Evaluation of a Cranial Ultrasound Scoring System for Prediction of Abnormal Early Neurodevelopment in Preterm Infants

INTRODUCTION

Cranial ultrasound (cUS) is widely used for the screening of preterm neonates to detect brain injury due to prematurity.^1,2 Infants found to have brain abnormalities on cUS are followed up with further imaging,³ focused neurodevelopment assessments,⁴ and early intervention provided when indicated.^5,6 Brain injuries that are predictive of early neurodevelopmental impairment in preterm infants include any grade of intraventricular hemorrhage (IVH), cerebral or cerebellar hemorrhage, white matter injury (WMI), ventriculomegaly and hydrocephalus, cystic changes, and signs of brain atrophy after any injury.^7,8

Late screening cUS examinations are performed at various time points in different institutions;⁹ 6-weeks, term–equivalent age (TEA) or discharge may be used as the last screening examinations. The late cUS examination is important to detect any evidence of WMI.

The use of a cUS scoring system (Appendix 1) for the quantification of brain injury in preterm neonates has been reported as a useful tool to consider using in the prediction of neurological outcome.^10,11 The scoring system includes measurements of the lateral ventricle, interhemispheric fissure, thickness of the corpus callosum and subarachnoid space. Several subjective assessments of injury are also included in the score items indicating signs of white matter loss, the presence of any cysts and deep gray matter injury. These assessments were obtained from a TEA cUS performed on the neonates. The cUS score provides a systematic approach to quantifying brain abnormality in a single score and the higher the score the more likely the expected outcome of adverse neurodevelopment. However, the data was limited on the validity of this technique as a reliable method of predicting neurological outcome and we, therefore, studied a group of high-risk neonates to assess its usefulness in predicting adverse early neurodevelopmental outcomes at 3–4-months corrected age (age from the original due date).

Early neurodevelopment assessment provides screening for early features of cerebral palsy (CP) or early developmental delay in infants who are born preterm and require intervention.¹²

The aim of this study was to evaluate and compare the parameters of two methods (cUS scoring vs conventional cUS reporting) to characterize abnormal cUS findings for early adverse neurodevelopmental outcomes—any CP or any adverse neurodevelopment assessed at 3–4-months corrected age.

MATERIALS AND METHODS

The cUS screening is routinely performed on preterm infants, who are born in less than 32 weeks gestational age (GA) or below 1500 gm in our unit. The screening includes a late or last cUS examination when neonates are 6-weeks postnatal age (PNA). Our unit does not perform a TEA cUS as part of the screening protocol. Infants born before 29 weeks GA or with a birth weight of less than 1000 gm were selected for inclusion in this retrospective cohort study as they were the group who had routine inpatient and outpatient neurodevelopmental assessments in our unit. The data available for the study was obtained from the time of the establishment of the neurodevelopment clinic in 2018 until 2022. This includes screening for early features of CP or developmental delay at 3–4-months corrected age.^4,12 Neonates were excluded from the study if they did not have a 6-weeks PNA cUS examination prior to their early neurodevelopment clinic assessment. Ultrasound examinations were performed to routine screening standards of our unit at the time and neurological examinations performed on each neonate in the study in a dedicated early neurodevelopment clinic.^12,13

The cUS score was calculated retrospectively by two investigators (GM and KVH) from the ultrasound examination reports using the cUS scoring system reported previously,^10,11 and both scorers were blinded to the early neurodevelopment clinic outcomes. In this study, a cUS score of 10 or lower was considered normal or test negative, whereas a score of more than or equal to 11 was considered abnormal or test positive.¹⁰ If the scores were different and one scorer considered the cUS examination was normal, while the other scorer scored as abnormal (10 vs ≥11) a consensus was reached after deliberation. Any cUS abnormality was considered positive if there was the presence of any abnormality, such as IVH, periventricular leukomalacia (PVL), cerebral or cerebellar hemorrhage and negative if the cUS examinations demonstrated no abnormality,⁸ Similarly, the severe CUA test was considered positive if there was IVH grade 3 or above present,^14,15 cystic PVL,¹⁶ cerebral (parenchymal) or cerebellar hemorrhage. Otherwise, the test outcome was considered negative.

The cUS examinations included static coronal and sagittal images obtained through the anterior fontanelle and axial plane images were obtained through the mastoid fontanelle.^1,17,18 All examinations were performed by in-house trained and credentialled sonographers and the examinations were reported by consultant pediatric radiologists. No extra projections were required to calculate the cUS scores and it was, therefore, possible to use available images and imaging reports.

The neurodevelopment assessments included a video assessment of general movements [general movements assessment (GMA) at fidgety age],¹⁹ Hammersmith Infant Neurological Examination (HINE),²⁰ and medical examination for the outcomes of early features of CP, high risk of CP, developmental delay, abnormal GMA, and suboptimal HINE.

RESULTS

Two hundred and forty-two preterm infants were included in the study with the median GA being 26.5 weeks (IQR, 4 weeks). The median body weight was 880 grams (IQR, 356.5 gm). Any CUA was reported in 93/242 (38%) infants and 17/93 (18%) had severe CUAs. Furthermore, 85/242 (35%) infants had an abnormal score of more than or equal to 11. The median score was similar between the two scorers (medial 10 vs 10) and there was no disagreement (κ = 1). At the early neurodevelopment clinic assessments, there were 33/242 (14%) infants diagnosed with CP/high risk of CP and 124/242 (51%) had some form of developmental delay.

DISCUSSIONS

We report the practical use of a cUS scoring system¹⁰ to predict early neurodevelopmental outcomes in a cohort of extremely preterm neonates. The cUS score in our study had no difference in sensitivity for the prediction of CP but lower sensitivity for any neurodevelopment delay, as compared to CUA. However, the specificity of using the cUS score was higher for both CP and any delay. Compared to severe CUA, the score had significantly higher sensitivity for predicting CP and any delay but had significantly lower specificity for CP and any delay. Overall, the cUS score performed the same when compared to any CUA but had higher sensitivity and lower specificity when compared to severe CUA when predicting abnormal early neurodevelopment in extremely preterm infants.

The literature reporting the use of cUS scoring is sparse and our study adds to the evidence on this emerging technique. The scoring system is the only cUS scoring system documented. The study by Skiöld et al. carried out on 84 infants found that agreement between the cUS scoring system and MRI scores was good. Sensitivity was the same for cUS and MRI in predicting CP (75%) and severe cognitive delay (100%). Specificity for CP was the same for MRI compared with cUS (97% vs 90%).¹⁰ Our study in comparison had lower sensitivity than theirs for predicting CP (57% vs 75%). The specificity to predict CP in our study was lower than Skiöld et al. (68% vs 90%), although the ultrasound examinations were performed at different time points with our infants examined at 6-weeks PNA while the Skiöld et al. infants were all TEA when the cUS score was calculated. The sample size of the Skiöld was however smaller than our study (84 vs 242).

This cUS score has been reported as having similar sensitivity to MRI brain in a study on infants at TEA.¹⁰ The score includes an expanded list of brain injuries compared with other studies^21,22 and is therefore more in line with MRI scoring systems.^23–25 This includes subtle abnormalities like corpus callosum thinning and delayed folding of the cortex as well as the sequelae of IVH, periventricular hemorrhagic infarction (PVHI), ventriculomegaly and injury to the white matter including any signs of atrophy. Although their study used the cUS score to compare cUS findings at TEA with MRI, it is reasonable to suggest that it could also be of value when comparing cUS examinations. Unlike the MRI and cUS studies on infants at TEA, our study evaluated data from ultrasound examinations performed at 6-weeks PNA, as this is the final screening examination performed at our unit. We acknowledge that further white matter volume loss could have occurred in the infants after the 6-weeks cUS examination and therefore not captured by the study. Further studies to assess later cUS score at TEA should be therefore investigated in a similar group of infants.

A strength of our study was the 100% agreement between the two scorers when deciding if a score was normal or abnormal (abnormal ≥ 11) and our study had a considerable larger cohort compared to the previous Skiöld study. Although our study has a relatively big sample size, compared with the Skiöld study, some CI are very wide and the study would benefit from a larger patient group. As the cUS score was possible to be calculated at a later date than when the examination was performed, it could be calculated by an independent operator in an auditing or research setting.

Another limitation of the study was assessing the cUS score in a dichotomous way to calculate the diagnostic accuracy parameters. The cUS score is a continuous measure with a score of 10 being normal and any score above 10 abnormal—the higher the score the more likely there is brain injury. However, we have used it as a categorical variable as a distinction between normal and abnormal is required to calculate the sensitivity and specificity. Therefore, we also performed an ROC analysis, which shows that the cut-off used in the study agrees with the previous study.¹⁰ The AUC of the ROC curve of 0.5 indicates that the performance of the tests are not significantly different for predicting early abnormal neurodevelopment and is of very limited value.

The late cUS examinations at our institution were performed at 6-weeks PNA, according to our protocol, rather than at TEA as described by Skiöld et al. This meant that the age of the neonates in our study was varied unlike those in the published study. It may indicate that this scoring system is only predictive when used for term-aged neonates. We plan to conduct future studies, which look at late screening time points to evaluate this further.

All infant’s neurodevelopment was assessed early at 3−4-months corrected age. The infants will also be undergoing later neurodevelopment assessments at two years of age. Early diagnosis allows for early intervention and treatment,^5,13 and any adverse neurodevelopment would be confirmed with longer term follow-up. As the study shows the overall cUS scores are quite low and therefore ultrasound poor at predicting abnormal neurodevelopment it is crucial there is follow-up in all early preterm infants. Severe CUA has good specificity but very poor sensitivity and therefore it is important it is not considered as sufficient to use alone.

CONCLUSIONS

The use of a cUS scoring system to predict early neurodevelopmental outcome is similar to any reported CUA for predicting outcomes, however, when compared to severe CUA, it had better sensitivity but poor specificity for predicting abnormal early neurodevelopmental in extremely preterm neonates.

ACKNOWLEDGMENTS

The authors would like to thank Keith Van Haltren for his contribution as second scorer of the cUS images. Ethics approval was obtained for the project from Monash Human Research Ethics Committee (ERM Reference No. 81908).

ORCID

Glenda McLean https://orcid.org/0000-0001-9336-2366

REFERENCES

1. Hwang M, Tierradentro–Garcia LO, Hussaini SH, et al. Ultrasound imaging of preterm brain injury: Fundamentals and updates. Pediatr Radiol 2022;52(4):817–836. DOI: 10.1007/s00247-021-05191-9.

2. Mohammad K, Scott JN, Leijser LM, et al. Consensus approach for standardizing the screening and classification of preterm brain injury diagnosed with cranial ultrasound: A Canadian perspective. Front Pediatr 2021;9:618236. DOI: 10.3389/fped.2021.618236.

3. van Wezel–Meijler G, Steggerda SJ, Leijser LM. Cranial ultrasonography in neonates: Role and limitations. Semin Perinatol 2010;34(1):28–38. DOI: 10.1053/j.semperi.2009.10.002.

4. Novak I, Morgan C, Adde L, et al. Early, accurate diagnosis and early intervention in cerebral palsy. JAMA Pediatrics 2017;171(9):897–907. DOI: 10.1001/jamapediatrics.2017.1689.

5. Hadders–Algra M, Boxum AG, Hielkema T, et al. Effect of early intervention in infants at very high risk of cerebral palsy: A systematic review. Dev Med Child Neurol 2017;59(3):246–258. DOI: 10.1111/dmcn.13331.

6. Byrne R, Noritz G, Maitre NL, Group NCHED. Implementation of early diagnosis and intervention guidelines for cerebral palsy in a high-risk infant follow-up clinic. Pediatr Neurol 2017;76:66–71. DOI: 10.1016/j.pediatrneurol.2017.08.002.

7. Brouwer AJ, Groenendaal F, Benders MJNL, et al. Early and late complications of germinal matrix-intraventricular haemorrhage in the preterm infant: What is new? Neonatology 2014;106(4):296–303. DOI: 10.1159/000365127.

8. Inder TE, de Vries LS, Ferriero DM, et al. Neuroimaging of the preterm brain: Review and recommendations. J Pediatr 2021:276e4–287e4. DOI: 10.1016/j.jpeds.2021.06.014.

9. McLean G, Malhotra A, Lombardo P, et al. Cranial ultrasound screening protocols for very preterm infants. Ultrasound Med Biol 2021;47(7):1645–1656. DOI: 10.1016/j.ultrasmedbio.2021.03.006.

10. Skiöld B, Hallberg B, Vollmer B, et al. A novel scoring system for term–equivalent-age cranial ultrasound in extremely preterm infants. Ultrasound Med Biol 2019;45(3):786–794. DOI: 10.1016/j.ultrasmedbio.2018.11.005.

11. Horsch S, Skiöld B, Hallberg B, et al. Cranial ultrasound and MRI at term age in extremely preterm infants. Arch Dis Child Fetal Neonatal Ed 2010;95(5):F310-4. DOI: 10.1136/adc.2009.161547.

12. King AR, Machipisa C, Finlayson F, et al. Early detection of cerebral palsy in high‐risk infants: Translation of evidence into practice in an Australian hospital. J Paediatr Child Health 2021;57(2):246–250. DOI: 10.1111/jpc.15191.

13. Connors R, Sackett V, Machipisa C, et al. Assessing the utility of neonatal screening assessments in early diagnosis of cerebral palsy in preterm infants. Brain Sci 2022;12(7):847. DOI: 10.3390/brainsci12070847.

14. Papile LA, Burstein J, Burstein R, 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(4):529–534. DOI: 10.1016/s0022-3476(78)80282-0.

15. Volpe JJ. Intraventricular hemorrhage in the premature infant: Current concepts. Part I. Ann Neurol 1989;25(1):3–11. DOI: 10.1002/ana.410250103.

16. de Vries LS, Benders MJ, Groenendaal F. Progress in neonatal neurology with a focus on neuroimaging in the preterm infant. Neuropediatrics 2015;46(4):234–241. DOI: 10.1055/s-0035-1554102.

17. Hand IL, Shellhaas RA, Milla SS, et al. Routine neuroimaging of the preterm brain. Pediatrics 2020;146(5):e2020029082. DOI: 10.1542/peds.2020-029082.

18. Meijler G, Steggerda SJ. Neonatal Cranial Ultrasonography, 3rd edition. Cham: Springer International Publishing; 2019.

19. Einspieler C, Prechtl HF. Prechtl’s assessment of general movements: A diagnostic tool for the functional assessment of the young nervous system. Ment Retard Dev Disabil Res Rev 2005;11(1):61–67. DOI: 10.1002/mrdd.20051.

20. Romeo DM, Cioni M, Scoto M, et al. Neuromotor development in infants with cerebral palsy investigated by the Hammersmith Infant Neurological Examination during the first year of age. Eur J Paediatr Neurol 2008;12(1):24–31. DOI: 10.1016/j.ejpn.2007.05.006.

21. Hintz SR, Barnes PD, Bulas D, et al. Neuroimaging and neurodevelopmental outcome in extremely preterm infants. Pediatrics 2015;135(1):e32–e42. DOI: 10.1542/peds.2014-0898.

22. Edwards AD, Redshaw ME, Kennea N, et al. Effect of MRI on preterm infants and their families: A randomised trial with nested diagnostic and economic evaluation. Arch Dis Child Fetal Neonatal Ed 2018;103(1):F15–F21. DOI: 10.1136/archdischild-2017-313102.

23. Kidokoro H, Neil JJ, Inder TE. New MR imaging assessment tool to define brain abnormalities in very preterm infants at term. Am J Neuroradiol 2013;34(11):2208–2214. DOI: 10.3174/ajnr.A3521.

24. Woodward LJ, Anderson PJ, Austin NC, et al. Neonatal MRI to predict neurodevelopmental outcomes in preterm infants. N Engl J Med 2006;355(7):685–694. DOI: 10.1056/NEJMoa053792.

25. Goeral K, Kasprian G, Huning BM, et al. A novel magnetic resonance imaging-based scoring system to predict outcome in neonates born preterm with intraventricular haemorrhage. Dev Med Child Neurol 2022;64(5):608–617. DOI: 10.1111/dmcn.15116.

APPENDIX 1: CRANIAL ULTRASOUND SCORING SYSTEM.¹