Citations

OBJECTIVE This study aimed to evaluate local control (LC) of tumors, patient overall survival (OS), and the safety of stereotactic radiosurgery (SRS) for esophageal cancer brain metastases (EBMs). METHODS This retrospective cohort study used data from 15 International Radiosurgery Research Foundation facilities encompassing 67 patients with 185 EBMs managed using SRS between January 2000 and May 2022. The median patient age was 63 years, with a male predominance (92.5%). Most patients (64.2%) had a single brain metastasis, while 7.5% had more than 5 metastases. The median tumor volume was 0.9 cm 3 , and the median margin dose delivered to the tumor was 20 Gy. RESULTS The median OS post-SRS was 15.2 months, with 1- and 2-year OS rates of 65.7% and 32.3%, respectively. A significant inverse correlation was found between the number of EBMs and OS in the univariable analysis. LC rates at 1 and 2 years were 89% and 76%, respectively. Adverse radiation effects (AREs) were observed in 17.9% of patients, with 13.4% being mild and transient and 4.5% severely symptomatic (Common Terminology Criteria for Adverse Events grade 3). New intracranial disease developed in 58.2% of patients, with 1- and 2-year rates of 58% and 73%, respectively. CONCLUSIONS SRS for EBMs demonstrated high survival rates and effective tumor control, with a low incidence of severe AREs. These findings highlight the potential role of SRS in the multidisciplinary multimodality management paradigm of EBM.

ABSTRACT Background Tricuspid valve‐in‐valve (TVIV) implantation has traditionally included post‐procedure overnight in‐hospital observation with next‐day discharge. Recently, our practice has shifted toward same‐day discharge for most patients after vascular access site hemostasis and adequate recovery from anesthesia. However, data on the safety of this approach remains scarce. Aims Our study aims to describe our practice change and to identify factors that may lead to overnight observation post‐TVIV. Methods We retrospectively reviewed medical records of patients who underwent uncomplicated TVIV implantation as an outpatient procedure at Mayo Clinic between 2011 and 2023. Results Of the 83 patients who underwent TVIV from 2011 to 2023, 65 (78%) were included in the study. The first 39 patients in our institutional series were observed overnight as standard practice. 18 (69%) of the most recent 26 patients in our series were discharged the same day; three patients who remained in the hospital overnight were because of provider preference and/or patient preference without any specific medical concern. Of all 47 patients who were observed overnight, one patient was treated for fluid overload, one patient had non‐sustained ventricular tachycardia on telemetry that resulted in adjustment of antiarrhythmic therapy, two patients had continued bleeding from the access sites and was observed for another day. Among the two of them, one also developed atrial tachyarrhythmias requiring initiation of additional antiarrhythmic therapy. Out of 18 patients discharged the same day, one patient had an ED presentation within 24 h due to access site bleeding which was resolved promptly. There was no mortality or procedural‐related readmission. Conclusion Same‐day discharge may be considered for adult patients who have undergone uncomplicated TVIV with adequate hemostasis and recovery after anesthesia.

Background: PV is a myeloproliferative neoplasm that leads to excessive production of erythrocytes and increased platelet and leukocyte counts. The current standard of care is to use therapeutic phlebotomy (TP) with or without cytoreductive therapy (CRT) to treat PV. Over time, the CRT dose often needs to be increased to treat PV progression. Preliminary results from the phase 2 REVIVE study (NCT04057040) demonstrated that rusfertide was superior to placebo in achieving hematocrit (Hct) levels <45% and reducing or eliminating the need for TP in patients who were phlebotomy-dependent prior to study entry (Kremyanskaya et al. EHA2023. [Abstract LB2710]; Kremyanskaya et al. New Engl J Med 2024;390:723-35).Final data from REVIVE will be presented. Methods: In Part 1 (Weeks 1-29), subcutaneous rusfertide (starting dose: 20 mg q1w) was initiated and titrated to control Hct to <45%. In Part 2 (Weeks 29-41; blinded randomized withdrawal phase), patients were randomized to receive rusfertide or placebo. Randomized patients were eligible to participate in the open-label extension (OLE) portion of the study (Part 3), in which all patients could receive rusfertide and investigators could make CRT dose adjustments. Results: In Part 1, 70 patients were enrolled; 59 were randomized in Part 2. Fifty-eight patients continued to Part 3. As of 9 July 2024, 50 (71.4%), 37 (52.9%), and 17 (24.3%) patients received rusfertide for ≥2, ≥2.5, or ≥3 years, respectively.Overall, median (range) patient age was 57.5 years (27-77); the majority were male (70%) and had high-risk disease (55.7%). Overall, 52.9% and 47.1% of patients were treated with TP alone or TP plus CRT, respectively. Prior to enrollment, the estimated average (mean) phlebotomy rate (EAPR) in patients who enrolled on study was 8.5/year, which was reduced to <1.0/year in Part 1. In Part 2, the EAPR was <1.0/year and 6.6/year in the rusfertide and placebo groups, respectively. For patients who continued onto Part 3 and received rusfertide, the EAPR remained at <1.0/year. In Parts 1, 2, and 3, the mean (SD) weekly average rusfertide dose was 30.7 mg (12.2), 38.5 mg (21.3), and 44.9 mg (20.3), respectively. Rusfertide consistently maintained Hct <45%, including in patients receiving therapy for ≥3 years. Platelets increased following initiation of rusfertide therapy and stabilized over time; mean leukocyte counts remained stable. Prior to enrollment, iron-related parameters were consistent with systemic iron deficiency. Throughout the study, rusfertide resulted in the normalization of mean serum ferritin levels, increased mean transferrin saturation, and increased mean serum iron levels. Overall, mean corpuscular volume increased slightly over the duration of the study. Throughout the study, there was no evidence of increased markers of inflammation (eg, C-reactive protein or interleukin 6). The most common (≥20%) treatment-emergent adverse events (TEAEs) were injection site reactions (85.7%), fatigue (38.6%), COVID-19 (32.9%), pruritus (34.3%), arthralgia (30.0%), dizziness (30.0%), nausea (24.3%), headache (24.3%), and anemia (21.4%). Grade 3 TEAEs occurred in 25.7% of patients.There were no Grade 4 or 5 TEAEs. One patient developed acute myeloid leukemia after treatment discontinuation. After more than 150 patient-years of rusfertide exposure, malignancies (most of which were skin cancers) were reported in 11 patients (all had risk factors that may have contributed to the development of these malignancies). There was no obvious correlation between increasing exposure to rusfertide and malignancies reported. Six thrombotic events (5 arterial and 1 venous) occurred in 5 patients with high-risk PV. Conclusions: In REVIVE, rusfertide added to TP with or without CRT provided long-term durable control of Hct and decreased the need for TP. Patients are eligible to roll over to the phase 2 THRIVE OLE study (NCT06033586), which will continue to assess the long-term safety and efficacy of rusfertide for a total period of up to 5.8 years. The randomized phase 3 VERIFY study (NCT05210790) is ongoing and will evaluate rusfertide vs placebo in PV patients.

Background: Positron emission tomography (PET) scan is the imaging modality of choice for initial staging and response assessment of follicular lymphoma (FL). Quantitative parameters such as standardized uptake value (SUV) have been investigated as putative prognosticators in FL, with high baseline SUVmax predicting aggressive disease. To date, there is no data to guide treatment choices in patients (pts) with high SUVmax but with no documented evidence of histologic transformation (HT). Specifically, it remains unclear whether anthracycline-containing regimens, such as R-CHOP, are more beneficial than R-Bendamustine (BR) in these pts. Therefore, we aimed to compare the efficacy of frontline R-CHOP and BR in newly diagnosed FL pts based on the SUVmax at baseline PET (bPET). Patients and Methods: Pts with newly diagnosed, biopsy-proven FL1-2 and FL3A who had a bPET scan and who received either first-line R-CHOP or BR within 6 months of diagnosis between 2002 and 2022 were evaluated. Analysis was restricted to pts with age < 70 years to reduce confounding of age with treatment selection. Cohorts included were the FOLL12 (n=562, Luminari et al JCO 2022) and the R-chemotherapy arm of the RELEVANCE (n=216, Morschhauser et al NEJM 2018) clinical trials, the prospectively enrolled Molecular Epidemiology Resource (MER) and the Lymphoma Epidemiology of Outcomes (LEO) cohorts (n=288) and a retrospective cohort from 13 Australian centers (n=219). Choice of chemotherapy regimen was per treating physician. Maintenance was per protocol (FOLL12/RELEVANCE) or per treating physician (MER/LEO/Australia). Event-free survival (EFS) and overall survival (OS) were defined as time from treatment initiation to progression, relapse, retreatment or death; both were censored at 72 months due to varying follow-up across cohorts. SUVmax was grouped by tertile within each cohort to account for potential differences in SUVmax distributions and patient populations across cohorts. Kaplan Meier curves and multivariable Cox models cohort-stratified and adjusted for clinical factors (FLIPI, ECOG PS, age, grade 3A) were used to evaluate outcomes. Results: The median age at diagnosis was 57 years (IQR 49-64). 651 (51%) pts were treated with R-CHOP and 634 (49%) with BR. Within the study cohorts, there was a prevalence of cases receiving BR compared to R-CHOP in the Australian (90% vs 10%) and MER/LEO (68% vs 32%) cohorts, while the opposite was observed in the RELEVANCE (16% vs 84%) and FOLL12 (37% vs 63; p<0.001) trials. Pts receiving R-CHOP were slightly younger (median 56 vs 58 years, p=0.005), more frequently had grade 3A (17% vs 7%, p<0.001), elevated LDH (27% vs 22%, p=0.029), bulky disease > 7 cm (44% vs 37%, p=0.019), lower ECOG PS (PS >1 7% vs 13%, p<0.001), and higher rate of high risk FLIPI (37% vs 30%, p=0.005). Baseline SUVmax was slightly lower in R-CHOP (median 10.7, IQR 7.9-14.9) vs BR (median 11.0, IQR 8.2-14.2) treated pts (p=0.0045). At a median follow-up of 56 months (range, 1-72 months), 415 pts (32%) had an event and 103 (8%) died. Higher SUVmax was not associated with EFS (middle tertile HR=0.96, 95% CI 0.76-1.22; upper tertile HR=0.86, 95% CI 0.68-1.09; p=0.45) or OS (middle tertile HR=1.28, 95% CI 0.80-2.05; upper tertile HR=0.97, 95% CI 0.59-1.59; p=0.44) in multivariable models stratified on cohort and adjusted for treatment and clinical factors. We then compared outcomes by treatment within tertiles of SUVmax. In multivariable models, R-CHOP had modestly inferior EFS to BR in low (HR=1.27, 95% CI 0.84-1.93) and middle tertiles (HR=1.49, 95% CI 0.99-2.25), and modestly superior OS to BR in low (HR=0.59, 95% CI 0.23-1.47) and middle (HR=0.79, 95% CI 0.37-1.71) tertiles. Both EFS (HR=1.06, 95% CI 0.71-1.58) and OS (HR=1.02, 95% CI 0.45-2.30) for R-CHOP were similar to BR in the highest tertile. No comparisons identified a statistically significant benefit for either therapy in multivariable models when evaluating SUVmax tertile subsets. Conclusions: In this large multicenter study of newly diagnosed FL pts with bPET who received frontline immunochemotherapy, higher baseline SUVmax was not associated with inferior outcomes after adjusting for treatment and clinical factors. Within the caveats of retrospective studies and confounding of treatment selection, these data provide no clear insights for SUVmax-based therapy selection in first-line follicular lymphoma pts treated with immunochemotherapy.

Cytokine release syndrome (CRS) commonly occurs during initiation of bispecific T-cell engaging therapy (BsAbs) for multiple myeloma (MM) (NEJM 2022;387:495, NEJM 2022;387:2232, Nat Med 2023;29:2259). Pre-medications, step-up dosing (SUD), and monitoring minimize high-grade toxicity risk, while protocolized outpatient (OP) management of BsAbs and toxicities could reduce resource utilization and improve patient satisfaction. At Mayo Clinic, SUD, clinical assessment and grade 1 CRS management occurs primarily as an OP for patients with MM receiving BsAbs. This study aims to assess the current OP management practices of BsAbs and related toxicities including patient outcomes, admission rates, and resource utilization. Patients with MM who received SUD from August 23, 2023, to March 29, 2024, were included. Baseline demographics, patient outcomes, and CRS management details are summarized. Resource utilization outcomes included level of care, length of hospital stay (LOS), and high-cost drug administration. 38 patients received MM BsAbs during the study timeframe. 34 patients initiated OP SUD (teclistamab n=17 and talquetamab n= 17). The median age was 71 years, 59% were male, 82% were white, and 88% had Medicare coverage. Median number of prior lines of therapy was 6. There were 317 OP encounters with a median of 10 encounters per patient. Of these, 72% were planned in-person assessments, 5% were unplanned clinic visits for potential toxicity-related symptom assessment, and 20% were phone visits that obviated in-person assessment. 66 of 68 doses of talquetamab (total of 1,628 mg) and 48 of 50 doses of teclistamab (total of 2,048 mg) were administered and billed in the OP setting. 24 patients (71%) that initiated OP SUD developed CRS (grade 1 n=16, grade 2 n=7). Management of the initial CRS event included steroids and acetaminophen, with or without tocilizumab. 2,600 mg of tocilizumab administered in the OP setting. Recurrent CRS was seen in 5 patients and also treated entirely OP. 4 patients developed ICANS (grade 1 n=3, grade 3 n=1). Hospitalization following OP SUD initiation occurred in 18 (53%) patients at some point during treatment. Indications included grade 2 CRS (n=6), ICANS (grade 1 n=3), refractory CRS (n=7), infection (n=1), GI bleed (n=1), planned observation for comorbidities or prior CRS (n=2), loss of caregiver (n=1), atrial fibrillation (n=1), heart failure (n=1), and pain crisis (n=1). Median LOS was 2 days. Recurrent admission occurred in 41% and median LOS was also 2 days for subsequent admissions. Two patients were admitted to the intensive care unit for hypotension or arrhythmia. 1 patient required dialysis. In total for SUD during the study time frame, 51 mg of teclistamab, 176 mg of talquetamab, and 8,240 mg of tocilizumab were billed under an IP encounter, which likely was due to inpatient administration or hospital admission within 72 hours of receiving the OP SUD. This descriptive study of our OP BsAb management practice demonstrated that 47% of patients remained OP, even with a CRS event. The CRS rate did not surpass those in landmark clinical trials and aligns with real-world data previously reported (Blood Cancer J 2024;14:35.). Availability and use of patient monitoring capabilities along with lack of sentinel outcomes emphasizes the safety of an OP BsAb practice. Drivers of financial burdens and savings include location and reimbursement of high-cost drugs and admission rates. Median LOS was less than half of what's anticipated with inpatient SUD. Inpatient administration of high-cost drugs can be more costly compared to OP due to associated medical costs and insufficient reimbursement in a population primarily insured by Medicare which reimburses inpatient mediations based on the Diagnostic Related Group (DRG) unless the medication has New Technology Add on Payment (NTAP). Using expected Medicare reimbursement rates and wholesale acquisition costs while excluding variable DRG reimbursement and NTAP impacts for teclistamab, the total net revenue improvement for shifting use of all teclistamab, talquetamab, and tocilizumab OP is $115,004 for the period analyzed (CMS.gov 2024). In conclusion, OP management of patients undergoing SUD for MM BsAbs is safe and has substantial reductions in resource utilization and costs.

Introduction: NPM1 mutations, seen in up to 30% of adults with acute myeloid leukemia (AML), often co-occur with FLT3 internal tandem duplication (FLT3-ITD). Measurable residual disease (MRD) testing before allogeneic hematopoietic cell transplant (alloHCT) for persistence of mutated NPM1 and/or FLT3-ITD can identify patients with AML in complete remission (CR) at highest risk of relapse and death. Quantitative RT-PCR assays are currently recommended for NPM1 MRD testing, while FLT3-ITD MRD testing is DNA-based using next generation sequencing (NGS). DNA-based NGS assays for mutated NPM1 MRD testing may have several advantages but require validation. We performed NPM1 MRD detection in blood collected from patients with AML during first CR (CR1) prior to alloHCT on a subset of patients from the Pre-MEASURE study (PMID: 36881031) using a highly-sensitive, commercially available DNA-based NPM1 NGS MRD assay. Methods: Of the 531 patients aged 18 or older from the Pre-MEASURE study who received an alloHCT for NPM1mutated AML during CR1 at a CIBMTR reporting site in the USA between the years of 2013-2019, a subset of 190 were randomly selected to be included in this analysis. A commercially available research testing kit (IVS, Invivoscribe, San Diego, CA), which can detect DNA-based NPM1 MRD, was used to establish a workflow following clinical testing standards and validated to detect NPM1 insertion variants down to a variant allele fraction (VAF) of at least 0.005%. Results were compared to those previously reported using an anchored multiplex PCR-based (AMP) targeted NGS assay (detection limit 0.01%) and FLT3-ITD MRD results, where applicable. Overall survival (OS) and cumulative incidence of relapse were estimated with the day of transplant as time 0 using Kaplan-Meier estimation (log-rank tests), Fine and Gray model, and Cox proportional hazards models. Results: NPM1 NGS MRD testing was successful for 186 (98%) patients, of which 48 (26%) relapsed (median: 4.0 months; range: 0.7-52.5) and 64 (34%) died (median: 9.7 months; range: 0.4-58.9) post-alloHCT. The IVS assay detected NPM1 insertion variants in 71 patients with a median VAF of 0.0026% (range:0.0002-2.1%). Using a VAF threshold of ≥0.01% for MRD positivity, the IVS test showed that 24 patients tested positive for NPM1 MRD pre-alloHCT, which was associated with significantly increased rates of relapse (52% vs 20% at 3yrs; HR=4.3; P<0.001) and decreased OS (34% vs 71% at 3yrs; HR=3.6; P<0.001) compared to testing negative, in close alignment with those previously reported using the AMP assay. After removing the VAF threshold, an additional 47 patients were classified as NPM1 MRD positive, which remained associated with increased rates of relapse (40% vs 15% at 3yrs; HR=3.4; P<0.001) and decreased OS (50% vs 75% at 3yrs; HR=2.9; P<0.001). In multivariable analysis, residual NPM1 MRD burden prior to alloHCT was associated with risk of relapse and death in a dose-dependent manner. Patients were further subdivided based on their reported FLT3-ITD mutational status at baseline. Increased risk of relapse was seen for NPM1 MRD positive patients regardless of baseline FLT3-ITD. 123 (66%) patients were co-mutated for FLT3-ITD at baseline and previously tested for FLT3-ITD MRD using the IVS assay (PMID: 38696205). Defining MRD based on the presence of NPM1 vs NPM1 and/or FLT3-ITD revealed equivalent levels of relapse (45% vs 44%) and OS (45% vs 46%) at 3 years, with NPM1 MRD evaluation identifying 5 additional relapses. These results were confirmed in the full Pre-MEASURE cohort (n=317) using the AMP assay. NPM1 MRD positive patients receiving nonmyeloablative conditioning or reduced-intensity conditioning (RIC) without melphalan (mel) had increased risk of relapse or death compared to patients receiving myeloablative conditioning or RIC with mel, regardless of FLT3-ITD co-mutational status (3yrs: relapse 87% vs 55%, P=0.006; OS 15% vs 42%, P=0.013). Conclusions: In patients with NPM1 mutated AML from the Pre-MEASURE study, we show that detection of residual NPM1 variants in pre-transplant blood during CR1 using a highly sensitive DNA-based assay is associated in a dose-dependent manner with a significantly increased risk of relapse and death after allo-HCT, which can be mitigated in part by conditioning regimen. In patients co-mutated for both FLT3-ITD and NPM1 at diagnosis, NPM1 should be prioritized as a target for NGS-MRD if only one test is available.

Background: FL3B is the least common among FL subtypes, accounting for 5-10% of cases. Almost half of FL3B tumors have concurrent diffuse large B cell lymphoma (DLBCL) in the diagnostic tissue biopsy, while FL3B rarely coexists with FL1-2 or FL3A. Several lines of evidence suggest that FL3B displays distinct biological and clinical characteristics resembling aggressive DLBCL rather than the indolent FL1-3A. Accordingly, DLBCL anthracycline-containing regimens, such as R-CHOP, are the mainstay of therapy for FL3B with or without coexisting DLBCL. However, while some studies show that R-CHOP therapy in FL3B was associated with 5-year overall survival (OS) rate of 54%, others report outcomes comparable to FL3A with 79-90% survival. Additional controversy involves the clinical implication of pure FL3B (FL3Bp) vs FL3B with concurrent DLBCL (FL3Bc). Thus, it remains unclear if FL3B clinically resembles DLBCL or lower-grade FL and whether FL3Bp has a different prognosis from FL3Bc. To address these questions, we performed a pooled study of two large prospective multicenter study cohorts to compare the clinical characteristics and outcomes of newly diagnosed FL3B with FL1-2, FL3A and DLBCL patients, all of whom were uniformly treated with frontline R-CHOP. Patients and Methods: Weevaluated 469 newly diagnosed patients with FL1-2 (n=216), FL3A (n=170) and FL3B (n=83) who received R-CHOP within 6 months of diagnosis and 739 de novo DLBCL (GCB n=455 and non-GCB n=284) who were prospectively enrolled into the Molecular Epidemiology Resource (MER) and the Lymphoma Epidemiology of Outcomes (LEO) cohorts between 2002 and 2021. FL3B cases were designated as FL3Bc and FL3Bp based on the presence or absence of coexisting DLBCL, respectively. Histologic diagnosis was determined by an expert pathologist at each participating center and centrally reviewed. Event free survival (EFS) and OS were defined as time from diagnosis to progression, relapse, retreatment or death, and early progression defined as failing to achieve EFS at 24 months (EFS24). Associations between clinical features and EFS or OS were evaluated using Kaplan Meier curves and Cox proportional hazards models unadjusted and adjusted for age and FLIPI (comparisons with FL1-2 and FL3A) or IPI (comparisons with DLBCL). Results: The median age was 61 years (range, 18-94 years). Among FL3B patients, 21 (25%) had FL3Bc and 62 (75%) FL3Bp. Baseline characteristics and outcomes were broadly similar between the two FL3B subtypes. We then compared all FL3B patients to those with lower-grade FL. Patients with FL3B showed similar clinical features to those with FL1-2 and FL3A, with exception for a lower tumor burden. After R-CHOP, FL1-2 patients had an inferior EFS (HR 2.13 95%CI 1.31-3.46) than those with FL3B, whereas there was no difference compared to FL3A. OS was similar across all FL grades. Although FL1-2 patients failed to achieve EFS24 more frequently than FL3B and FL3A (34% vs 18% and 15%, p<0.001), FL3B patients who did fail EFS24 had almost three folds higher risk of subsequent mortality (HR 2.84 95%CI 1.35-6.01) than other FLs. At 5-year follow-up, FL3B patients had twice the risk of relapse with an aggressive subtype (i.e., DLBCL, high-grade B-cell lymphoma) than those with FL1-2 and FL3A, who instead often recurred with indolent disease (i.e., FL1-2, FL3A; p<0.001). We further compared FL3B patients to those with DLBCL classified based on cell-of-origin into GCB and non-GCB subtypes. At diagnosis, patients with FL3B had more favorable clinical features than those with GCB and non-GCB DLBCL. Compared to FL3B, patients with non-GCB DLBCL had an inferior EFS (HR 1.73 95%CI 1.07-2.79) and OS (HR 2.04 95%CI 1.11-3.75), while those with GCB DLBCL did not show a significant difference in EFS (HR 1.41 95%CI 0.88-2.25) or OS (HR 1.69, 95%CI 0.93-3.06) likely due to insufficient statistical power. Patients with FL3B failed EFS24 less frequently than patients with GCB or non-GCB DLBCL (17% vs 25% and 32%; p=0.01). Among patients who failed EFS24, those with FL3B had a similar subsequent OS as both GCB (HR 1.41 95%CI 0.73-2.72) and non-GCB DLBCL (HR 1.27 95%CI 0.65-2.47). Conclusions: We report that FL3Bp and FL3Bc appear clinically similar. While most FL3B patients have a prognosis comparable to FL1-2 and FL3A, 18% exhibit aggressive behavior with poor outcome reminiscent of DLBCL. Future studies with a larger population will be needed to confirm our findings.

Background: A donor search prognosis score based on HLA type and race/ethnicity can classify patients into 3 groups based on the likelihood of finding an HLA-matched (8/8) unrelated donor (MUD): Very Likely (44% of the population), Likely (41%) and Very Unlikely (15%). Previous studies showed that >90% of Very Likely, 26% of Likely and <10% of Very Unlikely patients used a MUD for their hematopoietic cell transplants (HCT), with the rest using other donor sources (mismatched unrelated donor, cord blood, or haploidentical relative). For some patients, prolonged efforts to identify a MUD might be futile and delay HCT, leading to increased morbidity, greater chance for relapse and disease resistance, and worse HCT outcomes. BMT CTN 1702 (NCT03904134) was the first national study to track the entire donor search identification process. Designed as a biologic assignment study, the goal was to estimate and compare overall survival between two arms: patients who are Very Likely to find a MUD versus those who are Very Unlikely to find a MUD. It tested whether provision of a donor search prognosis score to centers could prevent prolonged futile searches for the Very Unlikely group and improve their transplant rates and outcomes, compared to the Very Likely group. Methods: Eligible diagnoses included acute leukemia, myelodysplastic syndromes, lymphoma, acquired aplastic anemia and sickle cell disease. Patients could be consented at any time after diagnosis until an unrelated donor search was started. Consented patients became evaluable only after it was determined that no appropriate HLA-identical siblings were available. The primary endpoint was survival time after the patient was declared evaluable, regardless of whether HCT occurred, and was compared between the Very Unlikely and Very Likely groups. The target enrollment was 1732 evaluable patients. Because this was a biologic assignment trial and not a randomized trial, formal comparisons of overall survival between the Very Likely and Very Unlikely groups were done using a covariate-adjusted Cox proportional hazards model. The following covariates were a priori specified for adjustment in the model, according to the statistical analysis plan: sex, age, Karnofsky Performance Score (KPS), race/ethnicity, disease, disease status, and interval from consent to evaluability. Results: 1751 evaluable patients were enrolled at 47 centers: 44% female, median age 59.4 years (range 0.6-81.3 years), 73% were non-Hispanic White, 10% Hispanic White, 8% Black, 5% Asian/Native Hawaiian/Pacific Islander, 0.5% American Indian/Alaska Native. Among the evaluable patients, 1181 received a HCT, 827 died, and 924 are still alive with or without HCT, with a median of 25 months of follow up (range=6-54 months [mos], inter-quartile range [IQR]=25-34 mos). The distribution was: 958 Very Likely, 517 Less Likely, and 276 Very Unlikely. Individuals in the Very Unlikely group were younger and were less likely to be non-Hispanic White. The unadjusted cumulative incidences of undergoing HCT by 6 mos and 2 years [yrs], in the Very Unlikely group were 51% and 61% vs. 60% and 69% in the Very Likely Group; after adjustment, there was not a significant difference in the likelihood of undergoing HCT (Dehn et al, submitted). There was no difference in median time from evaluability to HCT: 3.4 mos Very Unlikely and 3.3 mos Very Likely. Unadjusted survival rates from the time of evaluability for the Very Unlikely and the Very Likely groups were: 6 mos (85% vs 84%), 1 yr (69% vs. 69%) and 2 yrs (55% vs. 56%). In the multivariate Cox model, there was no significant difference in the risk of death (HR = 1.07 for Very Unlikely vs. Very Likely, 95% confidence interval 0.86-1.33, p=0.56) adjusting for significant clinical covariates: age, KPS, disease status, and time from consent to evaluability were all significantly associated with overall survival. There was no significant center effect and no interaction between donor search prognosis group and clinical covariates. Conclusion: In this large multicenter BMT CTN study, using the donor selection strategy described, there was no difference in adjusted time to HCT or survival for patients Very Likely versus Very Unlikely to find an 8/8 MUD. These results suggest that patients Very Unlikely to identify a MUD should proceed to HCT using the best available alternative donor rather than delaying HCT to try to identify a MUD.