar, John L Cleveland, Andrew Kuykendall, Rami S Komrokji, Seongseok Yun, Dae Hyun Lee Published: 2026 ABSTRACT AbstractBackgroundMyelofibrosis (MF) is a subgroup of Philadelphia chromosome-negative myeloproliferative neoplasms that are associated with an increased risk of cardiovascular disease, including pulmonary hypertension. Here we report the real-world prevalence, risk factors, and clinical outcomes of eRVSP in MF patients.MethodsA retrospective, single-center cohort study was conducted on 208 patients with MF that were diagnosed between 2013 and 2023. Elevated right ventricular systolic pressure (eRVSP) was defined as RVSP > 35 mmHg. Major adverse cardiac events (MACE) were defined as new-onset congestive heart failure, coronary artery disease requiring intervention, cerebrovascular events, or cardiovascular death after MF diagnosis. Univariate and multivariable Cox proportional hazard ratio regression model with eRVSP status as time-dependent covariate was used to estimate overall survival.ResultsEchocardiograms were performed in 208 MF patients, with RVSP estimates available in 156 patients (75%). eRVSP was present in 61 patients (39.1%). Patients with eRVSP were older (65 vs. 62 years, p = 0.053) and had higher rates of baseline hypertension (65.6% vs. 43.2%, p = 0.01), atrial fibrillation (16.4% vs. 4.2%, p = 0.02). MACE after diagnosis of MF occurred in 43 (20.7%) patients and was more frequent in eRVSP patients (36.1% vs. 15.8% p = 0.007), and this was predominantly driven by increased rates of new-onset congestive heart failure (27.9% vs. 8.4%, p = 0.003). In multivariable analysis, the presence of time-dependent eRVSP was associated with reduced survival (aHR: 4.41; 95%CI:2.84–6.85) after adjustment for clinically relevant covariates.ConclusionseRVSP was prevalent among MF patients undergoing echocardiogram and this was associated with increased cardiovascular morbidity, particularly HF. Furthermore, eRVSP was associated with inferior survival and underscores the need for targeted screening and management in MF patients with cardiovascular risk factors and diseases. Prospective studies with baseline cardiovascular assessment and echocardiograms in all MF patients will establish true prevalence and may further elucidate the morbidity and mortality implication of eRVSP in MF patients.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40959-026-00474-2. Introduction Myelofibrosis (MF) is a subgroup of Philadelphia chromosome-negative myeloproliferative neoplasms (MPN) having an increased risk of cardiovascular disease (CVD), including coronary artery disease (CAD), congestive heart failure (CHF), and pulmonary hypertension (PH) [1, 2]. The prevalence and risk factors of PH in MF patients are poorly defined, although the prevalence of PH in MF is highest within MPN subtypes (ranging from 3.8–14%) [1–5]. Other than advanced age and pre-existing CVD as contributing factors, the etiology of PH in MF patients is largely unknown [4]. Hypotheses regarding PH pathophysiology have ranged from it being due to somatic mutation of hematopoietic stem cells (i.e. JAK2 mutation), pro-inflammatory cytokines, venous thromboembolism (VTE, including chronic thromboembolic pulmonary hypertension), or left-sided heart disease based on elevated LV filling pressure[4, 6–8]. There are conflicting results regarding the association of PH and overall survival in patients with MF[1, 2]. Echocardiographic assessment of right ventricular systolic pressure (RVSP) through the tricuspid regurgitation jet provides a suggestion of pulmonary hypertension, although not sufficient to confirm the diagnosis of PH without invasive hemodynamic measurements [9]. Therefore, in our study, we use the term “elevated RVSP” (eRVSP) as our study is based on echocardiographic measures only. Given the limited understanding of PH in large cohorts of MF, the aim of this study is to investigate the prevalence, clinicopathologic risk factors, and clinical outcomes of eRVSP in MF. METHODS This was a retrospective, single-center cohort study of patients with a diagnosis of MF with echocardiogram who presented to Moffitt Cancer Center from 2013 to 2023. This study was approved by the Institutional Review Board of Moffitt Cancer Center (MCC18864). We retrospectively reviewed baseline CV comorbidities at the time of diagnosis or when first seen at our institution. CVD at baseline was defined as heart failure, coronary artery disease or stroke. Major adverse cardiac event (MACE) was defined as a new development of CHF, CAD requiring intervention, or stroke following diagnosis of MF. New-onset heart failure was defined as a clinical syndrome characterized by the inability of the heart to pump blood at an adequate volume to meet tissue metabolic requirements, or the ability to do so only at an elevation in the filling pressure diagnosed with an echocardiogram with structural and/or functional cardiac abnormality (systolic dysfunction with LVEF < 49%, or LVEF > 50% with diastolic dysfunction of any grade), elevated BNP levels, and/or objective evidence of pulmonary/systemic congestion. Acute coronary syndrome was defined as myocardial infarction or unstable angina with evidence of coronary artery disease on invasive or non-invasive testing. Stroke or TIA was defined as an episode of neurological dysfunction caused by infarction based on pathological, imaging, or other objective evidence of ischemic injury in a defined vascular distribution. The timing of TTE in relation to diagnosis of MF differed (ranging from baseline to clinical indication after diagnosis of MF, such as development cardiovascular manifestation, acute myeloid leukemia or evaluation for allogeneic stem cell transplant [SCT]). eRVSP was defined as right ventricular systolic pressure (RVSP)>35mmHg, estimated through tricuspid regurgitant (TR) jet[10]. Baseline MF characteristics and outcomes were retrospectively reviewed from medical records as previously described[11, 12]. Overall survival was defined as the duration between the date of diagnosis of MF until death or censored at the date of last patient contact. To investigate how factors associated with performing echocardiogram may influence observe prevalence of eRVSP, we performed inverse probability weighting (IPW) analysis adjusting for factors associated with performing echocardiogram in MF patients. To investigate the overall survival in MF patients with eRVSP, we performed a time-dependent analysis with diagnosis of eRVSP as a time-dependent variable because echocardiogram was performed during different time point during the natural history of MF. Cox proportional hazard ratio (HR) regression models were used to estimate overall survival with multivariable analysis. Diagnosis of eRVSP and allogeneic SCT status was considered a time-dependent covariate. Competing risk analysis was performed to evaluate the association between eRVSP and MACE, treating non-cardiovascular death as a competing event. A p-value of < 0.05 was considered statistically significant. RESULTS Within the 208 MF patients who had an echocardiogram performed, 156 patients (75%) had an echocardiogram with calculable RVSP estimates (Tables 1 and Supplemental Table 1). There were 61 patients (39.1%) with eRVSP with RVSP > 35mmHg. Patients with eRVSP had a median RVSP of 45mmHg (Interquartile range: 40-52mmHg). eRVSP group was older (65 vs. 62 years in no eRVSP; p = 0.053). In an exploratory analysis, we compared eRVSP, normal RVSP and inadequate TR jet (unable to assess RVSP) group (Supplemental Table 2), which showed comparable demographic and outcome with normal RVSP. There were no differences in sex, MF subtype, the Dynamic International Prognostic Scoring System (DIPSS) plus classification, transfusion dependency, or platelet levels between the eRVSP and no eRVSP. eRVSP had significantly higher rates of baseline hypertension (65.6% vs. 43.2%; p = 0.01), atrial fibrillation (16.4% vs 4.2%; p = 0.021), compared to no eRVSP. eRVSP had numerically higher rates of baseline CHF (6.6% vs 1.1%; p = 0.15) and VTE (18.0% vs 9.5%; p = 0.189). eRVSP group had a numerically higher rate of baseline CVD (29.5% vs 16.8%, p = 0.095; Supplemental Table 3). Patients with eRVSP had higher rates of reduced LVEF (< 49%), compared to normal RVSP (15.0% vs. 1.1%, P = 0.003). Further, patients with eRVSP had higher rates of elevated LV filling pressure, estimated by E/e’ ≥14, compared to normal RVSP (55.3% vs. 20.8%; P = 0.001), both summarized in Table 1. We stratified based on different eRVSP threshold by categorizing the eRVSP into mild eRVSP (35-44mmHg) and moderate or higher eRVSP (≥45mmHg), summarized in Supplemental Table 4. Patients with moderate or higher eRVSP had an increased incidence of MACE (48.1%, P = 0.001), compared to normal RVSP (14.7%) and mild eRVSP (23.5%). Table 1Clinical characteristics and outcomes based on pulmonary hypertensionTotal with EchoNormal RVSPeRVSPP value(N = 208)(N = 95)(N = 61)Age at Diagnosis63.1 ± 10.661.9 ± 10.565.3 ± 10.70.053Sex (Male)121 (58.2%)54 (56.8%)34 (55.7%)1Type of myelofibrosis0.98 Primary MF124 (59.6%)53 (55.8%)35 (57.4%) Post-PV37 (17.8%)18 (18.9%)11 (18.0%) Post-ET47 (22.6%)24 (25.3%)15 (24.6%)DIPSS +0.192 014 (6.9%)8 ( 8.9%)4 ( 6.7%) 164 (31.7%)34 (37.8%)14 (23.3%) 287 (43.1%)31 (34.4%)30 (50.0%) 337 (18.3%)17 (18.9%)12 (20.0%)Bone Marrow Fibrosis0.07 01 (0.7%)1 ( 1.6%)0 ( 0.0%) 16 (4.3%)3 ( 4.8%)0 ( 0.0%) 244 (31.2%)14 (22.6%)17 (44.7%) 390 (63.8%)44 (71.0%)21 (55.3%)Transfusion Dependency49 (23.6%)18 (18.9%)17 (27.9%)0.268Platelet count313.3 ± 254.1325.3 ± 264.7310.7 ± 237.10.732JAK2 Inhibitor Use140 (67.3%)72 (75.8%)39 (63.9%)0.157Progression to AML37 (17.8%)14 (14.7%)16 (26.2%)0.117Allogeneic stem cell transplant97 (46.6%)56 (58.9%)14 (23.0%) < 0.001 CV Disease at Baseline50 (24.0%)16 (16.8%)18 (29.5%)0.095 • CAD39 (18.8%)15 (15.8%)12 (19.7%)0.683 • CHF6 (2.9%)1 ( 1.1%)4 ( 6.6%)0.15 • Stroke/TIA13 (6.2%)1 ( 1.1%)5 ( 8.2%)0.066Hypertension110 (52.9%)41 (43.2%)40 (65.6%) 0.01 Hyperlipidemia72 (34.6%)33 (34.7%)18 (29.5%)0.614Diabetes34 (16.3%)16 (16.8%)9 (14.8%)0.902Atrial fibrillation24 (11.5%)4 ( 4.2%)10 (16.4%) 0.021 Left Ventricular EF- LVEF ≥50%197 (95.2%)94 (95.2%)51 (85.0%)- LVEF 30–49%9 (4.3%)1 (1.1%)8 (13.3%)- LVEF < 29%1 (0.5%)0 (0%)1 (1.7%) 0.003 LV Filling PressureN = 146- E/e’ ≤14101 (69.2%)57 (79.2%)17 (44.7%)- E/e’ >1445 (30.8%)15 (20.8%)21 (55.3%) 0.001 Cardiovascular OutcomeMajor Adverse Cardiovascular Event (MACE)43 (20.7%)15 (15.8%)22 (36.1%) 0.007 New Onset CHF28 (13.5%)8 ( 8.4%)17 (27.9%) 0.003 Acute Coronary Syndrome7 (3.4%)5 ( 5.3%)2 ( 3.3%)0.851 CAD with revascularization5 (2.4%)3 ( 3.2%)1 ( 1.6%)0.947 Ischemic Stroke/ TIA4 (1.9%)1 ( 1.1%)2 ( 3.3%)0.696 CV Death4 (1.9%)2 ( 2.1%)1 ( 1.6%)1Data are presented as mean ± Standard deviation or number (percentage), depending on the type of data. P values less than 0.05 are bolded. Transthoracic echocardiograms (TTE) were done according to standard protocol of our institution. Right ventricular systolic pressure (RVSP) was estimated through tricuspid regurgitant (TR) jet. There were 52 patients (25%) who did not have sufficient TR jet to estimate RVSPAbbreviations: eRVSP elevated right ventricular systolic pressure, PV polycythemia vera, ET essential thrombocytopenia, DIPSS Dynamic International Prognostic Scoring System, CRP C-reactive protein, WBC white blood cell, AML Acute myeloid leukemia, TIA Transient ischemic attack, CAD Coronary artery disease, DVT/PE Deep vein thrombosis/ pulmonary embolism, CHF Congestive heart failureP values less than 0.05 are bolded Clinical characteristics and outcomes based on pulmonary hypertension Data are presented as mean ± Standard deviation or number (percentage), depending on the type of data. P values less than 0.05 are bolded. Transthoracic echocardiograms (TTE) were done according to standard protocol of our institution. Right ventricular systolic pressure (RVSP) was estimated through tricuspid regurgitant (TR) jet. There were 52 patients (25%) who did not have sufficient TR jet to estimate RVSP Abbreviations: eRVSP elevated right ventricular systolic pressure, PV polycythemia vera, ET essential thrombocytopenia, DIPSS Dynamic International Prognostic Scoring System, CRP C-reactive protein, WBC white blood cell, AML Acute myeloid leukemia, TIA Transient ischemic attack, CAD Coronary artery disease, DVT/PE Deep vein thrombosis/ pulmonary embolism, CHF Congestive heart failure P values less than 0.05 are bolded MACE after MF diagnosis occurred in 43 patients (20.7%), most commonly new-onset CHF (13.5%). eRVSP group had a higher incidence of MACE (36.1% vs. 15.8%; p = 0.007), predominantly driven by a higher incidence of new-onset CHF (27.9% vs. 8.4%; p = 0.003). In a multivariable Fine-Gray subdistribution hazards model that treats non-cardiovascular death as a competing event, MF patients with eRVSP had a higher risk of MACE (sHR 1.96; 95% CI: 1.02–3.76; Supplemental Table 5). Notably, within our cohort of all MF patients (N = 629), 208 patients (32.5%) underwent echocardiogram (Fig. 1A). When compared to MF patients who did not have echocardiogram, those who had echocardiogram were younger, had high-grade bone marrow fibrosis, higher JAK2 inhibitor use, and increased incidence of future allogeneic HSCT and progression to AML (Supplemental Table 6). Within those who had progression to AML, 37(58.7%) had echocardiogram performed. Of those who underwent allogeneic SCT, 97(91.5%) had an echocardiogram performed. IPW analysis of weighted prevalence of eRVSP was performed (Supplemental Fig. 1). After IPW to account for factors associated with performing echocardiogram, the adjusted prevalence of eRVSP increased to 48.4%. Patients who underwent allogenic SCT had lower weighted eRVSP prevalence compared to those without allogeneic SCT (19.7% vs. 53.0%). The Moffitt Cancer Center MF cohort and survival of myelofibrosis patients with eRVSP. A Flow chart of patients with myelofibrosis and echocardiogram data of the Moffitt Cancer Center cohort. B Time-dependent survival based on the presence of eRVSP in myelofibrosis. Simon-Makuch plot of overall survival in patients with myelofibrosis are shown. All patients start in the no eRVSP state (green line). When eRVSP is diagnosed, they transition to post-eRVSP state (orange line). Cumulative incidence of death is shown in Y-axis. The x-axis depicts time since myelofibrosis diagnosis in years The median overall survival of the total cohort with MF was 7.8 years. Notably, the presence of eRVSP as time dependent covariate was associated with increased risk of death in Simon-Makuch plots (Fig. 1B). In multivariable analysis, the presence of time-dependent eRVSP was associated with reduced survival (aHR: 4.41; 95%CI:2.84–6.85; Supplemental Table 7) after adjustment for clinically relevant covariates (Age, sex, DIPSS+, allogeneic SCT and baseline CVD). Finally, we investigated if the presence of clonal hematopoiesis-related somatic mutations was associated with eRVSP in MF. The proportion of driver mutations for MF (JAK2, CALR, or MPL) were not different in eRVSP (Supplemental Table 8). DISCUSSION In our MF cohort that underwent echocardiogram, the prevalence of eRVSP was relatively high at 39.1%. eRVSP was associated with systemic hypertension, atrial fibrillation, and older age [4]. Interestingly, factors with possible pathophysiologic links for eRVSP, such as JAK2 inhibitor use, transfusion dependency, or VTE were not associated with eRVSP. Overall, these findings indicate that the presence of risk factors for left heart disease is associated with eRVSP in patients with MF, and echocardiogram evaluation may be warranted in MF patients with cardiovascular risk factors. As only a subset of MF patients underwent echocardiogram, we hypothesized overestimation of eRVSP prevalence may occur. However, the weighted prevalence of eRVSP increased to 48.4%. This paradoxical finding is likely from selection process for echocardiogram. Patients planned for allogeneic SCT over-represent those undergoing echocardiogram (45%) and have lower eRVSP prevalence. SCT candidates are likely healthier and less likely to have baseline cardiovascular disease. Therefore, the IPW-adjustment appropriately down-weights this overrepresented population, revealing a higher prevalence of eRVSP in general MF population. Our real-world evidence suggests that routine baseline echocardiogram may be considered for all MF patients, particularly in those not considered for SCT who may be systematically not undergo echocardiogram despite potentially higher eRVSP risk. Clonal hematopoiesis of indeterminant potential (CHIP) is an age-related phenomenon and is associated various cardiovascular disease. In contrast to our initial hypothesis that the presence of CHIP-related mutation is associated with eRVSP, the presence of these mutations was not associated with eRVSP. Overall, these findings indicate there is no association between MF-driver or common somatic CHIP-related mutations with eRVSP, suggesting that these mutations alone are not sufficient to drive eRVSP development. The interplay between these mutations and other eRVSP mechanisms, such as left-sided heart disease or non-CHIP mediated pro-inflammation, requires further investigation [6]. However, given the limited number of patients with CHIP-related mutations, the study is underpowered to show association between somatic mutations and adverse cardiovascular events. The presence of eRVSP was associated with increased MACE, most notably new onset CHF. Pathophysiologic links between eRVSP and CHF are well established, where eRVSP increases right ventricular afterload that leads to right-sided dysfunction and, in some cases, left-sided heart failure. Thus, as eRVSP is a marker of underlying cardiopulmonary pathology, its identification maybe important for risk stratification and prognostication in MF. However, it is also possible that as eRVSP is a marker for diastolic dysfunction, it is possible that subclinical or evolving CHF led to eRVSP, rather than eRVSP being a driver of HF. Future studies with serial echocardiograms and longitudinal clinical outcomes data would address these questions. Further, eRVSP as time-dependent covariate was independently associated with inferior survival. It is unclear whether eRVSP directly or indirectly contributes to increased mortality. The presence of eRVSP may have led to clinical decision not to pursue potentially cardiotoxic anti-cancer therapy. Regardless, the strong association of eRVSP with increased MACE and mortality underscores the need for systematic screening and targeted management in MF patients with an increased risk of CVD. There are several limitations to our study. First, there is selection bias, as we only investigated those MF patients who had an echocardiogram, not the whole cohort with MF. We do not have cardiovascular outcomes in those without an echocardiogram, as our study aim was to investigate the role of eRVSP in outcomes of myelofibrosis patients. To address this, we performed IPW analysis for the weighted prevalence of eRVSP. However, this should be interpreted cautiously, as we cannot directly verify eRVSP in those without echocardiogram. Also, 25% of patients without calculable RVSP due to insufficient TR jet usually have a lower chance of significant eRVSP and may overestimate the prevalence of eRVSP. Second, we did not ascertain the clinical manifestations of eRVSP seen on the echocardiogram. However, even with possible “subclinical” eRVSP, they were still associated with other CVD, such as CHF, hypertension, atrial fibrillation, and stroke. Also, we have not stratified further with left-sided heart disease due to insufficient data. Further, we do not have the data on the reason for the echocardiogram, including whether these patients were in setting of acute illness like sepsis. However, significant amount of patients underwent echocardiogram as a workup for allogeneic stem cell transplant evaluation. Third, the timing of echocardiograms for patients varied in relation to MF. To overcome immortality bias, we considered eRVSP status as time-dependent variable which allows an understanding of survival after diagnosis of eRVSP. Further, we only investigated the first available echocardiogram after diagnosis of myelofibrosis. Therefore, we do not have data on serial changes in RVSP. We do acknowledge that the echocardiograms were performed at variable time points, including different clinical indications. CONCLUSIONS In conclusion, eRVSP was prevalent among MF patients undergoing echocardiogram and this was associated with increased cardiovascular morbidity, particularly HF. Furthermore, eRVSP was associated with inferior survival and underscores the need for targeted screening and management in MF patients with cardiovascular risk factors and diseases. Prospective studies with baseline cardiovascular assessment and echocardiograms in all MF patients will establish true prevalence and may further elucidate the morbidity and mortality implication of eRVSP in MF patients. Supplementary Information Supplementary Material 1. Acknowledgements This study was supported by grants from the Florida Department of Health Casey DeSantis Cancer Innovation Fund (ID# MOABE) and from the American Heart Association (24CDA1271837; 10.58275/AHA.24CDA1271837.pc.gr.193589) and Research Support Fund from University of South Florida to DHL. J.L.C. is supported by the Cortner-Couch Endowed Chair from the University of South Florida School of Medicine. These studies are also supported by Collaborative Data Services Core at Moffitt Cancer Center & Research Institute (NCI grant P30 CA076292) and by monies from the State of Florida to Moffitt Cancer Center. Abbreviations elevated right ventricular systolic pressure polycythemia vera essential thrombocytopenia Dynamic International Prognostic Scoring System C-reactive protein white blood cell Acute myeloid leukemia. TIA: Transient ischemic attack CAD: Coronary artery disease Deep vein thrombosis/ pulmonary embolism Congestive heart failure Authors' contributions Conceived and designed study : SY and DHLAcquired data : ABS, EW, SSelvakumar, SShetty, HK, AC, CB, GM, JY, KL, SK, VP, NAA, AK, RSS, SY, DHLDrafted or revised manuscript : ABS, JLC, AK, SY, DHLApproved the final version: ABS, EW, SSelvakumar, SShetty, QM, TS, HK, AC, CB, GM, JY, KL, SK, VP, NAA, JS, TN, JEL, MA, JLC, AK, RSK, SY, DHL. Funding Support: This study was supported by grants from the Florida Department of Health Casey DeSantis Cancer Innovation Fund (ID# MOABE) and from the American Heart Association (24CDA1271837; 10.58275/AHA.24CDA1271837.pc.gr.193589) to DHL. J.L.C. is supported by the Cortner-Couch Endowed Chair from the University of South Florida School of Medicine. These studies are also supported by Collaborative Data Services Core at Moffitt Cancer Center & Research Institute (NCI grant P30 CA076292) and by monies from the State of Florida to Moffitt Cancer Center. Data availability No datasets were generated or analysed during the current study. Declarations Competing interests TN: Clinical trial support to the institution by Novartis, Clinical trial support (drug only supply) to the institution by Karyopharm, Consultant to Alexion. JEL: Consultancy: Dedham Group, Jasper Therapeutics, Astellas, Dava Oncology, Novartis, Agios/Servio, Boxer Capital, Jazz, BerGenBio, Millenium Pharma/ Takeda, ElevateBio Management, Daiichi Sankyo, AbbVie, Servier; Research Funding: Syntrix Pharmaceuticals, Celgene/ Bristol Myers Squibb.AK: Research Support: Protagonist, Blueprint Medicines Corporation, GSK - Sierra Oncology, Prelude Pharmaceuticals, Bristol Myers Squibb, Morphosys; Consultancy: Imago Biosciences, Incyte, Blueprint Medicines Corporation, Novartis, Abbvie, GSK - Sierra Oncology, Bristol Myers Squibb, PharmaEssentia, CTI Biopharma; Honoraria: Imago Biosciences, Incyte, Blueprint Medicines Corporation, Novartis, Abbvie, GSK - Sierra Oncology, Bristol Myers Squibb, PharmaEssentia, CTI Biopharma; Speakers Bureau: Imago Biosciences, Incyte, Blueprint Medicines Corporation, Novartis, Abbvie, GSK - Sierra Oncology, Bristol Myers Squibb, PharmaEssentia, CTI Biopharma. RK: Consultancy/honoraria: Servier, Geron, Bristol Myers Squibb , Sbio, PharmaEssentia, Rigel Speakers Bureau: Servier, Sbio , Pharma Essentia, RigelDHL: Consultant fee from Lexicon PharmaceuticalsThe remaining authors decare that they have no competing interests to disclose. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Seongseok Yun and Dae Hyun Lee contributed equally to this work. Contributor Information Seongseok Yun, Email: Seongseok.yun@moffitt.org. Dae Hyun Lee, Email: Daehyun.lee@pennmedicine.upenn.edu. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement No datasets were generated or analysed during the current study.