Khera, Nandita

Abstract Background There has been an increase in overall utilization as well as improvement in overall survival (OS) after hematopoietic cell transplantation (HCT) for most hematologic disorders over time. It is not known if these changes have impacted racial/ethnic minorities equally. In this study, we describe changes in the volume and rate of autologous (auto) and allogeneic (allo) HCT and examine trends in survival in different racial/ethnic groups from 2009 to 2018. Methods Data were obtained from the Center for International Blood and Marrow Transplant Research (CIBMTR) which collects patient data on most autologous and allogeneic HCT recipients in the US. We compared the number of auto and allo transplants in 4 different racial/ethnic groups: Non-Hispanic Whites (NHWs), African Americans (AAs), Hispanics and 'Others' across five 2-year cohorts from 2009 to 2018. 'Others' included Asians, Pacific Islanders and Native Americans as well as non-US residents. Since numbers in individual categories were too low to be meaningful, they were grouped together as 'Others'. Multi race individuals and those with either race or ethnicity missing were excluded. Cox proportional hazards models were used to examine trends in overall mortality after auto HCT in adult patients (for NHL, HD, and MM), and allo HCT among adult and pediatric patients (for AML, ALL, lymphoma including chronic lymphocytic leukemia, and MDS/MPD) in the 4 different racial ethnic groups over 5 study time periods. These models were adjusted for sociodemographic and clinical variables (age, sex, Karnofsky/Lansky score, HCT-comorbidity index, disease, disease status, geographic region, insurance, marital status, education, distance to the transplant center, median income derived from patient's ZIP code). Graft type, donor, conditioning, GVHD prophylaxis, donor-recipient CMV match and donor-recipient sex match were included for allo HCT models. Results A total of 80,080 (59,666 NHWs, 10,683 AAs, 7,110 Hispanics) patients received an auto HCT and 60,412 (44,272 NHWs, 4,964 AAs, 7,826 Hispanics) received an allo HCTs from 2009 to 2018. The volume of auto and allo transplants increased overall, as well as within each race/ethnicity group, over the study periods. In addition, there was a significant change in the proportion of patients in each racial/ethnic group undergoing an auto HCT, with a lower proportion of NHWs and increased proportion of minorities over time (from 78%, 12% and 8% in 2009 -2010 to 71%, 15% and 10% in 2017-2018 for NHWs, AAs and Hispanics respectively; p<0.001). Similar trends were observed for allo HCT: from 76%, 8% and 12% in 2009-2010 to 71%, 9% and 14% in 2017-2018 of all allo HCTs for NHWs, AAs and Hispanics respectively; p<0.001. While there was an increase in the rate of transplants over the years across all racial/ethnic groups, the rate was higher in AAs and Hispanics as compared to NHWs (Table 1). There was no significant difference in overall mortality amongst the three main racial ethnic groups, though there was decrease in mortality over time for auto HCT and adult allo HCT patients. In the allo HCT pediatric group, AAs had higher overall mortality as compared to NHWs across all time periods. (Table 2) Figure 1 shows temporal trends in two-year survival by race and ethnicity. Conclusions This large population-based study shows an improvement in the rates and survival of auto and allo HCT in adults over time. Changes in the proportion of different racial/ethnic groups undergoing HCT are higher than rate of change of population growth (3% decline for NHWs, no change in AAs and 2% increase in Hispanics from 2010 to 2019 per William H. Frey analysis of census population estimates released June 25, 2020). Progress is reflected in higher rates of increase in racial/ ethnic minority groups and comparable survival across the racial ethnic groups in our cohort as compared to two prior registry studies (Baker et al. JCO 2005 and Baker et al. BBMT 2009) for most HCT patients. However, in pediatric patients undergoing allo HCT, the gap in survival between NHWs and AAs has continued to persist over the years. With increasing diversity of the US population, we need to strengthen our efforts for improving access to, and outcomes of HCT to narrow the gap for all patients requiring HCT, irrespective of their sociodemographic characteristics. Figure 1 Figure 1. Disclosures Wood: Koneksa Health: Consultancy, Current equity holder in publicly-traded company; Teladoc: Consultancy; Pfizer: Research Funding. Ailawadhi: Karyopharm: Consultancy; AbbVie: Consultancy; Genentech: Consultancy; Takeda: Consultancy; GSK: Consultancy, Research Funding; Xencor: Research Funding; Cellectar: Research Funding; Medimmune: Research Funding; Ascentage: Research Funding; Pharmacyclics: Consultancy, Research Funding; Amgen: Consultancy, Research Funding; Janssen: Consultancy, Research Funding; Bristol Myers Squibb: Consultancy, Research Funding; BeiGene, Ltd.: Consultancy; Sanofi: Consultancy; Oncopeptides: Consultancy. Saber: Govt. COI: Other.

Abstract Background: Despite the promise of targeted agents in hematology, the high price of cancer therapies is a rapidly evolving problem. Until now, the increase in price of targeted anti-cancer treatments used in common hematological malignancies has not been evaluated. Here, we report patterns in price changes from 2015-2019 for multiple in-class anticancer medications for common hematologic malignancies. Methods: We utilized the publicly available Medicare Part D provider utilization and payment database from 2015 to 2019. We extracted drug prices (using generic names) for commonly used targeted anticancer agents for acute myeloid leukemia, chronic myeloid leukemia, chronic lymphocytic leukemia, B-cell lymphomas, and multiple myeloma. The primary outcome was the correlation of average change in Medicare spending per dosage unit among the multiple brand-name medications within each class available. We additionally calculated compound annual growth rates (CAGRs) [i.e. mean annual growth rates over a specified period of time] for medication costs within each class, and compared it with the consumer price index (CPI) (a measure of the average change over time in the prices of consumer items) and inflation rate. Agents without in-class competitor were excluded. CPI and inflation rates came from U.S. Bureau of Labor Statistics. Results: The study included 6 BCR-ABL inhibitors (1 generic), 2 BTK inhibitors, 2 IDH inhibitors, 2 FLT3 inhibitors, 3 IMIDs, 3 PI, 2 PI3K inhibitors, and 5 anti-CD20 monoclonal antibodies. The median (range) Pearson correlation coefficient values for drugs within each class were -0.155 (-0.984-0.992) for BCR-ABLi, -0.362 for BTKi, 0.234 (0.215-0.999) for IMIDs, 0.899 (-0.034-0.988) for PIs, and 0.954 (0.847-0.999) for anti-CD20 antibodies. The median correlation coefficient for BCR-ABLi was 0.751 if generic imatinib was excluded. The correlation coefficient between 2 nd and 3 rd generation IMIDs was 0.999. Non-generic BCR-ABLi and anti-CD20 antibodies showed strong linear association in price increase between two drugs within the same class. Due to drug novelty, coefficient could not be calculated for therapies with 2 or fewer data points (midostaurin, gilteritinib, enasidenib, ivosidenib, idelalisib, duvelisib). There was no significant correlation between expenditure for BTKi, PIs, and IMIDs. The median CAGRs in costs over this 5-year period were: were 6.29% for BCR-ABLi, 18.36% for BTKi, 2.69% for IDHi, 4.23% for FLT3i, 10.63% for IMIDs, 5.11% for PIs, 5.79% for PI3Ki, and 5.85% for anti-CD20s. The median CAGR in costs for modern precision-driven cancer therapeutic classes outpaced CPI (2.26%/year), and the average inflation rate (1.90%/year). Conclusions: Increase in cost within the same class should be weighed against incremental clinical benefit for the patients. For non-generic BCR-ABLi and anti-CD20 antibodies, despite there being several agents, the rise in drug expenditures correlated closely, calling into question the true value of within-class competition. There is an urgent need for drug pricing reform given the average expenditure of Medicare part D, and ultimately out-of-pocket costs for our patients with cancer continues to trend upwards. Increased advocacy efforts are needed to ensure precision therapeutics remains an attainable and sustainable goal. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

Abstract Background: Self-reported financial hardship (FH) amongst cancer patients is increasingly becoming a challenge for patients, caregivers, and healthcare providers. FH not only leads to financial struggles, significant lifestyle changes, and emotional distress, but also contributes to treatment noncompliance, affecting clinical outcomes. As treatment costs rise, it is crucial to develop efficient methods to proactively identify and alleviate FH in hematology practice. One potential approach is utilizing automated processes to identify those at highest risk of FH. At Mayo Clinic, screening for FH involves using a single financial strain question 'How hard is it for you to pay for the very basics like food, housing, medical care, and heating?' which all cancer patients answer annually as part of the institution's Social Determinants of Health (SDOH) assessment. Answers are on a five-point scale including not hard at all, not very hard, somewhat hard, hard, and very hard. In this study, we assess the prevalence and predictors for FH (denoted by a response of "Very hard" "Hard" or "Somewhat hard") amongst the Mayo Clinic hematologic malignancy patient population. Our study objective was to determine if this automated process could identify those at risk for FH. Methods: Patients who received care for hematologic malignancies (lymphoma, leukemia, plasma disorders, myelodysplastic/myeloproliferative disorders, and other heme malignancies) at any of the Mayo Clinic cancer centers (Minnesota, Arizona, and Florida) and who had completed the SDOH screen at least once were included in this study. The electronic medical record (EMR) and Mayo Clinic Cancer Registry were utilized to extract demographic and disease variables. Patient's home zip code was used to determine rural/urban residence, distance from cancer center, and the Area Deprivation Index (ADI), a measure of socioeconomic disadvantage based on home zip code (ranging from 1-100, with 100 representing the most disadvantaged). Multivariable logistic regression modeling was used to examine predictor variables for FH in this patient population. Results: The final cohort included 10,024 patients from 2018 to 2020. Median age was 64.6 years (IQR 58.1,73.7), 58% were male, and 79% married. Race/ethnicity composition was 94% White (n=9,268), 2.5% Black (n=246), 0.4% American Indian/Alaskan Native (44), and 4% Hispanic (n=360). Fifty-six percent of patients had Medicare and 41% had commercial insurance. Fifty percent were retired, 40% were working/students, and 72% were urban residents. Mean ADI was 41.2. Fifty-six percent of patients had lymphomas, 23.5% had plasma cell disorders, 8.5% had leukemias, 6.8% had other hematological malignancies, and 5.5% had myelodysplastic/myeloproliferative neoplasms. FH was reported by 12.8% (n=1286) of the patients. Table 1 shows the results of the multivariable model. A significantly higher likelihood of endorsing FH was noted in Hispanic vs non-Hispanics, Black and American Indian/Alaskan Native groups vs whites, Disabled/Unemployed vs working, Medicaid, Medicare, and Self-Pay groups vs commercial insurance, higher ADI (5 th quintile vs 1 st), and myelodysplastic/myeloproliferative disorder and other hematologic malignancy vs lymphoma patients. Older age, being retired, and living farther from the cancer center were associated with significantly less likelihood of endorsing FH. Conclusion: Our study used automated data extraction from the EMR to efficiently identify predictors of FH in hematologic cancer patients. Employing a dichotomized and automated "flag" for FH, particularly if incorporated in the EMR, could ease the identification of SDOH issues, facilitate timely connection to appropriate resources, and help provide better patient-centered care. Figure 1 Figure 1. Disclosures Ailawadhi: Sanofi: Consultancy; Cellectar: Research Funding; Karyopharm: Consultancy; Ascentage: Research Funding; Genentech: Consultancy; Janssen: Consultancy, Research Funding; BMS: Consultancy, Research Funding; Beigene: Consultancy; GSK: Consultancy, Research Funding; AbbVie: Consultancy; Medimmune: Research Funding; Pharmacyclics: Consultancy, Research Funding; Takeda: Consultancy; Amgen: Consultancy, Research Funding; Xencor: Research Funding. Fonseca: OncoTracker: Consultancy, Membership on an entity's Board of Directors or advisory committees; Takeda: Consultancy; BMS: Consultancy; Mayo Clinic in Arizona: Current Employment; Aduro: Consultancy; AbbVie: Consultancy; GSK: Consultancy; Merck: Consultancy; Juno: Consultancy; Scientific Advisory Board: Adaptive Biotechnologies: Membership on an entity's Board of Directors or advisory committees; Patent: Prognosticaton of myeloma via FISH: Patents & Royalties; Novartis: Consultancy; Bayer: Consultancy; Celgene: Consultancy; Caris Life Sciences: Membership on an entity's Board of Directors or advisory committees; Kite: Consultancy; Janssen: Consultancy; Amgen: Consultancy; Pharmacyclics: Consultancy; Sanofi: Consultancy. Griffin: Exact Sciences: Research Funding.

Abstract Introduction: Allogeneic stem cell transplantation (allo-SCT) can be curative for patients with hematologic malignancies. Despite optimization of supportive measures, the risk of non-relapse mortality is a particularly important consideration in older individuals. Sarcopenia, defined as low skeletal muscle mass, has been shown to be associated with decreased overall survival in cancer patients and may identify individuals at higher risk of treatment-related complications. Methods: Using Mayo Clinic Arizona's hematopoietic stem cell transplantation database, patients aged 60 years or older who underwent allo-SCT between 2008 and 2018 were identified. Patients who underwent abdominal computed tomography (CT) at our institution within 90 days prior to their transplant date were included in the study. The skeletal muscle area (SMA) was obtained using a fully automated algorithm for segmenting abdominal CT images to quantify body composition based on a deep learning approach (Weston et al., Radiology, 2019). The skeletal muscle index (SMI) was then calculated using the SMA and the subject's height at the time the CT was obtained. Sarcopenia was defined based on previously described consensus (Cruz-Jentoft et al., Age Ageing, 2010; Magudia et al., Radiology, 2021). T-tests and Fisher's exact tests compared patient characteristics between groups, where appropriate. Kaplan-Meier methods were used to construct survival times. Log-rank tests compared survival distributions and Cox proportional hazards models estimated hazard ratios. Results: A total of 79 patients met inclusion criteria (Table 1). Of these, 50 (63%, n=50/79) had sarcopenia, and 29 (37%, n=29/79) had normal skeletal muscle mass. Among the sarcopenia group, 43 (86%, n=43/50) had SMI measurements. No differences were seen between groups in age at transplant, sex, Karnofsky performance scale, disease risk index, or conditioning regimen intent. There was no statistically significant difference in overall survival between those with sarcopenia and those with normal skeletal muscle mass (log-rank p=0.18). Median survival was 3.3 years (95% CI: 0.8, not estimable (NE)) for those with sarcopenia and 3.6 years (95% CI: 1.3, NE) for those with normal skeletal muscle mass. Given the infrequency of events, the upper bounds of the median survival confidence intervals were not estimable. The unadjusted hazard ratio (HR) for sarcopenia versus normal skeletal muscle mass was 1.65 (95% CI: 0.79, 3.47; p=0.18). Males had a higher mean SMI than females, as expected (50.4 versus 37.5, p<0.001; mean difference 13.0, 95% CI: 8.9, 17.1). Without considering sex, higher or lower SMI was not associated with survival time in a Cox model in those with sarcopenia (HR: 0.97, 95% CI: 0.92, 1.03; p=0.32). Sex alone was also not associated with any survival benefit among these 43 patients. After adjusting for the difference seen in SMI between sexes, on average, those with higher SMI were associated with a survival benefit (HR: 0.92, 95% CI: 0.84, 0.99; p=0.039). This amounts to roughly an 8% reduction in hazard rate as SMI increased. Conclusions: The presence of sarcopenia alone was not associated with overall survival in patients 60 years and older undergoing allo-SCT for hematologic malignancy. Rating sarcopenia severity using SMI may assist in identifying high risk individuals with sarcopenia as lower sarcopenia severity was associated with improved overall survival. Further studies are warranted to explore the possible impacts of interventions directed at improving skeletal muscle mass on transplant-related outcomes. Figure 1 Figure 1. Disclosures Gaulin: DeciBio: Honoraria; Olson Research Group: Honoraria. Jain: CareDx: Other: for advisory board participation; CTI Biopharma: Research Funding; Syneos Health: Research Funding; Bristol Myers Squibb: Other: for advisory board participation; Targeted Healthcare Communications: Consultancy. Palmer: PharmaEssentia: Research Funding; Incyte: Research Funding; Sierra Oncology: Consultancy, Research Funding; CTI BioPharma: Consultancy, Research Funding; Protagonist: Consultancy, Research Funding.

Abstract Introduction: Allogeneic hematopoietic cell transplantation (alloHCT) is the only potentially curative treatment available for acute myelogenous leukemia (AML). It is a medically complicated and resource intensive procedure for which there are patient-, provider- and system-related factors that may impact its utilization. While Medicare coverage for alloHCT may help address a major financial barrier to access, there is still an unmet need due to other factors. This study intended to examine trends and factors associated with utilization of alloHCT and to estimate unmet need for alloHCT among Medicare beneficiaries with AML. Methods: This retrospective cohort study included all patients with a diagnosis of AML identified in the Medicare claims data from 2010 through 2016. Primary patient selection criteria included: primary or secondary diagnosis of AML, age limit (65-74), and continuous enrollment (for at least 180 days after AML diagnosis in Part A and Part B fee-for-service programs). To study trends in utilization, the transplant rates were calculated as the number of patients who received an alloHCT within 180 days and one year of diagnosis (numerator) divided by the total number of patients diagnosed with AML within each diagnosis year (denominator). Transplant rates within one year of diagnosis were further adjusted by patient characteristics, including age group, sex, race, residential region, and Elixhauser Comorbidity Index (ECI). A multivariable logistic regression was utilized to identify factors associated with the likelihood of receiving alloHCT within one year of diagnosis. Two approaches were applied to estimate unmet need for alloHCT. The first approach (Approach 1) used claims data to identify the potential need for alloHCT among patients who achieved complete remission for at least 90 days; patients who achieved 90-day remission but did not receive alloHCT at any time point were considered to have unmet need. In the second approach (Approach 2), the total number of patients diagnosed with AML in the claims data was run through the National Marrow Donor Program (NMDP) methodology , which takes estimates of risk level, response to treatment, comorbidity, and early mortality into consideration. Overall estimated need and unmet need from 2010-2015 and over different time periods were evaluated for both approaches. Results: Among the 5,974 patients diagnosed with AML from March 1, 2010 through June 30 th, 2016, 1226 patients (21%) received an alloHCT by the end of 2016. Trend analysis results suggest that utilization of alloHCT increased from 2010 to 2016 (P < 0.0001) (Figure 1). The likelihood of receiving alloHCT within one year of diagnosis was found to be associated with diagnosis year, age, race, geographic region, ECI, and population-level median household income (Table 1). Both approaches estimated that approximately 36% of the diagnosed patients were in need of alloHCT between 2010 and 2015. The overall unmet need was estimated as 59% and 43% based on the claims data approach and the NMDP methodology, respectively. Despite the differences in estimated unmet need between the two approaches, the unmet need for alloHCT was found to trend down over time (Figures 2 & 3). Discussion and Conclusions: Medicare coverage facilitates access to the health care system and receipt of health services when a need for treatment or care is recognized. Utilization of alloHCT has increased over time among Medicare beneficiaries with AML. However, there are persistent differences in utilization of alloHCT by age, race, geographic region, comorbidity, and socioeconomic status, indicating disparities in access to alloHCT among this population. The results of estimated need and unmet need for alloHCT may be affected by the lack of cytogenetics and molecular information in administrative claims. To minimize this limitation, we used NMDP methodology for validation and found a similar estimation of potential need for alloHCT and downward trend in unmet need. Although the unmet need for alloHCT for AML has improved over time, policy efforts, research, and continued education are needed to close the gap between the actual utilization of alloHCT and unmet need for this potentially curative treatment. Figure 1 Figure 1. Disclosures Saber: Govt. COI: Other.

7 Background: Despite the promise of precision oncology, the cost-effectiveness of targeted treatments is debated. Until now, the increase in price of oral targeted anti-cancer treatments used in common malignancies has not been evaluated. Here, we report patterns in price changes from 2015-2019 for multiple oral anti-cancer medications for common solid tumor malignancies. Methods: We utilized the publicly available Medicare Part D provider utilization and payment database from 2015 to 2019. We extracted drug prices (using generic names) for commonly used targeted oral anticancer agents for lung, breast, and prostate cancer. The primary outcome was the correlation of average change in Medicare spending per dosage unit among the multiple brand-name medications within each class available. We additionally calculated compound annual growth rates (CAGRs) [i.e. mean annual growth rates over a specified period of time] for medication costs within each class, and compared it with the consumer price index (a measure of the average change over time in the prices of consumer items). Results: The study included 6 EGFR inhibitors (1 generic), 5 ALK inhibitors, 2 BRAF inhibitors, 3 hormonal agents (1 generic), 3 CDK4/6 inhibitors, 2 PARP inhibitors, and 7 anti-androgen agents (2 generic). The median (range) Pearson correlation coefficient values for drugs within each class were 0.967 (0.915-0.978) for EGFRi, 0.981 (0.966-0.989) for ALKi, 0.996 for BRAFi, 0.994 (0.992-0.999) for CDK4/6i, 0.855 for PARPi, and 0.442 (-0.522-0.962) for anti-androgens. Except for anti-androgens, all other drug classes showed strong linear association in price increase between two drugs within the same-class. A coefficient could not be calculated for therapies with 2 or fewer data points ( i.e., generic erlotinib, dacomitinib, generic abiraterone, apalutamide, and darolutamide). There was no significant correlation between expenditure for anti-estrogen agents. The median (range) CAGRs in costs over this 5-year period were: were 4.56% for EGFRi, 6.40% for ALKi, 2.58% for BRAFi, 5.48% for hormonal agents, 5.21% for CDK4/6i, 27.29% for PARPi, and 34.8% for anti-androgen agents. Conclusions: The median CAGR in costs for modern oral precision driven cancer therapeutic classes mostly outpaced CPI (2.26%/year), and the average inflation rate (1.90%/year). Increase in cost within the same class should be weighed against incremental clinical benefit for the patients. For most classes despite there being multiple agents, the rise in drug expenditures correlated closely, calling into question the true value of within class competition. There is an urgent need for drug pricing reform given the average expenditure of Medicare part D, and ultimately out of pocket costs for our patients with cancer continues to trend upwards. Increased advocacy efforts are needed to ensure precision therapeutics remains an attainable and sustainable goal.

Lay Summary People who have advanced myelodysplastic syndromes (MDS) may live longer if they get a bone marrow transplant (BMT) instead of other therapies. However, only 15% of people with MDS actually get BMT. Experts say community physicians and transplant physicians should team up with insurance companies and patient advocacy groups to 1) spread this news about lifesaving advances in BMT, 2) ensure that everyone can afford health care, 3) provide emotional support for patients and families, 4) help patients and families get transportation and housing if they need to travel for transplant, and 5) improve care for people of under‐represented racial and ethnic backgrounds.

Abstract There is a lack of large comparative study on the outcomes of reduced intensity conditioning (RIC) in acute myeloid leukemia (AML) transplantation using fludarabine/busulfan (FB) and fludarabine/melphalan (FM) regimens. Adult AML patients from Center for International Blood and Marrow Transplant Research who received first RIC allo-transplant between 2001 and 2015 were studied. Patients were excluded if they received cord blood or identical twin transplant, total body irradiation in conditioning, or graft-versus-host disease (GVHD) prophylaxis with in vitro T-cell depletion. Primary outcome was overall survival (OS), secondary end points were leukemia-free survival (LFS), nonrelapse mortality (NRM), relapse, and GVHD. Multivariate survival model was used with adjustment for patient, leukemia, and transplant-related factors. A total of 622 patients received FM and 791 received FB RIC. Compared with FB, the FM group had fewer transplant in complete remission (CR), fewer matched sibling donors, and less usage of anti-thymocyte globulin or alemtuzumab. More patients in the FM group received marrow grafts and had transplantation before 2005. OS was significantly lower within the first 3 months posttransplant in the FM group (hazard ratio [HR] = 1.82, P < .001), but was marginally superior beyond 3 months (HR = 0.87, P = .05). LFS was better with FM compared with FB (HR = 0.89, P = .05). NRM was significantly increased in the FM group during the first 3 months of posttransplant (HR = 3.85, P < .001). Long-term relapse was lower with FM (HR = 0.65, P < .001). Analysis restricted to patients with CR showed comparable results. In conclusion, compared with FB, the FM RIC showed a marginally superior long-term OS and LFS and a lower relapse rate. A lower OS early posttransplant within 3 months was largely the result of a higher early NRM.

e19023 Background: Acute GVHD (aGVHD) is a major cause of mortality and morbidity following stem cell transplantation (HCT). In effort to prognostic patients with aGVHD, the Mount Sinai Acute GVHD International Consortium (MAGIC) combined two biomarkers, ST2 and REG3α to generate an algorithm known as the Ann Arbor (AA) biomarker risk. This is now available as a commercial test to help predict risk for severe acute GVHD and non-relapse mortality when done at the onset of symptomatic aGVHD. Our institution began utilizing these lab values and here we report our “real world” experience. The primary aim was to correlate AA risk probabilities with recently defined endpoints including day 28 overall response rate (ORR) and 6 month freedom from treatment failure (6mFFTF) and overall survival (OS). Methods: 58 HCT patients who underwent HCT between October 2018 and November 2019 and had the biomarker assay at the onset of aGVHD were identified from the institutional database and clinical information about disease, transplant and outcomes was obtained from the medical records. Clinical staging of aGVHD was documented per standardized published guidelines. Day 28 ORR was defined as complete response by day 28. 6mFFTF was defined as a patient being alive, without relapse of underlying disease or addition of new systemic therapy for aGVHD, and prior to the development of cGVHD. Results: When comparing low (n = 39), intermediate (n = 11), and high (n = 8) AA risk, there was no difference in age ( p =0.254), primary disease ( p =0.10), DRI ( p =0.994), donor source ( p =0.193), conditioning regimen intent ( p =0.537), GVHD prophylaxis ( p =0.351) ATG use ( p =0.06). Low AA risk was associated with lower grades of aGVHD at onset (grade I = 15%, grade II = 80%, grade III = 5%, grade IV = 0%), whereas high AA risk was associated with higher grades of aGVHD at onset (grade I = 0%, grade II = 50%, grade III = 25%, grade IV = 25%) ( p =0.009). For low, intermediate, and high AA risk the day 28 ORR was 69%, 45% and 29% ( p = 0.083); the 6mFFTF was 82%, 82%, and 25% ( p = 0.0001); 6 month OS was 92%, 100%, and 50% ( p = 0.0006) respectively. Conclusions: Our study shows a high correlation between AA risk and outcomes. High risk AA risk was associated with inferior outcomes This study was limited by the retrospective single institutional experience with relatively small numbers. However, our findings using the commercial assay mirror that seen with the MAGIC studies. Our “real world data” can serve as a foundation and continued validation for institutional implementation of AA risk stratification for aGVHD.