Clinical Neurophysiology xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies Ze D. Jiang ⇑,1, Cui Wang 1, Cao Chen Division of Neonatology, Children’s Hospital, Fudan University, Shanghai 200032, China
a r t i c l e
i n f o
Article history: Accepted 15 March 2014 Available online xxxx Keywords: Brainstem impairment Evoked potentials Neurodevelopment impairment Necrotizing enterocolitis Preterm babies
h i g h l i g h t s The effect of neonatal necrotizing enterocolitis (NEC) on the immature auditory brainstem remains
not fully understood. In maximum length sequence brainstem auditory evoked response, preterm NEC babies showed a
significant increase in wave V latency, I–V and particularly III–V intervals, with relatively normal wave I and III latencies and I–III interval. Neonatal NEC adversely affects myelination of the more rostral or central regions of the immature brainstem, resulting in delayed or impaired neural conduction in auditory brainstem.
a b s t r a c t Objective: Necrotizing enterocolitis (NEC) is a devastating neonatal disease, often leading to long term neurodevelopmental impairment. The effect of NEC on the immature brain remains not fully understood. We test the hypothesis that NEC adversely affects functional integrity, particularly neural conduction, of the preterm brainstem. Methods: Thirty-two preterm NEC babies (30–36 weeks gestation) were recruited at term age. Maximum length sequence brainstem auditory evoked response was recorded and analysed with click rates 91–910/ s at term age. The results were compared with normal term babies and age-matched healthy preterm babies. Results: Wave V latency, I–V and III–V intervals, and III–V/I–III interval ratio differed significantly among the three groups of babies at all click rates 91–910/s. Compared with normal term babies, preterm NEC babies showed significant increase in all these MLS BAER variables at all rates, with no apparent abnormalities in wave I and III latencies and I–III interval. All these abnormalities were more significant at higher than at lower click rates. No notable abnormalities were seen in wave amplitudes. Compared with age-matched healthy preterm babies, NEC babies showed similar abnormalities, although the abnormalities were relatively less significant. Conclusions: MLS BAER components that mainly reflect neural conduction in the more central regions of the auditory brainstem were abnormal in preterm NEC babies, although those components that mainly reflect peripheral function were generally normal. Significance: Neonatal NEC adversely affects myelination of the more rostral or central regions of the immature brainstem, resulting in delayed or impaired neural conduction, but spares the more peripheral regions. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
⇑ Corresponding author. Address: Neonatal Unit, Department of Paediatrics, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, United Kingdom. Tel.: +44 1865 580172. E-mail address: [email protected] (Z.D. Jiang). 1 These authors contribute equally to the research.
1. Introduction As the most common and devastating neonatal gastrointestinal emergency, necrotizing enterocolitis (NEC) has been reported to be a major neonatal risk for long term neurodevelopmental
http://dx.doi.org/10.1016/j.clinph.2014.03.015 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
2
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
impairment (Hintz et al., 2005; Martin et al., 2010; Neu and Walker, 2011; Rees et al., 2007; Schulzke et al., 2007; Shah et al., 2008,2012; Stoll et al., 2004; Ta et al., 2011). Survivors of preterm NEC babies have a greater incidence of poor neurodevelopmental outcome than babies born with prematurity alone. On the other hand, very little is known about short term neurological outcome in babies after NEC. Delineation of the functional status of the immature brain shortly after the occurrence of NEC and understanding of the short-term effect of NEC on the immature brain is of great importance for developing neuroprotective strategy targeting at particular neuropathology, improving neurodevelopmental outcome of NEC babies. Whether neonatal NEC affects functional integrity of the immature auditory brainstem remains to be determined. It is well known that development of the immature auditory brainstem highly depends on axonal development and myelination of the brainstem auditory pathway (Jiang, 2010; Moore, 1987; Moore et al., 1995, 1997; Ponton et al., 1996). The immature brainstem and auditory system are susceptible to various perinatal unfavourable factors or conditions (Wilkinson and Jiang, 2006). Babies who suffer NEC are often associated multiple unfavourable factors or conditions, including prematurity, low birth weight, hypoxia– ischemia, early nutrition deprivation or deficiencies, systematic infection episode or inflammation, etc. (Tarnow-Mordi et al., 2011; Volpe, 2008). These associated factors may impair the white matter and myelination of the immature brain, affecting later neurodevelopment (Volpe, 2008; Dammann and Leviton, 2004; Ehrenkranz et al., 2011; Poindexter et al., 2006). Some magnetic resonance imaging (MRI) studies showed that at term date preterm NEC babies is uniformly associated with white matter injury, and that adverse neurodevelopment in preterm NEC babies is mediated by white matter abnormality (Shah et al., 2008; Van de Bor et al., 1990; Volpe, 2008). It is conceivable that myelination of the immature auditory brainstem would also be adversely affected in preterm NEC babies. We, therefore, hypothesize that the functional integrity, particularly neural conduction that is closely related to myelination, of the neonatal auditory brainstem is adversely affected by NEC and the associated unfavourable factors. We first conducted a pilot study of general auditory function in a group of preterm NEC babies using the brainstem auditory evoked response (BAER) (Jiang et al., 2012a). The data were compared with those of normal term babies for any abnormalities in NEC babies. No major abnormality in peripheral hearing was found. However, there was a slight increase in the I–V interval that suggests a possible auditory abnormality in neonatal NEC, which was never described in the literature. This interesting finding provided a major impetus for us to conduct a full study of functional integrity of the auditory brainstem in a new group of preterm babies who suffered NEC during the preterm period. The functional integrity was examined using a relatively novel method—the maximum length sequence brainstem auditory evoked response (MLS BAER), which has several major advantages, compared with conventional BAER, i.e. the BAER processed using conventional averaging techniques (Eysholdt and Schreiner, 1982; Jiang, 2012; Lasky, 1997; Picton et al., 1992). The MLS technique enables us to gain more insights into the functional properties of the brainstem and the auditory pathway, and have a more thorough description of the electrophysiological behaviour of the auditory brainstem and its development (Jiang, 2012; Jirsa, 2001; Lasky, 1997). During the last decade, a number of studies have documented that the MLS BAER is a valid method to improve the detection of neuropathology that involves the immature auditory brainstem (Jiang, 2012; Jiang et al., 2000, 2003, 2005, 2007, 2009a,b, 2010, 2012b,c; Wilkinson et al., 2007). Neonatal NEC occurs particularly frequently in babies who are born prematurely or preterm with less birth weight (Hintz et al.,
2005; Martin et al., 2010; Rees et al., 2007; Schulzke et al., 2007; Shah et al., 2012; Stoll et al., 2004; Tarnow-Mordi et al., 2011). Since preterm birth per se could exert certain effect on the immature auditory brainstem, there is a need to exclude this possible confounding effect so that the effect of NEC on the BAER can be assessed more reliably and accurately. Thus, in this reported study the MLS BAER data of preterm NEC babies were not only compared with those of normal term babies to detect any abnormalities, but also compared with those of age-matched healthy or low-risk preterm babies who had no NEC. Any differences identified between NEC and healthy preterm babies should be attributed to the adverse effect of NEC on the immature brainstem.
2. Subjects and methods 2.1. Subjects The babies studied were recruited from the neonatal unit of the Children’s Hospital of Fudan University between October 2011 and February 2013. Informed consent was obtained from the parents of each baby before study entry. The medical chart of each baby was reviewed by one of the authors. At time of MLS BAER study (37–42 weeks postconceptional age—PCA), all babies were in a relatively stable clinical condition. Excluded from the study entry were those babies who had congenital malformation or chromosomal anomalies, congenital or perinatal infection of the central nervous system, and syndromes that can affect the auditory function (Down’s syndrome, Cornelia de Lange syndrome, and Ondine’s syndrome). Except for NEC in the study group, none of our babies had any major perinatal problems or complications that may directly or indirectly affect the central nervous system, particularly the brainstem auditory pathway, to minimize any confounding effects. These mainly included grades III and IV intraventricular haemorrhage, periventricular leukomalacia documented on head ultrasound, hypoxia–ischemia, severe hyperbilirubinemia that required blood transfusion and bronchopulmonary dysplasia (Jiang, 2012; Wilkinson and Jiang, 2006). Similar to our previous MLS BAER studies, any babies who had a BAER threshold P40 dB normal hearing level (nHL), suggesting peripheral hearing loss, were also excluded to minimize any significant effect of peripheral hearing impairment on the measurements of MLS BAER components (Jiang et al., 2003, 2005, 2010, 2012b,c; Wilkinson et al., 2007). The threshold in all subjects was determined with conventional BAER at 21/s click.
Table 1 Subjects’ data in preterm NEC babies, healthy preterm (HP) controls, and normal term (NT) controls.
Gender (n) Male/female
NEC
Healthy preterm
Normal term
16/16
18/16
17/21
Gestational age (weeks) Mean ± SD 32.8 ± 1.4 Range 30–36
32.5 ± 1.5 30–36
39.3 ± 1.3** 37–42
Postconceptional age (weeks) Mean ± SD 39.5 ± 1.4 Range 37–41
39.8 ± 1.4 37–42
39.6 ± 1.7 37–42
Birth weight (g) Mean ± SD Range
2019 ± 404** 1,460–3320
3414 ± 458** 2515–4425
9.2 ± 3.2 0–20
9.7 ± 3.8 0–20
1611 ± 361 1165–2620
BAER threshold (dB nHL) Mean ± SD 10.1 ± 4.1 Range 5–30 **
P < 0.01 (t-test) for comparison between NEC and control babies.
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
The study group was comprised of 32 preterm babies who suffered NEC during the preterm period, with 5 born as small-for-gestational-age (birth weight below 5th centile). They were all clinically, radiologically or histologically diagnosed as NEC with at least stage 2a (presence of pneumatosis intestinalis on abdominal X-ray films) of the modified Bell’s classification of NEC (Bell et al., 1978; Volpe, 2008). The stage of NEC of each NEC baby was verified by one of the authors. Medical management included antibiotic usage, withdrawal of feeds and resuscitative measures. Surgical intervention for NEC was at the discretion of pediatric surgeon. Eleven of the NEC babies received surgical intervention (laparotomy, peritoneal drainage or both). The healthy preterm control group was comprised of 34 preterm babies who did not have any of notable perinatal problems. The normal term control group was comprised of 37 healthy babies who were born at term and without any perinatal problems. At the time of testing, monaural BAER thresholds were all 0.05). This was also the case of normal term and healthy preterm controls. Thus, analysis of variance could be used for comparison of MLS BAER data between different groups. However, in order to minimize any possible insignificant differences, analysis of covariance (ANCOVA) was performed for comparison of MLS BAER data between different groups of babies. All MLS BAER variables, for each stimulus condition, were used as dependent variables in separate ANCOVAs. The potential confounders (PCA, gender and the click intensity above BAER threshold, which could affect MLS BAER measurements) were used as covariates to minimize any confounding effects, arising from possible differences in in these potential confounders between different groups of babies. The Holm–Bonferroni method was used both for the original and additional corrections (alpha level 0.05) to judge statistical significance for all comparisons of each dependent variable between different groups. 3. Results There were no significant differences between the preterm NEC babies and the healthy preterm babies in gender, gestational age, PCA, and BAER threshold, except for birth weight that was lower in the NEC babies than in the healthy babies (p < 0.01, Table 1). There were also no significant differences between the preterm NEC babies and the normal term babies in gender, PCA, and BAER threshold, but gestational age was smaller and birth weight that was lower in the NEC babies than in the term babies (p < 0.01, and 0.01, Table 1). Ten (31%) NEC babies had culture proved sepsis. The hearing level (i.e. the dB above the thresholds of individual subjects) was 51.7 ± 4.7 dB nHL for the preterm NEC group, which did not differ significantly from for the healthy preterm babies (50.3 ± 4.8 dB nHL), and the normal term babies (49.7 ± 5.5 dB nHL). No missing MLS BAER waves were found in the healthy preterm and normal term controls. In the preterm NEC babies, major MLS BAER components (waves I, III and V) were all reliably identified at all click rates, except that wave I was less clear at 910/sec in 2 babies which was then excluded from data analysis. 3.1. Comparison among preterm NEC babies, normal term and healthy preterm controls There were no statistically significant differences among the three groups of babies in wave I and III latencies and the I–III interval at any click rates. However, wave V latency differed significantly among the three groups at all repetition rates of clicks
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
4
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
(p < 0.01 at 91/s and 227/s, and p < 0.001 at 455/s and 910/s). The I–V and III–V interpeak intervals also differed significantly among the three groups at all click rates (all p < 0.001). Similarly, the III–V/I–III interval ratio differed among these groups at all rates (p < 0.01 at 91/s, and p < 0.001 at 227/s, 455/s and 910/s). These differences were generally more significant at higher repetition rates than at low rates. The amplitudes of waves V did not differ significantly among the three groups of babies at almost all click rates, except at 227/ s at which the amplitude differed significantly (p < 0.01). Neither wave I amplitude nor wave III amplitude differed significantly among the three groups at any click rates. The V/I and V/III amplitude ratios differed significantly among the three groups at 91/s (p < 0.05) and 227/s (p < 0.01), respectively. 3.2. Comparison of preterm NEC babies with normal term babies The preterm NEC babies showed a significant increase in wave V latency at all click rates 91–910/s (p < 0.01 at 91/s and 227/s, and p < 0.001 at 455/s and 910/s, Fig. 1). Both the I–V and the III–V intervals in the preterm NEC babies were significantly increased (p < 0.01 at 91/s, and p < 0.001 at 227/s, 455/s and 910/s, Fig. 2). This was also the case of the III–V/I–III interval ratio (p < 0.01 at 91/s, and p < 0.001 at 227/s, 455/s and 910/s). However, wave I and III latencies and the I–III interval were similar to those in the term babies at all rates, without any statistically significant difference (Figs. 1 and 2). No marked differences were found in wave amplitudes between the preterm NEC babies and the normal term babies at most click rates. Wave I amplitude was similar in the two groups at all 91–910/s (Fig. 3). Wave III and V amplitudes in the preterm NEC babies were slightly, though statistically significantly, reduced at 227–910/s (all p < 0.05, Fig. 3), respectively. No difference was found between the two groups of babies in the V/III amplitude ratio at any rates. The V/III amplitude ratio in the preterm NEC babies was decreased at 227/s and 910/s (p < 0.01 and 0.05). 3.3. Comparison of preterm NEC babies with healthy preterm babies Wave V latency in the preterm NEC babies were significantly increased at higher click rates (p < 0.05 at 91 and 227/s, and p < 0.001 at 455/s and 910/s, Fig. 1). The I–V intervals were increased more significantly at higher than at lower click rates (p < 0.05 at 91/s, p < 0.01 at 227/s, and p < 0.001 at 455/s and 910/s, Fig. 2). The same was true of the III–V interval (p < 0.01 at 91/s and 227/s, and p < 0.001 at 455/s and 910/s, Fig. 2). The III–V/I–III interval ratio in the preterm NEC babies was also increased, particularly at very high rates (p < 0.05 at 91/s and 227/s, and p < 0.001 at 455/s and 910/s). On the other hand, wave I and III latencies and the I–III interval did not show any significant differences from the healthy babies (Figs. 1 and 2). Compared with those in the healthy preterm babies, the amplitudes of MLS BAER wave components in the NEC babies tended to be reduced at higher click rates (Fig. 3). No significant difference was found in wave I amplitudes between the two groups of babies. However, both wave III and V amplitudes in the NEC babies were reduced at higher click rates 227–910/s (all p < 0.05, Fig. 3). The V/I amplitude ratio was decreased at 91/s and 455/s (p < 0.01 and 0.05). No significant difference was seen in the V/III amplitude ratio between the two groups at any rates. 4. Discussion In the pilot study, we compared general auditory function in preterm NEC babies with a gestation between 30 and 34 weeks
and normal term babies (Jiang et al., 2012a). The auditory function was assessed using conventional BAER—a widely used objective method to assess brainstem auditory function and particularly suitable for very young subjects (Eggermont and Salamy, 1988; Hall III, 2007; Lasky, 1997; Lasky et al., 2012; Moore, 1987; Volpe, 2001; Wilkinson and Jiang, 2006). The BAER was recorded at 21/sec and 60 dB nHL to examine general auditory function. The present study was carried out using the MLS BAER in a new group of preterm NEC babies with a gestational age between 30 and 36 weeks. The MLS BAER was recorded with clicks at much higher repetition rates (91–910/sec) and an intensity level slightly higher than 40 dB above the threshold of each subject for more reliable and accurate assessment of central auditory function. The data of preterm NEC babies were analysed in much more detail, and compared with those of normal term babies for any abnormalities due to NEC. These data were further compared with those of age-matched healthy or low-risk preterm babies with no NEC to excluding any confounding effect due to preterm birth. Our results indicate that there is delayed or impaired myelination in the more central or rostral regions of the immature auditory brainstem, with a relatively normal peripheral regions. This novel finding has never been reported before, and provided some new insight into the mechanisms of neurological impairment after neonatal NEC. 4.1. Impaired neural conduction in the immature auditory brainstem after NEC Like conventional BAER, the MLS BAER reflects electrophysiological activity of a large number of sequentially activated neurons at successively higher levels of the brainstem auditory pathway following acoustic stimulation (Eysholdt and Schreiner, 1982; Jiang, 2012; Lasky, 1997; Picton et al., 1992). Wave latencies and, in particular, interpeak intervals in the MLS BAER are primarily related to nerve conduction velocity associated with axonal diameter, myelination and synaptic function, and synchronization along the brainstem auditory pathway (Jiang, 2012). The present study revealed that at term date the preterm NEC babies had clear abnormalities in MLS BAER, including a significant increase in wave V latency, the I–V and III–V intervals, and the III–V/I–III interval ratio. These abnormalities suggest a significant delay in neural conduction, mainly reflecting delayed or impaired myelination and possible synaptic dysfunction in the auditory brainstem. The high repetition rates of click stimuli in MLS BAER provide a stronger temporal/physiological challenge to auditory neurons, and permit a more exhaustive sampling of physiological recovery or ‘‘fatigue’’ than is possible with much slower rates in conventional BAER (Jiang, 2012). This enables MLS BAER to have a potential to detect some, particularly early or subtle, neuropathology that may not be detected by conventional BAER. The changes in MLS BAER with varying click rate reflect neural processes concerning the efficacy of central synaptic transmission, as well as neural synchronisation and metabolic status of auditory neurons in the brainstem (Jiang, 2012; Lasky, 1997). The finding that MLS BAER abnormalities in the preterm NEC babies were more significant at higher than at lower repetition rates of clicks suggests that at high rate there is a significant increase in temporal dispersion due to delayed or impaired myelination, and a decrease in the efficacy of synaptic transmission in the immature brainstem. Preterm birth per se may exert certain effect on the BAER (Henderson-Smart et al., 1991; Jiang et al., 2002; Li et al., 2011; Pasman et al., 1992). Neonatal NEC occurs mainly in babies who are born at preterm. To reliably and accurately delineate the effect of NEC on the auditory brainstem, it is important to exclude any possible confounding effect exerted by preterm birth on our MLS BAER results. Thus, we also compared the MLS BAER data in preterm NEC babies with those in age-matched healthy preterm
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
5
Fig. 1. Boxplot (bold line across the box, median; box, 25th and 75th centile; extensions, the largest and smallest values) of wave I, III and V latencies at different click rates in MLS BAER, recorded at term date, in preterm NEC babies, healthy preterm (HP) controls, and normal term (NT) controls. ##p < 0.01, and ###p < 0.001 for comparison between NEC babies and NT babies; ⁄p < 0.05, and ⁄⁄⁄p < 0.001 for comparison between NEC babies and HP babies.
babies without NEC. The comparison revealed some significant differences, which were similar to, though slightly less significant than, those shown in the comparison between the preterm NEC babies and the normal term controls. Clearly, after excluding any possible effect of preterm birth on our MLS BAER results, our preterm babies after NEC were still associated with delayed or
impaired neural conduction along the brainstem. This further confirms that neonatal NEC adversely affects neural conduction of the immature brainstem. On the other hand, MLS BAER wave amplitudes and amplitude ratios in the preterm NEC babies did not show any marked abnormalities when compared with the normal term controls, though
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
6
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
Fig. 2. Boxplot (bold line across the box, median; box, 25th and 75th centile; extensions, the largest and smallest values) of I–V, I–III and III–V intervals at different click rates in MLS BAER, recorded at term date, in preterm NEC babies, healthy preterm (HP) controls, and normal term (NT) controls (see text for p values of statistical significance between groups). ##p < 0.01, and ###p < 0.001 for comparison between NEC babies and NT babies; ⁄p < 0.05, ⁄⁄p < 0.01, and ⁄⁄⁄p < 0.001 for comparison between NEC babies and HP babies.
were slightly smaller when compared with the healthy preterm controls at higher click rates. The relatively normal wave amplitudes are in sharp contrast to the significantly increased wave V latency and I–V and III–V intervals. It appears that neonatal NEC does not have any significant effect on the neural origin of wave amplitudes, although it has a significant adverse effect on the
neural origin of wave V latency and I–V and III–V intervals. In BAER, wave latencies and intervals are more related to neural conduction and myelination of the auditory brainstem, while wave amplitudes are more related to neuronal function (Jiang, 2012; Jiang and Wilkinson, 2012). The differential findings between MLS BAER wave latencies and intervals and wave amplitudes in
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
7
Fig. 3. Boxplot (bold line across the box, median; box, 25th and 75th centile; extensions, the largest and smallest values) of wave I, III and V amplitudes at different click rates in MLS BAER, recorded at term date, in preterm NEC babies, healthy preterm (HP) controls, and normal term (NT) controls (see text for p values of statistical significance between groups). #p < 0.05 for comparison between NEC babies and NT babies; ⁄p < 0.05 for comparison between NEC babies and HP babies.
our NEC babies suggest that there is delayed or impaired myelination and neural conduction in the immature brainstem but there is no severe neuronal pathology. This also supports the notion that in some neuropathology BAER wave amplitudes may not change in parallel with wave latencies and intervals, and can yield information that is somewhat independent of that from wave latencies and intervals (Jiang and Wilkinson, 2012; Jiang et al., 2008).
4.2. NEC adversely affects the more rostral or central regions of the brainstem but spares the more peripheral regions The significant increase in wave V latency and I–V interval, with a normal wave I latency, in the preterm NEC babies suggests that neural conduction along the central, specifically brainstem, auditor pathway is delayed or impaired after NEC, although neural
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
8
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
conduction along the peripheral auditory pathway is basically normal. Since there were no apparent abnormalities in wave III latency and the I–III interval, the significant increase in wave V latency and the I–V interval must be essentially produced by the significant increase in the III–V interval that reflects neural conduction at the more rostral (or central) regions of the auditory brainstem (Jiang, 2012; Jiang et al., 2009c). These findings indicate that the delayed or impaired neural conduction occurs at the more rostral regions of the brainstem. It appears that neonatal NEC adversely affects myelination and neural conduction in the more rostral or central regions of the brainstem, but does not have any marked effect on the more peripheral regions. 4.3. Possible risk factors for impaired neural conduction after NEC Neonatal NEC is often associated with various unfavourable perinatal conditions or risk factors, including prematurity, low birth weight, hypoxia–ischemia, inflammation and nutritional deficiencies, etc. (Claud, 2009; Dammann and Leviton, 2004; Dror and Allen, 2008; Edwards and Tan, 2006; Ehrenkranz et al., 2011; Neu and Walker, 2011; Poindexter et al., 2006; Stoll et al., 2004; Volpe, 2008). These unfavourable conditions can adversely affect the immature brainstem and auditory pathway, resulting BAER abnormalities (Henderson-Smart et al., 1991; Jiang, 2012; Wilkinson and Jiang, 2006). For instance, our previous MLS BAER studies have well documented that perinatal hypoxia–ischemia results in major MLS BAER abnormalities (Jiang et al., 2000, 2003, 2008, 2010). In the course of NEC, cerebral hypoxia–ischemia is often induced by inflammation reaction or ischemia reperfusion during intestinal injury (Volpe, 2008). This can disturb the development of axon and myelination of the immature brain, including the brainstem and central auditory pathway, resulting in neurophysiological impairment (Jiang, 2010). In the present study, due to the relatively small number of subjects we did not go for a detailed analysis of the associated risk factors that may be responsible for the MLS BAER abnormalities in our NEC babies. However, birth weight was significantly lower in the preterm NEC babies than in the healthy preterm babies. It is well recognized that, compared with those with a normal birth weight, babies born with a low birth weight are more likely to have perinatal problems and neurodevelopmental deficits. For instance, such babies have a higher incidence of hyperbilirubinemia than those born with normal birth weight. Owing to immature metabolic function a smaller increase in serum bilirubin levels than in mature babies can result in kernicterus. Low birth weight babies are very prone to life-threatening infections. They are also disposed to episodes of acidosis or hypoxia that often occurs in NEC babies (Neu and Walker, 2011). Shah et al. (2008) reported that 44% of their very preterm NEC babies had sepsis. In our NEC babies, 31% were associated with sepsis. It is likely that the impaired neural conduction in the brainstem seen in our preterm NEC babies is due to the collective effects of multiple perinatal factors (e.g. low birth weight, hypoxia–ischemia, inflammation and nutritional deficiencies, etc.) that are associated with NEC, although some factors may be more important than the others. Shah studied prospectively 192 unselected preterm infants (gestational age
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies Ze D. Jiang ⇑,1, Cui Wang 1, Cao Chen Division of Neonatology, Children’s Hospital, Fudan University, Shanghai 200032, China
a r t i c l e
i n f o
Article history: Accepted 15 March 2014 Available online xxxx Keywords: Brainstem impairment Evoked potentials Neurodevelopment impairment Necrotizing enterocolitis Preterm babies
h i g h l i g h t s The effect of neonatal necrotizing enterocolitis (NEC) on the immature auditory brainstem remains
not fully understood. In maximum length sequence brainstem auditory evoked response, preterm NEC babies showed a
significant increase in wave V latency, I–V and particularly III–V intervals, with relatively normal wave I and III latencies and I–III interval. Neonatal NEC adversely affects myelination of the more rostral or central regions of the immature brainstem, resulting in delayed or impaired neural conduction in auditory brainstem.
a b s t r a c t Objective: Necrotizing enterocolitis (NEC) is a devastating neonatal disease, often leading to long term neurodevelopmental impairment. The effect of NEC on the immature brain remains not fully understood. We test the hypothesis that NEC adversely affects functional integrity, particularly neural conduction, of the preterm brainstem. Methods: Thirty-two preterm NEC babies (30–36 weeks gestation) were recruited at term age. Maximum length sequence brainstem auditory evoked response was recorded and analysed with click rates 91–910/ s at term age. The results were compared with normal term babies and age-matched healthy preterm babies. Results: Wave V latency, I–V and III–V intervals, and III–V/I–III interval ratio differed significantly among the three groups of babies at all click rates 91–910/s. Compared with normal term babies, preterm NEC babies showed significant increase in all these MLS BAER variables at all rates, with no apparent abnormalities in wave I and III latencies and I–III interval. All these abnormalities were more significant at higher than at lower click rates. No notable abnormalities were seen in wave amplitudes. Compared with age-matched healthy preterm babies, NEC babies showed similar abnormalities, although the abnormalities were relatively less significant. Conclusions: MLS BAER components that mainly reflect neural conduction in the more central regions of the auditory brainstem were abnormal in preterm NEC babies, although those components that mainly reflect peripheral function were generally normal. Significance: Neonatal NEC adversely affects myelination of the more rostral or central regions of the immature brainstem, resulting in delayed or impaired neural conduction, but spares the more peripheral regions. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
⇑ Corresponding author. Address: Neonatal Unit, Department of Paediatrics, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, United Kingdom. Tel.: +44 1865 580172. E-mail address: [email protected] (Z.D. Jiang). 1 These authors contribute equally to the research.
1. Introduction As the most common and devastating neonatal gastrointestinal emergency, necrotizing enterocolitis (NEC) has been reported to be a major neonatal risk for long term neurodevelopmental
http://dx.doi.org/10.1016/j.clinph.2014.03.015 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
2
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
impairment (Hintz et al., 2005; Martin et al., 2010; Neu and Walker, 2011; Rees et al., 2007; Schulzke et al., 2007; Shah et al., 2008,2012; Stoll et al., 2004; Ta et al., 2011). Survivors of preterm NEC babies have a greater incidence of poor neurodevelopmental outcome than babies born with prematurity alone. On the other hand, very little is known about short term neurological outcome in babies after NEC. Delineation of the functional status of the immature brain shortly after the occurrence of NEC and understanding of the short-term effect of NEC on the immature brain is of great importance for developing neuroprotective strategy targeting at particular neuropathology, improving neurodevelopmental outcome of NEC babies. Whether neonatal NEC affects functional integrity of the immature auditory brainstem remains to be determined. It is well known that development of the immature auditory brainstem highly depends on axonal development and myelination of the brainstem auditory pathway (Jiang, 2010; Moore, 1987; Moore et al., 1995, 1997; Ponton et al., 1996). The immature brainstem and auditory system are susceptible to various perinatal unfavourable factors or conditions (Wilkinson and Jiang, 2006). Babies who suffer NEC are often associated multiple unfavourable factors or conditions, including prematurity, low birth weight, hypoxia– ischemia, early nutrition deprivation or deficiencies, systematic infection episode or inflammation, etc. (Tarnow-Mordi et al., 2011; Volpe, 2008). These associated factors may impair the white matter and myelination of the immature brain, affecting later neurodevelopment (Volpe, 2008; Dammann and Leviton, 2004; Ehrenkranz et al., 2011; Poindexter et al., 2006). Some magnetic resonance imaging (MRI) studies showed that at term date preterm NEC babies is uniformly associated with white matter injury, and that adverse neurodevelopment in preterm NEC babies is mediated by white matter abnormality (Shah et al., 2008; Van de Bor et al., 1990; Volpe, 2008). It is conceivable that myelination of the immature auditory brainstem would also be adversely affected in preterm NEC babies. We, therefore, hypothesize that the functional integrity, particularly neural conduction that is closely related to myelination, of the neonatal auditory brainstem is adversely affected by NEC and the associated unfavourable factors. We first conducted a pilot study of general auditory function in a group of preterm NEC babies using the brainstem auditory evoked response (BAER) (Jiang et al., 2012a). The data were compared with those of normal term babies for any abnormalities in NEC babies. No major abnormality in peripheral hearing was found. However, there was a slight increase in the I–V interval that suggests a possible auditory abnormality in neonatal NEC, which was never described in the literature. This interesting finding provided a major impetus for us to conduct a full study of functional integrity of the auditory brainstem in a new group of preterm babies who suffered NEC during the preterm period. The functional integrity was examined using a relatively novel method—the maximum length sequence brainstem auditory evoked response (MLS BAER), which has several major advantages, compared with conventional BAER, i.e. the BAER processed using conventional averaging techniques (Eysholdt and Schreiner, 1982; Jiang, 2012; Lasky, 1997; Picton et al., 1992). The MLS technique enables us to gain more insights into the functional properties of the brainstem and the auditory pathway, and have a more thorough description of the electrophysiological behaviour of the auditory brainstem and its development (Jiang, 2012; Jirsa, 2001; Lasky, 1997). During the last decade, a number of studies have documented that the MLS BAER is a valid method to improve the detection of neuropathology that involves the immature auditory brainstem (Jiang, 2012; Jiang et al., 2000, 2003, 2005, 2007, 2009a,b, 2010, 2012b,c; Wilkinson et al., 2007). Neonatal NEC occurs particularly frequently in babies who are born prematurely or preterm with less birth weight (Hintz et al.,
2005; Martin et al., 2010; Rees et al., 2007; Schulzke et al., 2007; Shah et al., 2012; Stoll et al., 2004; Tarnow-Mordi et al., 2011). Since preterm birth per se could exert certain effect on the immature auditory brainstem, there is a need to exclude this possible confounding effect so that the effect of NEC on the BAER can be assessed more reliably and accurately. Thus, in this reported study the MLS BAER data of preterm NEC babies were not only compared with those of normal term babies to detect any abnormalities, but also compared with those of age-matched healthy or low-risk preterm babies who had no NEC. Any differences identified between NEC and healthy preterm babies should be attributed to the adverse effect of NEC on the immature brainstem.
2. Subjects and methods 2.1. Subjects The babies studied were recruited from the neonatal unit of the Children’s Hospital of Fudan University between October 2011 and February 2013. Informed consent was obtained from the parents of each baby before study entry. The medical chart of each baby was reviewed by one of the authors. At time of MLS BAER study (37–42 weeks postconceptional age—PCA), all babies were in a relatively stable clinical condition. Excluded from the study entry were those babies who had congenital malformation or chromosomal anomalies, congenital or perinatal infection of the central nervous system, and syndromes that can affect the auditory function (Down’s syndrome, Cornelia de Lange syndrome, and Ondine’s syndrome). Except for NEC in the study group, none of our babies had any major perinatal problems or complications that may directly or indirectly affect the central nervous system, particularly the brainstem auditory pathway, to minimize any confounding effects. These mainly included grades III and IV intraventricular haemorrhage, periventricular leukomalacia documented on head ultrasound, hypoxia–ischemia, severe hyperbilirubinemia that required blood transfusion and bronchopulmonary dysplasia (Jiang, 2012; Wilkinson and Jiang, 2006). Similar to our previous MLS BAER studies, any babies who had a BAER threshold P40 dB normal hearing level (nHL), suggesting peripheral hearing loss, were also excluded to minimize any significant effect of peripheral hearing impairment on the measurements of MLS BAER components (Jiang et al., 2003, 2005, 2010, 2012b,c; Wilkinson et al., 2007). The threshold in all subjects was determined with conventional BAER at 21/s click.
Table 1 Subjects’ data in preterm NEC babies, healthy preterm (HP) controls, and normal term (NT) controls.
Gender (n) Male/female
NEC
Healthy preterm
Normal term
16/16
18/16
17/21
Gestational age (weeks) Mean ± SD 32.8 ± 1.4 Range 30–36
32.5 ± 1.5 30–36
39.3 ± 1.3** 37–42
Postconceptional age (weeks) Mean ± SD 39.5 ± 1.4 Range 37–41
39.8 ± 1.4 37–42
39.6 ± 1.7 37–42
Birth weight (g) Mean ± SD Range
2019 ± 404** 1,460–3320
3414 ± 458** 2515–4425
9.2 ± 3.2 0–20
9.7 ± 3.8 0–20
1611 ± 361 1165–2620
BAER threshold (dB nHL) Mean ± SD 10.1 ± 4.1 Range 5–30 **
P < 0.01 (t-test) for comparison between NEC and control babies.
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
The study group was comprised of 32 preterm babies who suffered NEC during the preterm period, with 5 born as small-for-gestational-age (birth weight below 5th centile). They were all clinically, radiologically or histologically diagnosed as NEC with at least stage 2a (presence of pneumatosis intestinalis on abdominal X-ray films) of the modified Bell’s classification of NEC (Bell et al., 1978; Volpe, 2008). The stage of NEC of each NEC baby was verified by one of the authors. Medical management included antibiotic usage, withdrawal of feeds and resuscitative measures. Surgical intervention for NEC was at the discretion of pediatric surgeon. Eleven of the NEC babies received surgical intervention (laparotomy, peritoneal drainage or both). The healthy preterm control group was comprised of 34 preterm babies who did not have any of notable perinatal problems. The normal term control group was comprised of 37 healthy babies who were born at term and without any perinatal problems. At the time of testing, monaural BAER thresholds were all 0.05). This was also the case of normal term and healthy preterm controls. Thus, analysis of variance could be used for comparison of MLS BAER data between different groups. However, in order to minimize any possible insignificant differences, analysis of covariance (ANCOVA) was performed for comparison of MLS BAER data between different groups of babies. All MLS BAER variables, for each stimulus condition, were used as dependent variables in separate ANCOVAs. The potential confounders (PCA, gender and the click intensity above BAER threshold, which could affect MLS BAER measurements) were used as covariates to minimize any confounding effects, arising from possible differences in in these potential confounders between different groups of babies. The Holm–Bonferroni method was used both for the original and additional corrections (alpha level 0.05) to judge statistical significance for all comparisons of each dependent variable between different groups. 3. Results There were no significant differences between the preterm NEC babies and the healthy preterm babies in gender, gestational age, PCA, and BAER threshold, except for birth weight that was lower in the NEC babies than in the healthy babies (p < 0.01, Table 1). There were also no significant differences between the preterm NEC babies and the normal term babies in gender, PCA, and BAER threshold, but gestational age was smaller and birth weight that was lower in the NEC babies than in the term babies (p < 0.01, and 0.01, Table 1). Ten (31%) NEC babies had culture proved sepsis. The hearing level (i.e. the dB above the thresholds of individual subjects) was 51.7 ± 4.7 dB nHL for the preterm NEC group, which did not differ significantly from for the healthy preterm babies (50.3 ± 4.8 dB nHL), and the normal term babies (49.7 ± 5.5 dB nHL). No missing MLS BAER waves were found in the healthy preterm and normal term controls. In the preterm NEC babies, major MLS BAER components (waves I, III and V) were all reliably identified at all click rates, except that wave I was less clear at 910/sec in 2 babies which was then excluded from data analysis. 3.1. Comparison among preterm NEC babies, normal term and healthy preterm controls There were no statistically significant differences among the three groups of babies in wave I and III latencies and the I–III interval at any click rates. However, wave V latency differed significantly among the three groups at all repetition rates of clicks
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
4
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
(p < 0.01 at 91/s and 227/s, and p < 0.001 at 455/s and 910/s). The I–V and III–V interpeak intervals also differed significantly among the three groups at all click rates (all p < 0.001). Similarly, the III–V/I–III interval ratio differed among these groups at all rates (p < 0.01 at 91/s, and p < 0.001 at 227/s, 455/s and 910/s). These differences were generally more significant at higher repetition rates than at low rates. The amplitudes of waves V did not differ significantly among the three groups of babies at almost all click rates, except at 227/ s at which the amplitude differed significantly (p < 0.01). Neither wave I amplitude nor wave III amplitude differed significantly among the three groups at any click rates. The V/I and V/III amplitude ratios differed significantly among the three groups at 91/s (p < 0.05) and 227/s (p < 0.01), respectively. 3.2. Comparison of preterm NEC babies with normal term babies The preterm NEC babies showed a significant increase in wave V latency at all click rates 91–910/s (p < 0.01 at 91/s and 227/s, and p < 0.001 at 455/s and 910/s, Fig. 1). Both the I–V and the III–V intervals in the preterm NEC babies were significantly increased (p < 0.01 at 91/s, and p < 0.001 at 227/s, 455/s and 910/s, Fig. 2). This was also the case of the III–V/I–III interval ratio (p < 0.01 at 91/s, and p < 0.001 at 227/s, 455/s and 910/s). However, wave I and III latencies and the I–III interval were similar to those in the term babies at all rates, without any statistically significant difference (Figs. 1 and 2). No marked differences were found in wave amplitudes between the preterm NEC babies and the normal term babies at most click rates. Wave I amplitude was similar in the two groups at all 91–910/s (Fig. 3). Wave III and V amplitudes in the preterm NEC babies were slightly, though statistically significantly, reduced at 227–910/s (all p < 0.05, Fig. 3), respectively. No difference was found between the two groups of babies in the V/III amplitude ratio at any rates. The V/III amplitude ratio in the preterm NEC babies was decreased at 227/s and 910/s (p < 0.01 and 0.05). 3.3. Comparison of preterm NEC babies with healthy preterm babies Wave V latency in the preterm NEC babies were significantly increased at higher click rates (p < 0.05 at 91 and 227/s, and p < 0.001 at 455/s and 910/s, Fig. 1). The I–V intervals were increased more significantly at higher than at lower click rates (p < 0.05 at 91/s, p < 0.01 at 227/s, and p < 0.001 at 455/s and 910/s, Fig. 2). The same was true of the III–V interval (p < 0.01 at 91/s and 227/s, and p < 0.001 at 455/s and 910/s, Fig. 2). The III–V/I–III interval ratio in the preterm NEC babies was also increased, particularly at very high rates (p < 0.05 at 91/s and 227/s, and p < 0.001 at 455/s and 910/s). On the other hand, wave I and III latencies and the I–III interval did not show any significant differences from the healthy babies (Figs. 1 and 2). Compared with those in the healthy preterm babies, the amplitudes of MLS BAER wave components in the NEC babies tended to be reduced at higher click rates (Fig. 3). No significant difference was found in wave I amplitudes between the two groups of babies. However, both wave III and V amplitudes in the NEC babies were reduced at higher click rates 227–910/s (all p < 0.05, Fig. 3). The V/I amplitude ratio was decreased at 91/s and 455/s (p < 0.01 and 0.05). No significant difference was seen in the V/III amplitude ratio between the two groups at any rates. 4. Discussion In the pilot study, we compared general auditory function in preterm NEC babies with a gestation between 30 and 34 weeks
and normal term babies (Jiang et al., 2012a). The auditory function was assessed using conventional BAER—a widely used objective method to assess brainstem auditory function and particularly suitable for very young subjects (Eggermont and Salamy, 1988; Hall III, 2007; Lasky, 1997; Lasky et al., 2012; Moore, 1987; Volpe, 2001; Wilkinson and Jiang, 2006). The BAER was recorded at 21/sec and 60 dB nHL to examine general auditory function. The present study was carried out using the MLS BAER in a new group of preterm NEC babies with a gestational age between 30 and 36 weeks. The MLS BAER was recorded with clicks at much higher repetition rates (91–910/sec) and an intensity level slightly higher than 40 dB above the threshold of each subject for more reliable and accurate assessment of central auditory function. The data of preterm NEC babies were analysed in much more detail, and compared with those of normal term babies for any abnormalities due to NEC. These data were further compared with those of age-matched healthy or low-risk preterm babies with no NEC to excluding any confounding effect due to preterm birth. Our results indicate that there is delayed or impaired myelination in the more central or rostral regions of the immature auditory brainstem, with a relatively normal peripheral regions. This novel finding has never been reported before, and provided some new insight into the mechanisms of neurological impairment after neonatal NEC. 4.1. Impaired neural conduction in the immature auditory brainstem after NEC Like conventional BAER, the MLS BAER reflects electrophysiological activity of a large number of sequentially activated neurons at successively higher levels of the brainstem auditory pathway following acoustic stimulation (Eysholdt and Schreiner, 1982; Jiang, 2012; Lasky, 1997; Picton et al., 1992). Wave latencies and, in particular, interpeak intervals in the MLS BAER are primarily related to nerve conduction velocity associated with axonal diameter, myelination and synaptic function, and synchronization along the brainstem auditory pathway (Jiang, 2012). The present study revealed that at term date the preterm NEC babies had clear abnormalities in MLS BAER, including a significant increase in wave V latency, the I–V and III–V intervals, and the III–V/I–III interval ratio. These abnormalities suggest a significant delay in neural conduction, mainly reflecting delayed or impaired myelination and possible synaptic dysfunction in the auditory brainstem. The high repetition rates of click stimuli in MLS BAER provide a stronger temporal/physiological challenge to auditory neurons, and permit a more exhaustive sampling of physiological recovery or ‘‘fatigue’’ than is possible with much slower rates in conventional BAER (Jiang, 2012). This enables MLS BAER to have a potential to detect some, particularly early or subtle, neuropathology that may not be detected by conventional BAER. The changes in MLS BAER with varying click rate reflect neural processes concerning the efficacy of central synaptic transmission, as well as neural synchronisation and metabolic status of auditory neurons in the brainstem (Jiang, 2012; Lasky, 1997). The finding that MLS BAER abnormalities in the preterm NEC babies were more significant at higher than at lower repetition rates of clicks suggests that at high rate there is a significant increase in temporal dispersion due to delayed or impaired myelination, and a decrease in the efficacy of synaptic transmission in the immature brainstem. Preterm birth per se may exert certain effect on the BAER (Henderson-Smart et al., 1991; Jiang et al., 2002; Li et al., 2011; Pasman et al., 1992). Neonatal NEC occurs mainly in babies who are born at preterm. To reliably and accurately delineate the effect of NEC on the auditory brainstem, it is important to exclude any possible confounding effect exerted by preterm birth on our MLS BAER results. Thus, we also compared the MLS BAER data in preterm NEC babies with those in age-matched healthy preterm
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
5
Fig. 1. Boxplot (bold line across the box, median; box, 25th and 75th centile; extensions, the largest and smallest values) of wave I, III and V latencies at different click rates in MLS BAER, recorded at term date, in preterm NEC babies, healthy preterm (HP) controls, and normal term (NT) controls. ##p < 0.01, and ###p < 0.001 for comparison between NEC babies and NT babies; ⁄p < 0.05, and ⁄⁄⁄p < 0.001 for comparison between NEC babies and HP babies.
babies without NEC. The comparison revealed some significant differences, which were similar to, though slightly less significant than, those shown in the comparison between the preterm NEC babies and the normal term controls. Clearly, after excluding any possible effect of preterm birth on our MLS BAER results, our preterm babies after NEC were still associated with delayed or
impaired neural conduction along the brainstem. This further confirms that neonatal NEC adversely affects neural conduction of the immature brainstem. On the other hand, MLS BAER wave amplitudes and amplitude ratios in the preterm NEC babies did not show any marked abnormalities when compared with the normal term controls, though
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
6
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
Fig. 2. Boxplot (bold line across the box, median; box, 25th and 75th centile; extensions, the largest and smallest values) of I–V, I–III and III–V intervals at different click rates in MLS BAER, recorded at term date, in preterm NEC babies, healthy preterm (HP) controls, and normal term (NT) controls (see text for p values of statistical significance between groups). ##p < 0.01, and ###p < 0.001 for comparison between NEC babies and NT babies; ⁄p < 0.05, ⁄⁄p < 0.01, and ⁄⁄⁄p < 0.001 for comparison between NEC babies and HP babies.
were slightly smaller when compared with the healthy preterm controls at higher click rates. The relatively normal wave amplitudes are in sharp contrast to the significantly increased wave V latency and I–V and III–V intervals. It appears that neonatal NEC does not have any significant effect on the neural origin of wave amplitudes, although it has a significant adverse effect on the
neural origin of wave V latency and I–V and III–V intervals. In BAER, wave latencies and intervals are more related to neural conduction and myelination of the auditory brainstem, while wave amplitudes are more related to neuronal function (Jiang, 2012; Jiang and Wilkinson, 2012). The differential findings between MLS BAER wave latencies and intervals and wave amplitudes in
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
7
Fig. 3. Boxplot (bold line across the box, median; box, 25th and 75th centile; extensions, the largest and smallest values) of wave I, III and V amplitudes at different click rates in MLS BAER, recorded at term date, in preterm NEC babies, healthy preterm (HP) controls, and normal term (NT) controls (see text for p values of statistical significance between groups). #p < 0.05 for comparison between NEC babies and NT babies; ⁄p < 0.05 for comparison between NEC babies and HP babies.
our NEC babies suggest that there is delayed or impaired myelination and neural conduction in the immature brainstem but there is no severe neuronal pathology. This also supports the notion that in some neuropathology BAER wave amplitudes may not change in parallel with wave latencies and intervals, and can yield information that is somewhat independent of that from wave latencies and intervals (Jiang and Wilkinson, 2012; Jiang et al., 2008).
4.2. NEC adversely affects the more rostral or central regions of the brainstem but spares the more peripheral regions The significant increase in wave V latency and I–V interval, with a normal wave I latency, in the preterm NEC babies suggests that neural conduction along the central, specifically brainstem, auditor pathway is delayed or impaired after NEC, although neural
Please cite this article in press as: Jiang ZD et al. Neonatal necrotizing enterocolitis adversely affects neural conduction of the rostral brainstem in preterm babies. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.03.015
8
Z.D. Jiang et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
conduction along the peripheral auditory pathway is basically normal. Since there were no apparent abnormalities in wave III latency and the I–III interval, the significant increase in wave V latency and the I–V interval must be essentially produced by the significant increase in the III–V interval that reflects neural conduction at the more rostral (or central) regions of the auditory brainstem (Jiang, 2012; Jiang et al., 2009c). These findings indicate that the delayed or impaired neural conduction occurs at the more rostral regions of the brainstem. It appears that neonatal NEC adversely affects myelination and neural conduction in the more rostral or central regions of the brainstem, but does not have any marked effect on the more peripheral regions. 4.3. Possible risk factors for impaired neural conduction after NEC Neonatal NEC is often associated with various unfavourable perinatal conditions or risk factors, including prematurity, low birth weight, hypoxia–ischemia, inflammation and nutritional deficiencies, etc. (Claud, 2009; Dammann and Leviton, 2004; Dror and Allen, 2008; Edwards and Tan, 2006; Ehrenkranz et al., 2011; Neu and Walker, 2011; Poindexter et al., 2006; Stoll et al., 2004; Volpe, 2008). These unfavourable conditions can adversely affect the immature brainstem and auditory pathway, resulting BAER abnormalities (Henderson-Smart et al., 1991; Jiang, 2012; Wilkinson and Jiang, 2006). For instance, our previous MLS BAER studies have well documented that perinatal hypoxia–ischemia results in major MLS BAER abnormalities (Jiang et al., 2000, 2003, 2008, 2010). In the course of NEC, cerebral hypoxia–ischemia is often induced by inflammation reaction or ischemia reperfusion during intestinal injury (Volpe, 2008). This can disturb the development of axon and myelination of the immature brain, including the brainstem and central auditory pathway, resulting in neurophysiological impairment (Jiang, 2010). In the present study, due to the relatively small number of subjects we did not go for a detailed analysis of the associated risk factors that may be responsible for the MLS BAER abnormalities in our NEC babies. However, birth weight was significantly lower in the preterm NEC babies than in the healthy preterm babies. It is well recognized that, compared with those with a normal birth weight, babies born with a low birth weight are more likely to have perinatal problems and neurodevelopmental deficits. For instance, such babies have a higher incidence of hyperbilirubinemia than those born with normal birth weight. Owing to immature metabolic function a smaller increase in serum bilirubin levels than in mature babies can result in kernicterus. Low birth weight babies are very prone to life-threatening infections. They are also disposed to episodes of acidosis or hypoxia that often occurs in NEC babies (Neu and Walker, 2011). Shah et al. (2008) reported that 44% of their very preterm NEC babies had sepsis. In our NEC babies, 31% were associated with sepsis. It is likely that the impaired neural conduction in the brainstem seen in our preterm NEC babies is due to the collective effects of multiple perinatal factors (e.g. low birth weight, hypoxia–ischemia, inflammation and nutritional deficiencies, etc.) that are associated with NEC, although some factors may be more important than the others. Shah studied prospectively 192 unselected preterm infants (gestational age
