Lymphocyte subsets in healthy children from birth through 18 ...

doi:10.1016/j.jaci.2003.07.003       Copyright © 2003 American Academy of Allergy, Asthma and Immunology. Published by Mosby, Inc.

Basic and clinical immunology

Lymphocyte subsets in healthy children from birth through 18 years of age*1

The pediatric AIDS clinical trials group P1009 study

William T. Shearer MD, PhD , a, Howard M. Rosenblatt MDa, Rebecca S. Gelman PhDb, c, Rebecca Oyomopito MSb, Susan Plaeger PhDd, e, E. Richard Stiehm MDe, Diane W. Wara MDf, Steven D. Douglas MDg, Katherine Luzuriaga MDh, Elizabeth J. McFarland MDi, Ram Yogev MDj, Mobeen H. Rathore MDk, Wende Levy MSl, Bobbie L. Graham BAm and Stephen A. Spector MDn

a Baylor College of Medicine, Houston, Tex, USAb Harvard School of Public Health, Boston, Mass, USAc Dana-Farber Cancer Institute, Boston, Mass, USAd National Institute of Allergy and Infectious Diseases, Bethesda, Md, USAe University of California, Los Angeles, calif, USAf University of California, San Francisco, calif, USAg Children's Hospital of Philadelphia, Philadelphia, Pa, USAh University of Massachusetts, Worchester, Md, USAi University of Colorado Health Science Center, Denver, Colo, USAj Children's Memorial Hospital, Chicago, Ill, USAk University of Florida Health Science Center, Jacksonville, fla, USAl Social and Scientific Systems, Rockville, Md, USAm Frontier Science and Technology Research Foundation, Amherst, NY, USAn University of California, San Diego, Calif, USA

Received 31 March 2003;  Revised 6 June 2003;  accepted 8 July 2003.  Available online 13 November 2003. Journal of Allergy and Clinical Immunology Volume 112, Issue 5 , November 2003, Pages 973-980

Abstract

Background

Peripheral blood lymphocyte subsets need to be determined in a large, urban, minority-predominant cohort of healthy children to serve as suitable control subjects for the

interpretation of the appearance of these cells in several disease conditions, notably pediatric HIV-1 infection.

Objective

We sought to determine the distribution of lymphocyte subsets in healthy urban-dwelling infants, children, and adolescents in the United States.

Methods

Lymphocyte subsets were determined by means of 3-color flow cytometry in a cross-sectional study of 807 HIV-unexposed children from birth through 18 years of age.

Results

Cell-surface marker analysis demonstrated that age was an extremely important variable in 24 lymphocyte subset distributions measured as percentages or absolute counts—eg, the CD4 (helper) T cell, CD8 (cytotoxic) T cell, CD19 B cell, CD4CD45RACD62L (naive helper) T cell, CD3CD4CD45RO (memory helper) T cell, CD8HLA-DRCD38 (activated cytotoxic) T cell, and CD8CD28 (activation primed cytotoxic) T cell. The testing laboratory proved to be an important variable, indicating the need for using the same laboratory or group of laboratories to assay an individual's blood over time and to assay control and ill or treated populations. Sex and race-ethnicity were much less important.

Conclusion

The results of this study provide a control population for assessment of the effects of HIV infection on the normal development and distribution of lymphocyte subsets in children of both sexes, all races, and all ethnic backgrounds from birth through 18 years of age in an urban population. This study's findings will also prove invaluable in interpreting the immune changes in children with many other chronic diseases, such as primary immunodeficiency, malignancy, rheumatoid arthritis, and asthma.

Author Keywords: Lymphocyte subsets; 3-color flow cytometry; infants; children; adolescents; healthy control subjects

MLE, Maximum likelihood estimate; NIAID, National Institute of Allergy and Infectious Diseases; NICHD, National Institute of Child Health and Human Development; PACTG, Pediatric AIDS Clinical Trials Group; PICL, Pediatric Immunology Core Laboratory; WBC, White blood cell count

Article Outline

• Methods • Study design, eligibility, and sample size (see e-text in online repository at www.mosby.com/jaci) • Flow cytometry • Statistical methods (see e-text in online repository at www.mosby.com/jaci ) • Data exclusions • Subsets used as end points• Results • Comparison of race, ethnicity, and sex of study subjects with patients enrolled in PACTG treatment studies • Change in lymphocyte subset distribution with age (see e-text in online repository at www.mosby.com/jaci) • Regressions of end points on age group, PICL, and other subject and laboratory characteristics (see e-text in Online Repository at www.mosby.com/jaci )• Discussion • Acknowledgements • References

Cell-surface markers (clusters of differentiation) of peripheral blood lymphocytes have provided an extraordinary insight into the ontogeny and state of activation of the human immune system.[1] Attempting to understand the pathogenesis of HIV infection and AIDS, for example, has prompted careful scrutiny of these cellular differentiation pathways in children, which are distinct from those observed in adults with HIV infection. [2, 3 and 4] There is not sufficient information on the peripheral blood lymphocyte subsets in healthy children, particularly information describing cell maturation and activation.

In the study of pediatric HIV infection, the principal immunologic change documented in large, multicenter, longitudinal studies has been the CD4 T-lymphocyte count, which decreased rapidly in infected infants and children[5, 6, 7, 8, 9, 10, 11, 12, 13 and 14] compared with that in HIV-exposed (but HIV-noninfected) control subjects. [15, 16, 17, 18 and 19] Use of these control subjects was justified on the basis of their lack of apparent HIV infection and with the assumption that their immune systems resembled those of absolute control subjects (children without HIV exposure and HIV infection). Some of the existing literature on the early disappearance of CD4 T lymphocytes in HIV-infected children included historical and concurrent control subjects, but such subjects, for the most part, did not match the HIV-infected and HIV-exposed children with regard to socioeconomic factors, racial and ethnicity factors, and maternal hard drug use. [20, 21, 22, 23 and 24] In one study cited by others that did include suitable control subjects, the data collected were limited to CD4, CD8, and CD38 T-lymphocyte and B-cell determinations. [25] Several other studies of various lymphocyte subsets of healthy infants and children have been published, but none were of sufficient size, minority dominance, or age

distribution (ie, birth through 18 years of age) for comparison with HIV-infected children in the United States. [26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42 and 43]

Methods

Study design, eligibility, and sample size (see e-text in online repository at www.mosby.com/jaci)

The Pediatric AIDS Clinical Trials Group (PACTG) P1009 study was an observational, cross-sectional assessment of lymphocyte subsets in HIV-negative, HIV-nonexposed healthy children aged 0 through 18 years of age. To be eligible, the child could not have a diagnosis of HIV infection nor exposure to HIV, acute or chronic infectious disease, any clinically significant disease or findings in the medical history that might compromise the study measures (eg, diabetes mellitus, asthma, rheumatoid arthritis, and cystic fibrosis), be taking prescription medications other than vitamins, or be pregnant or breast-feeding a baby. Allowable medications were restricted to decongestants, antihistamines, cough syrup, vitamins, and nutritional supplements. (We are not aware that these treatments produce changes in lymphocyte subsets in healthy children.)

There were 7 age strata in this study: 0 to 3 months (defined to be birth to 91 days old); 3 to 6 months (ie, 92-182 days old); 6 to 12 months (ie, 183-364 days old); 1 to 2 years (ie, from the first birthday to the day before the second birthday); 2 to 6 years (ie, from the second birthday to the day before the sixth birthday); 6 to 12 years (ie, from the sixth birthday to the day before the 12th birthday); and 12 to 18 years (ie, from the 12th birthday to the day before the 19th birthday). Within each age stratum, the aim was to obtain about 50% female and 50% male children and separately to obtain a racial-ethnic mix similar to the mix of patients entering the PACTG studies. Subject enrollment in the study began in August 1999 and concluded by July 2001 for 6 age strata and in January 2002 for 3- to 6-month-old children. No changes in laboratory performance, as assessed by means of monthly quality assessment programs with blinded specimens, were observed during this time.

The protocol estimated that up to 120 children would need to be entered in each stratum to obtain 90 eligible children with assay results. A total of 851 subjects were entered in the study. The accrual goal of at least 90 subjects' data was reached for all subsets for all age groups, except for 8 subsets in the 3- to 6-month age group, which had between 81 and 89 subjects with usable data.

The accrual goal was chosen so that the 95% CI on the median of each subset percentage would be no wider than about 10% and that the 95% CI on the 5th and 95th percentile for each age group would be no wider than about 15%. However, this calculation assumed the distributions of each subset would be Gaussian (normal) and would have an SD 3 times as large as that for the CD4 T-cell percentage for that age group, and neither of these assumptions turned out to be correct.

Accrual was by PACTG site and Pediatric Immunology Core Laboratory (PICL; see e-text in Online Repository).

Flow cytometry

Table I lists the lymphocyte subsets that were analyzed. Whole EDTA-anticoagulated blood was shipped at room temperature by means of overnight courier to the assigned core laboratory. Complete blood counts and differential counts were performed at the enrolling sites by using the automated instrumentation normally used at that institution. Acceptable samples were stained with a consensus whole-blood lysis technique and (to assure uniformity in reagents) mAbs that were pretitered and formulated in kits specifically designated for this study (BD Pharmingen Biosciences, La Jolla, Calif). Flow cytometer set-up, gating, and marker placement in each core laboratory was dictated by guidelines established by the NIAID Division of AIDS and the advanced flow cytometry working group of the PACTG in collaboration with the technical division of the manufacturer of the flow reagents. [44 and 45] Flow analysis was performed with a standardized 3-color analysis protocol that gated on cells stained with one fluorochrome (generally peridinin chlorophyll protein or CyChrome conjugated to CD45, CD4, or CD8), followed by 2-color analysis of cells stained with the prescribed remaining fluorochromes (FITC and phycoerythrin). The relative proportions of each subset were expressed as the percentage of the anchor gated population.

TABLE I. Peripheral blood lymphocyte subsets measured in the P1009 study

The presence of a surface marker on a cell is assumed unless the marker is followed by a negative sign.

When both positive and negative signs are used, this indicates a total cell population.

Statistical methods (see e-text in online repository at www.mosby.com/jaci)

Percentiles (10th, median, and 90th) of end point distributions are based on percentiles of observed (not transformed) data, not calculated from some distributional assumption.

All variables (age groups, PICL, sex, race-ethnicity, type of red fluorochrome, and immunization in the last 30 days) used in linear regressions of end points were binary (ie, equal to 0 or 1). These linear regressions used ordinary least squares and were unweighted. In the regressions partial F tests were used to test whether all coefficients associated with a particular category of binary variables were 0 (eg, no effect of age) after using the Holm adjustment for the 24 regressions (percentage end points adjusted separately from count end points).

In the first set of regression models, lymphocyte subset percentages were not transformed, and lymphocyte subset counts, white blood cell counts (WBCs), and lymphocyte counts were log base 10 transformed. The normality of regression residuals within age groups was tested by using the Shapiro-Wilks test with the Holm adjustment. Because so few end points had normally distributed residuals, the maximum likelihood estimate (MLE) of the Box-Cox power transformation for each end point was calculated (ie, "left-hand-side-only" transform). The Box-Cox transformation of end point Y is defined as follows:

(yp − 1)/p.All 50 regressions were then rerun by using the best Box-Cox transformation for each end point.

Data exclusions

Subjects were excluded from all analyses if they were ineligible (one patient who had a concurrent upper respiratory infection at the time of blood draw) and from all analyses except for WBCs and lymphocyte counts if they had no lymphocyte subset data at all (37 subjects). For each analyzed subset, subjects were excluded if they had no lymphocyte subset data for the relevant tube (ranged from 3-28 subjects, depending on the marker) or if the PICL did not use the specified fluorochrome for the anchor marker (79-106 subjects depending on the marker, all assayed at one PICL).

Data were also excluded on 3 subjects with unusual WBCs because the intent of the protocol was to exclude patients with infections or other diseases. For this purpose, the "normal" ranges for WBCs for children of various ages were slightly modified from those given in Nathan and Orkin[46 and 47] because it was feared that some racial and ethnic groups might have a wider range of WBCs. (The lower range for each age was taken to be 50% of that in Nathan and Orkin, and the upper range was taken to be Nathan and Orkin's

upper range plus 50% of their lower range). Because there were so few subjects with unusual WBCs, data on 12 patients with no complete blood count results were kept in the analysis of subset percentages.

After these data omissions had been taken into account, depending on the marker, between 676 and 807 subjects had data available for the subset percentages, and between 644 and 800 subjects had data available for the subset counts.

Subsets used as end points

Before analysis, it was decided that analysis would be restricted to 24 lymphocyte subsets (Table I). During analysis, it was decided that WBCs and lymphocyte counts would also be analyzed. Subset counts were obtained by multiplying subset percentages times anchor marker percentages (eg, CD3CD4) of total CD45 lymphocyte population times the absolute lymphocyte count (WBC times lymphocyte percentage obtained by local hematology laboratories).

Results

Comparison of race, ethnicity, and sex of study subjects with patients enrolled in PACTG treatment studies

The accrual goals on the P1009 study were 50% female and 50% male children in each age group and 58% African-American, 28% Hispanic, 12% white, and 2% other race-ethnic group in each age group; these accrual goals were based on the distribution of children entering PACTG treatment studies in 1999. To assess the degree to which these goals were achieved in the context of ongoing recruitment of patients into PACTG protocols, we compared 805 subjects with CD4CD45RACD62L subset results analyzed on P1009 with 993 HIV-positive patients entered into all PACTG protocols during 2001. The results revealed very similar sex and race-ethnicity distributions for P1009 versus PACTG enrollments: for male sex, 52% versus 51%; for African American race-ethnicity, 53% versus 60%; for Hispanic race-ethnicity, 30% versus 29%; for white race-ethnicity, 15% versus 10%; and for "other" race-ethnicity, 2% versus less than 1%, respectively.

Change in lymphocyte subset distribution with age (see e-text in online repository at www.mosby.com/jaci)

In addition to medians, we tabulated the 10th and 90th percentiles because the lowest 5% (or highest 5%) was usually composed of patients assayed at a single laboratory. Twenty-four lymphocyte subset distributions by age are contained in Table II (percentage) and Table III (cell counts per microliter), which also lists WBC and lymphocyte values. Children experience a gradual decrease in the percentage of CD4 T cells from a median of 52% at 0 to 3 months of age to a median of 37% at 6 to 12 years of age but then an increase in the adolescent years to a median of 41% at 12 to 18 years of age ( Table II; Fig E1 [in Online Respository at www.mosby.com/jaci], A and C). Because of the relative

lymphocytosis in infants, the corresponding CD4 T-cell counts remain high for 12 months and then begin to decrease (TABLE II and TABLE III; Fig E1, B and D ).

(110 K)

Application_1.pdf     Acrobat PDF file 1. Lymphocyte subsets in healthy children from birth through 18 years of age: The pediatric AIDS clinical trials group P1009 study

TABLE II. Subset percentages of peripheral blood lymphocytes in healthy children: Distribution by age

See legend to Table I.

Values are the percentages of the cells expressing the indicated markers for the lymphocyte population defined by the anchor marker presented as medians (10th and 90th percentiles). For example, in the 3/4/45RO subset the value reflects the percentage of CD3+ T cells that are also positive for CD4 and CD45RO.

TABLE III. Cell-subset counts of peripheral blood lymphocytes in healthy children: Distribution by age

Values are presented as medians (10th and 90th percentiles). Subset counts (number of cells per microliter × 10−3) were obtained by multiplying subset percentages times anchor marker percentages (ie, CD3CD4 or CD3CD8) of total CD45 lymphocyte population times the absolute lymphocyte count (WBC times lymphocyte percentage). Also, see legends to TABLE I and TABLE II.

There are important changes in percentages of CD8 T cells and CD19 B cells, which each increase dramatically: CD8 T cells from a low (median of 16%) at 3 to 6 months to a maximum (median of 26%) at 12 to 18 years (Table II) and CD19 B cells from a median level of 15% at 0 to 3 months to a sustained plateau (medians of approximately 25%) from 3 months to 2 years of life before returning to a low median of 14% at 12 to 18 years ( Table II).

Also compelling are the changes in distribution of the reciprocal set of maturation CD4 T-cell subsets: the naive helper (CD4CD45RACD62L) and memory helper (CD3CD4CD45RO) subsets (Table II; Fig E2 [in Online Repository at www.mosby.com/jaci], A and B). Reflecting the release from the thymus gland, the CD4CD45RACD62L naive T cells gradually decrease from a high (median of 89%) at 0 to 3 months to a low (median of 51%) at 12 to 18 years. The antigen-committed CD3CD4CD45RO memory T-cell population increases from a low of 8% (median) at 3 to

6 months to a high of 28% (median) at 12 to 18 years. The activated cytotoxic CD8HLA-DRCD38 T-cell population increased to a maximum median percentage of 15% at 1 to 2 years of life and then decreased to a value of 7% at 12 to 18 years (Table II; Fig E2, C ). The activation-primed cytotoxic CD8CD28 T-cell population gradually decreased from a median of 76% at birth to a median of 58% at 12 to 18 years ( Table II; Fig E2, D ).

Regressions of end points on age group, PICL, and other subject and laboratory characteristics (see e-text in Online Repository at www.mosby.com/jaci)

Table IV shows the results of tests of coefficients. Age was highly significant in all count regressions and almost all percentage regressions (and significant with the Holm adjustment in all percentage regressions). Interlaboratory variation was highly significant in most percentage or count regressions. The range of coefficients for the 7 PICL laboratory variables was greater than the range of coefficients for the 7 age groups for 9 of the 24 subset percentage regressions and 11 of the 24 subset log count regressions (eg, CD3 and CD16/56). Sex was seldom significant (significant only for CD4 T-cell percentage and 8/45RA T-cell percentage). Race-ethnicity was also not often significant (only 3/4−/45RO T-cell percentage and CD8, 4/45RA/62L, 8/45RA/62L, 4/45RA, 8/28, and 3/4–/45RO T-cell counts were significant).

TABLE IV. Regression tests cell-subset coefficients for statistical analysis of peripheral blood lymphocyte percentages and counts with respect to age, laboratory, sex, and race-ethnicity

H, Highly significant, P < .0001; S, significant using Holm method but P ≥ .0001; N, not significant after Holm adjustment for multiple comparisons.

The only endpoint for which the test of normality was not rejected for any age group was CD4 T-cell percentage. Because so many regressions did not have Gaussian-distributed residuals, the MLE of the best Box-Cox power transformation of each of the 50 end points was found; these are shown (together with their 95% CIs) in Table E1 (in Online Repository at www.mosby.com/jaci). Only the CD4 percentage CI included 1.0, and therefore only for this endpoint is no transformation best. Only the CD8 T-cell percentage CI included 0.0, and therefore only for this endpoint was the log transformation the best. None of the count end points included the log transformation in the CI. All regressions were rerun by using the Box-Cox MLE transformation to see whether this would qualitatively change any of the significance levels. They did not.

Discussion

In this large-scale cross-sectional study of lymphocyte subsets, there was a large 10th to 90th percentile range for all phenotypes. The age groups with the largest ranges were often the children younger than 1 year of age (9/24 subset percentages and 15/26 subset counts). In the regression analysis for percentages and log counts of the lymphocyte subsets, age groups were highly significant for all subsets. The testing laboratory variables were almost as significant as age groups for all end points, except CD8 T-cell percentage and CD3, CD4, CD8, and CD4/CD38 T-cell counts. In addition, the range of coefficients for laboratory variables could be larger than the range of coefficients for age groups, indicating that the difference between the results of 2 different laboratories analyzing the same blood could be larger than the biggest difference between age groups. Therefore pediatric studies should use the same laboratory for following a patient's results over time and use the same laboratory (or group of laboratories) to assess treatment-induced differences in these lymphocyte subsets. However, it might be possible to use literature values on healthy children in a qualitative way to assess whether the children in other studies started or ended treatment with lymphocyte subset values far out of the literature "normal values."

Sex proved not to be an important variable in our regression analysis, and aside from a few lymphocyte subsets, race-ethnicity was not an important variable. Thus the use of these data on healthy urban children might find application to all children, regardless of sex, race, and ethnic background.

This study represents the largest collection of pediatric data to date that describes the relative distribution of percentages and cell counts for the principal lymphocyte subpopulations in healthy children from birth through 18 years of age. The ranges and medians of lymphocyte subsets in this study might help to formulate hypotheses on the normal ontogeny of immune cells from birth through adolescence. The estimates of power transformations in Table E1 might permit the PACTG and groups studying other pediatric diseases to use smaller sample sizes, because deciding on a good transformation to Gaussian distributions usually takes far more patients than simply testing the difference between 2 groups and to avoid erroneous assumptions that results have a Gaussian distribution (eg, the common log transformation for counts was not a good transformation for any lymphocyte subset counts in this study).

Publication of these data obtained by use of PACTG-certified PICLs will enable investigators enrolling children into clinical trials with highly active retroviral therapy to evaluate immune restoration by comparing lymphocyte subsets with these cells in true control children rather than HIV-exposed children.[48, 49, 50 and 51]

Although this study of normal pediatric lymphocyte populations was designed for use in pediatric HIV infection, it will be of benefit for the assessment of immune suppression in other chronic illnesses of children living in large cities, such as primary immunodeficiency and malignancy associated with bone marrow stem cell transplantation,[52 and 53] rheumatoid arthritis, [54] and asthma. [55] Thus the benefit of this HIV-associated study of immune cells will go far beyond its original design.

Acknowledgements

We thank the investigators, the study staff, and the families who participated in this Pediatric AIDS Clinical Trials Group Study. This study represents a cooperative effort by the American Academy of Asthma, Allergy, and Immunology and the National Institutes of Health Pediatric AIDS Clinical Trials Group (PACTG). Two institutes of the National Institutes of Health participated in this study, the NIAID and the National Institute of Child Health and Human Development (NICHD). Six Pediatric Immunology Core Laboratories (PICLs) funded by NIAID and one by NICHD performed the lymphocyte subset analysis.

References

1. DE Lewis and GR Harriman, Cells and tissues of the immune system. In: RR Rich, TA Fleisher, BL Kotzin, WT Shearer and HW Schroeder, Jr, Editors, Clinical immunology: principles and practice (2nd ed.), Mosby, London (2001), pp. 2.1–2.21.

2. WT Shearer, TC Quinn, P LaRussa et al., Viral load and disease progression in infants infected with human immunodeficiency virus type 1. N Engl J Med 336 (1997), pp. 1337–1342. Abstract-EMBASE | Abstract-MEDLINE   | $Order Document | Full Text via CrossRef

3. AP Kourtis, C Ibegbu, AJ Nahmias et al., Early progression of disease in HlV-infected infants with thymus dysfunction. N Engl J Med 335 (1996), pp. 3431–3436.

4. RL Rabin, M Roederer, Y Maldonado, A Petru and LA Herzenberg, Altered representation of naive and memory CD8 T cell subsets in HlV-infected children. J Clin Invest 95 (1995), pp. 2054–2060. Abstract-MEDLINE | Abstract-EMBASE | Abstract-Elsevier BIOBASE   | $Order Document

5. A d'Arminio Montfore, R Novati, M Galli et al., T-cell subsets and serum immunoglobulin levels in infants born to HlV-seropositive mothers: a longitudinal evaluation. AIDS 4 (1990), pp. 1141–1144.

6. M Martino, PA Tovo, L Galli et al., Prognostic significance of immunologic changes in 675 infants perinatally exposed to human immunodeficiency virus. J Pediatr 119 (1991), pp. 702–709.

7. DS Stein, JA Korvick and SH Vermund, CD4+ lymphocyte cell enumeration for prediction of clinical course of human immunodeficiency virus disease: a review. J Infect Dis 165 (1992), pp. 352–363. Abstract-EMBASE | Abstract-MEDLINE   | $Order Document

8. LM Mofenson, J Bethel, J Moye, Jr, P Flyer and R Nugent, Effect of intravenous immunoglobulin (IVIG) on CD4+ lymphocyte decline in HlV-infected children in a Clinical trial of IVIG infection prophylaxis. J Acquir Immune Defic Syndr 6 (1993), pp. 1103–1113. Abstract-EMBASE | Abstract-MEDLINE   | $Order Document

9. NJ Waecker, Jr, DP Ascher, ML Robb et al., Age-adjusted CD4+ lymphocyte parameters in healthy children at risk for infection with the human immunodeficiency virus. Clin Infect Dis 17 (1993), pp. 23–25.

10. A Viagano, P Stasi, A Salvaggio, P Marchisio and N Principi, Patterns of T lymphocyte changes in early diagnosis of vertically acquired HIV infection. Pediatric AlDS and HIV Infection: Fetus to Adolescent 4 (1993), pp. 192–194.

11. S Plaeger-Marshall, P Hultin, J Bertolli et al., Activation and differentiation antigens on T cells of healthy, at risk, and HIV infected children. J Acquired Immune Defic Syndr 6 (1993), pp. 984–993. Abstract-EMBASE | Abstract-MEDLINE   | $Order Document

12. M Gesner, V Papaevangelou, M Kim et al., Alteration in the proportion of CD4 T lymphocytes in a subgroup of human immunodeficiency virus-exposed-uninfected children. Pediatrics 93 (1994), pp. 624–630. Abstract-EMBASE | Abstract-Elsevier BIOBASE | Abstract-MEDLINE   | $Order Document

13. WV Raszka, Jr, GA Meyer, NJ Waecker et al., Brief reports: variability of serial absolute and percent CD4+ lymphocyte counts in healthy children born to human immunodeficiency virus 1 -infected parents. Pediatr Infect Dis J 13 (1994), pp. 70–73.

14. KC Rich, D Brambilla, J Pitt et al., Lymphocyte phenotyping in infants: maturation of lymphocyte subpopulations and the effects of HIV infection. Clin Immunol Immunopathol 85 (1997), pp. 273–281. Abstract | Abstract + References | PDF (184 K)

15. The European Collaborative Study, Age-related standards for T lymphocyte subsets based on uninfected children born to human immunodeficiency virus 1-infected women. Pediatr Infect Dis J 11 (1992), pp. 1018–1026.

16. RE McKinney and CM Wilfert, Lymphocyte subsets in children younger than 2 years old: normal values in a population at risk for human immunodeficiency virus infection and diagnostic and prognostic application to infected children. Pediatr Infect Dis J 11 (1992), pp. 639–644. Abstract-MEDLINE | Abstract-EMBASE   | $Order Document

17. Y Yanase, T Tango, K Okumura, T Tada and T Kawasaki, Lymphocyte subsets identified by monoclonal antibodies in healthy children. Pediatr Res 29 (1986), pp. 1147–1151. Abstract-EMBASE | Abstract-MEDLINE   | $Order Document

18. SD Douglas, B Rudy, L Muenz et al., Peripheral blood mononuclear cell markers in antiretroviral therapy-naïve HIV infected and high risk HIV seronegative adolescents. AIDS

13 (1999), pp. 1629–1635. Abstract-EMBASE | Abstract-Elsevier BIOBASE | Abstract-MEDLINE   | $Order Document | Full Text via CrossRef

19. SD Douglas, B Rudy, L Muenz et al., T lymphocyte subsets in HIV infected and high risk HIV uninfected adolescents in the REACH cohort: evidence for potent retention of naïve T cell production capacity in adolescents. Arch Pediatr Adolesc Med 154 (2000), pp. 275–280.

20. RP Falcao, SJ Ismael and EA Donadi, Age-associated changes of T lymphocyte subsets. Diagn Clin Immunol 5 (1987), pp. 205–208. Abstract-MEDLINE   | $Order Document

21. GF Babcock, AF Taylor, BA Hynd, RM Sramkoski and JW Alexander, Flow cytometric analysis of lymphocyte subset phenotypes comparing normal children and adults. Design Clin Immunol 5 (1987), pp. 175–179. Abstract-MEDLINE   | $Order Document

22. FM Erkeller-Yuksel, V Deneys, B Yuksel et al., Age-related changes in human blood lymphocyte subpopulations. J Pediatr 120 (1992), pp. 216–222. Abstract-EMBASE | Abstract-MEDLINE   | $Order Document

23. C Ibegbu, TJ Spira, S Nesheim et al., Subpopulations in T and B cells in perinatally HlV-infected and noninfected age-matched children compared with those in adults. Clin Immunol Immunopathol 71 (1994), pp. 7–32.

24. TW McCloskey, T Cavaliere, S Bakshi et al., Immunophenotyping of T lymphocytes by three-color flow cytometry in healthy newborns, children, and adults. Clin Im