High prevalence of breastmilk‐acquired cytomegalovirus infection in jaundiced infants

Abstract Background Our objective was to evaluate the prevalence and different diagnostic methods of breastmilk (BM)‐acquired cytomegalovirus (CMV) infection in a pathologically jaundiced cohort. Methods A total of 400 infants confirmed with pathological jaundice at The People's Hospital of Qingyang City were screened for BM‐acquired CMV infection between February 2018 and February 2019. A total of 300 infants were finally enrolled in our study. CMV infection was confirmed by detecting both CMV‐DNA in various samples using FQ‐PCR and CMV‐IgM with chemiluminescence. Clinical and other laboratory data were collected from these infants during their hospitalization or regular visits. Results Ninety‐eight (32.67%) subjects were confirmed to be BM CMV‐DNA–positive, and 18 (18.37%) were diagnosed with a BM‐acquired CMV infection. All 18 (100%) infants with a BM‐acquired CMV infection were CMV‐DNA–positive in urine, while 5 (27.78%) cases and 11 (61.11%) cases were confirmed in plasma and peripheral blood mononuclear cells (PBMCs), respectively. Only 6 (33.33%) infants were CMV‐IgM–positive. Birthweight, direct bilirubin, aspartate aminotransferase, and the viral load in BM of the BM‐acquired CMV group were higher than those in the non‐infected group (P < .05). Low birthweight and viral load in BM were risk factors for BM‐acquired CMV infection. Detecting CMV‐DNA in urine samples exhibited better performance than the other methods for screening BM‐acquired CMV infections. Conclusions Our study found a high prevalence of BM‐acquired CMV infection in jaundiced infants, and detecting CMV‐DNA in a urine sample was the most sensitive method for disease screening.


| INTRODUC TI ON
Cytomegalovirus (CMV) (formal name: human herpesvirus 5) infection is common in China, and a primary CMV infection often occurs in infants. 1,2 CMV has the biological properties of latency and activation, and the infection remains throughout life. As a weak pathogen, CMV is not usually pathogenic in the population with normal immune function, and most cases of CMV infection are asymptomatic. CMV generally causes serious consequences in the immunosuppressed population, particularly fetuses and neonates with developmental immunodeficiencies. [3][4][5][6] CMV infection is frequently seen in mothers, and CMV can be excreted via breastmilk (BM) in 13%-27% of cases. BM is the best source of nutrition for neonates. The American Academy of Pediatrics has reported that BM feeding is very important for babies ≤6 months old, and BM is irreplaceable, particularly in premature infants; thus, CMV infections via BM are worthy of research attention. 7,8 As a common disease during the neonatal period, neonatal jaundice is mostly a physiological phenomenon during the normal growth process, but is occasionally a clinical manifestation of some latent disease.
As shown by clinical analysis, pathological jaundice is mostly caused by an infection, of which the most common is a CMV infection. [9][10][11] A BMacquired CMV infection usually causes no clinical symptoms in healthy neonates and has a low probability of causing deafness and nervous system sequela. However, the risk of serious sequela from CMV may increase in pathologically jaundiced infants. 12

| Diagnosis of CMV infection
The diagnosis of CMV infection can be performed in the fol- All laboratory data were obtained under conditions of a negative control, a positive control, and an indoor quality control.

| Processing of different types of samples before the CMV-DNA test
Urine sample 10 mL of mixed urine from three urinations (including urina sanguinis) was collected from subjects, sealed, and sent for examination. After mixing, 1 mL of the urine sample was centrifuged at 12 000 rpm for 5 minutes, and the supernatant was discarded.

| Collection of blood samples and tests of the laboratory
The venous blood was drawn TBIL, DBIL, adenosine deaminase (ADA), total bile acid (TBA), alkaline phosphatase (ALP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), and gamma-glutamyltransferase (γ-GGT) were detected with an AU2700 full-automatic biochemical analyzer (Japan); the complete cell count was measured with a Mindray BC-5800 blood cell analyzer. 100-300 × 10 9 /L. When the test results exceeded the upper limit of the references, an increase in the indices was indicated.

| Data analysis
Data were analyzed with SPSS (version 20.0) software (SPSS Inc).
Proportions were calculated by the chi-square test or Fisher's exact test, and numerical data were assessed by the t test. The diagnostic performance was assessed by the area under the receiver operating characteristic curve (AUC). In this study, an AUC > 0.7 was considered useful. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. P-values < .05 were considered significant.

| Comparison of clinical characteristics and laboratory indices between the BM-acquired CMVinfected and non-infected groups
The birthweight, DBIL, AST, and viral load in BM were significantly different between the BM-acquired CMV-infected and non-infected groups (P < .05). However, age, gestational age, onset time of jaundice, TBIL, ADA, TBA, ALP, ALT, γ-GGT, WBC, and PLT were not significantly different between these groups (P > .05; Table 1).

| Risk factors for BM-acquired CMV infection
The following possible risk factors for BM-acquired CMV infection were analyzed: small-for-gestational-age infants, low-birthweight infants, premature rupture of fetal membranes, BM feeding start age, onset time of jaundice, blood transfusion history, and BM CMV viral load. The probability of a low-birthweight infant and the BM CMV viral load were significantly higher in the infected group than those in the non-infected group (P < .05). No significant differences were observed in small-for-gestational-age infants, premature rupture of the fetal membranes, BM feeding start age, onset time of jaundice, and blood transfusion history between the two groups (P > .05).  Table 2).

| Diagnostic performance of the different test methods for BM-acquired CMV infection
As shown in Table 3, the diagnostic performance for BM-acquired CMV infection was achieved best by the urine CMV-DNA test with the maximum AUC, better by the PBMCs CMV-DNA test. The diagnostic value of plasma CMV-DNA and serum CMV-IgM was similar, which showed high specificity but lack of sensitivity. This study demonstrated that birthweight, DBIL, AST, and viral load in BM increased significantly in the infected group, and the differences between the infected group and the non-infected group were significant (P < .05). A single-factor analysis showed a higher risk for BM-acquired CMV infection in jaundiced infants with a lower birthweight and a higher BM CMV viral load, which is basically the same as the results of Martins-Celini. 21 Gunkel et al revealed that the severity of CMV infection is associated with extremely preterm babies (gestational age <26 weeks). 22