A cross-sectional descriptive study was conducted to identify the HbS haplotypes of SCD children in Abidjan and their responses to HU treatment as they relate to their respective haplotypes.

The participants to this study belong to a cohort of SCD children that were given HU treatment in Abidjan between November 2017 and April 2019. They had the homozygous SCD phenotype SSFA2 or the sickle-β˚-thalassemia phenotype SFA2. They were aged 5 to 15 years, and their parents gave their informed consent to their participation to this study. The cohort was built at the Teaching Hospital of Yopougon, in Abidjan, Côte d’Ivoire. The clinical hematology ward of the aforementioned hospital provided medical care during the entire study period. All participants received HU medication upon inclusion. Biological analyses were performed at the central laboratory of the same hospital for hematological assays (HbS phenotyping and complete blood cells count) and biochemical assays (blood sugar, ASAT, ALAT, lactate dehydrogenase or LDH, total and conjugated bilirubin, urea and creatinine assessment). The haplotyping tests were run at the Molecular Biology Unit of the CeDReS at the Teaching Hospital of Treichville, Abidjan, Côte d’Ivoire. Data were collected from patients using an interview form administered by a research nurse. Eligible children who did not consent to the study and those who requested to quit after consenting were excluded from it. The participants did not stop taking their regular medication such as antibiotics, folic acid, malaria prophylaxis, and non-steroid anti-inflammatory drugs.

The study was designed and conducted following the Declaration of Helsinki. It was reviewed and approved by the Ivorian National Ethical Committee for Health and Life Sciences Research under the registration number 108-23/MSHPCMU/CNESVSV-km.

Two types of Blood samples were collected per patient. One with ethylene diamine tetra-acetate (EDTA) as anticoagulant were drawn from fasting children for at least 10 hours at the inclusion and 12 months into HU treatment. The first part of the EDTA sample was used for the hematological assays on the automated blood cells counter XT of Sysmex, and for alkaline and acid electrophoreses of hemoglobin [20] . And the last part of it was used to perform a PCR-RFLP method that used five different primer sets amplifying specific regions within the human β gene cluster (Gγ, Aγ, the first and second sites inside Ψβ, δ), and three restriction enzymes that were HindIII, HindII and HincI [6] . A second blood sample was collected without an anticoagulant for the biochemical analyses of the aforementioned variables on the automated Cobas E411 of Roche.

Probabilities of events were determined, and margins of error were calculated using GraphPad 6 v 2 software. Descriptive analyses described the profile of the studied population. The statistical parameters of the associations between the hemoglobin S haplotypes and the elements of HU treatment or the different phenotypes of SCD were the Pearson’s chi-square test. A result was considered statistically significant for a p-value < 0.05.

In the cohort, more than 80.0% were male with sex ratio of 2.25 (Table 1).

That distribution was similar for the two hemoglobin phenotype groups recruited into the study. The average age was 9 ± 3 years. More than 90% was schooled children from preschool to senior high school. At least half of them had missed some school time due to SCD prior to their recruitment into the study. There were 19 out of 31 in the SSFA2 phenotype group and 3 out of 6 in the SFA2 phenotype group. Most of the participants with SSFA2 phenotype (51.5%) were born to a family that had another child with SCD. That proportion dropped to 33.3% for the children with the SFA2 phenotype.

Of note, 74.3% was not immunized against frequently found infectious diseases for which a vaccine is available (Table 1). At the time of the recruitment, more

than 60.0% of the participants had eating and sleeping disorders in both phenotype groups (Table 2).

After 12 months of HU, that proportion decreased 3-fold (p < 0.0001). At least half of the participants had abdominal and joint pains as well as jaundice at the inclusion that persisted over the 12 months of HU treatment. Hospitalization and blood transfusion were common among the SCD affected children at the inclusion (65.0%). The treatment almost eliminated those occurrences in both hemoglobin phenotypes (p < 0.001). Four participants died during the course of their HU treatment accounting for 10.2%. There were some side effects associated with the treatment. They were dominated by black and tarry stools and nausea in the participants to the study. However in terms of kidney and liver toxicity, the HU treatment was overall well tolerated after 12 months of treatment. The study participants’ kidney and liver functions were not damaged. The most frequent organ damage encountered was the enlargement of the spleen in the children with a hemoglobin phenotype SSFA2 (54.5%). HU treatment reduced the proportion children with enlarged spleen to 6.9% after 12 months.

One of the main benefits of the HU is its ability to increase HbF gene expression resulting in more HbF in the blood. At the inclusion, half of the participants had HbF levels below 10%. HU treatment decreased that proportion to respectively 20.7% and 0% in the SSFA2 and SFA2 phenotype groups. With regards to the actual levels of HbF, the average was largely higher in the SSFA2 phenotype

group than in the SFA2 phenotype group (9.0% versus 17.6%). There were four cases of death under HU treatment, all in the SSFA2 phenotype group (12.1%). This corresponds to an overall mortality rate of 10.2% for the entire cohort.

The SCD diagnosis mainly occurred after the 2 years old mark accounting for 66.7% (Table 2). At the inclusion, the median hemoglobin level was just above 6.0 G/dL along with normochromia (around 50.0%) and normocytosis (around 30.0%). After 12 months of treatment, the hemoglobin level rose to just above 8.0 G/dL of blood. The median white blood cells (WBC) counts were around 12.5 × 10³/dL of blood showing hyperleukocytosis in more than 70.0% of the participants. After the treatment, the median WBC counts was 9.5 × 10³/dL of blood indicating a decrease. Nine participants had high platelets count (PQ) above 4.5 × 10⁵/dL of blood (23.1%). Overall, PQ was normal at inclusion, and remained unchanged after the 12 months of HU treatment.

At their inclusion into the study (Table 1), the distribution of the different hemoglobin phenotypes was 33 cases of the homozygous sickle cell disease phenotype SSFA2 (84.6%) and 6 cases of the sickle cell β˚-thalassemia phenotype SFA2 (15.4%).

Overall, 77 chromosomes were analyzed in this cohort (Table 3).

Out of the 39 participants, the distribution of the HbS haplotypes was 25 cases of heterozygous Benin/atypical haplotype (64.1%) and 13 cases of heterozygous Benin/Senegal (33.3%) and one case of homozygous Senegal (2.6%). Among the four participants who passed away during HU treatment three had the heterozygous Benin/atypical haplotype and one had the heterozygous Benin/Senegal haplotype (Table 4).

Using adjusted logistic regressions; relationships were sought between hemoglobin haplotypes considered as a dependent variable and explanatory variables of the study such as increase in HbF levels, death of the participant, participant missing school time, increase in hemoglobin levels, and decrease in WBC counts (independent variables).

Factors Influenced by hemoglobin S haplotypes

Missing school time and decrease in the number of WBC were statistically associated with having the hemoglobin S haplotype Benin (Table 4). In univariate analyses, no relationships were found between the Hb phenotypes, HbF increase, the death of the participants and the hemoglobin levels (Table 4).

Association between hemoglobin F levels obtained with electrophoresis and specific hemoglobin S haplotypes was previously reported [21] [22] [23] [24] . However other reports have shown that among homozygous patients for a given haplotype, HbF levels varied up until the age of 10 years, and remain so for the entire adult life [25] [26] [27] . Therefore, the genotyping with PCR-RFLP conducted in this study allows for the accurate estimation of the proportions of HbS haplotypes within SCD children in Abidjan.

The average age of the participants to the current study is consistent with previously published data in children [2] [6] [28] . Some studies have tried to assess the impact of genetic modifiers on SCD mainly outside Western Africa [2] [5] [7] [14] [15] [23] [24] . They’ve shown that among other factors that modulate responses to treatment the hemoglobin haplotypes may play a significant role. The current study demonstrated that both the Benin and the Senegal haplotypes were likely to favorably respond to HU treatment in Abidjan. Although a small proportion of Benin haplotype did not respond to treatment. This is consistent with previous reports on HU therapy [8] [16] [17] [19] .

SCD has proven quite deadly in children from Benin and Nigeria, two countries in Western Africa where respectively 50% and 90% of homozygous SSFA2 SCD children die [26] [27] . In the current study, the mortality rate was significantly lower than those rates. Indeed, it has been reported that morbidity and mortality are dependent on the quality of care and the phenotypic hemoglobin profile of the specific child [4] [26] [27] [28] . The difference could be attributed to HU treatment. The time to the diagnostic of SCD indicated late SCD diagnosis compared to that observed in more advanced countries [28] . This points to the lack of perinatal genetic testing in Côte d’Ivoire. Previous reports on benefits of HU treatment in this category of patients have shown consistent improvement of the health of the SCD children with very little adverse effects of the drug itself [17] [18] [19] [24] .

More and more reports on hemoglobin S genotyping with the PCR-RFLP method detects the variants of the classically known hemoglobin haplotypes [5] [6] [16] [24] [28] . This study uncovered only one case of Senegal haplotype homozygote as described before in the literature whereas the others genotypes were associated with a variety of haplotypes that did match the generally accepted definition. Those new haplotypes were referred to as atypical haplotypes. These observations point to a molecular evolution of the populations that requires to gather more genetic data through sequencing in Africa in order to assess their impact on the clinical and biological presentation of SCD in that context. In Abidjan, SCD children are more likely to be of the Benin/atypical heterozygote haplotype. That piece of data is consistent with previous reports based upon inferences from phenotypic studies that suggest the Benin haplotype as the prominent haplotype in Côte d’Ivoire [21] .

All the haplotypes seemed to favorably respond to the HU treatment. However, some children with the Benin/atypical heterozygote haplotype that had the lowest HbF levels before treatment, and those levels did not increase their levels above 10% after 12 months of treatment. This observation was previously described in Brasilia where live a considerable black population [19] .