Atlanto-occipital assimilation: the role of ex-, basi-, and suboccipital bones in the physiopathology of type I basilar invagination. Patient series

Abstract

BACKGROUND

Atlas assimilation is the most common congenital anomaly of the craniocervical junction and is consistently associated with type I basilar invagination (BI1). Its precise bone composition and distribution, however, remain poorly defined.

OBSERVATIONS

All BI1 patients demonstrated atlas assimilation, which was absent in controls. The condyle-mass complex was the most consistent pattern, forming a continuous monoblock in all cases. Atlas anterior arch–basion fusion was universal but asymmetrical in 2 patients, while posterior arch fusion was complete in 5 patients and partial in 3. Compared with controls, BI1 patients exhibited elongated suboccipital, shortened clivus, shortened exoccipital, and widened lateral masses.

LESSONS

CT with multiplanar and 3D reconstructions provides a comprehensive evaluation of atlanto-occipital assimilation in BI1. The condyle-mass complex was the most consistent assimilation site, while anterior and posterior arch fusions were variable. Atlanto-occipital assimilation in BI1 appears to extend beyond midline structures, involving consistent exoccipital–atlas lateral mass nonsegmentation.

https://thejns.org/doi/10.3171/CASE25765

The purpose of this report is to examine the bone composition of atlanto-occipital assimilation throughout its circumference, including arches and lateral masses in type I basilar invagination (BI1), using CT imaging with multiplanar reconstruction (MPR) and 3D reconstruction techniques in patients with BI1 compared with normal controls.

Craniocervical junction (CCJ) structures embryologically originate from the last two occipital and the first two cervical somites, developing into the occipital bone, the atlas, and the axis. The occipital bone and atlas remain united during part of their fetal development, being separated by a delicate process of somitic segmentation.¹

In normal development, the anterior arch of the atlas stays in front of the foramen magnum, leaving free space to accommodate its neural, vascular, and CSFd structures. Disorders in segmentation (atlanto-occipital assimilation) cause the atlas to remain assimilated to the occipital bone, which is considered the most prevalent malformation of the CCJ.2–⁵ When the atlas is assimilated, it remains in line with the casion, dislocating the odontoid process in the direction of the foramen magnum, competing for the free space with the other structures of the foramen magnum (Figs. 1 and 2).⁶

FIG. 1.

A: The anatomy of BI1 showing the atlas assimilated to the basion and opisthion. The virtual normal position of the basion in relation to the odontoid process is shown by the dotted lines. The odontoid process is occupying the foramen magnum area. B: A model showing the normal anatomy (normal segmented CCJ). Note that the anterior arch of the atlas is in front of the basion, which is aligned with the odontoid process. In both situations, the odontoid process lies behind the atlas. Red and blue represent the assimilated parts in each model.

FIG. 2.

CT scans. In the left column, normally segmented CCJ is shown. In the right column, BI1 with atlanto-occipital assimilation is shown. The exoccipital length (A) (for comparison between the assimilated cases and controls) was measured from the jugular tubercles to the lowest part of C1. The anterior arch of the atlas length (B) is the distance between the lateral masses of C1. The lateral mass width is indicated by C, the clivus length by D, and the suboccipital length by E. The tentorium cerebelli and its median insertion on the protuberantia occipitalis interna are shown by the dashed arrow. The atlanto-condylar and atlantoaxial articular clefts are indicated by the dashed lines. In the lower left, note that the occipital condyle (C_O) and the lateral mass of the atlas (C₁) are normally segmented with a clearly visualized joint space. In the lower right, the occipital condyle and the lateral mass of the atlas form a single unsegmented block (condyle-mass complex [CO-C1]), with no articular cleft observed between them.

Basilar invagination (BI) is a condition characterized by the abnormally high position of the cervical spine and is frequently associated with atlas assimilation.^5,6 One of the BI types (BI1) has been consistently associated with atlas assimilation.^2,6

The composition of the assimilated elements has been described in small numbers in anatomical studies.⁷The use of CT with multiplanar reformatting visualizes more precisely the bony composition of the cranial base beyond the anatomical surface and could eventually provide greater pathophysiological elucidation in BI.^8,9,10,11

Study Description

A case series case-control study was conducted using a radiological institutional database. Cases were eligible if they demonstrated BI1, defined as both Chamberlain’s and McRae’s line violation by the odontoid process on imaging (Fig. 2).^8,12^,13

Age and sex distribution data were obtained from the examinations, which were anonymized for analysis. The data were compared with those from examinations considered normal (without BI or any alteration identified or reported on imaging) that were performed for the control group, matched by age and sex.

Exclusion criteria consisted of diagnoses of secondary causes of BI (basilar impression), rheumatoid arthritis, infections, tumors, inherited developmental disorders (e.g., osteogenesis imperfecta), and known genetic syndromes.

Imaging Techniques

Atlas assimilation was studied through volumetric CT via MPR and 3D reconstructions.

Analysis of Atlas Assimilation

The atlanto-occipital assimilation points were assessed along the full 360° circumference of the atlas, in the following distributions: between the right and left anterior hemiarches and the basioccipital bone, between the right and left posterior hemiarches and the supraoccipital bone, and between the right and left lateral masses and the exoccipital bone (respective occipital condyles). The diagnosis of atlas assimilation was defined as the absence of cortical bone or free space between the arches of the atlas and the occipital bone, and the lack of articular space (Figs. 2 and 3).

FIG. 3.

Upper Row: Parasagittal CT scans at the condyle-mass complex. Lower Row:Coronal CT scans. On the left, note the shortening distance between the atlas lateral masses. There is no separation between the exoccipital and C1 lateral masses (condyle-mass complex).

Six possible assimilation locations were evaluated: right anterior atlas hemiarch, left anterior atlas hemiarch, right posterior atlas hemiarch, left posterior atlas hemiarch, right lateral atlas mass, and left lateral atlas mass. Each topography of atlas assimilation was analyzed and classified as nonassimilated or assimilated. The conclusions were recorded in agreement by both observers, with any discrepancies resolved through discussion.

To evaluate the associated effect on the posterior fossa dimensions, the following linear measurements were obtained between the groups in the following bone structures (Fig. 2):

1. Exoccipital: From the jugular tubercle (tuberculum jugulare) to the lowest part of the occipital condyle. For comparison between the assimilated cases and controls, this distance was measured from the jugular tubercles to the lowest part of C1.

2. Suboccipital: From the posterior border of the foramen magnum (opisthion) to the internal occipital protuberance, on the internal surface of the squama of the occipital bone in line with the posterior median insertion of the tentorium cerebelli.

3. Clivus length: In the midsagittal plane, from the dorsum sellae to the basion.

4. Anterior arch of the atlas: Because of its arciform nature, it was replaced by the distance between the lateral masses of C1.

5. Widths of the lateral masses of C1: Between their lateral and medial cortices at the level of the transverse foramen of the atlas. The measurements that occurred bilaterally (exoccipital and diameters of the lateral masses of C1) were unified by the means of the right and left values.

Statistical Analysis

Sex and ages are described by count, means, and standard deviations. The calculations were made using Excel (Microsoft). Differences in the distribution of sexes were compared using the chi-square test, and age differences were compared using the Mann-Whitney U-test.

The normality of the distributions of discrete data was evaluated using the Shapiro-Wilk test in the R statistical environment (R Foundation for Statistical Computing). Two variables showed abnormal distribution, and the data were then compared using the Mann-Whitney U-test for independent samples using the online calculator Social Science Statistics (https://www.socscistatistics.com). The significance level was set at a p value < 0.05.

The demographic data and assimilation details are described in Table 1.

TABLE 1.

Linear dimensions in BI1 and control groups

Type	Age (yrs)	Sex	AAAL (mm)	EXO (mm)	CL (mm)	SO (mm)	LM (mm)
BI1	56	M	9	26.5	41	55	16.4
BI1	65	F	12	25	37	43	14.05
BI1	59	F	16	24	40	37	15.2
BI1	65	M	16	18	37	50	20.9
BI1	80	F	16	24	44	46	15.05
BI1	60	F	12	24	30	50	15.95
BI1	70	F	13	25.5	42	49	15.7
BI1	59	M	16	28	38	40	19.05
CTRL	81	F	16	38	43	36	13.75
CTRL	60	F	15	26	38	40	13.3
CTRL	64	F	13	36.5	46	34	14,1
CTRL	70	F	14	37.5	48	35	13.9
CTRL	64	M	14	35	42	40	14.2
CTRL	60	F	17	35	42	43	14.6
CTRL	55	M	14	38.5	49	40	16.45

AAAL = anterior arch of the atlas length; CL = clivus length; CTRL = control; EXO = exoccipital length; LM = C1 lateral mass width; SO = suboccipital length.

Fifteen cases and controls were evaluated. In the BI1 group, 3 males and 5 females were encountered, while in the control group, there were 5 females and 2 males. There was no statistical difference between age and sex between groups. In the BI1 group, the mean age was 64.25 ± 7.78, and in the control group, it was 64.85 ± 8.5 (U = 26.5, p = 0.45).

Atlanto-Occipital Assimilation

In all cases of BI1, there was assimilation of some part of the atlas to the occipital bone (Fig. 3). In no case in the control group was there any assimilation.

• Assimilation between the occipital condyles and the lateral masses of C1: The most frequently observed assimilation was between the occipital condyles and the lateral masses of the atlas (condyle-mass complex). In all cases, there was assimilation between the exoccipital (condyle) and the lateral mass of the atlas (Figs. 2 and 4).

• Assimilation between the anterior arches and the basion (Fig. 4): All the anterior arches were assimilated. However, there was some heterogeneity in this assimilation. In 2 cases, on one side the arches were completely assimilated, while on the contralateral sides the arches were partially assimilated, revealing certain asymmetry in the craniocervical segmentation defect.

• Assimilation between the posterior arches and the occipital bone (Fig. 4): The posterior arches were bilaterally assimilated in 5 of 8 cases. In 2 cases, one of the posterior arches was completely assimilated, while the contralateral arch was not completely assimilated. In 1 case, one arch was fully assimilated, and the contralateral arch was only partially assimilated. In 4 cases (50%), the posterior arches were unfused in the midline (C1 dysraphism).

FIG. 4.

3D-CT reconstructions obtained in the ventral (upper row) and dorsal (lower row) perspectives.

Linear Dimensions

The mean suboccipital distances in the BI1 and control groups were 47.14 ± 1.89 mm and 39 ± 3.26 mm, respectively (U = 5.5, p = 0.00914). The mean clivus distances in the BI1 and control groups were 36.71 ± 4.6 mm and 43.28 ± 4.49 mm, respectively (U = 11, p = 0.04846). The mean exoccipital distances between the right and left sides of the BI1 and control groups were 29.14 ± 2.76 mm and 34.28 ± 4.64 mm, respectively (U = 8, p = 0.02068). The mean distances between the lateral masses in the BI1 and control groups were 13.42 ± 2.69 mm and 14.42 ± 1.27 mm, respectively (U = 18.5, p = 0.24196). The mean diameters of the lateral masses of the atlas in the BI1 and control groups were 16.17 ± 2.21 mm and 13.61 ± 0.74 mm, respectively (U = 2, p = 0.00248).

Patient Informed Consent

As this study was based on an image data bank, anonymized, the project received permission without patient informed consent. This study was approved by the research ethics committee of Hospital do Servidor Público Estadual.

Discussion

Observations

BI is a disease in which the odontoid process is cranially and posteriorly displaced.¹⁴ BI is classified as type I (BI1) and type II (BI2).¹⁵ BI1 was associated with atlas assimilation and upward odontoid foramen magnum violation, which predispose to craniocervical instability. In BI2, the odontoid is abnormally high relative to the skull base, secondary to occipital bone underdevelopment.

Understanding BI physiopathology is essential for determining the most appropriate treatment. Segmentation defects overload the atlantoaxial joint, potentially causing craniocervical instability in BI1, unlike BI2, where instability frequently does not occur.

Fixation is generally the preferred treatment in type I. In contrast, as BI2 involves underdevelopment of the skull base, posterior fossa decompression without fixation appears to be a more suitable approach.

The most constant location of atlas assimilation to the occipital bone encountered in this study was between the lateral masses and the condyles, determining a single monoblock from the jugular tubercle to the C1–2 joint fissure, occurring in all cases (condyle-mass complex).

The anterior arch of the atlas was also consistently fused to the basioccipital bone. The fact that in some cases it is completely assimilated except in the midline can be explained embryologically because this portion of the atlas originates from the hypochordal bow, while the rest of the atlas originates from the first cervical sclerotome.^1,6–13,16 However, the assimilation was more heterogeneous than that of the condyle-mass complex.

The posterior arch of the atlas was completely assimilated in 5 of 8 cases, completely segmented in 1, and, in 2 cases, assimilation did not occur on one side. The assimilation of the posterior arch may restrain adequate migration of the vertebral arches until the posterior midline, leading to C1 dysraphism.

The linear dimensions of the BI1 and control groups were statistically different, even in this limited study sample. The clivus and exoccipital measurements in the BI1 group were significantly smaller than those of the control group. However, the distance between the internal occipital protuberance and the opisthion was greater in the BI1 group. It was not possible to separate the posterior arch of assimilated C1 from the occipital bone itself, extending down the measurement of the occipital bone (posterior edge of the foramen magnum), due to assimilation. The atlas lateral masses (condyle-mass complex) were significantly wider in the BI1 group, occupying almost the entire anterior margin of the foramen magnum, shortening the space between the lateral masses and, consequently, the anterior arches (Fig. 3).

To date, the pathophysiology of BI1 has been proposed as alterations of the central pillar and changes in the lateral pillar, centered around the axis.¹ The obtained data extended the pathophysiology of BI1 away from the midline to a more lateral malformation, to the condyle-mass complex, which was the most constant finding. The presence of the condyle-mass complex laterally, and the ventral and dorsal assimilation, is a manifestation of segmentation failure.^17,18 In this case, the atlas becomes a caudal extension of the skull, lowering the foramen magnum (Figs. 1–4).

Recently, the phenomenon known as craniovertebral slip-back has been described (Fig. 1).⁶ In a normal individual, the anterior arch of the atlas is located in front of the basion, and the odontoid process is aligned with the basion. When the atlas is assimilated, the craniospinal slip-back alignment is impaired, the atlas is joined to and in line with the basium, and the odontoid projects toward the foramen magnum.

From a biomechanical perspective, the lack of craniovertebral articulation (condyle-atlas) shifts the joint movement to the next effective junction for movement, which is C1–2, overloading this joint.

Limitations and Observations

The number of evaluated cases was limited. Nevertheless, the consistency of the differences found between the groups was statistically significant. Even so, a larger number of cases and controls is desirable.

The presence of the condyle-mass complex should be sought in cases of BI1, and CT studies with MPR should be used in the evaluation of patients. The presence of atlanto-occipital assimilation in conjunction with the projection of the odontoid process into the foramen magnum leads to the diagnosis of BI1 and suggests the use of CVJ stabilization.

The three parts of the occipital bone, the basi-, ex-, and suboccipital, should be studied when evaluating the pathophysiology of BI1.

Lessons

These findings refine previous concepts that linked the median anterior arch of the atlas assimilations with BI1, demonstrating that nonsegmentation occurs more frequently in the lateral region—the exoccipital bone (condyle–mass complex)—at the lateral margin of the foramen magnum.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: both authors. Acquisition of data: both authors. Analysis and interpretation of data: both authors. Drafting the article: both authors. Critically revising the article: both authors. Reviewed submitted version of manuscript: both authors. Approved the final version of the manuscript on behalf of both authors: Botelho. Administrative/technical/material support: Maciel. Study supervision: Botelho.

Correspondence

Ricardo V. Botelho: Instituto de Assistência Médica ao Servidor Público Estadual de São Paulo–IAMSPE, São Paulo, Brazil. bitbot@uol.com.br.