Minimally Invasive Surgical Techniques in Adult Degenerative Spinal Deformity: A Systematic Review

Background Minimally invasive surgery (MIS) approaches have the potential to reduce procedure-related morbidity when compared with traditional approaches. However, the magnitude of radiographic correction and degree of clinical improvement with MIS techniques for adult spinal deformity remain undefined. Question/purposes In this systematic review, we sought to determine whether MIS approaches to adult spinal deformity correction (1) improve pain and function; (2) reliably correct deformity and result in fusion; and (3) are safe with respect to surgical and medical complications. Methods A systematic review of PubMed and Medline databases was performed for published articles from 1950 to August 2013. A total of 1053 papers were identified. Thirteen papers were selected based on prespecified criteria, including a total of 262 patients. Studies with limited short-term followup (mean, 12.1 months; range, 1.5–39 months) were included to capture early complications. All of the papers included in the review constituted Level IV evidence. Patient age ranged from 20 to 86 years with a mean of 65.8 years. Inclusion and exclusion criteria were variable, but all required at minimum a diagnosis of adult degenerative scoliosis. Results Four studies demonstrated improvement in leg/back visual analog scale, three demonstrated improvement in the Oswestry Disability Index, one demonstrated improvement in treatment intensity scale, and one improvement in SF-36. Reported fusion rates ranged from 71.4% to 100% 1 year postoperatively, but only two of 13 papers relied consistently on CT scan to assess fusion, and, interestingly, only four of 10 studies reporting radiographic results on deformity correction found the procedures effective in correcting deformity. There were 115 complications reported among the 258 patients (46%), including 37 neurological complications (14%). Conclusions The literature on these techniques is scanty; only two of the 13 studies that met inclusion criteria were considered high quality; CT scans were not generally used to evaluate fusion, deformity correction was inconsistent, and complication rates were high. Future directions for analysis must include comparative trials, longer-term followup, and consistent use of CT scans to assess for fusion to determine the role of MIS techniques for adult spinal deformity.


Introduction
Minimally invasive (MIS) spine surgery has recently been at the forefront of innovations in spine surgery [4,6,9,10,13,15,16,30,31]. MIS spine surgery not only implies one performed through a smaller incision, but also an approach that seeks to reduce approach-related morbidity associated with traditional open spine surgery.
Traditional open surgical correction of adult degenerative scoliosis can be associated with perioperative risk and a prolonged recovery period [8,12,17,19,20,29,40]. Perioperative morbidity is compounded by the complexity of patients with adult degenerative scoliosis as well as patient-specific comorbidities [11,21,26,27,32]. MIS techniques, in comparison to open traditional surgery, may reduce approach-related and overall morbidity and so are attractive both to patients and surgeons [1,18,31,33,38]. However, the magnitude of radiographic correction and degree of clinical improvement with MIS techniques for adult spinal deformity remain ill defined. Moreover, minimally invasive techniques have a learning curve and pose their own set of unique challenges, technical limitations, and complications [2,13,14,41].
We therefore sought to systematically review the available literature on MIS approaches for adult spinal deformity, specifically to determine whether they (1) improve pain and function; (2) reliably correct deformity and result in fusion; and (3) are safe with respect to surgical and medical complications.

Literature Review
A systematic PubMed and Medline database search was performed for published articles related to MIS techniques addressing adult degenerative scoliosis. MIS herein denotes an alternative surgical technique with the specific aim of reduced iatrogenic tissue damage incurred during the exposure process as compared with conventional open surgical approaches, thereby seeking to reduce perioperative morbidity. The search was limited to clinical studies in the English language from 1950 to August 2013 with the following key terms: ''minimally invasive'', ''surgery'', ''adult'', ''spine'', ''deformity'', and ''scoliosis''. The search criteria are detailed subsequently: The searches yielded 147, 527, 77, and 302 papers, respectively (Fig. 1). Titles and abstracts generated by the search were subsequently reviewed and manuscripts were excluded from full-text review according to the following exclusion criteria: anatomical descriptions, case reports, commentaries, literature reviews, and studies addressing congenital or adolescent idiopathic scoliosis. Given the extremely limited data on MIS treatment of adult degenerative scoliosis, minimum followup was not one of the inclusion/exclusion criteria. We included papers with short-term followup to capture early complications.
Subsequently, the quality of the selected studies was judged by using the methodological index for nonrandomized studies (MINORS). The studies were further evaluated by the coauthors using the following criteria: completeness of operative data and perioperative complications, use of 36-inch standing radiographs, application of the current understanding of sagittal balance and spinopelvic parameters, length of followup C 12 months (again, we used this as a criterion for evaluating study quality; studies with shorter followup were included to be able to identify early complications), reporting of a minimum of two health-related quality-of-life (HRQoL) questionnaires (eg, visual analog scale [VAS] and Oswestry Disability Index [ODI]), and statistical analysis for significance of radiographic and HRQoL outcome data.
Included studies were reviewed by each of the four authors. Disagreements were adjudicated by consensus.
The initial search yielded 1053 results. After title and abstract review, 22 papers were selected for full-text review. Of these, 13 were selected for inclusion in the review with a total of 262 patients. All studies were Level IV evidence: 11 retrospective case series, one prospective case series, and one retrospective cohort. MINORS [36] scores ranged from 5 to 10 for the case series and 13 for the study that included a control (Table 1). There were no prospective controlled or randomized trials.

Patient Selection
Patient age ranged from 20 to 86 years with a mean of 65.8 years (Table 2). Inclusion and exclusion criteria were variable between studies. All required at minimum a diagnosis of adult degenerative scoliosis and some level of disability with one study also including a small number of patients with idiopathic scoliosis and iatrogenic scoliosis [5]. Only one study described specific indications for MIS rather than open treatments of their patients: those who had relative contraindications to a posterior approach, that is, previous surgery, age older than 65 years, and presence of comorbidities were treated with an MIS approach [26]. One group of authors attempted to delineate a schema for MIS modality selection based primarily on severity of sagittal plane deformity [15].

Discussion
With an ever-increasing array of surgical options, minimally invasive spine surgery techniques are altering treatment paradigms for a variety of spine disorders, including the complex field of adult degenerative scoliosis. In making the decision to treat a patient with a minimally * The MINORS scoring system includes additional items for studies with control arms. This was the only study to include a control arm. For the sake of comparison with the other studies, the score for the first eight items was included followed by the total score for all items; MINORS = methodological index for nonrandomized studies.   invasive rather than traditional open technique, the surgeon must be aware of both the advantages and limitations of the selected surgery. This systematic review found the literature on MIS approaches to adult spinal deformity to be both scant and rather preliminary. In general, followup was short, comparator groups absent, and basic methodological approaches such as the use of CT scans to assess fusion and full-length radiographs to assess deformity correction to have been inconsistently used. Although scores for pain and function generally improved, radiographic improvements were more modest and complications relatively frequent. Because, with the exception of one very small study, there were no comparisons made in these studies with traditional approaches, it is difficult to say whether  these approaches delivered on the promise of safer or comparably effective surgery. This study had a number of limitations. First, as noted, control groups were absent. This makes it impossible to determine whether the improvements observed are better than, worse than, or comparable to more traditional approaches. No comparisons could be made that control for levels of spine operated on or the number or combination of modalities used. Second, followup in these studies was very short, and this limits our ability to comment on either the durability of the corrections observed or the reoperation frequencies with any confidence. Only one study had greater than 2 years mean followup [4]. To establish MIS techniques as a useful tool in optimizing the surgical strategy and patient outcomes, a greater number of clinical studies demonstrating durable results is needed. Third, without CT scans, it is difficult to know for sure whether fusion has been achieved in these patients; only two of 13 studies used CT scans exclusively for the assessment of fusion. Fourth, this review is limited in its capacity to articulate the true incidence of complications. In analyzing complication rates, all complications were added together and then divided by the total number of patients. This type of analysis assumes a one-to-one correspondence of complications to patients not accounting for the possibility of multiple complications in a single patient, thus likely representing an overestimate of actual complication percentages; this may be offset by the fact that the followup was very short, and some complications (such as chronic and subacute complications such as pseudarthrosis or junctional disease [7,44]) and reoperations will therefore not be well represented in the studies we surveyed. One particular study noted an alarmingly high complication rate of 87.5%. We believe that this rate stemmed from a technique-specific complication. The authors of this paper used hyperlordotic cages without release of the anterior longitudinal ligament. Thus, six of the eight patient cohort experienced anterior endplate damage, and one further patient developed asymptomatic cage subsidence.
The specific MIS techniques described in the studies were generally separated into anterior column support and posterior instrumentation ( Table 6). The approach to the anterior column, discectomy, and interbody fusion was accomplished using some variant of a minimally invasive far lateral approach. Examples include extreme lateral interbody fusion (Fig. 2) (XLIF 1 ; Nuvasive, San Diego, CA, USA) and direct lateral interbody fusion (DLIF 1 ; Medtronic). In addition to an MIS lateral interbody fusion, three studies added the use of hyperlordotic cages to increase correction of sagittal deformity and restore lumbar lordosis with or without the use of an anterior longitudinal ligament release (Fig. 3) [15,16,30]. As a result of anatomical constraints, the lumbosacral junction is not accessible through a far lateral approach necessitating the use of an alternate technique. Three MIS techniques were used either exclusively or alternately in multiple studies for interbody fusion at the lumbosacral junction and in certain cases at L4-5. These include AxialLIF 1 , MIS transforaminal lumbar   interbody fusion (TLIF), and anterior lumbar interbody fusion. With the exception of one study that used lateral plates [13], instrumentation exclusively took the form of posterior percutaneous pedicle screw fixation. Beginning with Wang in 2012 [42], percutaneous screw fixation to the ilium demonstrated use in select cases. Of note, several studies included cases in which patients were treated with standalone interbody fusions without additional instrumentation [4,9,15,25,30].

Clinical Outcomes
From the studies we reviewed, it appears the VAS scores for leg/back pain are emerging as the primary tools for assessing clinical outcomes. Unfortunately, from the variability in reporting, it was not possible to combine all patient data points into a single analysis. With the exception of Anand et al. [4], the studies in the review are limited both in terms of cohort size and mean length of followup. Higher powered studies with adequate control groups will be needed to fully elucidate the use of MIS techniques in improving patient outcomes.

Radiographic Results: Deformity Correction and Fusion
Standing full-length radiographs are critical in the assessment of sagittal balance and spinopelvic parameters in a patient with adult degenerative scoliosis [3], but only half of the studies in the review obtained them. Of the studies  that did, only Deukmedjian et al. [15] included a complete preoperative and postoperative measurement of all parameters of sagittal balance and spinopelvic harmony. Tormenti et al. was the only study to show a loss of mean lumbar lordosis; however, three patients in the cohort had hyperlordosis ([ 60°) and one loss of lordosis [39]. Three studies included in their radiographic analysis reduction of apical vertebral translation [5,10,39]; however, to our knowledge, no evidence has been published in the literature demonstrating a correlation between apical vertebral translation and clinical outcome. Anand et al. did analyze sagittal balance but did not include any preoperative or postoperative analysis of spinopelvic parameters [5].
Restoration of sagittal balance and spinopelvic harmony has been demonstrated to be a strong predictor of functional outcome and represents a critical component in planning and evaluating deformity correction surgery [3,23,24,28,35,37]. We hope that with more widespread awareness of the importance of spinopelvic parameters and availability of diagnostic technologies, there will be future studies analyzing global spinal balance. Furthermore, only two of the included 13 studies consistently used CT scans for assessment of fusion [10,15]. CT scans provide exquisite bony detail compared with dynamic radiographs and have higher sensitivity for detecting early hardware failure. We hope that in future studies, use of CT scans for assessment of fusion will become more widespread. Use of allograft has increased concomitantly with MIS techniques to enhance fusion rates. Ten of the 13 studies in the review used some form of biologic agent to enhance fusion rate [4-6, 9, 10, 13, 16, 25, 39, 43]. A recent systematic review and meta-analysis by Rodgers et al. using previously unavailable internal data from the manufacturer demonstrated enhanced fusion rates in patients undergoing anterior fusion supplemented with rhBMP-2 compared with iliac crest bone graft [34]. However, using the same data, Fu et al. found that fusion rates for rhBMP-2 were comparable [22]. Both studies did find an increased risk of genitourinary complications with the use of rhBMP-2 [22,34]. At present, indications for the use of rhBMP-2 versus bone graft remain to be clarified [22,34].

Complications
Given both the limited cohort size and novelty of some of the techniques, some potential complications may not have been encountered. For example, advanced techniques such as anterior longitudinal ligament release carry an identifiable risk of great vessel injury [16], although none were reported. Safety of the neural structures in deformity surgery is of paramount importance. At first glance, a 14.3% neurologic complication rate appears high. However, the majority of these are transient and related to the lateral approach. A thorough working knowledge of lumbar plexus anatomy and use of continuous electromyographic neuromonitoring should serve to minimize the incidence of these approach-related complications. One of the aims of MIS treatment is a reduction in perioperative complications in a population predisposed to comorbidity. Only Isaacs et al. reported 28.3% of patients having at least one comorbidity before surgery [25].

Conclusions
At present, there exists a glaring paucity of studies investigating MIS treatment of adult degenerative scoliosis that meets the standards used for evaluating traditional deformity surgery. Work remains to be done in producing more robust studies with longer followup to determine durability of correction, subsidence rates, and improvement of quality of life. To make concrete claims about the efficacy of MIS treatment of deformity, studies with control groups treated with traditional deformity surgery are necessary. Consistent use of CT scans for assessment of fusion is needed because this is the main purpose of these surgical procedures. In addition, further study is needed to delineate the role of advanced techniques such as anterior longitudinal ligament release and use of hyperlordotic cages. Finally, given that adult degenerative scoliosis affects predominantly elderly patients, more data with larger cohorts fitting this demographic are needed to assess if MIS techniques reduce the incidence of age-related complications in patients undergoing spine surgery.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited. Preoperative sagittal balance, ODI, and VAS back/leg pain were 116 mm, 50%, and 4.7, respectively. The patient was treated with the T10-L5 MIS lateral approach, L5-S1 ALIF, T12L1, L23, L34 anterior longitudinal ligament release with a hyperlordotic cage placed at those levels, and T10-S1 percutaneous pedicle screws. Two-year followup sagittal balance, ODI, and VAS were 34 mm, 2%, and 2, respectively.