To the Editor: In their article “Optimizing Endoscopic Approaches for Sphenoid Lateral Recess Cerebrospinal Fluid Leaks: Proposal of a New Algorithm,” Manogaran et al1 present what they describe as a novel 4-tier anatomic classification of lateral sphenoid encephaloceles based on their anatomic location with respect to Foramen Rotundum (FR). The authors reference one of our papers where we validated our own 4-tier classification of lateral sphenoid encephaloceles, which we also classified based on the anatomic relationship between the encephaloceles and FR.2 Of note, the authors failed to reference our original paper where we first introduced our classification system.3 In our classification, we described 4 types of lateral sphenoid encephaloceles: Type 1 is the classic Sternberg canal leak, which rarely if ever occurs and likely represents a misnomer for this pathology. Type 2 is an encephalocele that occurs medial to FR. Type 3 is an encephalocele that lies lateral for FR. Type 4 is an encephalocele that passes through and enlarges FR. We previously validated this classification in a multicenter study of 49 patients. In this smaller study, the authors have kept the first 3 types of encephalocele locations but redefined Type 4 as a “far lateral” encephalocele that occurs farther away from FR than Type 3 encephaloceles. They then go on to present an algorithm for fixing these leaks based on the distance between the Vidian canal and V2, with the primary point being that if the distance is >4 mm, they recommend considering a contralateral precaruncular approach, a surgery they have never performed in vivo but only in cadavers. Although we commend the authors on tackling this difficult neurosurgical problem and also their excellent repair results with no failures in any of their cases and no instances of facial numbness or dry eye, we wish to discuss a few aspects of their reclassification of lateral sphenoid encephalocele and their algorithm for treatment, which we find somewhat problematic. First, the authors have eliminated our description of encephaloceles that pass directly through and enlarge FR. We feel this is an important category since in their classification, it is not clear whether such an encephalocele would be Type 3 or Type 4. If Type 3 is medial and Type 4 is lateral, how do they categorize those that pass directly through and enlarge FR? In our study, we found that 16% of encephaloceles passed through and enlarged FR. For this reason, we created a Type 4. The authors also present multiple subtypes of their Type 4 encephaloceles, some that destroy the entire floor, some that are higher than FR, and some pass lateral to the sphenoid. In our classification, these are all just Type 3 lesions, i.e., lateral to FR. The other issue with their schema, which divides encephaloceles into Type 3 (lateral to FR) and Type 4 (far lateral to FR) is that there is no objective distinction between these 2 categories. How far lateral does the encephalocele have to be to become Type 4? This ambiguity is problematic and will reduce intersubject reliability in classifying encephaloceles. One could argue that the reason for one vs another classification is that it makes a difference in what type of repair is chosen. The authors make this argument, claiming that the Type 4 lesions are challenging to repair endonasally and are “ideal” for a precaruncular approach, a surgery they have never performed on a human being. We wish to caution the authors about recommending surgical approaches that they have not validated in the operating room. That said, we agree that lateral sphenoid encephaloceles that are either large and pass through FR or are lateral to FR are more challenging to repair via the endonasal approach, often requiring sacrifice of the Vidian nerve and increasing the risk of damaging V2. Indeed, in our multicenter study, symptoms of Vidian and V2 numbness occurred only in Type 3 and Type 4 encephaloceles. For this reason, we have suggested that the lateral transorbital approach might be safer and less morbid method for fixing these lesions. We have performed this approach in a few patients and reported our results.4 We stand by our 4-tier classification and hope that further experience and publications on this subject will find value in our contribution. We also wish to offer that the lateral transorbital approach, rather than the precaruncular approach, will be more effective at managing encephaloceles that are lateral to FR and difficult to reach endonasally.
Theodore H. Schwartz (Wed,) studied this question.