Abstract
Background:
Resistance to malaria infection may be conferred by erythrocyte genetic variations including glucose-6-phosphate dehydrogenase (G6PD) deficiency and lack of Duffy antigens. In red blood cell (RBC) transfusion, G6PD deficiency may shorten transfusion survival. Because Duffy-null units are commonly transfused in sickle cell disease (SCD) due to antigen matching protocols, we examined whether Duffy-null donor RBC units have a higher prevalence of G6PD deficiency.
Materials and Methods:
Pediatric patients with SCD on chronic transfusion therapy were followed prospectively for multiple transfusions. RBC unit segments were collected to measure G6PD activity and RBC genotyping. Decline in donor hemoglobin (ΔHbA) following transfusion was assessed from immediate post-transfusion estimates and HbA measurements approximately 1 month later.
Results:
Of 564 evaluable RBC units, 59 (10.5%) were G6PD deficient (23 severe, 36 moderate deficiency); 202 (37.6%) units were Duffy-null. G6PD deficiency occurred in 40 (19.8%) Duffy-null units versus 15 (4.5%) Duffy-positive units (p<0.0001). In univariate analysis, the fraction of Duffy-null RBC units per transfusion was associated with greater decline in HbA (p=0.038); however, in multivariate analysis, severe G6PD deficiency (p=0.0238) but not Duffy-null RBC (p=0.0139) were associated with ΔHbA.
Conclusion:
Selection of Duffy-null RBC units may result in shorter in vivo survival of transfused RBCs due to a higher likelihood of transfusing units from G6PD deficient donors.
Keywords: sickle cell disease, glucose-6-phosphate dehydrogenase deficiency, Duffy antigen
INTRODUCTION
Several genetic polymorphisms that influence the erythrocyte phenotype are believed to have undergone evolutionary selection by conferring resistance to malaria infection. Among these are sickle hemoglobin (HbS), glucose-6-phosphate dehydrogenase (G6PD) deficiency, and variant antigens in blood group systems such as Duffy-null (Fya-b-). These polymorphisms can also influence selection of donor red blood cells (RBC) for transfusion, particularly in sickle cell disease (SCD), and the outcomes of those transfusions.
G6PD deficiency is one of the most common RBC disorders worldwide, affecting approximately 5% of the global population.1 G6PD is essential to protecting the erythrocyte from hemolysis due to oxidative stress. The vast majority of individuals with G6PD deficiency are asymptomatic, thus they may be unaware of their G6PD deficiency status and are as likely to donate blood as G6PD replete individuals.2 However, G6PD deficient RBC units may be more susceptible to RBC storage lesions, resulting in decreased posttransfusion recovery, increased reactive oxygen species accumulation, and post-transfusion hemolysis.3–5
The Duffy antigen is a receptor for circulating cytokines and Plasmodium vivax, allowing parasitic entry to RBCs. Glycophorin antigens in the MNS system are receptors for Plasmodium falciparum. The Duffy-null and glycophorin B-deficient RBC (S-s-U-), phenotypes are found in 68% and 5% of African individuals, respectively, and convey malaria resistance.6,7
Alloimmunized patients with SCD commonly receive prophylactic matching for several RBC antigens including Duffy, Kidd, and S to prevent further alloimmunization. Additionally, units are selected to be HbS-negative for several reasons, including reduced survival of sickle trait RBC after transfusion.8 Previously, we demonstrated that the clearance of donor hemoglobin A was accelerated after transfusion of G6PD deficient units and in SCD patients who received extended antigen matched RBCs. We aimed to determine if donor RBCs with the Duffy-null phenotype or other antigen-negative phenotypes had a higher rate of G6PD deficiency which would impact transfusion efficacy.
METHODS
A prospective, observational study of children with SCD receiving chronic transfusion therapy was conducted at Children’s Healthcare of Atlanta and Children’s National Medical Center.9 All RBC units were HbS negative, pre-storage leukoreduced, and serologically matched for C/c, E/e, and K antigens (limited phenotype matching). For patients with RBC alloimmunization, RBC units were additionally matched for Fya, Jkb, and antigens to which they had alloantibodies (extended phenotype matching). A transfusion episode consisted of 1 – 3 RBC units, based upon clinical protocol for the individual patient.
Segments of donor units were tested for G6PD activity level no later than their date of expiration. The quantitative G6PD assay (Trinity Biotech, Berkeley Heights, NJ) was performed as previously described, in which the normal mean G6PD activity of stored RBC units was 6.5 U/g Hb.10 G6PD deficiency was defined as <60% of the normal mean (<3.9 U/g Hb), while severe G6PD deficiency was <10% of the normal mean activity (<1.3 U/g Hb). Donor genomic DNA was extracted from segments of leukoreduced RBC units, subjected to PCR amplification, and analyzed using the prototype HEA-LR BeadChip assay (Immucor, Norcross, GA) which detects single nucleotide polymorphisms associated with 35 antigens in 11 blood group systems.11 If HEA-LR results were indeterminant, available donor phenotype or genotype records from the blood supplier (LifeSouth Community Blood Centers, Gainesville, FL) were reviewed.
Statistical Analyses
The frequencies of moderate, severe, or no G6PD deficiency in donor RBC were compared to specific RBC minor antigens including Duffy via chi-squared test or Fisher’s exact test, as appropriate for sample size. For each transfusion episode, the number of G6PD deficient units transfused was represented as a fraction of the full transfusion volume (for transfusions of 1, 2, or 3 RBC units), as follows: 0: no deficient units, 1/3: 1 of 3 units deficient, ½: 1 of 2 units deficient, 2/3: 2 of 3 units deficient, 1: all units deficient. The ratio of Duffy-null units and S-negative and U-negative units per transfusion episode was similarly represented. The decline in donor hemoglobin (HbA) after a transfusion episode was determined by estimation of immediate post-transfusion HbA and calculation of HbA prior to the next chronic transfusion episode, as previously described.9 The associations of HbA decline (ΔHbA) with G6PD deficiency and with the Duffy-null phenotype were assessed individually by generalized linear model. Multivariate linear mixed effects modeling of ΔHbA was used to examine G6PD deficiency ratio, Duffy-null ratio, patient splenectomy, and phenotype matching category (limited or extended). Analysis was performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).
RESULTS
Six-hundred RBC units, transfused to 63 patients with hemoglobin SS SCD, were tested for G6PD activity. Of those, 564 (94.0%) units had evaluable RBC antigen phenotype results. While HEA-LR testing failed in 80 (13.0%) units; antigen phenotype or genotype was provided from blood supplier records for 27 units with HEA-LR failure. Additionally, 158 units had indeterminant results for ≥1 antigen in the 35-antigen panel, thus the number of units with phenotype results per specific antigen.
Severe G6PD deficiency was present in 23 (4.1%) units, and moderate G6PD deficiency in 36 (6.4%) units. G6PD deficiency occurred in 40 (19.8%) Duffy-null units versus 15 (4.5%) Duffy-positive units (p<0.0001). G6PD deficiency was also significantly more common in S-negative units (12.8% vs. 5.3%, p=0.019) and U-negative units (66.7% vs. 9.5%, p=0.014). Comparison of G6PD deficiency to other RBC antigen frequencies is shown in Table 1.
Table 1.
Comparison of G6PD activity by RBC minor antigens, in donor RBC units
| Antigen | Severe G6PD deficiency | Moderate G6PD deficiency | Normal G6PD | p-value | |
|---|---|---|---|---|---|
| C | Positive | 1 (1.7%) | 4 (6.7%) | 55 (91.7%) | 0.60 |
| Partial* | 1 (8.3%) | 1 (8.3%) | 10 (83.3%) | ||
| Negative | 19 (4.1%) | 31 (6.7%) | 411 (89.2%) | ||
|
| |||||
| c | Positive | 20 (4.0%) | 34 (6.7%) | 452 (89.3%) | 1.00 |
| Negative | 1 (3.6%) | 2 (7.1%) | 25 (89.3%) | ||
|
| |||||
| E | Positive | 2 (5.1%) | 3 (7.7%) | 34 (87.2%) | 0.76 |
| Negative | 19 (3.8%) | 33 (6.7%) | 443 (89.5%) | ||
|
| |||||
| e | Positive | 21 (4.1%) | 35 (6.9%) | 454 (89.0%) | 0.87 |
| Negative | 0 (0%) | 1 (4.2%) | 23 (95.8%) | ||
|
| |||||
| K | Positive | 0 | 0 | 3 (100%) | 1.00 |
| Negative | 23 (4.1%) | 36 (6.4%) | 500 (89.5%) | ||
|
| |||||
| Jsa | Positive | 4 (9.3%) | 4 (9.3%) | 35 (81.4%) | 0.10 |
| Negative | 19 (3.8%) | 31 (6.1%) | 457 (90.1%) | ||
|
| |||||
| Jsb | Positive | 23 (4.2%) | 36 (6.6%) | 489 (89.2%) | 1.00 |
| Negative | 0 | 0 | 6 (100%) | ||
|
| |||||
| Fy | Positive | 3 (0.9%) | 12 (3.6%) | 320 (95.5%) | <0.0001 |
| Null | 18 (8.9%) | 22 (10.9%) | 162 (80.2%) | ||
|
| |||||
| Jka | Positive | 23 (4.7%) | 33 (6.8%) | 431 (88.5%) | 0.18 |
| Negative | 0 (0%) | 3 (4.8%) | 60 (95.2%) | ||
|
| |||||
| Jkb | Positive | 8 (2.8%) | 16 (5.6%) | 261 (91.6%) | 0.19 |
| Negative | 15 (5.8%) | 17 (6.6%) | 226 (87.6%) | ||
|
| |||||
| M | Positive | 16 (3.8%) | 30 (7.1%) | 379 (89.2%) | 0.45 |
| Negative | 7 (5.3%) | 6 (4.5%) | 120 (90.2%) | ||
|
| |||||
| N | Positive | 15 (3.7%) | 27 (6.7%) | 359 (89.5%) | 0.61 |
| Negative | 8 (5.7%) | 9 (6.4%) | 123 (87.9%) | ||
|
| |||||
| S | Positive | 4 (1.9%) | 7 (3.4%) | 196 (94.7%) | 0.019 |
| Negative | 18 (5.5%) | 24 (7.3%) | 288 (87.3%) | ||
|
| |||||
| s | Positive | 21 (4.2%) | 29 (5.8%) | 447 (89.9%) | 1.00 |
| Negative | 1 (2.5%) | 2 (5.0%) | 37 (92.5%) | ||
|
| |||||
| U | Positive | 22 (4.1%) | 29 (5.4%) | 488 (90.5%) | 0.014 |
| Negative | 0 (0%) | 2 (66.7%) | 1 (33.3%) | ||
Due to indeterminate results on HEA-LR, some units were missing antigen results for 1 or more antigens.
Partial C indicates detection of the RHCE* ce(733G) and (1006T) polymorphisms on at least 1 RHCE allele; the HEA-LR platform is not designed to detect other RHCE variants or partial antigens.
Fy positive indicates Fya and/or Fyb positive.
Twenty (32%) patients received extended phenotype matching (including Fya), although only 1 patient had a previous Duffy antibody (anti-Fya). Patient Duffy genotypes are shown in Table 2. There were 52 (83%) patients with the RBC phenotype Fyb- due to the FY*-67T>C silencing mutation in the promotor GATA box but are not at risk for anti-Fyb formation.
Table 2.
Duffy genotypes of transfused patients with SCD
| FY gene | Number of GATA box mutations | RBC Duffy phenotype | Duffy phenotype of other cells | Number of patients |
|---|---|---|---|---|
| *B/B | 2 | Fy(a−b−) | Fy(a−b+) | 46 (73%) |
| *B/B | 1 | Fy(a−b+) | Fy(a−b+) | 10 (16%) |
| *A/B | 1 | Fy(a+b−) | Fy(a+b+) | 6 (9.5%) |
| *A/B | 0 | Fy(a+b+) | Fy(a+b+) | 1 (1.5%) |
Of 222 transfusion episodes evaluated (160 limited phenotype matched, 62 extended phenotype matched), those with extended matching were more likely to have Duffy-null units as compared to transfusions with limited phenotype matching (59.7% vs. 42.5%, p=0.022). In univariate generalized linear analysis, Duffy-null RBC (as a fraction of Duffy negative units per transfusion episode) was associated with greater decline in HbA post-transfusion (ΔHbA); however when accounting for the effects of G6PD deficiency and other covariates in multivariate analysis (Table 3), Duffy-negative RBC were not independently associated with HbA decline (p=0.139). In the multivariate mixed effects model, HbA decline was associated only with severe G6PD deficiency ratio (p=0.0238) and splenectomy (p=0.0023).
Table 3.
Multivariate linear mixed effects model of post-transfusion HbA decline
| Univariate General Linear Model | Multivariate Mixed Effects Model | |||
|---|---|---|---|---|
| Parameter estimate (g/dL/week) |
p-value | Parameter estimate (g/dL/week) |
p-value | |
| No splenectomy | 0.107 | <0.0001 | 0.096 | 0.0022 |
| Extended RBC antigen matching | 0.066 | 0.0045 | 0.050 | 0.380 |
| Transfusion characteristics (fraction of units/transfusion) | ||||
| Severe G6PD deficiency | 0.267 | 0.0013 | 0.162 | 0.0238 |
| Moderate or severe G6PD deficiency | 0.102 | 0.034 | ||
| Fy-null | 0.066 | 0.038 | 0.055 | 0.139 |
| S-negative | 0.061 | 0.054 | ||
| U-negative | −0.24 | 0.28 | ||
The decline in HbA (ΔHbA) was represented as the change in HbA (g/dL) from immediate post transfusion to the next scheduled transfusion time point, divided by the number of days between the 2 post-transfusion time points. ΔHbA units were then converted to g/dL/week for this analysis. For the multivariate model, only severe G6PD deficiency (not moderate G6PD deficiency) of donor units was introduced to the model.
DISCUSSION
Alloimmunization is a major concern in transfusion therapy in SCD, and accordingly, phenotype matching of all transfusions in SCD for C/c, E/e, K is recommended. Phenotype matching is extended to include Duffy, Kidd, and S antigens after a patient develops RBC alloantibodies.12 As the majority of Black individuals lack both Fya and Fyb expression on RBC, extended phenotype matching for the Fya antigen is more likely to result in selection of RBC units that are Duffy-null. Here we show that because Duffy-null RBC donors are more likely to have G6PD deficiency, RBC units from these donors may have accelerated post-transfusion clearance. While the Duffy-null phenotype is not independently associated with RBC clearance, it is a confounding factor, and selection of Duffy-null donors increases the likelihood of receiving a G6PD-deficient donor unit.
Review of a large transfusion service inventory previously showed G6PD deficiency among only 0.3% of RBC units, but in over 12% of D+C-E- phenotype-selected units.10 Our analysis of antigen matched units selected for chronic transfusion in SCD shows that Duffy-null units are more than 4 times more likely to be G6PD deficient than Duffy-positive units, and that approximately 1 in 5 Duffy-null units transfused to SCD patients may be G6PD deficient. Currently G6PD deficiency is not an exclusion for blood donation. However, G6PD deficient donor RBCs have been shown to have lower 24-hour post-transfusion recovery.3 Further, G6PD deficient units are associated both with an accelerated decline in HbA and subsequent increased reticulocytosis and sickle hemoglobin in SCD patients, demonstrating the detrimental effect of G6PD deficient units on chronic transfusion efficacy for patients with SCD.13
Selecting RBC units for transfusion in SCD has many considerations. Phenotype matching reduces RBC alloimmunization, but may select for units more likely to be from G6PD deficient donors and thus negatively impact transfusion efficacy. Recognizing strong selective pressures from malaria resulting in high prevalence of G6PD deficiency, the Duffy-null phenotype, and other blood group antigen variants, further investigation is warranted to determine the presence of other associations between G6PD deficiency and RBC phenotypes common in donors who are of African ethnicity.
ACKNOWLEDGEMENTS
This work was supported by an unrestricted grant from Immucor. Marianne Yee received funding from the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number 1K23HL146901–01A1.
Footnotes
Conflict of Interest statement: The authors declare no competing financial interests.
REFERENCES
- 1.Nkhoma ET, Poole C, Vannappagari V, Hall SA, Beutler E. The global prevalence of glucose-6-phosphate dehydrogenase deficiency: a systematic review and meta-analysis. Blood Cells Mol Dis. 2009;42(3):267–278. [DOI] [PubMed] [Google Scholar]
- 2.Glucose-6-phosphate dehydrogenase deficiency. WHO Working Group. Bulletin of the World Health Organization. 1989;67(6):601–611. [PMC free article] [PubMed] [Google Scholar]
- 3.Francis RO, D’Alessandro A, Eisenberger A, et al. Donor glucose-6-phosphate dehydrogenase deficiency decreases blood quality for transfusion. J Clin Invest. 2020;130(5):2270–2285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Karafin MS, Francis RO. Impact of G6PD status on red cell storage and transfusion outcomes. Blood Transfus. 2019;17(4):289–295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Tzounakas VL, Kriebardis AG, Georgatzakou HT, et al. Data on how several physiological parameters of stored red blood cells are similar in glucose 6-phosphate dehydrogenase deficient and sufficient donors. Data Brief. 2016;8:618–627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Meny GM. An update on the Duffy blood group system. Immunohematology. 2019;35(1):11–12. [PubMed] [Google Scholar]
- 7.Dankwa S, Chaand M, Kanjee U, et al. Genetic Evidence for Erythrocyte Receptor Glycophorin B Expression Levels Defining a Dominant Plasmodium falciparum Invasion Pathway into Human Erythrocytes. Infect Immun. 2017;85(10). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Osei-Hwedieh DO, Kanias T, Croix CS, et al. Sickle Cell Trait Increases Red Blood Cell Storage Hemolysis and Post-Transfusion Clearance in Mice. EBioMedicine. 2016;11:239–248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yee MEM, Josephson CD, Winkler AM, et al. Hemoglobin A clearance in children with sickle cell anemia on chronic transfusion therapy. Transfusion. 2018;58(6):1363–1371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Francis RO, Jhang J, Hendrickson JE, Zimring JC, Hod EA, Spitalnik SL. Frequency of glucose-6-phosphate dehydrogenase-deficient red blood cell units in a metropolitan transfusion service. Transfusion. 2013;53(3):606–611. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dang JP, Maurice CB, Orengo L, Stack G. Blood Group Genotype Testing on Leukoreduced RBC Segments: Establishing the Feasibility of Hospital-Based Donor Genotyping. AABB; September 2015, 2015; Anaheim, CA. [Google Scholar]
- 12.Chou ST, Alsawas M, Fasano RM, et al. American Society of Hematology 2020 guidelines for sickle cell disease: transfusion support. Blood Adv. 2020;4(2):327–355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sagiv E, Fasano RM, Luban NLC, et al. Glucose-6-phosphate-dehydrogenase deficient red blood cell units are associated with decreased posttransfusion red blood cell survival in children with sickle cell disease. Am J Hematol. 2018;93(5):630–634. [DOI] [PMC free article] [PubMed] [Google Scholar]