Evidence & Methods
Lumbar fusion is commonly performed for degenerative conditions and isthmic spondylolysis. A relatively high reoperation rate has been demonstrated with extension of fusion to adjacent levels in many of these surgeries. Risk factors for progressive disease at adjacent segments have not been well delineated.
In this retrospective study of patients undergoing posterior lumbar interbody fusion, the authors found a mean 2.5% annual incidence of surgery for adjacent-level degenerative processes. However, older patients having longer initial fusions, fusions to L5, and adjacent-level laminectomies appeared to be associated with an increased rate of additional surgery.
This study emphasizes that not all fusions, even if one technique is considered (PLIF), have similar downstream risk for adjacent-level disease. The information is helpful for informed consent. However, some considerations remain. The authors' surgical technique (PLIF) may have impacted their findings. Using repeat surgery as a surrogate indicator of progressive disease is problematic and may underestimate disease progression in older patients (ie, it is possible fewer patients will, or be able to, have additional surgery) or overestimate the disease progression in patients with ill-defined pain syndromes (ie, discogenic pain) who tend to aggressively pursue surgical care.
—The Editors
Thirty years ago, Ehni asserted that, “fusion generates a conflict between immediate benefit and late consequences” [1]. There remains a widely held belief that creation of rigid sections in the lumbar spine will lead to excessive stress on and premature degeneration of the motion segments adjacent to an arthrodesis [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23]. This belief is supported by in vitro evidence of increased stresses and intradiscal pressures at segments adjacent to a lumbar spinal fusion [24], [25], [26], [27], [28], [29]. If true, it is important that clinicians, their patients, and health economists have an accurate knowledge of the incidence and risk factors for the development of adjacent segment disease (ASD) after lumbar arthrodesis. Such knowledge will enable a better understanding of the role and place for nonfusion stabilization technology [21], [30], [31], [32].
A number of clinical studies have examined the prevalence of symptomatic ASD at various time points after lumbar fusion [2], [3], [4], [5], [15], [17], [22], [23], [33], [34], [35]. Harrop et al. [36] recently reported the results of a systematic review of the published rates of radiographic adjacent segment degeneration and symptomatic ASD after lumbar spinal fusion. They emphasized the difference between these two conditions and noted a range in the reported prevalence rates of ASD, between 0% and 36%, over various time intervals.
In 1999, Hilibrand et al. [37] reported the results of their detailed study of both the incidence and prevalence of ASD in 374 patients after anterior cervical fusion. They defined “annual incidence” as the percentage of patients in whom new disease developed during the course of a given year of follow-up, having been disease free at the start of that year, and found this to be 2.9%. They defined “prevalence” as the percentage of all patients in whom symptomatic ASD developed within a given period of follow-up. With the exception of a limited investigation by Ghiselli et al. [3], a similar study to that of Hilibrand et al. does not appear to have been carried out in the lumbar spine.
The aim of the present study was to determine annual incidence and prevalence figures for the lumbar spine in a large consecutive patient cohort. The substantial cohort size and uniform surgical technique aided subgroup analysis and investigation of potential risk factors for surgery for adjacent segment disease (SxASD). Similar Kaplan–Meier (K-M) [38] statistical methodology to Hilibrand et al. [37] was used to determine annual incidence and prevalence. Potential surgical and demographic risk factors for SxASD were explored using Cox proportional-hazards (Cox) regression modeling [39]. To reduce potential inaccuracy in retrospective diagnosis of clinically significant ASD, the authors chose to examine rates of surgery for ASD using the end point of further surgical intervention rather than the presence of symptoms or radiographic signs of degeneration.