INTRODUCTION
Paraspinal musculature acts as both the power generator and stabilizing force for spinal movement. In the presence of adolescent idiopathic scoliosis (AIS), the pathologic curvatures of the vertebral column change the symmetry, power dynamics, and tissue composition of the paraspinal musculature [1-3].
As with any traditional open approach, extensive iatrogenic damage of the soft tissues and paraspinal musculature occurs during long, single incision deformity correction surgeries. Scar tissue formation, ischemia and denervation caused by dissection and prolonged retraction has been associated with degenerative changes in paraspinal muscular adipose compositions, resulting in loss of contractile strength and overall instability [4,5]. Minimally invasive surgical correction of AIS offers a potentially less morbid alternative to the traditional open approach. Three smaller incisions and tube dilators minimize dissection and wide retraction in an attempt to preserve the paraspinal musculature. These minimally invasive surgery (MIS) techniques can significantly decrease hospital length of stay and were associated with faster recovery of pain and Oswestry Disability Index (ODI) scores while still providing the same degree of correction as traditional, open techniques [6-9].
Serum creatine kinase (CK) and aldolase concentrations increase rapidly after any significant muscle injury, including spinal surgery [10-12]. At the initial phase of tissue injury, these proinflammatory cytokines trigger an inflammatory response [13,14]. The degree of increase of a proinflammatory cytokine is proportional to the severity of tissue injury and is counterbalanced by an equally extensive compensatory anti-inflammatory reaction [13-16].
Several past studies have compared open versus MIS spinal fusion techniques with respect to length of surgical incisions, blood loss or postoperative CK and aldolase levels [17-21]. Kim et al. [20] investigated the tissue injury quantitatively after mini-open, single level lumbar fusion in 20 patients with lumbar spinal stenosis. They concluded that mini-open lumbar fusion may reduce morbidity related to muscle injury, as both serum CK and aldolase were significantly higher in patients that underwent the traditional open approach during the acute postoperative period. Regev et al. [21] assessed the extent of soft tissue dissection and retraction during the MIS and open posterior spinal surgical exposures. They concluded that mini-open posterior lumbar interbody fusion (PLIF) was less disruptive of the multifidus muscle (MF) than open PLIF. This study was limited by a small sample size (n=20) and did not measure the MF at the fusion level because of the interference by artifacts produced by the metal screws and rods. Few quantitative analyses have assessed the changes laboratory and radiographic parameters of the paraspinal muscles in the 2 techniques (MIS vs. open). Most of these studies are retrospective in nature with a smaller sample size which limits the levels of evidence that they present and hence generalizability of their findings.
This prospective study aims to quantitatively compare the impact on the paraspinal musculature and soft tissues between MIS and open cohorts through analysis of tissue injury markers (serum CK, aldolase levels) and muscle atrophy within the paraspinal musculature at 1-year postoperative follow-up. Paraspinal muscle atrophy was evaluated by paraspinal muscle area (mm2) using axial T2 magnetic resonance imaging (MRI) at the T6–7 levels. These levels (T6–7) were chosen for the MRI muscle measurement as these were instrumented for every patient in our cohort. Our null hypothesis was that the MIS technique may be associated with significant reductions in tissue injury reactions postoperatively compared to traditional open techniques.
MATERIALS AND METHODS
The University of Texas Austin Institutional Review Board (2018-01-0155) approval was obtained for this prospective study conducted from 1st, January 2015 to 30th, November 2020. When a patient fulfilled the study criteria, the clinician provided an explanation of the study and obtained consent from the patient. Patients were assigned to either open or MIS cohort based on the surgical techniques (MIS vs. open). The surgeon typically offered both MIS and open surgical approaches to patients. For those with a flexible curve under supine traction x-ray, classified as Lenke type 1 with instrumentation initiating at T3, MIS was recommended. Conversely, the traditional open approach was suggested for patients with a rigid spine and planned instrumentation initiating at T1–2 spinal levels. The determination of what level to initiate instrumentation was influenced by factors such as the extent of spinal curvature, flexibility of the deformity, and specific surgical goals for each patient. The surgeon carefully considered these factors to determine the most appropriate level to initiate instrumentation for each case, aiming to achieve optimal correction and stabilization while minimizing risks and complications. Despite surgeon’s recommendation, ultimately the choice of approach, whether MIS or open, was determined by the patient's preference.
The inclusion criteria for the study were: (1) AIS patients undergoing spinal deformity corrections (≥3 spinal levels treated) between 1st, January 2015 to 30th, November 2020 with the capacity to provide informed consent; (2) age at surgery between 10 and 18 years; (3) the lowest instrumented level equal to L2. Exclusion criteria included: (1) patients less than 10 years old or older than 18 years old; (2) posterior fixation extending beyond L2; (3) patients who did not undergo spinal deformity surgical corrections; (4) presence of chronic medical conditions e.g., endocrine, renal, cardiac or autoimmune disease, neuromuscular scoliosis and congenital scoliosis; (5) current or recent history of malignancy or infectious disease; (6) previous spine surgery at the same level within 6 months; (7) any contraindications for MRI; and 8) the inability to provide informed consent.
1. Surgical Techniques
All surgical procedures were performed by a single surgeon at a university hospital using a uniform surgical approach and fixation strategy. In each case, total intravenous anesthesia was administered, and the mean arterial pressure was kept above 65 mmHg. Tranexamic acid was administered for bleeding prophylaxis with a loading bolus of 15 mg/kg over the first hour followed by a continuous infusion of 1 mg/kg/hr for the remainder of the case.
For patients undergoing the traditional open approach, a single midline incision was employed. A combination of Cobb dissection and electrocautery were used to achieve a subperiosteal separation of the paravertebral muscles to the level of the transverse processes. Freehand pedicle screws insertion technique with intraoperative fluoroscopy and conventional radiography was used as the chief fixation strategy with occasional, supplementary interlaminar hooks. In open surgical procedures, the spinal curvature often exhibits greater rigidity compared to MIS. When correcting this rigid curvature, there is a potential for exerting increased force, which can result in heightened soft tissue damage. The primary challenge is to balance achieving the desired correction with minimizing collateral damage to surrounding tissues. Based on our clinical experience, we recommend that surgeons meticulously evaluate each case to determine the optimal level of correction that maximizes outcomes while minimizing risks to the patient's soft tissues.
The MIS technique started with 3 incisions, each spanning approximately 2 to 3 levels, extending into the fascia. Each incision was then undermined with bilateral fascial incisions just lateral to the MFs and then dissected down to the lateral spine at each level. Bilateral submuscular tunnels were created via the Wiltse approach, exposing lateral lamina and facets. Pedicle screws were used as the chief fixation strategy, with rods inserted through either the superior or inferior incision and fed along the construct. The rods are then locked in place with set screws once final alignment is achieved.
All 44 patients, whether utilizing the open or MIS technique underwent posterior column osteotomies (PCOs) with the mean number of PCO levels as 6. These were performed sequentially at different spinal levels by resecting the inferior/superior facets, facet capsule, and part of the lateral ligamentum flavum, as described by Smith-Peterson. The MIS approach spared the midline ligament from resection and PCO technique was carious out via burr and Kerrison, resecting the inferior and superior facets in line and anterior joint capsule. Intracanal bleeding was controlled with Gelfoam and pledgets. PCOs were done at the periapical levels of the deformity corresponding to the levels around the planned vertical column resection.
Meticulous hemostasis was sought at each stage of surgery using a combination of mono- and bipolar electrocautery, bone wax, thrombin soaked cottonoids and hemostatic matrix. A drain was placed deep to the fascia. Cell saver was used for most (95%) of the open cases and about 50% of the MIS cases. Transfusions of allogeneic blood in our cohort were guided by a clinical protocol designed to keep the hematocrit above 25.
Age, sex, body mass index (BMI), operative time (skin-skin time), length of hospital stays, number of spinal levels treated, estimated blood loss (EBL), number of transfusions and complications were recorded. Cobb angle, sagittal balance, proximal curve (pelvic tilt), lumbar lordosis (LL), pelvic incidence (PI), PI–LL mismatch, visual analogue scale (VAS), ODI, and scoliosis research society-22r (SRS-22r) were collected preoperative (preop), 6 months, 12 months, and 24 months postoperatively (postop).
Tissue injury was represented by serum enzyme levels of CK and aldolase. Peripheral blood samples were collected from all patients preop on the day of the surgery, and postop on day one, 2nd, 3rd, 4th, discharge, day 14–21 postop (3-week visit), 3 month and 6 months after surgery.
Paraspinal muscle area (mm2) using axial T2 MRI at the T6–7 disc level of patients who underwent MIS and open deformity correction for AIS was measured. Preoperative MRI was obtained for each patient which was compared with postoperative MRI performed at 12 months postoperatively. A total of 17 patients were included in the MIS group and 16 patients in the Open group for the MRI measurements. A measurement tool included in the PACS software (Synapse ver. 5.0, FUJIFILM Medical Systems U.S.A. Inc., Stamford, CT, USA) was used to obtain vertical and horizontal diameter of left and right paraspinal muscles and an area of an ellipse was calculated. Examples of the measurements are included in Figures 1 and 2 of MIS and Open group, respectively. We calculated absolute values of left and right side as well as mean for preoperative and postoperative MRI and a difference between the 2 was also calculated. Student t-test was performed using the mean differences between preoperative and postoperative measurements of paraspinal muscle on MRI comparing MIS and Open group as well the difference found when normalized by bodyweight. MRI measurements were performed following a standardized MRI protocol and similar level among all patients. This was done to minimize inaccuracies from metal artifact as well as reduce any measurement bias. The specific levels where the MIS groups were exposed were not recorded.
Characteristics of subjects were analyzed using means and standard deviations calculations. Student t-tests were used to compare the means of continuous variables between the 2 groups. The chi-square contingency table was used to compare the dichotomous values between the groups. Two different multiple regressions were modeled with surgery types (mini- vs. conventional open technique) as independent variables to analyze the effect of these variables on enzyme concentrations. CK and aldolase concentrations were considered separately as dependent variables for the 2 models. Test between subjects analysis of variance were used to compare the differences in SRS-22r, ODI, and VAS scores between the 2 groups. The level of significance was set at α=0.05. All the statistical analyses were performed using SPSS ver. 15.0 (SPSS Inc., Chicago, IL, USA).
2. Power Analysis
Based on our initial power analysis (with at least 80% power, effect size of 50% and normal values of aldolase ~1 – 7.5 U/L and CK ~22 – 198 U/L) we needed 20–25 patients per group. To offset any sample size effects on the study due to loss to follow-ups or subject voluntary withdrawal, we kept enrollment open for an additional 10 subjects (5 open and 5 treatment groups). However, there was no loss to follow up in the study and, subsequently, no need to enroll additional patients.
RESULTS
Over the date range, 44 patients met the inclusion criteria. Mean patient age for the MIS and Open groups were 16.1 years (±1.2 years) and 15.3 years (±2.6 years), respectively. 95.2% were female in the MIS group and 82.6% in the Open group. Both the groups had similar BMI (MIS vs. open: 21.2±3.2 vs. 23.2±4.1). Table 1 depicts the basic demographics and clinical data of our cohort. The majority (91.2%) of the patients in the MIS group were Lenke type 1 while the Open group had 52.2% of the cases as Lenke type 1, followed by 39.1% as Lenke type 2. The mean flexibility of the index curve was similar for the 2 groups (52.8% vs. 55.1%).
The mean number of spinal levels treated in MIS and Open groups were similar (8.7±0.72 vs. 9.2±1.1). There was a significant difference between the groups with respect to mean EBL (MIS vs. open: 142.1±102.4 mL vs. 397.1±281.2 mL, p<0.05). Mean cell saver did differ significantly between the 2 groups (68.1±51.2 mL vs. 213±158 mL, p<0.05). 95% of the patients in the Open group received cell saver while about 50% of the MIS cases received cell saver return. Mean operative time was significantly greater for the MIS cases as compared to their counterparts (344.2±31.8 minutes vs. 313±30.4 minutes, p<0.05). Length of stay was comparatively higher for the Open group as compared to MIS (3.4±0.83 days vs. 4.2±0.77 days, p>0.05). These values are depicted in Table 1.
The radiographic measurements for the 2 groups at different time points are shown in Table 2. The mean preop Cobb angle was similar between the 2 groups (53.1°±4.1° vs. 54.5°±3.6°). There was a significant % mean correction (preop vs. 24 months) in both the groups (MIS vs. open: 74.7% vs. 76.5%). Similarly, the mean sagittal balance at 24 months postop improved in both the groups as compared to the preop. Mean % PT correction (preop vs. 24 months) in both the groups improved significantly (55.5% vs. 72%, p<0.05). The mean PI–LL mismatch was similar between the 2 groups. Figures 3 and 4 show examples of standing posterior-anterior radiographs of one of MIS and open cases at preop, 1 and 2 years postop.
As compared to the baseline values, the mean patient-reported outcome (PRO) scores improved significantly in both the groups. The mean VAS scores (preop vs. 24 months) improved from 2.6±1.6 to 0.57±0.43 and 1.9±1.6 to 0.83±0.54 in the MIS and Open groups, respectively. Similarly, the ODI scores improved from 14.3±12.5 to 4.6±3.6 and 12.3±9.9 to 5.4±4.62 in the MIS and Open groups, respectively. In both MIS and Open groups, the SRS scores improved (preop vs. 24 months) from 3.7±0.79 to 4.6±0.52 and 3.6±0.57 to 4.4±0.58. Table 3 shows these outcome values for the 2 groups at different time points.
1. Aldolase
In both the groups, the mean aldolase values at postop day 1 were significantly higher than baseline values. The Open group at postop day 1 had significantly higher aldolase concentration value than MIS group (18.2±7.6 mU/mL vs. 12.9±4.2 mU/mL, p<0.05). At postop day 2, day3 and day 4, the Open group had comparatively higher values than their counterpart. The difference in mean aldolase values between the 2 groups decreased at different postoperative time points. Eventually the aldolase values returned to the baseline values in both the groups. Figure 5 shows the aldolase values for the 2 groups at different time points.
2. Creatine Kinase
The mean CK enzyme values increased significantly at postop day 1 in both groups as compared to the baseline preop values. At postop day 1, the mean CK value was significantly greater in the Open group than their counterpart (3,003.1±60.1 IU/L vs. 1,649.4±40.6 IU/L, p<0.05). The differences between the 2 groups remained significant on postop day 2, 3, and 4 (p<0.05). Thereafter, CK values returned to the preop values in both the groups. Figure 6 shows the CK values for the 2 groups at different time points.
3. Complications
In the MIS group, 1 patient had wound drainage 1-week postsurgery which resolved without any intervention. Another patient developed wound infection postop requiring incision and drainage and antibiotics. Two patients in the Open group developed durotomies which were repaired intraoperatively. No other complications were observed in either group.
4. Regression Analysis
Two separate regression analysis were conducted for the 2 groups with postop day 1, CK, and aldolase as dependent variables. We chose the postop day 1 enzyme values as the dependent variable because these values peaked on this day and there was a significant difference in the enzyme values between the MIS and Open groups. For the MIS group, the analysis showed that none of the independent variables had any significant effect on the outcome variables (CK and aldolase). Interestingly, for the Open group, the amount of cell saver transfusion and EBL had significant effect on postop day 1 aldolase levels (p=0.007). None of the variables had any significant effect on the CK levels.
5. MRI Findings
Table 4 shows the MRI measurements for MIS and Open group. The obtained values were subsequently normalized by the patient’s weight in kilograms that were measured at preoperative and postoperative clinic visits near the time of their MRI as represented in Table 5 MIS and Open group.
On average, the MIS group had a 5.1% loss of muscle mass in the postoperative MRI’s compared to preoperative MRIs. When normalized by bodyweight given the time difference between the 2 MRI’s, this difference was nearly 9%. On average, the Open group had a 10.3% loss of muscle mass in the postoperative MRI’s compared to preoperative MRIs. When normalized by bodyweight given the time difference between the 2 MRI’s, this difference was 14.1%. A Student t-test performed using mean differences of paraspinal muscle mass on MRI between MIS and Open group yielded a p-value of 0.27, and 0.46 for normalized by bodyweight group.
A lower level of muscle mass loss was seen in the MIS group compared to Open group overall. This could be due to less soft tissue disruption experienced by the MIS group. However, these differences were not statistically significant, possibly due to being underpowered.
DISCUSSION
Paraspinal muscle injuries during spinal surgery can lead to rapid increases in aldolase and CK levels [10-12]. Previous research has highlighted how MIS results in less MF muscular injury compared to open surgery [17-21]. However, many of those studies are retrospective analyses with small sample sizes, limiting the generalizability of the findings. Additionally, none of these studies examined whether MIS and open surgery result in differing levels of tissue trauma in patients with AIS.
In our prospective analysis, we found that patients with AIS who underwent MIS had decreased tissue markers and muscle atrophy when compared to those treated with open surgery. Although its magnitude decreased over time, the initial difference in tissue marker levels between the 2 surgical approaches supports our hypothesis that MIS is associated with less tissue trauma than open surgery in the treatment of AIS. However, despite the difference in CK and aldolase levels, postoperative ODI, VAS, and SRS-22r scores were similar between the MIS and open surgical cohorts at each follow-up interval. This indicates a lack of correlation between perioperative tissue trauma and postoperative symptomatic improvement.
These findings aligned well with the pre-existing literature. In a 2010 prospective study, Shunwu et al. [22] assessed whether minimally invasive transforaminal lumbar interbody fusion (TLIF) resulted in reduced postoperative morbidity compared to its open surgical counterpart. Similar to our study, the authors found a significantly decreased CK level in minimally invasive TLIF in the early postoperative period. However, this study only recorded CK levels at the 3- and 7-day postoperative intervals, thus restricting observation of the trend throughout the early postoperative course. Additionally, this study reported significant differences between the postoperative ODI and VAS scores between the MIS and open surgical cohorts [22]. Though, both differences were below the minimally clinically important difference values of 12.8 and 1.2 respectively, indicating a lack of clinical significance [23].
1. Differences in Perioperative Blood Loss, Operative Time, and Hospital Length of Stay Between MIS and Open Surgery
Our study found that MIS procedures resulted in increased operative time, decreased blood loss, and shorter length of stay, despite treating a similar number of spinal levels as open surgery. These results further support our hypothesis, as many studies have highlighted how decreased tissue trauma is associated with less EBL and faster recovery times, regardless of the increased operative time [20-22,24]. Additionally, MIS is associated with a steeper learning curve, therefore operative times can be expected to become more comparable with the open approach as surgeons gain more experience [4,20,25]. This is beneficial to patients with AIS, as they can be educated on the potential for MIS to result in less healthcare costs related to blood transfusions and hospital room fees.
A comparative analysis conducted by Tsutsumimoto et al. [25] in 2009 reported no significant difference in perioperative blood loss between minimally invasive PLIF and open PLIF, despite mini-open PLIF resulting in decreased CK levels. It is noteworthy that the authors only had 10 patients in the mini-PLIF group, who were the first to undergo the procedure. Therefore, the comparable blood loss between the MIS and open surgical cohorts in this study could be attributed to differences in experience with both procedures, as noted by the authors.
Minimally invasive techniques generally involve smaller incisions and less extensive dissection, leading to reduced intraoperative blood loss. This can be particularly beneficial for patients with comorbidities that increase the risk of complications from blood transfusions. Phan et al. [26] performed a systematic review and meta-analysis, concluding that MIS techniques are associated with significantly less blood loss compared to open surgery. This reduction in blood loss also correlated with a lower need for blood transfusions. Giannadakis et al. [27] reported similar findings in their study of minimally invasive lumbar fusion, where patients had less intraoperative blood loss and a lower incidence of transfusion-related complications.
Reduced tissue damage translates to faster recovery times, which often results in shorter hospital stays. This is a crucial factor in reducing healthcare costs and minimizing the risk of hospital-acquired infections. Parker et al. [28] conducted a comparative effectiveness and cost utility analysis, showing that patients undergoing MIS-TLIF had shorter hospital stays by approximately 2 days compared to those undergoing open TLIF. The reduced length of stay was attributed to quicker mobilization and less postoperative pain. Smith et al. [29] also reported shorter hospitalization periods for patients who underwent minimally invasive lumbar fusion surgeries, highlighting the efficiency of MIS in promoting faster recovery. The reduced postoperative pain and faster recovery times associated with MIS enable patients to return to their daily activities and work sooner. This is a significant benefit in terms of both quality of life and economic productivity.
Terman et al. [30] demonstrated that patients who underwent minimally invasive lumbar spine surgery returned to work and normal activities faster than those who had open surgery. The study emphasized the importance of minimally invasive techniques in facilitating quicker functional recovery. Shin et al. [31] found that the time to return to daily activities was significantly shorter for patients who underwent minimally invasive procedures, with many returning to their preoperative activity levels within weeks.
2. Limitations
There are some limitations of this study: (1) The patients were not randomized rather patients chose the surgical option after discussion with the surgeon. Despite this both the groups had similar radiographic findings. (2) Both surgical approaches resulted in complications. In the MIS group, infectious complications may have led to elevated tissue inflammatory markers unrelated to perioperative tissue trauma. In the open surgical group, repair of cerebrospinal fluid leaks may have caused additional tissue trauma beyond what is typically seen in uncomplicated correction of AIS. Therefore, both complications may have caused tissue trauma markers to be artificially elevated at different times during the postoperative recovery period. (3) Metal artifact can pose a challenge during measurement and impact accuracy. In our study, standardized MRI protocol was used for both patient groups. In addition, a similar level was used for all patients with measurement along similar identified fascial planes to minimize impact from metal artifact. (4) Tissue injury markers were not matched up with PROs. This decision was primarily due to our standard clinical practice. We verified early PRO data, including VAS and ODI scores at 3 weeks and 3 months postoperatively. However, no significant differences were observed between the MIS and open surgical groups in these measures. We routinely collect SRS-22r scores preoperatively and at 6, 12, and 24 months postoperatively. Additionally, we examined postoperative morphine milligram equivalent dosages between the 2 groups at various time points, finding similar dosages. (5) The study does not report long-term follow-up results of MIS for AIS. Further long-term studies, ideally with 5 to 10 years of follow-up, are needed to determine whether the initial benefits of MIS translate into durable outcomes in terms of curve correction, patient-reported outcomes, and muscle preservation.
Despite these limitations, this study contributes to the existing literature by providing stronger evidence compared to prior retrospective studies. We compared the MIS and open surgery techniques in AIS patients, using rigorous protocols like tissue injury markers (CK and aldolase), MRI-based assessments of paraspinal muscle atrophy and PROs. By focusing on AIS patients with varying levels of spinal flexibility, this study reflects real-world clinical decision-making, where the choice between MIS and open techniques is influenced by curve flexibility. Additionally, the comprehensive tracking of early postoperative markers and functional outcomes over a 2-year follow-up provides detailed temporal data. The study also offers a novel perspective by normalizing MRI-based muscle atrophy assessments by patient weight, a method not thoroughly explored in previous AIS correction studies. While earlier research highlighted differences in tissue injury between open and minimally invasive techniques, this study's prospective design and targeted approach based on curve flexibility present new and clinically relevant insights.
CONCLUSION
In patients with AIS, MIS is associated with decreased tissue injury markers and muscle atrophy than open surgical correction in in the early postoperative period. While this difference may be associated with the decreased blood loss and shorter hospital stays seen in MIS, it did not result in a significant difference in the symptomatic improvement of AIS when compared to open surgery. Though the final outcomes were similar, this early data may assist clinical decision and foster additional, larger, prospective studies addressing these questions. It seems, assuming similar efficacy and safety, that, a priori, using the least invasive approach is rational. Additionally, it appears MIS scoliosis correction patients may benefit from decreased surgical morbidity through reduced blood loss, shorter lengths of stay and a more rapid recovery. However, it is important to acknowledge that our study focused on symptomatic improvement starting at the 6-month interval. Therefore, further research needs to be done to explore the effect of differing levels of perioperative tissue trauma on early postoperative symptomatic improvement in patients with AIS.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
