Assessment of Relationship between Clinical Manifestation of Chiari Malformation Type I and Volumetric Measurement of Cerebellar Tonsils Herniation within Foramen Magnum

Background

The description of CMI was initially based on pathological and anatomical anomaly related to the cerebellar tonsils [1]. Based on anatomical definition; CMI is characterized by hindbrain deformity and elongation of the cerebellar tonsils resulting into their descent into the foramen magnum. The classic definition of CMI is radiological and is defined as herniation of the cerebellar tonsils below the foramen magnum of >3 mm in children and> 5 mm in adults [2]. There are questions about the most accurate radiological criteria to diagnose CMI which remain unanswered. Is CMI a sagittal disease? should we only rely on the cut off descent of cerebellar tonsils of 5mm below foramen magnum in sagittal plane in diagnosis of CMI ? Can the descent of cerebellar tonsils in coronal plane reflect the severity of CMI? Can the volume of cerebellar tonsils herniation reflect better the severity of CMI? Can T/F volume ratio (Cerebellar tonsil volume/Foramen magnum volume) reflect better the severity of CMI patients?

Data Management and Statistical Analysis

To assess the relationship between clinical manifestation of CMI and cerebellar tonsils, we computed various non- parametric statistical tests and hypothesis testing to analyze correlation and variation of cerebellar tonsils measurement in sagittal, coronal and axial planes, cerebellar tonsils volume, and the T/F volume ratio (cerebellar tonsils volume/ foramen magnum volume).

We run Spearman Rank correlation test to analyze correlation between headache severity and both sagittal and coronal measurements (right and left), correlation between total cerebellar tonsils volume and foramen magnum, correlation between cerebellar tonsils volume and headache severity, myelopathy severity and tonsils volume, and T/F volume ratio and myelopathy. The hypothesis tests using Wilcoxon rank sum tests was also constructed to assess the significance of variation between right and left cerebellar tonsils measurements (coronal and sagittal measurements) in diagnosis of CMI, cerebellar volume (left and right), and sagittal and coronal measurements. We computed descriptive analysis to determine demographic characteristics of the study participants, occurrence of signs and symptoms among patients with CMI. Before running every test, we assessed the skewness of the data by variable and made a log-transformation whenever suitable likely, we found none. When we found that the data were skewed, we went on running and look for suitable non-parametric tests. We did all the calculations in Python 3 using scipy. Stats, Wilcoxon, Pearsonr, Seaborn and Matplotlib, pyplot Packages and Panda library.

Results

The study included 96 patients based on calculated sample size and statistical power of the study. CMI was most common in female with male to female ratio of 1:1.5. The majority of patients with CMI in our study was in fourth and fifth decade with age ranges: less than 20 years of age, between 20-30 years of age ,between 40-50 years of age and above 50 years representing 7.3%, 12.5%, 39.6%, 32.3% and 8.3% respectively.

Myelopathy followed by sensory loss was the common finding .Upper motor neuron lesion like Hoffman sign, spasticity and clonus was found to be 8% and 18% respectively. The motor weakness as result of CMI was found in 15% of the study population. Scoliosis as result of CMI was found in 5% while syringomyelia was found in 8% in study population.

Click to enlarge

Click to enlarge

Headache and myelopathy severity was graded according to NRS and mJOA respectively. Patients with CMI mostly presented with severe headache, moderate headache with 58.3% and 41.7% respectively according to pain NRS scale in the study. The severity of myelopathy was classified into mild, moderate and severe according to mJOA. Most patients with CMI was mild myelopathic with 81.2% followed by moderate myelopathy with 11.5 % and severe myelopathy in 7.3% of study subjects (Figures 4 & 5).

The measurement of herniated cerebellar tonsils was done in 3D; coronal measurement, sagittal measurement, axial measurement and volume of herniated cerebellar tonsils were calculated.

The diameter of FM was measured and the volume of FM calculated using sphere formula and transverse diameter of FM diameter respectively. The findings are the following:

• The left median coronal measurement was 11.2 mm with interquatile (IQR) range (9.3 mm- 13.1 mm).
• The right median coronal measurement was 11.4 mm with interquatile range of (10.3 mm-14.1mm).
• The left median sagittal measurement was 7.8 mm with interquatile range of (6.2 mm -10.4mm).
• The right median sagittal measurement was 8.8 mm with interquatile range of (7.2mm-11.3mm).
• The median transverse diameter of FM was 38.25 mm with interquatile range of (36.2mm-40.63mm).
• The median volume of foramen magnum was 29,571.3 mm3 with interquartile range of (24,774.6mm- 35,088.1mm).
• The median of total cerebellar tonsils volume was 2,427.61mm3 with interquartile range of (1,706.01mm3 - 3,625.2mm3).
• The median and interquartile range of T/F volume ratio (cerebellar tonsils volume/FM volume) was 0.082(0.053 0.177) respectively.

There was a relationship between myelopathy severity and the volume of herniated cerebellar tonsils as well as the relationship between T/F volume ratio and myelopathy severity according to mJOA; there was also a correlation between myelopathy severity and the volume of herniated cerebellar tonsils. The more patient had an increase of volume of herniated cerebellar tonsils, the more the severity of myelopathy on mJOA scale (Figure 6).

Click to enlarge

There was a crescendo relationship between T/F volume ratio (cerebellar tonsils volume/Foramen magnum volume) and severity of myelopathy. The increase of volume ratio is proportional to the severity of myelopathy on mJOA scale (Figure 7).

Click to enlarge

Discussion

The majority of patient with CMI in our study was in fourth and fifth decade with age ranges: less than 20 years of age, between 20-30 years of age, between 40-50 years of age and above 50 years representing 7.3%, 12.5%, 39.6%, 32.3% and 8.3 % respectively. Aska Arnautovic and colleagues in their review of study on demographics, operative treatment and outcome showed that peak age of presentation in the pediatric studies were 8 years, followed by 9 years, and in the adult series,41 years, followed by 46 years old with female to male ratio 1.3:1 [5]. James R Houston, et al. in their study on evidence for sex difference in morphological abnormality in CMI report no sex difference of morphometric abnormalities of CMI malformation [6].

Patients with CMI mostly presented with severe headache and moderate headache with 58.3% and 41.7 % respectively according to pain numeric rating scale in our study. There was no correlation between headache severity and cerebellar descent measurement in both sagittal and coronal. In contrast there was proportional correlation between headache severity and T/F volume ratio (cerebellar tonsils volume/ foramen magnum volume ratio). The study by Heffez DS, et al. [7] reported that there is no correlation between symptoms severity with extent of cerebellar descent in CMI patients [7].

Jonathan Pindrik, et al. in their study on clinical presentation of CMI and syringomyelia in pediatric population report headache or neck pain in 27%-70%, motor or sensory deficits of the extremities, 6%–17%, Scoliosis 18%-50%, Ataxia or gait impairment, decreased coordination 4%–9%,Spasticity 6% [8].

The syringomyelia is define as cerebral spinal fluid cavitation within the spinal cord seen as high signal intensity on T2W image MRI sequence. Syringomyelia was found in 8% of our study population.

The pathogenesis of syringomyelia associated with CMI has not been clarified. Many theories explaining the presence of syringomyelia in CMI have been postulated in literature, however disorder in CSF pathway around FM is taught to be main cause of syringomyelia [9].

The prevalence of syringomyelia in CMI is variable among adults and pediatric population. The occurrence of a syringomyelia in association with CMI varies between 35 to 75% in pediatric patients while in adults syringx associated with CMI varies between 40 to 60% [10]. Aitken and colleagues identified syringomyelia in 12% of patients with CMI [11].

The scoliosis associated with CMI in our study is 5% of the study population.

The exact mechanism of scoliosis in association with CMI is not well known. It is thought to result from chronic compression of cervico-medullary junction by cerebellar tonsils, abnormality in posterior column tracts affect postural reflex resulting in difficulty sustaining appropriate posture there by subsequent scoliosis. However others authors associated the scoliosis with the presence of syringomyelia [12]. Mohammad Hassan A, et al. who report scoliosis resulting from isolated CMI in 3.0% [12]. On the other hand Palma Ciaramitaro, et al. found scoliosis in18%-50% of CMI [10].

Sahib SS, et al. [13] in their study on morphometric and volumetric in pediatric population with CMI found no clinical usefulness of 2D or 3D measurements that can help to differentiate symptomatic pediatric patients with CMI from those who are asymptomatic for CMI in their study, hence recommend another diagnostic criteria beyond classic measurement of tonsil position useful for detecting best candidates for surgical management in patients with CMI [13].

Our study showed a correlation between myelopathy severity and the volume of herniated cerebellar tonsils. There was a proportional increase of volume of herniated cerebellar tonsils and the severity of myelopathy. In addition, our study found a crescendo relationship between T/F volume ratio (cerebellar tonsils volume/foramen magnum volume) and severity of myelopathy. The increase of T/F volume ratio is proportional to the severity of myelopathy.

Currently, there is paucity of data in literature showing the relationship of severity of myelopathy and cerebellar tonsil volume as well as the relationship between T/F volume ratio (cerebellar tonsils volume/foramen magnum volume) and severity of myelopathy as discussed in our study. The T/F volume ratio (cerebellar tonsil volume/Foramen magnum volume) and cerebellar tonsils volume can be an additional tool to be used in diagnosis and predict the severity of CMI.

Conclusion

Many studies showed no correlation between cerebellar tonsils herniation and clinical severity in patients with CMI and suggest other diagnostic criteria beyond standard measurement of cerebellar tonsil position which can help in determining the best candidates who need surgical management for patients suffering from CMI. Our study is the first in current literature to show the correlation between the tonsil volume and T/F volume ratio (cerebellar tonsil volume/ foramen magnum volume) and headache severity using NRS as well as severity of myelopathy using mJOA scale. The study shed light on cerebellar tonsil volume and T/F volume ratio as predictors of severity in patients with CMI. Further studies are recommended to find out if cerebellar tonsils volume and T/F volume ratio (cerebellar tonsil volume/foramen magnum volume) can be used as diagnostic tool in patients with CMI.

Study Limitation and Strength

We should have used imbedded digital calculation of the volume instead of classic calculation of sphere formula which gives approximate result but not true results of calculation. The Foramen magnum is virtual therefore it is exact shape is debatable. The volumetric measurement of cerebellar tonsils and foramen magnum are ideal reflection of severity in patients with Chiari Malformation type I.