Determinants of hydroxychloroquine-mediated hemolytic anemia in G6PD-deficient COVID-19 patients

doi:10.21203/rs.3.rs-1672676/v1

PDF

Research Article

Determinants of hydroxychloroquine-mediated hemolytic anemia in G6PD-deficient COVID-19 patients

https://doi.org/10.21203/rs.3.rs-1672676/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted 19 May, 2022

You are reading this latest preprint version

Aims

To identify the factors contributing to hydroxychloroquine (HCQS)-mediated hemolytic anemia in G6PD deficient COVID-19 patients.

Materials and Methods

We have developed Multiple linear regression (MLR) and classification and regression tree (CART) models based on the published data (n = 13) and evaluated the impact of G6PD variants in exerting hemolysis.

Results

African Ancestry subjects had significant HCQS-mediated drop in hemoglobin (pre- vs. post-therapy Hb, g/dl: 12.75 ± 1.57 vs. 6.78 ± 0.62, p = 0.0008) than those from non-African ancestry (pre- vs. post-therapy Hb, g/dl: 13.27 ± 1.51 vs. 9.97 ± 2.34, p = 0.04). Diabetics had significant HCQS-mediated drop in hemoglobin (pre- vs. post-therapy Hb, g/dl: 12.96 ± 0.98 vs. 6.82 ± 0.76, p = 0.0007) than non-diabetics (pre- vs. post-therapy Hb, g/dl: 12.88 ± 2.13 vs. 9.13 ± 2.51, p = 0.05). MLR model explained 53.75% variability in HQCS-mediated hemoglobin drop when African Ancestry, diabetes, hypertension and azithromycin used as input variables. CART model efficiently explained HQCS-mediated hemolytic anemia by projecting African Ancestry and diabetes as the key predictors (R² = 1.00). Although, African Ancestry G6PD A (-) variant had lesser ligand propensity than G6PD Mediterranean, G6PD Y70H and G6PD Union, it is thermolabile similar to G6PD Orissa, G6PD Y70H and G6PD L235F.

Conclusions

African Ancestry diabetic patients are more prone for HQCS-mediated hemolytic anemia in G6PD deficient COVID-19 patients. MLR and CART models explain 53.75% and 100% variability in HQCS-mediated hemolytic anemia. In view of similar in silico profile of Indian G6PD variants to G6PD A (-), HCQS should be used with caution in diabetics during COVID-19 therapy.

Clinical Pharmacology

Hematology

Glucose-6-phosphate dehydrogenase

Diabetes

African Ancestry

COVID-19

Hydroxychloroquine

Hemolytic anemia

The G6PD (Glucose 6-phosphate dehydrogenase) regulates the pentose phosphate pathway. It reduces NADP + to NADPH, thus regulating the regeneration of glutathione. G6PD deficiency causes hemolytic anemia following infection or the use of certain medications.

The most prominent medications that induce hemolysis are antimalarials and sulfonylureas. G6PD deficiency is an X-linked disorder. G6PD-deficient hemizygous men and homozygous women are prone for oxidative stress-induced hemolysis. Few studies showed 8-aminoquinololine-induced hemolysis even in G6PD heterozygous women [1, 2]. The pharmacological relevance for G6PD deficiency is expanding. G6PD deficient subjects should avoid rasburicase, primaquine, dapsone, pegloticase, and methylene blue [3].

To tackle the COVID-19 pandemic, many hospitals have used drug repurposing. Hydroxychloroquine (HCQS) is one such drug that was effective in tackling this disease. But, many G6PD deficient COVID-19 patients experienced hemolysis following hydroxychloroquine therapy [4–12]. HCQS-therapy in COVID-19 resulted in an increased incidence of methemoglobinemia [13].

Few studies explored whether G6PD deficiency predisposes to COVID-19. The incidence of G6PD deficiency was 33% in Arab COVID-19 patients [14] while its incidence is < 1% in the general population [15]. G6PD deficient COVID-19 patients exhibited the need for mechanical ventilation due to lower PaO2/FiO2 [16]. N-acetyl cysteine reversed the HCQS-induced hemolysis in G6PD deficient COVID-19 [17].

The mechanism of HCQS-mediated hemolysis in G6PD deficient COVID-19 patients is unclear. We have conducted a meta-analysis to delineate the causes of hemolysis. Besides, we explored the G6PD variants that prevalent in our population, which can contribute towards HCQS-mediated hemolytic anemia. We have assessed the impact of these variants using in silico approaches.

Review of Literature

We did a literature review from Pubmed, Medline, and Google Scholar databases using the following key words: "G6PD deficiency", "COVID-19", "hydroxychloroquine" and, "hemolytic anemia". We retrieved demographic, clinical and laboratory details from these studies.

Univariate analysis

We have carried out paired t-test to ascertain HCQS-mediated drop in hemoglobin in each group. The impact of co-morbidities mainly diabetes, hypertension, obesity, renal dysfunction; co-medications mainly azithromycin, sulfa drugs, methylene blue; and ethnic origin was assessed.

Multiple linear regression

We have used www.wessa.net module to develop multiple linear regression (MLR) model. The variables that are found to be relevant based on univariate analysis were included as inputs and HCQS-mediated hemoglobin drop was used as outputs. The contribution of each variable towards the drop in hemoglobin is assessed based on the slope value. The actual drop in hemoglobin is plotted against the MLR-predicted drop in hemoglobin. The ‘R2’ value depicts percentage variability explained by this model.

Classification and regression tree (CART) model

We have used www.bigml.com module to develop CART model. This model was built with the key predictor at the apex of the tree and the subsequent predictors being the branches. This decision tree predicts the response variable by considering non-linear complex relationships.

Evaluation of G6PD variants prevalent in our population

We have recruited 1215 men in the age group of 18–72 year for this study from July 2012 to June 2021. These subjects were healthy and had no clinical implications. All the subjects consented to this study. Institutional Ethical Committee of Sandor Proteomic Pvt Ltd, Hyderabad (EC/SRP/008/2012, dated 26.6.12) approved the study protocol. We have collected whole blood samples in EDTA. None of them had chronic nonspherocytic hemolytic anemia (CNSHA).

Genotyping

We have used QIAamp DNA mini kit (Qiagen, USA) for DNA extraction and Infinium global screening array (GSA) v 1.0 for genotyping. The process involves Genome amplification, fragmentation, hybridization, enzymatic base extension, and fluorescent staining. The iScan system captured the bead chip images. The genome Studio V2011.1 facilitated the analysis and genotype calling. The sample dependent and sample independent controls ensured efficient quality control. We have excluded the markers with an SNP missing call rate of > 0.03 from the analysis.

In silico analysis

We have used SiFT, Provean, SNAP2 and Position-specific evolutionary preservation time (PSEP) to assess the deleterious nature of the identified G6PD variants. The crystal structure of G6PD in complex with structural NADP was used as the template (PDB ID: 6E08) to predict thermal and denaturant stability of G6PD variants using Cologne University Protein Stability Analysis (CUPSAT). Ligand Protein Interaction Comparison and Analysis (LPIcom) computed the propensity score. This score is the probability of interaction of each residue of protein for NADP. The score '0' indicates no probability. The score '9' indicates the highest probability. Galaxy site (http://galaxy.seoklab.org/cgi-bin/submit.cgi?type=SITE) explored ligand binding sites in the G6PD crystal structure. Besides, it modeled tertiary structures of variants. HHsearch facilitated the prediction of the binding ligands based on homology. LigDockCSA predicted the protein-ligand complex structures.

Analysis of published data

We have included 12 case reports of hydroxychloroquine-mediated hemolytic anemia in G6PD-deficient COVID-19 patients published during the period of 2020 – 2021. All were men with the mean age of 49.5 ± 18.3 yr. Out of these 12, seven patients were of African Ancestry (58.3%). Seven were diabetic (58.3%), two had hypertension (16.7%), two had cardiac problems (16.7%), three were obese (25%), two had renal problems (16.7%), five were using azithromycin (41.7%), two were on concomitant therapy with contra-indicated medications (sulfa drugs, methylene blue (16.7%). Two patients expired following the therapy. A 54 yr old man from USA expired with Coomb’s negative hemolysis following co-administration of methylene blue along with hydroxychloroquine and azithromycin. A 60 yr old African American man with acute kidney injury, electrolyte abnormalities, high anion gap, metabolic acidosis expired due to severe hemolytic crisis following HCQS therapy. N-acetyl cysteine was reported to be effective in reversing the hemolytic anemia induced by HCQS and sulfa in one patient. Folate was effective in reversing HCQS-mediated hemolysis in one patient. (Table 1)

Subjects with African Ancestry exhibited significant lowering of hemoglobin following HCQS therapy (pre- vs. post-therapy Hb, g/dl: 12.75 ± 1.57 vs. 6.78 ± 0.62, p=0.0008) than those from those with non-African ancestry (pre- vs. post-therapy Hb, g/dl: 13.27 ± 1.51 vs. 9.97 ± 2.34, p=0.04). Diabetics exhibited significant lowering of hemoglobin following HCQS therapy (pre- vs. post-therapy Hb, g/dl: 12.96 ± 0.98 vs. 6.82 ± 0.76, p=0.0007) than non-diabetics (pre- vs. post-therapy Hb, g/dl: 12.88 ± 2.13 vs. 9.13 ± 2.51, p=0.05). (Fig 1)

Multiple Linear regression

We have used African Ancestry, diabetes, hypertension, and use of azithromycin as the input variables and HCQS-induced hemolytic drop in hemoglobin (Hb) as output variable. This additive model suggests that in G6PD-deficient subjects with African ancestry are more prone for developing hemolytic anemia with 2.04 g/dl drop in hemoglobin following HCQS therapy for COVID-19. Diabetics will have 1.51 g/dl drop in hemoglobin. Hypertension and azithromycin have only minor contribution towards inducing hemolytic anemia. MLR model based on these variables explained 53.75% variability in HCQS-induced hemolytic anemia. (Fig 2) The mathematical representation of this model is shown below:

Classification and Regression Tree (CART) model

G6PD deficiency in African Ancestry subjects was found to be the key predictor of HCQS-mediated hemolytic anemia. The second predictor being the diabetes followed by obesity. Hypertension and azithromycin use being the mild contributors. These variables cumulatively explain all the contributors towards HCQS-mediated hemolytic anemia (R²: 1.00). G6PD deficient African Ancestry subjects with diabetes and obesity exhibited maximum drop in hemoglobin following HCQS-therapy (8.6 g/dl). G6PD deficient African Ancestry non-diabetic subjects exhibited maximum drop in hemoglobin when Azithromycin is co-administered concomitantly (7.3 g/dl). The drop in hemoglobin is lowest in subjects of non-African Ancestry without obesity and concomitant use of Azithromycin (2.4 g/dl). (Fig 3)

G6PD variant analysis in our population

We have used Infinium global screening array to identify the G6PD variants in our population, which are likely to be risk factors for HCQS-mediated hemolytic anemia. This array covers 56 G6PD variants. The SNP missing call rate was 0.0025±0.005. Out of the 1215 men screened, 52 had G6PD deficiency. G6PD Mediterranean (n=19) and G6PD Kerala-Kalyan (n=18) are the most common variants identified. G6PD Orissa was observed in six subjects. G6PD Mahidol, G6PD Viangchan /Jammu, G6PD Chatham were observed in two subjects each. G6PD Y70H, G6PD L235F and G6PD Union were observed in one subject each.

In silico studies (Table 2)

G6PD wild

We have used the crystal structure of G6PD (PDB ID: 6E08) as a reference for the wild protein. The NADP interacts with 38G, 40S, 41G, 42D, 43L, 112Y,141A, 142L, 143P, 170E, 199I, 200D, 201H, 246R, 437Y residues. NADP forms 8 hydrogen bonds i.e., one each with 43L, 112Y, 199I; three with 171K, and two with 201H. (Supplementary Fig 1)

G6PD Mediterranean (S188F)

This is predicted to be deleterious by SIFT, Provean and SNAP2 scores. Its position specific evolutionary conservation (PSEP) time is 220 million years suggesting possibly damaging nature. The ligand propensity score is 9. In the S188F, the NADP binding site is devoid of 38G, 200D, and 246R residues. Hence, there is a disruption of the H-bond with 43L and the formation of new bonds with 40S, 72R, 141A, and 147Y.

G6PD Kerala-Kalyan (E317K)

This is predicted to be deleterious by Provean score. Its PSEP time is 361 million years suggesting possibly damaging nature. The ligand propensity score is 0. In the E317K variant, the NADP binding site is devoid of 200D and 246R. This results in the loss of H-bond with 43L and the formation of new H-bonds with 40S, 42D, and 72R.

G6PD Orissa (A44G)

This is predicted to be deleterious by SIFT, Provean and SNAP2 scores. Its PSEP time is 4200 million years suggesting probably damaging nature. It is found to be thermolabile (DG: -3.79 Kcal/mol). In the A44G variant, the NADP binding site is devoid of 199I, 200D, and 246R residues. This resulted in the loss of H-bonds with 43L, 112Y, 199I and the formation of new H- bonds with the 40S, 72R, 142L, and 147Y.

G6PD Mahidol (G163S)

This is predicted to be deleterious by SIFT, Provean and SNAP2 scores. Its PSEP time is 4200 million years suggesting probably damaging nature. The ligand propensity score is 7. In G163S variant, the NADP binding site is devoid of 199I, 200D, and 246R residues. As a result, there is a loss of H-bonds with 43L and 199I and the formation of new H-bonds with 40S, 72R, 142L, and 147Y.

G6PD Viangchan/Jammu (V291M)

This is predicted to be deleterious by SIFT, Provean and SNAP2 scores. Its PSEP time is 4200 million years suggesting probably damaging nature. The ligand propensity score is 2. In the V291M variant, the NADP binding site is devoid of 199I, 200D, 201H, and 246R residues. There is disruption of H-bonds with 199I and 201H and new H-bond formation with 40S, 42D, 47K, 72R, 142L, and 437Y.

G6PD Chatham (A335T)

This is predicted to be deleterious by SIFT and SNAP2. Its PSEP time is 324 million years suggesting possibly damaging nature. Its ligand propensity score is 2. In the A335T variant, the NADP binding site is devoid of 38G, 40S, 112Y, 200D, and 246R residues. As a result, there is disruption of H-bond at 112Y and formation of new H-bonds with 42D, 72R, 141A, 142L, and 147Y.

G6PD Y70H

This is predicted to be deleterious by SIFT, Provean and SNAP2. Its PSEP time is 361 million years suggesting possibly damaging nature. It has high thermolability (DG: -8.17 Kcal/mol). Its ligand propensity score is 9. In the Y70H variant, the catalytic binding site is devoid of 38G, 40S, 41G, 112Y, 141A, 142L, 143P, and 246R residues. This led to the loss of H-bonds with 43L, 112Y, and 199I and the formation of new H-bonds with 47K and 89K residues.

G6PD L235F

This is predicted as neutral variant by SIFT, Provean and SNAP2. However, its PSEP time is 455 million years suggesting probably damaging nature. It has thermolability (DG: -3.04 Kcal/mol). Its ligand propensity score is 2. In the L235F variant, there is disruption of all the NADP binding sites except for 201H and 437Y residues. This results in the formation of New H-bonds with K47, 51T, 54W, and K205 residues.

G6PD Union (R454C)

This is predicted to be deleterious by SIFT, Provean and SNAP2. Its PSEP time is 4200 million years suggesting probably damaging nature. Its ligand propensity score is 9. In the R454C variant, the NADP binding site is devoid of 199I, 200D, 201H, 246R, and 437Y residues. This leads to disruption of H-bond with 199I and the formation of new H-bonds with 40S, 42D, 72R, 140L, and 142L.

The meta-analysis of the published data on HCQS-mediated hemolytic anemia in G6PD deficient COVID-19 revealed African Ancestry and diabetes as the most significant determinants of hemolytic anemia. This is further confirmed by MLR and CART models. MLR model explained 53.75% variability in the drop of hemoglobin following HCQS therapy by considering African Ancestry, diabetes, hypertension and Azithromycin use as the predictors. CART model demonstrated maximum drop in hemoglobin in G6PD deficient African Ancestry diabetic subjects with obesity followed by G6PD deficient African Ancestry non-diabetic subjects with concomitant administration of Azithromycin. The drop in hemoglobin is lowest in subjects of non-African Ancestry without obesity and concomitant use of Azithromycin. Methylene blue is contraindicated medication in G6PD deficient subjects and hence its use along with HCQS might have triggered severe hemolytic crisis causing death in one patient. The other patient had acute renal failure with significant metabolic acidosis, which might have contributed to death in addition to HCQS-mediated hemolytic anemia.

In view of increased susceptibility of HCQS-mediated hemolytic anemia in African Ancestry G6PD deficient subjects, the possible G6PD variants causing this hemolysis should be explored. The rs1050828 (V68M) variant (A-) is one such variant with high frequency in Africans [18] and African Americans [19]. The allelic effect size of this variant for G6PD activity was − 0.198 (-0.213 to -0.183) [20]. The position-specific evolutionary preservation (PSEP) time for this variant was 1037 millions of years (my) thus suggesting probably damaging nature of this variant consistent with SIFT and SNAP2 scores. This variant has thermolabile nature. This variant is also predominant in certain ethnic groups of India, specifically among Siddis of Karnataka, India [21].

A meta-analysis of 33 studies representing 16003 patients demonstrated increased COVID-19 severity (OR: 2.75, 95% CI: 2.09–3.62) and mortality (OR; 1.90, 95% CI: 1.37–2.64) in diabetic patients [22]. A meta-analysis of 189 studies representing 57563 COVID-19 patients revealed significant lowering of hemoglobin, red blood cell count, and higher ferritin in severe cases compared to moderate cases [23]. Lower hemoglobin levels were observed with older age, higher percentage of diabetes, hypertension and overall comorbidities, consistent with our observation [23]. Another meta-analysis representing 35879 veterans showed greater risk of hospitalization, ICU admission and mortality among diabetic subjects [24]. Diabetics using sulfonylurea had higher risk of hospitalization while those on insulin had higher risk of hospitalization and death [24]. In vitro studies depicted that thiazolidine derivatives inhibit SARS-CoV-2 protease at 0.01 µM [25]. Sitagliptin treatment at the time of hospitalization was reported to reduce mortality with improved clinical outcome in COVID-19 [26]. SGLT2 inhibitors were found to be effective in preventing vital organ damage in COVID-19 [27]. Sulfonyl ureas and pioglitazone may increase the risk for hemolysis in G6PD deficient subjects.

N-acetyl cysteine and folate were effective in minimizing HCQS-mediated hemolysis. N-acetyl cysteine therapy in sickle cell anemia was reported to decrease the percentage of dense cells, increase red cell glutathione and reduce risk for vaso-occlusion [28]. N-acetyl cysteine therapy in beta thalassemia was reported to reduce hemolysis and phagocytosis of RBC by macrophages, increase reduced glutathione content of RBC, platelets and neutrophils thus reducing the oxidative stress [29]. Oral administration of NAC in COVID-19 prevents the severity of COVID-19 by different mechanisms; i) blocking the angiotension-converting enzyme 2, which hampers the penetration of SARS-COV-2 into cells, ii) acting as an anti-oxidant; iii) serving as anti-inflammatory [30]. The efficacy of folate supplementation in reversing the hemolytic anemia in G6PD deficiency was demonstrated earlier also [31].

Our findings projecting S188F, E317K, and A44G as the most frequent G6PD variants in Indian population corroborated with a multi-ethnic study [32]. G6PD A(-) and G6PD Mediterranean variants are reported to cause structural changes in G6PD enzyme inducing conformational instability leading to the loss of binding of one or both substrates [33]. G6PD Mediterranean was reported to cause hemolytic anema in Palestinian children [34]. G6PD Viangchan and G6PD Viangchan + Mahidol were reported to exhibit 10- and 18-folds reduction in NADP + catalytic efficiency [35]. Coinheritance of G6PD Kerala Kalyan with Chuvash polycythemia was shown to cause hemolytic erythrocytosis in a 32-year-old male [36]. The ligand propensity score is maximum in G6PD Mediterranean, G6PD Y70H and G6PD Union, hence likely to have significant reduction in NADP + catalytic efficiency. G6PD Orissa, G6PD Y70H and G6PD L235F variants exhibit thermolabile nature.

African Ancestry G6PD deficient diabetic patients are more prone to developing HCQS-mediated hemolytic anemia following COVID-19. MLR and CART models explained the variability in HCQS-mediated hemoglobin drop by considering ethnic origin, diabetes, hypertension, obesity and azithromycin use. These models help in predicting the likelihood of hemolytic anemia so that early intervention strategies such as N-acetyl cysteine or folate supplementation can be initiated to minimize the life-threatening side effects. We have compared G6PD A (-) variant predominant in subjects with African Ancestry with other G6PD variants prevalent in India. G6PD Mediterranean is the most common G6PD variant that may induce hemolytic crisis in Indian subjects.

Acknowledgements

The authors are grateful to the management of Yoda LifeLine Diagnostics Pvt Ltd, Hyderabad; Sandor Speciality Diagnostics Pvt LTd, Hyderabad; and the Deanship of Scientific Research, King Saud University for funding through Vice Deanship of Scientific Research Chairs for funding this study.

Conflicts of Interest Statement

All the authors hereby declare no conflicts of interest.

Author Contributions

The study was conceptualized and designed by SMN and VKK. The acquisition, analysis and interpretation of the data was carried out by SMN, VKK, TH and SAA. SMN and TH drafted this manuscript. VKK and SAA carried critical revision. The final version was approved by all the authors. All authors agreed to be accountable for all aspects of the work assuring accuracy and integrity of the work.

Chu CS, Bancone G, Nosten F, White NJ, Luzzatto L. Primaquine-induced haemolysis in females heterozygous for G6PD deficiency. Malar J. 2018 Mar 2;17(1):101.
Chu CS, Freedman DO. Tafenoquine and G6PD: a primer for clinicians. J Travel Med. 2019 Jun 1;26(4):taz023.
Belfield KD, Tichy EM. Review and drug therapy implications of glucose-6-phosphate dehydrogenase deficiency. Am J Health Syst Pharm. 2018 Feb 1;75(3):97–104.
De Franceschi L, Costa E, Dima F, Morandi M, Olivieri O. Acute hemolysis by hydroxycloroquine was observed in G6PD-deficient patient with severe COVD-19 related lung injury. Eur J Intern Med. 2020 Jul;77:136–137.
Beauverd Y, Adam Y, Assouline B, Samii K. COVID-19 infection and treatment with hydroxychloroquine cause severe haemolysis crisis in a patient with glucose-6-phosphate dehydrogenase deficiency. Eur J Haematol. 2020 Sep;105(3):357–359.
Maillart E, Leemans S, Van Noten H, Vandergraesen T, Mahadeb B, Salaouatchi MT, De Bels D, Clevenbergh P. A case report of serious haemolysis in a glucose-6-phosphate dehydrogenase-deficient COVID-19 patient receiving hydroxychloroquine. Infect Dis (Lond). 2020 Sep;52(9):659–661.
Aguilar J, Averbukh Y. Hemolytic Anemia in a Glucose-6-Phosphate Dehydrogenase-Deficient Patient Receiving Hydroxychloroquine for COVID-19: A Case Report. Perm J. 2020;24:20.158.
Mastroianni F, Colombie V, Claes G, Gilles A, Vandergheynst F, Place S. Hydroxychloroquine in a G6PD-Deficient Patient with COVID-19 Complicated by Haemolytic Anaemia: Culprit or Innocent Bystander? Eur J Case Rep Intern Med. 2020 Aug 18;7(9):001875.
Aydemir D, Dağlıoğlu G, Candevir A, Kurtaran B, Bozdogan ST, Inal TC, Ulusu NN. COVID-19 may enhance risk of thrombosis and hemolysis in the G6PD deficient patients. Nucleosides Nucleotides Nucleic Acids. 2021;40(5):505–517.
Naymagon L, Berwick S, Kessler A, Lancman G, Gidwani U, Troy K. The emergence of methemoglobinemia amidst the COVID-19 pandemic. Am J Hematol. 2020 Aug;95(8):E196-E197.
Laslett N, Hibbs J, Hallett M, Ghaneie A, Zemba-Palko V. Glucose-6-Phosphate Dehydrogenase Deficiency-Associated Hemolytic Anemia and Methemoglobinemia in a Patient Treated With Hydroxychloroquine in the Era of COVID-19. Cureus. 2021 May 25;13(5):e15232.
Chaney S, Basirat A, McDermott R, Keenan N, Moloney E. COVID-19 and hydroxychloroquine side-effects: glucose 6-phosphate dehydrogenase deficiency (G6PD) and acute haemolytic anaemia. QJM. 2020 Dec 1;113(12):890–891.
Palmer K, Dick J, French W, Floro L, Ford M. Methemoglobinemia in Patient with G6PD Deficiency and SARS-CoV-2 Infection. Emerg Infect Dis. 2020 Sep;26(9):2279–81.
AlJishi JM, Alhajjaj AH, Alkhabbaz FL, AlAbduljabar TH, Alsaif A, Alsaif H, Alomran KS, Aljanobi GA, Alghawi Z, Alsaif M, Al-Tawfiq JA. Clinical characteristics of asymptomatic and symptomatic COVID-19 patients in the Eastern Province of Saudi Arabia. J Infect Public Health. 2020 Dec 17;14(1):6–1.
Warsy AS, El-Hazmi MA. G6PD deficiency, distribution and variants in Saudi Arabia: an overview. Ann Saudi Med. 2001 May-Jul;21(3–4):174–7
Youssef JG, Zahiruddin F, Youssef G, Padmanabhan S, Ensor J, Pingali SR, Zu Y, Sahay S, Iyer SP. G6PD deficiency and severity of COVID19 pneumonia and acute respiratory distress syndrome: tip of the iceberg? Ann Hematol. 2021 Jan 13:1–7.
Ibrahim H, Perl A, Smith D, Lewis T, Kon Z, Goldenberg R, Yarta K, Staniloae C, Williams M. Therapeutic blockade of inflammation in severe COVID-19 infection with intravenous N-acetylcysteine. Clin Immunol. 2020 Oct;219:108544.
da Rocha JEB, Othman H, Tiemessen CT, Botha G, Ramsay M, Masimirembwa C, Adebamowo C, Choudhury A, Brandenburg JT, Matshaba M, Simo G, Gamo FJ, Hazelhurst S; as members of the H3Africa Consortium. G6PD distribution in sub-Saharan Africa and potential risks of using chloroquine/hydroxychloroquin based treatments for COVID-19. Pharmacogenomics J. 2021 Dec;21(6):649–656.
Lo KS, Wilson JG, Lange LA, Folsom AR, Galarneau G, Ganesh SK, Grant SF, Keating BJ, McCarroll SA, Mohler ER 3rd, O'Donnell CJ, Palmas W, Tang W, Tracy RP, Reiner AP, Lettre G. Genetic association analysis highlights new loci that modulate hematological trait variation in Caucasians and African Americans. Hum Genet. 2011 Mar;129(3):307–17.
Shah SS, Macharia A, Makale J, Uyoga S, Kivinen K, Craik R, Hubbart C, Wellems TE, Rockett KA, Kwiatkowski DP, Williams TN. Genetic determinants of glucose-6-phosphate dehydrogenase activity in Kenya.,BMC Med Genet. 2014 Sep 9;15:93.
Devendra R, Gupta V, Biradar SS, Bhat P, Hegde S, Hoti SL, Mukherjee MB, Hegde HV. G6PD A- is the major cause of G6PD deficiency among the Siddis of Karnataka, India. Ann Hum Biol. 2020 Feb;47(1):55–58.
Kumar A, Arora A, Sharma P, Anikhindi SA, Bansal N, Singla V, Khare S, Srivastava A. Is diabetes mellitus associated with mortality and severity of COVID-19? A meta-analysis. Diabetes Metab Syndr. 2020 Jul-Aug;14(4):535–545.
Taneri PE, Gómez-Ochoa SA, Llanaj E, Raguindin PF, Rojas LZ, Roa-Díaz ZM, Salvador D Jr, Groothof D, Minder B, Kopp-Heim D, Hautz WE, Eisenga MF, Franco OH, Glisic M, Muka T. Anemia and iron metabolism in COVID-19: a systematic review and meta-analysis. Eur J Epidemiol. 2020 Aug;35(8):763–773.
Wander PL, Lowy E, Beste LA, Tulloch-Palomino L, Korpak A, Peterson AC, Young BA, Boyko EJ. Risk factors for adverse outcomes among 35 879 veterans with and without diabetes after diagnosis with COVID-19. BMJ Open Diabetes Res Care. 2021 Jun;9(1):e002252.
Petrou A, Zagaliotis P, Theodoroula NF, Mystridis GA, Vizirianakis IS, Walsh TJ, Geronikaki A. Thiazole/ Thiadiazole/ Benzothiazole Based Thiazolidin-4-One Derivatives as Potential Inhibitors of Main Protease of SARS-CoV-2. Molecules. 2022 Mar 28;27(7):2180.
Solerte SB, D'Addio F, Trevisan R, Lovati E, Rossi A, Pastore I, Dell'Acqua M, Ippolito E, Scaranna C, Bellante R, Galliani S, Dodesini AR, Lepore G, Geni F, Fiorina RM, Catena E, Corsico A, Colombo R, Mirani M, De Riva C, Oleandri SE, Abdi R, Bonventre JV, Rusconi S, Folli F, Di Sabatino A, Zuccotti G, Galli M, Fiorina P. Sitagliptin Treatment at the Time of Hospitalization Was Associated With Reduced Mortality in Patients With Type 2 Diabetes and COVID-19: A Multicenter, Case-Control, Retrospective, Observational Study. Diabetes Care. 2020 Dec;43(12):2999–3006.
Fernandez-Fernandez B, D'Marco L, Górriz JL, Jacobs-Cachá C, Kanbay M, Luis-Lima S, Porrini E, Sarafidis P, Soler MJ, Ortiz A. Exploring Sodium Glucose Co-Transporter-2 (SGLT2) Inhibitors for Organ Protection in COVID-19. J Clin Med. 2020 Jun 28;9(7):2030.
Pace BS, Shartava A, Pack-Mabien A, Mulekar M, Ardia A, Goodman SR. Effects of N-acetylcysteine on dense cell formation in sickle cell disease. Am J Hematol. 2003 May;73(1):26–32.
Amer J, Atlas D, Fibach E. N-acetylcysteine amide (AD4) attenuates oxidative stress in beta-thalassemia blood cells. Biochim Biophys Acta. 2008 Feb;1780(2):249–55.
De Flora S, Balansky R, La Maestra S. Rationale for the use of N-acetylcysteine in both prevention and adjuvant therapy of COVID-19. FASEB J. 2020 Oct;34(10):13185–13193.
Pamuk GE, Dogan Celik A, Uyanik MS. Brucellosis triggering hemolytic anemia in glucose-6-phosphate dehydrogenase deficiency. Med Princ Pract. 2009;18(4):329–31.
Sukumar S, Mukherjee MB, Colah RB, Mohanty D. Molecular basis of G6PD deficiency in India. Blood Cells Mol Dis. 2004 Sep-Oct;33(2):141–5.
Sirdah M, Reading NS, Vankayalapati H, Prchal JT. A computational study of structural differences of binding of NADP + and G6P substrates to G6PD Mediterraneanc.563T, G6PD A-c.202A/c.376G, G6PD Cairoc.404C and G6PD Gazac.536A mutations. Blood Cells Mol Dis. 2021 Jul;89:102572.
Sirdah M, Reading NS, Perkins SL, Shubair M, Aboud L, Prchal JT. Hemolysis and Mediterranean G6PD mutation (c.563 C > T) and c.1311 C > T polymorphism among Palestinians at Gaza Strip. Blood Cells Mol Dis. 2012 Apr 15;48(4):203–8.
Boonyuen U, Chamchoy K, Swangsri T, Saralamba N, Day NP, Imwong M. Detailed functional analysis of two clinical glucose-6-phosphate dehydrogenase (G6PD) variants, G6PDViangchan and G6PDViangchan + Mahidol: Decreased stability and catalytic efficiency contribute to the clinical phenotype. Mol Genet Metab. 2016 Jun;118(2):84–91.
Jamwal M, Mallik N, Aravindan AV, Jain A, Sharma P, Malhotra P, Das R. Hemolytic erythrocytosis: an amalgamated phenotype from coinherited Chuvash polycythemia and G6PD Kerala-Kalyan with acquired transient stomatocytosis. Ann Hematol. 2021 Aug;100(8):2107–2109.

Table 1. Studies depicting the association of G6PD deficiency with COVID-19 related complications

Author name	Age (yr)	Gender	Comorbidities	G6PD U/g Hb	Medications	Pre-treatment Hb (g/dl)	Post-treatment Hb (g/dl)	Study findings related to COVID-19 infection in G6PD deficient subjects
Maillart et al, 2020	65	M	Type 2 diabetes	0.2	Hydroxychloroquine, Azithromycin	13.3	7.2	Acute respiratory distress syndrome, acute hemolysis and acute renal failure
Beauverd et al, 2020	68	M	Type 2 diabetes	2.5	Hydroxychloroquine	12.0	6.5	Acute respiratory distress syndrome, acute hemolysis
Franceschi et al, 2020	72	M	Ischemic cardiomyopathy	---	Lopinavir, Hydroxychloroquine	15.0	12.5	Acute hemolysis
Aguilar et al, 2020	51	M	Type 2 diabetes, hypertension, obesity	Deficient	Hydroxychloroquine	14.5	5.9	Acute hemolysis
Mastroianni et al, 2020	32	M	Obesity	<0.2	Hydroxychloroquine	10.2	7.7	Hemolytic anemia. Cessation of HQCS followed by folate supplementation improved hemoglobin levels.
Naymagon et al, 2020	54	M	Diabetes	Deficient	Azithromycin, Hydroxychloroquine, Methylene blue		……	Coomb’s negative hemolysis leading to death
Ibrahim et al, 2020	44	M	Obesity	0.5	H/o hemolytic reaction to sulfa drugs One dose of Hydroxychloroquine	12.6	7.9	Acute hemolysis following hydroxychloroquine therapy. N-acetyl cysteine was effective in blocking the hemolysis and bringing down liver enzymes, CRP and Ferritin in an adult
Palmer K et al, 2020	62	M	Type II diabetes, hypertension	0.8	Amoxicillin/clavulanic acid, heparin, amlodipine, metformin	15.5	5.2	Acute hemolysis. Two blood transfusions, oxygen therapy and folic acid supplementation improved the clinical condition.
Aljishi et al, 2021	50±24	M:F 28:54	22 diabetics	27 G6PD deficient			----	No different in the incidence of G6PD deficiency in symptomatic and asymptomatic patients (32.4% vs. 33.3%)
Youssef et al, 2021	53.5	M:F 3:3	1 diabetic	5.6 (1.8 – 9.3)			8.1 (6.6 -11.0)	Low hemoglobin levels, lowest PaO2/FiO2 and prolonged duration of mechanical ventilation observed in G6PD-deficent COVID-19 patients compared to G6PD-normal COVID-19 patients.
Aydemir et al, 2021	44	M	------	0.91	Favipiravir, convalescent plasma			Thrombocytopenia and lymphopenia following two doses of convalescent plasma. Methyl prednisolone treatment followed by Anakinra improved patient’s signs, symptoms and laboratory findings
Laslett N et al, 2021	60	M	Acute kidney injury, electrolyte abnormalities, high anion gap metabolic acidosis	19.8	Antibiotics, hydroxychloroquine	14.1	6.8	Severe hemolytic crisis followed by death during hydroxychloroquine therapy.
Chaney et al, 2020	57	M	NIDDM, hypertension, hypercholesterolemia, gastro-esophageal reflux disease and Glaucoma	2.8	Piperacillin tazobactam, azithromycin, hydroxychloroquine, ceftriaxone	12.4	6.6	Hemolytic anemia following hydroxychloroquine administration
Pelle MC et al, 2020	86	F	Hypertension, anxiety-depressive syndrome	-----	Hydroxychloroquine, Azithromycin	10.6	8.3	Autoimmune hemolytic anemia followed by myocardial infarction during hydroxychloroquine therapy

Table 2. Comparison of G6PD A(-) with G6PD variants prevalent in India using in silico tools

Canonical	Variant	Provean	SIFT	SNAP2	PSEP time (my)	Thermal stability (DG, Kcal/mol)	Ligand Propensity
G6PD Mediterranean	S188F	-4.22	0.017	87	220	-0.62	9
G6PD Kerala-Kalyan	E317K	-2.68	0.105	23	361	0.67	0
G6PD Orissa	A44G	-2.93	0.001	76	4200	-3.79	0
G6PD Mahidol	G163S	-4.34	0.048	55	4200	0.69	7
G6PD Viangchan	V291M	-2.58	0	92	4200	0.62	2
G6PD Chatham	A335T	-0.84	0.003	89	324	-0.65	2
	Y70H	-4.47	0	95	361	-8.17	9
	L235F	-2.44	0.075	37	455	-3.04	2
G6PD Union	R454C	-7.13	0	87	4200	-11	9
G6PD A (-)	V68M	-1.56	0.007	73	1037	-4.03	2

PSEP: position-specific evolutionary preservation; my: million years

SupplementaryFig.docx

PDF

Version 1

posted 19 May, 2022

You are reading this latest preprint version

Determinants of hydroxychloroquine-mediated hemolytic anemia in G6PD-deficient COVID-19 patients

Status:

Version 1

Abstract

Aims

Materials and Methods

Results

Conclusions

Figures

Introduction

Materials And Methods

Review of Literature

Univariate analysis

Multiple linear regression

Classification and regression tree (CART) model

Evaluation of G6PD variants prevalent in our population

Genotyping

In silico analysis

Results

Discussion

Conclusions

Declarations

References

Tables

Supplementary Files

Status:

Version 1