Background

Degenerative lumbar spondylolisthesis (DLS) refers to anterolisthesis of one vertebral body over another vertebral body secondary to osteoarthritic degeneration, leading to spinal canal stenosis [1]. DLS is an aging-related disease, and its incidence is increasing under the background of the global population aging [2]. Each year, 39 million individuals are diagnosed with DLS, accounting for a global prevalence of 0.53% [3]. DLS may be accompanied with low back pain, radiculopathy, or neurogenic claudication [2].

Surgeries have been regarded as the standard treatment modality for intractable cases [4]. The proportion of lumbar surgeries increases by more than two-fold not only because of elevated prevalence of degenerative lumbar spine disease but also because of improved surgical techniques, good outcomes, and increased hospitals and surgeons [5, 6]. However, due to complications (such as fusion failure, persistent pain, and infection), progressive degenerative changes-related diseases, or an unrelated previous surgeries, some patients require reoperation [7]. Despite improvements in surgical skills and techniques, the reoperation rate is still unimproved, with a 10-year reoperation rate of about 15% [7]. Given the high prevalence and chronicity of DLS, understanding the risk factors for reoperation is important [8]. Park et al. have revealed the longitudinal trends in the lumbar reoperation rate, and the reoperation was associated with demographics, comorbidities, primary surgery type, and preoperative spinal pathology [9]. Noh et al. haven found lifestyle-related factors, such as smoking, drinking, and exercise, were associated with the higher rate of reoperation [10]. However, results of studies on the risk factors for reoperation of DLS patients remain controversial. Rabah et al., have reported that diabetes was related to greater risk of reoperation [11], while Khan et al. reported no significant association between diabetes and reoperation [12]. In the study performed by Zhong et al., obesity was found to be associated with a higher incidence of unplanned reoperations [8]. Nevertheless, Kuo et al. found that obesity was not significantly associated with the reoperation [13].

Considering the controversial results, we aimed to perform a systematic review and meta-analysis to evaluate the incidence and risk factors of reoperation in DLS patients for the purpose of improving surgical outcomes and prognosis.

Methods

Literature search strategy

This study was performed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [14]. Two researchers (YZC and YZ) conducted the literature search from September 28, 2022 to October 28, 2022 in Pubmed, Embase, Cochrane Library, and Web of Science. Consensus was reached by discussion; if consensus cannot be reached by discussion, a third researcher (XHF) was consulted. Search terms were “degenerative spinal diseases” OR “degenerative spondylolisthesis” OR “degenerative lumbar spondylolisthesis” OR “degenerative cervical spondylolisthesis” AND “spinal surgery” OR “fusion” OR “reoperation” OR “repeat surgery” OR “risk factor”.

Selection criteria

Studies meeting the following inclusion criteria were selected: (1) population: DLS patients; (2) patients undergoing lumbar surgeries, including decompression surgeries and fusion surgeries; (3) outcome: reoperation rate and risk factors; and (4) studies: cohort studies. The population included DLS patients and DLS patients with lumbar spinal stenosis (LSS). Fusion surgeries included posterolateral lumbar fusion (PLF) and lumbar interbody fusion (LIF). Reoperation was defined as the secondary lumbar surgeries due to progression of lumbar degenerative changes or postoperative instability [8]. Risk factors included body mass index (BMI), sex, age, diabetes, smoking, and more bleeding.

Studies were excluded by meeting one of the following criteria: (1) animal studies; (2) other degenerative spinal diseases including lumbar spine stenosis, degenerative disc disease, and degenerative cervical spondylosis; (3) not English articles; (4) unable to extract data; (5) case reports, conference abstracts, letters, reviews, and meta-analysis.

Data extraction and critical appraisal

The following data were extracted: the first author, year of publication, country, study design, patients, definition of spondylolisthesis, sample size, age, sex, BMI, disease duration, surgery types, follow-up time, number of reoperations, reasons for reoperation, reoperation methods, and risk factors of reoperation. The Newcastle–Ottawa Scale (NOS) was applied to evaluate the quality of cohort studies [15]. This scale consisted of three items: selection, comparability, and outcome. This scale was scored a total of 9 points, and divided into low quality (0–3 points), fair quality (4–6 points), and high quality (7–9 points) [15].

Statistical analysis

All statistical analyses were performed using Stata15.1 software (Stata Corporation, College Station, TX, USA). Rate was used as the effect index in the analysis of reoperation rate. Odds ratio (OR) was used as the effect index for categorical data, and effect size was represented as 95% confidence interval (CI). Heterogeneity test was performed for each outcome effect size, and results were quantified as I-squared (I2). Random-effect model was used for analysis if I2 ≥ 50%, and fixed-effect model was used if I2 < 50%. For the high heterogeneity (I2 ≥ 50%), subgroup analysis was conducted based on study design, patients, surgery types, follow-up time, and quality of studies to explore the source of heterogeneity. Sensitivity analysis was carried out for all outcomes. Begg test was used to assess publication bias for the outcome included more than 10 articles. Trim-and-fill analysis was used to adjust the publication bias. P < 0.05 was considered statistically significant.

Results

Identification of studies and characteristics of patients

A total of 7,662 articles were searched from Pubmed (n = 1026), Embase (n = 1349), Web of Science (n = 4441), and Cochrane Library (n = 846). After removing the duplicates, 5,251 articles remained. Further, 5,150 articles were excluded due to publishing as reviews or meta-analyses (n = 929), conference abstracts (n = 310), animal trials (n = 22), case reports (n = 226), and letters (n = 9), not English articles (n = 112), and topic not meeting the requirements (n = 3542). In the remaining 101 articles, we further excluded 3 articles unable to extract data, 29 articles reporting other degenerative spinal diseases, and 30 articles with topic not meeting the requirements. Finally, 39 cohort studies were retained in this meta-analysis [7, 8, 10,11,12,13, 16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48], with 27 retrospective cohort studies and 12 prospective cohort studies. The flow diagram of our searching was displayed in Fig. 1. In the included studies, 28 studies were assessed as fair quality, and 11 studies were assessed as high quality. Characteristics of the included studies were presented in Table 1.

Fig. 1
figure 1

Flowchart of identifying studies

Full size image
Table 1 Characteristics of included patients
Full size table

Overall results and subgroup analysis results of reoperation rate

This meta-analysis showed a 10% (95%CI: 8%-12%) of reoperation rate in DLS patients (Fig. 2). A high heterogeneity was observed in the results (I2 = 99.3%). To explore the source of heterogeneity, subgroup analyses were performed. In study design subgroup, 10% of reoperation rate was found in both prospective cohort study (95%CI: 7%-13%) and retrospective cohort study (95%CI: 8%-13%). In patients subgroup, DLS patients showed 11% of reoperation rate (95%CI: 8%-13%) and DLS patients with LSS showed 10% of reoperation rate (95%CI: 6%-13%). In surgery types subgroup, the reoperation rate was 11% (95%CI: 9%-13%) in patients undergoing decompression, 10% (95%CI: 7%-12%) in patients undergoing fusion, 9% (95%CI: 5%-13%) in patients undergoing decompression and fusion, and 7% (95%CI: 3%-11%) in patients undergoing other surgeries. In follow-up time subgroup, the reoperation rate was 9% (95%CI: 6%-12%), 12% (95%CI: 9%-14%), and 10% (95%CI: 6%-15%) at follow-up time < 5 years, between 5 to 10 years, and ≥ 10 years, respectively. In study quality subgroup, there was 11% (95%CI: 9%-13%) of reoperation rate in studies with fair quality and 7% (95%CI: 5%-10%) of reoperation rate in studies with high quality. The overall and subgroup analysis results were shown in Table 2.

Fig. 2
figure 2

Forest plot regarding to reoperation rate

Full size image
Table 2 Meta analysis of reoperation rate
Full size table

Meta-analysis of risk factors for reoperation

The meta-analysis showed that obesity (OR = 1.91, 95%CI: 1.04–3.51, I2 = 53.1%), diabetes (OR = 2.01, 95%CI: 1.43–2.82, I2 = 0%), and smoking (OR = 1.51, 95%CI: 1.23–1.84, I2 = 0%) were associated with an increased risk of reoperation. Age (OR = 0.99, 95%CI: 0.95–1.03, I2 = 78.4%), sex (OR = 1.31, 95%CI: 0.83–2.05, I2 = 60.4%), and more bleeding (OR = 0.86, 95%CI: 0.07–10.22, I2 = 87.5%) were not associated with the reoperation. The overall results were demonstrated in Table 3. Forest plots regarding to obesity, diabetes, and smoking were demonstrated in Fig. 3A, B, and C, respectively.

Table 3 Risk factors for the reoperation of DLS patients after surgeries
Full size table
Fig. 3
figure 3

Forest plots regarding to obesity (A), diabetes (B), and smoking (C)

Full size image

Systematic review of risk factors for reoperation

This systematic review examined two literatures about the obesity. Chan et al. carried out a retrospective cohort study of obesity and reoperation after lumbar surgery [23]. As expected, significant higher risk of reoperation was found in patients who were obese [23]. Similar evidence was supported by Rabah et al. that an increase of one unit in BMI was associated with 4% increased risk of reoperation [11]. Moreover, study of Chan et al. showed addition of fusion was associated with higher risk of reoperation [23]. Rabah et al. found operative time > 5 h to be associated with an increased risk of reoperation [11]. A study consisted of 5-year follow-up indicated that having an index of fusion operation and perioperative complications was associated with the increased odds of reoperation [13]. Compared to intervertebral fusion, patients undergoing posterolateral fusion had 4.02-times risk of reoperation [8]. In addition, Gerling et al. have reported that patients with 2/3 moderate or severe stenotic levels, predominant back pain, no physical therapy, and greater leg pain score at baseline indicated higher reoperation rate [26].

Assessment of publication bias and sensitivity analysis

Sensitivity analysis was performed by sequentially removing the study to assess the robustness of overall results. All results of sensitivity analysis were consistent with those of the main analysis (Tables 2 and 3). By funnel plot, we detected an evidence of publication biases (Z = 2.91, P = 0.004) (Table 2, Supplementary Fig. 1A). Therefore, a trim-and-fill method was utilized to fill the missing data to eliminate the impact of publication bias. Funnel plot with missing data filled was demonstrated in Supplementary Fig. 1B. Before filled, reoperation rate was 10% (95%CI: 8%-12%). After filled, reoperation rate was 11% (95%CI: 9%-13%).

Discussion

The reoperation rate of DLS patients undergoing lumbar surgeries remains high in spite of improved surgical skills and techniques; therefore, exploring risk factors of reoperation is important [7, 8]. Considering the controversial results in the risk factors [8, 11,12,13], we performed a systematic review and meta-analysis based on currently available studies to analyze the reoperation rate and risk factors. In this study, we found a 10% of reoperation rate in DLS patients after lumbar surgeries. Obesity, diabetes, and smoking were identified as risk factors for the reoperation.

Several previous studies have reported the reoperation rate after lumbar surgeries in DLS patients [7, 49,50,51,52]. The reoperation rate was reported as 12.4% from 1990 to 1993 and 14.0% from 1997 to 2000 [51]. Ghogawala et al. proved that the reoperation rate was 15% at 1 year after the surgery in DLS patients only undergoing decompression [52]. In the present studies, the reoperation rate was found nearly the same as that reported in previous studies [7]. The reoperation rate in DLS patients was 15.7% at the mean follow-up of 8.2 years [7]. For patients undergoing fusion procedures, the cumulative reoperation rate was 14% [49]. Another report demonstrated that the reoperation rate ranged from 5.8% to 16.3% according to the type of surgeries [50]. Similar to the studies mentioned above, in our study, the reoperation rate of DLS was 10%, ranging from 8 to 12%. Our results may be useful for clinicians to evaluate the reoperation rate.

Identifying risk factors of reoperation for patients after lumbar surgeries is of clinical interest. In this study, obesity, diabetes, and smoking were found to be associated with higher risk of reoperation. Rabah et al. and Chan et al. have confirmed that smoking status was associated with greater risk of reoperation [11, 23]. Also, there were several studies reporting the positive association between obesity and reoperation of patients undergoing lumbar surgeries [8, 23]. This can be explained by that obese patients were more likely to be frail [53, 54], and frail patients had 56% increased odds of reoperation after lumbar surgery [23].

Animal studies have long recognized the close association between diabetes and lumbar spine disorders [55,56,57]. Diabetic models have revealed some harmful changes, such as increase of toxic end products of glycation, expression of matrix metalloproteinases 2 related to extracellular matrix degradation, and hyperglycemia-induced intervertebral disc inflammation, promoting intervertebral disc degeneration process [58,59,60]. Studies have revealed that diabetes was closely associated with degenerative lumbar spine disorders [61, 62]. Park et al. have found the influence of diabetes on the prevalence of lumbar spine surgeries, indicating that diabetes may be a factor aggravating lumbar spine disorders [62]. In Park et al. study, patients with diabetes underwent more lumbar surgeries than those without diabetes [62]. Their finding suggested that diabetes was significantly associated with the increased number of lumbar spine surgeries, and this finding is of critical importance because it revealed that diabetes may be an incentive for the increase of the severity of lumbar spine disorders, which ultimately led to the necessity of surgeries [62]. In this meta-analysis, diabetes was identified as a risk factor for the reoperation of DLS patients undergoing lumbar surgeries. This was consistent with the findings from Zhong et al. [8] Our findings suggested that when treating DLS patients with diabetes, physicians should pay more attention to glycemic control for the purpose of decreasing the risk of reoperation.

This meta-analysis explores the reoperation rate and risk factors for the reoperation. Results show that there is 10% of reoperation after lumbar surgeries, and obesity, diabetes, and smoking are found to increase the risk of reoperation. Our findings suggest that DLS patients should control glycemic level and weight, and reduce smoking to decrease the risk of reoperation. There are some limitations in this study. First, all fusion techniques (PLF and LIF) were put together. Due to the limitations of the included studies, it is unable to further analyze the reoperation rate in DLS patients undergoing the single fusion technique. Second, the number of studies reporting the risk factors of reoperation is relatively small, and some outcomes can only be qualitatively described, which may affect the stability of the results. Third, the risk of reoperation may be different according to the severity of lumbar spondylolisthesis and the first surgical methods; however, data provided in the currently available studies are insufficient to further analyze. Future meta-analysis including more relevant studies are needed to verify our findings and to explore the effect of lumbar spondylolisthesis severity and the first surgical methods on the risk of reoperation.

Conclusion

Our meta-analysis found 10% of reoperation rate in DLS patients undergoing lumbar surgeries, and identified obesity, diabetes, and smoking as risk factors for the reoperation. Our findings suggested that patients should improve glycemic level and weight, and quit smoking to reduce the reoperation after lumbar surgery.