Introduction
Degenerative spondylolisthesis (DS) involves anterior subluxation (anterolisthesis) of the superior vertebra relative to an adjacent vertebra. It is caused by intersegmental instability at the lumbar motion segment because of the degeneration of the facet joint complex and disc degeneration [1,2]. Lumbar DS is a common cause of spinal canal stenosis and contributes to radiculopathy and neurogenic claudication in the adult population, particularly in patients aged >50 years [2–4]. As the population ages and the incidence of DS increases, the rate of chronic disability due to low back pain increases concomitantly, contributing to the steady growth of years lost to disability [5,6]. Studies comparing surgical and nonsurgical treatments for DS have demonstrated positive clinical outcomes and functional improvement in patients who underwent surgery [7–11]. However, the optimal surgical technique for symptomatic DS, specifically comparing decompression surgery with or without fusion, remains controversial.
DS accounts for the largest proportion of lumbar instrumented fusion procedures, and the rates of fusion surgeries have increased exponentially in the last decade [12–15]. Despite this, no consensus has been reached regarding the necessity of instrumentation and fusion in addition to standard decompression in treating symptomatic lumbar DS [16,17]. Instrumented fusion with decompression is often performed to address vertebral instability and is therefore a common treatment strategy lumbar DS [18–21]. However, fusion surgery is associated with additional risks and is more complex than decompression alone [16,22,23]. Instrumentation raises concerns regarding cost-effectiveness and implant-related complications [24,25]. In addition, recent research has not established the decompression and fusion surgery over decompression surgery alone, as measured using the Oswestry Disability Index and the 6-minute walk test at 2- and 5-year follow-up periods [16,26]. A recent systematic review and meta-analysis suggested that fusion does not add value for most patients with lumbar DS; however, these studies did not consider disc height, facet angle, and extent of slippage [16–19].
Given the lack of consensus regarding the preferred treatment modality, this study aimed to investigate the differences in the 30-day perioperative morbidity and mortality rates, as well as postoperative complication rates, among patients undergoing decompression alone or decompression with fusion for lumbar DS. This study used national data from the American College of Surgeons National Quality Improvement Program (ACS-NSQIP) database from 2015 to 2020. We hypothesize that between the two surgical treatment groups, the cohort undergoing laminectomy or laminotomy with fusion (LWF) will have an increased risk of perioperative complications.
Materials and Methods
Ethics statement
Informed consent was not required for this study because it utilizes anonymized data from the ACS-NSQIP national database. Institutional review board approval was not necessary for the completion of this study.
Study design
This retrospective cohort study used the ACS-NSQIP database, which is a medical chart-informed data registry on the outcomes and complications of surgery that aims to assist surgeons and hospitals in providing better data-driven patient care. NSQIP provides data 300 days postoperation.
Patient selection and data collection
Patient data were obtained from the database from 2015 to 2020. The participant use data files were merged using IBM SPSS Statistics for Windows ver. 28.0 (IBM Corp., Armonk, NY, USA), and variable coding was standardized to ensure the accuracy of the statistical results. The inclusion criteria were as follows: adults aged ≥65 years who presented with spondylolisthesis of the lumbar or lumbosacral region (the 10th revision of the International Classification of Diseases [ICD10]: M43.16 and M43.17) and had undergone surgery including laminectomy and LWF or laminotomy without fusion (LWOF) (CPT codes 63047, 63030, and 22612) (Table 1). For this study, all fusion cases involved instrumented fusion using pedicle screws and included posterior and posterolateral approaches.
For comparative analysis, patients were stratified into two cohorts: (1) Laminectomy or LWOF: 63047 (laminectomy, facetectomy and foraminotomy [unilateral or bilateral with decompression of the spinal cord, cauda equine, and/or nerve root(s), single vertebral segment]) or 63030 (laminotomy [hemilaminectomy], with decompression of the nerve root(s), including partial facetectomy, foraminotomy, and/or excision of a herniated intervertebral disc). (2) Laminectomy or LWF: 63047+22612 (laminectomy, facetectomy and foraminotomy [unilateral or bilateral with decompression of the spinal cord, cauda equina and/or nerve root(s), single vertebral segment AND posterior, posterolateral or lateral transverse process technique arthrodesis procedures on the spine (vertebral column)]) OR 63030+22612 (laminotomy [hemilaminectomy], with decompression of nerve root(s), including partial facetectomy, foraminotomy and/or excision of herniated intervertebral disc AND posterior, posterolateral or lateral transverse process technique arthrodesis procedures on the spine [vertebral column]).
Independent variables
Patient demographics such as age and sex were extracted directly from the NSQIP. Comorbidities included diabetes, hypertension, congestive heart failure, chronic obstructive pulmonary disease, smoking status, steroid use, American Society of Anesthesiologist (ASA) physical status classification, and preoperative functional status. Diagnoses of these conditions were defined according to the predefined ICD cohorts within the NSQIP database. These variables were extracted directly from NSQIP except for the ASA classification, which was extracted as a numerical variable to allow for statistical analysis, and the preoperative functional health status, which was recoded into a binary variable classified as either “dependent” or “independent.”
Postoperative outcome and complication variables
NSQIP collects outcome (dependent) variables up to 30 days postoperatively. The outcome variables extracted for this study included estimated mortality, estimated morbidity, reintubation, reoperation, readmission, operation time (minutes), and length of hospital stay (LOS). The variables for postoperative complications included deep wound infection, pneumonia, deep vein thrombosis, pulmonary embolism, urinary tract infection, myocardial infection, bleeding requiring transfusion, stroke, sepsis, renal failure, cardiac arrest, superficial infection, and renal insufficiency.
Statistical analysis
IBM SPSS Statistics for Windows version 28.0 (IBM Corp., Armonk, NY, USA) was used for all statistical analyses. Patient demographics, preoperative comorbidities, and perioperative characteristics were explored for their association with postoperative complications. A comparative analysis between the two cohorts was conducted using the chi-square test and analysis of variance (ANOVA) for categorical and continuous variables, respectively. In addition, the analysis of outcomes between the two groups was stratified by patient preoperative ASA physical status classification system score, ranging from 1 to 4. ASA 1 indicates the patient is in good health; 2, the patient has a mild but well-managed or treated condition; 3, the patient has a serious condition that affects overall health; and 4, presence of a severe and life-threatening condition [27]. By organizing the statistical output by ASA groups, ANOVA with Tukey’s post hoc test and chi-square analyses were conducted with the significance established at p<0.05. Patients with incomplete data for any variable were not included in the analysis.
Results
Patient demographics
A total of 4,120 patients met the outlined inclusion criteria. Of these, 2,384 (58%) underwent LWOF and 1,736 (42%) underwent LWF. The mean age of the LWOF group was 72.4±5.4 years and that of the LWF group was 73.0±5.6 years (p<0.001). Male patients accounted for 38.5% and 35.8% in the LWOF and LWF groups, respectively (p=0.085). No significant difference was found in the incidence of the medical comorbidities recorded between the two groups. However, hypertension was most prevalent comorbidity, which was present in 1,700 patients (71.3%) in the LWOF group and 1,248 (71.9%) in the LWF group. In addition, the preoperative functional status of the patients and the mean ASA score were not significant between the two groups. The characteristics of the two cohorts are presented in Table 2.
Multivariate analysis
A significant relationship was found between the treatment group and rates of several postoperative complications (Table 3). Higher mean mortality, mean morbidity, mean length of stay, and bleeding risk necessitating transfusion were found to be associated with LWF as opposed to LWOF. All variables for each complication subgroup are presented in Table 3.
Multivariate ANOVA
In the ASA 1 subcohort, 64.0% of the patients underwent LWF, whereas in the ASA 2, 3, and 4 subcohorts, 41.6%, 42.0%, and 46.8% underwent LWF, respectively. An analysis of the postoperative complications between the two treatment groups stratified by the ASA score demonstrated an overall higher complication rate occurring in those with higher scores, with an increasing complication rate in accordance with the ASA score (Table 4). No complications occurred in the ASA 1 subcohort. In the ASA 2 subcohort, 120 (12.6%) complications were reported in the LWOF group, whereas 100 (14.7%) were reported in the LWF group. In the ASA 3 subcohort, 235 (17.2%) complications were reported in the LWOF group, whereas 204 (20.6%) were reported in the LWF group. In the ASA 4 subcohort, 19 (38.0%) complications were reported in the LWOF group, whereas 13 (29.5) were reported in the LWF group.
For the ASA 2 and ASA 3 subcohorts, a significant increase was noted in the mean mortality (p<0.001 for ASA 2 and 3) and morbidity (p<0.001 for ASA 2 and 3) in the LWF group compared with that in the LWOF group. When assessing individual complications between the two groups, the average LOS (p<0.001) and rates of pulmonary embolism (p=0.028) were significantly higher in the LWF group when stratified into the ASA 2 subcohort. The bleeding risk was significantly higher for the LWF group in the ASA 3 subcohort (p=0.027).
Discussion
Decompression with instrumented fusion is accepted as a surgical treatment for lumbar DS with mechanical low back pain that results in lumbar spinal stenosis and instability. However, decompression alone may result in noninferior outcomes for lumbar DS, as recent evidence has not strongly supported an advantage to fusion [16,17,26,28,29]. This study investigated the differences in the 30-day perioperative morbidity and mortality rates as well as the postoperative complication rates in patients undergoing decompression with or without fusion for DS.
This retrospective review of 4,120 patients demonstrated significant differences in the complication variables between the two cohorts, with the LWF group bearing a significantly higher complication risk profile including higher rates of mean mortality, morbidity, mean LOS, bleeding necessitating transfusion, and pulmonary embolism. When stratified by the ASA score, a linear increase was noted in postoperative complication rates with higher ASA scores in both groups. In addition, in the ASA 2–3 subcohorts, the postoperative complication rate was higher in patients undergoing decompression with fusion versus decompression alone. This corresponded to a significant increase in the mean morbidity for decompression with fusion versus decompression alone in the ASA 2–3 subcohorts; however, the mean morbidity was not significantly different between the two surgical approaches in the ASA 4 subcohort despite decompression alone having a greater percentage of postoperative complications. This trend was likely due to a substratification of patient operative risk within the ASA 4 class cohort itself that was amplified with the relatively small subcohort. These data, along with this further understanding, can be used to guide prognostic indications for patients with degenerative lumbar spondylolisthesis in the immediate 30-day postoperative period and thus improve patient selection for surgery.
A potential concern about performing decompression alone for lumbar DS is iatrogenic instability necessitating revision surgery. In a standard laminectomy, the midline ligamentous structures are detached, and the lamina, hypertrophic ligamentum flavum, and facet that contribute to lumbar stenosis are resected. Performing a wide laminectomy for full decompression may contribute to the development of instability after surgery. Reoperation rates after decompression without fusion range from 28% to 34% [17,26]. Conversely, the SLIP trial (Spinal Laminectomy versus Instrumented Pedicle Screw trial) demonstrated a lower rate of reoperation over 4 years in patients who underwent fusion rather than decompression alone (14% versus 34%, p=0.05) [17]. Despite being a compelling evidence, multiple trials have demonstrated a lack of substantial benefit from the addition of fusion to decompression for lumbar DS. The consequences of vertebral subluxation in this setting are not yet elucidated, as the natural history of untreated DS has not been shown to result in subluxation progression or clinical symptoms in some studies. In addition, Försth et al. [16] argue that reoperation is more likely selected when surgical options remain, such as fusion, as is the case when decompression without fusion is performed as the index procedure.
These data must be evaluated considering the limitations in the study design. As a retrospective review of a national database, this study is inherently at risk of selection bias. The period of evaluation for revisions and postoperative complications is only 30 days postoperatively; thus, this study was unable to evaluate long-term sequelae, such as revision surgery for iatrogenic instability in decompressions alone, nonunion, or adjacent segment disease for decompression and fusion surgeries. Second, the NSQIP database does not permit distinguishing between mobile and nonmobile lumbar degenerative spondylolisthesis, which is an inherent data limitation. In addition, an association between several postoperative complications and surgical technique in patients with lumbar DS does not indicate an absolute criterion determining which technique is better suited for each patient, as the decision ultimately depends on the clinical picture and imaging findings. This study demonstrates that patient-specific factors, including ASA, and weighing early perioperative complications in laminectomy and fusions against the potential for future instability with decompression alone should all be considered in the surgical decision-making process.
Conclusions
To our knowledge, this study is one of the largest comparing decompressions versus decompression with fusion in patients with lumbar DS who are indicated for surgery. The results indicate that decompression with fusion may lead to worse perioperative outcomes. However, surgeon discretion should be used on a case-by-case basis in patients with lumbar DS to balance the risk of a potentially higher need for reoperation, which has been previously demonstrated in multiple trials in patients undergoing decompression alone. Our findings also indicate that a higher ASA score is predictive of early postoperative complications. Thus, the ASA can be leveraged as a prognostic tool to decrease the risk of short-term postoperative complications for patients undergoing decompression and fusion in lumbar DS.
Key Points
The choice of decompression with fusion or decompression alone for the management of degenerative spondylolisthesis (DS) is controversial.
Decompression with fusion is commonly performed because of vertebral instability; however, fusion is more complex and has additional risks than decompression alone—particularly in cases of nonmobile spondylolisthesis.
This retrospective study presents that adding fusion to decompression may lead to worse 30-day perioperative outcomes, including mean length of stay, pulmonary embolism, and bleeding risk necessitating transfusion.
A higher American Society of Anesthesiologist (ASA) score is predictive of early postoperative complications following lumbar spine surgery for degenerative spondylolisthesis; therefore, ASA can be leveraged as a prognostic tool to decrease the risk of short-term postoperative complications.
Decompression alone is an appealing treatment option for lumbar DS, particularly for patients with higher ASA scores and those at higher risk.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
