Abstract: Introduction. Elderly patients aged 65 or older with acute myeloid leukemia (AML) are often ineligible for hematopoietic stem cell transplantation (HSCT) and generally have a poor prognosis. The prognostic risk classification based on NCCN Guidelines Version3. 2017; NCCN 2017 (O'Donnell MR, JNCCN. 2017) is widely performed; however, the impact of this classification on the prognosis of such elderly AML patients is unclear. While nutritional status assessment using controlling nutritional status (CONUT score) based on serum level of albumin (Alb), total-cholesterol (T-chol) and total lymphocyte count (TLC) predicts prognosis of elderly patients with solid tumor (Liu X, BMC Cancer. 2018), the prognostic significance of nutritional status in elderly patients with AML remains to be clarified. Methods. Hokkaido Leukemia Net (HLN) is prospective cohort study collecting AML samples from hospitals of North Japan Hematology Study Group (NJHSG). In this study, we focused on newly diagnosed AML patients aged 65 or older treated without HSCT, and investigated cytogenetic and molecular abnormality of leukemic cells including FLT3-ITD, NPM1, CEBPA, and KIT. We stratified the patients into favorable, intermediate, and adverse risk group based on NCCN 2017. In order to adjust the assessment of nutritional status for hematopoietic malignancy, we modified the CONUT score eliminating TLC from evaluation criteria (modified-CONUT score, Table) and defined patients with score 3 or more at diagnosis as high group. We evaluated the impacts of NCCN 2017 and modified-CONUT scores on overall survival (OS) in these patients. The study procedures were in accordance with the Helsinki Declaration and institutional ethical guidelines, conducted under the auspices of the institutional ethics committee, and approved by the institutional review boards. Results. Overall, 181 patients with newly diagnosed AML patients aged 65 or older enrolled in HLN between April 2010 and March 2018. Seven patients undergone HSCT were excluded and 174 patients were reviewed (Age 65-93, median 71; male 104, female 70). In this cohort, classification based on NCCN 2017 successfully divided the prognosis of the patients for 2-year and 5-year OS [2-year OS; favorable group, 59.9%; intermediate group, 43.8%; adverse group, 8.1%, 5-year OS; favorable group, 41.5%, intermediate group, 19.7%; adverse group, 4.1%, P=0.00258, Figure A]. On 112 patients who had available records of serum Alb levels and T-chol levels at diagnosis, OS in patients with high modified-CONUT score was significantly lower than the low score group [2-year OS; low score group, 50.3%; high score group, 18.5%; 5-year OS; low score group, 23.5%; high score group, 9.24%, P=0.00203, Figure B] . In a univariate analysis, adverse group in NCCN 2017 and high modified-CONUT score were associated with poor 2-year OS. A multivariate analysis demonstrated that　adverse group in NCCN 2017 and high modified-CONUT score were independently associated with poor 2-year OS (adverse group in NCCN 2017; HR, 2.464 ; 95% CI, 1.514 to 4.012, P=0.0002854, high modified-CONUT score; HR, 1.664 ; 95% CI, 1.051 to 2.635, P=0.02976; log-rank). Altogether, we demonstrated that risk stratification based on NCCN 2017 and modified-CONUT score are both effective for predicting prognosis in elderly patients with newly diagnosed AML. Conclusion. The prognostic risk classification based on AML disease status using NCCN Guidelines 2017 effectively stratify prognosis of elderly patients with AML. Moreover, new assessment scoring of patients' nutrition status based on modified-CONUT score can easily stratify elderly patients with newly diagnosed AML. Figure Disclosures Teshima: Novartis: Honoraria, Research Funding.

Abstract: Elderly patients aged 65 or older with acute myeloid leukemia (AML) have poor prognosis. The risk stratification based on genetic alteration has been proposed in national comprehensive cancer network (NCCN) guideline but its efficacy was not well verified especially in real world elderly patients. The nutritional status assessment using controlling nutritional status (CONUT) score is a prognostic biomarker in elderly patients with solid tumors but was not examined in elderly AML patients. We performed prospective analysis of genetic alterations of 174 patients aged 65 or older with newly diagnosed AML treated without hematopoietic stem cell transplantation (HSCT) and developed simplified CONUT (sCONUT) score by eliminating total lymphocyte count from the items to adapt AML patients. In this cohort, both the NCCN 2017 risk group and sCONUT score successfully stratified the overall survival (OS) of the elderly patients. A multivariable analysis demonstrated that adverse group in NCCN 2017 and high sCONUT score were independently associated with poor 2-year OS. Both risk stratification based on NCCN 2017 and sCONUT score predict prognosis in the elderly patients with newly diagnosed AML.

Abstract: Introduction. Older patients aged over 65 with acute myeloid leukemia (AML) are often ineligible for hematopoietic stem cell transplantation (HSCT) and generally have a poor prognosis. The personalized strategy for appropriate treatment for these patients has not yet been completely established. While C-reactive-protein (CRP) to albumin ratio (CAR) based on serum level of CRP and albumin (Alb) predicts prognosis of patients with hematological malignancies treated with HSCT (Miyashita E, BBMT. 2019), the prognostic significance of this inflammatory and nutritional assessment in transplant-ineligible elderly patients with AML remains to be clarified. Methods. Hokkaido Leukemia Net (HLN) is prospective cohort study collecting AML samples from hospitals of North Japan Hematology Study Group (NJHSG). In this study, we focused on older (≥65 years) patients with newly diagnosed AML treated without HSCT and investigated cytogenetic and molecular abnormality of leukemic cells including FLT3-ITD, NPM1, CEBPA, and KIT. We stratified the patients into favorable, intermediate, and adverse risk group based on NCCN 2017. Based on the result of blood sample test, we assessed CAR [CRP (mg/dL) / Alb (g/dL)] at diagnosis for each patient. Since the median CAR at diagnosis was 0.52 in this cohort, we defined CAR & lt; 0.52 and ≥ 0.52 as CAR low and CAR high, respectively. We evaluated the impacts of NCCN 2017 and CAR on overall survival (OS) in these patients. The study procedures were in accordance with the Helsinki Declaration and institutional ethical guidelines, conducted under the auspices of the institutional ethics committee, and approved by the institutional review boards. Results. In total, 197 patients with newly diagnosed AML patients aged 65 or older enrolled in HLN between April 2010 and March 2019. Nine patients undergone HSCT were excluded, and 188 patients were reviewed (Age 65-93, median 72; male 119, female 69). In this cohort, classification based on NCCN 2017 successfully divided the prognosis of the patients for 2-year and 5-year OS [2-year OS; favorable group, 58.1%; intermediate group, 29.4%; adverse group, 12.6%, 5-year OS; favorable group, 33.7%, intermediate group, 8.74%; adverse group, 9.44%, P=0.000079, Figure A]. OS in patients with CAR 0.52 or more was significantly lower than the low score group [2-year OS; CAR low group, 43.4%; CAR high group, 21.5%; 5-year OS; CAR low group, 21.0%; CAR high group, 9.96%, P=0.00013, Figure B] . In a univariate analysis, higher age, adverse group in NCCN 2017, high CAR, and not reached complete remission (CR) after first induction chemotherapy were associated with poor 2-year OS. A multivariate analysis demonstrated that adverse group in NCCN 2017 and high CAR were independently associated with poor 2-year OS (adverse group in NCCN 2017; HR, 1.671; 95% CI, 1.157 to 2.421, P=0.007, high CAR; HR, 1.572; 95% CI, 1.093 to 2.276, P=0.01618; log-rank). Combined with NCCN 2017, while there were no significant prognostic impacts of CAR at diagnosis in patients with favorable or adverse risk groups (Figure C, E), we found that in intermediate risk group the patients with high CAR had poorer prognosis than patients with low CAR [2-year OS; CAR low group, 43.7%; CAR high group, 15.9%, P=0.0000658; Figure D]. Altogether, we demonstrated that high CAR predicts poor prognosis in the transplant-ineligible elderly patients with newly diagnosed AML and improves risk stratification of the patients with NCCN 2017 intermediate risk group. Conclusion. CAR is a simple and easily evaluable parameter for predicting outcomes of transplant-ineligible elderly patients with newly diagnosed AML, independently from NCCN disease risk classification. Moreover, CAR can further stratify the prognosis of patients with NCCN 2017 intermediate risk group and improve the predictive risk stratification for elderly AML patients. Figure 1 Figure 1. Disclosures Kondo: SANWA KAGAKU KENKYUSHO CO.,LTD.: Consultancy; Astellas Pharma Inc.: Consultancy, Honoraria; Otsuka Pharmaceutical Co., Ltd.: Honoraria, Research Funding; Novartis Pharma KK: Honoraria; Bristol-Myers Squibb Company: Honoraria; Sumitomo Dainippon Pharma Co., Ltd.: Honoraria. Teshima: Merck Sharp & Dohme: Membership on an entity's Board of Directors or advisory committees; Nippon Shinyaku Co., Ltd.: Research Funding; Pfizer Inc.: Honoraria; Fuji pharma CO.,Ltd: Research Funding; Kyowa Kirin Co.,Ltd.: Honoraria, Research Funding; TEIJIN PHARMA Limited: Research Funding; Bristol Myers Squibb: Honoraria; Janssen Pharmaceutical K.K.: Other; CHUGAI PHARMACEUTICAL CO., LTD.: Research Funding; Astellas Pharma Inc.: Research Funding; Takeda Pharmaceutical Company: Honoraria, Membership on an entity's Board of Directors or advisory committees; Novartis International AG: Membership on an entity's Board of Directors or advisory committees, Other, Research Funding; Gentium/Jazz Pharmaceuticals: Consultancy; Sanofi S.A.: Research Funding.

Abstract: Introduction Since Core-binding factor acute myeloid leukemia (CBF-AML) is categorized as a favorable cytogenetic risk group, allogenic hematopoietic stem cell transplantation (HSCT) is generally not recommended during the first complete remission (CR1). However, approximately 30-50% of CBF-AML patients relapse, and about 50% of relapsed patients remain incurable. Therefore, in real-world clinical setting, the indication of HSCT even in CR1 is considered based on prognosis prediction and minimal residual disease (MRD) status. With respect to prognosis prediction by chromosomal abnormalities, the prognostic impact of loss of sex chromosome (LOS) in the patients with CBF-AML is still controversial. Previous studies suggest that the impact of LOS in predicting the prognosis of CBF-AML may depend on treatment strategies including HSCT. In addition, although CBF-AML develops LOS at a high rate of 20-40%, it is still unclear whether LOS is age-related or neoplastic. Methods Hokkaido Leukemia Net (HLN) is prospective cohort study collecting AML samples from hospitals of North Japan Hematology Study Group (NJHSG).Cases of CBF-AML that were registered in HLN from January 2010 to December 2019 were used for analysis in the present study. All the patients received either intensive or low-intensity chemotherapy. HSCT was performed at the discretion of the attending physicians even in CR1. Conventional cytogenetic analysis was performed at the time of diagnosis and after each treatment. FLT3-ITD and KIT exon17 mutations were analyzed using genome DNA template at the central laboratory. The results of genetic analysis were returned to physicians so that they could reflect the genetic result in their treatment choice. We evaluated the impacts of several factors including advanced age, poor response to chemotherapy, LOS, and KIT exon17 mutation on overall survival (OS) and relapse-free survival (RFS) in these patients. This study was conducted in compliance with ethical principles based on the Declaration of Helsinki and was approved by the institutional review board of Hokkaido University Hospital. Results A total 96 CBF-AML patients including 62 patients with RUNX1/RUNX1T1 and 34 patients with CBFβ/MYH11 were enrolled in the present study. Intensive induction chemotherapy using '7+3' cytarabine plus anthracycline and low-intensity induction chemotherapy were performed in 90 patients and 6 patients, respectively. In 90 patients who underwent intensive induction chemotherapy, the CR rate after one cycle of induction therapy was 73.3%. Total CR rate as the best response was 91.7% for all CBF-AML patients. HSCT in CR1 was performed in 15 patients (15.6%) at the discretion of the attending physician based on genetic alterations, poor response to chemotherapy, and sustained detection of MRD after chemotherapy. There was no significant difference in 5-year OS (RUNX1/RUNX1T1 64.9% vs CBFβ/MYH11 52.9%; P=0.546) or 5-year RFS (63.6% vs 53.3%; P=0.364) between the CBF subtypes. LOS was significantly more frequent in patients with RUNX1/RUNX1T1 than in patients with CBFβ/MYH11 (35.5% vs 5.9%; P=0.001). Multivariate analyses for 5-year OS in CBF-AML patients revealed that age of 50 years or older (HR: 3.46, 95% CI: 1.47-8.11, P=0.004) and receiving 2 or more induction cycles (HR: 3.55, 95% CI: 1.57-8.05, P=0.002) were independently associated with worse OS and that LOS was independently associated with better OS (HR: 0.09, 95% CI: 0.01-0.71, P=0.022). We next sought to clarify whether LOS was age-related or neoplastic based on the transition of chromosome results. All the 24 patients with LOS achieved CR, and chromosome analysis at CR showed a normal karyotype in all of the 21 patients for whom chromosome analysis was performed. In the remaining 3 patients, only FISH analysis for RUNX1/RUNX1T1 or CBFβ/MYH11 was performed. Furthermore, chromosome analysis showed mosaics of metaphase cells with and without t(8;21) in 9 patients at diagnosis. In all of those patients, metaphase cells without t(8;21) showed a normal karyotype. Conclusion In our real-world clinical setting in which '7+3' induction was performed and the indication of HSCT was determined by prognostic factors and MRD status, LOS correlated to prolonged survival in CBF-AML patients. Furthermore, the transition of chromosome results revealed that LOS was not age-related physiological loss but part of neoplastic chromosomal abnormalities. Figure 1 Figure 1. Disclosures Teshima: CHUGAI PHARMACEUTICAL CO., LTD.: Research Funding; Janssen Pharmaceutical K.K.: Other; Novartis International AG: Membership on an entity's Board of Directors or advisory committees, Other, Research Funding; Gentium/Jazz Pharmaceuticals: Consultancy; Takeda Pharmaceutical Company: Honoraria, Membership on an entity's Board of Directors or advisory committees; Kyowa Kirin Co.,Ltd.: Honoraria, Research Funding; Astellas Pharma Inc.: Research Funding; Merck Sharp & Dohme: Membership on an entity's Board of Directors or advisory committees; TEIJIN PHARMA Limited: Research Funding; Nippon Shinyaku Co., Ltd.: Research Funding; Bristol Myers Squibb: Honoraria; Fuji pharma CO.,Ltd: Research Funding; Pfizer Inc.: Honoraria; Sanofi S.A.: Research Funding. Kondo: Astellas Pharma Inc.: Consultancy, Honoraria; Otsuka Pharmaceutical: Honoraria, Research Funding; Novartis Pharma KK: Honoraria; Bristol-Myers Squibb Company: Honoraria; Sumitomo Dainippon Pharma: Honoraria; Sanwa Kagaku Kenkyusho CO.,LTD: Consultancy.

Abstract: Acute myeloid leukemia (AML) patients older than 65 years have a poor prognosis. Recently, CAR (C-reactive-protein/albumin ratio) has been actively reported as a prognostic index reflecting the nutritional and inflammatory status of elderly patients with solid tumors, but the usefulness of this index as a prognostic indicator in transplant-ineligible elderly AML patients has not been investigated. We studied genetic alterations and CARs in 188 newly diagnosed AML patients aged 65 years or older who were treated in a multicenter setting and had treated without HSCT. Both NCCN 2017 risk group, reflecting the genetic component of the tumor, and CAR, reflecting the inflammatory and nutritional status of the patient, successfully stratified the overall survival (OS) of the patients (2-year OS; CAR low vs high, 42.3% vs 17.8%, P   〈  0.001). Furthermore, in multivariate analysis, NCCN 2017 poor group and high CAR were extracted as independent poor prognostic factors predicting 2-year OS in the current study. We found, for the first time, that CAR at diagnosis predicted the prognosis of elderly patients with newly diagnosed AML treated without HSCT.