Degenerative changes in the cervical spine are part of the normal aging process and are nearly ubiquitous in older people 1. They are generally asymptomatic 23. Spondylosis, with the development of osteophytes, occurs as part of the degenerative process. This can lead to the development of clinical symptoms in some individuals if the osteophytes impinge on neural structures such as the nerve root or spinal cord. If this encroachment occurs in the lateral recess or lateral canal it can lead to radiculopathy. If it occurs in the central canal it can cause myelopathy. However, encroachment in either of these regions can also be asymptomatic with regard to myelopathy 14. For example, Matsumoto, et al 1 assessed 497 asymptomatic subjects and found posterior disc protrusion with compression of the spinal cord in 7.6%. While this figure was presented in the abstract of the paper, no details were provided as to how this compression was measured. However, the figure was similar to that of Teresi, et al 5 who found cord compression on MRI in 7 of 100 asymptomatic subjects. Cord compression without myelopathy has also been found on CT myelography 6.

Cervical spondylotic myelopathy (CSM) is the most common cause of spinal cord dysfunction in older individuals and usually develops insidiously 7. However, it has been reported to develop after trauma 89101112131415. Some authors have suggested that individuals who have asymptomatic spondylotic encroachment on the cervical spinal cord are at increased risk of acute myelopathy if they experience minor trauma such as a fall or motor vehicle collision 1617. This has led some surgeons to recommend decompression surgery for the purpose of preventing this trauma-induced myelopathy in presumed susceptible individuals 1819. For example, Epstein 18 stated "Patients under 65 years of age, if mildly symptomatic or at risk for quadriplegia with even mild trauma, may warrant early decompression". However, he did not provide evidence-based recommendations as to how to determine risk of quadriplegia or the level of risk that would warrant surgery in the absence of frank myelopathy.

The authors, all non-surgical spine specialists, have had patients consult them for second opinion after being recommended this type of surgery. Each of these patients was asymptomatic with regard to cervical myelopathy (though they had neck pain), but cervical MRI had revealed cervical spondylosis which encroached on, and compressed, the spinal cord. It was reported in each of these cases that the surgeon making the recommendation did so based on the view that the spinal cord encroachment placed the patient at risk of spinal cord injury if he or she were to experience even relatively minor trauma. These patients expressed a desire for a non-surgical opinion as to whether such surgery is truly advisable. This is apparently a frequent enough occurrence in the experience of other spine specialists to have warranted a "Curve/Countercurve" piece in a recent issue of Spine Line, a publication of the North American Spine Society 19.

Evidence-based medicine calls for the clinician to provide counseling and treatment that is based on the best available evidence, combined with clinical experience and patient preference 202122. The purpose of this review is to investigate whether the scientific literature can be used to inform the surgical and non-surgical spine specialist regarding how to advise patients who have spondylotic encroachment on the cervical spinal cord in the absence of frank myelopathy.

The following databases were searched up to May 31, 2008: Medline, Cinahl, Embase and MANTIS. Searches of the authors' own libraries were also conducted. Finally, citation searches of relevant articles and texts were conducted manually. The search terms used for the database searches can be found in table 1.

The search yielded 1881 citations. Relevant papers were retrieved and reviewed by two independent reviewers. Studies that were deemed relevant were those that investigated the risk of spinal cord injury from minor trauma in patients with pre-existing spondylotic central canal encroachment and those that reported on outcomes and complications to cervical decompression surgery, with or without fusion. Case reports and small case series were excluded. Also excluded were studies reporting risk of spinal cord injury resulting from major trauma and studies involving individuals who had narrowing of the central canal from sources other than degenerative changes. In cases in which systematic reviews of the literature were found, the individual studies included in the reviews were not reviewed separately, unless this was necessary to clarify information that was not readily apparent from the systematic review.

The role of preventive surgery in patients with asymptomatic cervical spinal cord encroachment has been a point of controversy amongst surgeons. Riew, in a point-counterpoint piece, 19 argued that the risk of myelopathy in patients with asymptomatic encroachment on the cervical spine is not worth the risk of surgery. Combining data from the Paralyzed Veterans of America, National Library of Medicine, and the US Census, he estimated the "worst case scenario" risk of myelopathy in this patient population to be 1:2100. He argued that even if the risk of serious complication from surgical decompression was 1:1000, this would be twice the risk of myelopathy after trauma 19. As has been pointed out in the present paper, however, the studies Riew cited on which he based the assumption of risk were of inadequate design to assess true risk 2526. However, this point only strengthens his recommendation against surgery in this population. Others 18 have argued that because of the potentially catastrophic nature of spinal cord injury after trauma, decompression surgery is appropriate in this patient population. The purpose of this study is to assess the evidence regarding this risk and attempt to compare what is known about this risk with what is known about the risk of surgery. It is hoped that all spine clinicians can take an evidence-based approach to counseling patients with this condition.

All studies that related to the risk of spinal cord injury in patients with asymptomatic encroachment located in the search were case reports, case series or retrospective cross-sectional studies. None were case-control or prospective cohort studies. Thus, while it can be said that there may be an association between the presence of asymptomatic cord encroachment and spinal cord injury after trauma, no firm conclusions can be drawn about causation. Case-control or prospective cohort studies would be necessary to make this determination 35. Also, in the majority of cases the size of the central canal was measured with radiographs. Recent evidence indicates poor correlation between radiographically-determined central canal size and that determined by MRI 36. Because the studies were of inadequate design to assess risk and used inadequate measurement methods, the present authors did not feel that it was of benefit to undergo a formal critical appraisal of the studies.

Bednarik, et al 3738 have studied risk factors for the development of CSM in individuals with asymptomatic spondylotic cord compression using a prospective cohort design. In their initial study of 66 subjects with this condition who were followed for 2–8 years 37, they found that 13 subjects (19.7%) developed symptomatic CSM. The only risk factors for the progression to CSM in this cohort were symptomatic radiculopathy at baseline, electromyographic (EMG) evidence of anterior horn lesion at baseline and abnormal somatosensory evoked potentials (SSEP) at baseline. In a more recent publication with a larger sample size (n = 199) and longer follow period (2–12 years, median 44 months) 38 they found that 45 subjects (22.6%) developed symptomatic CSM. Baseline symptomatic radiculopathy, EMG evidence of anterior horn cell lesion and abnormal SSEP were found to be risk factors for the development of CSM during the follow up period. There was a tendency toward increased risk in males vs females and in those with abnormal motor evoked potentials, but these did not reach statistical significance (p = 0.072 and p = 0.112, respectively). Factors in their model that were not found to increase risk of the development of CSM were age, type of compression (spondylosis, disc herniation or the combination of both), number of stenotic levels, decreased cross sectional area of the spinal canal, decreased Pavlov ratio and hyperintense signal within the spinal cord on T2-weighted MRI image. They did not include exposure to trauma in their analysis, however, when re-analyzing the data they found relatively few exposures to trauma and that these had no impact on development of CSM (Bednarik J, personal communication 26^th June 2008).

In all the surgical studies found in the search, the subjects had symptomatic myelopathy. No outcome studies were found that included asymptomatic subjects thought to be at risk. Thus, the role surgery plays in preventing spinal cord injury in asymptomatic subjects thought to be at risk is not known. It is also not known whether the complication rate of decompression surgery in patients with asymptomatic cord encroachment would be the same as in those with myelopathy. However, as the reported postsurgical complications generally relate to the surgery itself and not to the myelopathy (see Table 1), it is not likely that the complication rate would be substantially different in asymptomatic individuals as compared to symptomatic individuals.

Based on this review of the literature, it remains to be determined whether an individual with cervical spinal cord encroachment, without signs or symptoms of myelopathy, is at increased risk of spinal cord injury after trauma. It also remains to be determined what the magnitude is of any increased risk. This determination would require population-based case-control or, preferably, prospective cohort studies. With these designs, bias can be minimized and statistical conclusions can be drawn regarding risk 35. Until such studies have been performed, it cannot be stated with certainty that individuals with the findings discussed here are at increased risk of trauma-induced myelopathy.

Because of this, there is currently no substantial evidence upon which to base a recommendation for prophylactic decompression surgery in this patient population. However, evidence-based medicine calls for recommendations to be individually directed and to take into account scientific evidence combined with clinical experience and patient preference 202122. There may be individual variations in a particular case, such as severe canal encroachment, low signal change within the spinal cord on T1 weight images with high signal on the T2 weighted images (which has been found to correlate with poor surgical outcome) 39, ossification of the posterior longitudinal ligament or persistent engagement in high-risk activities, which may influence one's recommendation. Also it may be advisable for the non-surgical spine specialist to counsel patients who have asymptomatic cord encroachment to avoid high-risk activities, particularly those that could involve high-acceleration extension injury. Given the fact that post-traumatic myelopathy has been reported to be associated with falls in the elderly 40, it would be reasonable for elderly patients with this finding to be provided prevention strategies, including exercises for improved balance, in order to lessen the likelihood of falling 41.

Asymptomatic cervical spondylotic spinal cord encroachment is fairly common. It has been said that individuals with this finding are at increased risk of severe myelopathy if they experience minor trauma. In some cases, prophylactic decompression surgery has been recommended. However, there is no good evidence that these individuals are at increased risk and, given the potentially serious complications of surgery, the evidence does not allow for firm and broad recommendations to be made regarding prophylactic surgery. Population-based case-control or prospective cohort studies are needed to determine whether the magnitude of any risk in this patient population justifies surgical intervention.

The authors declare that they have no competing interests.

DRM conceived of the research idea, supervised the literature search and data extraction process and was the principle writer of the manuscript. CMC and JKG conducted the literature searches and were involved in data extraction. All authors reviewed and made editorial changes in the manuscript. All authors read and approved the final manuscript.