Association between the choice of the conditioning regimen and outcomes of allogeneic hematopoietic cell transplantation for myelofibrosis

Abstract

Allogeneic hematopoietic cell transplantation (allo-HCT) remains the only curative treatment for myelofibrosis. However, the optimal conditioning regimen either with reduced-intensity conditioning (RIC) or myeloablative conditioning (MAC) is not well known. Using the Center for International Blood and Marrow Transplant Research database, we identified adults aged ≥18 years with myelofibrosis undergoing allo-HCT between 2008-2019 and analyzed the outcomes separately in the RIC and MAC cohorts based on the conditioning regimens used. Among 872 eligible patients, 493 underwent allo-HCT using RIC (fludarabine/ busulfan n=166, fludarabine/melphalan n=327) and 379 using MAC (fludarabine/busulfan n=247, busulfan/cyclophosphamide n=132). In multivariable analysis with RIC, fludarabine/melphalan was associated with inferior overall survival (hazard ratio [HR]=1.80; 95% confidenec interval [CI]: 1.15-2.81; P=0.009), higher early non-relapse mortality (HR=1.81; 95% CI: 1.12-2.91; P=0.01) and higher acute graft-versus-host disease (GvHD) (grade 2-4 HR=1.45; 95% CI: 1.03-2.03; P=0.03; grade 3-4 HR=2.21; 95%CI: 1.28-3.83; P=0.004) compared to fludarabine/busulfan. In the MAC setting, busulfan/cyclophosphamide was associated with a higher acute GvHD (grade 2-4 HR=2.33; 95% CI: 1.67-3.25; P<0.001; grade 3-4 HR=2.31; 95% CI: 1.52-3.52; P<0.001) and inferior GvHD-free relapse-free survival (GRFS) (HR=1.94; 95% CI: 1.49-2.53; P<0.001) as compared to fludarabine/busulfan. Hence, our study suggests that fludarabine/busulfan is associated with better outcomes in RIC (better overall survival, lower early non-relapse mortality, lower acute GvHD) and MAC (lower acute GvHD and better GRFS) in myelofibrosis.

Introduction

Myelofibrosis is a chronic myeloproliferative neoplasm arising either de novo (primary) or secondary to antecedent essential thrombocytosis or polycythemia vera. Despite the recent advances in disease biology and treatment options such as Janus activating kinase (JAK) inhibitors, allogeneic hematopoietic cell transplantation (allo-HCT) remains the only potentially curative option.^1-3 The availability of reduced intensity conditioning (RIC) and the choice of donors have expanded the scope of allo-HCT for these patients who are often older adults.⁴ While several factors influence outcomes of allo-HCT, conditioning intensity and conditioning regimen are aspects that could be tailored to improve the outcomes. Currently, both myeloablative conditioning (MAC) and RIC platforms are available for allo-HCT in myelofibrosis.^5-12 A large study from the European Group for Blood and Marrow Transplant (EBMT) compared the outcomes of allo-HCT with RIC versus MAC in myelofibrosis and demonstrated comparable results with both approaches, but better graft-versus-host disease (GvHD)-free -and relapse-free survival (GRFS) with MAC.⁹ However, the optimal conditioning regimen either with RIC or MAC is not well known. While some studies have previously compared different RIC regimens with varying results,^10-12 similar comparative studies with MAC are lacking and no studies have demonstrated a survival difference based on the conditioning regimen. Hence, we sought to determine the outcomes of allo-HCT for myelofibrosis based on the choice of the conditioning regimen, separately with RIC and MAC.

Methods

Study objective

Our objectives were to compare the overall survival, dis-

ease-free survival, non-relapse mortality, relapse, incidence of acute GvHD, chronic GvHD and GRFS based on the choice of the conditioning regimen used with RIC or MAC.

Data source

CIBMTR is a combined research program of the Medical College of Wisconsin and the National Marrow Donor Program. It comprises a voluntary network of more than 450 transplantation centers worldwide that contribute data on consecutive allo-HCT to a centralized statistical center.¹³ Observational studies conducted by the CIBMTR are performed in compliance with all applicable federal regulations pertaining to the protection of human research participants. Patients provided written informed consent for research. The Institutional Review Boards of the Medical College of Wisconsin and the National Marrow Donor Program approved this study.

Study population

Adults aged ≥18 years with a diagnosis of myelofibrosis (chronic phase) who underwent allo-HCT between the period 2008-2019 and data reported to the CIBMTR were identified. The cohort was then selected to focus on the most common conditioning regimens used in RIC (fludarabine/busulfan vs. fludarabine/melphalan] and MAC (fludarabine/busulfan vs. busulfan/cyclophosphamide] setting (Online Supplementary Figure S1). Conditioning regimens were classified in the CIBMTR dataset based on prior published data.^14,15 The donor groups included matched related donors, eight of eight (HLA-A, -B, -C and -DRB1) matched unrelated donors and seven of eight matched unrelated donors. Key exclusion criteria were allo-HCT from haploidentical donor, syngeneic donor, cord blood, and ex vivo T-cell depleted or CD34 selected grafts. In addition, 51 patients in fludarabine/busulfan MAC group who received post-transplant cyclophosphamide (post-Cy) were excluded as there were no such corresponding patients in busulfan/cyclophosphamide MAC group.

Statistical analysis

Baseline characteristics were summarized using descriptive statistics with median and range for continuous variables and proportions for categorical variables. Outcomes were compared separately in RIC and MAC cohorts based on the conditioning regimens. Definitions of the outcomes are provided in the Online Supplementary Appendix. Cumulative incidence estimates were calculated for competing risks outcomes including acute GvHD, chronic GvHD, non-relapse mortality, and relapse. Kaplan-Meier method was used to estimate the probabilities for survival. In order to evaluate for other relevant factors that could influence the outcomes, multivariable Cox regression analysis was used (see below for the variables included). The proportional hazards assumption was examined and covariates that violate the proportional hazards assumption were added as time-dependent covariates. In the absence of binary endpoints, hazard ratio (HR) and confidence limits were reported. A pairwise comparison within the non-reference groups was also performed in multivariable models to demonstrate their effect and shown as contrasts. Variables included in multivariable analysis were age, race/ethnicity, disease subtype (primary vs. post essential thrombocythemia [ET] or polycythemia vera [PV]), dynamic international performance scoring system (DIPSS) score, hematopoietic cell transplantation comorbidity index (HCT-CI), Karnofsky performance scale (KPS), systemic symptoms, splenic radiation, splenomegaly, interval from diagnosis to alloHCT, ruxolitinib use pretransplant, donor-recipient HLAmatch, sex match, cytomegalovirus (CMV) match, stem cell source, GvHD prophylaxis (tacrolimus based vs. cyclosporine based vs. post-Cy vs. others), use of antithymocyte globulin (ATG)/alemtuzumab, and year of transplant. A stepwise selection method was used to identify the final model with a significance level of 0.05 and only variables reaching that statistical significance were shown. In addition, adjusted univariate estimates were provided for outcomes that were significantly associated with conditioning regimen. Fine and Gray model was used for analysis of non-relapse mortality, GvHD and relapse.¹⁶ Center effect was tested using the score test proposed by Commenges and Andersen and marginal Cox models were used for further adjustments.¹⁷ Center effect was noted to be significant only for chronic GvHD and was adjusted accordingly. Missing category was included in the models as one group to avoid loss of data and power.¹⁸ All analyses were performed at a two-sided significance level of 0.05 using SAS 9.4 (SAS Institute, Cary, NC).

Results

Baseline characteristics

Of 872 eligible patients, 493 underwent allo-HCT using RIC (fludarabine/busulfan n=166, fludarabine/melphalan n=327) and 379 using MAC (fludarabine/busulfan n=247, busulfan/cyclophosphamide n=132). Key baseline characteristics of the patients are summarized (Table 1; Online Supplementary Tables S1 and S2; unadjusted univariate estimates in Online Supplementary Tables S3 and S4). In the RIC cohort, compared to fludarabine/busulfan patients, fludarabine/melphalan patients had longer median interval from diagnosis to allo-HCT (37 vs. 22 months, P=0.02), lower proportion with antithymocyte globulin/alemtuzumab use (25% vs. 52%, P<0.01), and higher proportion with pretransplant ruxolitinib use (61% vs. 49%, P=0.03). In the MAC cohort, compared to fludarabine/busulfan patients, busulfan/cyclophosphamide patients had younger age (median age 55 vs. 60 years, P<0.01), higher proportion with low-intermediate risk disease (61% vs. 54%, P=0.03), higher proportion with bone marrow graft (12% vs. 4%, P<0.01), lower proportion with antithymocyte globulin/alemtuzumab use (5% vs. 45%, P<0.01), and lower proportion with pretransplant ruxolitinib use (43% vs. 59%, P<0.01). Median follow-up of the cohort was 26 (range, 3-150) months.

Overall survival

In multivariable analysis (Table 2), overall survival in the RIC setting was significantly worse with fludarabine/melphalan (HR=1.80; 95% CI: 1.15-2.81; P=0.009, 2-year adjusted overall survival 54.4% vs. 60.9%) as compared to fludarabine/busulfan (Figure 1). In the MAC setting, overall survival was not significantly different between based on the conditioning regimen (busulfan/cyclophosphamide HR=1.14; 95% CI: 0.75-1.71; P=0.54) (Figure 2). Other factors significantly associated with overall survival were donorrecipient HLA match (higher risk with unrelated donors in the MAC setting) and the use of antithymocyte globulin/alemtuzumab (higher risk in the RIC setting) (Online Supplementary Tables S5 and S6).

Disease-free survival

In multivariable analysis (Table 2), disease-free survival was not significantly different based on the conditioning regimen used in RIC (fludarabine/melphalan HR=1.03; 95% CI: 0.77-1.38; P=0.85) or MAC (busulfan/cyclophosphamide HR=1.03; 95% CI: 0.77-1.38; P=0.83) settings (Online Supplementary Figures S2 and S3). Other factors significantly associated with disease-free survival were Karnofsky performance status (higher risk with lower score in MAC) and pretransplant ruxolitinib use (higher risk in MAC) (Online Supplementary Tables S5 and S6).

Table 1.Key baseline characteristics.

Non-relapse mortality

In the RIC setting, there was a significantly higher risk of early non-relapse mortality with fludarabine/melphalan as compared to fludarabine/busulfan (17.4% vs. 4.3%, HR=1.81; 95% CI: 1.12-2.91; P=0.01). Beyond 6 months the risk of non-relapse mortality was low with fludarabine/melphalan (HR=0.46; 95% CI: 0.23-0.91; P=0.02) (Table 2; Online Supplementary Figure S4) (cut-off of 6 months was chosen due to non-proportional hazard). No significant differences in non-relapse mortality were seen with the MAC-based on the conditioning regimens (busulfan/cyclophosphamide HR=1.36; 95% CI: 0.83-2.21; P=0.22) (Online Supplementary Figure S5). The other factor significantly associated with non-relapse mortality was donor-recipient HLA-match (higher risk with unrelated donors in MAC) (Online Supplementary Tables S5 and S6).

Relapse

The risk of relapse was not significantly different based on the conditioning regimen used in RIC or MAC (RIC - fludarabine/melphalan HR=0.85; 95% CI: 0.64-1.12; P=0.25; MAC - busulfan/cyclophosphamide HR=0.92; 95% CI: 0.64-1.32; P=0.65) (Online Supplementary Figures S6 and S7; Online Supplementary Tables S5 and S6). Other factors significantly associated with relapse were Karnofsky performance status (higher risk with poor score in MAC), pretransplant ruxolitinib use (higher risk in MAC) and year of transplant (higher risk with recent period in RIC).

Graf-versus-host disease

In the RIC setting, fludarabine/melphalan was associated with a significantly higher risk of acute GvHD grade 2-4 (fludarabine/melphalan 40%, fludarabine/busulfan 35.3%, HR=1.45; 95% CI: 1.03-2.03; P=0.03) and grade 3-4 (fludarabine/melphalan 21.8%, fludarabine/busulfan 12.1%, HR=2.21; 95% CI: 1.28-3.83; P=0.004) (Online Supplementary Figures S8 and S9). In the MAC setting, busulfan/cyclophosphamide was associated with a significantly higher risk of acute GvHD grade 2-4 (busulfan/cyclophosphamide 58.9%, fludarabine/busulfan 34.4%; HR=2.33; 95% CI: 1.67-3.25; P<0.001) and grade 3-4 (busulfan/cyclophosphamide 32.6%, fludarabine/busulfan 11.9%; HR=2.31; 95% CI: 1.52-3.52; P<0.001) (Online Supplementary Figures S10 and S11). Chronic GvHD was significantly associated with donor-recipient HLA-match (higher risk with 7/8 matched unrelated donors in RIC) and pretransplant ruxolitinib use (lower risk in MAC), but not by the conditioning regimen (Online Supplementary Tables S5 and S6).

Table 2.Multivariable analysis of outcomes based on conditioning regimen.

Graf-versus-host disease-free relapse-free survival

In the RIC setting, GRFS was not significantly different between fludarabine/busulfan and fludarabine/melphalan (HR=1.11; 95% CI: 0.90-1.35; P=0.32) (Online Supplementary Figure S12). However, in the MAC setting, busulfan/cyclophosphamide was associated with significantly inferior GRFS (HR=1.94; 95% CI: 1.49-2.53; P<0.01) (2-year adjusted probability 5.1% vs. 19.4%) as compared to fludarabine/busulfan (Table 2; Figure 3). Other factors significantly associated with GRFS included recipient age (in MAC) and donorrecipient HLA-match (higher risk with unrelated donors in MAC) (Online Supplementary Tables S5 and S6).

Figure 1.Overall survival with reduced-intensity conditioning. RIC: reduced-intensity conditioning; Flu: fludarabine; Bu: busulfan; Mel: melphalan.

Figure 2.Overall survival with myeloablative conditioning. MAC: myeloablative conditioning; Flu: fludarabine; Bu: busul-fan; Mel: melphalan.

Engrafment

The rates of neutrophil engraftment (30 days) were significantly better with fludarabine/busulfan in RIC (fludarabine/busulfan 95.1% vs. fludarabine/melphalan 92.4%; P=0.006) and MAC (fludarabine/busulfan 95.2% vs. busulfan/cyclophosphamide 87.2%; P=0.02). The rate of platelet engraftment (100 days) was better with fludarabine/busulfan in the RIC setting (RIC - fludarabine/busulfan 84.4% vs. fludarabine/melphalan 73.9%; P<0.001; MAC -fludarabine/busulfan 86.1% vs. busulfan/cyclophosphamide 83.7%; P=0.27).

Additional analyses

In the RIC cohort, we investigated whether the outcomes differed based on the dose of melphalan (100 vs. 140 mg/m²) used in fludarabine/melphalan group. As shown in the Online Supplementary Table S7, the outcomes did not significantly vary based on the dose of melphalan (shown as contrasts between melphalan 100 vs. 140 mg/m²).

Discussion

Our study highlights the significant differences in outcomes of allo-HCT for myelofibrosis based on the choice of the conditioning regimen. Fludarabine/busulfan conditioning was associated with superior overall survival, lower early non-relapse mortality and lower acute GvHD (all with RIC), and lower acute GvHD and superior GRFS with MAC. A key aspect of conditioning strategy is its ability be tailored in order to improve the outcomes. Events such as non-relapse mortality and GvHD that affect the morbidity and mortality after allo-HCT could be influenced by the conditioning strategy and efforts to minimize these complications are vital to improve the long-term success. Although RIC and MAC platforms are clinically decided based on factors such as age, comorbidities, performance status, and other aspects that are often not modifiable, our results illustrate the influence of common conditioning regimens used in these settings and provides valuable information for choosing the appropriate regimen in clinical practice.

Prior retrospective studies have evaluated the impact of conditioning intensity and regimen in myelofibrosis, albeit with variable results and certain key differences compared to our study.^5-12 A study by Robin et al. included 160 patients with myelofibrosis from two European centers (Paris [fludarabine/busulfan] or Hamburg [fludarabine/melphalan]), but with antithymocyte globulin given for all patients who received fludarabine/busulfan conditioning.¹¹ Another CIBMTR study by Gupta et al. included only patients with primary myelofibrosis and RIC (fludarabine/TBI vs. fludarabine/melphalan vs. fludarabine/busulfan) between 1997-2010 with a relatively younger patient population (median age 55 years).¹⁰ Hence, the differences in the study population, the nature of the cohort (registry- vs. individual center-based), treatment received and variations in time period included could have contributed to the differences in results noted between the current study and prior studies. To date, prospective studies of conditioning regimen in myelofibrosis are single-arm or comparative studies with smaller sample size.^19,20,21 For example, a phase II study by Patriarca et al. prospectively compared fludarabine/busulfan and fludarabine/thiotepa for allo-HCT in 60 patients with myelofibrosis and showed similar outcomes with both these regimens.²¹ Hence, our study addresses the knowledge gap in this area using a larger dataset with a comparison of commonly reported conditioning regimens. Unfortunately, due to the limited number of patients receiving other less common conditioning regimens such as fludarabine/thiotepa, these regimens could not be compared in our study. Additionally, given the results of a large EBMT study showing no difference in overall survival between MAC and RIC,⁹ we did not compare the outcomes of MAC versus RIC in our analysis which also helped to minimize the heterogeneity in comparisons.

Figure 3.Graf-versus-host-disease-free relapse-free survival with myeloablative conditioning. MAC: myeloablative conditioning; CY: cyclophosphamide; GvHD: graft-versus-host disease; Flu: fludarabine; Bu: busulfan; Mel: melphalan.

Apart from the conditioning regimen, factors such as donor-recipient HLA-match, performance status and use of antithymocyte globulin/alemtuzumab influenced the outcomes similar to prior studies. The imbalances in baseline characteristics were adjusted in multivariable models and there were no significant interactions noted between the baseline characteristics and main effect (conditioning regimen). Antithymocyte globulin/alemtuzumab was associated with worse overall survival in RIC and was more commonly used with fludarabine/busulfan regimens. Despite this, an early survival advantage was noted with fludarabine/busulfan in RIC. The association between the outcomes and factors such as the route of busulfan administration (oral vs. intravenous, targeted vs. non-targeted; data not shown) and the dose of melphalan (in RIC) were also investigated and none was found. In MAC, ruxolitinib prior to allo-HCT was associated with higher risk of relapse, inferior disease-free survival, higher risk of acute GvHD and lower risk of chronic GvHD. Although prior studies indicate the feasibility and safety of ruxolitinib therapy prior to alloHCT,^22,23 we could not evaluate the possible mechanisms behind these differences due to limited information on the duration, dose, response, and other aspects of ruxolitinib therapy. Other factors such as the role of splenectomy, spleen size or splenic radiation therapy and their association with outcomes could not be evaluated due to the small number of patients with those interventions.

Despite the large sample size, our study is limited by the retrospective design and lack in-depth information on factors such as genomic mutations and therapies for myelofibrosis given pre- and post-allo-HCT that could affect the outcomes.^24,25 The lack of detailed information on genomic mutations precluded further analyses and calculation of molecular risk scores (such as MIPSS70, MYSEC-PM etc.). For example, a study by Gagelman et al. investigated the prognostic significance of somatic mutations in myelofibrosis patients undergoing allo-HCT and identified that ASXL1 and non-CALR/MPL driver mutations were associated with poor outcomes. This study also established a prognostic model with variables such as patient age, performance status, white blood count, platelet count, HLAmismatched donor and molecular mutations. However, due to the lack of information on these aspects, we could not apply this scoring system in our study.²⁴ We also could not assess the reasons behind the choice of individual conditioning regimens used for these patients, understanding that centers could have their preferences while choosing conditioning regimens. However, we evaluated for center-effects in multivariable analyses and adjustments were made accordingly. As our study mainly focused on patients with chronic phase myelofibrosis, the role of conditioning strategy in advanced-phase disease (accelerated/blast phase) was not evaluated. Due to the nature of the GvHD reporting in the dataset, chronic GvHD was analyzed as a whole outcome without further stratification (mild, moderate, severe).

Our study demonstrates that fludarabine/busulfan-based conditioning is associated with superior overall survival, lower early non-relapse mortality, and lower acute GvHD with RIC and lower acute GvHD and superior GRFS with MAC. The results provide valuable information for tailoring the conditioning strategies to minimize non-relapse mortality and GvHD and improve survival. Prospective comparative studies are warranted to confirm these results and identify the ideal conditioning regimen in myelofibrosis.

Footnotes

• Received September 6, 2022
• Accepted February 1, 2023

Correspondence

G.S. Guru Murthy gmurthy@mcw.edu

Disclosures

GSGM reports the following all outside the submitted work: honoraria from Cardinal Health, DAVA Oncology and Curio science; advisory board membership of TG Therapeutics, consultancy for Gilead, Cancerexpert now, Qessential and Techspert. AA reports research funding from Incyte Corporation. TB reports honorarium from Pfizer Hematology and Oncology. JC reports participation on a Data Safety Monitoring Board for Allovir, Inc.; financial relationships with Actinium Pharmaceuticals, Bluebird Bio Inc., Dynavax Pharma, Atyr Pharmac, Gamida Cell, Miragen Therapeutics, Mustang Bio, Novavax, Ovid Therapeutics, Sorrento Therapeutics, TG Therapeutics, Vaxart Inc, and Veru Inc., outside the submitted work. BD reports institutional research funding with Takeda, Janssen, Angiocrine, Pfizer, Poseida, MEI, Sorrento; consultancy with Jazz, Celgene, and Gamida Cell. SG reports financial relationships as speaker with Seattle Genetics and KITE Pharma; and advisory board member with Kadmon, BMS, Sanofi, Astrazeneca, Kite, Daiichi Sankyo and Astellas. VG reports consultancy work for Novartis, Incyte, BMS-Celgene, Sierra Oncology, Morphosys, Pfizer, and Takeda, and received research grant through institution from Novartis and Incyte. MRG reports having worked as a PI with multiple pharmaceutical sponsors and as both consultant and PI for Incyte (manufacturer of ruxolitinib). NH reports as advisory board member of Novartis. TJ reports honoraria for advisory board participation for Care Dx. Bristol Myers Squibb, and Incyte; honoraria for lecture at APP Oncology Summit. TN reports research support (clinical trial support) to the institution by Novartis; research support (drug supply only) to the institution for clinical trial by Karyopharm. GO reports financial relationships with Incyte (grant), BMS (personal fees), Incyte (personal fees), Novartis (personal fees), and Pfizer (personal fees). RM reports research support and stock ownership with Incyte. DRA reports as a consultant and on speaker bureau for Incyte (makers of ruxolitinib used for treatment).

Contributions

GM, WS, SK and NE conceived and designed the study, collected and assembled the data, and wrote the manuscript. All authors performed data analysis and interpretation of data; and GM, WS and SK provided final approval of the manuscript. GM and WS had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The views expressed in this article do not reflect the ofcial policy or position of the NIH, the Department of the Navy, the Department of Defense, the HRSA, or any other agency of the US Government.

Data-sharing statement

CIBMTR supports accessibility of research in accord with the National Institutes of Health (NIH) data-sharing policy and the National Cancer Institute (NCI) Cancer Moonshot public access and data-sharing policy. The CIBMTR only releases de-identified datasets that comply with all relevant global regulations regarding privacy and confidentiality.

Funding

The CIBMTR is supported primarily by Public Health Service U24CA076518 from the National Cancer Institute (NCI), the National Heart, Lung and Blood Institute (NHLBI) and the National Institute of Allergy and Infectious Diseases (NIAID); HHSH250201700006C from the Health Resources and Services Administration (HRSA); and N00014-20-1-2705 and N00014-20-1-2832 from the Ofce of Naval Research; support is also provided by Be the Match Foundation, the Medical College of Wisconsin, the National Marrow Donor Program, and from the following commercial entities: AbbVie; Accenture; Actinium Pharmaceuticals, Inc.; Adaptive Biotechnologies Corporation; Adienne SA; Allovir, Inc.; Amgen, Inc.; Astellas Pharma US; bluebird bio, inc.; Bristol Myers Squibb Co.; CareDx; CSL Behring; CytoSen Therapeutics, Inc.; Daiichi Sankyo Co., Ltd.; Eurofins Viracor, DBA Eurofins Transplant Diagnostics; Fate Therapeutics; Gamida-Cell, Ltd.; Gilead; GlaxoSmithKline; HistoGenetics; Incyte Corporation; Iovance; Janssen Research & Development, LLC; Janssen/Johnson & Johnson; Jasper Therapeutics; Jazz Pharmaceuticals, Inc.; Kadmon; Karius; Karyopharm Therapeutics; Kiadis Pharma; Kite Pharma Inc; Kite, a Gilead Company; Kyowa Kirin International plc; Kyowa Kirin; Legend Biotech; Magenta Therapeutics; Medac GmbH; Medexus; Merck & Co.; Millennium, the Takeda Oncology Co.; Miltenyi Biotec, Inc.; MorphoSys; Novartis Pharmaceuticals Corporation; Omeros Corporation; OncoImmune, Inc.; Oncopeptides, Inc.; OptumHealth; Orca Biosystems, Inc.; Ossium Health, Inc; Pfizer, Inc.; Pharmacyclics, LLC; Priothera; Sanofi Genzyme; Seagen, Inc.; Stemcyte; Takeda Pharmaceuticals; Talaris Therapeutics; Terumo Blood and Cell Technologies; TG Therapeutics; Tscan; Vertex; Vor Biopharma; Xenikos BV. The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.