CONGENITAL CYTOMEGALOVIRUS INFECTION IN A HIGHLY SEROPOSITIVE SEMI-URBAN POPULATION IN INDIA

Abstract

To determine the incidence and natural history of congenital cytomegalovirus (CMV) infection in a population of women with near universal serologic reactivity for CMV, a prospective study of 423 women attending the antenatal clinic of the Comprehensive Rural Health Center in northern India was conducted. All 9 (2.1%) CMV positive infants were born to mothers who were CMV seropositive at the first antenatal visit. One child had hepatosplenomegaly at birth and another child had mild unilateral hearing loss at 4 months of age.

Cytomegalovirus (CMV) is a common cause of congenital infection and a frequent cause of hearing loss in children. The frequency of congenital CMV infection among different populations has been estimated to vary between 0.2% and 5.4%.^1,2 The infection is subclinical or asymptomatic at birth in about 90% of cases but can eventually cause sequelae including sensorineural hearing loss in 10–15% of children.¹ Populations from developing countries including India and low-income populations from developed countries are highly seropositive for CMV infection, with rates ranging from more than 80% to nearly 100%.^1–3

A limited number serological studies, conducted at tertiary care hospitals, have attempted to define the incidence of congenital CMV infections in India. All have depended solely on an IgM enzyme-linked immunosorbent assay for diagnosis of congenital infection. IgM antibodies to CMV could be detected in 18–20% of neonates with clinically suspected congenital infection and in cord blood of 1.8% of all live births in the few such studies published from India.^4,5 We report the results of a prospective study of highly seropositive women attending the antenatal clinic of the Comprehensive Rural Health Center (CRHC) in northern India, based on virus detection using a polymerase chain reaction (PCR) for CMV DNA in neonatal saliva to describe the incidence and clinical features of congenital CMV infection.⁶

METHODS

A total of 1473 women attending the antenatal clinic of the CRHC at Ballabgarh, Haryana, India, were enrolled sequentially in the study at the time of their first antenatal visit. After obtaining written informed consent, a peripheral blood sample was collected from each study subject, along with clinical history and demographic information.

Only the women who returned to the CRHC for delivery (n = 423) were included in the analysis of congenital infection. These women had a mean age of 22.3 ± 2.8 years (median age, 22 years; range, 17–34 years), a median gravidity of 2, and a median parity of 0; these parameters were similar to the 1473 women originally enrolled (mean age, 22.4 ± 3.1 year; median age, 22 years; age range, 17–35 years; median gravidity, 2; and median parity, 0). Sera collected from these women at their first antenatal visit were tested for CMV IgG antibodies (Biokit, Spain). Saliva samples were collected at birth from the neonates born to these mothers, using a cotton-tipped swab in viral transport medium.⁶ Sera collected at the first antenatal visit from mothers of infants with a positive CMV PCR in saliva were tested for the presence of CMV IgM antibodies (Biokit, Spain).

The saliva was tested for the detection of cytomegalovirus DNA by a nested PCR. DNA was extracted from the saliva using commercial column kits (Qiagen Inc., Los Angeles, CA). An external PCR was performed to amplify the DNA of the gB gene of CMV (amplicon size, 501), followed by a nested PCR to obtain a final amplicon of 249 bp, adapting primers, and protocols described previously.^7,8 Annealing for both external and nested PCR was done at 55°C. Amplified products were resolved using agarose gel electrophoresis.

A urine sample from babies found to be positive for CMV DNA was obtained for confirmation of congenital CMV infection by PCR within the first 3–4 weeks of birth at the first follow-up visit. A thorough clinical examination was performed at this clinic visit. During subsequent follow-up visits, the CMV positive babies were tested for hearing function at the age of 4 months by auditory brain-stem evoked response audiometry (ABR). Sensorineural hearing loss was defined as air conduction thresholds of >25 dB on ABR.⁹

Maternal demographic information and newborn information were recorded on standardized case report forms. The categorical variables were compared with Fisher exact test, and continuous variables were analyzed using a 2-sided t test or Wilcoxon rank sum test, as appropriate.

RESULTS

Of the 423 women who returned to the CRHC for delivery, CMV IgG antibodies could be detected in the first antenatal serum sample in 419 women [99.0%; 95% confidence interval (CI): 97.6–99.7].

Nine (2.1%; 95% CI: 1.0–4.0) neonates were positive for CMV DNA in saliva specimens obtained at birth. All cases of congenital CMV infection were confirmed by a PCR performed on urine samples obtained at the time of their enrollment visit in the follow-up study.

All 9 infants who had congenital CMV infection were born to women with serological reactivity for CMV in the first antenatal specimen (enrollment serum). The antenatal sera from the mothers of 9 PCR positive infants were negative for CMV IgM antibodies.

The clinical and anthropometric details of the CMV positive neonates are shown in Table 1. On clinical examination, hepatosplenomegaly was found at birth in 1 of the 9 babies. The remaining infants had no clinical abnormalities on physical examination. Sera collected from all babies at birth tested negative for syphilis by the Venereal Disease Research Laboratory and for rubella IgM antibodies (Biokit, Spain). During a follow-up visit at the age of 4 months, ABR evaluation revealed a mild unilateral hearing loss (at 40 dB) in 1 of the 8 asymptomatic infants at birth. The infant with hepatosplenomegaly had normal hearing.

TABLE 1: Characteristics of the CMV Positive Babies Born to Seropositive Mothers (n = 9)

Various maternal and neonatal parameters including maternal age, gravidity, parity, gestational age, birth weight, head circumference, length, and APGAR scores were compared between the CMV positive infants and the CMV negative infants. There were no significant differences between the CMV positive group and the CMV negative group, except gravidity (2-tailed P value = 0.04; Wilcoxon rank sum scores), which showed a median of 1 (range, 1–2) in the CMV positive infants, versus a median of 2 (range, 1–6) in the CMV negative infants.

DISCUSSION

The paucity of reliable data from developing countries on the incidence and natural history of congenital CMV infection becomes quite evident from the recent review and meta-analysis of the epidemiology of congenital CMV infection.¹⁰ In this report, we describe the findings of a prospective virus detection-based study of congenital CMV infection in India, a country with a very high CMV seroprevalence.³ Unlike the few previous reports from India, which derived the data from suspected cases of congenital CMV infection presenting to tertiary care hospitals or on live births occurring at such hospitals, this study examined a low-income population that is highly seropositive for CMV infection seen at a rural health center.^4,5 While the previous reports based their diagnosis on serological assays, we used a PCR technique for the detection of CMV DNA in saliva specimens from newborns. This technique seems to be at least as sensitive in the detection of congenital CMV infection as detection of viruria.⁶ Nevertheless, positive neonates detected in the newborn screening assays were subsequently confirmed by PCR performed on their urine samples.

Studies from various populations with high maternal seroprevalence rates, such as Ivory Coast, Brazil, Chile, Mexico, and Gambia have shown an incidence of congenital CMV infection between 0.89 and 5.4%.^2,10 High maternal seroprevalence rates have been consistently associated with a high incidence of congenital infection, even in populations from developed countries.^10,11 Our study further substantiates this finding, as the incidence of congenital CMV infection was found to be 2.1% (95% CI: 1.0–4.0) despite a seroprevalence rate of about 99% in our study population. Various studies have demonstrated that nonprimary CMV infection (reinfection or reactivation) in mothers with preexisting antibodies can also lead to symptomatic infections at birth as well as sequelae similar to those seen after primary infections.^12,13 Though the number of CMV positive neonates in our study was small (9 of 423 live births), all positive infants were born to mothers with preexisting antibodies. None of the 9 mothers of infected infants had CMV IgM antibodies at the time of enrollment. These values are similar rates reported from studies in Northern Europe and North America.^1,10

When the maternal demographics and neonatal characteristics of the CMV positive and CMV negative groups were compared, none showed a significant difference between the 2 groups except gravidity (P = 0.04). Although other studies have shown that babies born to primigravid mothers had a higher incidence of congenital infection than those born to multigravid mothers, the small number of infected infants in our study limits the significance of this finding.¹⁴

A major limitation of this study is that of the 1473 women originally enrolled, only 423 returned for delivery to the CRHC. Approximately 50% of women still undergo home deliveries in rural and semi-urban India.¹⁵ It is also common for women to return to their parental home for delivery. It is possible that the incidence of congenital CMV infection documented in our study may not represent a true population-based estimate because of the large number of women who did not return to the CRHC for delivery. However, this possibility is less likely because of the similar demographic characteristics between the 2 groups.

ACKNOWLEDGMENTS

The authors acknowledge the valuable contributions of Dr. Sunil Kumar Verma, Mr. Sanjeev Kumar, and Ms. Shelly in the field work for the study.