Sickle cell disease severity scoring: cross-validation between a disease severity score and a paediatric severity score

• Main Author: Coelho, Andreia
• Publication Date: 2012
• Other Authors: Dias, Alexandra, Morais, Anabela, Nunes, Baltazar, Faustino, Paula, Lavinha, João
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10400.18/1115
• topic: Sickle Cell Disease; Severity Scores; Fetal Hemoglobin; Doenças Genéticas

Sickle cell disease (SCD) is a monogenic disorder under polygenic and environmental control. This aetiopathogenic architecture leads to marked clinical heterogeneity with the emergence of multiple and diverse subphenotypes, which makes the patients severity stratification particularly difficult. A number of severity scores have been proposed, aiming at the integration of many clinical dimensions into a meaningful single synthetic measure of morbidity and/or risk of death within a given period. As part of a wider research on the development and validation of vaso-occlusion early predictors in SCD, we have analysed (i) the correlation between two scores, namely a disease severity score (DSS) and a paediatric severity score (PSS), and (ii) the association of scores to a number of genotypic and phenotypic markers in a series of 99 paediatric SS patients followed-up in two large general hospitals in Greater Lisbon area. The evaluated patients were mostly (97%) of Sub-Saharan African ancestry and presented an M/F ratio of 1.17, a median current age of 9.9 years, and a median follow-up/patient of 5.0 years. Both DSS and PSS displayed a non-normal (p<0.01) multimodal distribution. The Spearman’s ρ correlation coefficient between DSS and PSS was 0.280 which is significant at the 2-tailed 0.01 level. Although statistically significant, this weak positive correlation, coupled with a fair inter-rater agreement (κ value) of 0.281, suggests the compared severity scores, although overall convergent, are measuring different aspects of the phenotype, at different developmental stages (paediatric versus adult) and with different weights. Regarding the association studies, statistically significant relationships were observed (i) of DSS to a beta-globin gene cluster polymorphism, leukocyte and reticulocyte counts, RDW and HbF and (ii) of PSS to HbF. In conclusion, recently developed SCD severity scores are not yet the effective tool needed for patient stratification in genotype/phenotype (including response to medical interventions) association studies, as well as in the discovery and validation of prognosis markers of the largely unpredictable SCD clinical course. In particular, the marked reduction of sepsis incidence in younger SCD patients in developed countries is just an illustration of the type of problems novel improved severity scores should address. Acknowledgements: This study was partially funded by FCT grants PIC/IC/83084/2007 and CIGMH. The authors have no competing interests.