Letter to the Editor: “Clivopalate angle: a new diagnostic method for basilar invagination at magnetic resonance imaging”.

by José J. C. Nascimento, Eulâmpio J. S. Neto, Severino A. Araújo-Neto, Paula R. B. Diniz (josejailson64@gmail.com)

Clivopalate angle: a new diagnostic method for basilar invagination at magnetic resonance imaging

Dear Dr. Y. Menu,
Editor-in-Chief of the European Radiology,

With interest, we had the honor to read the paper by Lichao Ma et al titled “Clivopalate angle: a new diagnostic method for basilar invagination at magnetic resonance imaging”, recently published in European Radiology. This study evaluated at MRI the diagnostic accuracy of clivopalate angle (CPA) for “basilar invagination” (BI) comparing this measure with clivodens angle (CDA) and clivoaxial angle (CXA).

Respectfully, we have made some observations that we believe are important. The study introduction emphasizes the importance of the odontoid distance to Chamberlain line (LC) [1] in BI diagnosis. The research protocol describes that the BI group (n = 112) had clinical manifestations of brainstem and “with protrusion of the odontoid tip > 5.0 mm above the CL on sagittal MR images”. The LC measurement was performed by a single experienced radiologist. In diagnostic accuracy studies the reference standard must be robust enough to strengthen the definition of groups. The method could have provided some evidence of intra or inter-examiner reliability for the LC.

In one of the study justifications, the authors discuss the difficulty of identifying the anatomical landmarks of LC in MRI, as well as the risk of bias in the LC accuracy due to anatomical variations of the odontoid process. In fact, the results “showed five types of morphological variations (Fig. 3) in 50 cases in the BI group and in 18 cases in the control”. Because it is a study of diagnostic accuracy specifically for BI, the sampling of this group could have used more than one craniometric criterion, such as other classical parameters and contact or compression of the odontoid tip on brainstem and spinal cord evaluated by experienced radiologists [2, 3]. Although the BI group is symptomatic, Table 2 indicates that slightly more than 50% of these patients showed other comorbidities, such as Chiari malformation and syringomyelia, which can occur without BI and can simulate similar clinical manifestations of BI [4].

Table 2 shows that 70 patients with BI had atlantoaxial subluxation. In the classification discussed by Goel in 2004 [5], these patients should be classified as BI of type A, which is characterized by a joint alteration frequently observed in patients with rheumatoid arthritis and trauma. Figure 5d shows a case of BI type A in which the patient has a well-developed cranial base. The presence of other abnormalities of craniovertebral junction was evaluated and showed in Table 2. It does not describe presence of occipital components hypoplasia (basioccipital, occipital condyles), which are frequently observed in patients with BI of type B [2, 5, 6]. Although, Figure 3f represents a common case of hypoplasia of basioccipital (clivus), with the basion approximately at the height of the pontine bulb groove.

The new parameter CPA proposed by authors uses only anatomical landmarks of the skull base (clivus by Wackenheim line and hard palate plane), which would not necessarily be altered in patients with atlantoaxial subluxation. According to the results “no structural abnormality of the hard palate was identified in either group”. In theory, parameter CPA would not have anatomically explicit relationships with the pathophysiology of BI type A. Commonly, patients with BI type B show a greater plane inclination at hard palate on midsagittal MRI and hypoplasia of clivus, as previously discussed. This would be an appropriate factor for to use of CPA in the BI type B. It would be interesting if the study protocol had tested the objectives using BI as A and B types, since are conditions with different pathophysiology and treatment [3, 4, 7, 8].

Finally, it is desirable that future studies involving diagnostic accuracy follow the Standards for Reporting Diagnostic Accuracy Studies (STARD) [9], in order to strengthen evidence-based medicine