Published online August 1, 2008
PEDIATRICS Vol. 122 No. 2 August 2008, pp. 331-339 (doi:10.1542/peds.2007-2308)
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pahwa, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pahwa, S.
Related Collections
Right arrow Infectious Disease & Immunity
Right arrowRelated AAP Red Book topics:
Human Immunodeficiency Virus...
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Facebook   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?

ARTICLE

CD4+/CD8+ T Cell Ratio for Diagnosis of HIV-1 Infection in Infants: Women and Infants Transmission Study

Savita Pahwa, MDa, Jennifer S. Read, MDb, Wanrong Yin, MSc, Yvonne Matthews, MAc, William Shearer, MDd, Clemente Diaz, MDe, Kenneth Rich, MDf, Hermann Mendez, MDg, Bruce Thompson, PhDc for the Women and Infants Transmission Study

a Departments of Microbiology and Immunology, and Pediatrics, University of Miami, Miller School of Medicine, Miami, Florida
b Pediatric, Adolescent, and Maternal AIDS Branch, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland
c Clinical Trials & Surveys Corp, Baltimore, Maryland
d Department of Pediatrics, Baylor College of Medicine, Houston, Texas
e Department of Pediatrics, University of Puerto Rico, San Juan, Puerto Rico
f Department of Pediatrics, University of Illinois at Chicago, Chicago, Illinois
g Department of Pediatrics, State University of New York, Brooklyn, New York


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
OBJECTIVE. In this study, we tested the hypothesis that the CD4+/CD8+ T cell ratio could predict HIV infection status in HIV-exposed infants.

METHODS. CD4+/CD8+ T cell ratios were determined from data for live-born singleton infants who had been prospectively enrolled in the Women and Infants Transmission Study. Data for 2208 infants with known HIV infection status (179 HIV-infected and 2029 uninfected infants) were analyzed.

RESULTS. Receiver operating characteristic curves indicated that the CD4+/CD8+ T cell ratio performed better than the proportion of CD4+ T cells for diagnosis of HIV infection as early as 2 months of age, and this relationship was unaffected by adjustment for maternal race/ethnicity, infant birth weight, gestational age, and gender. At 4 months of age, 90% specificity for HIV diagnosis was associated with 60% sensitivity. For ease of use, graphical estimates based on cubic splines for the time-dependent parameters in a Box-Cox transformation (L), the median (M), and the coefficient of variation (S) were used to create LMS centile curves to show the sensitivity and specificity of CD4+/CD8+ T cell ratios in HIV-infected and uninfected infants until 12 months of age. At 6 months of age, a simplified equation that incorporated sequential CD4+/CD8+ T cell ratios and hematocrit values resulted in improved receiver operating characteristic curves, with 94% positive predictive value and 98% negative predictive value. The positive and negative predictive values remained above 90% in simulated infant populations over a wide range of HIV infection prevalence values.

CONCLUSIONS. In the absence of virological diagnosis, a presumptive diagnosis of HIV infection status can be made on the basis of CD4+/CD8+ T cell ratios in HIV-1-exposed infants after 2 months of age; sensitivity and specificity can be improved at 6 months by using a discriminant analysis equation.

Key Words: HIV diagnosis • perinatal HIV transmission • HIV-exposed infants • CD4+/CD8+ T cell ratio

Abbreviations: MTCT—mother-to-child transmission • LMS—least-mean squares • PCR—polymerase chain reaction • ROC—receiver operating characteristic • WITS—Women and Infants Transmission Study

Mother-to-child transmission (MTCT) of the virus is the main cause of pediatric HIV infection and occurs in ~25% of infants born to HIV-infected women who do not receive interventions for prevention of MTCT.1 In many parts of the developing world where breastfeeding is normative, postnatal transmission of HIV contributes to alarmingly high rates of MTCT.2 Many HIV-1-infected pregnant women do not have access to programs for prevention of MTCT, and their infants remain at high risk of infection.3 Regardless of the availability of programs for prevention of MTCT, determination of the HIV infection status of infants born to HIV-infected women is critically important for proper care of both HIV-infected and uninfected infants.4–6

HIV nucleic acid amplification assays, including DNA polymerase chain reaction (PCR) assays, represent the standard for the diagnosis of HIV infection among HIV-exposed infants. For example, HIV DNA PCR assays have >99% sensitivity and specificity at 6 weeks, even in resource-poor settings.7 However, HIV nucleic acid testing is not available in many resource-poor areas of the world where there is a large burden of HIV infection. In these settings, serological testing may be the only type of testing available, resulting in delays in the diagnosis of HIV-exposed infants. Such delays are attributable to the fact that infection status cannot be established until seroreversion after decay of passively transferred maternal antibodies in the case of uninfected children or until development of HIV-associated clinical symptoms or documented persistence of anti-HIV antibodies beyond 18 months (the cutoff point for decay of maternal antibodies) in the case of HIV-infected children.6 Until appropriate HIV virological assays for diagnosis of HIV infection become universally available, it is critical to develop alternative diagnostic algorithms for perinatally HIV-exposed infants, for diagnosis as soon after birth as possible. Several approaches are currently under investigation.6,8,9 The present study addresses this problem from an immunologic perspective and evaluates whether the CD4+/CD8+ T cell ratio may be a useful diagnostic tool for HIV-exposed infants.

Clinical sites in many parts of the world, although lacking HIV nucleic acid testing facilities, have access to T cell subset analysis. Typically, CD4+ T cell levels are used to monitor the degree of immunosuppression in HIV-infected infants, children, and adults.10 Concurrent with the loss of CD4+ T cells, a predominant effect of HIV infection is expansion of CD8+ T cells,11,12 which occurs as a component of the immune activation that is a pathogenic feature in HIV infection. We hypothesized that an inverse CD4+/CD8+ T cell ratio would likely be a more-sensitive discriminator of HIV infection status than depletion of CD4+ T lymphocytes alone.

In the present study, we analyzed CD4+/CD8+ T cell ratios for a prospective cohort of infants with perinatal HIV exposure for whom infection status had been established with HIV-1 DNA PCR assays. The objectives of our study were (1) to determine whether the CD4+/CD8+ T cell ratio in HIV-exposed infants could discriminate between infected and uninfected infants, (2) to determine the earliest age at which such discrimination could be accomplished, and (3) to determine whether the discriminating power could be enhanced by using additional simple laboratory measures.


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Women and Infants Transmission Study
The Women and Infants Transmission Study (WITS) was a prospective cohort study of HIV-infected women and their children conducted at multiple sites in the United States (Boston and Worcester, MA; Houston, TX; Chicago, IL; and New York City, NY) and San Juan, Puerto Rico. The design and objectives of this study were described previously.13 Briefly, beginning in 1989, HIV-infected pregnant women were screened and recruited into the study, which followed a standard, institutional review board-approved protocol. Each enrolled woman provided written informed consent for her participation and that of her child. Diagnostic testing of each infant was performed according to a fixed schedule (at birth and at 1, 2, 4, 6, 9, and 12 months of age). The diagnosis of HIV infection in children required 2 positive HIV-1 DNA PCR assays. Early in the WITS, 3 negative assays were required to declare a child uninfected, with ≥1 test being performed after the age of 30 days. This rule was later changed to 2 negative assay results, with 1 occurring after 30 days of age and 1 occurring after 4 months of age. Flow cytometry was performed by using standardized monoclonal antibody panels (Becton Dickinson, San Jose, CA) at National Institutes of Health-certified study laboratories.14 The natural history of CD4+ and CD8+ T cell subsets in the WITS cohort was reported previously.15

Study Population for Analysis
The study population for this analysis was restricted to live-born infants with known HIV-1 infection status (infected or uninfected) who were singletons or the first-born children of multiple-gestation pregnancies and who were born during the first enrollment in the WITS for their mothers.

Statistical Methods
A power analysis was performed to assess the possibility of detecting statistically significant results for the proposed study. Considering 1 set of results for each visit, we assumed that data for 150 HIV-infected infants and 2500 non–HIV-infected infants would be part of the analysis. Under this assumption, it would be possible to detect a change from 0.7 to 0.8 in 2 receiver operating characteristic (ROC) curves, for the proportion of CD4+ T cells and the CD4+/CD8+ T cell ratio, with 80% power when testing at the {alpha} = .05 level.

Means, SDs, and proportions, where presented, were calculated by using SAS 8 (SAS Institute, Cary, NC). ROC curves, with the WITS definition of HIV infection as the standard, were generated for each visit by using the CD4+ T cell proportion, the ratio of the CD4+ T cell and CD8+ T cell proportions (CD4+/CD8+ T cell ratio), and HIV-1 DNA PCR measures collected at that visit. To account for the correlated nature of the paired data in comparisons of ≥2 adjusted ROC curves (constructed with data from the same individuals), a nonparametric approach based on generalized U statistics methods was used to correct for the correlation.16 The hypothesis that the 2 empirical ROC curves would be equal was tested.16,17 Multivariate logistic regression models were fitted, with WITS-defined HIV infection status as the dependent variable and the proportion of CD4+ T cells and CD4+/CD8+ T cell ratio as the independent variables. The adjusted analyses also incorporated other covariates. Longitudinal profiles of CD4+/CD8+ T cell ratio percentiles for HIV-infected and uninfected infants were generated by using the least-mean squares (LMS) method.18

A linear discriminant analysis was performed to determine whether CD4+/CD8+ T cell ratios and hematocrit levels at different times during infancy could improve on the univariate ROC results for the CD4+/CD8+ T cell ratio data at 6 months of age. Specifically, the CD4+/CD8+ T cell ratios at 1 month, 2 months, 4 months, and 6 months of age and hematocrit values at 6 months of age were used to develop a simplified discriminant equation, as follows: D(CD4+/CD8+ T cell ratio,hematocrit level) = 6-month CD4+/CD8+ T cell ratio + 4-month CD4+/CD8+ T cell ratio – 1-month CD4+/CD8+ T cell ratio + (0.3 x 6-month hematocrit level).

By choosing 2 Bayesian prior probabilities for the equation, we were able to establish 2 regions that allowed the classification of a sizeable number of infants as infected or uninfected. To develop the discriminant function into a diagnostic tool, the following cutoff points were determined: <9, HIV-infected; 9–12, indeterminate; >12, non–HIV-infected. The equation was evaluated in the same way as the CD4+/CD8+ T cell ratio, and the values for the ROC were determined.


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Size and Characteristics of the Study Population
Between 1989 and 2005, 3484 women were enrolled in the WITS and 2842 live births occurred. The HIV infection status was determined for 2664 of those 2842 live-born infants. Of the 2664 infants with known infection status, 2565 were singletons and 49 were the first-born children of multiple-gestation pregnancies; 2208 were born to women during their first enrollment in the WITS. Therefore, the study population consisted of 2208 infants, of whom 179 were HIV-1-infected and 2029 were uninfected. Figure 1 depicts the derivation of the study population and the number of infants (overall and according to HIV infection status) who completed study visits during the first 12 months in the study. Eleven percent of the population was white, 48% black, 35% Hispanic, and 6% other. Seventeen percent of infants were of low birth weight. The median gestational age was 38 weeks (range: 25–43 weeks). The overall rate of MTCT of HIV-1 was 8.11%, and the rate decreased over time (on or before February 28, 1994, 20.54%; from March 1, 1994, to July 31, 1996, 7.19%; on or after August 1, 1996, 2.27%).


Figure 1
View larger version (23K):
[in this window]
[in a new window]

  		FIGURE 1 Derivation of the study population and median CD4+/CD8+ T cell ratios at each study visit.

 
CD4+/CD8+ T Cell Ratio and HIV Infection Status in Infancy
The numbers of infants at each study visit with complete flow cytometry data, as well as the median CD4+/CD8+ T cell ratios for HIV-1-infected and uninfected infants, are shown in Fig 1. Nonparametric comparisons were conducted to test the hypothesis that the areas under the ROC curves for the proportion of CD4+ T cells and the CD4+/CD8+ T cell ratio were equal. In unadjusted and adjusted analyses, infant HIV-1 infection status served as the response variable and the proportion of CD4+ T cells or the CD4+/CD8+ T cell ratio as the independent variable. Figure 2 shows the ROC curves according to visit (birth and 1, 2, 4, 6, and 12 months) for the CD4+/CD8+ T cell ratio at that visit, for determination of HIV-1 infection status. More-pronounced convexity in the upper left indicates greater diagnostic capability; this is measured in terms of the area under the curve, computed at each visit on the basis of logistic regression modeling, which is noted at the lower right of each curve. Early ROC curves show areas under the curve close to 0.7. With the 1-month ROC curve as an example, the area under the curve is 0.71. For the CD4+/CD8+ T cell ratio to have 95% sensitivity, the specificity of the CD4+/CD8+ T cell ratio is 15%. By the time of the 4-month visit, the area under the curve has improved to 0.866. The curve shows that, for the 4-month CD4+/CD8+ T cell ratio to have a sensitivity of 90%, the specificity of the ratio is 60%. Adjusting these analyses for maternal race/ethnicity and infant gender, birth weight, and gestational age did not change the overall relationship between CD4+/CD8+ T cell ratios and HIV infection status of the infants (Fig 3).


Figure 2
View larger version (25K):
[in this window]
[in a new window]

  		FIGURE 2 ROC plots of proportion of CD4+ T cells versus CD4+/CD8+ T cell ratio (univariate analyses): A, birth; B, 1 month; C, 2 months; D, 4 months; E, 6 months; F, 12 months.

 

Figure 3
View larger version (25K):
[in this window]
[in a new window]

  		FIGURE 3 ROC plots of proportion of CD4+ T cells versus CD4+/CD8+ T cell ratio (adjusted for other covariates): A, birth; B, 1 month; C, 2 months; D, 4 months; E, 6 months; F, 12 months.

 
The P values for the generalized U statistics comparing the correlated ROC curves (proportion of CD4+ T cells and CD4+/CD8+ T cell ratio) are indicated for each plot. There was significant evidence that the CD4+/CD8+ T cell ratio had better diagnostic accuracy than the proportion of CD4+ T cells in discriminating between HIV-infected and uninfected infants at ≥2 months of age, in both univariate and multivariate analyses (2 months: univariate, P = .018; multivariate, P = .001; 4 months: univariate, P = .047; multivariate, P = .036; 6 months: univariate, P < .0001; multivariate, P = .0002; 12 months: univariate, P = .0005; multivariate, P = .007). There was a marginal difference between the 2 diagnostic markers at birth (P = .054).

Figure 4 uses LMS curves to present percentiles for CD4+/CD8+ T cell ratios as a function of infants' ages. The curves were generated at the 5th, 10th, 25th, 50th, 75th, 90th, and 95th percentiles according to HIV-1 infection status (infected and uninfected), by using the LMS method.18 For any value of y, the specificity (100 minus the percentile for the uninfected infants) and sensitivity (the percentile for the infected infants) for any age can be obtained. From birth to 4 months of age, the CD4+/CD8+ T cell ratio decreased for HIV-infected infants but increased slightly for uninfected infants. After 4 months of age, the discriminatory power (separation of the 2 sets of percentile measurements) of the ratio stabilized, with the LMS percentile curves showing a slight decrease in the values of the ratio from 4 months of age to 12 months of age for both HIV-1-infected and uninfected infants.


Figure 4
View larger version (62K):
[in this window]
[in a new window]

  		FIGURE 4 Reference percentile curves for CD4+/CD8+ T cell ratio according to HIV-1 infection status: 5th, 10th, 25th, 50th, 75th, 90th, and 95th percentiles.

 
To assess whether maternal receipt of antiretroviral drugs was associated with outcome measures, we determined the values of CD4+/CD8+ T cell ratios for infected and uninfected infants according to in utero exposure to maternal antiretroviral drug therapy. Figure 5A shows the mean of CD4+/CD8+ T cell ratios from birth to 6 months according to HIV infection status and maternal antiretroviral drug usage. The association of maternal antiretroviral drug therapy and infant hematocrit levels also was evaluated (Fig 5B), because in utero exposure to antiretroviral drugs has been reported to influence hematologic indices, albeit transiently.19,20 Maternal antiretroviral drug usage did not seem to affect the CD4+/CD8+ T cell ratio or hematocrit level among HIV-infected or uninfected infants.


Figure 5
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 5 Mean of CD4+/CD8+ T cell ratio (A) and mean of hematocrit percentage (B) according to age in HIV-1–infected and uninfected infants whose mothers did or did not receive antepartum or intrapartum antiretroviral drug treatment.

 
Discriminant Analysis Incorporating CD4+/CD8+ T Cell Ratio and Hematocrit Level
To determine whether we could improve the diagnostic potential of the CD4+/CD8+ T cell ratio, we developed a discriminant analysis algorithm that could be used to make a presumptive diagnosis of HIV-1 infection on the basis of CD4+/CD8+ T cell ratio and hematocrit data at time points up to 6 months of age. There were 88 HIV-infected and 955 uninfected infants with complete data for this analysis. The CD4+/CD8+ T cell ratios for these subjects at 1 month, 2 months, 4 months, and 6 months of age were significantly related to HIV infection status (P < .0001). The hematocrit levels at 4 and 6 months of age also were significantly related to HIV infection status (P < .0001). The discriminant function was developed as described above and, with evaluation of the resulting discriminant equation in the same way as for the CD4+/CD8+ T cell ratio, the values for the ROC increased at the 6-month evaluation to 0.895 and 0.899 in the unadjusted and adjusted analyses, respectively. These ROC values were superior to the ROC values obtained with CD4+/CD8+ T cell ratios alone.

With application of this discriminant diagnosis rule to the WITS data, the presumptive diagnosis and actual infection status in the study cohort are shown in Table 1. Sixteen, 54, and 18 HIV-infected infants were classified as HIV-infected, indeterminate, and uninfected, respectively. Similarly, 1, 144, and 810 uninfected infants were classified as HIV-infected, indeterminate, and uninfected, respectively. The positive and negative predictive values for the discriminant function applied to the WITS data were 94% and 98%, respectively.


View this table:
[in this window]
[in a new window]

  		TABLE 1 HIV-1 Infection Status Based on Discriminant Analysis and Actual HIV-1 Infection Status

 
Effects of Prevalence of HIV Infection on Diagnosis
Developing and testing the discriminant function with the same population can lead to bias in terms of the degree to which the classification agrees with the final HIV infection status of the infant. Furthermore, the agreement table was developed with a fixed prevalence of HIV infection (as observed in the WITS cohort). To address this potential bias and to investigate how different prevalence values could affect the classification probabilities of the discriminant function, we simulated infant populations with different prevalences, as shown in Fig 6, in the following way. We constructed a simulation population of 5000 infants by stratifying the WITS population into HIV-infected and uninfected infants. Then we randomly selected prespecified numbers of HIV-infected and uninfected infants with replacement to obtain the simulated population of 5000 infants. For instance, to obtain a simulated population with a 10% prevalence of HIV infection, we selected with replacement 500 infants from the HIV-infected group in the WITS population and 4500 infants from the uninfected group in the WITS population. Figure 6 also shows the numbers and proportions of infants who were classified as HIV-1-infected or uninfected, ranging from 62% (when the prevalence of HIV infection was 50%) to 84%. The positive and negative predictive values remained above 90% over a wide range of prevalences of HIV infection.


Figure 6
View larger version (11K):
[in this window]
[in a new window]

  		FIGURE 6 Positive predictive value (PPV) and negative predictive value (NPV) based on discriminant function for simulated data with various prevalences of HIV-1 infection among infants. a Numbers and proportions of children who could be classified by using CD4+/CD8+ T cell ratio and hematocrit measurements.

 

   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In this study, we evaluated a surrogate measure, the CD4+/CD8+ T cell ratio, for its ability to discriminate between HIV-infected and uninfected infants born to HIV-infected women. Our analyses suggest that, in situations where virological diagnosis is not feasible and facilities exist for CD4+ and CD8+ T cell determinations, the CD4+/CD8+ T cell ratio can play a valuable role in providing a presumptive diagnosis of HIV infection and can aid in the care of HIV-exposed infants born to HIV-infected women.

The rationale for selecting the CD4+/CD8+ T cell ratio as a potential diagnostic test was based on several known facts about HIV pathogenesis. It is well established that proportions of CD4+ T cells are normally extremely large, relative to proportions of CD8+ T cells, in infants at birth and during the first several months of life21,22 and they gradually decrease with age.21,23 Although CD4+ T cell counts of HIV-infected infants decrease faster than those of uninfected infants,15–18,24 depletion of CD4+ T cells in HIV-infected infants does not become apparent early in life except in rare instances where immunosuppression is evident at birth.25,26 In contrast to CD4+ T cells, CD8+ T cells show greater turnover and are greatly influenced by the immune activation that accompanies HIV infection.27,28 The ensuing CD8+ T cell expansion begins early in HIV-1 infection and contributes to the altered CD4+/CD8+ T cell ratio that becomes evident even in acute HIV infection in adults29,30 and precedes detection of loss of CD4+ T cells. The expansion of CD8+ T cells and the known relationship of the CD4+/CD8+ T cell ratio to HIV disease progression in HIV-infected adults31 led us to hypothesize that a similar CD8+ T cell expansion occurs in HIV-infected infants, making the CD4+/CD8+ T cell ratio a more-attractive measure than CD4+ T cell counts for diagnosis. Such a presumption assumes that the infant is otherwise healthy, without evidence of concurrent coinfection (fungal, parasitic, or viral) that could produce CD8+ T cell expansion. An additional advantage of the CD4+/CD8+ T cell ratio is that it does not matter whether the ratio is derived from proportions or absolute numbers of CD4+ and CD8+ T cells.

To assess the value of the CD4+/CD8+ T cell ratio as a diagnostic test, we made use of data from a cohort of North American infants born to HIV-infected mothers, with known HIV infection status based on virological testing. First, we compared the proportions of CD4+ T cells and CD4+/CD8+ T cell ratios in known HIV-infected and uninfected infants. At birth, the difference between the 2 measures was marginal; starting at 2 months of age, there was significant evidence that the CD4+/CD8+ T cell ratio had better accuracy than the proportion of CD4+ T cells in discriminating between HIV-infected and uninfected infants, in both univariate and multivariate analyses. The discriminatory power at ≥4 months of age resulted in a trade-off of ~60% specificity for 90% sensitivity or 65% sensitivity for 90% specificity, still far less than the values obtained with PCR results (sensitivity: 99%; specificity: 98%).

To simplify the use of a given CD4+/CD8+ value at a particular age by a health care provider, percentile curves for CD4+/CD8+ T cell ratios were generated according to HIV infection status (infected or uninfected) as a function of the age when determinations of proportions of CD4+ T cells and CD8+ T cells were made. As expected, the CD4+/CD8+ T cell ratio was high in both HIV-infected and uninfected infants at birth. Although there was considerable overlap at intermediate values, values at the upper or lower ends of the range of CD4+/CD8+ T cell ratios were clearly able to discriminate between HIV-1-infected and uninfected infants. Importantly, the curves provide information about the sensitivity and specificity of a given ratio for diagnosis of HIV infection at all ages to 12 months. With the caveat that ideally data should be developed for the local population, the percentile curves presented here can be used cautiously by health care providers to determine the potential risk of the infant being infected and can be combined with other parameters, including clinical judgment, to facilitate management decisions.

The predictive value of the CD4+/CD8+ T cell ratio was improved through incorporation of hematocrit levels, a simple laboratory marker that is relatively easy to measure in low-resource areas. Markers such as total lymphocyte counts, hemoglobin levels, and serum albumin levels were shown previously to be useful surrogates for disease progression.32 We developed a simple discriminant analysis algorithm by using CD4+/CD8+ T cell ratios at different time points and hematocrit levels at 4 and 6 months of age. By varying the Bayes prior probabilities before the analysis, it was possible to develop a testing strategy that had a 94% positive predictive value and a 98% negative predictive value. The ability to rule out HIV infection was particularly strong with this analysis. Although such information would not completely eliminate the need to continue monitoring for HIV infection, it could greatly minimize monitoring requirements and alleviate the anxiety of the caregiver. In situations where safe alternatives to breastfeeding are available, the probable diagnosis of not being infected through perinatal MTCT would allow measures to be taken to prevent infection after birth. The enhanced ability to make a presumptive diagnosis of HIV infection would be of paramount importance for establishing the plan for treatment and monitoring according to local guidelines. A limited number of infants are expected to fall into the indeterminate category according to discriminant values. Subjects in this category ideally would be subjected to virological testing for a definitive diagnosis, which would limit nucleic acid testing to only a few cases, instead of all HIV-exposed infants. Without virological testing, determination of the status of indeterminate cases would depend on clinical staging or serological testing during follow-up monitoring.

It has been reported that prenatal antiretroviral prophylaxis can transiently influence hematologic indices of HIV-exposed infants.19 However, we did not find any differences in either CD4+/CD8+ T cell ratios or hematocrit levels between HIV-infected infants whose mothers did versus did not receive antiretroviral drugs. Also, no differences were observed for uninfected infants whose mothers did versus did not receive antiretroviral drugs. Therefore, this type of discriminant analysis is not affected by antiretroviral drug intervention among the mothers.

The predictive value of diagnostic tests is influenced by the prevalence of the particular condition. The WITS cohort was established before routine implementation of antiretroviral prophylaxis for prevention of MTCT of HIV but continued enrollment in the current era, when both antiretroviral prophylaxis and cesarean section before labor and before membrane rupture are used for prevention of such transmission. In a simulated population, with the prevalence of HIV infection among infants ranging from 1% to 50%, the diagnosis of HIV infection by using both CD4+/CD8+ T cell ratios and hematocrit levels could be made for 61% to 84% of infants by 6 months of age. The CD4+/CD8+ T cell ratio performed better with increasing age, and it might be worthwhile to evaluate the ratio as a tool for monitoring disease progression or responses to treatment.

We conclude that the CD4+/CD8+ T cell ratio performs well as a diagnostic assay for HIV infection in infants, and its diagnostic potential may be enhanced when it is used in conjunction with other clinical or laboratory assessments (eg, hematocrit levels). Several new cost-effective methods for obtaining CD4+ and CD8+ T cell counts are being investigated,33 which might lead to broader access to CD4+/CD8+ T cell ratio data. The diagnostic utility of the CD4+/CD8+ T cell ratio in HIV-exposed infants has received limited attention to date. This is one of the first studies using a large database of HIV-exposed infants for whom virological diagnoses were established with DNA PCR assays and CD4+/CD8+ T cell ratios were determined. An earlier small study using CD4+/CD8+ T cell ratios for HIV-exposed infants has been reported,34 in which there was a strong correlation between lower CD4+/CD8+ T cell ratios and positive HIV-1 DNA PCR results. Analysis of a large dataset35 resulted in similar conclusions as the present report. Our results suggest that a study in resource-poor situations in which CD4+/CD8+ T cell ratios, and virological diagnosis can be performed would be required to validate the findings of our study. With higher rates of MTCT of HIV, the sensitivity and specificity are expected to be higher. In resource-poor settings without routine availability of virological assays for infant diagnoses, establishing a presumptive diagnosis of HIV infection with surrogate measures such as the CD4+/CD8+ T cell ratio early in life may allow timely intervention for HIV-exposed infants, to improve outcomes.


   ACKNOWLEDGMENTS
 
The WITS study was supported by the following grants from the National Institutes of Health: U01 AI 034858, U01 DA 015054, U01 DA 015053, U01 AI 034841, U01 HD 036117, U01 HD 041983, N01 AI 085339, U01 AI 050274–01, GCRC RR000188, and GCRC RR000645.

Principal investigators, study coordinators, and program officers for WITS were as follows: Clemente Diaz and Edna Pacheco-Acosta (University of Puerto Rico, San Juan, PR); Ruth Tuomala, Ellen Cooper, and Donna Mesthene (Boston/Worcester site, Boston, MA); Phil LaRussa and Alice Higgins (Columbia Presbyterian Hospital, New York, NY); Sheldon Landesman, Herman Mendez, and Ava Dennie (State University of New York, Brooklyn, NY); Kenneth Rich and Delmyra Turpin (University of Illinois at Chicago, Chicago, IL); William Shearer and Norma Cooper (Baylor College of Medicine, Houston, TX); Joana Rosario (National Institute of Allergy and Infectious Diseases, Bethesda, MD); Kevin Ryan (National Institute of Child Health and Human Development, Bethesda, MD); Vincent Smeriglio and Katherine Davenny (National Institute on Drug Abuse, Bethesda, MD); and Bruce Thompson (Clinical Trials & Surveys Corp, Baltimore, MD). The Scientific Leadership Core included Kenneth Rich (principal investigator) and Delmyra Turpin (study coordinator).


   FOOTNOTES
 
Accepted Dec 12, 2007.

Address correspondence to Savita Pahwa, MD, Department of Microbiology and Immunology, University of Miami Miller School of Medicine, 1580 NW 10th Ave, BCRI 712, Miami, FL 33136. E-mail: spahwa{at}med.miami.edu

This work was presented in part at the 14th Conference on Retroviruses and Opportunistic Infections; February 25-28, 2007; Los Angeles, CA (abstract 686).

The authors have indicated they have no financial relationships relevant to this article to disclose.


What's Known on This Subject

A previous study using the CD4+/CD8+ T cell ratio for diagnosis among HIV-exposed infants was performed, but the sample size was very small.

 

What This Study Adds

Early diagnosis of HIV infection in infants with perinatal HIV exposure is critical for clinical management decisions. For resource-poor countries where virological diagnosis is not feasible, this study offers another way to make a presumptive diagnosis of HIV infection.

 


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Read JS. Prevention of mother-to-child transmission of HIV. In: Zeichner SL, Read JS, eds. Textbook of Pediatric HIV Care. Cambridge, England: Cambridge University Press; 2005:111–133
  2. Coovadia HM, Rollins NC, Bland RM, et al. Mother-to-child transmission of HIV-1 infection during exclusive breastfeeding in the first 6 months of life: an intervention cohort study. Lancet. 2007;369 (9567):1107 –1116[CrossRef][Web of Science][Medline]
  3. Little K, Newell ML, Luo C, Ngongo N, Borja MC, McDermott P. Estimating the number of vertically HIV-infected children eligible for antiretroviral treatment in resource-limited settings. Int J Epidemiol. 2007;36 (3):679 –687[Abstract/Free Full Text]
  4. Paintsil E, Andiman WA. Care and management of the infant of the HIV-1-infected mother. Semin Perinatol. 2007;31 (2):112 –123[CrossRef][Web of Science][Medline]
  5. Little K, Thorne C, Luo C, et al. Disease progression in children with vertically-acquired HIV infection in sub-Saharan Africa: reviewing the need for HIV treatment. Curr HIV Res. 2007;5 (2):139 –153[CrossRef][Web of Science][Medline]
  6. Nielsen K, Bryson YJ. Diagnosis of HIV infection in children. Pediatr Clin North Am. 2000;47 (1):39 –63[CrossRef][Web of Science][Medline]
  7. Sherman GG, Cooper PA, Coovadia AH, et al. Polymerase chain reaction for diagnosis of human immunodeficiency virus infection in infancy in low resource settings. Pediatr Infect Dis J. 2005;24 (11):993 –997[CrossRef][Web of Science][Medline]
  8. Fiscus SA, Wiener J, Abrams EJ, Bulterys M, Cachafeiro A, Respess RA. Ultrasensitive p24 antigen assay for diagnosis of perinatal human immunodeficiency virus type 1 infection. J Clin Microbiol. 2007;45 (7):2274 –2277[Abstract/Free Full Text]
  9. Patton JC, Akkers E, Coovadia AH, Meyers TM, Stevens WS, Sherman GG. Evaluation of dried whole blood spots obtained by heel or finger stick as an alternative to venous blood for diagnosis of human immunodeficiency virus type 1 infection in vertically exposed infants in the routine diagnostic laboratory. Clin Vaccine Immunol. 2007;14 (2):201 –203[Abstract/Free Full Text]
  10. Chiappini E, Galli L, Tovo PA, et al. Virologic, immunologic, and clinical benefits from early combined antiretroviral therapy in infants with perinatal HIV-1 infection. AIDS. 2006;20 (2):207 –215[Web of Science][Medline]
  11. Chinen J, Easley KA, Mendez H, Shearer WT. Decline of CD3-positive T-cell counts by 6 months of age is associated with rapid disease progression in HIV-1-infected infants. J Allergy Clin Immunol. 2001;108 (2):265 –268[CrossRef][Web of Science][Medline]
  12. Shearer WT, Easley KA, Goldfarb J, et al. Evaluation of immune survival factors in pediatric HIV-1 infection. Ann N Y Acad Sci. 2000;918 :298 –312[Web of Science][Medline]
  13. Sheon A, Fox H, Rich K, et al. for The Women and Infants Transmission Study Group. The Women and Infants Transmission Study (WITS) of maternal-infant HIV transmission: study design, methods, and baseline data. J Women's Health. 1996;5 (3):69 –78
  14. Calvelli T, Denny TN, Paxton H, Gelman R, Kagan J. Guideline for flow cytometric immunophenotyping: a report from the National Institute of Allergy and Infectious Diseases, Division of AIDS. Cytometry. 1993;14 (7):702 –715[CrossRef][Web of Science][Medline]
  15. Rich KC, Brambilla D, Pitt J, et al. Lymphocyte phenotyping in infants: maturation of lymphocyte subpoplations and the effects of HIV infection. Clin Immunol Immunopathol. 1997;85 (3):273 –281[CrossRef][Web of Science][Medline]
  16. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44 (3):837 –845[CrossRef][Web of Science][Medline]
  17. Mandrekar JN, Mandrekar SJ. Statistical methods in diagnostic medicine using SAS software. In: Proceedings of the Thirtieth Annual SAS Users Group International Conference. Cary, NC: SAS Institute; 2005;12:505–513
  18. Cole TJ, Green PJ. Smoothing reference centile curves: the LMS method and penalized likelihood. Stat Med. 1992;11 (10):1305 –1319[Web of Science][Medline]
  19. Pacheco SE, McIntosh K, Lu M, et al. Effect of perinatal antiretroviral drug exposure on hematologic values in HIV-uninfected children: an analysis of the Women and Infants Transmission Study. J Infect Dis. 2006;194 (8):1089 –1097[CrossRef][Web of Science][Medline]
  20. Mussi-Pinhata M, Rego MA, Freimanis L, et al. Maternal antiretrovirals and hepatic enzyme, hematologic abnormalities among human immunodeficiency virus type 1-uninfected infants: the NISDI perinatal study. Pediatr Infect Dis J. 2007;26 (11):1032 –1037[Web of Science][Medline]
  21. McKinney RE, Jr, Wilfert CM. 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. 1992;11 (8):639 –644[Web of Science][Medline]
  22. Shearer WT, Rosenblatt HM, Gelman RS, et al. Lymphocyte subsets in healthy children from birth through 18 years of age: the Pediatric AIDS Clinical Trials Group P1009 study. J Allergy Clin Immunol. 2003;112 (5):973 –980[CrossRef][Web of Science][Medline]
  23. Shearer WT, Easley KA, Goldfarb J, et al. Prospective 5-year study of peripheral blood CD4, CD8, and CD19/CD20 lymphocytes and serum Igs in children born to HIV-1 women: the P2C2 HIV Study Group. J Allergy Clin Immunol. 2000;106 (3):559 –566[CrossRef][Web of Science][Medline]
  24. Shearer WT, Rosenblatt HM, Schluchter MD, Mofenson LM, Denny TN. Immunologic targets of HIV infection: T cells: NICHD IVIG Clinical Trial Group, and the NHLBI P2C2 Pediatric Pulmonary and Cardiac Complications of HIV Infection Study Group. Ann NY Acad Sci. 1993;693 :35 –51[CrossRef][Web of Science][Medline]
  25. Nahmias AJ, Clark WS, Kourtis AP, et al. Thymic dysfunction and time of infection predict mortality in human immunodeficiency virus-infected infants: CDC Perinatal AIDS Collaborative Transmission Study Group. J Infect Dis. 1998;178 (3):680 –685[Web of Science][Medline]
  26. Kourtis AP, Ibegbu C, Nahmias AJ, et al. Early progression of disease in HIV-infected infants with thymus dysfunction. N Engl J Med. 1996;335 (19):1431 –1436[Abstract/Free Full Text]
  27. Hazenberg MD, Stuart JW, Otto SA, et al. T-cell division in human immunodeficiency virus (HIV)-1 infection is mainly due to immune activation: a longitudinal analysis in patients before and during highly active antiretroviral therapy (HAART). Blood. 2000;95 (1):249 –255[Abstract/Free Full Text]
  28. Sachsenberg N, Perelson AS, Yerly S, et al. Turnover of CD4+ and CD8+ T lymphocytes in HIV-1 infection as measured by Ki-67 antigen. J Exp Med. 1998;187 (8):1295 –1303[Abstract/Free Full Text]
  29. Mattapallil JJ, Letvin NL, Roederer M. T-cell dynamics during acute SIV infection. AIDS. 2004;18 (1):13 –23[CrossRef][Web of Science][Medline]
  30. Zaunders J, Carr A, McNally L, Penny R, Cooper DA. Effects of primary HIV-1 infection on subsets of CD4+ and CD8+ T lymphocytes. AIDS. 1995;9 (6):561 –566[Web of Science][Medline]
  31. Phillips AN, Sabin CA, Elford J, Bofill M, Lee CA, Janossy G. CD8 lymphocyte counts and serum immunoglobulin A levels early in HIV infection as predictors of CD4 lymphocyte depletion during 8 years of follow-up. AIDS. 1993;7 (7):975 –980[Web of Science][Medline]
  32. Mofenson LM, Harris DR, Moye J, et al. Alternatives to HIV-1 RNA concentration and CD4 count to predict mortality in HIV-1-infected children in resource-poor settings. Lancet. 2003;362 (9396):1625 –1627[CrossRef][Web of Science][Medline]
  33. Gomo E, Ndhlovu P, Vennervald BJ, Nyazema N, Friis H. Enumeration of CD4 and CD8 T-cells in HIV infection in Zimbabwe using a manual immunocytochemical method. Cent Afr J Med. 2001;47 (3):64 –70[Medline]
  34. Zijenah LS, Katzenstein DA, Nathoo KJ, et al. T lymphocytes among HIV-infected and -uninfected infants: CD4/CD8 ratio as a potential tool in diagnosis of infection in infants under the age of 2 years. J Transl Med. 2005;3 (1):6[CrossRef][Medline]
  35. Shearer WT, Pahwa S, Read JS, et al. CD4/CD8 T cell ratio predicts HIV infection in infants: the NHLBI P2C2 HIV study. J Allergy Clin Immunol. 2007;120 (6):1449 –1456[CrossRef][Web of Science]
PEDIATRICS (ISSN 1098-4275). ©2008 by the American Academy of Pediatrics

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Facebook Facebook   Add to Reddit Reddit   Add to Technorati Technorati   Add to Twitter Twitter    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pahwa, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pahwa, S.
Related Collections
Right arrow Infectious Disease & Immunity
Right arrowRelated AAP Red Book topics:
Human Immunodeficiency Virus...
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Facebook   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?