Human Cytomegalovirus UL144 Is Associated with Viremia and Infant Development Sequelae in Congenital Infection

Specimen and patient details.The National Virus Reference Laboratory of Ireland detected 29 individuals with cCMV by detection of early antigen fluorescent foci (DEAFF) analysis (12) of urine during 2006 and 2007. cCMV infection was confirmed by detection of the virus within the first 3 weeks of life. Ethical approval for all patients was granted from the source hospitals. CMV-positive patients identified during the same period were included in the study for comparison (n = 57), specifically, postnatally infected infants (postnatally acquired CMV [pCMV]; n = 19/57). Adult CMV infection was detected in patients with HIV infection, posttransplantation patients, and patients undergoing dialysis and in chronic respiratory tract infections.

Plasma viral load.The patient PVL was measured using the Cobas Amplicor HCMV Monitor Test (Roche Diagnostics, United Kingdom) according to the manufacturer's protocol. The assay's quantitation range was between 600 and 1,000,000 copies/ml. Samples with amplification for CMV DNA but with viral loads of <600 copies/ml were deemed positive, but the viral load was not recorded. Following a positive DEAFF test result, 23/29 cCMV patients requested PVL testing.

HCMV genotyping.Urine specimens were extracted on the Roche MagNA pure using the total nucleic acid kit (Roche Diagnostics, United Kingdom). The UL144 gene (737 bp) was amplified from the extracted urine as described by Lurain et al. (16). The amplicons were purified with a QiaQuick PCR purification kit (Qiagen, United Kingdom) and sequenced on an ABI Prism 3.10 using the Big Dye sequencing kit (Applied Biosystems) according to the manufacturer's instructions. Sequence trace files were analyzed by Lasergene SeqMan Pro software version 7.2.1 (DNAStar). Sequences were aligned in Clustal W (http://www.ebi.ac.uk/Tools/clustalw/index.html ), and maximum-likelihood trees were constructed in PAUP* version 4.0 Beta 10 (Sinauer Associates, Inc., United Kingdom). The model of evolution was selected using Modeltest version 3.7 (21). Bootstrap resampling was carried out for 1,000 replicates of the data set. Genotypes were determined from the phylogenetic tree and Basic Local Alignment Search Tool analysis (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi ).

Luminex assay; cytokine and chemokine detection.Multianalyte cytokine and chemokine profiling was performed using the Luminex-100 system and the XY platform (Luminex Corp., TX). Plasma or serum specimens received within 2 weeks of the initial DEAFF-positive specimen (n = 20/29 cCMV patients) were analyzed for the presence of the following cytokines and chemokines by using a human cytokine Lincoplex kit according to the manufacturer's instructions (Millipore, Miami): interleukin 1α (IL-1α), IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12, tumor necrosis factor alpha (TNF-α), gamma interferon (IFN-γ), and macrophage inflammatory protein 1α (MIP-1α). No serum samples from healthy neonates were used as controls for ethical reasons. Neonatal cytokine and chemokine control levels were determined using HCMV-negative healthy adult serum specimens and from published neonatal and pediatric cytokine “control” values (14, 17, 22, 23). In addition, cCMV cytokine and chemokine levels were compared to those of other, noncongenital HCMV infections identified during the same period.

Clinical manifestation of disease and developmental outcome.Clinical information was obtained retrospectively by chart review for 18/29 cCMV patients. Infants were clinically assessed at birth for evidence of the following possible manifestations of cCMV infection: intrauterine growth retardation (IUGR), microcephaly, hepatosplenomegaly, lethargy, hypotonia, poor suck, seizures, and dysmorphic features. At 12 to 18 months postinfection, the Bayley III criteria were used to assess surviving children for developmental delay. Specific cCMV clinical manifestations, such as SNHL and mental retardation, were recorded. The patients were then subdivided into groups that had either cCMV clinical manifestations and developmental delay or normal development at 12 to 18 months postinfection.

Statistical analysis.Statistical analysis was performed using Statview software, version 5.0.1. Mann-Whitney U, Spearman's correlation, and chi-square analyses were used where appropriate, and a P value of less than 0.05 was considered statistically significant.

Nucleotide sequence accession numbers.The GenBank accession numbers for all the Irish HCMV sequences are GQ855960 to GQ856042.

HCMV genotype.The UL144 gene was successfully amplified from 27/29 urine specimens from the cCMV-infected infants, and the sequences were genotyped following the construction of a maximum-likelihood phylogenetic tree (Fig. 1 and Table 1). All HCMV UL144 genotypes were detected (Table 2), and within each genotypic group, approximately 2 to 6 nucleotide differences were observed between strains. UL144 A consisted of two subtypes, A1 and A2, and there were approximately 10 nucleotide changes between the subtypes within the region sequenced. All genotypes were found to be circulating throughout the 2-year period, and there was no significant increase in a specific genotype at any time, indicating that no particular strain(s) dominated during the study period.

• Open in new tab
• Download powerpoint

FIG. 1.

Maximum-likelihood phylogenetic tree based on UL144 gene sequences from Irish congenital human cytomegalovirus infections identified during 2006 and 2007. UL144 gene sequences from cCMV and pCMV infections were genotyped using a phylogenetic tree (black). Adult and reference HCMV sequences are shown in gray. The GenBank accession numbers for the reference sequences (gray and boxed) are as follows: NW004 (AF179196), NW007 (AF179197), NW009 (AF179199), NW012 (AF179200), NW013 (AF179201), NW016 (AF179204), NW020 (AF179206), NW024 (AF179209), PT001 (AF084976), PT002 (AF084977), PT015 (AF084990), PT016 (AF084991), PT017 (AF084992), PT019 (AF084994), PT023 (AF084998), PT025 (AF085000), PT028 (AF085005), and PT030 (AF085004).

UL144 genotypes A, B, and C were detected within the cCMV patient population at 33.3%, 29.6%, and 25.9%, respectively (Table 1). Three recombinant strains were also detected, 2 UL144 A/B sequences (7.4%) and 1 UL144 A/C sequence (3.7%). The cCMV genotypes were compared with those identified from other HCMV infections, specifically, those detected in pCMV infection and adult immunodeficient patients. No particular strain was associated with any HCMV-positive group. For cCMV and pCMV infections, UL144 A was the most prevalent strain and was detected at 33.3% and 42.1%, respectively. UL144 B was the second most common genotype present in cCMV (29.6%)- and pCMV (31.6%)-infected patients. In adult HCMV infection, the opposite was observed. UL144 B was the most prevalent (39.5%), followed closely by UL144 A (31.5%). UL144 C was the least common genotype detected in all patient groups analyzed: cCMV (25.9%), pCMV (21.1%), and adult HCMV (15.8%) infection. Recombinant strains were rare, but both A/B and A/C were detected; however, no UL144 B/C recombinants were identified.

Plasma viral load in cCMV infection.Assessment of the PVL was recommended following confirmation of cCMV infection, and follow-up blood specimens for 23 of the 29 congenitally infected infants were sent for analysis (Table 1). HCMV viremia was confirmed in 20/23 patients tested, and the viral loads ranged from 600 to >1,000,000 copies/ml (median, 2,770 copies/ml; mean, 57,963 copies/ml). The patients for whom a PVL was determined were analyzed according to their associated UL144 genotypes (n = 18/20). The median PVLs for infections caused by genotypes UL144 A (n = 7; PVL range, 600 to >1,000,000 copies/ml) and UL144 C (n = 4; PVL range, 708 to 975,175 copies/ml) were 19,800 and 8,635 copies/ml, respectively. UL144 A and C were significantly associated with high PVLs (P < 0.04; Mann-Whitney U test) compared with the PVLs achieved following UL144 B infection (n = 5; PVL range, 600 to 2,870 copies/ml) (Fig. 2). Too few recombinant strains were identified to be included in this analysis (UL144 A/C, n = 1; UL144 A/B, n = 1); however, the PVL for the one A/C strain was 25,600 copies/ml, which is within the high PVL range of this study (Table 1).

• Open in new tab
• Download powerpoint

FIG. 2.

Association of the human cytomegalovirus UL144 genotype with the plasma viral load. The HCMV PVL was significantly associated with UL144 genotypes A and C (P < 0.04; Mann-Whitney U test) compared to genotype B. PVLs were investigated for 23 cCMV patients. The UL144 genotypes were as follows: A (n = 7/23), B (n = 5/23), C (n = 7/23), A/B (n = 1/23), A/C (n = 1/23), and not assigned (n = 2/23). No HCMV was detected in the plasma of 3/23 cCMV patients. For each genotype, the median PVL was as follows: UL144 A, 11,620 copies/ml; UL144 B, 600 copies/ml; and UL144 C, 2,670 copies/ml. For each genotype, the mean PVL was as follows: UL144 A, 135,631 copies/ml; UL144 B, 1,004 copies/ml; and UL144 C, 198,631 copies/ml.

cCMV clinical manifestations and developmental outcome.Clinical manifestations of cCMV infection at birth, such as jaundice, microcephaly, hepatosplenomegaly, and intracranial abnormalities, were used to categorize HCMV infection as symptomatic. No specific HCMV genotype was significantly linked with symptomatic infection, and all UL144 genotypes were detected in symptomatic neonates. However, as there is no neonatal HCMV screening program in Ireland, the majority of newborns included in the study were symptomatic at the time of testing, and only 3 out of the 18 patients for whom the full clinical data were available were asymptomatic. It should be noted, though, that all 3 asymptomatic cases were caused by infection with a UL144 B HCMV strain (Table 1).

For the above reasons, our analysis focused on the prediction of cCMV outcome at 12 to 18 months postinfection as opposed to the manifestations of disease at initial diagnosis. Long-term clinical conditions, including SNHL (bilateral or unilateral), mental retardation, and developmental delay at 12 to 18 months postinfection, were used to determine if the infected neonate had continuing cCMV sequelae. Follow-up clinical data were available for the surviving cCMV neonates (16/18). Patient developmental delay was categorized based on communicative development (receptive and expressive), motor development, and global general development. Following assessment, a total of 11 children had some form of developmental delay, 6 of whom developed SNHL. Five patients had normal development at 12 to 18 months postinfection. Finally, 2 cCMV-related deaths were recorded (Table 3).

All UL144 genotypes were associated with cCMV-related manifestations and developmental delay; however, only patients infected with genotypes UL144 B and A/B (n = 5) could be described as having a normal and unaffected clinical outcome (Table 2). In addition, a significant association between cCMV clinical sequelae and UL144 A and C (n = 7) was observed (P < 0.04; chi-square analysis). Furthermore, the two cCMV-related deaths were found to be caused by a UL144 A (PVL, 23,100 copies/ml) and a UL144 C (no CMV DNA detected in the blood) strain. Unfortunately, the sample population size precluded analysis of individual cCMV outcomes with genotype.

cCMV-induced cytokine response.The levels of 13 cytokines and chemokines were determined in sera or plasma of the cCMV patients (n = 20), the pCMV infants (n = 9), and a proportion of adult HCMV-infected individuals (n = 27). IL-8 and IL-10 were significantly increased following cCMV infection (Table 1) (P < 0.02 and P < 0.0001, respectively; Mann-Whitney U analysis) and in all HCMV-positive individuals (Fig. 3). In addition, a significant positive correlation between IL-10 levels and PVLs was observed (P < 0.03; r = 0.527; n = 17; Spearman correlation). IL-10 levels, PVLs, and clinical data were available for 10 cCMV patients, and high PVLs and IL-10 levels were associated with cCMV long-term clinical manifestations (Fig. 4). No significant differences in the levels of all other cytokines and chemokines assayed were detected (data not shown).

• Open in new tab
• Download powerpoint

FIG. 3.

Cytokine responses to cCMV infection. Circulating IL-8 and IL-10 were significantly increased following cCMV infection with respect to healthy adult control levels (P < 0.015 and P < 0.0001, respectively; Mann-Whitney U test). The healthy adult control cytokine levels were compared with published neonatal and pediatric cytokine control levels and were within the ranges given (14, 17, 22, 23). cCMV cytokine levels were also compared to the levels detected in pCMV-infected infants and an adult immunodeficient HCMV-positive population. Data are shown as box and whisker plots. The top and bottom of the boxes represent the 75th and 25th percentiles, respectively, and the band in the middle of the boxes represents the 50th percentile (median). The error bars represent the 5th and 95th percentiles.

• Open in new tab
• Download powerpoint

FIG. 4.

The plasma viral load is correlated with IL-10 levels. IL-10 levels and PVLs are plotted with respect to the accompanying patient clinical data (n = 10). N, no concerns following cCMV infection (n = 3); CM, cCMV-related clinical manifestations (n = 7). IL-10 levels were significantly correlated with the PVL (P < 0.03; r = 0.527; Spearman correlation).