Original Paper Received: April 25, 2013 Accepted after revision: August 23, 2013 Published online: November 15, 2013

Neonatology 2014;105:27–32 DOI: 10.1159/000355306

High Rate of Symptomatic Cytomegalovirus Infection in Extremely Low Gestational Age Preterm Infants of 22–24 Weeks’ Gestation after Transmission via Breast Milk Katrin Mehler a André Oberthuer a Ruth Lang-Roth b Angela Kribs a  

 

 

Department of Neonatology, Children’s Hospital, and b Department of Paediatric Audiology, University of Cologne, Cologne, Germany  

 

Key Words Human cytomegalovirus · HCMV infection · ELBW preterm infants · Breast milk

Abstract Background: Very immature preterm infants are at risk of developing symptomatic or severe infection if cytomegalovirus is transmitted via breast milk. It is still a matter of debate whether human cytomegalovirus (HCMV) infection may lead to long-term sequelae. Objectives: We hypothesized that symptomatic and severe HCMV infection transmitted via breast milk affects extremely immature infants at a very high rate. Methods: In 2012, untreated breast milk was fed to extremely low birth weight infants after parental informed consent was obtained. We retrospectively analyzed data on HCMV infection of infants born in 2012 between 22 and 24 weeks of gestation. Results: 17 infants were born to HCMV IgG-seropositive mothers. 11 (65%) of these were diagnosed with symptomatic infection. In all cases, thrombocytopenia was the reason to analyze the infant’s urine. HCMV infection was diagnosed at a median time of 12 weeks after birth. In 5 (45%) infants, thrombocytopenia was the only symptom and resolved without antiviral therapy or platelet transfusion. 6 (55%) infants developed sepsis-like disease with mildly elevated CRP values and showed signs of respiratory failure. 3 (27%) were able to be stabilized on CPAP, 3 (27%) had to be

© 2013 S. Karger AG, Basel 1661–7800/14/1051–0027$39.50/0 E-Mail [email protected] www.karger.com/neo

intubated and mechanically ventilated. 4 children were treated with ganciclovir and/or valganciclovir. 55% failed otoacoustic emissions and/or automated auditory brainstem response testing at discharge. Conclusions: In very immature infants born at the border of viability and suffering from multiple preexisting problems, HCMV infection may trigger a severe deterioration of the clinical course. © 2013 S. Karger AG, Basel

Introduction

Human breast milk has been known for decades to be a potential source of human cytomegalovirus (HCMV) infection in preterm infants. Between 4 and 6 weeks postpartum, HCMV can be isolated from breast milk in up to 100% of HCMV IgG-positive mothers [1, 2]. Very immature preterm infants are at the greatest risk of developing symptomatic HCMV infection via breast milk [3–6]. Previous reports on HCMV infection in preterm infants via breast milk have suggested that infected infants did not have any long-term sequelae [7, 8] but new data from 2012 and 2013 showed a negative influence of HCMV infection on cognitive development [9, 10]. In most cases, HCMV-infected infants were not clinically ill [11] with severe illness occurring at a median rate of only 0.7% [11]. Katrin Mehler, MD Department of Neonatology University of Cologne, Children’s Hospital Kerpenerstrasse 62, DE–50937 Cologne (Germany) E-Mail katrin.mehler @ uk-koeln.de

Downloaded by: UCSF Library & CKM 169.230.243.252 - 3/24/2015 6:43:23 PM

a

 

Patients and Methods Infants born in 2012 between 22 and 24 weeks of gestation who survived until discharge were included in the study and analyzed retrospectively. Maternal HCMV IgM and IgG seroprevalence is routinely tested in our hospital in the event of preterm delivery and was done in all mothers before birth. In HCMV IgG-positive mothers an informed consent form explaining the advantages of feeding untreated breast milk and the risks of HCMV transmission and infection was signed by all mothers within 1 week after birth. Subsequently, all infants of mothers who agreed to feed untreated milk were fed breast milk that was neither frozen nor pasteurized. Mothers brought their milk from home in cooling bags or pumped at the ward and stored the milk in a refrigerator on the ward. Nurses checked whether the amount stored in the refrigerator was sufficient for the infants’ needs. Surplus milk was frozen until needed at –20 ° C, lacking amounts were thawed. Mothers were encouraged to start breastfeeding if infants were clinically stable and 28– 30 weeks of gestational age. Gestational age, birth weight and complications of prematurity (bronchopulmonary dysplasia at 36 weeks, ileostomy, intraventricular hemorrhage grade 3 or 4, hydrocephalus with shunt, retinopathy requiring surgery) were analyzed in infants of HCMV IgG-positive mothers. According to local policies, infants’ urines were qualitatively analyzed for HCMV by PCR if symptoms suspicious of HCMV infection were present. If urine PCR showed positive results, quantitative analysis of viral copies from blood was performed. In case of newly recognized thrombocytopenia, urine analysis for HCMV copies was part of the routine diagnostic workup. We did not analyze breast milk for presence of HCMV. If hydrocephalus with shunt was present, ce 

28

 

Neonatology 2014;105:27–32 DOI: 10.1159/000355306

rebral fluid was analyzed for HCMV. At the time of HCMV infection clinical symptoms such as thrombocytopenia, sepsis like syndrome, respiratory failure, hepatitis and colitis were recorded. At discharge, hearing was assessed with the transient-evoked otoacoustic emissions (TEOAE) test and automated auditory brainstem response (AABR) audiometry at 35 dB. Infants who did not pass the test were referred to the pedaudiology department for further testing. The study protocol was approved by the local ethics committee.

Results

In 2012, 22 mothers gave birth to 26 infants at 22–24 weeks of gestation who survived until discharge. 10 infants born at 22–24 weeks of gestation died. 7 died within 48 h and never received breast milk. 3 infants, who died on days 3, 9 and 10 had received breast milk but were born to HCMV IgM- and IgG-negative mothers. No mother was tested HCMV IgM-positive. 16 (73%) mothers of surviving infants were tested HCMV IgG-positive and gave birth to 17 infants included in our study. While all HCMV-negative mothers were of German origin, HCMV-seropositive mothers came from either Germany (n = 4), Turkey (n = 5), Russia (n = 2), Africa (n = 2), Morocco (n = 1) or Greece (n = 2). 11/17 (65%) infants born to a HCMV-positive mother were diagnosed with symptomatic infection. In all cases, thrombocytopenia (platelet count 20 mg/l occurred in 2 infants with sepsis-like syndrome. 1 infant had hepatitis with elevated GOT and Mehler/Oberthuer/Lang-Roth/Kribs

Downloaded by: UCSF Library & CKM 169.230.243.252 - 3/24/2015 6:43:23 PM

The discrepancy of high HCMV transmission rates and only very few reported infants with severe clinical illness is probably the main reason why there is no consensus on how breast milk of HCMV-positive mothers should be pretreated before it is fed to preterm infants. Based on the good outcome of HCMV-infected preterm infants reported so far [7, 8], we decided to refrain from routine holder pasteurization in 2012. This is in line with international recommendation issued for example by the American Academy of Pediatrics [12]. Before 2012, all mothers were tested for HCMV and breast milk was (Holder) pasteurized from day 10 if they were IgG-positive. We did not pasteurize colostrum and stopped pasteurization if infants were >32 weeks of gestation. After stopping general pasteurizing in 2012, we observed a high incidence of symptomatic (65%) and more importantly, severe (35%) HCMV infection in very immature infants born at 22–24 weeks of gestation. We assumed that the rates of symptomatic and severe infection of infants born to HCMV-positive mothers are significantly higher in extremely immature preterm infants than previously reported [11]. In this report, we retrospectively summarized these cases of HCMV infection of extremely low gestational age preterm infants.

Table 1. Characteristics of infants of HCMV IgG-positive mothers

Gestational age 22 weeks (%) Gestational age 23 weeks (%) Gestational age 24 weeks (%) Birth weight, mean (range) IVH grade 3 or 4 (%) Hydrocephalus with shunt (%) Ileostomy for FIP/NEC (%) ROP with surgery (%) BPD at 36 weeks (%) Weight gain, g/day (mean ± SD) Weight at discharge, g (mean ± SD) Duration of hospital stay, days (mean ± SD) Corrected age at discharge, days (mean ± SD) Head circumference at discharge, cm (mean ± SD) TEOAE/AABR failure at discharge

CMV infection (11)

No CMV infection (6)

p value*

3 (27) 5 (45) 3 (27) 541 (390–713) 2 (18) 2 (18) 2 (18) 0 6 (54) 13 (2) 2,487 (359) 152 (35) 34 (28) 32 (1.4) 6 (54)

1 (16) 5 (83) 0 560 (370–775) 3 (50) 0 3 (50) 0 3 (50) 13 (2) 2,472 (577) 146 (21) 29 (17) 32 (1.8) 1 (17)

0.250 0.749 0.205 0.404 0.205 0.627 0.634 0.949 0.701 0.697 0.884 0.160

* p values calculated using the χ2 test (nominal variables) or t test (continuous variables).

Patient No.

Neutrophils/mm3 (trough)

CRP, mg/l

GOT, U/l

GPT, U/l

Bilirubin, mg/dl (conjugated)

Platelets/mm3 (trough)

1 2 3 4 5 6 7 8 9 10 11

0.95 0.65 1.55 19.4 11.45 12.25 0.8 1.5 1.9 1.3 0.85

12.6 6.6 4.2 15 21.6 10.7 38.6 1 mg/dl) (table 2). In 2 infants with severe disease (patients 4 and 5) and posthemorrhagic hydrocephalus cerebral spinal fluid from a Rickham capsule was analyzed. In patient 5, HCMV PCR from CSF was positive and liquor protein levels were slightly elevated (1.5 mg/dl). Yet, CSF parameters did not differ from samples drawn from the Rickham capsule before HCMV infection occurred. In patient 4, HCMV PCR from liquor was negative. Four children were treated with ganciclovir (10 mg/ kg · day–1, two daily doses) and/or valganciclovir (10 mg/ kg · day–1, two daily doses). Of these, 3 had additional severe symptoms (table 3). 2 infants had severe colitis with

abdominal distension and tenderness, gastric losses, poor feeding tolerance, lactic acidosis and dilated bowel on X-ray investigation. Both also had to be intubated and ventilated for respiratory failure. Infant 4 also developed pneumonitis with positive HCMV PCR in tracheal aspirates. Both patients were treated with ganciclovir and valganciclovir for 3 weeks. HCMV copy count (EDTA-stabilized blood) decreased from 187,000 to 384 copies/ml (infant 4) and from positive (low copy count) to negative (infant 10) within 3 and 2 weeks of treatment, respectively. Infant 5 in whom HCMV PCR from liquor from the Rickham capsule was positive was treated with intravenous ganciclovir for 3 weeks followed by oral valganciclovir

HCMV Infection via Breast Milk in ELBW Infants

Neonatology 2014;105:27–32 DOI: 10.1159/000355306

     

     

29

Downloaded by: UCSF Library & CKM 169.230.243.252 - 3/24/2015 6:43:23 PM

Table 2. Laboratory parameters of infected infants at the time of HCMV diagnosis

Table 3. Signs and symptoms, treatment and follow-up of HCMV-infected infants

Patient Gestational No. age, weeks

Birth weight, g

Age at diagnosis, weeks

Copy count/ml

Sepsis-like syndrome

Other symptoms

(Val)ganciclovir

no no no colitis, pneumonitis hepatitis; CSF-PCR positive for HCMV no no no no colitis no

yes no no yes yes

1 2 3 4 5

22 24 23 23 22

390 640 540 525 490

12 12 9 12 8

14,700 730 182,000 187,000 9,020

no no yes yes yes

6 7 8 9 10 11

22 23 23 24 24 23

430 540 413 713 635 590

8 8 9 12 12 17

101,000 101,400 2,020 positive* positive* 1,600

yes yes no no yes no

no no no no yes no

treatment for another 3 weeks. HCMV copy count in EDTA blood decreased from 9,020 to 721 copies/ml within 3 weeks and HCMV PCR in CSF was negative after 4 weeks of treatment. 1 child born at 22 weeks of gestation was asymptomatic besides thrombocytopenia but had a high HCMV copy count and was treated because of extreme immaturity. HCMV copy count fell from 14,700 to 264 copies/ml within 3 weeks. Importantly, we did not observe any adverse effects in any infant treated with ganciclovir/valganciclovir. Full blood count and liver enzymes were checked once per week and showed normal results. At discharge only 45% of infected infants passed otoacoustic emissions testing and/or automated auditory brainstem response audiometry, and 55% failed. These were referred to the pedaudiology department for further testing and none of these children showed a hearing impairment >20 dB hearing loss. In 1 patient, who was transferred to another hospital near to the parents’ home, no results were accessible. In contrast, all infants, with the exception of 1, who were born to HCMV IgG-positive mothers but who were not infected passed testing at discharge (83%).

Discussion

In the present study we report on high rates of symptomatic HCMV infections (65%) and severe sepsis-like illness (35%) in a cohort of extremely immature infants exposed to HCMV in breast milk. 30

Neonatology 2014;105:27–32 DOI: 10.1159/000355306

Severe illness is at present estimated to occur in 0–13.8% of premature infants [11, 13]. Extremely low birth weight and a gestational age

High rate of symptomatic cytomegalovirus infection in extremely low gestational age preterm infants of 22-24 weeks' gestation after transmission via breast milk.

Very immature preterm infants are at risk of developing symptomatic or severe infection if cytomegalovirus is transmitted via breast milk. It is still...
93KB Sizes 0 Downloads 0 Views