Letter to the Editor: “Diagnostic performance and interobserver agreement of the callosal angle and Evans’ index in idiopathic normal pressure hydrocephalus: a systematic review and meta-analysis”

by Pavel Ryska, Ondrej Slezak, Anders Eklund, Jan Malm, Jonatan Salzer, Jan Zizka (jan.zizka@regionvasterbotten.se)

Diagnostic performance and interobserver agreement of the callosal angle and Evans’ index in idiopathic normal pressure hydrocephalus: a systematic review and meta-analysis

Dear Editor,

We have, with great interest, read the recently published European Radiology article by Park H.Y. et al: Diagnostic performance and interobserver agreement of the callosal angle and Evans’ index in idiopathic normal pressure hydrocephalus: a systematic review and meta-analysis [1].

This meta-analytical review cites among others our research article which has appeared in the Journal of the Neurological Sciences in 2020 [2]. Hereby, we would like to suggest some corrections and comments to the meta-analytical review recently published [1].

First, we would like to raise awareness that callosal angle [3] and simplified callosal angle [4] should be considered two separate and independent biomarkers. The reasons are: they are defined by differently angulated sections (perpendicular to the bi-commissural line versus parallel to the brain stem vertical line) as well as by different anatomical landmarks (posterior commissure versus corpus callosum midpoint). Pooled analysis of these two independent biomarkers may induce bias and, according to our understanding, these two individual biomarkers should always be assessed independently and separately. Not only different anatomical definitions of these two biomarkers, yet also different cutoff levels proposed for callosal angle (90°) and simplified callosal angle (123°) suggest that they cannot be regarded as the same entity. These differences may also offer an explanation of the statement in the Results section: “When a sensitivity analysis was performed based on six studies, excluding a single study using a different method of measuring the callosal angle, the degree of heterogeneity was decreased.” Although additional pooled sensitivity analysis performed after exclusion of the data obtained from simplified callosal angle measurements has been provided in the review, the results of the callosal angle pooled sensitivity and specificity analysis published in the Abstract have also been derived from the simplified callosal angle data which may be considered inappropriate.

Second, meta-analysis of the simplified callosal angle diagnostic performance has been done based on the results of a single, retrospective study published in 2015 [4]. From that point of view, omitting available results of simplified callosal angle diagnostic performance analysis published in other, prospective study [2] may be viewed as a limitation of this meta-analysis [1].

Third, prospectively acquired data on Evans index diagnostic performance (n = 80) published in [2] have not been included in the meta-analysis although they have met the same inclusion criteria as the callosal angle data which (on the contrary) have been included in the meta-analytical review [1]. Furthermore, this prospective study [2] has offered four individual Evans index diagnostic performance analyses derived from four different and separately evaluated anatomical planes (bi-callosal, bi-commissural, hypophysis-fastigium and brain stem vertical line) which, to the best of our knowledge, have not been published before. The fact that those available literary data fulfilling the inclusion criteria have not been included in the meta-analysis might be considered as another limitation of this systematic review [1].