Molecular Testing in Sickle Cell Disease: From Newborn Screening to Transfusion Care

ABSTRACT

Sickle cell disease (SCD) is one of the most frequent monogenic diseases worldwide and a highly heterogeneous and complex disease. SCD care carries several challenges. This includes early and accurate diagnosis as well as optimal red blood cell transfusion matching in this population carrying a high risk of alloimmunization. For decades, molecular biology has used hemoglobin and SCD as models for the development of several molecular tools. Such tools can now be used for various aspects of SCD care. Molecular diagnosis is notably the root of noninvasive prenatal testing. In postnatal diagnosis, including newborn screening, molecular approaches can overcome several limitations of protein‐based methods. Simple approaches such as polymerase chain reaction can be used as a high‐throughput and low‐cost screening test. Moreover, combining sequence and deletion analyses allows for a comprehensive study of the β‐globin locus, resolving complex cases. In transfusion care, genotyping for blood group determination has been shown to be more accurate compared to protein‐based serological testing. Future development of molecular testing in SCD includes their use as prognostic tools and recent molecular diagnosis approaches. However, despite carrying major advantages, molecular testing may also present some limitations, such as high cost, limited accessibility in many countries, and limited information using targeted approaches. Molecular testing has a different pattern of advantages and limitations than protein‐based analyses. Therefore, the optimal use of molecular testing is frequently not as a standalone approach but in combination with protein‐based techniques. The optimal combination depends on the resources available and the clinical challenge, to ultimately improve SCD care.

1. Introduction

1.1. Epidemiology

Sickle cell disease (SCD) is one of the most frequent monogenic recessive diseases worldwide. The Global Burden Disease initiative recently estimated that 7.74 million (95% confidence interval [₉₅CI]: 6.51–9.2) individuals live with SCD. There is a global trend toward higher SCD prevalence leading to more than 500 000 babies born yearly with this disease and an overall 41.4% increase (₉₅CI₅: 38.3–44.9) since 2000 [1]. SCD mortality is high with an estimate of 376 000 (₉₅CI: 303 000–467 000) disease‐related death [1]. Despite major improvement in care, life expectancy of people with SCD living in rich countries remains significantly shorter than the general population, being currently estimated to be around 60 years. The main causes of death are sudden death, pulmonary hypertension, kidney failure, and infections. A major element while considering SCD care including the place of molecular testing, is SCD global distribution, which is highly heterogeneous worldwide, as 80% of individuals with SCD live in Sub‐Saharan Africa [2]. The other regions with high prevalence are India, the Caribbean, and the Middle East. This distribution reflects a lower mortality rate of individual carriers of SCD alleles in heterozygous state infected by plasmodium falciparum (protection that is not found in individuals with SCD).

1.2. Molecular Basis of SCD

SCD is a group of monogenic recessive diseases due to a constitutional (germline) missense variant (HBB:c.20A>T, p.Glu7Val, rs334) in the β‐globin gene, which codes for the β subunit of hemoglobin (Hb). This gain‐of‐function variant leads to an abnormal β‐globin chain (β^S‐globin) and ultimately to an abnormal Hb molecule, the HbS, composed of two α‐globin chains and two β^S‐globin chains (α₂β^S ₂). SCD, however, is a basket term encompassing several molecular variations. β^S‐globin allele homozygosity is the most frequent combination and is frequently named sickle cell anemia (SCA) [3]. Compound heterozygosity of β^S‐globin and null β‐thalassemia alleles leads to S/β⁰ thalassemia with a similar pathophysiology to SCA. In both SCA and S/β⁰ thalassemia, no HbA is produced. S/β⁺‐thalassemia occurs in the case of compound heterozygosity of β^S‐globin and β‐thalassemia alleles that retain residual HbA production, although reduced compared to wild‐type alleles. The severity of S/β⁺‐thalassemia depends on the residual quantity of HbA produced, thus it depends on the specific molecular alteration. Compound heterozygosity of β^S‐globin allele with another β‐globin variant at the same locus (HBB:c.19G>A, p.Glu7Lys), which leads to the production of the abnormal Hb C, leads to HbSC disease with a specific expression pattern. Other compound heterozygosities can occur with rarer β‐globin variants (e.g., HbSD, HbSO‐Arab et HbSE), with variable severity profiles.

1.3. Overview of SCD Pathophysiology

Compared to wild‐type Hb (HbA), HbS has lost one negative charge and gained hydrophobicity. This alters Hb dimer–tetramer assembly, resulting in HbS instability and polymerization upon deoxygenation, which is the root of the numerous pathophysiological pathways involved in SCD. The polymerization‐depolymerization cycle will distort red blood cells (RBC), leading to several alterations and eventually its destruction (hemolysis). In addition, the RBC alterations lead to increased RBC adhesion, nitrite oxide depletion, increased oxidative stress, sterile inflammation, and endothelial activation. These phenomena lead to vaso‐occlusion, ischemic‐reperfusion injuries, vascular remodeling, and organ dysfunction [4].

1.4. Expression of SCD and Prognosis Factors

Patients with SCD present a chronic hemolytic anemia that can be associated with episodes of pain, most often referred to as vaso‐occlusive crises, with an extremely variable severity. SCD can lead to acute life‐threatening complications such as acute chest syndrome and stroke. Chronic, progressive organ dysfunction that can lead to dramatic complications such as cerebrovascular disease, pulmonary hypertension, and nephropathy is the major determinant of morbidity and mortality in adults [3, 4]. In addition, the progressive splenic atrophy leads to hyposplenia and increased susceptibility to encapsulated bacteria, which may cause fatal infections.

A major clinical challenge is that expression patterns and SCD severity are extremely variable across patients [3]. Many research efforts have been led to identify modifiers of this severity [5]. A large amount of data clearly demonstrated that fetal Hb (HbF) was a major disease modifier, with increased HbF protecting against almost all SCD‐related complications [6, 7]. HbF levels are largely genetically determined, with a few single‐nucleotide polymorphisms (SNPs) responsible for a large part of HbF variance [8]. Moreover, a very high level of HbF (> 30%) can occur in the setting of various molecular alterations, mainly deletions in the β‐globin locus. In these settings, called hereditary persistence of fetal hemoglobin (HPFH), these patients do not have the phenotype of SCD, as no SCD‐related complications occur. The presence of α‐thalassemia also affects the majority of SCD complications, but with variable effects [6]. A few other genetic modifiers have been validated but are specific to some complications [7].

1.5. Clinical Care of Individuals With SCD

In addition to acute complications management, the clinical care of patients with SCD implies various levels of intervention. Infection prevention has been shown to be a major aspect, greatly reducing mortality [9]. Hydroxyurea is the main disease‐modifying therapy, but few other drugs have been approved in some countries, and several are under study. RBC transfusions, mainly as exchange transfusions, are often used in SCD, with the main goal of reducing HbS percentage to prevent its sickling. They can be used to treat acute complications such as stroke or as a chronic transfusion program with regular transfusions to prevent complications. However, none of these approaches are curative. Hematopoietic stem cell transplantation, either with allogenic or genetically corrected autologous cells, is the only curative treatment. Nevertheless, only a few patients benefit from them due to lack of accessibility, poor health condition, and/or lack of suitable donors.

Hemoglobin studies and specifically those dedicated to SCD have been seminal in the field of human genetics and molecular medicine [10]. These models have been serving to apply new technologies and concepts for several decades. This is not currently changing as SCD is a major model for current gene therapys approaches, despite their limitations in terms of accessibility. Moreover, molecular tools can be used for several aspects of SCD management. Indeed, for all the mentioned components of SCD diagnosis and care (prenatal and postnatal diagnosis, prognosis, and transfusion care), molecular testing can play a significant role in SCD and will now be discussed (Figure 1).

FIGURE 1.

Molecular testing uses in sickle cell disease care.

2. Molecular Toolbox in SCD Care

Molecular (or genetic) techniques are heterogeneous in terms of approach and provide different and sometimes not overlapping information. The most frequently used tools in SCD are described below.

Polymerase chain reaction (PCR) has revolutionized molecular biology and medicine. Coincidentally with this review's topic, the initial motivation underlying PCR development was to develop a prenatal test for SCD [11]. In SCD care as in most other settings, the amplification of the targeted fragment allowed by PCR can be used in a wide diversity of approaches. Bidirectional allele‐specific amplification (ASA) produces fragments of different lengths depending on whether the targeted allele is present. The analysis can be performed in a single tube, and the product can be revealed using gel‐based electrophoresis, allowing low‐cost detection of SCD [12]. Alternative detection of the product, such as high‐resolution melting can also be used [13].

Amplification refractory mutation system (ARMS) and PCR‐sequence‐specific primer (PCR‐SSP) are similar to ASA but result in PCR amplification only in the presence of the targeted allele (rs334 in our case). The use of several primers allows discriminating wild type, β^S, and β^C alleles [14]. Allele‐specific oligonucleotides PCR has been used in SCD but is supplanted by the previous ones in most cases. Gap‐PCR is an alternative technique that allows detecting predetermined deletion and can be used for analyzing thalassemia‐related deletions, including HPFH [15].

Restriction fragment length polymorphism (RFLP) is a technique using the capacity of restriction enzymes to cut DNA at specific sites, thus specifically of some alleles. This approach has been used with several restriction enzymes for SCD prenatal testing [16]. Alternative techniques using oligonucleotide probes have been initially developed as prenatal tests for SCD to allow faster testing.

DNA microarray relies on hybridization with multiple probes to determine the genotype at predetermined loci. Their main advantage is the analysis of multiple loci simultaneously. As such, it has been mainly used in genome‐wide association studies, including in SCD, rather than for SCD diagnosis purposes [17].

Multiplex Ligation‐dependent Probe Amplification (MLPA) is an approach using probe sets to analyze copy number variations (and eventually methylation) of selected genes. It is therefore not suitable for detecting single‐nucleotide variants not specifically targeted by probes in the assay, but is highly reliable for deletions (including small ones) as can be found in β‐thalassemia and HPFH [18]. Contrary to gap‐PCR which identifies only preselected deletions, MLPA can identify new deletions [15].

Sanger sequencing is the first method that allowed agnostic analysis of DNA sequence. It is still the gold standard for single‐nucleotide variants and indels, including SCD. However, methods with higher throughput are more frequently used.

Pyrosequencing is a “sequencing by synthesis” method with a higher throughput than Sanger sequencing. While reliable in most of the SCD cases, this technique is not completely accurate given its difficulty to detect single‐nucleotide change [19].

Next‐generation sequencing (NGS) comprises short‐read (second‐generation sequencing) and long‐read (third‐generation sequencing) technologies. Short‐read NGS is nowadays the most frequently used technologies to analyze single‐nucleotide variants, indels, and to a lesser extent, copy number variations. Depending on the capture panel used, it can analyze a selected number of exons of several genes (targeted panel), whole exome, or whole genome sequencing. Of note, previous amplification of DNA using PCR product is usually performed before sequencing. Several of the existing approaches have been reported in SCD, including SNAP‐Seq, an amplicon‐based method developed for β‐globin sequencing [20]. By analyzing unfragmented reads of > 10 kilobases (compared to 200–600 bases with short‐read sequencing), long‐read sequencing are method of choice for analyzing simultaneously single‐nucleotide variants and deletions. Given the complex variations that can occur at the β‐globin locus, long‐read sequencing may be useful to provide a comprehensive analysis.

3. Molecular Testing as Postnatal Newborn Screening and Diagnosis Tool

3.1. Newborn Screening Utility

Newborn screening is one of the cornerstones of SCD care. It has been implemented in most of the countries with the financial and logistic resources available and is expanding in economically disadvantaged countries. Its major benefit is the early introduction of antibiotic prophylaxis and vaccination against encapsulated germs. Indeed, hyposplenia begins as early as the age of two months. The effect of newborn screening on mortality has been well demonstrated to improve specialized care enrollment, decrease hospitalization and pain episodes, and reduce mortality rates [21]. To be effective, newborn screening has to be universal as targeted screening misses some affected newborns.

Newborn screening relies on the same tools as those used later in life for diagnostic purposes. Protein‐based chemistry methods are the main approaches used, but many alternative tools exist or are under study. Molecular diagnosis is an alternative that has several advantages.

3.2. Why Use Molecular Testing: Limitations of Routine Protein‐Based Chemistry and Emerging Approaches

Protein‐based chemistry approaches rely on the detection of each specific Hb type, allowing the quantification or semi‐quantification of each Hb variants. Several techniques exist, the most frequent being gel‐based electrophoresis, high‐pressure liquid chromatography (HPLC; also called high performance liquid chromatography), capillary zone electrophoresis (CZE), isoelectric focusing (IEF), and gel‐based electrophoresis. All these techniques rely on the differences in charge and/or size of each Hb variants to discriminate them and protein‐based analyses are appropriate for most cases of SCD. Identification of HbS variant and exclusion of HbA is sufficient for the diagnosis of SCA and S/β⁰ thalassemia. As well, the presence of HbS variant along with other variant such as Hb C or other Hb variant allows to retain a diagnosis. In other settings such as HPFH or S/β⁺ thalassemia, quantification of the Hb variant is required. They can be used with large number of samples, making them suitable for newborn screening and diagnosis purpose. HPLC and CZE are the most sensitive technique and favored where available, but centers with limited testing routinely rely on CZE and gel‐based electrophoresis [22]. Mass spectrometry approaches are accurate protein‐based approach but require specific expertise and equipment limiting its wide use in newborn screening and diagnosis.

Each protein‐based approach has specific advantages and limitations but shares similar drawbacks [22]. First, there is no means to discriminate potential cells from a donor from patients' own cells. Thus, they should be performed > 3 months after a transfusion to accurately determine the endogenous phenotype. Second, several Hb variants can share the same electric charge and/or size and cannot be distinguished, resulting in potential misdiagnosis [22]. As each approach has its own pattern of resolution and non‐resolution of Hb variants, two approaches are frequently used together to enhance diagnostic accuracy. Third, they are poorly suitable for prenatal/preimplantation diagnosis as they require fetal blood sampling through cordocentesis. While reliable, cordocentesis is a procedure that may be associated with a high rate of fetal loss [23]. Fourth, given that β‐globin expression begins at low levels during the third trimester of pregnancy, such tests have greater limitations in detecting β‐globin variants in premature neonates, which may lead to false negative results. This, combined with the frequent transfusions in this population, limits the accuracy of protein‐based newborn screening in premature neonates. Finally, the Hb variants' kinetics in the first years of life hampers faithful distinction between SCD and HPFH in newborn screening and diagnosis during this period. Indeed, HbF represents most of the Hb present at birth and is then progressively replaced by HbA (or its variant, HbS, depending on the genotype). In SCD, the steady state of residual HbF production is delayed and can occur only after 5–6 years of life. In HPFH, HbF remains high all lifelong. Thus, the Hb patterns of SCD and HPFH are indistinguishable at birth and until several years of life. Given that SCD is a severe disease requiring early intervention while HPFH is a benign variation that does not require any specific care, this is a major limitation of these approaches. Usually, all patients are managed as having SCD until the diagnosis of HPFH can be made. This may lead to unnecessary medicalization and stress for these patients.

3.3. Role of Molecular Testing Approaches for Screening and Diagnosis

The wide variety of molecular tools available can be used for SCD screening and diagnosis with two different purposes: providing a low‐cost, high‐throughput screening method or allowing a comprehensive analysis of β‐globin locus to confer an accurate diagnosis (Figure 2).

FIGURE 2.

Choice of postnatal molecular testing approaches according to clinical need. ARMS, amplification refractory mutation system; ASA, allele‐specific amplification; MLPA, multiplex ligation‐dependent probe amplification; PCR, polymerase chain reaction.

Considering the aim of an easily accessible tool for screening, alternative to protein‐based approaches, PCR‐based techniques are the most frequently used [24]. They provide low‐cost detection of SCD, using either ASA or ARMS [12, 14]. These approaches can be used in both prenatal and postnatal testing [25]. They have also been used for preimplantation testing. Using the appropriate primers and/or control, these approaches can detect β^S and β^C alleles. Despite less being widely used, RFLP has been reported in prenatal testing [16]. This approach is also feasible to diagnose SCD from buccal swab.

Molecular testing can be used to provide an accurate diagnosis by analyzing more comprehensively the β‐globin sequence. Indeed, despite the molecular alteration of SCD being a well‐known substitution, the molecular biology of SCD is complex. The β‐globin locus can carry many alterations, ranging from SNPs to rare variants and including both single base alterations and larger deletions. Depending on the variants present in cis or in trans of the β^s variant, the diagnosis can be different (e.g., SCD vs. HPFH) or the expression can be influenced (e.g., SCA vs. S/β⁺‐thalassemia). Thus, several tools can be used mainly depending on the question, but also on the setting and the resources available. This may require the combination of several techniques. For example, HPFH diagnosis is frequently made by combining deletion analysis by MLPA and single‐nucleotide variant analysis by Sanger or NGS. Long‐read sequencing using nanopore technology is a promising alternative. It has been recently reported as efficient for SCD diagnosis [26]. This approach may also even be suitable for low‐resource settings and allow a comprehensive approach by identifying other variants in the β‐globin gene.

Compared to protein‐based approaches, molecular tests have several advantages. As they are accurate even in case of recent (< 3 months) transfusion, they are technique of choice for newborn screening after intrauterine transfusion or postnatal diagnosis after a transfusion [27]. Moreover, they are not sensitive to the high level of HbF present at birth and impeding accurate discrimination of some Hb variants, and is a method suitable independently of the term of birth. Indeed, molecular testing allows discriminating SCD for HPFH at birth [28]. It can also precisely discriminate S/β⁺‐thalassemia from sickle cell trait [29]. While protein‐based approaches remain the most widely used for newborn screening [30], a two‐step approach could be used to add molecular testing in case of equivocal results and offer accurate diagnosis.

3.4. Molecular Testing for Noninvasive Prenatal Testing

Prenatal testing is widely requested by couples at risk of having a child with SCD. Invasive prenatal testing made on chorionic villus sampling or amniocentesis uses the same approaches as newborn screening and postnatal diagnosis. However, noninvasive prenatal testing using fetal cell‐free DNA (circulating in the maternal bloodstream as early as 8 weeks) requires specific approaches. Indeed, accurate diagnosis of recessive conditions can be challenging [31]. As both maternal and fetal alleles are simultaneously analyzed, it may be difficult to accurately determine the fetal‐specific allele frequency (fetal fraction).

Several NGS‐based approaches have been studied in SCD with various strategies to discriminate the maternal and fetal origin of sequenced alleles. One approach separately determined fetal fractions and sequenced β‐globin locus [32]. While promising, this technique had limited resolution in the case of low fetal fraction (< 4%). Another approach used Unique Molecular Identifiers coupled to relative mutation dosage to infer fetal fraction from a heterozygous mother [33]. It also suffered from difficulties concluding in the case of small fetal fractions. One other approach tested in SCD coupled β‐globin locus analysis with SNP capture with promising results despite not all samples being resolved [34]. Droplet digital PCR has also been studied in the case of SCD and also carried a significant rate of inconclusive result [35].

4. Molecular Testing as a Potential Prognosis Tool

4.1. Single‐Gene Analysis of Prognosis Factors Such as α‐Thalassemia

The α‐globin subunits are encoded by two genes in cis located on chromosome 16. Non‐functional α‐globin results in α‐thalassemia, the severity of which is proportional to the number of genes affected. In individuals with SCD, despite non‐visible on hemoglobin electrophoresis, concomitant inheritance of α‐thalassemia can lower the concentration of HbS as fewer α‐globin chains are available to form Hb tetramers. As the reduced HbS concentration limits its propensity to polymerize, α‐thalassemia could reduce SCD clinical severity. However, the studies are discordant on whether α‐thalassemia reduces SCD severity, and the effect may even be opposite depending on the pathophysiology of complications [6]. Thus, the utility of analyzing α‐thalassemia as a prognostic factor in individuals with SCD remains currently unclear. If performed, this analysis is mainly based on molecular testing investigating deletions that are the most frequent molecular alterations leading to α‐thalassemia, especially in African‐ancestry individuals. Gap‐PCR is widely used to investigate targeted deletions such as the common α‐globin‐α^3.7 deletion [36]. The use of MLPA is also possible and has the advantage of also detecting α‐globin gene duplications (resulting in 3 α‐globin genes in cis) [37]. Other PCR‐ and non‐PCR‐based techniques are also available [38]. For non‐deletional α‐thalassemia, DNA sequencing analysis is the preferred method despite RFLP or PCR‐based approaches such as ARMS having been reported [39].

Other single‐gene analyses could be performed but are not part of routine care. For example, glucose‐6‐phosphate dehydrogenase deficiency (G6PD) is an X‐linked disorder frequently coinherited with SCD and could affect SCD severity [40]. However, this effect is controversial depending on studies [40]. Biochemical analyses for G6PD are the most frequently used, but given the high genotype–phenotype correlation, molecular analysis can be useful. As G6PD deficiencies are mainly due to missense mutations, molecular analyses rely on DNA gene sequencing like Sanger sequencing or NGS.

4.2. Polygenic Scores (PGS)

Complex traits are influenced by a large number of variants, but the individual effect of each of these variants is usually small. Thus, accurate prediction of these traits using genetic data requires integrating the effect of multiple variants. PGS correspond to the sum of all variants (usually SNPs) associated with a phenotype, each variant being weighted by its individual effect. These scores allow refining the risk or the expressivity of many outcomes. Such PGS have been designed in SCD for HbF level and pain episodes [8]. Although PGS are not currently used in clinics, the genotyping of the corresponding SNPs is easy to perform, notably through microarray, and could allow a more personalized medicine for SCD individuals.

4.3. Pharmacogenomics

Pharmacogenomics aims to tailor drug choice, dosage or regiments according to the patient's genotype. A few drugs, like warfarin, currently use this information, but the growing evidence will likely broaden their use. In SCD, current efforts aim to understand the pharmacogenomics of the high variability of HbF increase with the use of hydroxyurea, the most widely used disease‐modifying drug. As for PGS only genotyping of the SNPs of interest is required.

4.4. Somatic Variants

All the genetic analyses presented here target constitutional (germline) variants. However, somatic (acquired) variants stochastically appear in all tissues through life with an age‐related prevalence and can be associated with various phenotypes. Clonal hematopoiesis (CH) is the most studied category of somatic variants and corresponds to the expansion of a hematopoietic stem cell clone carrying a somatic variant conferring a selective advantage [41]. CH is associated with a wide range of outcomes, some positive (such as lower risk of Alzheimer disease) and many negative, such as higher risk of cardiovascular disease, hematological malignancy, and death [41]. Two studies suggest that CH could be more frequent in individuals with SCD, while another did not find an association [42, 43, 44]. The reason for this increased prevalence is unclear but suggests CH could modify SCD phenotype, and further studies are required to investigate this potential biomarker. Nowadays, CH is researched using NGS, the most sensitive approaches using error‐corrected targeted panels. Of note, somatic uniparental disomy may lead to β^S‐globin allele clonal homozygosity in an individual with constitutional heterozygosity [45].

5. Blood Group Molecular Testing

5.1. The Unique Challenge of Alloimmunization in SCD

Alloimmunization occurs when a patient develops antibodies against a foreign antigen (alloantibodies). This can occur if one is exposed to a blood group antigen that is not expressed. Alloimmunization can lead to acute hemolytic transfusion reaction (HTR) occurring within 24 h of transfusion, which is due to a blood group mismatch that is usually detectable by routine blood group analyses, or delayed HTR (DHTR), that can occur up to 30 months and usually due to an alloantibody undetectable at the time of transfusion. Given the risk of severe HTR, alloimmunization is a pervasive consideration in blood transfusion. Alloimmunization poses, however, a unique and major challenges in individuals with SCD. Indeed, it occurs more frequently in this population, can lead to unique complications and may jeopardize the few therapeutic options available in some settings [46].

Historically, alloimmunization has been described to occur in up to 30%–50% of adults with SCD [46, 47, 48]. In spite of improvement in transfusion practice such as using extended phenotypically matched for C, E, and K antigens, alloimmunization still occur at a persistent and significant rate of ~15% [49]. The reason of a persistent high frequency of alloimmunization is multifactorial. The main component is the discrepancy in antigen distribution between African‐ancestry individuals with SCD and non‐African‐ancestry blood donors in Europe and North America. The prevalence of high number of blood antigens, including in the rhesus system, is highly different between African‐ancestry and European‐ancestry populations. The discrepancy is further increased by the high genetic diversity within African‐ancestry populations. In addition, alloimmunization risk may be increased by SCD‐specific factors such as chronic inflammation with recurrent flare‐ups and immune dysfunction.

DHTR are underrecognized in patients with SCD and may be initially misdiagnosed as a pain crisis. In SCD, DHTR can also lead to hyperhemolysis syndrome (“bystander hemolysis”). In this complication, in addition to the transfused RBCs, the patient's own RBCs are destroyed even though they do not express the mismatched antigen. This complication can be life‐threatening and require intensive immunosuppression to safely transfuse other RBCs [49, 50].

The presence of alloimmunization may be a major issue in patients with SCD as it hampers feasibility of further transfusion. In particular, risks of transfusion reaction in an alloimmunized patient is significant, and search and obtention of a matched RBC may represent major challenge for blood banks [50]. As a result, alloimmunized patients may suffer for delay in obtaining RBC and present an increased mortality [51]. Although alloimmunization does not prevent allogeneic or gene‐corrected autologous hematopoietic stem cell transplantation, it increases the complexity of the procedure [47].

Thus, preventing alloimmunization is a crucial part of SCD care. Standard phenotyping approaches may be limited. Therefore, molecular analyses are increasingly used.

5.2. Why Use Molecular Testing: The Limitation of Phenotyping Approach

In addition to standard ABO/RhD analysis, extended antigen profile is recommended in patients with SCD, including at least C/c, E/e, K, Jk^a/Jk^b, Fy^a/Fy^b, M/N, and S/s antigens [49]. This is meant to reduce the alloimmunization incidence in this population as the most common alloantibodies in SCD target these antigens [47]. Serological phenotyping based on hemagglutination is the main approach to analyze these antigens. It is the root of routine ABO/RhD analyses and can be used for virtually any antigens. However, serological analyses have several limitations, especially in SCD population (Table 1). First, although automated assays exist [52], most analyses are manually performed, time‐consuming and retains a subjective component despite standardization. Second, it is unreliable in case of recent (< 3 months transfusion), which occur more frequently in patients with SCD. Third, phenotyping technic may confound two antigens with limited difference but not interchangeable and at risk for alloimmunization. The most frequent examples are RH gene variants, the most frequent being RHD variants. These variants can result in reduced expression of D protein and/or epitope loss (they are sometimes referred to as weak and partial D phenotypes, respectively) [53]. A reduced D phenotype proteins lead to a weak or absent hemagglutination to anti‐D reagent despite the presence of D protein and carrying individuals are not at higher risk of alloimmunization. Variants associated with epitope loss can be detected by anti‐D reagents but nevertheless lead to anti‐D alloimmunization against the missing epitope [53]. RHD gene diversity is important in SCD population as up to 22%–29% of African‐American may carry RHD variants [54]. Serological testing can also be misleading in case of GATA site variants inside the DARC gene. This variant abolishes Duffy glycoprotein with an erythroid‐specific manner but retains the expression on other cells. Thus, patients carrying such variant has Fy(b−) blood group but do not carry an alloimmunization risk against Fy(b+) RBC [55]. This distinction may be clinically relevant in patients with SCD given the high prevalence of Duffy variants in African‐ancestry individuals. Finally, it may be difficult to distinguish an alloantibody from an autoantibody and patients with SCD present an increased risk of autoimmunity, including autoimmune hemolytic anemia.

TABLE 1.

Advantages and limitations of blood group analysis approaches.

Serological protein‐based testing Advantages ‐Low cost ‐Widely accessible ‐Level of analysis relevant to alloimmunization which is driven by protein epitopes Limitations ‐Manually‐performed, time‐consuming testing in most of cases ‐Cannot distinguish transfused from own patients' cells ‐Analysis limited to the epitopes targeted by the reagents: miss epitope loss such as some RHD variants ‐A sufficient epitopes expression is required: miss low expression epitopes such as some RHD variants ‐Analysis limited to RBC: miss epitopes present on other cells and preventing from alloimmunization ‐May be difficult to distinguish autoantibodies from alloantibodies

Molecular testing Advantages ‐More accurate than serological protein‐based testing ‐Automated processing ‐Insensitive to transfusion ‐Identify RHD variants ‐Identify variants resulting in epitopes expression on non‐RBC cells Limitations ‐More expensive ‐Not widely available in routine ‐Require bioinformatics expertise ‐Most tests are targeted on SNPs and miss de novo alterations ‐Most tests do not analyze gene deletions

Abbreviations: RBC, red blood cell; SNPs, single‐nucleotide polymorphism.

Thus, serological testing has limitations that are exacerbated in patients with SCD and may lead to considering molecular testing.

5.3. The Use of Molecular Testing in SCD Care

As most blood group antigens result from SNP [48], genotyping through molecular testing is an easily accessible tool, and a wide variety of molecular approaches can be used that are not specific to SCD. Several low‐throughput PCR‐based approaches are suitable. RFLP has been the first approach to be used but can be somewhat limited by incomplete digestion of PCR products. Many PCR‐SSP primers have been developed, and these tests are now routinely used in many laboratories and can be multiplexed. Other PCR‐based approaches available include real‐time PCR, PCR‐ELISA, PCR‐sequence‐specific oligonucleotide, fluorescent‐based Taq‐Man technology, melting curve analysis, and pyrosequencing. Several high‐throughput DNA‐array systems have also been developed; some are research‐based while others are commercially available. They rely on multiplex PCR and mainly fluorescence‐based automated detection of the products through an array or beads. These arrays can incorporate a large number of SNP, and one version using ~879 000 SNPs allowed a ≥ 95% accurate prediction of blood groups [56]. These targeted approaches only detect selected SNP and cannot identify de novo variants, deletions, or complex rearrangements, and a combination of techniques may be required to provide a comprehensive analysis. NGS with selected enrichment of blood group sequences has also been used, and WGS could provide useful agnostic analysis. However, it cannot be routinely used given the associated cost [57].

In patients with SCD, the use of RBC genotyping has been shown to carry a lower risk of error than serological typing [58], to reduce alloimmunization [54], and to resolve D variants issues [59]. As it allows resolving the various issues mentioned before with phenotyping (Table 1), it is therefore recommended by the American Society of Hematology [49]. It is systematically used in some countries to optimize RBC transfusion in the SCD population [60]. However, as it is not routinely available in all countries and is associated with increased costs, it is only used in selected patients in many settings [50]. Of note, RBC genotyping can be performed on buccal swab.

Molecular testing is a powerful tool, shown to be more efficient than serological testing [61], but may be erroneous in predicting serological blood groups in some situations. For example, contrary to most blood groups that are due to SNPs, many inactivating variants may lead to O group [48]. Thus, accurate prediction of O groups based on DNA requires comprehensive analysis of these variants to avoid erroneous typing.

Finally, molecular testing could be used as a prognosis tool for alloimmunization. Indeed, a genetic predisposition is likely as not all at‐risk patients with SCD experience alloimmunization [62]. Several loci have been associated with alloimmunization, including in HLA, TRIM21, CD81, CTLA4, and Fcγ receptor genes. If further studies confirm and quantify the part of such a genetic component, PCR‐, array‐, or NGS‐based molecular analyses could be used to stratify alloimmunization risk.

6. Conclusion

Molecular analyses provide unique information compared to more widely used techniques such as protein‐based approaches and blood group phenotyping. However, they were not replaced by molecular testing despite their tremendous development in this field over the last decades. Indeed, they provide complementary information to molecular testing and may be less expensive. As no approach is currently both comprehensive and easily accessible, they appear as complementary tools in SCD. Molecular testing is mainly used as complementary testing for selected situations, but it may be used as a first‐line test, for example for newborn screening. The ever‐lowering cost of NGS and the development of long‐read sequencing will provide additional tools for molecular testing in SCD. In parallel, the identification of genetic variations associated with SCD expression may allow using molecular information to tailor patients' management.

Pincez T. and Pastore Y. D., “Molecular Testing in Sickle Cell Disease: From Newborn Screening to Transfusion Care,” International Journal of Laboratory Hematology 48, no. 2 (2026): 295–304, 10.1111/ijlh.14513.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.