I read the articles by Watters et al (1) and Dix et al (2) with great interest. Having had experience with a similar case very recently, I felt obliged to emphasize a few points.
First, Watters et al seem to have overlooked at least 10 cases of spontaneous spinal cord herniation (SSCH) in their review, four of which were fortunately included by Dix et al. I know of six more well-described SSCH cases that I would like to draw to the attention of AJNR readers (3–6). The main issue I raise is Watters et al lack of criteria for differentiating spontaneous cord herniation with history of irrelevant or trivial trauma from true post-traumatic spinal cord herniation. Reviewing the literature on spinal cord herniation of all types, I was able to see that when there was a history of slightest trauma, everything focused on this hard-to-get information.
Three cases included in Table 2 of Watters et al's article labeled as post-traumatic cord herniation (7–9) had their injury more than 3 decades prior to their presentation or the onset of symptoms and I think all three of these cases in fact are SSCH cases. Wortzman et al noted that their patient “…had fractured his pelvis…and suffered several minor injuries to his back” 36 years prior to presentation (7). We do not know the level of these minor injuries, but the herniation at T7 is obviously quite distant to the pelvis. Findings in Borges et al's second patient suggest a similar contradiction. This 68-year-old man was hit by shrapnel from a grenade explosion 40+ years prior to symptom onset or his presentation (8). Again, there are no specific data about the level of injury that one can correlate with the herniation at T3 level. The third patient was reported by Urbach et al. It is somewhat more natural for Watters et al to include this case as post-traumatic, because the very author of the article defines a childhood trauma as the cause both in the title and the abstract (9). This 44-year-old patient sustained “a blunt spinal injury without fracture, but with temporary paraparesis,” yet there was no mention of level of injury or mention of type of paraparesis (spastic versus flaccid). The case was labeled as post-traumatic because Urbach et al thought they discarded all other causes.
The second case reported by Watters et al is similar to these, and I do not believe that case is post-traumatic. Their patient experienced a whiplash injury with initial torso hyperextension followed by torso flexion and immediate onset of lumbar pain. I do not think such a trivial event focused in the lumbar area can be the objective cause of a cord herniation at the T6 level, which is a level heavily protected by the support of the rib cage.
I would like to emphasize that if there is not closed or penetrating injury at the relevant level, one cannot scientifically attribute trauma as the cause of the dural defect and hence, the herniation. Likewise, when there is a fractured vertebra (eg, compression fracture) with the herniation at the same level, like the one reported by Baur et al (5), or a penetrating (ie, stab) injury (10) exactly at the level of herniation, only then can one convincingly call this post-traumatic cord herniation.
I suggest that Watters et al may have confused the readership by discussing all types of cord herniation together and by defining new neuroradiologic signs that have no relevance to the diagnosis or differential diagnosis of transdural spinal cord herniation. They say “nuclear trail sign” suggested a previously herniated disk fragment. Such a relationship was previously suggested but has not been confirmed. Both Borges et al (8) and Hausmann and Moseley (11) thought the cause of myelopathy (hence the cause of the dural defect) was the disk herniation, but no surgeon has so far been able to detect any extruded disk remnant while reducing the herniated portion of the spinal cord. The same holds true for a “syrinx associated with post-traumatic cord herniation.” Trauma is a well known cause of syringomyelia, and when trauma is significant, it can cause both.
I agree, however, that SSCH is an underdiagnosed entity and that, with increasing awareness, more and more cases will be revealed because sagittal MR imaging is almost pathognomonic.
References
- Copyright © American Society of Neuroradiology