|
|
| |||||||||||||||||||||||||||||||||||||||
Background Among infants with maternally transmitted human immunodeficiency virus (HIV) infection, there are two patterns of disease progression. In about a fifth of these infants there is a rapid progression to profound immunodeficiency, whereas in the majority the disease progresses much more slowly.
Methods We studied the clinical and biologic characteristics of the mothers of infants infected with HIV type 1 (HIV-1) in the French Prospective Multicenter Cohort. Infection in the children was confirmed by serologic tests at the age of 18 months or by death from the acquired immunodeficiency syndrome at an earlier age. Only the 162 infected infants who could be followed for at least 18 months or until death were included in the analysis.
Results The risk of opportunistic infections or encephalopathy in the first 18 months was 50 percent in the infants of mothers with class IV disease, according to the Centers for Disease Control and Prevention classification, and 14 percent in the infants of mothers with class II or III disease (relative risk, 3.6; 95 percent confidence interval, 1.8 to 7.3; P<0.002). Forty-four percent of the former infants and 9 percent of the latter died before 18 months (relative risk, 4.7; 95 percent confidence interval, 2.1 to 10.4; P<0.002). The risk of death correlated inversely with the mother's CD4+ cell count and directly with her HIV-1 p24 antigen level at delivery. There was also a direct correlation between the mother's CD4+ cell count and that of the infant at one, three, and nine months of age (correlation coefficient at nine months [n = 44], 0.48; P<0.002). HIV-1 p24 antigen was detected more often in the infants whose mothers also had the antigen.
Conclusions In infants whose HIV infection is maternally acquired, the rate of disease progression varies directly with the severity of the disease in the mother at the time of delivery.
Methods
The design of this prospective cohort study of infants born to HIV-infected mothers, which began in 1986, has been described in detail elsewhere7. After informed consent has been obtained, mother-infant pairs are enrolled in the study if the mother's serologic status is known at or before the time of delivery. The 62 participating centers agree to include all eligible patients whose infection is diagnosed locally. For this analysis, data on the mother, which were forwarded to the coordinating center, included the mode of infection, geographic origin (European or African), clinical class of disease according to the classification of the Centers for Disease Control and Prevention (CDC),8 CD4+ cell count, and HIV type 1 (HIV-1) p24 antigen level. The last three measures were all assessed within the month preceding delivery. The CD4+ cell count was available only for women followed at certain centers. Follow-up data for the infants were collected at regular intervals (at birth; 3, 9, 15, and 18 months after birth; and then every 6 months). The outcomes analyzed in this study were opportunistic infections (CDC pediatric class P2D1), HIV-related encephalopathy (CDC class P2B), and death9. Laboratory evaluations included serologic tests for HIV (the enzyme-linked immunosorbent assay and the Western blot test), detection of HIV-1 p24 antigen in serum (by the direct method without dissociation of antibody-antigen complexes), and the circulating CD4+ cell count.
Diagnosis of HIV Infection in the Infants
This analysis included only infants who were followed for more than 18 months and who were clearly infected with HIV, as evidenced by either serologic status at 18 months or death due to AIDS before 18 months. Siblings of infected infants and twins were excluded from the analysis.
Statistical Analysis
Quantitative variables are expressed as means ±SD. The Kaplan-Meier method was used to establish life-table curves. Standard deviations were determined with Greenwood's formula. The relation between maternal characteristics and the clinical status of the infants was determined by comparing the incidence of events in the first 18 months of life, since all the infants were born at least 18 months before the cutoff date for the analysis. Fisher's exact test was used to compare these rates of incidence. Pairs of survival curves were compared with the log-rank test. All P values are two-tailed, and the threshold of significance was set at 0.05. The 95 percent confidence limits for relative risks were estimated according to the Mantel-Haenszel method for prospective cohort studies. The analysis was performed with SAS software (SAS Institute, Cary, N.C.).
Results
At the time of the analysis, 1244 infants born to HIV-1-infected mothers were included in the cohort. Of the 940 infants who had been followed for more than 18 months, 162 were infected with HIV-1; 70 children were still alive and being followed at 36 months of age.
Description of the Mothers
The mode of infection was known for 159 women: 85 (53 percent) were believed to have been infected through sexual contact, 71 (45 percent) through intravenous drug abuse, and 3 (2 percent) through blood transfusions. Thirty of the 159 women (19 percent) were from sub-Saharan Africa. The clinical status at the time of delivery was known for 145 women: 108 (74 percent) had class II disease, according to the CDC classification (seropositivity with no symptoms), 21 (14 percent) had class III disease (persistent polyadenopathy), and 16 (11 percent) had class IV disease (AIDS and other conditions). At the time of delivery, HIV-1 p24 antigen was detected in serum from 16 of 108 women (15 percent). The mean CD4+ count (data available for 58 mothers) was 461.2 ±325.2 cells per cubic millimeter. Since the beginning of the cohort study, the distribution of these variables has been stable, apart from the predominant mode of infection, which has gradually shifted from intravenous drug abuse to heterosexual contact. The CD4+ cell count and the frequency of HIV-1 p24 antigen detection were not correlated with geographic origin or the mode of infection (data not shown).
The Risk of Opportunistic Infections, Encephalopathy, and Death in the Infants
The cumulative risk of an opportunistic infection or encephalopathy reflected the characteristic profile of the disease in children -- that is, the cumulative incidence of these conditions increased rapidly in the first few months of life and then more slowly after 12 months. The risk of an opportunistic infection was 13 ±3 percent at 18 months and 15 ±3 percent at 36 months. The first 24 opportunistic infections were due to Pneumocystis carinii in 11 cases, cytomegalovirus in 6, Candida albicans in 4, and other organisms in 3. The risk of encephalopathy was 10 ±2 percent at 18 months and 11 ±3 percent at 36 months. Kaplan-Meier plots for these two outcomes combined differed little from the plots for the individual outcomes, since both conditions often develop in the same child. The overall risk of HIV-related encephalopathy or an opportunistic infection was 18 ±3 percent at 18 months and 20 ±3 percent at 36 months (Figure 1).
|
|
Neither the mode of infection in the women nor their geographic origin was related to the course of the disease in the infants. The risk of opportunistic infections or encephalopathy at 18 months, as well as the risk of death, was the same whether the mother was African or European (Table 1 and Table 2). In contrast, the risk of opportunistic infections or encephalopathy and the risk of death were clearly related to the mother's clinical status. The risk of serious manifestations of disease at 18 months was 50 percent in the infants born to mothers with class IV disease at delivery, as compared with 14 percent in those whose mothers had class II or III disease (relative risk, 3.6; 95 percent confidence interval, 1.8 to 7.3; P<0.002) (Figure 1 and Table 1). Forty-four percent of the infants whose mothers had class IV disease died before the age of 18 months, as compared with 9 percent of the infants whose mothers had class II or III disease (relative risk, 4.7; 95 percent confidence interval, 2.1 to 10.4; P<0.002) (Figure 2 and Table 2). The risk of early, severe manifestations and the risk of death were higher in infants whose mothers had HIV-1 p24 antigen at the time of delivery. The maternal CD4+ count (>400 or 
400 cells per cubic millimeter) was significantly related to the risk of death (P<0.008). It was just above the threshold of significance in relation to opportunistic infections or encephalopathy, probably because the sample of women analyzed was too small (Table 1 and Table 2). Similar results were obtained when the Kaplan-Meier curves for each maternal characteristic were compared (Figure 1 and Figure 2).
|
|
In addition, given the impact of prophylactic treatment on some opportunistic infections, as well as the impact of antiretroviral therapy, we confirmed that the infants born to mothers with class IV disease were not undertreated in comparison with the infants whose mothers had less advanced disease; in fact, the opposite was found. Treatment with trimethoprim-sulfamethoxazole and zidovudine was started earlier and was more frequent among the 16 children born to mothers with class IV disease. At 18 months, 15 of the children had received trimethoprim-sulfamethoxazole and 13 had received zidovudine, as compared with 69 and 81, respectively, of the 130 children born to mothers with class II or III disease.
CD4+ Cell Counts and Detection of p24 Antigen in Mothers and Infants
The CD4+ count (expressed as a percentage) in the mothers at the time of delivery correlated with that in the infants at one month of age (n = 12; r = 0.61; P<0.04), at three months (n = 29, r = 0.45; P<0.02), and at nine months (n = 44; r = 0.48; P<0.002). Similarly, HIV-1 p24 antigen was detected more frequently in the infants of mothers with detectable antigen at the time of delivery. The p24 antigen was detected in 31 percent of the infants whose mothers had detectable antigen, as compared with 7 percent of the infants whose mothers did not have it (n = 96, P<0.03).
Discussion
We are beginning to understand the relation between the clinical and biologic characteristics of pregnant women and the risk of vertical HIV transmission10. The rate of transmission to infants increases in parallel with the clinical severity of the maternal infection and is inversely proportional to the maternal CD4+ cell count. Our data show a clear relation between maternal characteristics at the time of delivery and the course of HIV disease in infants. The risk of a severe, early form of the disease (i.e., opportunistic infections, HIV-related encephalopathy, or both) was three to four times higher in infants whose mothers had class IV disease than in infants whose mothers were asymptomatic or had only persistent lymphadenopathy. This relation was confirmed when the biologic markers of HIV-1 disease in the mothers and infants were considered.
Our analysis was concerned with the first few years of life and thus provides information mainly on the severe form of HIV disease that occurs in the first 12 months after birth. We did not take into account the later risk of opportunistic infections or other AIDS-defining events, particularly lymphoid interstitial pneumonia, which is not associated with a high mortality rate.
The relation between the course of HIV infection in the person who transmits the virus and the course in the person who becomes infected is difficult to establish in the case of sexual transmission. This relation has been studied, however, in cases of transmission by blood transfusion. Ward et al. showed that the short-term risk of AIDS in a person infected in this way depends on the time required for the disease to develop in the blood donor11. Mother-infant transmission is another instance in which this type of correlation can be established. However, the interpretation of clinical data is not simple. In our study, intrafamilial transmission of opportunistic infections after birth was probably unlikely, since the risk of encephalopathy was the same as that of opportunistic infections in the infants, and laboratory values in the mothers influenced those in the infants immediately after birth. In contrast, simultaneous transmission during pregnancy of another pathogen, such as cytomegalovirus, whose replication in the maternal blood and transmission to the infant would be facilitated by the mother's immune deficiency, could influence the progression of HIV-1 disease. However, severe cytomegalovirus infections in the infants in our study accounted for less than one third of the initial opportunistic infections.
The main hypothesis linking maternal status to disease progression in infants involves the conditions under which HIV is transmitted to the infant. The inoculum may well be larger when the mother's disease is at an advanced stage. The presence of HIV-1 p24 antigen in the infants at birth -- an indirect and imperfect index of the viral load and degree of replication -- correlated with the presence of antigen in the mothers. Another approach is to study the phenotype of the viral strains transmitted to the infant. Several studies have suggested that the pathogenicity of HIV strains increases during the course of the infection, a shift marked by the formation of syncytial cells in vitro12,13. A comparison of the strains in mothers and infants is under way in several laboratories.
An additional hypothesis, which is not necessarily incompatible with the other two, is based on the timing of transmission. Arguments favoring both early and late (perinatal) transmission have been made, and both types may well occur. A relation between the time of transmission and the course of the disease in infants has been suggested, particularly to explain the two patterns of disease progression. The rare occurrence of thymic abnormalities in fetuses of HIV-infected women14 suggests that an interaction between the virus and the fetal immune system may account for the early and severe immune deficiency observed in 15 to 20 percent of cases. Similarly, the presence of the virus in fetal brain tissue may be related to the subsequent onset of encephalopathy15. It is possible that in mothers who have advanced disease, with a heavy intracellular and extracellular viral load, the fetuses are infected at an earlier stage of development, resulting in the onset of the disease soon after birth. However, reports of early opportunistic infections and encephalopathy in infants infected at birth by blood transfusions16 may suggest alternative hypotheses.
If the maternal viral load is a primary determinant of the progression of disease in the child, a therapeutic reduction in viral replication during pregnancy may slow the course of the disease in the infant, apart from the ability of such treatment to prevent transmission. While awaiting the results of intervention trials, physicians should take these findings into account when counseling HIV-infected pregnant women.
Supported by a grant from the Agence Nationale de Recherche sur le SIDA, Paris.
Source Information
From the Unite d'Immunologie-Hematologie Pediatrique, INSERM Unite 132 (S.B., C.G.), and the Laboratoire de Virologie (C.R.), Hopital Necker Enfants Malades; and INSERM Unite 292-Sante Publique, Epidemiologie, Reproduction Humaine, Hopital Bicetre (M.-J.M., J.-P.T.) -- both in Paris. A complete list of participants in the French Pediatric HIV Infection Study Group appears in the Appendix.
Address reprint requests to Dr. Blanche at Hopital Necker Enfants Malades, Unite d'Immunologie-Hematologie Pediatrique, 149, rue de Sevres, 75743 Paris CEDEX 15, France.
References
The following investigators participated in the study: M. Tardieu, Hopital Bicetre; G. Noseda and J.M. Huraux, Hopital Pitie-Salpetriere; M. Levine and E. Vilmer, Hopital Robert Debre; A. de Crepy and F. Simon, Hopital Bichat; A. Krivine and C. Francoual, Hopital Saint-Vincent; L. di Maria and C. Courpotin, Hopital Trousseau; C. Carlus Moncomble, Hopital Cite Universitaire; M. Burgard, C. Rouzioux, D. Girault, J.L. Stephan, S. Blanche, J. Terris, and F. Veber, Hopital Necker; G. Firtion and R. Henrion, Hopital Cochin; N. Ciraru-Vigneron and C. Bruner, Hopital Lariboisiere; F. Parnet Mathieu, Hopital Saint-Antoine; and F. Herve, Hopital Tenon -- all in Paris; C. Allisy, Argenteuil; M. Dandine, Aulnay; P. Labrune and M. Vial, Clamart; E. Lachassine and J. Gaudelus, Bondy; C. Floch, F. Mazy, F. Meier, and M. Robin, Colombes; M.C. Allemon, St. Denis; P. Talon, Montfermeil; D. Berterottiere, Neuilly; M.C. Rousset, Poissy; P. Wipff, Mantes; A. Vinas, Versailles; P. Narcy and D. Armengaud, Saint-Germain; F. Guillot, Villeneuve St. Georges; A. Leblanc and A. May, Evry; D. Douard and H. Fleury, Bordeaux; J. Brouard and F. Freymuth, Caen; D. Valle, G. Michel, and C. Tamalet, Marseilles; J. Nicolas, Montpellier; A. Deville, F. Monpoux, and J. Cottalorda, Nice; X. Hernandorena, Bayonne; V. Brossard, Rouen; and J. Tricoire, A. Robert, and J. Puel, Toulouse.
The coordinating team included M.J. Mayaux, C. Rouzioux, S. Blanche, C. Laurent, A. Doussin, and Y. Moudoub.
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
HIV-1 in Newborns
Speiser D. E., Wyler C.-A., Siegrist C.-A., Blanche S., Mayaux M.-J., Rouzioux C.
Extract |
Full Text
N Engl J Med 1994;
330:1905-1906, Jun 30, 1994.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
