To the Editors:
Glucose-6-phosphate dehydrogenase (G6PD) deficiency is the most common enzyme defect, being present in more than 400 million people worldwide.1 The global distribution of this disorder is similar to that of malaria, lending support to the so-called malaria protection hypothesis. G6PD deficiency is an X-linked genetic defect resulting from mutations in the G6PD gene, which cause functional variants with many biochemical and clinical phenotypes. Approximately 140 mutations have been described; most are single base changes, leading to amino acid substitutions. In most areas of high prevalence of G6PD deficiency, several polymorphic alleles are found, but in sub-Saharan Africa (SSA), the variant G6PD A- accounts for approximately 90% of G6PD deficiency.2,3
Of the 33 million people in the world infected with HIV, 22 million live in SSA.4 The HIV epidemic in western Europe is now widely related to the flow of migrants from high-burden countries, mainly SSA. In France, patients coming from this area account for 27% of newly diagnosed HIV infection.5
Exposure to sulfa drugs is frequent in HIV-positive patients, because trimethoprim/sulfamethoxazole is the drug most often used for Pneumocystis jiroveci pneumonia prophylaxis and treatment. Either dapsone/trimethoprim or clindamycin/primaquine is effective in the treatment of mild to moderate P. jiroveci pneumonia, but both regimens are contraindicated in G6PD deficiency. Atovaquone should be used in patients who are deficient in G6PD, but this drug is much too expensive for most patients in SSA.6 Moreover, the treatment of toxoplasmosis of the central nervous system is based on a regimen combining pyrimethamin/sulfadiazine.
We studied the prevalence of G6PD deficiency in the cohort of SSA male patients attending the HIV unit and looked retrospectively for hemolytic events when receiving sulfa.
From January 1 to December 31, 2008, all HIV-1-positive male patients coming from SSA attending the department were, after written, informed consent, prospectively tested for G6PD deficiency based on the estimation of enzyme activity by quantitative spectrophotometric analysis of the rate of NADPH production from NADP.1 Pyruvate kinase activity was measured for all G6PD-deficient patients. By analyzing the medical records, we looked retrospectively for hemolytic events, with or without clinical manifestations, when receiving sulfa. The baseline and the end of follow-up were defined, respectively, by the first time the patients attended the department and the time (in 2008) of testing for G6PD deficiency. Hemolysis was classified as certain (anemia, increased bilirubin [without taking simultaneously indinavir or atazanavir], increased lactate dehydrogenase, or increased reticulocytosis, or decreased haptoglobin), possible (two or more biologic abnormalities: anemia, increased bilirubin [with or without taking simultaneously indinavir or atazanavir], increased lactate dehydrogenase, increased reticulocytosis), or very unlikely (no anemia, bilirubin within the normal values, [or increased with taking simultaneously indinavir or atazanavir], and normal lactate dehydrogenase, reticulocytosis, or haptoglobin).
The comparison between G6PD-tested and -nontested patients and the comparison between deficient and nondeficient patients was made by the Mann-Whitney U test, except for percentages of AIDS (Fisher exact P value).
Of all adult HIV-1 patients (1256 patients) from the Internal Medicine Department of Lariboisiere Hospital during the period of January to December 2008, 714 (56.8%) were native of SSA (269 men, 445 women), 471 (38%) were white (383 men, 88 women) (native of western Europe , North Africa ), 26 (2.1%) were native of South America, 24 (1.9%) of the Caribbean, and 19 (1.5%) of Asia. Of the 293 male patients from SSA or the Caribbean, 190 (64.5%) were tested. The 103 (35.5%) remaining patients were checked in private laboratories, missed their appointments, or were lost to follow-up. None of the 293 patients refused to be tested for G6PD deficiency. The 103 nontested patients did not differ from the 190 tested patients with regard to age at study period (41.9 ± 9.3 years versus 42.1 ± 8.2 years, respectively, P = 0.89), duration of previous follow-up (65.5 ± 55.8 months versus 69.9 ± 42.9 months, respectively, P = 0.08), CD4+ T-cell count at the beginning of follow-up (305 ± 243 109/L versus 303 ± 215 109/L, respectively, P = 0.82), CD4+ T-cell count at study period (433 ± 230 109/L versus 432 ± 205 109/L, respectively, P = 0.93), HIV RNA at the beginning of follow-up (4.3 ± 1.1 log10 copies/mL versus 4.5 ± 1.1 log10 copies/mL, respectively, P = 0.35), lines of antiretroviral therapy during previous follow-up (2.1 ± 2.3 versus 2.1 ± 1.9, respectively, P = 0.27), and patients with AIDS at study period (30.1% versus 24.2%, respectively, P = 0.27), except for HIV RNA at the end of follow-up. which was significantly higher in nontested patients (2.8 ± 1.4 log10 copies/mL versus 2.3 ± 1.1 log10 copies/mL, respectively, P = 0.009). This could reflect poorer adherence to treatment attested by more frequent missing of their appointments, explaining the absence of testing for G6PD deficiency.
We identified 27 G6PD-deficient patients, 12.8% of the tested patients. G6PD levels were between 0.1 and 2.3 U/g hemoglobin (normal values, 7-11 U/g hemoglobin) with a mean of 1.2 ± 0.4 and a median of 1.2. Mean age was 45.5 ± 8.9 years (median, 43.0 years; range, 34-63 years). Prevalence of G6PD deficiency was calculated respectively: 2.3% in patients from Cameroon (one of 44), 16.3 % in patients from the Ivory Coast (seven of 43), 14% in patients from the Congo Republic (three of 21), 21.4% in patients from Senegal (three of 14), 7.7 % in patients from Mali (one of 13), and 38.5% in patients from Democratic Republic of the Congo (five of 13). Seventeen of the 27 G6PD deficient patients were exposed to sulfa drugs treatment. Table 1 describes their main clinical and biologic characteristics. Fourteen patients could be assessed for hemolysis based on biologic data. According to the early (during the first week after introduction of sulfa) and late (1 month after introduction of sulfa) evaluations of the biologic parameters, hemolysis was classified as possible in five of the 14 (36%) patients who had previously been challenged by sulfa and whose records were complete, allowing precise analysis. They corresponded to 15 challenges because one patient received TMP-SMX initially and then Pyr-SDZ as treatment for central nervous system toxoplasmosis. Time to occurrence of hemolysis was respectively 11, 15, 20, 34, and 37 days postsulfa. In the other nine patients (64%) who had previously been challenged by sulfa and whose records were complete, corresponding to 10 challenges, hemolysis was very unlikely.
Because a chronic subclinical hemolysis could speed up the course of HIV infection by inducing a permanent, eventually deleterious, activation of the immune system, we then verified that the nondeficient did not differ from deficient patients in any of the following parameters: CD4+ T lymphocytes (432 ± 206 versus 436 ± 199/mm3, P = 0.93) and HIV RNA (4.2 ± 4.9 versus 3.4 ± 4.0 log10/mL, P = 0.39) at the end of follow-up. G6PD-deficient patients were followed up for a significantly longer duration (91 ± 42.6 months versus 66.3 ± 42 months, P = 0.002) and were given more lines of antiretroviral treatment than nondeficient patients (mean 3.0 ± 2.5 versus 2.0 ± 1.8, P = 0.028, Mann-Whitney). Despite we were unable to support the hypothesis of permanent smoldering but deleterious inflammation (by measurements of C-reactive protein, d-dimers, or interleukin-6, for instance) in this retrospective study, it could not be ruled out.
We identified 12.8% of sub-Saharan HIV male patients with G6PD deficiency, in accordance with the high prevalence of this deficiency previously reported in SSA populations.7 Among the very different clinical phenotypes observed, according to functional variant and residual biochemical activity in red blood cell, the World Health Organization Class III G6PD A-(202A376G) is the most common African mutation.3 The 27 G6PD-deficient patients identified were probably A because of their sub-Saharan origin and of their residual enzyme activity of 10% to 40%.3
Among the drugs commonly used for prophylaxis or treatment of opportunistic infections in HIV-infected patients, several are contraindicated (sulfadiazin, sulfamethoxazole, trimethoprim); pyrimethamin alone does not seem to lead to hemolytic anemia.8
We used a retrospective design for this study, because a prospective design (giving trimethoprim/sulfamethoxazole as primary prophylaxis in HIV-infected patients previously identified as G6PD-deficient) seemed to us unethical.
Hemolytic anemia in G6PD A is self-limiting because only the older red blood cells are destroyed and young red blood cells have normal or near-normal enzyme activity. Our results-no acute clinical hemolysis despite severe deficiency with residual enzyme activity-found at least once at less than 1 U/g hemoglobin in 12 of 27 (44%) patients and lack of chronic subclinical hemolysis in most patients continuing to take sulfa were in accordance with relative contraindications of these drugs.8
In conclusion, systematic screening for G6PD deficiency in sub-Saharan HIV male patients does not seem very useful particularly in developing countries where alternative prophylaxis is not available as a result of its cost. However, G6PD deficiency screening could be useful in the Mediterranean region where G6PD variants are World Health Organization Class II and may be sensitive to drugs where those with milder defects (such as G6PD A-) are not.
We thank F. Galacteros, MD, PhD (Hôpital Henri Mondor, Créteil, France) for reading of the manuscript. N.M. was responsible for testing for G6PD deficiency; P.C. was responsible for laboratory testing. N.M., A.R., A.A., B.C., J.E., and J.F.B. conceived and designed the study and wrote the manuscript. B.C. and J.D.M. collected the data. G.S. was the study statistician.
Pierre O. Sellier, MD, PhD*<
Nathalie Mario, MD†
Agathe Rami, MD*
Annalisa Andreoli, MD‡
Beda M. Choho, MD*
Guy Simoneau, MD*
John Evans, MD§
Philippe Chappuis, MD, PhD¶
Jean-François Bergmann, MD, PhD*
*Department of Internal Medicine and Infectious Diseases, Lariboisière Hospital, Paris, France
†Department of Biochemistry A Saint-Antoine Hospital, Paris, France
‡Department of Clinical Haematology, Lariboisière Hospital, Paris, France
§SCOR, Paris la Défense, France
¶Department of Biochemistry, Lariboisière Hospital, Paris, France