P1418: determinants of haemoglobin in sickle cell disease patients in sub-saharan africa: major impact of the native country and independent effects of inflammation and haemolysis.

Topic: 26. Sickle cell disease Background: Sickle cell disease (SCD) is one of the most frequent monogenic disease worldwide leading to the production of the abnormal haemoglobin (Hb) S that causes chronic peripheral haemolysis. Anaemia has been identified as a morbidity and mortality factor of SCD in several studies, in both developed and undeveloped countries, but few studies have analysed the determinants of Hb level in the African context. Aims: The aim of this study was to analyse the factors associated with Hb level in African patients with SCD, especially disease-specific factors such as haemolysis but also non-specific factors such as inflammation and socio-demographic factors. Methods: We conducted a cross-sectional study nested in the multinational prospective observational cohort CADRE, initiated to describe the natural history of SCD in sub-Saharan Africa. We conducted our analyses in 3 groups of patients according to their phenotype, SS/Sβ0, SC and Sβ+. We studied the association between Hb level with demographic factors (country, sex, age, educational level), body mass index (BMI), haemolysis assessed by clinical icterus, bilirubin or LDH levels and inflammation assessed by leukocytes and platelets count, using a linear regression model. We performed sub-group analyses: 1/in the population with haemolysis biomarkers available i.e. LDH and/or bilirubin; 2/in adults and children, adjusting for patient education level in adults and father education level in children. Variables with ≤20% missing data were imputed with multiple imputations by chained equations. Results: 4206 patients were included, with various SCD phenotypes: SS/Sβ0 (n=3425, 81%), SC (n=600, 14%) and Sβ+ (n=181, 5%) with median Hb level of 7.9, 11.3 and 11.2 g/dL respectively. 56.3% of the patients were from West Africa (Mali, Senegal and Ivory Coast) whereas 43.7% were from Central Africa (Cameroon, Gabon and Democratic Republic of Congo). In multivariate analysis in SS/Sβ0 patients, central African countries (β=-0.5, p<0.001), lower BMI (β=0.15, p<0.001), clinical icterus (β=-0.24, p<0.001) and higher leukocytes counts (β=-0.21, p<0.001) were independently negatively associated with Hb level whereas in SC and Sβ+ patients, female gender (p<0.001), lower BMI (p=0.007 and p=0.02), clinical icterus (p=0.02 and p=0.016) and higher platelets counts (p<0.001) were. In the sub-group of patients with biological haemolysis evaluation, LDH and reticulocytes count were independently negatively associated with Hb level in all SCD phenotypes but more importantly in SS/Sβ0 patients. In the adult population, a higher educational level was positively associated with Hb level in SC patients only. In the pediatric population, father educational level was not associated with Hb level in all phenotype groups. We found significant interactions between: 1/sex and age class in all the models with a Hb level gap between male and female more important in adult patients compared to children and in SC and Sβ+ patients compared to SS/Sβ0 patients; 2/country and clinical icterus in SS/Sβ0 patients with a negative effect of icterus on Hb level more important in central African countries.Summary/Conclusion: Our results show that anaemia in SCD African patients is the result of disease-specific factors such as haemolysis level and also non-specific factors such as inflammation and socio-demographic factors. They also suggest the presence of non-measured confounding factors associated to the African region such as malnutrition, infectious burden and genetic factors other than the SCD phenotype such as beta globin haplotypes. Keywords: Epidemiology, Anemia, Sickle cell disease