Blood transfusion, specifically the transfusion of red blood cells (RBC), significantly contributes to the modern healthcare system. Every day, blood transfusion saves lives in developing countries where acute anaemia caused by malaria, sickle cell disease (SCD), pregnancy-related events and other trauma remains high or is on the rise. For example, in Burkina Faso, around 103 731 RBC units were used in 2017.¹ However, this did not meet the transfusion needs of the country. Moreover, this number is far lower than the theoretical needs of around 196 000–580 000 per the World Health Organization estimation method III (i.e. 1% – 3% of the 19.5 million inhabitants).²

Besides the chronic blood shortage, developing countries also face poor quality of blood products and their unsafe use.³ Indeed, residual risks of transfusion-transmitted infections remain high^4,5 due to inadequacies in blood donor selection and retention and laboratory screening of blood donations. Furthermore, blood transfusion adverse events are underestimated due to the weakness or nonexistence of haemovigilance and quality management systems.^6,7 Finally, although transfusion-transmitted infections and major blood groups matching errors are worrying, blood transfusion safety issues, alloimmunisation and the occurrence of alloantibody are also pressing issues, especially in multi-transfused patients such as those undergoing haemodialysis for chronic renal failure (CRF).^4,8,9,10

Anaemia is highly prevalent in end-stage renal disease patients, often non-regenerative normochromic normocytic anaemia caused by inadequate renal erythropoietin production. Erythropoietin infusions or other erythropoietin-stimulating agents manage anaemia in end-stage renal disease patients. The United States Food and Drug Administration recommends an erythropoietin haemoglobin target range of 100 g/L – 120 g/L¹¹ and expressly states that erythropoietin-stimulating agents should be used to increase haemoglobin only to the level necessary to avoid transfusion.^11,12,13 In 2016, an expert committee advocated for including erythropoietin-stimulating agents in the World Health Organization Model List of Essential Medicines to reduce the need for transfusions in patients with end-stage chronic kidney disease. Erythropoietin-stimulating agents prevent transfusion-related risks, facility requirements, and risk management costs in the event of possible harm (infections, haemosiderosis).¹⁴

However, erythropoietin-stimulating agents treatments are out of reach for most patients in our context. Therefore, RBC transfusion is used to manage CRF-related anaemia and SCD patients. Meanwhile, our country faces poor pre-transfusion compatibility practices; ABO and RhD matching is the only mandatory screening for RBC transfusions. No other blood group is considered, and no alloantibody screening or compatibility test is performed.¹⁵ Given this context, high RBC alloantibodies frequency is expected among transfused patients; however, there is a paucity of data on this. Thus, this study determined the frequency of anti-erythrocyte alloimmunisation and identified alloantibody specificities among CRF multi-transfused patients in Yalgado Ouedraogo Teaching Hospital, Ouagadougou, Burkina Faso.

Our study aimed at determining the frequency and the specificity of alloantibodies among the multi-transfused haemodialysis CRF patients at the teaching hospital Yalgado Ouedraogo of Ouagadougou (Burkina Faso). We found an alloimmunisation rate of 5.9% with antibodies mainly of the anti-Rh blood group antigens specificity.

This study overviews of RBC immunological risks among patients with chronic diseases who are lifelong blood transfusion patients. Although there have been recent changes in the blood transfusion system in Burkina Faso, including the replacement of multiple hospital-based blood banks with a centralised system, standardisation and harmonisation of practices,^6,18,19,20 improved blood collection and infectious disease screening, some improvements towards managing blood recipients are necessary. For example, compactibility screening is still limited to the ABO and RhD antigens, contrary to obtainable standards in high-income countries, where rare groups, at least Rh-Kell major antigens, are screened for before transfusion. Moreover, in Burkina Faso, alloantibody screening tests and laboratory compatibility tests using at least an indirect antibody test as recommended is not implemented: the patient’s plasma and a sample of the RBC units are tested for agglutination on a glass surface.

This study was the first in the country to use the gel column card method, one of the current best methods for alloantibody screening. Nevertheless, the study presents some limitations as complementary antibody identification techniques, such as enzyme-treated red cells reagents panels (papain, bromelain or other) or wide-range panels, were not used. The lack of complementary identification can explain the high rate (4 of 14; 28.7%) of alloantibody undetermined specificity (inconclusive antibody identification).²¹

The overall alloimmunisation rate of 5.9% among CRF patients undergoing haemodialysis on our study is consistent with the findings of two systematic reviews and meta-analysis studies conducted by Ngoma et al. in 2015 and Boateng et al. in 2019. These studies reported an overall alloimmunisation rate of 6.95% and 7.5% in sub-Saharan Africa.^22,23 Our results are similar to those of Kafando et al., who found an alloimmunisation rate of 4.2% among children transfused with Rh-Kell unmatched blood units.²⁴ Alloimmunisation rates in the same range were reported in Uganda (6.1%), Rwanda (6.4%), Sudan (4.0%) and Tanzania (4.1%). However, some reported higher rates: Uganda in 2010 (10.2%), Mali in 2013 (10.3%) and Nigeria in 2015 (9.3%).^{22,25,26,27,28} These results reflect the poor immunological safety of blood transfusions in sub-Saharan Africa, where blood transfusion is performed based only on the blood donor and recipient ABO and RhD antigens matching.

Alloimmunisation rates observed in our study and other studies from sub-Saharan Africa are lower than those observed in Europe and North America when they only screened for ABO and RhD. Prevalences ranged from 18% to 76% in the United Kingdom and United States.^{29,30,31,32,33,34} In France, the rate was about 30%.^35,36 Despite the mandatory donor and recipient Rh-Kell antigens matching before transfusion in these developed countries, alloimmunisation rates in those settings are higher than ours.^34,37,38 This serves as a reminder that the risk of alloimmunisation is multiparametric, depending on the population’s subgrouping or prevalent diseases.^30,36 Also, high rates of alloimmunisation could be due to antigen discrepancies between transfused RBC concentrates collected from donors with European ancestry and SCD recipients who are often of sub-Saharan African descent.^36,37 A similar hypothesis was assumed in some other countries with multi-ethnic groups, such as Iran.³⁹

In Burkina Faso, blood group antigen distribution is established for ABO and RhD within blood donors and patients.^40,41 There is no data about Rh subgroups or other important RBC antigens. It is known that significant differences in the distribution of blood group antigens within the country’s natural ethnic groups could exist. Sawadogo et al. found that the phenotype O was more frequent in the Central-West, Central and East regions corresponding to ‘Mossi’, ‘Gourounsi’, and ‘Gourmantché’ areas, whereas the phenotype A and AB were more prevalent in ‘Boucle du Mouhoun’ and ‘Hauts-Bassins’ regions and the ‘Bwaba’ and ‘Bobo’ areas. The phenotype O negative was infrequent in ‘Bwaba’.⁴⁰ These studies suggest that in Burkina Faso, with more than 50 ethnic groups, dominant blood groups vary between or are specific to particular ethnic groups. Thus, new studies should be conducted to establish blood subgroup frequencies and RBC matching strategies in the country.

In our study, the antibody specificity of four participants of 14 (28.6%) was indeterminate. This impairment could be due to the discrepancy between the European-sourced red cell reagent panel and our population. This situation highlights the need to implement local panels for RBC alloantibodies testing as with some other low- and middle-income countries.^42,43,44 Furthermore, Boateng et al. claim that creating and maintaining a database of phenotyped blood donors will facilitate the selection of matched blood components for emergency transfusions as seen in sub-Saharan Africa and help locally manufacture RBC reagents. Thus, RBC alloantibodies screening may become more economical and sustainable for multi-transfused patients, particularly patients with SCD in this zone.²²

The majority (85.7%) of the alloantibodies found in this study were anti-Rh group antigens. Anti-D and anti-C antibodies accounted for 35.7%, followed by anti-E and anti-C^w. In a previous study of children who received transfusions in Burkina Faso, anti-C and anti-E were the most frequent.²⁴ In our study, the two mainly represented antibodies were co-associated (Anti-D+C) in 4 of 14 patients. The predominance of Rh group antibodies was also reported in some other west African countries, as well as in Côte d’Ivoire,⁴⁵ Mali,²⁶ Senegal⁴⁶ and Nigeria,^27,47 but in these studies, anti-E was the most often encountered when compared to anti-D and anti-C. Surprisingly, we found Rh anti-D antibodies, which could be due to errors occurring during patients’ blood typing. Our hospital has reported as many as 46 blood typing errors yearly (unpublished data). Another hypothesis is that partial RhD antigen carriage is frequent in individuals with African ancestry. In this case, an RhD-positive patient can develop alloantibodies after receiving RhD-positive RBC, as reported by Chou et al.^30,48 The same hypothesis applies to partial C carriers.⁴⁹

This study tried to identify the risk factors associated with alloimmunisation. Neither gender, age, nor the number of blood units received was associated with alloimmunisation in our study. However, Ifeoma et al.⁴⁷ in Nigeria, Senghor et al.⁴⁶ in Senegal and Natukunda et al.⁵⁰ in Uganda have associated these factors with alloimmunisation; the small size of our sample might have prevented the observation of such associations.