Evaluation of Dural Parameters at C1 Level in Patients with Chiari 1 Malformation Following Foramen Magnum and C1 Posterior Arch Removal: Introduction of a Novel Concept to Decompress without Affecting Stability
Keywords: Atlanto axial instability, C1 laminectomy, Chiari 1 malformation, foramen magnum decompression, posterior arch C1, posterior fixation
Chiari 1 malformation and its management have been a continuous focus of controversies due to many reasons like the plethora of manifestation, incidental detection being one of them, and whether to intervene or not, association with multiple other pathologies, selection of adjuncts with FMD, and at the end, high recurrence of symptoms even though benign pathology.,
But even after all the controversies, FMD along with C1 laminectomy is the gold standard with multiple adjuncts to it, but the high rate of failure and incomplete resolution of symptoms in most, lead to further research and brought up the newer concept of central instability, wherein there is no overt mobility on imaging.
Goel proposed a hypothesis of Chiari malformation caused due to atlanto-axial stability (AAI) and is the protective mechanism to prevent further injury due to mobility, and these changes are reversed once instability is addressed. To achieve this, he promotes posterior fixation as treatment of choice, with exceptionally good results. In decompression of foramen magnum, the removal of the posterior arch of C1 vertebra along with duraplasty is considered to have a better outcome, whereas if the concept of instability of craniovertebral junction is considered, the maneuvring of removal of the posterior arch of C1 will be minimal but may even, in fact, exacerbate the instability, and could even worsen the symptoms.,
This is a prospective, single centre, non-randomized observational study. Patients with symptomatic Chiari type 1 malformation operated with FMDD and C1 laminectomy in the department of neurosurgery from July 2018 to August 2020 were included. Patients of Chiari 1 malformations with atlanto-axial instability (AAI) were excluded. All the patients were evaluated with Magnetic resonance imaging (MRI) of Craniovertebral junction (CVJ) and screening of whole spine, dynamic X-ray CVJ to rule out AAI.
All the patients were followed up at a minimum of six months. Clinical follow up focusing on the relief of symptoms, development of any new symptoms and complications were recorded. Radiological evaluation was performed by MRI CVJ with the screening of the whole spine as well as dynamic X-ray CVJ to look for development of new onset AAI post FMD. For clinical improvement visual analog scale (VAS) and Chicago Chiari outcome scale (CCOS) were calculated.
The dural dimensions were measured on T2WI (T2 Weighted) MRI images in a plane parallel to C1 vertebra, as acquisition was done in the desired plane or using MPR (multiplanar reformation or reconstruction) images, in order to obtain pre- and post-operative images at the exact level. The measurements were performed by two independent senior neuroradiologists with more than 10 years of experience. Any discrepancies were sorted out with combined repeat assessment by both neuroradiologists in order to avoid interobserver bias [Figure 1].
Calculation of the dural diameters at the level of C1 arch
Average diameter of the dura at the level of C1 was measured in both pre- and post-operative MRI images by averaging four diameters in perpendicular planes to each on axial T2 images in a plane parallel to C1 vertebra [Figure 2]. Anteroposterior (AP) diameter of the dura and the width of the posterior arch of C1 vertebra was measured by calculating the average of AP diameter at upper, middle, and lower levels of C1 vertebra in axial sections. Finally, the absolute dural expansion in AP was calculated by the difference between the average AP diameter in pre- and post-operative MRI scans.
Thus, the percentage (%) of dural expansion = (% width of posterior arch removed) = (absolute dural expansion in AP/width of posterior arch) × 100
Area spanned by dura at the level of C1 (cm2) on pre- and post-operative imaging
The area spanned by dura was calculated in pre- and post-operative images by drawing dura using free hand technique on axial MRI T2WI images parallel to C1 vertebra. Wherein, the area was calculated by centricity software encompassed within the marked segment. Besides these, we calculated an arbitrary measurement of pre-operative area spanned by dura including the C1 posterior arch. This was calculated by measuring the extent of C1 laminectomy (displacement distance between excised margins of C1 posterior arch) in postoperative scan. The anterior and lateral limits of dura were marked in a similar fashion as previously in pre-operative scan; however, the posterior limit was formed by the posterior limit of the posterior arch of C1 vertebra, with dura margin in linear continuity. The trajectory is drawn in ellipsoid shape with the projected length of posterior arch excised (in postoperative) forming the base of the ellipsoid shape [Figure 3].
Pre- and post-operatively dynamic X-ray CVJ were done to evaluate for new onset AAI.
The statistical analysis was done using STATA version 16 software. And Wilcoxon signed-rank and Paired t tests were used to analyze the results.
In our study, mean age was 25.3 years with range of 14–34 years, with 5 men and 4 women. Most of the patients presented with motor complaints (weakness - 88.88%) followed by pain (head and neck cumulative - 77.77%) and paresthesia (77.77%).
In our study, C1 was found occipitalized in 2 (22.22%) patients. Majority of patients (44.44%) were operated within 1–2 years of appearance of symptoms, with delayed presentation in the female population [Table 1].
Analysis of dural dimensions
Patients complaining preoperative (n = 6) occipitocervical pain had reduction in pain scores (VAS) postoperatively. A reduction in postoperative VAS (mean 3) compared to preoperative scores (mean 3) was noted in all the patients. There was post-operative improvement in all the parameters, with maximum difference seen in functionality (mean 3.78); no complications were observed. The mean CCOS was 14 (/16). Mean width of posterior arch of C1 vertebra was 3.9 ± 0.45 mm.
The average diameter of the dura at the level of C1 arch showed significant expansion on postoperative imaging (21.09 ± 1.53 mm) compared to the preoperative (20.1 ± 1.89 mm) dimensions. The preoperative AP diameter of the dura was 16.6 ± 2.3 mm, which increased to 18.8 ± 2.5 mm in postoperative period. This was statistically significant (P < 0.05). However, the AP expansion was less than the preoperative AP diameter with posterior arch included 20.73 ± 2.3 mm [Table 2].
This could have been due to the inclusion of two patients with occipitalized C1 and hence were excluded from the analysis. Despite the exclusion of these patients, the increase in average AP diameter of the dura was statistically significant (P = 0.015).
The average length of the extent of laminectomy (displacement of inner border of cut margin of posterior C1 arch) was 15.9 ± 2.9 mm in our study. Two patients with occipitalized C1 were the outliers and on excluding the outliers, the mean dimensions were 15.7 ± 2.4 mm.
The expansion in area spanned by thecal sac/dura at the level of C1 was 3.43 ± 0.85 cm2 on the preoperative imaging which increased to 3.97 ± 0.91 cm2 in postoperative period. This was statistically significant (P < 0.05), but the post-op area expansion was less than cumulative area of dura with C1 arch included 4.33 ± 0.83 cm2 in pre-op [Table 3].
On analyzing data excluding the occipitalised C1, there was still a statistically significant increase in area covered by C1 (P = 0.0156).
The correlation coefficient shows positive correlation between the average absolute AP dural expansion and the percentage of excised posterior arch [Figure 4].
There was no evidence of AAI as noted on postoperative X-ray CVJ in all the patients. There was a statistically significant increase in average dural AP diameter, area covered by dura at level of C1 in post-operative state, and statistically significant decrease in VAS scores in all the patients excluding the outliers.
A mean cut-off of 50.58% C1 posterior arch removal is essential in order to achieve the desired expansion of the thecal sac without actually removing the posterior arch of C1.
The expected result of decompression of the crowded CVJ is not reflected as dural expansion at level of C1, because of multiple reasons. To list a few, in postoperative period there is a reduction in tonsillar volume either due to direct tonsillectomy or hitching of tonsils, or upward migration of herniated tonsils into the posterior fossa due to availability of extra space, provided by FMD.
Goel et al. hypothesized that bony abnormalities like C1 assimilation and bony fusions could be secondary protective responses to longstanding AAI. Thus, fixation of the atlanto-axial joint prevents further injury by limiting CVJ mobility, and improve symptoms or at least halt the progression. In this study, the authors noted excellent results following fixation even without FMD., In another single institutional study, Salunke et al. showed promising results in terms of radiological resolution of the syrinx; however patients did not fare well clinically.
Even though a novel concept, the existence of central instability cannot be completely denied as the management based on it has yielded good results based on the abovementioned studies. The other proof of existence of microinstability at the atlanto-axial joint is the recurrence and incomplete resolution of symptoms after PFD in a substantial number of patients reported in the literature.
Considering the above concept of inherent microinstability at the atlantoaxial joint as a cause of CM1, it would be logical to avoid such an instability in the postoperative period. Therefore, in this study we aimed to focus on a novel way to achieve decompression without trying to affect the stability of CV junction. CV junction is formed by the occipital condyles, atlas (C1), axis (C2) vertebrae and their articulations and any process which violates these structures would aggravate the preexisting instability. Atlas is an atypical vertebra, formed of anterior and posterior arches, which provides attachment to muscles and ligaments. Although minimal, surgical excision of the posterior arch of C1 might contribute to some instability. Hypothetically, C1 laminectomy should aggravate atlanto-axial instability leading to deterioration of symptoms, which may be overshadowed initially by the FMD, but could progress over due course of time presenting with symptom relapse or recurrence, which is seen in a subset of patients with CM1.
In this study, there was presence of C1 assimilation in two (22.22%) patients, with difference in results. Thus, they constituted the outliers, thereby skewing the result. Hence the data was analyzed separately, excluding these patients, despite which the expansion in AP diameter (P = 0.0156) and area (P = 0.0156) was statistically significant.
The correlation coefficient between dural expansion and the extent of laminectomy showed a positive correlation (0.57), suggesting an increase in expansion with corresponding increase in the lateral limits of laminectomy. However, the laminectomy has a lateral limit beyond which it cannot be done.
It was observed that following laminectomy and lax duraplasty, the dural expansion did not extend beyond the limits of the posterior arch of the atlas. The restoration of cerebellar tonsils back into their natural position following FMD, simultaneous tonsillectomy or tonsillar tucking. The other reason could be the soft tissues (muscles, ligaments and fat), which might otherwise cause the dural compression in the absence of C1 arch.
Based on the above hypothesis, we found the resultant average absolute dural expansion (either on basis of linear expansion or area) to be ~50% width of the posterior arch despite complete laminectomy. Therefore, an internal coring of ~50% thickness of posterior arch of C1 can achieve enough room to accommodate for the dural expansion without affecting the postoperative stability. But in case of assimilated C1 the posterior arch needs to be excised completely as adequate duraplasty may not be possible without it.
Kumar et al. showed the need for complete C1 arch removal and thereby suggested indirectly that reduction in acute angulation of dura achieved good decompression. This however was not seen in our series, as all the patients had dural expansion but none beyond the posterior border of C1 arch. Hence, the likelihood of acute dural angulation is minimal with 50% of the posterior C1 arch cored internally.
Reports of spontaneous fracture of the anterior arch of C1 following FMD and C1 laminectomy have been documented in the literature. These patients required fusion to stabilize the atlantoaxial joints, further consolidating the hypothesis of instability that may occur after complete removal of posterior arch of C1. This might support our rationale of preserving the posterior arch of C1 by partial coring (50% of arch).,,
Various authors have recommended wider FMD and C1 laminectomy to achieve better results and minimal complications., Going by the hypothesis of Goel and colleagues, CM1 is a result of atlanto-axial stability, and in fact designated tonsillar herniation acts as a protective mechanism. Therefore, treatment of the cause of tonsillar herniation which is fixation of the instability would provide symptomatic relief rather than treating the effect., This study showed excellent outcomes with 97% of the patients becoming symptom free following fixation of the atlantoaxial joints.
We propose through this study, complete C1 laminectomy may not be necessary to achieve the desired decompression of the neural structures. The recurrence of the symptoms in patients with adequate FMD and C1 laminectomy radiologically, might be explained with the progressively worsening AAI. The target should be a wider length of coring, so that space for dural expansion is wide. This might lead to better outcomes both in terms of clinical radiological aspects. Small sample size and lack of long-term outcome witnessing future incidence of AAI are the limitations. However, this is a pilot study and future studies involving more number of patients with a long term follow up is needed to further enhance our knowledge in this regard.
Chiari 1 malformation represents a complex interplay of various factors and hence accounts for the nebulae of management strategies. The current understanding of the pathogenesis leads to AAI as the common causative factor and hence fixation of the CVJ would be the answer. In this study, we propose that FMD clubbed with partial preservation by internal coring of C1 posterior arch would lead to satisfactory outcomes and lesser recurrence rate by avoiding postoperative AAI.
AAD = Atlanto-axial dislocation, ACM 1 = Arnold Chiari 1 malformation, AP = anteroposterior, C1 = atlas vertebra, CCOS = Chicago Chiari outcome scale, CVJ = craniovertebral junction, FMD = foramen magnum decompression, MPR = Multiplanar reformation or reconstruction, pre op = pre-operative, post op = post-operative.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms. In the form the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3]