Pasteurization of Breastmilk Decreases the Rate of Postnatally Acquired Cytomegalovirus Infections,But Shows a Nonsignificant Trend to an Increased Rate of Necrotizing Enterocolitis in Very Preterm Infants

Abstract

Aim:

This study assessed whether feeding preterm infants unpasteurized breastmilk (1) decreases the rate of late-onset sepsis and necrotizing enterocolitis and (2) increases the rate of postnatally acquired cytomegalovirus infections.

Subjects and Methods:

Between January 2008 and July 2013, preterm infants below 32 completed weeks of gestational age admitted to the neonatal intensive care unit of Innsbruck Medical University (Innsbruck, Austria) (n=344) were eligible for the study. Of those, 323 fed breastmilk were retrospectively enrolled in the study. Two groups were formed, with 164 infants being fed unpasteurized and 159 infants being fed pasteurized breastmilk.

Results:

There was no significant difference in the rate of late-onset sepsis or necrotizing enterocolitis between the unpasteurized and pasteurized breastmilk groups (late-onset sepsis, 15.9% versus 15.1% [p=0.486]; necrotizing enterocolitis, 2.4% versus 4.4% [p=0.254]). The number of infants diagnosed with postnatally acquired cytomegalovirus infection was significantly higher in the unpasteurized group (39.3%) compared with the pasteurized group (4.2%) (p=0.008).

Conclusions:

Feeding preterm infants unpasteurized breastmilk increases the rate of postnatally acquired cytomegalovirus infections. However, we also demonstrate a nonsignificant trend to a decreased rate of necrotizing enterocolitis in the unpasteurized group, which needs to be confirmed in larger studies.

Background

Breastmilk is considered ideal nutrition for preterm infants, but in seropositive mothers it is also the main source of postnatally acquired cytomegalovirus (CMV) infection. Pasteurization of breastmilk prevents transmission of CMV and in October 2003 was recommended by the Austrian Society of Paediatrics to be applied to every milk sample from seropositive mothers being fed to preterm infants up to 34 weeks of gestational age. For now there has been no international consensus available on how to handle breastmilk fed to preterm infants.¹ In the last decade much emphasis has been put on the observation of short-term and long-term morbidity in preterm infants caused by postnatally acquired CMV infections.² Currently, there is growing evidence that most infected preterm infants either remain asymptomatic within the neonatal period or have only mild symptoms, which are self-limited and without sequelae.^3–6 Sensineural hearing loss has not been detected in children after postnatally acquired CMV infection.^7,8 Regarding long-term outcome, studies demonstrated that cognitive and motor function among affected infants are within the normal range, even though cognitive function might be slightly reduced compared with uninfected controls.^8,9

Pasteurization not only eliminates CMV, but also reduces bioactive and immunological components of the breastmilk like, for example, growth factors,¹⁰ lysozyme, and immunoglobulins,^11,12 which also depends on the method of pasteurization used.¹⁰ Pasteurization also has a detrimental impact on antibacterial properties of human milk.¹³ Bearing in mind the loss of potentially protective agents in the breastmilk, pasteurization might be the reason for increased rates of necrotizing enterocolitis (NEC) and late-onset sepsis (LOS) in preterm infants fed pasteurized donor milk compared with those fed exclusively unpasteurized breastmilk.¹⁴ NEC is a life-threatening disease with a high case-fatality rate¹⁵ and a significantly adverse effect on long-term neurodevelopmental outcome.¹⁶ Similarly, LOS is a major problem in neonatology with high mortality and morbidity rates.^17,18 The question arises whether concerns about postnatal transmission of CMV justify pasteurizing breastmilk and thereby withholding immunological and bioactive components of the breastmilk, which might play an important role in the prevention of NEC and LOS.

The aim of this study was to assess whether feeding preterm infants pasteurized breastmilk (1) decreases the rate of LOS and NEC and (2) increases the rate of postnatally acquired CMV infections.

Subjects and Methods

Study cohort

The study survey area was Tyrol, a state in western Austria with 680,000 inhabitants and about 7,000 live births per year. All infants (n=368) born between January 2008 and July 2013 with a gestational age below 32 completed weeks and admitted to the neonatal intensive care unit of Innsbruck Medical University (Innsbruck, Austria), the only neonatal intensive care unit in the geographical region, were retrospectively enrolled in the study. Infants were excluded from the study if they died within the first 2 days of life (n=3; two infants died from severe lung hypoplasia and one from trisomy 18) or were not born at Innsbruck Medical University but were secondarily admitted to its neonatal intensive care unit (n=24). In total, 341 infants were eligible for the study. Of them, 18 infants were identified to be fed exclusively formula and were post hoc–excluded from further analysis owing to the limited number of infants in this study group. The remaining 323 infants were fed breastmilk. Two groups were formed based on the routine processing of breastmilk in very preterm infants at our ward, which was pasteurization until October 2010 (study period I; January 2008–October 2010) and unpasteurized human milk after that (study period II; October 2010–July 2013). During the two time periods no further changes regarding storage of human breastmilk occurred. Consequently, 164 infants (study period II) were fed unpasteurized and 159 infants (study period I) pasteurized breastmilk.

Maternal and neonatal data

Maternal and neonatal data included mode of delivery, Apgar score, gestational age (full weeks of gestation), birth weight (g), sex, diagnosis of early-onset sepsis and LOS, NEC, intracerebral hemorrhage, periventricular leukomalacia, CMV infection, need for red blood cell concentrate (receiving at least one erythrocyte concentrate because of anemia during hospital stay), death, and days until full enteral feeding. Infants were discharged home when they demonstrated cardiorespiratory stability and when feeding patterns were established (without need for gavage feeding). Length of stay and age at discharge were recorded for all survivors (duration of hospital stay [in days]). NEC was defined according to the criteria of Bell et al.¹⁹ Intracerebral hemorrhage was classified according to the method of Papile et al.²⁰ Ultrasound examinations were routinely performed on the second and fifth day of life, thereafter every week, and finally every second or third week. A diagnosis of early-onset sepsis (<72 hours of birth) or LOS (>72 hours of birth) required signs of generalized infection, increased C-reactive protein level, and antibiotic therapy for 5 or more days. As CMV reactivation during lactation was shown to correlate well with immunoglobulin G seropositivity,²¹ seropositive mothers were regarded as being potentially infectious.⁵ CMV infection was regarded as postnatally acquired via breastmilk if the first urine sample of the infant taken in the first 5 days of life was CMV-negative by polymerase chain reaction and later (assessed after week 2 of life) became positive during the hospital stay.

For pasteurization of the breastmilk, the Holder method was applied using the LABU-Muttermilchpasteur^® (Labu Buchrucker GmbH, Ottensheim, Austria).

During the two time periods evaluated there was no change in medical care practice in the neonatal intensive care unit, especially including the preparation and practice of packed red blood cell transfusions and diagnosis of CMV infection. All procedures followed were in accordance with the ethical standards with the Helsinki Declaration of 1975, as revised in 1983. We state that no investigative review needs to be obtained for a retrospective data analysis in Tyrol, Austria.

Statistical analysis

Statistical analyses were performed with SPSS for Windows software (version 21.0; SPSS Inc., Chicago, IL). For group comparisons of continuous data, Student's t tests were performed; when the data were skewed, Mann–Whitney tests were applied. For comparison of categorical data Fisher's exact test was used. Data are presented as numbers with percentages, medians with interquartile range, or means with standard deviations or 95% confidence intervals. Results were deemed statistically significant when two-sided p was <0.05.

Results

Study subjects

In total, 341 infants were eligible for the study; 323 infants, of whom 164 were fed unpasteurized and 159 were fed pasteurized breastmilk, were further evaluated. Maternal and neonatal characteristics were similar for both groups, except for the time to achieve full enteral feeding, which was longer in the pasteurized group (p=0.016) (Table 1).

Table 1.

Maternal and Neonatal Characteristics of the Study Cohort, Divided into Two Groups Based on Pasteurization or Not of Breastmilk

Variable	Pasteurized (n=159)	Unpasteurized (n=164)	p value
Cesarean section (n)	151 (95.0%)	147 (89.6%)	0.056^a
Apgar score
1 minute	6 (5; 7)	6 (5; 7.5)	0.689^b
5 minutes	8 (7; 9)	8 (7; 9)	0.549^b
10 minutes	9 (8; 9)	9 (8; 9)	0.731^b
Gestational age (weeks)	29.5 (27.7; 30.7)	30.0 (28.2; 31.2)	0.081^b
Birth weight (g)	1,226.82±382.01	1,271.28±411.68	0.315^c
Male	87 (54.7%)	84 (51.2%)	0.302^a
ICH	25 (15.7%)	32 (19.5%)	0.228^a
Grade 3–4	4 (2.5%)	9 (5.5%)	0.141^a
PVL	3 (1.9%)	3 (1.8%)	0.643^a
Early-onset sepsis	18 (11.3%)	12 (7.3%)	0.147^a
Red blood cell concentrate	66 (41.5%)	81 (49.4%)	0.095^a
Died	3 (1.9%)	4 (2.4%)	0.517^a
Full enteral feeding (days)	16.54±14.57	13.47±5.95	0.016^c
Duration of hospital stay (days)	59.92±27.92	58.90±27.65	0.740^c

Data are n (%), mean±standard deviation, or median (25^th; 75^th percentile) values as indicated.

By Fisher's exact test.

By Mann–Whitney test.

By Student's t test.

ICH, intracerebral hemorrhage; PVL, periventricular leukomalacia.

Rate of LOS in the pasteurized and nonpasteurized groups

The overall rate of LOS was 15.4% in the study cohort (52 out of 341). There was no significant difference in the rate of LOS observed between the two study periods (15.2% versus 15.3%; p=1.000). The number of infants with LOS did not differ significantly between the two groups: 15.1% among those receiving pasteurized human milk and 15.9% among those receiving unpasteurized human milk (p=0.486).

Rate of NEC in the pasteurized and nonpasteurized groups

The overall rate of NEC was 3.2% (11 out of 341) in the study cohort. There was no significant difference in the rate of NEC observed between the two study periods (4.8% versus 1.7%; p=0.129). The rate of NEC did not differ significantly between the pasteurized (4.4%) and the unpasteurized (2.4%) groups (p=0.254).

Frequency of postnatally acquired CMV infections

In total, five of the 323 infants (1.5%) were diagnosed with congenital CMV infection. In only 16.0% (52 out of 323) of the infants were two CMV samples of the urine available. Of these, 52 infants 12 (23.1%) were diagnosed with postnatally acquired CMV infection. None of these infants showed clinical symptoms of severe sepsis-like syndrome. The number of infants diagnosed with postnatally acquired CMV infection was significantly higher in the unpasteurized group (11 out of 28) than in the pasteurized group (one out of 24) (p=0.008). We detected no statistically significant difference between the number of infants in need of red blood cell transfusion and the rate of postnatally acquired CMV infection (p=0.324).

Discussion

Years ago several studies postulated a negative impact on short- and long-term morbidity in very preterm infants caused by postnatally acquired CMV infection.^22,23 In Austria pasteurization of breastmilk has therefore been recommended by the Austrian Society of Paediatrics to be applied to every breastmilk sample being fed to preterm infants up to 34 gestational weeks from CMV-seropositive mothers (www.docs4you.at). Currently, there is growing evidence that most infants suffering from postnatally acquired CMV infection either remain asymptomatic or develop only mild symptoms, which are self-limited and without sequelae.^3,4,6,24

Pasteurization most effectively eliminates viral infectivity but concomitantly reduces significant immunological and bioactive components of the breastmilk.^10,12,25 This might be of major importance because feeding breastmilk has been shown to protect preterm infants from developing NEC and LOS.^15,26–29 NEC is a life-threatening disease with high fatality rates, putting survivors at risk for long-term sequelae.³⁰ Pike et al.³¹ demonstrated that NEC is not only associated with persistent bowel dysfunction, but also constitutes a risk for functional impairment, including motor, sensory, and cognitive outcomes in middle childhood. Similarly, LOS is independently associated with neurodevelopmental impairment in early childhood.^32,33 A recent study by Mitha et al.³⁴ showed that very preterm infants presenting with LOS had a higher risk for cerebral palsy at the age of 5 years than did uninfected infants, especially in cases of combined early-onset sepsis and LOS. Schanler et al.¹⁴ showed that preterm infants fed exclusively breastmilk had fewer episodes of LOS and/or NEC than did infants who were fed with pasteurized donor milk and preterm formula. We were able to confirm these results in our study, showing that the rate of NEC was slightly higher in the pasteurized than in the unpasteurized group. However, because the rate of NEC was overall low in our study cohort, namely, only 3.4%, this difference did not reach statistical significance and needs to be further investigated in larger trials. One should keep in mind that the beneficial effect of not pasteurizing breastmilk might be even more evident if evaluated in centers showing higher rates of NEC than we do. In contrast to their study, we were not able to show a significant difference in the rate of LOS between the groups.

From our results and previous studies published on this issue, we think that feeding unpasteurized breastmilk is protective and should therefore not be withheld from preterm infants. However, the question remains whether the positive effects of unpasteurized breastmilk outweigh the concerns about postnatally acquired CMV infection. The present study demonstrates that feeding preterm infants unpasteurized breastmilk significantly increases the rate of postnatally acquired CMV infections (4.2% versus 39.3%; p=0.008). However, infants with postnatally acquired CMV infection did not suffer from LOS more often than did uninfected infants (data not shown). In our study cohort we did not observe systemic inflammatory response syndrome, which led us to the clinical suspicion of postnatally acquired CMV infection.

This finding supports previous studies showing that postnatally acquired CMV infections produce mild symptoms and are not a reason for concern during the neonatal period. In a systematic review Kurath et al.³⁵ found that symptomatic postnatal CMV infections occur at a median rate of 3.7% in preterm infants, and episodes of severe sepsis-like syndrome associated with postnatally acquired CMV were even less frequent (0.7%). So far, few studies have been published showing postnatally acquired CMV infection as an independent risk factor for adverse neurodevelopmental outcome in preterm infants. There is growing evidence that sensineural hearing loss in infants with postnatally acquired CMV infection is uncommon.^7,8 Furthermore, motor function does not seem to be affected by postnatally acquired CMV infection. Just recently a study was published proposing a negative effect of postnatally acquired CMV infection on cognitive outcome in early childhood.⁹ However, as mentioned by the authors,⁹ that study needs to be interpreted with caution owing to the small number of infants included in the study and potential confounders in the matched-control study design.

The limitations of our study are that we could not investigate a longitudinal CMV status for all infants born in the study period, resulting in a rather small sample size of infants diagnosed with postnatally acquired CMV infection. Furthermore, it has been published that the effect of pasteurization on antimicrobial properties of human milk are different, depending on the method used (high temperature/short time method versus Holder pasteurization),^10,36 which we did not investigate in this retrospective study. Thus, based on the concerns in previous studies and our awareness of the limitations of the present study, we plan to prospectively reevaluate our analysis in a large cohort of very preterm infants and then also focus on a potentially adverse effect on neurodevelopmental outcome.

Conclusions

Pasteurization of breastmilk is a common practice at neonatal intensive care units worldwide, but guidelines for feeding of preterm infants and applied procedures differ immensely. Although feeding preterm infants unpasteurized breastmilk increases the rate of postnatally acquired CMV infections, the risk of serious complications or long-term sequelae seems to be rather low. The potential protective effects of unpasteurized breastmilk, especially against NEC and LOS (and associated complications and long-term sequelae), seem to outweigh the risks from acquired CMV infection in preterm infants. The observation in our study of a nonsignificant trend to decreased rates of NEC in preterm infants fed unpasteurized breastmilk is suggestive for this conclusion, but larger studies are needed to clarify the true risk–benefit ratio of providing unpasteurized breastmilk to preterm infants.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Omarsdottir

, Casper

, Akerman

, et al. Breastmilk handling routines for preterm infants in Sweden: A national cross-sectional study. Breastfeed Med, 2008; 3:165–170.

Omarsdottir

, Casper

, Zweygberg Wirgart

, et al. Transmission of cytomegalovirus to extremely preterm infants through breast milk. Acta Paediatr, 2007; 96:492–494.

Capretti

, Lanari

, Lazzarotto

, et al. Very low birth weight infants born to cytomegalovirus-seropositive mothers fed with their mother's milk: A prospective study. J Pediatr, 2009; 154:842–848.

Mussi-Pinhata

, Yamamoto

, do Carmo Rego

, et al. Perinatal or early-postnatal cytomegalovirus infection in preterm infants under 34 weeks gestation born to CMV-seropositive mothers within a high-seroprevalence population. J Pediatr, 2004; 145:685–688.

Miron

, Brosilow

, Felszer

, et al. Incidence and clinical manifestations of breast milk-acquired cytomegalovirus infection in low birth weight infants. J Perinatol, 2005; 25:299–303.

Nijman

, de Vries

, Koopman-Esseboom

, et al. Postnatally acquired cytomegalovirus infection in preterm infants: a prospective study on risk factors and cranial ultrasound findings. Arch Dis Child Fetal Neonatal Ed, 2012; 97:F259–F263.

Nijman

, van Zanten

, de Waard

, et al. Hearing in preterm infants with postnatally acquired cytomegalovirus infection. Pediatr Infect Dis J, 2012; 31:1082–1084.

Bevot

, Hamprecht

, Krageloh-Mann

, et al. Long-term outcome in preterm children with human cytomegalovirus infection transmitted via breast milk. Acta Paediatr, 2012; 101:e167–e172.

Goelz

, Meisner

, Bevot

, Hamprecht

, et al. Long-term cognitive and neurological outcome of preterm infants with postnatally acquired CMV infection through breast milk. Arch Dis Child Fetal Neonatal Ed, 2013; 98:F430–F433.

10.

Goelz

, Hihn

, Hamprecht

, et al. Effects of different CMV-heat-inactivation-methods on growth factors in human breast milk. Pediatr Res, 2009; 65:458–461.

11.

Koenig

, de Albuquerque Diniz

, Barbosa

, et al. Immunologic factors in human milk: The effects of gestational age and pasteurization. J Hum Lact, 2005; 21:439–443.

12.

Untalan

, Keeney

, Palkowetz

, et al. Heat susceptibility of interleukin-10 and other cytokines in donor human milk. Breastfeed Med, 2009; 4:137–144.

13.

Van Gysel

, Cossey

, Fieuws

, et al. Impact of pasteurization on the antibacterial properties of human milk. Eur J Pediatr, 2012; 171:1231–1237.

14.

Schanler

, Lau

, Hurst

, et al. Randomized trial of donor human milk versus preterm formula as substitutes for mothers' own milk in the feeding of extremely premature infants. Pediatrics, 2005; 116:400–406.

15.

Shah

, Meinzen-Derr

, Gratton

, et al. Hospital and neurodevelopmental outcomes of extremely low-birth-weight infants with necrotizing enterocolitis and spontaneous intestinal perforation. J Perinatol, 2012; 32:552–558.

16.

Rees

, Pierro

, Eaton

. Neurodevelopmental outcomes of neonates with medically and surgically treated necrotizing enterocolitis. Arch Dis Child Fetal Neonatal Ed, 2007; 92:F193–F198.

17.

Tsai

, Hsu

, Chu

, et al. Incidence, clinical characteristics and risk factors for adverse outcome in neonates with late-onset sepsis. Pediatr Infect Dis J, 2014; 33:e7–e13.

18.

Hofer

, Kothari

, Morris

, et al. The fetal inflammatory response syndrome is a risk factor for morbidity in preterm neonates. Am J Obstet Gynecol, 2013; 209:542.e1–542.e11.

19.

Bell

, Ternberg

, Feigin

, et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg, 1978; 187:1–7.

20.

Papile

, Burstein

, et al. Incidence and evolution of subependymal and intraventricular hemorrhage: A study of infants with birth weights less than 1,500 gm. J Pediatr, 1978; 92:529–534.

21.

Hamprecht

, Maschmann

, Vochem

, et al. Epidemiology of transmission of cytomegalovirus from mother to preterm infant by breastfeeding. Lancet, 2001; 357:513–518.

22.

Lawrence

. Cytomegalovirus in human breast milk: Risk to the premature infant. Breastfeed Med, 2006; 1:99–107.

23.

Paryani

, Yeager

, Hosford-Dunn

, et al. Sequelae of acquired cytomegalovirus infection in premature and sick term infants. J Pediatr, 1985; 107:451–456.

24.

Doctor

, Friedman

, Dunn

, et al. Cytomegalovirus transmission to extremely low-birthweight infants through breast milk. Acta Paediatr, 2005; 94:53–58.

25.

Hamprecht

, Maschmann

, Muller

, et al. Cytomegalovirus (CMV) inactivation in breast milk: Reassessment of pasteurization and freeze-thawing. Pediatr Res, 2004; 56:529–535.

26.

Underwood

. Human milk for the premature infant. Pediatr Clin North Am, 2013; 60:189–207.

27.

Ronnestad

, Abrahamsen

, Medbo

, et al. Late-onset septicemia in a Norwegian national cohort of extremely premature infants receiving very early full human milk feeding. Pediatrics, 2005; 115:e269–e276.

28.

Menon

, Williams

. Human milk for preterm infants: Why, what, when and how?. Arch Dis Child Fetal Neonatal Ed, 2013; 98:F559–F562.

29.

Lawrence

. The risks of not breastfeeding: New associations. Breastfeed Med, 2014; 9:237–238.

30.

Shah

, Doyle

, Anderson

, et al. Adverse neurodevelopment in preterm infants with postnatal sepsis or necrotizing enterocolitis is mediated by white matter abnormalities on magnetic resonance imaging at term. J Pediatr, 2008; 153:170–175, 175.e1.

31.

Pike

, Brocklehurst

, Jones

, et al. Outcomes at 7 years for babies who developed neonatal necrotising enterocolitis: The ORACLE Children Study. Arch Dis Child Fetal Neonatal Ed, 2012; 97:F318–F322.

32.

Schlapbach

, Aebischer

, Adams

, et al. Impact of sepsis on neurodevelopmental outcome in a Swiss National Cohort of extremely premature infants. Pediatrics, 2011; 128:e348–e357.

33.

Stoll

, Hansen

, Adams-Chapman

, et al. Neurodevelopmental and growth impairment among extremely low-birth-weight infants with neonatal infection. JAMA, 2004; 292:2357–2365.

34.

Mitha

, Foix-L'Helias

, Arnaud

, et al. Neonatal infection and 5-year neurodevelopmental outcome of very preterm infants. Pediatrics, 2013; 132:e372–e380.

35.

Kurath

, Halwachs-Baumann

, Muller

, et al. Transmission of cytomegalovirus via breast milk to the prematurely born infant: A systematic review. Clin Microbiol Infect, 2010; 16:1172–1178.

36.

Chantry

, Wiedeman

, Buehring

, et al. Effect of flash-heat treatment on antimicrobial activity of breastmilk. Breastfeed Med, 2011; 6:111–116.

Pasteurization of Breastmilk Decreases the Rate of Postnatally Acquired Cytomegalovirus Infections,But Shows a Nonsignificant Trend to an Increased Rate of Necrotizing Enterocolitis in Very Preterm Infants—A Preliminary Study

Abstract

Abstract

Aim:

Subjects and Methods:

Results:

Conclusions:

Background

Subjects and Methods

Study cohort

Maternal and neonatal data

Statistical analysis

Results

Study subjects

Rate of LOS in the pasteurized and nonpasteurized groups

Rate of NEC in the pasteurized and nonpasteurized groups

Frequency of postnatally acquired CMV infections

Discussion

Conclusions

Footnotes

Disclosure Statement

References