Reoperation Rates Following Instrumented Lumbar Spine Fusion

Study Design. A prospective cohort study. Objective. This study evaluated the cumulative reoperation rate and indications for reoperation following instrumented lumbar spine fusion (LSF). Summary of Background Data. LSF reduces disability and improves health-related quality of life for patients with several spinal disorders. The rate of instrumented LSF has drastically increased over the last few decades. The increased incidence of LSF, however, has led to increased reoperation rates. Methods. The data are based on the prospective LSF database of Tampere University Hospital that includes all elective indications for LSF surgery. A total of 433 consecutive patients (64% women, mean age 62 years) who underwent LSF in Tampere University Hospital between 2008 and 2011 were evaluated and indications for reoperations were rechecked from patient records and radiographs. The most common diagnosis for the primary surgery was degenerative spondylolisthesis and the mean follow-up time was 3.9 years. The cumulative incidence of reoperations and the “time to event” survival rate was calculated by Kaplan-Meier analysis. Results. By the end of 2013, 81 patients had undergone at least one reoperation. The cumulative reoperation rate at 2 years was 12.5% (95% confidence interval: 95% CI: 9.7–16.0) and at 4 years was 19.3% (95% CI: 15.6–23.8). The most common pathology leading to reoperation was adjacent segment pathology with a cumulative reoperation rate of 8.7% (95% CI: 6.1–12.5) at 4 years. The corresponding rates for early and late instrumentation failure were 4.4% (95% CI: 2.7–7.0) and 2.9% (95% CI: 1.9–7.1), respectively, and for acute complications, 2.5% (95% CI: 1.4–4.5). Conclusion. Although previous studies reported that early results of spinal fusion are promising, one in five patients required reoperation within 4 years after surgery. Patients and surgeons should be aware of the reoperation rates when planning fusion surgery. Level of Evidence: 4


Introduction
Lumbar spine fusions (LSF) are performed to treat several spinal disorders when conservative treatment fails. LSF is effective for treating degenerative and isthmic spondylolisthesis [1,2].
Prospective randomized studies revealed that posterolateral fusion is more efficient than an exercise program for short term improvent of function and pain relief in adult isthmic spondylolisthesis [3,4]. Weinstein et al. [2] reported that decompressive laminectomy (with or without fusion) has advantages over nonoperative treatment for treating degenerative spondylolisthesis. A Cochrane review, however, revealed no clear benefits of surgery over nonoperative treatment for lumbar spinal stenosis [5]. In a recent randomized controlled trial of 247 patients with lumbar spinal stenosis with or without degenerative spondylolisthesis, Försth et al. [6] reported no benefit of adding fusion to decompression surgery, thus, fusion in spinal stenosis surgery is controversial. In contrast a randomized controlled trial by Ghogawala et al. [7] revealed that the addition of fusion to laminectomy slightly improved outcomes in 66 patients with degenerative grade I spondylolisthesis. The value of spinal fusion for the treatment of chronic low back pain is controversial [8][9][10]. Spinal fusions are also performed in revision spine surgery and for the treatment of scoliosis and recurrent disc herniations [11].
The incidence of spinal fusion surgery has drastically increased over the past few decades [11,12], and thus the incidence of reoperations has also increased. Any indication resulting in a return to the operating room is considered a reoperation, which is an undesirable outcome of LSF that causes pain and inconvenience for the patient and as well as additional expense to the patient, society, and employer. Based on the timing, reoperations are classified as early or late. Early reoperations are caused by an acute complication (<3 months after surgery), and instrumentation failure or early nonunion (< 1 year). Indications for reoperation 1 year or more after fusion may be pseudoarthrosis, persistent pain or recurrence of symptoms or adjacent segment pathology (ASP) [13,14]. According to population-based analysis, cumulative reoperation rates accounting for any unplanned return reoperation rate as high as 35% during a mean 26-month follow-up of instrumented fusions from the thoracic spine to the pelvis.
The aim of the present study was to determine the overall rate, indications and timing of reoperations-following instrumented LSF-in a prospectively collected cohort in a single university hospital experience.

Materials and Methods
Since the beginning of 2008, all adult patients from the Pirkanmaa district undergoing elective LSF at Tampere University Hospital were invited to join a prospective follow-up study. The data were collected as part of standard clinical practice and operation details were recorded in a database. Formal approval to collect the data was granted by the Ethics Committee of Tampere University Hospital and all patients provided signed written consent prior to data collection. As Finland has a national health insurance system, all hospitals cover The information about reoperations performed by the end of 2013 was collected from the database, and indications for reoperations were confirmed from patient records, radiographs and magnetic resonance images. The time between the index operation and reoperation was recorded. The disability was assessed by using the Finnish validated Oswestry Disability pathology at the mobile adjacent segment next to the fused area, comprising disc degeneration, disc space narrowing, disc herniation, spinal stenosis caused by hypertrophic facet arthritis and/or osteophyte formation, spondylolisthesis, instability, scoliosis, or vertebral compression fracture as defined earlier in the classification system of Park et al. [18,19]. 7. Implant removal was also classified as reoperation.

Statistical Analysis
The data are presented as counts with percentages or as means with range or standard deviation (SD). We calculated the 95% confidence interval (Cl) for the major outcomes. We performed a time-to-event analysis using the product-limit estimate of the Kaplan-Meier cumulative survival function to estimate the cumulative reoperation rate. The α-level was set at 0.05. SPSS Statistics for Windows version 19.0 (IBM Corp., Armonk, NY) was used for the analysis.

Results
Of the 433 LSF patients, 81 underwent at least one reoperation by the end of 2013 with a mean (SD) follow-up time of 3.9 (1.2) years. There were 32 patients who required reoperation within one year of the index operation. The cumulative reoperation rate at 2 years (95%) was 12.5% (Cl: 9.7-16.0) and at 4 years, 19.3 % (Cl: 15.6-23.8; Figure 1). The most common indications for primary surgery were degenerative spondylolisthesis and spinal stenosis accounting for 65% of the patients ( Table 1). The primary fusion was longer than two levels in 35 % of patients (158/433) and 37 % in patients who underwent reoperation (30/81). In the patients, who were not reoperated the preoperative mean (SD) ODI was 46 (16) and at two years post fusion 21 (19). Correspondingly, among patients who required Late failures were detected in 10 of the reoperated cases. Pseudoarthrosis was detected in all cases by fragmentation in the fusion area, loosened screws, or rod breakages.
Late failures were reoperated a mean of 2.4 years after the index operation (Table 2).
Indications for reoperation are time-dependent and the most common indication depends on which time interval from the index surgery is evaluated. Within 1 and 2 years after the index operation, the most common reason for reoperation was early failure with cumulative reoperation rates of 3.4% (Cl: 2.0-5.6) and 3.9% (Cl: 2.4-6.2), respectively (Table 3). Cumulative reoperation rates for late failures were 1.3 % (Cl: 0.5-3.6) at 2 years and 2.9% (Cl: 1.9-7.1) at 4 years. When evaluating the 4-year period after the initial operation, ASP was the main cause of the first reoperation with an incidence of 8.7% (Cl: 6.1-12.5; Figure 2A-C).

Discussion
The present prospective study revealed cumulative reoperation rate for LSF of 12.5% at 2 years and 19.3% at 4 years. The most common indication for reoperation was ASP accounting for 8.7% of the reoperations.    [15,34]. One confounding factor is patients who died before the end of 2013 and thus the follow-up time remained shorter, but is natural when studying an older patient population. A study from South Korea excluded patients who died from the analysis [15].
A strength of this study is the population-based, nonselected, patient cohort, including almost every patient who underwent elective instrumented LSF between 2008 and 2011 among patients in the Pirkanmaa region. Population-based studies are less susceptible to selection or nonresponse biases than typical case series studies [13]. Another strength is the small number of drop-outs, small amount of missing data, and adherence to the study protocol. Almost all patients (98%) enrolled in our study completed the follow-up period, and the only missing follow-up data were due to patient death.
In our study, cumulative reoperation rates after instrumented LSF was 12.5% at 2 years and 19.3% at 4 years, which should be taken into account when planning fusion surgery. Although the rate is high, it comprehensively includes indications for primary fusion and reoperations. Reoperations are frequent after LSF, occurring in one in five subjects within 4 years. Therefore, the indication for instrumented LSF must be well defined and the risk for reoperation must be kept in mind and should be part of the preoperative consent discussion with the patient.
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.