Independent Prognostic Value of Flow Cytometry (FCM) in Myelodysplastic Syndromes (MDS) - Composition of a Prognostic FCM-Score for Overall Survival

BackgroundFlow cytometry (FCM) is a co-criterion in Myelodysplastic Syndromes (MDS) diagnostics. The aims of the present study were (1) to develop a composite prognostic FCM-Score for OS in MDS ; (2) to asess whether a computational algorithm could improve the identification of aberrant expression and cell population frequencies and (3) to validate the accuracy of the prognostic FCM-Score for OS in MDS in an independent cohort.MethodsFCM was performed in bone marrow (BM) of 399 patients cytomorphologically classified as MDS. Cell populations were identified. OS was assessed following univariate and multivariable Cox proportional hazards regression analysis using log-rank likelihood test to calculate FCM prognostic models. Kaplan Meier curves, and receiver operating characteristic curve (ROC) were used to test independent prognostic value of the models versus known diagnostic FCM-scores (Ogata-, FCSS-, iFS-score). T-REX pipeline was applied to check the feasibility of an unsupervised machine learning approach in identifying the FCM parameters. Validation of the prognostic score with best performance in an independent cohort of 110 MDS patients was performed.ResultsPrognostic FCM-scores were calculated based on the 9 FCM parameters with independent prognostic impact. FCM-scores, FCM-A: HR (95 %CI) 3.20 (2.15 - 4.48); FCM-B: 4.08 (2.54 - 6.55)), outperformed well known diagnostic scores (Ogata-score (2.00 (1.29 – 3.11))). FCM-scores allowed a better prognostic grading than IPSS-R (HR (95 %CI): 2.37 (1.61-3.49)). Kaplan Meier survival curves stratified by FCM-score A and B showed a highly significant overall survival benefit for patients with a low score (p<.0001) and FCM-A and FCM-B scores presented better discrimation capability than IPSS-R ((AUC): 0.69, and 0.71 vs. 0.62). T-REX pipeline was able to identify differences in expression of significant parameters between the low and high scoring patients. The validation of the OS prognostic score obtained presented good discrimination performance (c-stats 0.764; AUC 0.7463).ConclusionsA promising novel prognostic score based on distinct FCM characteristics which could predict overall survival in MDS patients was presented.Legal entity responsible for the studyThe authors.FundingHas not received any funding.DisclosureAll authors have declared no conflicts of interest. BackgroundFlow cytometry (FCM) is a co-criterion in Myelodysplastic Syndromes (MDS) diagnostics. The aims of the present study were (1) to develop a composite prognostic FCM-Score for OS in MDS ; (2) to asess whether a computational algorithm could improve the identification of aberrant expression and cell population frequencies and (3) to validate the accuracy of the prognostic FCM-Score for OS in MDS in an independent cohort. Flow cytometry (FCM) is a co-criterion in Myelodysplastic Syndromes (MDS) diagnostics. The aims of the present study were (1) to develop a composite prognostic FCM-Score for OS in MDS ; (2) to asess whether a computational algorithm could improve the identification of aberrant expression and cell population frequencies and (3) to validate the accuracy of the prognostic FCM-Score for OS in MDS in an independent cohort. MethodsFCM was performed in bone marrow (BM) of 399 patients cytomorphologically classified as MDS. Cell populations were identified. OS was assessed following univariate and multivariable Cox proportional hazards regression analysis using log-rank likelihood test to calculate FCM prognostic models. Kaplan Meier curves, and receiver operating characteristic curve (ROC) were used to test independent prognostic value of the models versus known diagnostic FCM-scores (Ogata-, FCSS-, iFS-score). T-REX pipeline was applied to check the feasibility of an unsupervised machine learning approach in identifying the FCM parameters. Validation of the prognostic score with best performance in an independent cohort of 110 MDS patients was performed. FCM was performed in bone marrow (BM) of 399 patients cytomorphologically classified as MDS. Cell populations were identified. OS was assessed following univariate and multivariable Cox proportional hazards regression analysis using log-rank likelihood test to calculate FCM prognostic models. Kaplan Meier curves, and receiver operating characteristic curve (ROC) were used to test independent prognostic value of the models versus known diagnostic FCM-scores (Ogata-, FCSS-, iFS-score). T-REX pipeline was applied to check the feasibility of an unsupervised machine learning approach in identifying the FCM parameters. Validation of the prognostic score with best performance in an independent cohort of 110 MDS patients was performed. ResultsPrognostic FCM-scores were calculated based on the 9 FCM parameters with independent prognostic impact. FCM-scores, FCM-A: HR (95 %CI) 3.20 (2.15 - 4.48); FCM-B: 4.08 (2.54 - 6.55)), outperformed well known diagnostic scores (Ogata-score (2.00 (1.29 – 3.11))). FCM-scores allowed a better prognostic grading than IPSS-R (HR (95 %CI): 2.37 (1.61-3.49)). Kaplan Meier survival curves stratified by FCM-score A and B showed a highly significant overall survival benefit for patients with a low score (p<.0001) and FCM-A and FCM-B scores presented better discrimation capability than IPSS-R ((AUC): 0.69, and 0.71 vs. 0.62). T-REX pipeline was able to identify differences in expression of significant parameters between the low and high scoring patients. The validation of the OS prognostic score obtained presented good discrimination performance (c-stats 0.764; AUC 0.7463). Prognostic FCM-scores were calculated based on the 9 FCM parameters with independent prognostic impact. FCM-scores, FCM-A: HR (95 %CI) 3.20 (2.15 - 4.48); FCM-B: 4.08 (2.54 - 6.55)), outperformed well known diagnostic scores (Ogata-score (2.00 (1.29 – 3.11))). FCM-scores allowed a better prognostic grading than IPSS-R (HR (95 %CI): 2.37 (1.61-3.49)). Kaplan Meier survival curves stratified by FCM-score A and B showed a highly significant overall survival benefit for patients with a low score (p<.0001) and FCM-A and FCM-B scores presented better discrimation capability than IPSS-R ((AUC): 0.69, and 0.71 vs. 0.62). T-REX pipeline was able to identify differences in expression of significant parameters between the low and high scoring patients. The validation of the OS prognostic score obtained presented good discrimination performance (c-stats 0.764; AUC 0.7463). ConclusionsA promising novel prognostic score based on distinct FCM characteristics which could predict overall survival in MDS patients was presented. A promising novel prognostic score based on distinct FCM characteristics which could predict overall survival in MDS patients was presented.