Co-inheritance of a-thalassemia dramatically decreases the risk of acute splenic sequestration in a large cohort of newborns with hemoglobin SC

Hemoglobin SC (HbSC) is the second most common variant of sickle cell disease worldwide after Hb SS. Hb C is caused by a mutation in the seventh codon of the HBB gene [HBB:c.19G>A(p.Glu7Lys)]. The β allele is associated with different β haplotypes that are named according to the region in which they originated,1 although such multicentric origin of the β allele has recently been proven to be unreliable.2 In addition to this anthropological importance, different β haplotypes can modulate the severity of sickle cell anemia.3 β haplotypes I, II, and III were first described by Boehm et al.4 There are no data in the literature on the association of β haplotypes with clinical and hematologic features. Likewise, the effect of co-inheritance of α-thalassemia (α-thal) in patients with HbSC is still unclear.

This study was based on a retrospective cohort of 461 newborns from Minas Gerais, Brazil (birth date: 01/01/1999 to 31/12/2012). Results on the clinical and hematologic profiles have been recently published.5 Methods for the determination of α-thal and β-globin gene cluster haplotypes, as well as detailed statistical methods are available in the Online Supplementary Appendix S1. P<0.01 was considered significant. The study was approved by the institutional research ethics committee (protocol 13327713.5.0000.5149).

Co-inheritance of α-thal was determined in 389 children; 77 (19.8%) were −α/αα and one (0.26%) −α/−α; 311 (79.9%) had wild HBA alleles (αα/αα). The patients were analyzed for a follow up of a total of 3,522.72 patient-years. There was no difference in follow up between the two groups (8.78 and 9.13 years for patients with or without co-inheritance of α-thal, respectively; P=0.51). The incidence of pain crises for all those 389 children was 52.1 episodes per 100 patient-years (95%CI: 49.8-54.6). The incidence of infections was 60.1 events per 100 patient-years (95%CI: 57.6-62.7), and that of red blood cell (RBC) transfusions, 4.4 events per 100 patient-years (95%CI: 3.73-5.15).

The risks of pain crises, infections, and RBC transfusion were, respectively, 1.73 (95%CI: 1.5-1.99; P<1×10-7), 1.31 (95%CI: 1.16-1.47; P=5×10), and 2.88- fold (95%CI: 1.6-5.18; P=9×10) higher for αα/αα children compared with those with α-thal (Table 1). To the best of our knowledge, this is the first clear demonstration that co-inheritance of α-thal in children with HbSC decreases the incidence of these three events. In children with HbSS, the effect of α-thal on the incidence of pain crises before hydroxyurea was introduced in the management of patients with sickle cell anemia is controversial.76 In a contemporary cohort of 250 children with HbSS, our group found that the incidence of pain crises before hydroxyurea was started was 82.3 (95%CI: 75.8-89.3) per 100 patient-years for those with α-thal compared with 76 (95%CI: 72.0-80.2) for those without α-thal (P=0.11) (Rezende et al., 2019, unpublished data).

Table 1.Association of co-inheritance of - α^3.7 thalassemia (one or two deleted genes) with clinical data from 389 children with hemoglobin SC.

Of the 78 children with co-inheritance of α-thal (−α/αα or −α/−α), only one (1.3%) had acute splenic sequestration crisis (ASSC) whereas 55 of 311 (17.7%) children without α-thal had at least one episode of ASSC (see Online Supplementary Appendix S1 for a definition of phenotype). The cumulative probability of ASSC in the groups with and without co-inheritance of α-thal was 1.3% (95%CI: 0-3.85%) and 24% (95%CI: 17.7-30.3%), respectively (P=0.0003) (Figure 1). As expected, a statistically significant association between co-inheritance of α-thal and lower levels of mean corpuscular volume (MCV), mean cell Hb (MCH), reticulocyte, and white blood cell (WBC) counts was found (P<0.001) (Table 2). In children with Hb SS, the rates of events seem to be similar in those with or without α-thal.6 In our cohort of 250 HbSS children (Rezende et al., 2019, unpublished data), there was no significant difference in the cumulative probability of ASSC between children with or without α-thal (42% and 38%, respectively; P=0.51).

Figure 1.Kaplan-Meier probability curve of the occurrence of the first episode of acute splenic sequestration, according to the presence or absence of co-inheritance of −α^3.7 thalassemia (one or two deleted genes).

Table 2.Association of co-inheritance of −α^3.7 thalassemia (one or two deleted genes) and mean laboratory results in 389 children with hemoglobin SC.

In the present study, one-fifth of children with HbSC co-inherited α-thalassemia. The frequency of this co-inheritance in São Paulo, Brazil, was recently shown to be 26.7%,8 but was as high as 35.2% in other reports.9 In a clinical study of 179 adults with HbSC in France, the prevalence was 27%.10 In Minas Gerais, the proportion of α-thal co-inheritance in children with Hb SS was 26.6%.11 Studies suggest that α-thal would limit the number of cells with high intracellular Hb concentration, with consequent attenuation of the clinical events, as observed in the present study and that of others.12 The most remarkable clinical effect in the present study was the protection against ASSC in HbSC-possessing children with co-inheritance of α-thal, but rates of painful crises, infections, and RBC transfusion were also significantly lower in children with α-thal. In 1991, Platt et al. attributed the slight increase in the pain rate among HbSS patients with α-thal to their higher hematocrits.7 In the present study, the mean Hb concentration for HbSC children with and without α-thal was 10.55 g/L and 10.48 g/L, respectively (P=0.50). In a multivariate regression model, the decrease of the pain rate in children who co-inherited α-thal was statistically significant (P=0.001), and independent of the non-significant positive effect of the Hb concentration level on the pain rate (P=0.14).

Haplotype determination was performed in 387 children (Online Supplementary Table S1): 211 (54.5%) were β CAR and 162 (41.9%) β Benin. Regarding β alleles, type I was predominant (305 cases; 78.8%), as usually reported.13104 The most common genotypes were CAR/I (42.6%) and Benin/I (34.9%). The HpaI restriction site (rs4426157) was “–/–” when the β haplotype was Benin (n=34) and "+/–" when it was CAR (n=50).

As expected, the prevalence of β CAR and Benin haplotypes in SC children was very similar to that already reported for our SS children in Minas Gerais.11 Many studies indicate a gradually less severe clinical outcome in patients with the Arab-Indian, Senegal, Benin, and CAR haplotypes,3 probably related to the observed gradual decline of the fetal Hb concentration. In the present study, there was a statistically significantly higher risk of pain crises and transfusion rate in children with the CAR haplotype, but the difference was modest (Online Supplementary Table S2). There was no difference in baseline hematologic laboratory results when the β CAR and β Benin haplotype groups were compared to each other (Online Supplementary Table S3). Some statistically significant clinical differences between β-I and β-II haplotype were observed (Online Supplementary Table S2), but the usefulness of this as a prognostic clinical tool seems to be irrelevant.

Another interesting result of the present study, corroborating the unicentric origin of HbC allele in West-Central Africa, was that the HpaI restriction site, characteristically negative for β4 was equally negative when associated with β -Benin and positive when associated to β -CAR. Nowadays, the Benin haplotype is predominant in West-Central Africa.41

In conclusion, the present study adds new and relevant information regarding the mitigating influence of co-inheritance of α-thal on the clinical and laboratory course of HbSC disease in a large cohort of newborns.