Summary: Mother-to-child HIV prevention trials in sub-Saharan Africa use the US National Institutes of Health Division of AIDS (DAIDS) grading scale to monitor hematologic toxicity. A recent study of nevirapine prophylaxis given for 6 months in breast-feeding Zimbabwean infants reported several cases of relative neutropenia in clinically well infants, raising concerns of drug toxicity. However, the DAIDS tables are based on normal blood counts for white infants, although there is evidence that black African infants may have lower absolute neutrophil counts (ANCs) than white infants. To establish normal hematologic values in black Zimbabwean infants and to quantify the apparent prevalence of relative neutropenia in this population, we evaluated HIV-uninfected healthy infants born to HIV-uninfected women at birth, 10 days, 6 weeks, 3, and 4 months of life. A physical examination and blood count were performed at each visit, and an HIV test was performed at the final visit. The ANC values were graded using the DAIDS table. A total of 145 healthy term infants satisfied the inclusion criteria. The mean ANC values for Zimbabwean infants were less than half of the corresponding standard values at all 5 time points (P < 0.0001). Using the DAIDS table in use at the time that the blood was collected, 57% of these healthy infants had relative neutropenia of any grade at birth, followed by 29% at day 10, 53% at 6 weeks, 32% at 3 months, and 37% at 4 months of life. Our data indicate that relative neutropenia exists in healthy black Zimbabwean infants. The guidelines for identifying toxicity were changed in December 2004. However, even by the new DAIDS tables, 43%, 23%, 24%, 42%, and 43% of these healthy babies had relative neutropenia at the time of the 5 visits. Future HIV prevention and treatment trials in sub-Saharan Africa should use normal hematologic values derived from African infants to avoid the overestimation of antiretroviral drug toxicity.
From the *University of Zimbabwe-University of California, San Francisco (UZ-UCSF) Collaborative Research Program in Women's Health, Harare, Zimbabwe; †Department of Pediatrics, Stanford University School of Medicine, Stanford, CA; Departments of ‡Pediatrics and §Obstetrics and Gynecology, University of Zimbabwe College of Health Sciences, Harare, Zimbabwe.
Received for publication May 13, 2005; accepted April 20, 2006.
This work was supported by the Fogarty AIDS International Training and Research Program (award no. D43 TW00003-17).
Presented in part at the XV International AIDS conference, July 11-16, 2004, Bangkok, Thailand (abstract no. TuPpB2040).
Informed consent was obtained from all subjects, and the study was approved by the Institutional Review Board and Human Subjects Committees at Stanford University School of Medicine and at the Medical Research Council of Zimbabwe.
Reprints: Jennifer Wells, MPH, UZ-UCSF Collaborative Program in Women's Health, 15 Phillips Ave, Belgravia, Harare, Zimbabwe (e-mail: email@example.com).
With increasing use of antiretroviral (ARV) agents for preventing mother-to-child transmission (MTCT) of HIV and treatment of HIV-infected children in resource-poor countries, monitoring drug toxicity is a significant challenge to the practicing clinician. US-funded MTCT clinical trials in sub-Saharan Africa use the Division of AIDS (DAIDS) grading scale to monitor hematologic toxicity.1,2 This scale is based on normal values derived predominantly from American infants.3 It is well recognized that children and adults of black African descent can have lower white blood cell (WBC) counts and absolute neutrophil counts (ANCs) compared with children of European-American ancestry.4-9 A study from Malawi showed lower levels of neutrophils at 6 weeks of age among breast-fed HIV-exposed infants who received short-term prophylaxis with nevirapine and zidovudine when compared with healthy infants born to HIV-uninfected mothers; however, these changes were minimal and seemed to have no clinical significance.10 A recent phase-1 and phase-2 study of nevirapine prophylaxis given for 6 months in breast-feeding black Zimbabwean infants reported several cases of relative neutropenia in clinically well infants.11 These low ANCs may have reflected the presence of relative neutropenia in this population rather than ARV drug toxicity. Furthermore, normative data for hematologic values in black African neonates are very limited.12-14 To our knowledge, there are no normative data on hematologic values for black Zimbabwean infants. The objectives of this study were to (1) establish normal ANC values in black Zimbabwean infants and (2) compare the normal values derived from black Zimbabwean infants with those published on American infants and to quantify the prevalence of relative neutropenia in this population.
Zimbabwe is a southern African country of approximately 10.4 million inhabitants whose capital city, Harare, has a population of 1.5 million. The study was conducted at 2 antenatal clinics in Chitungwiza, a satellite community 25 km south of Harare. Written informed consent was obtained from each mother. The protocol was approved by the Institutional Review Boards of Stanford University School of Medicine and the Medical Research Council in Zimbabwe.
Women presenting for antenatal care at the 2 study sites were offered pretest counseling, and those accepting HIV-1 testing were screened for HIV-1 infection using 2 rapid tests, Determine (Abbott Diagnostics, Abbott Park, IL) and Unigold (Trinity Biotech, Jamestown, NY). Women who tested negative for HIV-1 received posttest counseling and were provided information about the study. Women were eligible if they met the following criteria: at least 18 years, HIV-1 negative, and intended to attend well-baby visits at the study sites.
Infants born to enrolled HIV-uninfected women were evaluated at birth, 10 days (days 8-15), 6 weeks (±2 weeks), 3 months (±2 weeks), and 4 months (±2 weeks) of life. At each visit, a physical examination was conducted, and blood was collected for blood count. The birth sample was collected between 12 and 24 hours after delivery, just before discharge, from all the infants except those born on the weekend. Other exclusion criteria included infants born to mothers with complicated pregnancies or deliveries (eg, preeclampsia and premature), infants found to be HIV-infected, or infants whose HIV status at 4 months could not be evaluated. Blood samples of infants included in the study that were drawn at a time they had viral or bacterial infection or were receiving antibiotics were excluded from analysis for that particular visit. Likewise, blood samples of infants who were seen outside the visit window were also excluded from analysis for that particular visit.
Whole blood was collected using EDTA vacutainer tubes and was analyzed using an Abbott 3700 hematology analyzer. The ANC was determined by multiplying the WBC by the percentage of neutrophils. The neutrophil count was corrected for the nucleated red blood cells. The ANC obtained at various time points were graded according to the DAIDS Toxicity Grading Tables (Table 1).1,2
Diagnosis of HIV-1 Infection in Infants
Infant plasma samples obtained at 4 months of age were batched and tested for HIV-1 RNA using the Roche AMPLICOR Monitor test, version 1.5 (Roche Diagnostics Corp, Indianapolis, IN). Version 1.5 of this test includes primers designed to detect all HIV-1 group M subtypes.
Geometric means of the ANC, with 95% confidence intervals, were calculated for each age group, as well as medians and 90% reference interval (5%-95% interpercentile range); t tests were used for comparing the means of the logarithmically transformed data with the logarithm of the published means. For addressing the probable association between test results of the same infants at various times, the Pearson correlation coefficient, r, was computed for each of the 10 pairs of measurements.15 Data were analyzed using SAS statistical software version 8.2 (SAS Institute, Cary, NC).
A total of 220 HIV-1-uninfected women were enrolled antenatally; 37 mothers discontinued study participation before delivery. Of the 183 infants born into the study, 38 were excluded from analysis at all time points for the following reasons: complicated delivery (4), HIV positive (1), and HIV test not available at 4 months (33). A total of 145 healthy infants were included in the study. All 145 infants were born at term, with median birth weight at 3.05 kg and uneventful postnatal period. All infants were found to be HIV-1 uninfected by RNA polymerase chain reaction assay obtained at their visit at 4 months. The ANC values obtained at earlier study visits from infants who were excluded from analysis because of unavailable HIV test at 4 months were not significantly different from those of the included infants. All the children were black. Sex designation was missing from the records of 8 (6%) infants. Of the other 137 infants, 68 were boys and 69 were girls. Of the 725 expected samples, 52 (7%) were not collected (31 weekend births, 21 later missed visits). In addition, 89 collected samples were not used in the analysis for the following reasons: samples from sick infants (72), blood drawn outside the visit window (14), and sample degeneration (3). The analysis was done with 584 samples taken from 145 infants (55, 5 samples; 50, 4 samples; 30, 3 samples; 9, 2 samples; and 1, 1 sample). There were 552 samples with known sex: 53% were from girls, and 47% were from boys (P = 0.002). This phenomenon results from the higher incidence of sick visits among male infants; 68% of the samples excluded because of sick visits were related to males. The differences in the proportion of female samples per visit are small and statistically insignificant.
The number of infants analyzed at each time point is depicted in Table 2. The differences of ANC between boys and girls, as well as the difference from infants with missing sex, were small and statistically insignificant. The geometric mean ANC values for Zimbabwean infants were less than 50% of the excepted normal values for American infants (Table 2). The difference was statistically significant (P < 0.0001) for all the ages at which the comparison was done. Using the 1994 DAIDS table in effect at the time of the study,1 7% of the samples indicated severe/life-threatening relative neutropenia, whereas 41% of the samples indicated relative neutropenia of any grade. The corresponding percentages according to the current guidelines are 10% and 32%. The percentages of infants determined to have neutropenia, according to the 2004 revised DAIDS table,2 are presented by age in Table 2. Overall, 76% of these healthy infants would present with neutropenia in at least one of their visits, and 30% would present with severe/life-threatening neutropenia. Thirteen infants had severe/life-threatening neutropenia in at least half of their visits, 2 of whom had severe/life-threatening neutropenia in all their visits.
The ANC values obtained at various study visits for each infant were weakly correlated (|r| = 0.35). However, the results of the study are only marginally modified when 1 sample is randomly picked from each infant.
Infants with neutropenia pose a diagnostic dilemma to the practicing physician.16 Exposure to ARV prophylaxis, especially zidovudine, is associated with long-term reduced neutrophil counts at all ages in HIV-uninfected children.9 In addition, HIV-exposed infants may develop neutropenia because of drug toxicity from trimethoprim/sulfamethoxazole, often prescribed for prevention of Pneumocystis carinii pneumonia.16,17
We found that healthy black Zimbabwean infants have relative neutropenia that does not seem to decrease during the first few months of life. Numerous adult studies both in the United States and in sub-Saharan Africa have investigated ethnic and sex differences in hematologic parameters.5-7,18-20 Healthy individuals of African ancestry have lower WBC and neutrophil counts than whites.18 Reed and Diehl7 reported that healthy white adults in the United States had significantly greater mean concentrations of leukocytes and neutrophils than African Americans. Nduka et al19 compared the hematologic indices of 512 Africans and 196 individuals of European descent living in the same Nigerian environment for at least 1 year; WBC counts were similar, but a higher neutrophil count was found in the white individuals. Adewuyi et al20 found that white Zimbabwean men had significantly higher WBC counts than black men, and the difference was because of the significantly lower neutrophil counts in the black individuals. A recent study from Uganda showed that ANC declined significantly with age until the age of 13 years, with no differences by sex.8 Another study showed that exposure to ARV prophylaxis in fetal and early life is associated with long-term reduced neutrophil counts in both white and black children; however, black children always had significantly lower neutrophil counts than white children.9
Studies evaluating relative neutropenia among healthy African infants have yielded conflicting results.12 Scott-Emuakpor et al12 determined hematologic values for 402 African neonates and found that African neonates had lower values than their North American and European counterparts. Conversely, Mukiibi et al13 determined the hematologic parameters of 366 Malawian infants; analysis of the differential WBC count showed normal levels of neutrophils similar to those published in standard North American hematology textbooks for neonates of predominantly northern European descent.3 The reasons for the lower baseline neutrophil counts in individuals of African ancestry as compared with their northern European or American counterparts are unclear, although genetic, dietary, or environmental factors have been suggested.6 Whatever the cause, these individuals with relative neutropenia maintain consistently low ANC counts without evidence of increased susceptibility to infection or other adverse events.6
The results of this study clearly support the existence of relative neutropenia in healthy black Zimbabwean infants. It is critical to recognize the existence of relative neutropenia in our population, especially while interpreting low ANC in HIV-exposed infants receiving ARV prophylaxis to prevent MTCT of HIV or in HIV-infected infants receiving chronic treatment, and thus avoid attributing the finding of relative neutropenia to ARV drug toxicity. A larger study is needed to create appropriate population-based normative hematologic values in Zimbabwean infants. Future MTCT clinical trials in Africa should use normal values derived from African infants to avoid the overestimation of ARV toxicity.
The authors thank Dr Mike Simoyi, Dr Mary Bassett, study nurses (Agnes Munhenga, Sylvia Jena, Sabina Chiwara, and Jane Chirwa), study counselors, and all the mothers and infants who participated in the study.
1. National Institute for Allergy and Infectious Disease (NIAID). Division of AIDS toxicity table for grading severity of pediatric adverse experiences. In AIDS Clinical Trials Group (ACTG) toxicity tables, appendix II-B. NIAID; 2004; Bethesda, MD. Available at: http://rcc.tech-res-intl.com
2. National Institutes of Health. Division of AIDS (DAIDS) revised toxicity tables for grading severity of pediatric adverse experiences. US National Institutes of Health DAIDS HIV Vaccine and Research Program, version 1.0; December 2004; Washington, DC. Available at: http://rcc.tech-res-intl.com
3. Brugnara C. Reference values in infancy and childhood. In: Nathan DG, Orkin SH, Ginsburg D, et al, eds. Nathan and Oski's Hematology of Infancy and Childhood
, 6th ed, 2. Philadelphia: WB Saunders, 2003:1848.
4. Sadowitz PO, Oski FA. Differences in polymorphonuclear cell counts between healthy white and black infants: response to meningitis. Pediatrics
5. Rana SR, Castro OL, Haddy TB. Leukocyte counts in 7,739 healthy black persons: effects of age and sex. Ann Clin Lab Sci
6. Haddy TB, Rana SR, Castro O. Benign ethnic neutropenia: what is a normal absolute neutrophil count? J Lab Clin Med
7. Reed WW, Diehl LF. Leukopenia, neutropenia, and reduced hemoglobin levels in healthy American blacks. Arch Intern Med
8. Luganda ES, Mermin J, Kaharuza F, et al. Population-based hematologic and immunologic reference values for a healthy Ugandan population. Clin Diagn Lab Immunol
9. European Collaborative Study. Levels and patterns of neutrophil cell counts over the first 8 years of life in children of HIV-1-infected mothers. AIDS
10. Taha TE, Kumwenda N, Gibbons A, et al. Effect of HIV antiretroviral prophylaxis on hepatic and hematologic parameters of African infants. AIDS
11. Shetty AK, Coovadia HM, Mirochnick MM, et al, For the HIVNET 023 Study Team. Safety and trough concentrations of nevirapine prophylaxis given daily, twice weekly or weekly in breastfeeding infants from birth to 6 months. J Acquir Immune Defic Syndr
12. Scott-Emuakpor AB, Okolo AA, Omene JA, et al. Pattern of leukocytes in the blood of healthy African neonates. Acta Haematol
13. Mukiibi JM, Mtimavalye LAR, Broadhead RL, et al. Some hematological parameters in Malawian neonates. East Afr Med J
14. Mukiibi JM, Nkrumah FK, Kaur M, et al. Neonatal hematology in Zimbabwe. II. The red cell and white cell parameters. Cent Afr J Med
15. Armitage P, Berry G. Statistical Methods in Medical Research. 3rd ed. London, England: Blackwell; 1994:163-165.
16. Boxer LA. Neutrophil abnormalities. Pediatr Rev
17. Israel DS, Plaisance KI. Neutropenia in patients infected with human immunodeficiency virus. Clin Pharm
18. Bain BJ. Ethnic and sex differences in the total and differential white cell count and platelet count. J Clin Pathol
19. Nduka N, Aneke C, Maxwell-Owhochuku S. Comparison of some haematological indices of Africans and Caucasians resident in the same Nigerian environment. Haematologia (Budap)
20. Adewuyi J, Sigola LB, Keogh E. African neutropenia in Zimbabwe. Cent Afr J Med