Abstract: Older patients with acute myeloid leukemia (AML) have high relapse risk and poor survival after allogeneic hematopoietic cell transplantation (HCT). Younger patients may receive myeloablative conditioning to mitigate relapse risk associated with high-risk genetics or measurable residual disease (MRD), but older adults typically receive reduced-intensity conditioning (RIC) to limit toxicity. To identify factors that drive HCT outcomes in older patients, we performed targeted mutational analysis (variant allele fraction ≥2%) on diagnostic samples from 295 patients with AML aged ≥60 years who underwent HCT in first complete remission, 91% of whom received RIC, and targeted duplex sequencing at remission in a subset comprising 192 patients. In a multivariable model for leukemia-free survival (LFS) including baseline genetic and clinical variables, we defined patients with low (3-year LFS, 85%), intermediate (55%), high (35%), and very high (7%) risk. Before HCT, 79.7% of patients had persistent baseline mutations, including 18.3% with only DNMT3A or TET2 (DT) mutations and 61.4% with other mutations (MRD positive). In univariable analysis, MRD positivity was associated with increased relapse and inferior LFS, compared with DT and MRD-negative mutations. However, in a multivariable model accounting for baseline risk, MRD positivity had no independent impact on LFS, most likely because of its significant association with diagnostic genetic characteristics, including MDS-associated gene mutations, TP53 mutations, and high-risk karyotype. In summary, molecular associations with MRD positivity and transplant outcomes in older patients with AML are driven primarily by baseline genetics, not by mutations present in remission. In this group of patients, where high-intensity conditioning carries substantial risk of toxicity, alternative approaches to mitigating MRD-associated relapse risk are needed.

Abstract: Background: Clonal hematopoiesis (CH) is an age-related condition in which somatic mutations can be detected in the blood of healthy individuals. In the non-transplant setting, CH is associated with an elevated risk of developing hematologic malignancy and an increased risk of non-hematologic outcomes due to altered inflammatory signaling. During hematopoietic cell transplantation (HCT), CH in older donors can engraft in recipients and could therefore influence outcomes through effects on graft immunologic function or by causing donor cell leukemia. A definitive link between donor CH and recipient outcomes has not been established. We therefore evaluated the impact of CH in donors aged 40 years or older on recipient clinical outcomes in 1727 donor-recipient pairs. Methods: We performed targeted, error-suppressed sequencing of 46 genes on 1727 samples from donors age 40 and older. We defined CH as pathogenic mutations at variant allele fraction (VAF) 0.005 or greater. Median donor age was 51 (range 40-80) and median recipient age was 55 (range 1-78). There were 889 matched related donors (51.5%), 454 haploidentical donors (26.3%), 273 matched unrelated donors (15.8%), 71 mismatched unrelated donors (4.1%), and 38 mismatched related donors (2.2%). 929 recipients (53.8%) had myeloid malignancies, 718 (41.6%) had lymphoid malignancies, and 80 (4.6%) had non-malignant diseases. 1022 (59.2%) recipients received peripheral blood stem cell products and 703 (40.7%) received bone marrow. 672 recipients (38.9%) received post-transplant cyclophosphamide (PTCy) for graft versus host disease (GVHD) prophylaxis. Median follow-up for survivors was 6.0 years. Results: We identified CH in 388 of 1727 (22.5%) donor samples. Mutations in DNMT3A were most common (302 mutations in 253 donors), followed by TET2 (96 mutations in 89 donors), ASXL1 (n=22) and PPM1D (n=14). No other genes were mutated in more than 10 donors (0.5%). The presence of donor CH was independently associated with older donor age, but not with donor sex, donor-recipient relatedness, graft source, or recipient disease. The presence of donor CH at VAF 0.01 or greater was associated with improved progression-free survival (PFS) in a multivariable model that included donor and recipient age, HCT-CI, disease category, Disease Risk Index score, conditioning intensity, and donor type (HR for death or relapse 0.81, 95% CI 0.68-0.97, P=0.019). This effect was driven by DNMT3A mutations, which were independently associated with improved overall survival (HR for death 0.75; 0.59-0.95, P=0.016) and reduced risk of relapse (sHR 0.76; 0.59-0.97, P=0.029) in the same model. CH involving other gene mutations, including TET2 and genes other than DNMT3A/TET2, was not significantly associated with any recipient outcome. Smaller clones (VAF 0.005-0.01) had no effect on any outcome. The association between donor DNMT3A-CH and recipient outcomes was limited to those who did not receive PTCy for GVHD prophylaxis (Figure 1A-C). In the model described above, donor DNMT3A-CH in the absence of PTCy was independently associated with improved PFS (HR 0.61; 0.45-0.84, P=0.002), reduced risk of relapse (sHR 0.59; 0.39-0.9, P=0.014) and an elevated risk of chronic GVHD (sHR 1.36; 1.01-1.83, P=0.042) compared to those without DNMT3A-CH. In recipients who received PTCy, there was no significant effect of donor DNMT3A-CH on PFS, relapse, or cGVHD (1D). Eight recipients developed donor cell leukemia (DCL), for a cumulative incidence of 0.7% at 10 years. In seven of these cases, we identified gene mutations in the corresponding donor products that matched the mutations found in the subsequent DCL, including 2 with TP53 mutations, 3 with splicing factor mutations, and 2 with germline DDX41 mutations that were present in both donor and recipient. No recipients who received products with sole DNMT3A-CH developed DCL. Conclusions: In HCT donors age 40 or older, the presence of DNMT3A clonal hematopoiesis at VAF & gt;/= 0.01 is independently associated with prolonged overall and progression-free survival in transplant recipients. This effect is driven by reduced risk of disease relapse and confined to recipients who do not receive PTCy, suggesting that it is mediated at least in part by effects on donor T cells. The risk of direct evolution of DNMT3A-CH to donor cell leukemia is low, and most DCLs were traced to atypical donor CH involving MDS-associated genes or germline risk alleles. Disclosures Nikiforow: Novartis: Honoraria; Nkarta Therapeutics: Honoraria; Kite/Gilead: Honoraria. DeZern:MEI: Consultancy; Abbvie: Consultancy; Astex: Research Funding; Celgene: Consultancy, Honoraria. Ritz:Rheos Medicines: Consultancy; Infinity Pharmaceuticals: Consultancy; Amgen: Research Funding; Equillium: Research Funding; Kite Pharma: Research Funding; Avrobio: Consultancy; Falcon Therapeutics: Consultancy; TScan Therapeutics: Consultancy; Talaris Therapeutics: Consultancy; LifeVault Bio: Consultancy. Soiffer:VOR Biopharma: Consultancy; Kiadis: Membership on an entity's Board of Directors or advisory committees; Novartis: Consultancy; Juno: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees; alexion: Consultancy; Be the Match/ National Marrow Donor Program: Membership on an entity's Board of Directors or advisory committees; Gilead: Consultancy; Mana Therapeutics: Consultancy; Precision Bioscience: Consultancy; Cugene: Consultancy; Rheos Therapeutics: Consultancy. Lindsley:Jazz Pharmaceuticals: Consultancy, Research Funding; Takeda Pharmaceuticals: Consultancy; MedImmune: Research Funding; Bluebird Bio: Consultancy.

Abstract: Arsenic trioxide (ATO) and all-trans retinoic acid (ATRA) combination therapy yields high complete remission and disease-free survival rates in acute promyelocytic leukemia (APL). ATO is dosed on actual body weight and high ATO doses in overweight patients may contribute to increased toxicity. We performed a retrospective, two-center study comparing toxicities in patients who received the Lo-Coco et al ATRA/ATO regimen with capped ATO, ≤10 mg/dose, and non-capped ATO, 〉 10 mg/dose. A total of 44 patients were included; 15 received doses ≤10 mg and 29 received 〉 10 mg. During induction, there was no difference in the incidence of grade ≥3 hepatotoxicity, grade ≥3 QTc prolongation, neurotoxicity, and cardiac toxicity between groups. In consolidation, patients receiving 〉 10 mg/dose experienced a greater incidence of neurotoxicity (66.7% vs 22.2%; p = 0.046). Capping doses saved $24634.37/patient and reduced waste of partially-used vials. At a median follow-up of 27 months, no disease relapses occurred in either group. This represents an opportunity to improve the safety profile of this highly effective regimen.

Abstract: INTRODUCTION: Patients with homozygous sickle cell anemia (SCA) have frequent cardiopulmonary complications. The clinical prevalence and consequences of intracardiac or intrapulmonary right-to-left shunts in SCA are unknown. Here we report a retrospective examination of this complication, and clinical and laboratory correlates at clinical baseline, in an adult population with SCA. These shunts may be of particular relevance, due to the susceptibility of these patients to thrombosis and the role that desaturated hemoglobin plays in the underlying pathophysiology of this disease. METHODS: This single-institution study included 153 patients with homozygous HbSS, who are followed at Case-UH Medical Center, Cleveland. Clinical and laboratory data were gathered on patients who underwent an echocardiogram with bubble contrast. Echocardiograms were reviewed to confirm the presence, characteristics, and degree of a right-to-left shunt. Immediate (intracardiac) or delayed (intrapulmonary) shunts were identified; the latter were quantitated as 〈 5, 5-15, or 〉 15 bubbles visualized in the left heart. Continuous variables were described, using mean ± standard deviation (SD) and nominal variables were described as N (%). The relationship between clinical characteristics and echo results was determined using linear regression. All analyses were undertaken using SAS v9.4 The SAS Institute, Cary, NC. n.b. Compound heterozygous sickle cell disease patients showed this complication rarely, and were not examined systematically. RESULTS: 82/153 (53.6%) of studied SCA patients were female. Mean (SD) age in years was 32.2 (11.8). 90 (58.8%) patients had an echo with bubble study. In our population, 27 (17.6%) patients had an intracardiac shunt, 41 (26.8%) had an intrapulmonary shunt, and 22 (14.4%) had no shunt present; 63 (41.2%) patients did not have a bubble study. Mean (SD) LDH was 523.3 (318.1) for those with intracardiac shunts and 570.5 (205.4) for those with intrapulmonary shunts. Mean (SD) LDH for all shunt patients combined was 551.3 (255.9). Mean LDH (SD) for patients either with no shunt or no contrast echocardiogram was 415.5 (193.0). In a combined analysis, patients with shunts (N=68) showed a statistically higher baseline mean LDH when compared to a mean LDH of patients with either a negative bubble study or no bubble study (N=85), (551.3 vs 415.5, p 〈 0.001). No difference was seen in prevalence of clinical outcomes of acute chest syndrome or stroke between groups. Laboratory markers of disease activity, such as absolute reticulocyte count or baseline hemoglobin were not different in shunt versus non-shunt patients. 8 patients had 〈 5 bubbles in the left atrium or ventricle, 7 had 5-15 and 25 〉 15 bubbles. There was no evidence that this had an effect at clinical baseline, but the impact of such shunts during crises or multi-organ failure has not yet been assessed. A small number of individuals with shunts showed significant oxygen desaturation with exercise, but these were not statistically distinct from other groups. CONCLUSIONS: 68 patients with HbSS, or almost half of the adult SCA population, had right-to-left shunting on echocardiograms, either intracardiac or intrapulmonary, when evaluated at clinical baseline. These patients had significantly higher LDH than did patients with no bubble study or a negative bubble study, when evaluated in a combined post hoc analysis. The pathophysiology of these shunts is not known, but these have the potential to delay reoxygenation of sickle hemoglobin at clinical baseline and to facilitate right-to-left passage of fat or thrombotic emboli during clinical exacerbation. We are currently evaluating the potential impact of these shunts in patients with HbSS prospectively, both at baseline and during clinical exacerbation. Disclosures Schilz: Genetech: Speakers Bureau; Bayer: Consultancy, Speakers Bureau; Gilead: Consultancy, Speakers Bureau; Actelion: Consultancy, Speakers Bureau; United Therapeutics: Consultancy, Research Funding; Merck: Research Funding; Arena: Research Funding; Eiger: Research Funding.

Abstract: Platelet dosing has been studied in adult oncology inpatients, but there is almost no published evidence to guide platelet dosing for adult outpatients. We evaluated transfusion indices after 1 unit and 2 unit apheresis platelet transfusions at our hospital to determine whether a benefit to 2‐unit transfusions could be detected. Materials and Methods A retrospective chart review was conducted of all adult oncology patients who received an outpatient platelet transfusion over a 16‐month period (July 2016–November 2017). Pre‐ and post‐transfusion platelet count, and chronology of subsequent platelet transfusions were compared. Results A total of 8467 platelet transfusions were administered to 602 patients during the study period. 59·8% of patients ( n  = 360) were transfused interchangeably with one or two platelets throughout the study period. The primary study population were comprised of these patients. On average, a 2‐unit platelet transfusions resulted in a higher immediate post‐transfusion platelet count (43 vs. 37 x 10 3 /μl, P  〈   0·001) and a lower corrected count increment (9707 vs. 14 060, P   〈  0·001). Transfusion with 2 platelets did not increase the number of days between outpatient transfusions (median; 4 vs. 4, P  = 0·959) or the platelet count at the time of next transfusion (11 vs. 11 x 10 3 /μl, P  =   0·147). Conclusion Among adult, oncology outpatients that were transfused interchangeably with one or two units of platelets, transfusion with two platelets did not offer a durable improvement in platelet count or impact the subsequent transfusion schedule.

Abstract: Abnormal erythrocyte adhesion owing to polymerization of sickle hemoglobin is central to the pathophysiology of sickle cell disease (SCD). Mature erythrocytes constitute & gt;80% of all erythrocytes in SCD; however, the relative contributions of erythrocytes to acute and chronic vasculopathy in SCD are not well understood. Here, we showed that bending stress exerted on the erythrocyte plasma membrane by polymerization of sickle hemoglobin under hypoxia, enhances sulfatide-mediated abnormal mature erythrocyte adhesion. We hypothesized that sphingomyelinase (SMase) activity, which is upregulated by accumulated bending energy, leads to elevated membrane sulfatide availability, and thus, hypoxic mature erythrocyte adhesion. We found that mature erythrocyte adhesion to laminin in controlled microfluidic experiments is significantly greater under hypoxia than under normoxia (1856 ± 481 vs 78 ± 23, mean ± SEM), whereas sickle reticulocyte (early erythrocyte) adhesion, high to begin with, does not change (1281 ± 299 vs 1258 ± 328, mean ± SEM). We showed that greater mean accumulated bending energy of adhered mature erythrocytes was associated with higher acid SMase activity and increased mature erythrocyte adhesion (P = .022, for acid SMase activity and P = .002 for the increase in mature erythrocyte adhesion with hypoxia, N = 5). In addition, hypoxia results in sulfatide exposure of the erythrocyte membrane, and an increase in SMase, whereas anti-sulfatide inhibits enhanced adhesion of erythrocytes. These results suggest that the lipid components of the plasma membrane contribute to SCD complications. Therefore, sulfatide and the components of its upregulation pathway, particularly SMase, should be further explored as potential therapeutic targets for inhibiting sickle erythrocyte adhesion.

Abstract: Acute promyelocytic leukemia (APL) is a subtype of acute myeloid leukemia (AML), classified by a translocation between chromosomes 15 and 17 [t(15;17)], that is considered a true oncologic emergency though appropriate therapy is considered curative. Therapy is often initiated on clinical suspicion, informed by both clinical presentation as well as direct visualization of the peripheral smear. We hypothesized that genomic imprinting of morphologic features learned by deep learning pattern recognition would have greater discriminatory power and consistency compared to humans, thereby facilitating identification of t(15;17) positive APL. By applying both cell-level and patient-level classification linked to t(15;17) PML/RARA ground-truth, we demonstrate that deep learning is capable of distinguishing APL in both discovery and prospective independent cohort of patients. Furthermore, we extract learned information from the trained network to identify previously undescribed morphological features of APL. The deep learning method we describe herein potentially allows a rapid, explainable, and accurate physician-aid for diagnosing APL at the time of presentation in any resource-poor or -rich medical setting given the universally available peripheral smear.

Abstract: Background: Patients with acute myeloid leukemia (AML) presenting with hyperleukocytosis have frequent complications and early mortality. Pulmonary, central nervous system, and cardiovascular complications are common. Attempts to improve outcomes in the 10% of AML patients arriving with hyperleukocytosis have included leukapheresis. No prospective randomized study has supported the use of leukapheresis, and retrospective reports have revealed persistently poor outcomes as well as emerging concerns of acquired coagulopathy and worsening hypoxemia after leukapheresis. While many centers use a leukapheresis protocol processing two blood volumes, Johns Hopkins protocol routinely processes three-five blood volumes. This study aimed to assess coagulopathy, hypoxemia, and mortality with large volume leukapheresis. Methods: 32 patients with newly diagnosed AML treated with large volume leukapheresis for WBC depletion are included in this report. Demographic, clinical, laboratory, and apheresis-related data were collected. Coagulopathy and hypoxemia-related metrics were evaluated within 6 hours before leukapheresis and within 6 hours of completion of leukapheresis. Descriptive and inferential statistics (chi square and Mann-Whitney U test) were used to compare pre and post-leukapheresis findings and assess clinical outcomes. Results: Twenty-nine of 32 (93.8%) patients presented with symptomatic leukostasis (with pulmonary and/or CNS symptoms in 26/29). Median blood volume processed was 14.8 liters (range 4-23.4L). Mean platelet count decreased from 60x109/L to 37x109/L after leukapheresis (p 〈 0.001). Comparing supplemental oxygen requirements before and after leukapheresis, 1 patient had decreased requirements, 6 increased, and 25 were unchanged. Median change in prothrombin time (PT) was an increase of 1 second (range: -2.4 to +7.7 seconds; p=0.13). Twenty-six of 26 evaluable patients had a decline in fibrinogen post-pheresis, with median reduction of 84 mg/dL (range: -12 to -483 mg/dL; p=0.04). Twelve of 18 (66.7%) patients 65 years and older died by day 30, compared with 3/14 〈 65 years-old (p=0.01). Fifteen of 32 (46.9%) arrived with acute renal failure on admission, 10 of whom died by day 30. All 8 patients requiring hemodialysis during their initial admission died prior to day 30. 12/32 (37.5%) required mechanical ventilation, 10 of whom died by day 30. Overall, 10/32 (31.3%) died by day 7, and 15/32 (46.9%) died by day 30. The most common primary cause of death was multiorgan failure including both renal failure and hypoxemia in 8 patients. Two patients died of confirmed intracranial hemorrhage, and 2 died with clinical suspicion for intracranial hemorrhage but were too unstable for imaging prior to death. Two deaths were attributed to ischemic stroke, and 1 patient died with isolated refractory hypoxemia. Conclusions: Patients with AML presenting with hyperleukocytosis have a very high mortality, particularly when complicated by symptomatic leukostasis. Similar to Van de Louw's report, we observed worsening coagulopathy and a subgroup with increased oxygen requirements after leukapheresis. While our sample size is too small to draw broad conclusions, we were not able to identify a group clearly benefiting from leukapheresis. We did not find evidence that larger volume leukapheresis decreased complications or mortality. These results should lead to caution when considering leukapheresis for patients with newly diagnosed AML, particularly in those presenting with a severe coagulopathy. Table Disclosures Webster: Pfizer: Consultancy; Amgen: Consultancy; Genentech: Research Funding. Gojo:Amphivena: Research Funding; Amgen Inc: Consultancy, Honoraria, Research Funding; Juno: Research Funding; Merck: Research Funding; Jazz: Consultancy, Honoraria; Novartis: Consultancy, Honoraria; Abbvie: Consultancy, Honoraria. Pratz:Boston Biomedical: Consultancy; Millenium/Takeda: Research Funding; Agios: Consultancy, Research Funding; Astellas Pharma: Consultancy, Research Funding; AbbVie: Consultancy, Research Funding. Smith:Celgene: Consultancy; Jazz: Consultancy; Pfizer: Consultancy; Agios: Consultancy; Novartis: Consultancy. DeZern:Astex Pharmaceuticals, Inc.: Consultancy; Celgene: Consultancy. Levis:Amgen: Consultancy, Honoraria; Agios: Consultancy, Honoraria; Astellas: Consultancy, Research Funding; FUJIFILM: Consultancy, Research Funding; Menarini: Consultancy, Honoraria; Novartis: Consultancy, Research Funding; Daiichi Sankyo Inc: Consultancy, Honoraria.

Abstract: Medical diagnosis and therapy often rely on laboratory testing. We observed mistaken testing in evaluations for hemophagocytic lymphohistiocytosis (HLH) that led to delays and adverse outcomes. Physicians were mistakenly ordering interleukin-2 and quantitative natural killer cell flow cytometry, rather than soluble interleukin 2 receptor (sIL2R) or qualitative natural killer functional tests in the evaluation of patients suspected to have HLH. We initiated a prospective quality improvement project to reduce mistaken testing, reduce delays in correct testing due to mistaken ordering, and improve HLH evaluations. This consisted of provider education, developing an evaluation algorithm, and ultimately required systems interventions such as pop-ups and removal of the mistaken tests from the electronic ordering catalog. Active education reduced mistaken testing significantly in HLH evaluations from baseline (73.3% vs 33.3%, P  = .003, relative risk reduction (RRR) 54.5%), but failed to meet the pre-specified RRR cutoff for success (70%). Education alone did not significantly reduce the proportion of HLH evaluations with delays in sIL2R testing (23.3% vs 7.4%, P  = .096). Mistaken testing increased after the active intervention ended (33.3% vs 43.5%, P  = .390, with RRR 40.7% from baseline. Mistaken test removal was successful: mistaken testing dropped to 0% ( P   〈  .001, RRR 100%), saved $14,235 yearly, eliminated delays in sIL2R testing from mistaken testing (23.3% vs 0%, P  = .008), and expedited sIL2R testing after admission for HLH symptoms (14.6 days vs 3.8 days, P  = .0012). These data show systems controls are highly effective in quality improvement while education has moderate efficacy.