Pneumocystis carinii pneumonia and cytomegalovirus infection in children with vertically acquired HIV infection

Objectives: The outcome of Pneumocystis carinii pneumonia (PCP) in HIV-infected infants is poor, and the role of cytomegalovirus (CMV) co-infection in the course and outcome of PCP is unclear. This study describes the prevalence, clinical characteristics, management and changes in survival over time of vertically HIV-infected infants developing PCP and/or CMV infection.

Cited Here...: Data on children with HIV, born in the UK and Ireland and reported to the National Study of HIV in Pregnancy and Childhood, with PCP and/or CMV were combined with clinical information collected from reporting paediatricians.

Cited Here...: By April 1998, 340 vertically HIV-infected children had been reported, of whom 93 had PCP and/or CMV, as their first AIDS indicator disease; 85 (91%) were infants. Among infants with PCP, 79% were born to mothers not diagnosed as HIV infected, and there was an independent and statistically significant association with breast-feeding, being black African, and developing CMV disease. Median survival after PCP and/or CMV was significantly better in those born between 1993 and 1998 compared with those born before 1993 (P = 0.009), and worse than after other AIDS diagnoses (P = 0.01). Infants with dual infection were more likely to be ventilated (P = 0.003) and receive corticosteroids (P = 0.002) than those with PCP alone.

Conclusion: Although survival from PCP and CMV has improved over time, these remain serious and potentially fatal infections among infants in whom maternal HIV status is not recognized in pregnancy. Breast-feeding increases the risk of combined PCP and CMV infection, which is associated with severe disease.

Introduction

Among HIV-infected babies who have not received Pneumocystis carinii pneumonia (PCP) prophylaxisPCP is a common first AIDS indicator disease, usually occurring in the first 6 months of life ^[1,2], with reported survival rates of only 38-62%^[1-4]. Cytomegalovirus (CMV) is often isolated at the time of PCP diagnosis from HIV-infected adults ^[5] and children ^[4]. Without evidence from lung biopsy, it can be difficult to discriminate asymptomatic viral shedding from active disease ^[4-6].

In this study, we describe the epidemiology, clinical characteristics, management and changes in survival over time, of children born in the UK and Ireland who had PCP, CMV disease or both infections between 1989 and 1998, focusing on infants in particular.

Methods

Active surveillance of pregnant women and children with HIV infection in the UK and Republic of Ireland has been carried out since 1989 through the Royal College of Obstetricians and Gynaecologists and the British Paediatric Surveillance Unit to form the National Study of HIV in Pregnancy and Childhood (NSHIPC) ^[7]. Infected children are followed annually to monitor clinical progression.

Paediatricians who had reported cases of PCP, CMV or co-infection (PCP + CMV) in infants less than 12 months of age were asked to provide further details by a separate questionnaire. Information about diagnostic methods, specific treatments, mechanical ventilation, primary and secondary prophylaxis and clinical outcome were requested. When multiple diagnostic methods were reported, they were ranked from the most to the least definitive as follows: post-mortem findings; histology from biopsy specimens; bronchoalveolar lavage (BAL); tracheal aspirate and nasopharyngeal aspirate. For CMV, further tests were ranked: positive culture of virus from blood; detection of early fluorescent foci (DEAFF) testing in blood; detection of CMV DNA by polymerase chain reaction in blood; urine culture and urine DEAFF positivity. Data obtained from the questionnaires and NSHIPC were combined and analysed using the Statistical Analysis System (Cary, NCa, USA).

Survival was measured from the date of the first PCP or CMV diagnosis until the last follow-up visit or death. Survival from AIDS was analysed using Weibull survival models.

Results

By April 1998, 531 children with perinatally acquired HIV had been reported. A total of 183 (54%) of the 340 children born in the UK or Ireland had an AIDS diagnosis, of whom 104 (57%) were black African. Ninety-three children were reported to have PCP, CMV or PCP + CMV as their first AIDS indicator disease (including 22 previously reported ^[1]).

Information about feeding was available for 70 (84%) of the 83 children who presented with PCP as their first AIDS indicator disease. Only 19 (51%) with PCP alone were breast-fed compared with 29 (88%) who had co-infection with PCP and CMV (χ² = 9.17, P = 0.002). A diagnosis of PCP alone was not associated with ethnic origin. However, in a multivariate analysis, CMV infection in association with PCP was independently associated with being black African compared with other ethnic groups [odds ratio (OR) 3.8, 95% confidence interval (CI) 1.2-16.0] and with breast versus bottle-feeding (OR 4.5, 95% CI 1.1-12.5).

Infants with Pneumocystis carinii pneumonia and/or cytomegalovirus

Eighty-five (91%) children presenting with PCP or CMV as their first AIDS indicator disease were under 12 months (76 under 6 months, nine 6-12 months), compared with only 30% of children with other first AIDS indicator diseases. In 60 (79%) of the 76 infants with PCP or PCP + CMV, HIV infection was only diagnosed after presentation with PCP. Of the remaining 16, four received prophylaxis which failed, and 12 did not receive prophylaxis (seven presented before the publication of UK guidelines ^[8], in three, parents refused and in two the reason was unknown).

Method of diagnosis

Information on methods of diagnosis and management was available for 80 infants with PCP or CMV as a first (n = 78) or subsequent (n = 2) AIDS diagnosis (PCP = 37, CMV = 8, PCP + CMV = 35). PCP was diagnosed by BAL in 44 (61%) infants. In the remaining 28, the diagnosis was made at post-mortem (two), by lung biopsy (two), tracheal aspirate (four), nasopharyngeal aspirate (six) and on clinical grounds alone (14). Among infants with CMV, the majority had isolated lung disease. Retinitis occurred in nearly half of those with multiple site disease (Table 1). Only two (7%) infants with isolated lung disease had a definitive diagnosis compared with 10 (71%) with multiple site disease.

Table 1

Management

All but one of the 72 infants with PCP received high-dose co-trimoxazole therapy (one died before drug therapy commenced). Pentamidine was also given to 23 infants, either because of an allergic reaction to co-trimoxazole (nine) or because of failure to improve (14). One child received additional atovaquone. Of infants with PCP + CMV, 91% received adjuvant corticosteroids compared with 59% with PCP alone (χ² = 9.8, P = 0.002).

Fifty-three (66%) of the 80 infants were ventilated, of whom 11 also received high-frequency oscillation. Three infants also received nitric oxide, seven surfactant and one both interventions. Infants with PCP and CMV co-infection were significantly more likely to be ventilated (χ² = 9.1, P = 0.003) and to receive high-frequency oscillation ventilation (χ² = 5.7, P = 0.02) than those with PCP alone. Overall, 74% of infants with a diagnosis of CMV received ganciclovir.

Survival

Median survival from the time of PCP diagnosis increased from 3.5 months (95% CI 1.8-6.8) for infants born before 1993 to 14.2 months (95% CI 6.0-33.3) for infants born subsequently (χ² = 6.7, P = 0.009). A trend towards poorer median survival with dual infection compared with PCP alone [1.9 months (95% CI 0.7-5.0) before 1993, increasing to 7.8 months (95% CI 3.7-16.4) afterwards), did not reach statistical significance (P = 0.3). Children with other AIDS diagnoses in infancy had significantly better survival rates than those with PCP or CMV disease, median survival 13.4 months (95% CI 5.5-32.9) before the end of 1992 and 54.4 months (95% CI 18.5-159.9) subsequently (χ² = 6.1, P = 0.01;Fig. 1). Infants requiring mechanical ventilation had significantly poorer survival, compared with those not ventilated (χ² = 9.2, P = 0.002). The use of steroids in infants with PCP (either alone or with CMV) did not significantly influence survival (χ² = 0.2, P = 0.6).

Fig. 1

Secondary prophylaxis

Thirty-nine infants survived PCP, of whom 35 (90%) received secondary prophylaxis (with co-trimoxazole, 32; dapsone, two; and pentamidine, one). Parents of two infants refused prophylaxis and two failed to receive prophylaxis for unknown reasons. Secondary CMV prophylaxis was prescribed to 16 of the 29 survivors of CMV disease, including all children with CMV retinitis (intravenous ganciclovir, five; foscarnet, one; cidofavir, one; and oral acyclovir, nine).

Discussion

In the decade from 1989, PCP was the commonest AIDS-defining diagnosis, occurring in 68% of reported infected infants born in the UK and Ireland. In almost all, the mother was unaware of her HIV diagnosis before the child developed PCP. In 1998, the uptake of interventions to reduce vertical transmission increased to over 95% among women who knew their HIV diagnosis in pregnancy, and transmission rates fell to an estimated 2%^[9]. In addition, and similar to other reports ^[10], PCP in infants born to mothers known to have HIV declined after guidelines were published advocating early PCP prophylaxis for all infants born to infected women ^[8,9].

Since the completion of this study and up to July 1999, 15 further infants were notified to the British Paediatric Surveillance Unit with PCP (five), CMV (two) and dual infection (eight). Four died during the acute illness, all with dual infections. Survivors were all taking highly active antiretroviral therapy (HAART); none have been reported with subsequent AIDS indicator diseases (P.A. Tookey, personal communication). Many also have undetectable viral loads and high CD4 cell counts ^[11]. It is apparent that if the infant survives the initial illness, the prognosis is much improved in the era of HAART ^[11]. However, HAART may not prevent death during the acute illness in infants not previously known to be at risk of HIV.

Among infants with PCP, we observed independent associations between breast-feeding, black African ethnicity and the risk of CMV. We restricted the analysis of risk factors for CMV to cases of PCP, as these infants had similar diagnostic tests and therefore a similar chance of identifying CMV. The acquisition of CMV infection in infancy is more common in infants breast-fed by CMV-seropositive mothers, and CMV seroprevalence is higher in black compared with white women ^[12]. Among HIV-infected infants, a higher rate of CMV acquisition has also been related to the degree of maternal immunosupression ^[13].

Improved survival between 1989 and 1998 may be the result of the earlier diagnosis and more aggressive treatment of PCP and CMV infections, as well as the availability of more antiretroviral drugs, which even before the availability of HAART have been shown to improve survival in children ^[14,15]. HAART, including protease inhibitor or non-nucleoside reverse transcriptor inhibitor drugs, was not available for children in the UK and Ireland until late 1997. In this study, survival was significantly worse for infants with PCP or CMV disease compared with other AIDS diagnoses. In contrast to previous studies in both adults and children ^[16-19], the use of adjuvant corticosteroids was not associated with improved survival from PCP. This finding may be partly explained by the high prevalence of CMV co-infection among children with PCP in our study (48%) compared with only 0-18% in previous studies ^[17-19]. Histological evidence of unsuspected CMV has been reported in adults and children at autopsy ^[20,21], but whether corticosteroids increase this is unclear ^[18,22,23]. In infants, CMV is a primary infection and corticosteroids could adversely affect the course of CMV disease.

The range of reported diagnostic methods for CMV lung disease illustrates the difficulty of making a definitive diagnosis of CMV disease in children, particularly with isolated lung rather than multiple organ disease. An association between CMV infection and HIV disease severity and progression has been reported in infants ^[12]. It remains unclear whether CMV is merely a marker for more severe immune suppression, or contributes by either enhancing HIV replication ^[24] or entry into cells ^[25].

No child in this study received primary CMV prophylaxis. Secondary prophylaxis was given to all infants with CMV retinitis and 43% of those with CMV isolated from other sites. Life-long CMV maintenance therapy with ganciclovir or foscarnet has been recommended in adults and children with CMV end-organ disease. However, as infants who recover from PCP or CMV disease commence HAART and develop undetectable HIV-RNA levels and substantial increases in CD4 lymphocye counts, secondary PCP or CMV prophylaxis may no longer be necessary.

This study of vertically HIV-infected infants shows that survival has improved over time among all infants developing AIDS, but the prognosis is worst in those with PCP or CMV, or both. The antenatal identification of HIV-infected women is a priority, as it not only reduces the number of infected children, but also reduces the risk of infected infants developing PCP, if primary prophylaxis is given ^[9,10]. Our data suggest that infants developing PCP should be investigated for CMV infection and retinitis, especially if they were breast-fed. In addition, anti-CMV therapy should be strongly considered in infants with PCP and evidence of CMV infection, especially in those who receive adjuvant corticosteroids. This strategy is already being followed in some London centres.

Acknowledgements

The authors would like to thank the British Paediatric Surveillance Unit of the Royal College of Paediatricians and Child Health, their paediatric respondents and especially those paediatricians who provided additional information on children with PCP or CMV. They also thank the obstetric respondents to the surveillance scheme, run under the auspices of the Royal College of Obstetricians and Gynaecologists, and acknowledge the support of colleagues at the Communicable Disease Surveillance Centre, Colindale and the Scottish Centre for Infection and Environmental Health.