Anderst et al. have investigated what happens to cervical range of motion above and below a C5-C6 anterior cervical discectomy and fusion in comparison with findings in asymptomatic control subjects who had not undergone an operation. I commend the authors for a well-done study. They used biplanar video fluoroscopy to assess total cervical spine motion and found that this technique was better able to capture the total range of motion, as compared with static flexion-extension radiographic views. At the C4-C5 level, they found that, while the fusion did not decrease the total range of adjacent-level motion, there was increased extension and decreased flexion motion. In addition, there was increased adjacent-segment translation at the levels both cranial and caudal to the fusion level.
The major strength of the study is that the authors utilized an established, reproducible, accurate, and useful technique for documenting motion changes in the cervical spine. They found that, even at only seven months postoperatively, there are subtle motion changes at the segments adjacent to a fusion. While it is easy to conjecture that these changes might result in accelerated degeneration, the clinical relevance of these changes and, more importantly, whether they have anything to do with degeneration that occurs at the adjacent level are not addressed by this study. Published reports of data on arthroplasty versus arthrodesis derived from U.S. Food and Drug Administration investigational device exemption trials failed to show that there is any difference between these two treatments in terms of the incidence of adjacent-segment reoperations at up to five years postoperatively1,2. Systematic literature review studies have failed to show a difference in adjacent-segment pathology between motion-preserving and arthrodesis procedures3. Other reviews have demonstrated no differences in adjacent-segment kinematics between arthrodeses and arthroplasties4. Therefore, the impact of a cervical arthrodesis on adjacent segments remains controversial.
The authors state that the patients could not be their own controls because they had pain preoperatively. However, it may still have been better to perform the initial studies on subjects with minimal neck pain and then repeat those studies on the same subjects seven months postoperatively. Patients who are in need of cervical spine surgery are likely to have greater degrees of disc degeneration than asymptomatic controls who do not need surgery. Therefore, it is possible that the operatively treated patients had the findings seen in this study inherently and not because of the fusion; there was no qualitative or quantitative assessment of the extent of their disc degeneration compared with that of the controls.
As the authors pointed out, seven months is such a short time postoperatively that true kinematic changes may have yet to manifest themselves. It may take one to two years or more to see the true scope of changes. In addition, it is possible that not all arthrodesis sites had fused solidly by seven months. It would therefore have been helpful to have divided the patient subset into those in whom the site had fused and those in whom it had not fused to see if there was a difference. This may not have been practical, since there were only ten patients. In addition, there were no data at the time of the study regarding the presence or absence of neck pain in the operatively treated patients, which obviously can alter the results. I would encourage the authors to repeat the study on the same subjects at two years, comparing their data not to those in asymptomatic controls but to the data in the same subjects at seven months. This would give us valuable longitudinal data.
There are a few minor points. The authors noted that static flexion and extension cervical spine radiographs are less able to determine the full range of cervical spine motion than are dynamic images made while the patient flexes and extends the neck. While some of this difference may be due to the inertia of moving one’s neck, it is difficult to believe that, at the one cycle per three seconds utilized, there is substantial inertial movement beyond the end point achieved by a patient who has been coached to fully flex and extend the neck. I suspect that much of the discrepancy has to do with submaximal efforts on the part of the patient and radiology technician during the attainment of the static flexion-extension views. It is not uncommon to have to repeat flexion-extension views with additional instructions to maximally flex and extend the neck in my practice. Therefore, I believe that this is a minor and unproven “advantage” of the dynamic studies over the static ones.
In conclusion, the authors identified subtle changes at the levels adjacent to a fusion at seven months postoperatively when compared with the findings in asymptomatic controls who had not been operated on. Patients without preoperative neck pain and the same group of patients evaluated postoperatively may have served as better control and experimental groups. Whether these subtle changes progress or have clinical relevance in accelerating adjacent-segment pathology remains to be elucidated.