L Ching, JCY Lee, KWK Chan, WH Chan

Abstract

Objective: Congenital cytomegalovirus (cCMV) infection is one of the most common congenital intrauterine infections worldwide, and it is well-known to be associated with sensorineural hearing loss (SNHL) and long-term adverse neurological outcomes in children. A magnetic resonance imaging (MRI) severity score by Lucignani et al is reported to be a good predictor for such outcomes. This study aims to show whether the MRI severity score can predict the neurological outcomes in cCMV infection patients. Method: Total 21 paediatric patients who were diagnosed to have cCMV infection and MRI performed between the year 2015 and 2024 in Queen Elizabeth Hospital were included. Images of their MRI brain were reviewed by two radiologists blinded to clinical data and gave an MRI severity score by Lucignani et al. Results: MRI score equal to or more than 2 is associated with adverse neurologic outcomes, while no association is demonstrated for other clinical parameters such as birth weight and gestational week. Conclusion: MRI severity score is shown to be associated with adverse neurological outcomes. Therefore, early MRI brain is helpful for telling prognosis in cCMV infection patients.

Introduction

Congenital cytomegalovirus (cCMV) infection is one of the most common congenital intrauterine infections worldwide, with an estimated incidence of approximately one in 200 infants.¹ It is well-known to be associated with sensorineural hearing loss (SNHL) and long-term adverse neurological outcomes in children.² Some of the patients will present with symptoms at birth, including low-birth weight, microcephaly or hepatosplenomegaly. For these symptomatic patients, antiviral is recommended in treating the infection and preventing the long-term sequelae.^3-7

Magnetic Resonance Imaging (MRI) brain is currently readily available and can assess intracranial abnormalities that are commonly associated with cCMV infection. Different scoring systems have been used to predict long-term outcomes of patients.^8-10 An MRI severity score by Lucignani et al.¹⁰ incorporates all abnormalities that are proven to be associated with cCMV infection and it grades the severity of each abnormality, therefore it is reported to be a good predictor for long-term SNHL and neurologic outcomes with score equals to or more than 2. This retrospective study aims to determine whether this MRI severity score can predict the SNHL and neurological outcomes in cCMV infection patients in our local population.

Methods

Research Participants
Antiviral therapy is recommended for symptomatic cCMV patients in the Department of Paediatrics, Queen Elizabeth Hospital (QEH) since 2015 according to international recommendations. All these patients are assessed by paediatric infectious diseases physicians for fitness and timing of starting and continuation of antiviral therapy. cCMV infection is diagnosed by positivity of Detection of Early Antigen Fluorescent Foci (DEAFF) test in urine for newborn infants within the first 3 weeks of life for those infants with symptoms compatible with cCMV infection. An internal cCMV patients log of all cCMV patients is kept in the Department of Paediatrics, QEH since 2015. All cCMV patients are having follow-up in the Department of Paediatrics, QEH till age of 18 years after they have completed audiology surveillance. A total of 24 patients were diagnosed cCMV infection between 1st January 2015 to 31st December 2024 in the Department of Paediatrics, QEH. MRI brain was performed in 21 of these patients. Three patients were excluded either due to defaulted follow-up or defaulted MRI brain appointment. Ethical approval for all stages of the research was obtained from the Central Institutional Review Board of Hospital Authority (PAED-2025-093).

Data Collection
Independent variables include maturity (gestational weeks), birth weight (kg) and the presence of symptoms were collected by reviewing clinical records. Adverse neurologic outcomes were defined as the presence of SNHL, and neurodevelopment outcomes including autistic spectrum disorder (ASD) and global developmental delay (GDD). SNHL in this study is defined as hearing threshold >25 dBnHL by either auditory brainstem response (ABR), play audiometry or pure tone audiometry (PTA) by audiologists.

MRI Acquisition

MRI Protocol
In our institution, either 1.5-T MRI MAGNETOM Avanto Fit (Siemens Healthineers, Erlangen, Germany) or 3-T MRI MAGNETOM Skyra scanner (Siemens Healthineers, Erlangen, Germany) are used to scan these patients. A summary of our institutional protocol (Table 1) is attached as reference.

Image Interpretation
Images of their MRI brain were reviewed by two radiologists blinded to clinical data and gave an MRI severity score by Lucignani et al (Table 2) which consisted of white matter abnormalities, cortical malformation, cerebellar abnormalities, any calcifications and any ventricular dilatation. All the sequences in the MRI examination were reviewed and the images were assessed and scored using the five domains. The minimum possible score was 0 and the maximum possible score was 13. Examples of abnormal MRI brains with their corresponding MRI scores are shown in Figures 1-3. In our study, ventricular dilatation is defined as fronto-occipital horn radio greater than 0.4.

Figure 1 8-month-old girl with cCMV infection presented with severe microcephaly. (a) T2 weighted axial (b) T1 weighted axial MRI images show abnormal T2 hyperintense, T1 hyperintense signals in the white matter of bilateral periventricular regions (arrow), which score 2 marks in the MRI severity score.

Figure 2 3-month-old boy with cCMV infection presented with generalised petechiae and hepatosplenomegaly. (a) T1 weighted axial MRI image shows excessive number of gyri over bilateral frontal lobes (arrow), suggestive of polymicrogyria. (b) T2 weighted axial MRI image shows oval-shaped T2 hyperintense lesions in the periventricular region of bilateral temporal lobes (arrowhead), suggestive of periventricular temporal cysts. (c) T2 weighted coronal MRI imagen shows ballooning of bilateral lateral ventricles (curved arrow), suggestive of ventriculomegaly. This patient has total score of 6 in the MRI severity score.

Figure 3 3-month-old boy with cCMV infection presented with microcephaly. (a) T1 weighted axial MRI image shows simple gyral pattern over bilateral frontal lobes (arrow). In addition, there is periventricular T1 hyperintense signal which shows blooming artifact in the (b) SWI axial MRI image (arrowhead), suggestive of extensive periventricular calcifications. (c) T2 weighted axial MRI image shows extensive T2 hyperintense signal over bilateral parietal white matter (curved arrow). (d) T2 weighted coronal MRI image shows ventriculomegaly. This patient has total score of 7 in the MRI severity score.

Statistical Tests
Different statistical tests have been used in this study. Kappa test is used to assess the inter-observer variability between two radiologists in assessment of the MRI brain. Mann-Whitney test is used to compare the difference between different parameters and their association with the development of neurological outcome. Chi-square test has been used to assess the association between the development of symptoms at birth and neurological outcomes. Fisher's exact test has been used to assess the association between MRI score and development of neurological outcomes.

Results

MRI Alterations
Two radiologists reviewed MRI brain of 21 patients and gave MRI severity score independently. There is good agreement between 2 radiologists by Kappa test (Kappa value 0.796). The combined MRI severity scores range from 0 to 6.5 and the median is 0.5. There are 10 patients who have adverse neurologic outcomes, and 11 patients have no adverse neurologic outcomes. Table 3 is the summary of the patients with their symptoms, age of performing MRI, MRI score, hearing outcome, neurodevelopmental outcome and year of follow up.

Correlation Between Clinical Rsk Factors and Neurological Outcome
By Mann-Whitney Test, the mean gestational week of cohort with neurological outcomes is 36.0 weeks, and the cohort without neurological outcome is 37.3 weeks, and there is no statistically significant association between them (p=0.751). The mean birth weight of the cohort with neurological outcome is 2.35 kg while that of the cohort without neurological outcome is 2.43 kg, and there is no statistically significant association between them (p=0.805) (Table 4).

Regarding the development of symptoms at birth, there are 20 patients that are symptomatic at birth, which 9 of them develop neurological outcome subsequently and 11 of them do not. Only 1 patient included was asymptomatic at birth, and the patient did not develop neurological outcome. These is no statistically significant association between being symptomatic at birth and the development of adverse neurological outcome (p=0.675) by Chi-square test (Table 5).

Correlation Between MRI Scoring and Clinical Outcome
The mean MRI severity score of 2.8 is associated with adverse neurologic outcomes by Mann-Whitney Test (p=0.003). MRI severity score equals to or more than 2 is associated with adverse neurologic outcomes by Fisher's exact test (p=0.001) (Table 6).

Discussion

Based on the literature, we have found the MRI severity score by Lucignani et al predicts the adverse neurological outcome in both the symptomatic and asymptomatic patients. The score gives a comprehensive assessment in different parts of the brain with degree of severity, that leads to difference in clinical severity.

The white matter lesions may be secondary to both hypoxia and inflammatory infiltration and are related to placental involvement by the viral infection.¹¹ Their appearance at MRI examination varies depending on the time of gestation in which the mother contracted the infection. In the first trimester of gestation, Cytomegalovirus has a specific tropism for the immature cells of germinal matrix, while after the 26th week of gestation, astrocytes are more affected.¹² Therefore, the infection in the first trimester may cause delayed myelination together with malformations of neuronal migration and cortical organisation. Late in pregnancy, the virus mostly affects oligodendrocytes, leading to reduced myelination, cerebral atrophy, and ventricular dilation.¹²

In this study, we have demonstrated that the MRI severity score equals to or more than 2 is associated with an increased risk of adverse neurologic outcomes in cCMV infection patients. Therefore, in conclusion, the presence of MRI abnormalities predicts an increased risk of adverse neurological outcome in cCMV infected infants, regardless of whether they are symptomatic at birth. This finding may suggest early, preferably within 2 months of first diagnosed cCMV infection, MRI brains should be performed in cCMV infection patients. The MRI severity score can be used as an important sole predictive factor for adverse neurologic outcomes for cCMV infection patients, so as to guide the clinicians to start the antiviral treatment and provide prognosis to patients' families.

This study is a single-center retrospective study. It may be prone to bias as the sample size is relatively small. Also, there was time interval between the diagnosis of cCMV infection and the performance of MRI limited by its availability, which the brain of the patient may be insulted during waiting for the MRI which may mislead the MRI score assessment.

In the future, further prospective multicentre studies can be considered to explore the result to further validate the results.

Declaration of Interest

All authors have disclosed no conflicts of interest.

References

1. Mussi-Pinhata MM, Yamamoto AY, Moura Brito RM, et al. Birth prevalence and natural history of congenital cytomegalovirus infection in a highly seroimmune population. Clin Infect Dis 2009;49:522-8.

2. Kenneson A, Cannon MJ. Review and meta-analysis of the epidemiology of congenital cytomegalovirus (CMV) infection. Rev Med Virol 2007;7:253-76.

3. Kimberlin DW, Jester PM, Sanchez PJ, et al. Valganciclovir for symptomatic congenital cytomegalovirus disease. N Eng J Med 2015;372;933-43.

4. Rawlison WD, Boppana SB, Fowler KB, et al. Congenital cytomegalovirus infection in pregnancy and the neonate: consensus recommendations for prevention, diagnosis, and therapy. Lancet 2017;17:e177-88.

5. Luck SE, Wieringa JW, Blazquez-Gamero D, et al. Congenital cytomegalovirus – A European expert consensus statement on diagnosis and management. Pediatr Inf Dis J 2017: 36;1205-13.

6. Pesch MH, Kuboushek K, McKee MM, Thorne MC, Weinberg JB. Clinical updates: Congenital cytomegalovirus infection. BMJ 2021;373:1-10.

7. Red Book: 2021–2024 Report of the Committee on Infectious Diseases (32nd Edition): Section 3, Chapter 42, Cytomegalovirus infection. p.294-300. https://publications.aap.org/redbook/book/347/chapter/5751423/Cytomegalovirus-Infection.

8. Walle CV, Keymeulen A, Schiettecatte E, et al. Brain MRI findings in newborns with congenital cytomegalovirus infection: results from a large cohort study. Eur Radiol 2021;31:8001-10.

9. Kachramanoglou C, Jan W, Jones B, et al. Diagnostic analysis of baseline brain MRI features in infants with congenital cytomegalovirus infection: a simplified scoring system. Clin Radiol 2021;76:942.e7-942.e14.

10. Lucignani G, Espagnet MCR, Napolitano A, et al. A new MRI severity score to predict long-term neurologic outcomes in children with congenital Cytomegalovirus infection. J Matern Fetal Neonatal Med 2021;34:859-66.

11. Capretti MG, Lanari M, Tani G, et al. Role of cerebral ultrasound and magnetic resonance imaging in newborns with congenital cytomegalovirus infection. Brain Dev. 2014;36:203-11.

12. Barkovich AJ, Lindan CE. Congenital cytomegalovirus infection of the brain: imaging analysis and embryologic considerations. AJNR Am J Neuroradiol 1994;15:703-15.