Evaluation of cerebral oxygenation and perfusion in small for gestational age neonates and neurodevelopmental outcome at 24–36 months of age

Study design

A prospective matched case-control study was conducted from April 2013 to April 2016 in the 2^nd NICU of the Aristotle University of Thessaloniki, Greece. The study was approved by the Ethical Committee of the institution and the Aristotle University of Thessaloniki, and a written parental consent was obtained for each neonate prior to the enrolment.

Population

We enrolled neonates of 28–36 weeks’ gestation that were SGA and an equal number of matched AGA controls. For each SGA neonate, an AGA control with similar gestation (±7 days) was enrolled. For the definition of SGA and AGA, we utilised Fenton’s growth charts [9]; SGA were neonates with birth weight <10^th centile for their gestation, while AGA were neonates with a birth weight ≥10^th and ≤90^th centile for gestation. Symmetrical SGA were neonates with both birth weight and head circumference <10^th centile, while asymmetrical SGA were those neonates with birth weight <10^th and head circumference ≥10^th centile for their gestation. The population of SGA neonates consisted of the subgroup of neonates that were IUGR based on the in-utero estimation of fetal weight or abdominal circumference below the 10^th percentile, and the non-IUGR neonates [3].

Neonates in any condition that might potentially influence cerebral oxygenation and perfusion were excluded from the study. Therefore, neonates with congenital central nervous system malformations, congenital cyanotic cardiac disease, hemodynamically significant patent ductus arteriosus, intraventricular haemorrhage (IVH) grade III–IV [10], post-haemorrhagic hydrocephalus, hypotension defined by a mean arterial pressure less than 2 standard deviation (SD) for gestation, severe hypocapnia in mechanical ventilated neonates with a partial pressure of carbon dioxide less than 35 mm Hg, perinatal stress defined by pH <7 and base excess ≤−12 mmol/L, or haemoglobinopathies were excluded. Cranial ultrasound was performed in all neonates in order to rule out IVH III–IV, while haemoglobinopathy screening was performed only upon clinical suspicion.

Data collection

The perinatal and neonatal characteristics of the neonates were recorded including gestational age, gender, birth weight, body length and head circumference, mode of conception, full course of antepartum steroid administration, maternal preeclampsia or gestational diabetes, intrauterine growth velocity, normality of the antenatal Dopplers, mode of delivery, resuscitation required, APGAR score in the first and fifth minute, respiratory distress syndrome, duration of invasive and non-invasive mechanical ventilation, day of enteral feeding initiation, development of bronchopulmonary dysplasia, necrotising enterocolitis, sepsis, cranial ultrasound findings at discharge, and discharge body weight. Sepsis was defined by a positive blood culture; respiratory distress syndrome and bronchopulmonary dysplasia were defined according to standard criteria; and necrotising enterocolitis according to modified Bell criteria [11]. Moreover, standard laboratory tests were performed on every study participant during the first, third and seventh day of life including haemoglobin (Hb), haematocrit (Hct), glucose and acid-base balance analysis. The vital signs were continuously recorded (Drager Siemens SC 9000XL Monitor, Solna, Sweden) including heart rate, mean arterial pressure and pulse oxygen saturation (SpO₂). Mean blood pressure was recorded via an oscillometric technique with an inflatable cuff with 5-min intervals during NIRS measurements.

Transcranial Doppler measurements

We performed a colour transcranial Doppler, utilising the CX50 Philips ultrasound machine (Philips Medical Systems Nederland B.V., Best, The Netherlands), through the temporal window with the S8-3 probe on each neonate in the first 12 postnatal hours, and repeated on the third and seventh day of life. The middle cerebral artery haemodynamics were evaluated, according to the guidelines of the International Society of Ultrasound in Obstetrics and Gynecology [12]. The mean velocity, peak systolic velocity and end diastolic flow were recorded, and the pulsatility index (PI) and resistance index (RI) were computed according to the formula: PI=(peak systolic velocity-end diastolic flow)/mean velocity and RI=(peak systolic velocity-end diastolic flow)/peak systolic velocity, respectively.

Near-infrared spectroscopy monitoring

The NIRO-200NX Cerebral Oximeter Monitor (Hamamatsu Protonics^® KK, Tokyo, Japan) was used to measure tissue oxygenation and perfusion. Two neonatal transducers containing a light-emitting diode and two sensors, the detection and the emission probe at a distance of 4 cm, were attached to each neonate. One was placed on the left and the other on the right frontoparietal forehead and held gently with a gauze bandage. The cerebral tissue oxygenation index (cTOI) and cerebral tissue haemoglobin index (cTHI) were computed using a spatially resolved spectroscopy algorithm. The cTOI is expressed as the percentage of oxygenated haemoglobin in real time and represents the mixed oxygenation of the cerebral compartments, as most (~75%) of the blood in the tissue is in the venous system [13]. The normal values of cTOI that have been proposed for healthy neonates depend on the different commercial devices and are estimated at 70–73% using NIRO-200NX. The cTHI, expressed in an arbitrary unit, provides a surrogate for total tissue Hb content within the cerebral tissue, and could be utilised as an indirect measure of blood flow and vasomotor activity. The cTHI represents the cumulative amount of oxygenated and non-oxygenated tissue Hb and is expressed in μmol. The cTHI reflects the cerebral perfusion and depends on the cardiac output and the cerebral vascular resistances.

Spectroscopic measurements were performed for 40 min each on the first, third and seventh postnatal day in all neonates. NIRS and SpO₂ signals were sampled at 2-s intervals and data were stored as numerical values in a personal computer. cTOI as well as cTHI were expressed as a median of 900 measurements, while cerebral fractional tissue oxygen extraction (cFTOE) was calculated from the formula SpO₂-TOI/SpO₂ [14].

Bayley-III Scale of Infant and Toddler Development

A neurodevelopmental assessment of SGA and AGA neonates that attended the follow-up appointment was performed at 24–36 months of age with the Bayley-III Scale of Infant and Toddler Development (BSID-III). The BSID-III includes the evaluation of five distinct areas of development: cognitive, communication (expressive and receptive language), motor (fine and gross), social/emotional and adaptive. Each scale is estimated with composite and subscale scores, in addition to normal centiles. The normal reference index that has been established for a composite score is 100±15, while for the subscale score it is 10±3. Values below 85 for composite and below 7 for subscale scores are considered as indicating neurodevelopmental impairment [15]. As BSID-III has not been standardized yet in the Greek population, a translated manual, questionnaire and evaluation form was adopted by the American edition, and the cognitive, communication and motor domains were evaluated.

Outcomes

The primary outcome was the evaluation of the cerebral oxygenation and perfusion in SGA and AGA neonates within the first, third and seventh postnatal day, with a subgroup analysis between SGA/IUGR and AGA neonates, and furthermore, the effect that perinatal parameters may have in cerebral oxygenation and perfusion.

The secondary outcome was the evaluation of any association between the cerebral oxygenation parameters or the clinical and laboratory characteristics and the neurodevelopmental outcome between the SGA and AGA neonates, assessed by the BSID-III index scores at 24–36 months of age.

Statistical analysis

The normality of the distributions of continuous variables was assessed by the Kolmogorov-Smirnov test. Continuous variables were expressed as mean±SD or median [interquartile range (IQR)] as appropriate. Comparisons of continuous variables between SGA and AGA neonates, as well as between SGA/IUGR and AGA neonates, on the first, third and seventh postnatal day were performed using Student’s unpaired t-test or the non-parametric Mann-Whitney test. Categorical variables were expressed as n (%) and compared with Pearson’s chi-square (χ²) test or Fisher’s exact test. The effect of perinatal parameters on cerebral oxygenation and perfusion was assessed with Pearson’s coefficient. Multivariate linear regression analysis was utilised for the adjustment of multiple confounders and the cerebral monitoring parameters associated with the BSID-III index scores at 24–36 months of age.

All performed tests were two-sided, and a P-value less than 0.05 was considered statistically significant (alpha 0.05). The data were analysed using SPSS Statistics Version 20.0 (IBM, Chicago, IL, USA). The power of the analysis was estimated using PASS 11.0 (NCSS, LLC. Kaysville, UT, USA), and sample sizes of 38 cases and 38 controls were estimated to achieve a 90% power to detect a difference of −10 on cTOI and 0.1 on cFTOE between the two groups with a significant level (alpha) of 0.05.

Cerebral oxygenation and perfusion and Doppler parameters

Cerebral oxygenation was expressed by the averaged value of cTOI and cFTOE and was equal between SGA and AGA neonates during the first, third and seventh postnatal day (Table 2). In subgroup analysis, male SGA neonates had higher cTOI on the first postnatal day compared to female SGA neonates (73±7% vs. 69±7%, P=0.04). Of note, cerebral perfusion that was expressed by the average cTHI was significantly higher in SGA neonates on the first postnatal day, but this difference was diminished on the subsequent measurements on the third and seventh postnatal days.

With regard to the colour Doppler parameters, no significant differences were recorded in the PI or RI between the two groups of neonates. The longitudinal analysis of cTOI, cTHI, PI and RI within the first seven postnatal days in SGA and AGA neonates is presented in Figure 1. A subgroup analysis was also performed between the SGA/IUGR and the AGA neonates with respect to cerebral oxygenation and perfusion and Doppler parameters, which also revealed a significant higher cTHI in the SGA/IUGR neonates during the first postnatal day (Supplementary Table 1).

Figure 1:

Longitudinal values of cTOI, cTHI, PI and RI within the first 7 days of life in SGA and AGA neonates.

cTOI, Cerebral tissue oxygenation index; cFTOE, cerebral fractional oxygen extraction; cTHI, cerebral tissue haemoglobin index; PI, pulsatile index; RI, resistance index; SGA, small for gestational age; AGA, appropriate for gestational age.

The effect of gestational age, size at birth, IUGR, gender and other examined parameters on cTOI, cFTOE and cTHI is presented in Table 3. Male gender had a significant effect (r=0.210, P=0.04) on the cTOI at birth.

BSID-III index scores

Overall, 46 (48%) neonates attended the follow-up and had a neurodevelopmental assessment at 24–36 months of age (Table 4). Twenty-three SGA infants were evaluated at a mean corrected age of 27.8±2.9 months and 23 AGA infants at a mean corrected age of 27.4±1.8 months. There were no significant differences between the SGA and AGA infants regarding the composite index scores in the cognitive (94±8 compared to 95±16, P=0.86), communication (103±17 compared to 94±20, P=0.13) and motor domains (109±13 compared to 104±18, P=0.24), respectively. None of the participants had developed cerebral palsy. Accordingly, the same findings were demonstrated when a subgroup analysis between SGA/IUGR and AGA neonates was performed (Supplementary Table 2).

Regression analysis adjusting for any other confounding factors revealed that cTOI and cTHI during the first week of life were not significantly associated with the composite index scores. Of note, the length of the mechanical ventilation and the event of late-onset sepsis were significant risk factors affecting the cognitive and the communication composite index scores of the BSID-III assessment, respectively (Table 5).