My name is Zoe Gogowalla. It is a distinct honor and privilege to present the early results from the Cervical Spondylotic Myelopathy Surgical Trial, Quality of Life, Complications, and Return to Work at the 2020 Annual Meeting of the American Association of Neurological Surgeons. CSMS was funded with federal grants from NIH and PCORI. Hospital cost analysis was funded by the Allen and Jackie Stewart Spine Research Center. I also own intellectual property and shares in a company called NIDIS, which has nothing to do with the CSMS trial. No randomized controlled trial in surgery or at any arena of medicine is performed by one individual. I am greatly indebted to the following individuals, Ed Benzel, Karen Freund, Rob Whitmore, Sandy Schwartz, Melissa Dunbar, Fred Barker, Bob Heary, Norma Taron, and Bill Butler for their enormous support and guidance throughout the process of performing the CSM trial. Background and rationale. Cervical Spondylotic Myelopathy represents one of the common conditions that neurosurgeons treat all over the world. The condition represents the most common cause of spinal cord dysfunction worldwide. Surgery for cervical myelopathy is associated with significant complication rates ranging from 15-18 percent of cases depending upon different trial results. Hospital charges exceed $2 billion a year in the United States. It's estimated that approximately 40,000 surgical cases are done per year in the United States. Yet the optimal approach for treating cervical myelopathy has not been established. In Asia and in Europe, cervical laminoplasty is commonly performed. Whereas in North America, ventral and dorsal fusion surgery is commonly performed with less utilization of laminoplasty for the treatment of cervical myelopathy. The CSMS randomized control trial aims to compare ventral versus dorsal surgery in a randomized fashion, but is weighted 2-3 to allow for a greater number of patients in the dorsal cohort to allow for a comparison between dorsal laminectomy and fusion versus laminoplasty. Our study population was carefully defined. Inclusion criteria were patients aged 45-80 years with two or more levels of cervical stenosis. Patients were excluded if they had segmental kyphotic deformity, kyphosis greater than five degrees in extension, structurally significant OPLL, previous cervical spine surgery, or ASA score greater than three. We published the clinical protocol for the cervical spondylotic myelopathy surgical trial in neurosurgery in 2014. In this published protocol, we describe a technique by which we reach clinical equipoise in each case. Each patient in the CSMS trial was reviewed by 15 cervical spine expert surgeons. Equipoise was defined as met if greater than 50 percent of those who reviewed the case voted for randomization, that is ventral versus dorsal surgery, and if less than 80 percent of experts voted for the same approach, ventral fusion versus dorsal fusion versus laminoplasty. The study design was a randomized control trial comparing ventral and dorsal surgery, but with a two-to-three weighting, giving more patients in the dorsal cohort to allow a pre-specified planned comparison of laminoplasty versus fusion strategies. In the ventral cohort, multi-level anterior cervical discectomy and or corpectomy was performed with fusion. In the dorsal cohort, laminoplasty or laminectomy with lateral mass fixation and fusion was performed. Our biostatisticians examined our pilot study data and with a 90 percent power, calculated a sample size of 159 patients to have enough power to detect a difference between ventral and dorsal surgery if one exists. The plan was to enroll patients from 15 major centers, NIH-funded study protocol development and PCORI-funded actual trial study execution. Our study aims, primary outcome was the one-year change in SF36, Physical Component Summary score, aiming for a minimal clinically important difference of five points. Secondary outcomes included validated patient-reported outcome tools, SF36 Physical Component Summary, Neck Disability Index, EuroQOL 5D, and modified JOA scores assessed at 3, 6, and 12 months after surgery. We also assessed complications, productivity and return to work, as well as outpatient health resource utilization, including radiology, physical therapy, and opioid utilization. Results of the cervical spondylotic myelopathy surgical trial are now presented. In terms of our study execution, we screened 458 patients, enrolled 269, and randomized 163 or 61 percent of enrolled patients. We had a 95 percent follow-up at one year. There were three percent or five patients who crossed over, resulting ultimately in 66 patients treated with ventral decompression infusion, 69 patients treated with dorsal decompression infusion, and 28 patients treated with laminoplasty. Baseline characteristics were comparable across all three surgical strategies. Age was 62 years on average. Male gender was roughly equal to female gender. The baseline SF36, modified JOA, EQ5D, and neck disability index scores were all comparable amongst all three treatment arms. Most importantly, the number of levels of cervical stenosis was comparable across all three treatment arms with an average of 2.8 levels of stenosis treated in the CSMS trial. Primary outcome measure was the one-year SF36 physical component summary score. As randomized, there was no difference between ventral versus dorsal surgery. However, when examined as treated, patients with laminoplasty had superior SF36 physical component summary scores at one year compared with ventral or dorsal fusion. Laminoplasty was associated with fewer complications, ventral fusion associated with a 42 percent complication rate, dorsal fusion with a 27.5 percent complication rate, and laminoplasty with an 11 percent complication rate. There were no significant differences in major complication rates amongst the three groups, and the majority of complications seen in the ventral fusion group were minor problems with dysphagia, which resolved within three months of surgery. What about neck pain after surgery? We observed that patients treated with dorsal fusion had more neck disability index scores after three months. However, all three groups were comparable at one year. Of note, laminoplasty had the lowest neck disability index scores at all time intervals after surgery. Return to work data were comparable between the three surgical strategies. However, return to work rate at three months was worse for patients treated with dorsal fusion. Health resource utilization was assessed in the outpatient setting. Patients treated with dorsal laminoplasty had more x-ray utilization, more ongoing physical therapy, and more opioid utilization relative to the fusion alternatives. Hospital charges were also less for patients treated with laminoplasty compared with ventral or dorsal fusion. There are limitations to the CSMS clinical trial. The promising results observed for laminoplasty patients should be interpreted with caution. Because patients treated with laminoplasty were selected by the treating surgeon amongst patients randomized to the dorsal arm of the trial with a potential for selection bias. A future randomized control trial would be important to validate the observations made in CSMS surgical trial. Our conclusions, CSMS is the first randomized control trial comparing ventral and dorsal surgery for CSM. Ventral and dorsal surgery showed no difference in primary outcome at one year. As treated in a non-randomized but pre-specified secondary analysis, laminoplasty was associated with superior health-related quality of life, as measured by the SF-36 physical component summary score, fewer complications, and less outpatient health resource utilization at one year. Thank you for the privilege of presenting this paper at the WNS annual meeting. Thank you.

Cervical Spondylotic Myelopathy Surgical Trial

ventral surgery

dorsal surgery

SF36 physical component summary score

laminoplasty