Insight into the historic contributions made to modern-day spine surgery provides context for understanding the monumental accomplishments comprising current techniques, technology, and clinical success. Only during the last century did surgical growth occur in the treatment of spinal disorders. With that growth came a renaissance of innovation, particularly with the evolution of spinal instrumentation and fixation techniques. In this article, the authors capture some of the key milestones that have led to the field of spine surgery today, with an emphasis on the historical advances related to instrumentation, navigation, minimally invasive surgery, robotics, and neurosurgical training.
JNSPG 75th Anniversary Invited Review Article
Corey T. Walker, U. Kumar Kakarla, Steve W. Chang, and Volker K. H. Sonntag
Michael A. Mooney, Scott Brigeman, Michael A. Bohl, Elias D. Simon, John P. Sheehy, Steve W. Chang, and Robert F. Spetzler
Overlapping surgery is a controversial subject in medicine today; however, few studies have examined the outcomes of this practice. The authors analyzed outcomes of patients with acutely ruptured saccular aneurysms who were treated with microsurgical clipping in a prospectively collected database from the Barrow Ruptured Aneurysm Trial. Acute and long-term outcomes for overlapping versus nonoverlapping cases were compared.
During the study period, 241 patients with ruptured saccular aneurysms underwent microsurgical clipping. Patients were separated into overlapping (n = 123) and nonoverlapping (n = 118) groups based on surgical start/stop times. Outcomes at discharge and at 6 months, 1 year, 3 years, and 6 years after surgery were analyzed.
Patient variables (e.g., age, smoking status, cardiovascular history, Hunt and Hess grade, Fisher grade, and aneurysm size) were similar between the 2 groups. Aneurysm locations were similar, with the exception of the overlapping group having more posterior circulation aneurysms (18/123 [15%]) than the nonoverlapping group (8/118 [7%]) (p = 0.0495). Confirmed aneurysm obliteration at discharge was significantly higher for the overlapping group (109/119 [91.6%]) than for the nonoverlapping group (95/116 [81.9%]) (p = 0.03). Hospital length of stay, discharge location, and proportions of patients with a modified Rankin Scale (mRS) score > 2 at discharge and up to 6 years postoperatively were similar. The mean and median mRS, Glasgow Outcome Scale, Mini–Mental State Examination, National Institutes of Health Stroke Scale, and Barthel Index scores at all time points were not statistically different between the groups.
Compared with nonoverlapping surgery, overlapping surgery was not associated with worse outcomes for any variable at any time point, despite the complexity of the surgical management in this patient population. These findings should be considered during the discussion of future guidelines on the practice of overlapping surgery.
Marco T. Reis, Phillip M. Reyes, BSE, Idris Altun, Anna G. U. S. Newcomb, Vaneet Singh, Steve W. Chang, Brian P. Kelly, and Neil R. Crawford
Lateral lumbar interbody fusion (LLIF) has emerged as a popular method for lumbar fusion. In this study the authors aimed to quantify the biomechanical stability of an interbody implant inserted using the LLIF approach with and without various supplemental fixation methods, including an interspinous plate (IP).
Seven human cadaveric L2–5 specimens were tested intact and in 6 instrumented conditions. The interbody implant was intended to be used with supplemental fixation. In this study, however, the interbody was also tested without supplemental fixation for a relative comparison of these conditions. The instrumented conditions were as follows: 1) interbody implant without supplemental fixation (LLIF construct); and interbody implant with supplemental fixation performed using 2) unilateral pedicle screws (UPS) and rod (LLIF + UPS construct); 3) bilateral pedicle screws (BPS) and rods (LLIF + BPS construct); 4) lateral screws and lateral plate (LP) (LLIF + LP construct); 5) interbody LP and IP (LLIF + LP + IP construct); and 6) IP (LLIF + IP construct). Nondestructive, nonconstraining torque (7.5 Nm maximum) induced flexion, extension, lateral bending, and axial rotation, whereas 3D specimen range of motion (ROM) was determined optoelectronically.
The LLIF construct reduced ROM by 67% in flexion, 52% in extension, 51% in lateral bending, and 44% in axial rotation relative to intact specimens (p < 0.001). Adding BPS to the LLIF construct caused ROM to decrease by 91% in flexion, 82% in extension and lateral bending, and 74% in axial rotation compared with intact specimens (p < 0.001), providing the greatest stability among the constructs. Adding UPS to the LLIF construct imparted approximately one-half the stability provided by LLIF + BPS constructs, demonstrating significantly smaller ROM than the LLIF construct in all directions (flexion, p = 0.037; extension, p < 0.001; lateral bending, p = 0.012) except axial rotation (p = 0.07). Compared with the LLIF construct, the LLIF + LP had a significant reduction in lateral bending (p = 0.012), a moderate reduction in axial rotation (p = 0.18), and almost no benefit to stability in flexion-extension (p = 0.86). The LLIF + LP + IP construct provided stability comparable to that of the LLIF + BPS. The LLIF + IP construct provided a significant decrease in ROM compared with that of the LLIF construct alone in flexion and extension (p = 0.002), but not in lateral bending (p = 0.80) and axial rotation (p = 0.24). No significant difference was seen in flexion, extension, or axial rotation between LLIF + BPS and LLIF + IP constructs.
The LLIF construct that was tested significantly decreased ROM in all directions of loading, which indicated a measure of inherent stability. The LP significantly improved the stability of the LLIF construct in lateral bending only. Adding an IP device to the LLIF construct significantly improves stability in sagittal plane rotation. The LLIF + LP + IP construct demonstrated stability comparable to that of the gold standard 360° fixation (LLIF + BPS).
Gregory J. Velat, Steve W. Chang, Adib A. Abla, Felipe C. Albuquerque, Cameron G. McDougall, and Robert F. Spetzler
Intramedullary, or glomus, spinal arteriovenous malformations (AVMs) are rare vascular lesions amenable to resection with or without adjuvant embolization. The authors retrospectively reviewed the senior author's (R.F.S.'s) surgical series of intramedullary spinal AVMs to evaluate clinical and radiographic outcomes.
Detailed chart and radiographic reviews were performed for all patients with intramedullary spinal AVMs who underwent surgical treatment between 1994 and 2011. Presenting and follow-up neurological examination results were obtained and graded using the modified Rankin Scale (mRS) and McCormick Scale. Surgical technique, outcomes, complications, and long-term angiographic studies were reviewed.
During the study period, 20 patients (10 males and 10 females) underwent resection of glomus spinal AVMs. The mean age at presentation was 30 ± 17 years (range 7–62 years). The location of the AVMs was as follows: cervical spine (n = 10), thoracic spine (n = 9), and cervicothoracic junction (n = 1). The most common presenting signs and symptoms included paresis or paralysis (65%), paresthesias (40%), and myelopathy (40%). Perioperative embolization was performed in the majority (60%) of patients. Pial AVM resection was performed in 17 cases (85%). Angiographically verified AVM obliteration was achieved in 15 patients (75%). At a mean follow-up duration of 45.4 ± 52.4 months (range 2–176 months), 14 patients (70%) remained functionally independent (mRS and McCormick Scale scores ≤ 2). One perioperative complication occurred, yielding a surgical morbidity rate of 5%. Three symptomatic spinal cord tetherings occurred at a mean of 5.7 years after AVM resection. No neurological decline was observed after endovascular and surgical interventions. No deaths occurred. Long-term angiographic follow-up data were available for 9 patients (40%) at a mean of 67.6 ± 60.3 months (range 5–176 months) following AVM resection. Durable AVM obliteration was documented in 5 (83%) of 6 patients.
Intramedullary AVMs may be safely resected with satisfactory clinical and angiographic results. The pial resection technique, which provides subtotal AVM nidus resection, effectively devascularized these lesions, as confirmed on postoperative angiography, without violating the spinal cord parenchyma, thereby potentially reducing iatrogenic injury.
S. Harrison Farber, David J. Mauler, Soumya Sagar, Mark A. Pacult, Corey T. Walker, Michael A. Bohl, Laura A. Snyder, Kristina M. Chapple, Volker K. H. Sonntag, Juan S. Uribe, Jay D. Turner, Steve W. Chang, and U. Kumar Kakarla
Anterior cervical discectomy and fusion (ACDF) is a common and robust procedure performed on the cervical spine. Literature on ACDF for 4 or more segments is sparse. Increasing the number of operative levels increases surgical complexity, tissue retraction, and risks of complications, particularly dysphagia. The overall risks of these complications and rates of dysphagia are not well studied for surgery on 4 or more segments. In this study, the authors evaluated their institution’s perioperative experience with 4- and 5-level ACDFs.
The authors retrospectively reviewed patients who underwent 4- or 5-level ACDF at their institution over a 6-year period (May 2013–May 2019). Patient demographics, perioperative complications, readmission rates, and swallowing outcomes were recorded. Outcomes were analyzed with a multivariate linear regression.
A total of 174 patients were included (167 had 4-level and 7 had 5-level ACDFs). The average age was 60.6 years, and 54.0% of patients (n = 94) were men. A corpectomy was performed in 12.6% of patients (n = 22). After surgery, 56.9% of patients (n = 99) experienced dysphagia. The percentage of patients with dysphagia decreased to 22.8% (37/162) at 30 days, 12.9% (17/132) at 90 days, and 6.3% (5/79) and 2.8% (1/36) at 1 and 2 years, respectively. Dysphagia was more likely at 90 days postoperatively in patients with gastroesophageal reflux (OR 4.4 [95% CI 1.5–12.8], p = 0.008), and the mean (± SD) lordosis change was greater in patients with dysphagia than those without at 90 days (19.8° ± 13.3° vs 9.1° ± 10.2°, p = 0.003). Dysphagia occurrence did not differ with operative implants, including graft and interbody type. The mean length of time to solid food intake was 2.4 ± 2.1 days. Patients treated with dexamethasone were more likely to achieve solid food intake prior to discharge (OR 4.0 [95% CI 1.5–10.6], p = 0.004). Postsurgery, 5.2% of patients (n = 9) required a feeding tube due to severe approach-related dysphagia. Other perioperative complication rates were uniformly low. Overall, 8.6% of patients (n = 15) returned to the emergency department within 30 days and 2.9% (n = 5) required readmission, whereas 1.1% (n = 2) required unplanned return to surgery within 30 days.
This is the largest series of patients undergoing 4- and 5-level ACDFs reported to date. This procedure was performed safely with minimal intraoperative complications. More than half of the patients experienced in-hospital dysphagia, which increased their overall length of stay, but dysphagia decreased over time.
Adib A. Abla, Andrew G. Shetter, Steve W. Chang, Scott D. Wait, David G. Brachman, Yu-Tze Ng, Harold L. Rekate, and John F. Kerrigan
The authors present outcomes obtained in patients who underwent Gamma Knife surgery (GKS) at 1 institution as part of a multimodal treatment of refractory epilepsy caused by hypothalamic hamartomas (HHs).
Between 2003 and 2010, 19 patients with HH underwent GKS. Eight patients had follow-up for less than 1 year, and 1 patient was lost to follow-up. The 10 remaining patients (mean age 15.1 years, range 5.7–29.3 years) had a mean follow-up of 43 months (range 18–81 months) and are the focus of this report. Five patients had undergone a total of 6 prior surgeries: 1 transcallosal resection of the HH, 2 endoscopic transventricular resections of the HH, 2 temporal lobectomies, and 1 arachnoid cyst evacuation. In an institutional review board–approved study, postoperative complications and long-term outcome measures were monitored prospectively with the use of a proprietary database. Seven patients harbored Delalande Type II lesions; the remainder harbored Type III or IV lesions. Seizure frequency ranged from 1–2 monthly to as many as 100 gelastic seizures daily. The mean lesion volume was 695 mm3 (range 169–3000 mm3, median 265 mm3). The mean/median dose directed to the 50% isodose line was 18 Gy (range 16–20 Gy). The mean maximum point dose to the optic chiasm was 7.5 Gy (range 5–10 Gy). Three patients underwent additional resection 14.5, 21, and 32 months after GKS.
Of the 10 patients included in this study, 6 are seizure free (2 after they underwent additional surgery), 1 has a 50%–90% reduction in seizure frequency, 2 have a 50% reduction in seizure frequency, and 1 has observed no change in seizure frequency. Overall quality of life, based on data obtained from follow-up telephone conversations and/or surveys, improved in 9 patients and was due to improvements in seizure control (9 patients), short-term memory loss (3 patients), and behavioral symptoms (5 patients); in 1 patient, quality of life remains minimally affected. Incidences of morbidity were all temporary and included poikilothermia (1 patient), increased depression (1 patient), weight gain/increased appetite (2 patients), and anxiety (1 patient) after GKS.
Of the approximately 150 patients at Barrow Neurological Institute who have undergone treatment for HH, the authors have reserved GKS for treatment of small HHs located distal from radiosensitive structures in patients with high cognitive function and a stable clinical picture, which allows time for the effects of radiosurgery to occur without further deterioration. The lack of significant morbidity and the clinical outcomes achieved in this study demonstrated a low risk of GKS for HH with results comparable to those of previous series.