The Future of Minimally Invasive Spine Surgery

The field of minimally invasive spine surgery (MIS) has enjoyed a staggering rate of advancements in techniques and technologies over the last 15 years. Even though innovation in this domain has not reached its apogee, the rate of change is likely slowing. Therefore, in considering the future of MIS, it seems most appropriate to honestly assess prior developments (MIS evolution), what techniques in MIS appear to be most effective (MIS present), and what efforts hold the most promise for the future (MIS future). MIS DEFINITIONS MIS has become a fundamental part of spinal surgical practice for a variety of reasons, including the promise of decreased morbidity and faster recovery times. A combination of better surgeon training and education and the wide availability of the necessary operative equipment has made most MIS techniques readily accessible. Furthermore, there has been an increased demand from patients seeking MIS treatment for their spinal diseases. This demand has been driven, in part, by advertising and product vendors. New technologies emerge each year, including recent devices that offer percutaneous approaches for central lumbar stenosis,1 foraminal stenosis,2 and sacroiliac joint dysfunction.3 However, this onslaught of new devices and techniques can produce as much confusion as enthusiasm because the evidence base for their use lags behind their introduction. The resultant poor understanding of the effectiveness and clinical role for various “MIS” procedures has even led to a threat to payment for services. Indeed, the BlueCross/BlueShield subsidiary Anthem has had a long-standing negative coverage decision against percutaneous and endoscopic spinal surgery as being “investigational and not medically necessary.”4 In 2009, United Healthcare declined to cover MIS direct lateral interbody fusion techniques but chose to cover “laser spine surgery.”5 All this serves to underscore the need for better clarity in the classification of MIS procedures. In response to these pressures, the American Association of Neurological Surgeons/Congress of Neurological Surgeons Section on Disorders of the Spine and Peripheral Nerves in 2011 elaborated a framework for better understanding of MIS procedures. Specifically, we set forth procedural definitions based primarily on the method of access and modality of visualization. This framework seeks to differentiate those procedures that achieve the same anatomic surgical objectives as existing open procedures from uniquely novel approaches. Therefore, the techniques are broken down into 3 broad categories: percutaneous, endoscopic, and minimal access. Percutaneous procedures are those that gain access via a percutaneous needle or cannula and for which the only mode of visualization is fluoroscopy. Endoscopic procedures also gain access via a percutaneous cannula, use solely the endoscopic camera for visualization, and require specialized instruments. Minimal-access techniques, on the other hand, rely on a tissue-sparing approach with tubular dilators and retractors that allow visualization of the anatomic structures with the naked eye with or without the use of adjunctive magnification (eg, loupes or microscope). Minimal access also permits the use of either conventional or specialized surgical instruments. It is this final category of procedures that have the most surgical relevance and can therefore be best compared with and equated to their typical open counterparts. The remainder of this article focuses largely on the minimal-access category of techniques when referring to MIS. MIS EVOLUTION The primary goal of MIS surgery is to reduce approach-related tissue injury and complications and thereby reduce postoperative pain, blood loss, and recovery time while still achieving the same clinical objectives.6 Preclinical, histological, serological, radiological, and clinical outcome data have shown the often profound iatrogenic tissue injury associated with typical open posterior spinal approaches.7 Further evidence has emerged demonstrating significant reductions in muscle injury, pain, and disability with MIS techniques compared with open techniques for lumbar fusion.8 It was with these tenets in mind that Foley and Smith9 began the modern era of MIS with the description of tubular microendoscopic discectomy in 1997. MIS techniques have since been routinely applied to the full spectrum of degenerative spinal conditions, including herniated disks of all regions of the spine,10-12 spinal stenosis in all regions of the spine,13,14 lumbar spondylolisthesis and instability,15 and even now spinal deformity, particularly with the advent of direct lateral transpsoas techniques.16,17 The direct lateral approaches have dramatically altered the morbidity associated with anterior approaches to the thoracic and lumbar spine and are therefore likely to be a critical tool in the evolution of MIS using more advanced techniques such as MIS direct lateral corpectomy (Figure 1). Finally, we have also demonstrated the safety and efficacy of the use of MIS approaches in the treatment of intradural spinal pathology, including neoplasms (eg, meningioma, schwannoma), syringomyelia, and arteriovenous malformations (Figure 2).18-21FIGURE 1: A, sagittal reconstruction of lumbar spine computed tomographic scan showing pathological fracture of L3 resulting from multiple myeloma. B, intraoperative photograph showing exposure through the minimally invasive surgery (MIS) retractor via a direct lateral transpsoas approach after complete L2-3 and L3-4 diskectomies and before completion of L3 corpectomy. C, intraoperative photograph as in B after placement of an expandable intervertebral cage. D, postoperative lateral lumbar spine radiograph demonstrating reconstruction after MIS L3 corpectomy and percutaneous pedicle screw instrumentation.FIGURE 2: A, intraoperative photograph under microscopic magnification of intradural type I dural arteriovenous fistula seen through minimally invasive surgery tubular retractor after hemilaminectomy. White arrow indicates the venous side of the fistulous connection at the root sleeve (rostral is to the left). B, intraoperative indocyanine green angiography (ICGA) of same field in A, with the white arrow indicating the fistula. C, ICGA after clip application to the fistula showing normalization of blood flow to the intradural venous system. D, photograph after ligation of the venous side of the fistula.One adjunctive technology that has greatly affected MIS is intraoperative image-guided navigation. Image-guided navigation should theoretically address the dual concerns of the potential for greater hardware malpositioning resulting from reduced intraoperative visualization of anatomic structures in MIS and intraoperative radiation exposures resulting from heavy reliance on fluoroscopy to perform MIS procedures. Early implementation of image-guided navigation into spine surgery was cumbersome and prone to error. However, the advent of intraoperative cone-based computed tomography, combined with modern navigation systems, now permits real-time 3-dimensional imaging and offers both significant reductions in occupational radiation exposures and increased accuracy, particularly for percutaneous spinal instrumentation (Figure 3).22 In fact, 2 recent meta-analyses have demonstrated the dominance of image-guided over conventional screw insertion for placement accuracy.23,24 Image-guided navigation clearly has an important role in the future of MIS, particularly as technological advancements make it more convenient, adaptable, and cost-effective.FIGURE 3: A, intraoperative photograph showing cone beam-based computed tomography unit in place around the patient ready for acquisition of images for intraoperative navigation. B, image-guided screwdriver with the screw attached. C and D, intraoperative screen captures from navigation computer system showing real-time 3-dimensional imaging during use of image-guided tap (C) and screwdriver (D).Another related development that is showing some early promise is MIS robotic spine surgery. Both cadaveric and clinical retrospective studies have suggested improvements in accuracy, operative time, and radiation exposure with the use of robot-assisted pedicle screw placement compared with traditional techniques.25-27 However, a recent prospective, randomized controlled trial (RCT) comparing robot-assisted screw insertion with conventional freehand screw insertion counterintuitively found greater accuracy with the freehand technique.28 Specific technical aspects of the robotic system studied may have contributed to this result, but an apt conclusion may be that robot-assisted spine surgery is still an application in its infancy as a result of sluggish innovation in instrumentation development and a lack of clear indications of superiority over current standard or image-guided techniques. Robot-assisted surgery may hold promise for “intracavitary” procedures, including thoracoscopic removal of paraspinal neoplasms29 or transnasal/transoral endoscopic approaches to the upper cervical spine.30 It is worth noting that the most minimally invasive spinal procedure available currently is stereotactic body radiotherapy. This technique of high-dose single or hypofractionated focused radiation treatment has begun to fundamentally alter our approach to spinal neoplastic disease. Stereotactic body radiotherapy relies on image guidance with real-time anatomic correlations in the treatment suite that result in treatment accuracy of 1 mm.31 Remarkable 85% to 90% overall local control rates up to 5 years for metastatic disease in particular have been reported for both primary and adjuvant spinal stereotactic body radiotherapy.32,33 Stereotactic body radiotherapy has evolved as the perfect intersection of radiobiological and surgical MIS principles and will be a cornerstone of spinal oncological care for the future. MIS PRESENT As we consider the future of MIS, we must continuously assess the techniques and practices that we currently use and that emerge newly on the scene. This requires an honest accounting of the evidence base as we decide which procedures ultimately show clinically superior outcomes and cost-effectiveness. Surgeons must avoid the “siren song” of applying MIS techniques solely to satisfy market demand or patient expectations. Rather, the decision to use MIS must be based on a solid foundation of clinical evidence that details the advantages and disadvantages of MIS compared with nonoperative and/or conventional open treatments and that relies on some form of comparative effectiveness research, an area of spinal surgery research that is still lacking.34 Similarly, in our eagerness to advance the technological aspects of spinal surgery, physician-investigators have come under heavy scrutiny for the influence industry relationships have on the reporting of scientific results. Just as there is a clear need for continued collaboration between surgeons and industry to foster advancements that benefit patients, there is also an absolute need for honest publication of experiences with new devices and technologies. A recent publication relating a single surgeon’s experience with the percutaneous MiLD (minimally invasive lumbar decompression) device for lumbar stenosis (Vertos Medical, Aliso Viejo, California) described a high failure rate of 60% of the treated patients requiring reoperation for persistent stenosis.1 The article generated a vigorous response from the manufacturer as noted in the lay press, in which “Vertos accused the principle investigator of scientific misconduct and violating ‘research ethics’ by…independently deciding to follow his patients for added time without seeking agreement from Vertos.”35 New technologies require close working relationships between doctors and industry, but the future of MIS certainly should not include attacks against clinician-researchers who attempt to honestly report their results, and the hope is that the example cited here is not a harbinger of things to come. As we seek to bolster the evidence base in MIS, it is important to ask biologically sound and clinically useful questions. Even when prodigious efforts are made to generate high-level evidence from RCTs, the conclusions drawn from the data may or may not be plausible or applicable to general spinal surgery practice.36 Perhaps the most compelling example of this problem is in the RCTs performed to examine MIS vs open microdiscectomy.37-42 As seen in the Table, the 6 RCTs performed to date (5 lumbar, 1 cervical) lack consensus in their findings for a host of reasons. Ultimately, the true benefits of MIS techniques are unlikely to be realized in this procedure that already has low complication rates and morbidity with rapid recovery rates. Thus, we find heterogeneous outcomes from underpowered studies with variable applications of the techniques in variably skilled surgeons’ hands. Discovering the true effect size for MIS requires an understanding of the roles played by surgeon experience, patient-specific factors (eg, obesity, comorbidities), and the type and magnitude of the procedure performed. Only then can appropriate studies be designed to address the most compelling clinical concerns.TABLE: Randomized Controlled Trials Comparing MIS and Open MicrodiskectomyaThere are some MIS procedures for which a significant amount of clinical data has accumulated to give clinicians an idea of the important role for these techniques. In the surgical treatment of lumbar spinal stenosis, for example, not only has unilateral laminotomy/internal laminectomy for bilateral decompression14 emerged as the MIS procedure of choice, but a number of prospective and retrospective studies with mid- to long-term outcomes suggest that this approach has a superior overall clinical success rate compared with open laminectomy.43-48 This technique can have a somewhat demanding early learning curve but once mastered can be the optimal treatment for this common degenerative condition (Figure 4).FIGURE 4: A and B, preoperative axial (A) and sagittal (B) T2-weighted magnetic resonance images of typical L4-5 severe degenerative lumbar stenosis. C and D, postoperative images at the same locations as in A and B after a minimally invasive surgical unilateral approach for bilateral decompression.For the treatment of degenerative lumbar spondylolisthesis, instability, and severe spondylosis, MIS transforaminal interbody fusion (TLIF) is now frequently used to achieve decompression and arthrodesis with minimal-access techniques (Figure 5). Indeed, this “workhorse” procedure has been in common use for over a decade.15,49 A recent review of the literature comparing MIS and open TLIF revealed important differences between the 2 approaches.50 Estimated blood loss and length of stay were consistently lower in the MIS cohorts (282 vs 693 cm3 and 5.6 vs 8.1 days, respectively), whereas duration of surgery was similar between the 2 groups. Complication rates were also different, with a greater incidence of surgical site infections, urinary tract infections, and other unclassified complications in the open cohorts and a higher number of new transient neurological deficits, unintended durotomies, and hardware complications in the MIS cohorts. The latter fact points out the initial learning curve associated with MIS that tapers off over time with a lower incidence of these types of complications with greater surgeon experience. Despite this early learning curve, comparative effectiveness studies have convincingly shown that MIS TLIF is associated with reduced costs over a 2-year period while generating equivalent improvements in quality-adjusted life-years compared with open TLIF.51 The same group also found shorter length of stay, reduced narcotic dependence, and earlier return to work times for the MIS TLIF patients.FIGURE 5: A, lateral upright lumbar spine flexion radiograph demonstrating unstable L4-5 spondylolytic spondylolisthesis. B, postoperative lateral upright lumbar spine radiograph after minimally invasive surgery transforaminal interbody fusion and percutaneous pedicle screws demonstrating a reduction of spondylolisthesis.It appears that the reduction in overall perioperative complications is one of the primary factors driving the observed dominance of MIS over open techniques for procedures like TLIF. Our own work has demonstrated a postoperative surgical site infections rate of 0.22% after 1338 MIS procedures (including a 0.74% rate after MIS instrumented arthrodesis) compared with the 2% to 5% rate typically reported in the literature for open procedures.52 Similar findings were reported by Parker et al,53 who reviewed the literature on this topic and found a statistically significantly lower incidence of surgical site infections after MIS (0.6%) compared with open (4.0%) TLIF. Their cost analysis estimated a savings of almost $100 000 per 100 TLIFs performed as a result of this improvement in the surgical site infection rate alone. Furthermore, these reductions in complications and recovery times attributable to MIS have expanded the cohort of patients eligible for spinal surgery. The work of multiple surgeons now has shown that MIS procedures can be performed safely and effectively in patients > 75 to 80 years of age, a population that in many cases was deemed poor candidates for spinal surgery in the past.54,55 MIS FUTURE Despite the accumulated data that have begun to demonstrate specific techniques and clinical situations in which MIS may be preferable to open surgery, the overall evidence base has relied heavily on retrospective studies with a relatively small number of patients and surgeons involved. A greater emphasis on prospective and comparative studies is required going forward as we seek to firmly establish the role of MIS in spinal surgery in the future. However, study design in spinal surgery is highly complex, and as mentioned above, the application of the traditional research “gold standard” RCT has significant limitations.56 The inherent heterogeneity of the populations studied, the pathologies encountered, and the procedures used in spinal surgery make the strict confines of a typical RCT unrealistic and ultimately lacking in external validity. It is clear that to answer the key clinical questions facing us today, different methods are required—specifically those that use larger data sets generated from larger populations of patients under typical “real-world” clinical conditions.57 One such solution is to answer the clarion call of neurosurgical leadership by participating in national data registries such as the National Neurosurgery Quality and Outcomes Database. With the inclusion of specific procedural data coupled to both cost and validated clinical outcome measures, more robust and generalizable research conclusions can be drawn. In particular for MIS, effectiveness and cost utility can be determined, quality and performance standards can be set, and clinical practice guidelines can be elaborated, all of which can finally serve to define which procedures are best in which patients when performed by which practitioners. In summary, the future of MIS is clearly bright. On the basis of the foundation of early solid evidence, MIS has found mainstream use. Looking forward, our goals should include the judicious application of new technology; the refinement of indications based on procedure, patient, and surgeon; the implementation of comprehensive universal training in MIS techniques at both the residency and postgraduate levels; and finally the generation of large data sets of patient-centered and cost outcomes proving that MIS procedures not only are effective in their own right but in a great number of cases may be superior in outcome and cost to any other intervention. For related video content, please access the Supplemental Digital Content: http://www.youtube.com/watch?v=JX-WhPhfuuE Disclosure Dr O’Toole has been a consultant for Globus Medical, Inc and Pioneer Surgical and has received royalties from Globus Medical, Inc.