Anomalies of the clivus of interest in dental practice: A systematic review.

PURPOSE: The clivus is a region in the anterior section of the occipital bone that is commonly imaged on large-volume cone-beam computed tomography (CBCT). There have been several reports of incidental clivus variations and certain pathological entities that have been attributed to the variations. This study aimed to evaluate the effects of these variations within the scope of dentistry.
MATERIALS AND METHODS: Medical databases (PubMed, Scopus, and Web of Science) were searched using a controlled vocabulary (clival anomalies, cone-beam CT, canalis basilaris medianus, fossa navicularis magna, clival variation). The search was limited to English language, humans, and studies published in the last 25 years. The articles were exported into RefWorks® and duplicates were removed. The remaining articles were screened and reviewed for supporting information on variations of the clivus on CBCT imaging.
RESULTS: Canalis basilaris medianus and fossa navicularis magna were the most common anomalies noted. Many of these variations were asymptomatic, with most patients unaware of the anomaly. In certain cases, associated pathologies ranged from developmental (Tornwaldt cyst), to acquired (recurrent meningitis). While no distinct pathognomonic aspects were noted, there were unique patterns of radiographic diagnosis and treatment modalities. Most patients had a normal course of follow-up.
CONCLUSION: Interpretation of CBCT volumes is a skill every dentist must possess. When reviewing large-volume CBCT scans, the clinician should be able to distinguish pathology from normal anatomic variations within the skull base. The majority of clivus variations are asymptomatic and will remain undetected unless incidentally noted on radiographic examinations.

Introduction

The human skull consists of 22 bones, with the main networks pieced together through fibrous connections called sutures. These sutures allow flexibility and growth between the large flat bones throughout maturation. Over time, these sutures eventually fuse, resulting in an individual's permanent anatomy. However, not all aspects of the skull follow this same pattern. A second way in which bones of the skull form a unity is referred to as synchondrosis, where joints with bony surfaces join through cartilage. Of these unions, there is none more impactful than the joint between the sphenoid and the occipital bone. The occipital bone is located on the most posterior-inferior aspect of the skull and houses the back part of the brain. Just like many of the other bones of the skull, the occipital bone is subdivided into separate parts based on certain characteristics, including distinct anatomy, grooves, or canals. The part affiliated with the spheno-occipital synchondrosis is the clivus.1

The clivus is located on the most anterior section of the occipital bone. This bony part of the cranial base slopes down from the dorsum sellae. While the clivus was previously considered unexceptional, some studies have shown that this aspect does in fact show repeating patterns in certain anomalies from person to person, the 2 most notable of which are canalis basilaris medianus (CBM) and fossa navicularis magna (FNM).1 CBM has been described radiographically as a channel-like osseous defect, usually less than 2 mm in diameter, with smooth borders found in the basiocciput clivus region.12 This variation has also been subcategorized into 6 different subtypes based on whether the channel through the clivus is complete or incomplete.3 In contrast, FNM has been reported as a notch-like, rounded osseous defect with corticated margins in the lower part of the clivus or basiocciput bone.45 These variations have been traced to 2 possible theorized origins, vascular or notochordal. The vascular origin theory states that these variations are caused by persistent enlargement of emissary veins, which are veins that connect vessels outside the cranium to vessels inside the cranium.4 In contrast, the notochordal origin theory, which is slightly more accepted, states that a remnant of the cephalic end of the notochordal canal prevents complete ossification of the clivus during formation.

With the substantial increase in imaging and technology used in the field of dentistry, this anatomic location has gained a crucial place in the diagnostic process. A particularly important advance in imaging is cone-beam computed tomography (CBCT), an important evaluation tool in the field of dentistry that allows the provider to view a 3-dimensional (3D) representation of the region of interest. This systematic review based on published descriptions of clivus anomalies aimed to arrive at an understanding of their features; any notable consequences of the finding, including age, development, and potential pathology; and how patients with clivus anomalies can be managed.14

Materials and Methods

Information sources

The search was conducted at the University of Pennsylvania School of Dental Medicine in Philadelphia, PA, USA using commonly available electronic databases (PubMed, Scopus, and Web of Science). The search was limited to English-language articles published in the past 25 years, with a focus on human studies only.

Search

The search strategy was a combination of MeSH (Medical Subject Headings) terms and free text words. In PubMed, the following search was performed: (clival anomalies) OR (((((canalis basilaris medianus) OR fossa navicularis magna) OR clivus CBCT) OR clivus variation) OR ((((“Cranial Fossa, Posterior” [Mesh]) AND “Cranial Fossa, Posterior/pathology” [Mesh])) AND Variation)). In Scopus and Web of Science, the search was: ((clivus AND CBCT) OR (Clivus AND Variation)).

Data collection process

The results of all 3 searches were exported into RefWorks® (ProQuest, Ann Arbor, MI, USA), where duplicates were deleted. The remaining articles' abstracts were screened by the authors for relevance to the topic of clivus variation and the use of CBCT imaging in diagnosis. The full-text articles were obtained and read for confirmation of direct relevance, as listed in the tables. Articles were further excluded due to a lack of significance or patient-based cases.

Data analysis

The case-based articles that were selected for this research were entered into a table and described under the following subheadings: demographics, variation, developmental or acquired aspects, radiographic findings, symptoms, treatment/follow-up, and clinical significance.

Study selection

From the original yield of 456 titles after duplicates were removed, 40 articles were selected for full-text analysis. Eleven studies were eventually chosen as case-based publications that contained information clival anomalies and CBCT imaging. These 11 studies were analyzed in terms of the above-stated criteria. In addition to those 11 studies, 5 articles were selected as essential background information (Fig. 1). These 5 articles were analyzed in terms of the prevalence, dimensions, and significant patterns noted. Three separate tables were created and analyzed. The reasons for exclusion included a lack of imaging, content unrelated to the clivus, a lack of supporting information, and irrelevance.

Fig. 1

PRISMA flow chart shows the search process and selection of final articles used for this research.

Characteristics of the case-based studies

The 11 studies chosen reported a total of 17 cases of rare clival variations. Of the 17 cases reported, 6 were related to CBM. Nine of the 17 cases described FNM, and the remaining 2 presented cases involved a craniopharyngeal canal and a normal clivus. Although the normal clivus case was included in a study that was chosen for this review, this specific case was not analyzed. The age of the patients ranged from 1 month to 68 years, with a mean of 26.8 years. The patients included 6 males and 11 females (Tables 1 and 2)

Table 1

Summary of case-based publications and their characteristics

CBM: canalis basilaris medianus, FNM: fossa navicularis magna, IV: intravenous, CPC: craniopharyngeal canal, CBCT: cone-beam computed tomography

Table 2

Summary of clinical and radiographic significance of the findings noted

CBCT: cone-beam computed tomography, CT: computed tomography, FNM: fossa navicularis magna, CBM: canalis basilaris medianus, DD: differential diagnosis, IV: intravenous, SOS: spheno-occipital synchondrosis.

Characteristics of the statistical studies

The 5 studies chosen used either dry skull or 3D CBCT imaging to determine the prevalence of clival variations. Three of the studies focused solely on FNM, 1 study focused solely on CBM, and 1 study focused on FNM, CBM, and craniopharyngeal canal. The population groups ranged in size from 168 subjects to 1059 subjects, although 1 study did not specifically include the number of patients analyzed. All 5 studies presented findings on the prevalence of the analyzed variations, and 3 of the studies also presented dimensional variations of the clivus (Table 3).

Table 3

Summary of the prevalence and presentation of variations of clival anomalies

CBM: canalis basilaris medianus, FNM: fossa navicularis magna, CPC: craniopharyngeal canal, CT: computed tomography, CBCT: cone-beam computed tomography

Methodological evaluation

Case selection was evaluated with a methodological assay quality chart. Cases were evaluated based on 8 items in 4 domains:6 selection, ascertainment, causality, and reporting. Only items applicable to the study were evaluated. Numerical scoring was not performed in this systematic review following the recommendation, and an overall judgment regarding the methodological quality and synthesis of the studies included was made in light of the questions deemed most crucial for each clinical scenario. For the methodological evaluation, overall judgments of acceptable or unacceptable were made (Table 4).

Table 4

Methodological evaluation and acceptability of the publications analyzed in this research

Results

The case studies largely fell into the 2 major categories of clival variations: CBM and FNM (Table 1). Furthermore, 3 of the cases fell into the additional subcategories of inferior incomplete CBM, inferior complete CBM, and superior complete CBM. It is important to note that the case-study articles were published due to the uniqueness of the pathology, and these findings therefore do not show a perfect correlation with the real prevalence of variations and pathologies.

In the FNM cases presented, 5 of the cases contained pathology associated with the variation, while 4 cases involved no pathology. Of the cases that involved pathology, 1 was found to have only a developmental pathology (a notochord remnant bone defect filled with lymphoid tissue of the pharyngeal tonsil).7 Two were found to have a developmental pathology with an additional acquired pathology.89 Prabhu et al.9 presented a case with a pharyngeal tonsil remnant that served as the route of acute clival osteomyelitis. Benadjaoud et al.8 presented a case with a Tornwaldt cyst that also served as the route of secondary osteomyelitis. Two cases were found to have an acquired pathology in the absence of a developmental pathology. One of these cases involved an intracranial infection and recurrent meningitis, while the other showed sinusitis and abducens nerve palsy.1011 In addition, Segal et al.10 presented a case with a unique thrombus in the left internal jugular vein.11 Finally, 4 cases were presented as strictly incidental findings with no pathology whatsoever.

In all of the cases involving an infection, including osteomyelitis, intracranial infection, and recurrent meningitis, the patient presented with fever, neck pain, and stiffness of the neck. The stiffness of the neck was largely concluded to be connected to swelling of the cervical lymph nodes. Additional symptoms such as change in consciousness,10 abducens nerve palsy,10 and throbbing headaches11 were noted. The patient with only a developmental pathology also presented with symptoms of sinusitis, which included fever and facial tenderness.7 The patients with no related pathology presented with no symptoms.

No treatment was deemed necessary for the patients who presented with no symptoms at the time, and no follow-up was noted.4 All of the patients who presented with infection were treated with intravenous (IV) antibiotic treatments, differing in type and duration. The IV antibiotic therapies included ceftriaxone and amoxicillin. The osteomyelitis and meningitis cases were treated with a mix of both IV and oral antibiotics,8911 while the intracranial infection was treated with IV antibiotics and anticoagulants due to jugular involvement. Moreover, all 3 infection cases were also treated surgically with drainage, and Alalade et al.11 mentioned repair of the clival defect. All cases showed a favorable course of follow-up, with the slight exception of only partial resolution of the venous sinus thrombosis.10 In the case of a remnant notochord reported by Beltramello and colleagues,7 further treatment was decided against. Their study concluded that the ability to differentiate this anatomic variant from a more severe pathology was crucial.

In the publications dealing with CBM, 2 cases were found to have only a developmental pathology (Tornwaldt nasopharyngeal cyst12 and recurrent nasopharyngeal polyps13). One case was found to have a developmental pathology (pharyngeal enterogenous cyst) with an acquired pathology (partial herniation of the bulb).14 One case was found to have only an acquired pathology (recurrent meningitis).2 Lastly, 2 cases involved strictly incidental findings.1

As with FNM, the 2 cases that presented with no pathology showed no additional signs or symptoms.14 In the cases reported by Syed et al.1 and Lohman et al.,12 the anomalies were found incidentally on CT scans and during the workup of a case of Parkinson disease, respectively. The remaining 3 patients all presented with symptoms ranging from extremely severe, such as vomiting, nystagmus, and bradycardia,14 to moderate, such as headaches, neck stiffness, and vision difficulty,2 or mild, such as snoring and nasal congestion.13

All 3 patients who presented with no symptoms were not deemed to require treatment, and no follow-up information was noted. The 2 patients who presented with developmental pathology were treated surgically, and no further complication or pathology was reported.1314 The case with only an acquired pathology was uniquely treated with IV steroids.2 The follow-up reported no relapse of meningitis to date, but the patient's left eye vision remained poor.2

Statistical outcomes

The statistical publications were included to provide information on the true prevalence of each of these variants (Table 3). Each of the studies was performed in a different population pool, yielding inconsistent results. Two of the 5 studies analyzed CBM, 4 of the 5 studies investigated FNM, and 1 study explored both.

In the FNM studies, the reported values varied to a much greater extent than in the CBM results. The reported prevalence was as low as 3% and as high as 27.5% of the population. Cankal and colleagues15 presented separate analyses of dry skulls and imaging studies, and found FNM in 5.3% of dry skulls, but only in 3% of CT scans. However, Magat5 presented an outlier study with an incredible reported 27.5% prevalence of FNM variations on CT scans. Their article mentioned the substantial difference from previous studies and related the change to differences in methodologies, ethnicities, and sample size. The ranges of dimensions of FNM varied as well, with the mean diameter ranging from 2.85 to 5.37 mm, the mean length ranging from 5.12 to 8.55 mm, and the mean depth being approximately 2.2 mm in all 3 studies.51516

In the statistical studies of CBM, the results were relatively close in prevalence. Bayrak and colleagues17 found that approximately 2.5% of people had this variation, while Currarino3 discovered the variation in 2–3% of adults and 4–5% of children. Currarino3 categorized CBM into 6 different subtypes: 3 complete subtypes (superior, inferior, and bifurcate) and 3 incomplete subtypes (superior recess, inferior recess, and long channel).

Discussion

With the increase in large-volume CBCT scans being taken in the dentistry field, the interpretation of these images has become significant, as missed findings can potentially increase liability in practice. While the area of the clivus may not be the immediate concern, this anatomic landmark is captured in a vast majority of films. During interpretation, the dental practitioner must be able to distinguish normal from abnormal findings, and consequently, decide whether an abnormality (if present) would affect the continuation of treatment. In this study, recurring patterns have been identified that will help clinicians in identifying and diagnosing anomalies within the clivus region and provide assistance in deciding on future treatment.

Each of the cases discussed provides a significant piece of information regarding when to include these variations in the differential diagnosis, what to look for on imaging, what symptoms to identify, when/how to treat, and what the prognosis is following certain treatments. For many of the cases, it was the first time that a certain pathology was published in association with clivus variation, and these cases therefore set precedents for handling these conditions in future patients (Table 2).

The radiographic findings were relatively consistent for each of the findings. The FNM cases showed a circular, notch-like corticated defect on the inferior portion of the basiocciput (Fig. 2). In patients with no clinical symptoms, the notch appeared with a well-defined and corticated periphery.1 However, in the cases associated with pathology, the CBCT readings showed an osteolytic process with cortical destruction,9 a bony dehiscence with “air bubbles” present,10 and evidence of clival osteomyelitis.8 In the cases of infection, a mass or abscess associated with the clivus was also commonly seen on the CBCT image.89 In the CBM cases (Figs. 3 and 4), there was slightly more variation in the appearance due to the existence of different subtypes. In a general sense, the findings of each case were described as a channel-like corticated radiolucency originating from the pharyngeal or superior aspect of the basiocciput to the intracranial aspect of the clivus. At first glance, this tract-like radiolucency could appear similar to a fracture of the clivus, but with the absence of any further osteolytic process. Asymptomatic cases were described as having these well-defined channels,1 whereas the cases associated with pathology did not show as well-defined of a pattern and often had a mass affiliated with the clivus.121314

Fig. 2

Mid-sagittal cone-beam computed tomographic image shows notch-like defect within the pharyngeal portion of the clivus. This is one of the classic appearances of fossa navicularis magna.

Fig. 3

Mid-sagittal cone-beam computed tomographic image shows notching of the superior surface of the clivus demonstrating canalis basalis medianus. The superior recess is 1 of the 6 forms of canalis basilaris medianus described in the literature.

Fig. 4

Mid-sagittal cone-beam computed tomographic image shows complete superior canalis basalis medianus.

A recurring pattern of symptoms affiliated with infection through the clival variation was consistently noted. A patient presenting with fever, neck pain, and stiffness of the neck, without a respiratory or other obvious infection, should be considered for CT imaging. If these symptoms present with clivus involvement, FNM or CBM should be considered in the examination and included in the differential diagnosis. Other common symptoms associated with these clival variations were nasal congestion, headache, and vision defects.

With respect to treatment, asymptomatic cases that did not cause any potential harm to the patients were left untreated.1412 A surgical approach was a common choice for symptomatic pathologies such as cysts,14 nasopharyngeal polyps,13 osteomyelitis,89 and recurrent meningitis.11 All infections were treated with IV antibiotics (such as ceftriaxone and amoxicillin) with the exception of 1 recurrent meningitis case that was treated with IV steroids.2

The course of follow-up was promising for all cases and treatments. The prognosis of these clival variant pathologies can be classified as good, with only vision defects having lasting effects.

This review highlights a couple of key takeaways. First, these variations could affect patients of all ages, including within the first months of life. Second, as a whole, the majority of clivus variations might remain asymptomatic and undiscovered unless radiographically examined. Most of the cases of clival variations were found incidentally on CBCT scans or during research on the prevalence of this condition, and thus required no specific treatment. However, this was not always the case, and symptomatic cases have been published with increasing frequency to promote a better understanding of diagnoses in this area. The findings showed that recurrent cases of infection, especially meningitis, appearance of osteolytic destruction without accompanying symptoms, or the common presenting symptoms of fever, neck pain, and neck stiffness without an accompanying focus, are all reasons to suspect the possibility of a clival variation. Finally, this review showed that while most treatments were consistent, there are different potential approaches. For example, Lohman et al. presented a case where the physician decided against surgical treatment for cyst removal,12 whereas Morabito and colleagues presented a similar case in which surgical treatment was chosen based on the patient's symptoms and prognosis.14 This diversity in approach is also seen from the fact that Jacquemin et al. used IV steroids,2 whereas Alalade et al. chose to use IV antibiotics.11

From the information gained, this review could conclude that the finding of a clival variation would not affect dental treatment. Moreover, there was no evidence to support any alteration or change in conventional treatment for patients with anatomical variants. However, it is important for dental practitioners to be trained in radiographic interpretation so that they can recognize and diagnose variants properly. There was no significant additional benefit of further testing, which is therefore discouraged. Due to the spontaneity of the effects of CBM and FNM, patients should be made aware of the variant in case of idiopathic complications in the future. However, most of these cases would be expected to remain asymptomatic throughout life. If complications do occur, there are proven surgical and non-surgical treatments that can correct the clival variation and address any secondary effects. Unnecessary testing and unwarranted surgical management of CBM and FNM are not recommended in general dental practice.