Arguments for the choice of surgical treatments in patients with lumbar spinal stenosis – a systematic appraisal of randomized controlled trials

In this review of arguments for or against choosing a specific type of treatment for a patient with lumbar spinal stenosis nine randomized trials (RCTs) were appraised. The main argument for decompression surgery only was a high success rate (three RCTs). Furthermore, decompression alone is less invasive than fusion surgery (one RCT) and maintains the relative stability of the spine (one RCT). The main argument for fusion was that it is a well-established technique with a high fusion success rate (two RCTs) and statistically significant improvement in clinical outcome compared to decompressive lumbar laminectomy alone (one RCT). Arguments for the choice of fusion with instrumentation were an increased fusion rate compared to fusion (two RCTs) and prevention of spondylolisthesis progression (two RCTs).

Main arguments against decompression alone were an increase in vertebral instability and a progression of spondylolisthesis (three RCTs). The main argument against fusion was that the indication for fusion in spinal stenosis has remained unclear (one RCT). In particular, degenerative changes (osteophytes, decreased disc height, calcified ligaments) increase the stability of the spine and thereby reduce the need for arthrodesis (one RCT). The main arguments against fusion with instrumentation were that the invasive procedures are associated with more complications as, e.g., device related osteoporosis (one RCT) and are more expensive (one RCT).

The main argument identified in this appraisal for and against decompression alone in patient with lumbar spinal stenosis was whether or not instability should be treated with (instrumented) fusion procedures. Increased vertebral instability may lead to progression and compression of nerve roots and therefore require reoperation. Further, by reading the original papers we noticed that key terms including instability as well as reoperation are not clearly and unambiguously defined. Authors of different papers used various definitions for these terms. For the interpretation of study results and the appraisal of the clinical implications, or the synthesis of the results of original studies in a systematic review it is necessary to know what the different terms denote and how the different concepts, e.g., instability, are operationalized and quantified.

Various definitions of the term instability have been published that vary among experts. The meaning is different for clinicians, radiologists, and bioengineers [18]. An example for a clinical definition of instability is presented by White and Panjabi [19,20] as follows: “The loss of the spine’s ability to maintain its pattern of displacement under physiologic loads so there is no initial or additional neurologic deficit, no major deformity, and no incapacitating pain”. A radiological definition reported by Sonntag and Marciano [21]: “Increased angulatory or translatory motion noted on active flexion-extension lateral or anteroposterior bending radiographs that exceeds 4 mm or 10° angulation […]” According to Frymoyer [22] from a bioengineers point of view segmental instability is defined as “a loss of spinal motion segment stiffness, such that force application to that motion segment produces greater displacement than would be seen in a normal structure, resulting in a painful condition, the potential for progressive deformity, and neurologic structures at risk”.

And even between physicians of the same discipline, opinions on what instability means, differ exemplified by a further definition from radiology reported by White and Panjabi [19]. They defined criteria for diagnosing instability from flexion-extension radiographs as “sagittal plane translation greater than 4.5 mm or greater than 15% of the vertebral body width, or sagittal plane rotation of greater than 15° at L1/L2, L2/L3 or L3/L4, greater than 20° at L4/L5, or greater than 25° at L5/S1” [23].

An additional hurdle is that the association between clinical signs and radiological findings remains controversial and challenging [18,24]. A study conducted by Pitkanen et al. [25] found only poor correlation between radiological abnormalities identified in functional radiographs and clinical signs suggesting lumbar instability. It remains unclear whether the radiological approach is poor, or the definition of clinical instability is not that appropriate [25].

Recently an international and interdisciplinary expert panel was unable to agree on a definition of instability and its clinical relevance in symptomatic lumbar degenerative spondylolisthesis [26]. As a consequence of a lack of a broadly accepted definition and method of quantification of instability, researchers use different methods to quantify instability. This impedes the appraisal of the clinical impact of study results and hinders the synthesis of results from original studies in a systematic review. There is a need for a broadly accepted definition to facilitate the meaningful interpretation of study results.

An important question that needs to be addressed in future studies is whether a second surgery is necessary as a consequence of the first intervention or because of the progression of a preexisting degenerative disease. Many authors don’t define how they classify reoperations and the descriptions of reasons for repeat surgery vary [27], e.g., Martin et al. [28] defined in their study reoperation as “any lumbar operation in a patient who had at least 1 previous lumbar spine procedure. It was not necessarily a repeat of the same procedure or performed at the same vertebral location, but in all cases, it was still within the lumbar region”.

For the interpretation of study results it is important for clinicians and researchers that the reasons for a follow-up surgery in studies are clearly defined. Early reoperation can be necessary because of a surgical complication, such as spinal fluid leak, hematoma, infection, neurologic deficit, or mislocated instrumentation. For a reader these reoperations are clearly related to the prior surgery. This relation is less evident at a later stage. Non-union or complications resulting from surgical implants can be a reason for a reoperation at a later stage. Nevertheless, reoperations at a later stage may be performed because of recurrent or persistent pain symptoms, pseudoarthrosis, or progressive degeneration at another spine level [28]. Therefore, a detailed description, classification or definition is important to compare the results of different studies.

The main limitation of this study is the small number of randomized controlled trials comparing surgical procedures and therefore limited discussion on arguments for or against specific techniques. While spinal stenosis is a prevalent disease in elderly patients and surgical interventions are performed on a regular base only nine RCTs were available for the current appraisal.

Another limitation of this study is the missing distinction of different types of decompression procedures used (e.g. complete facetectomy or complete laminectomy). The extent of laminectomy or facetectomy may influence postoperative instability and should be addressed in all studies.

It is important that researchers and clinicians agree on definitions for important key concepts such as instability and description of reasons for reoperation. Clear definitions of concepts and methods for the quantification or categorization facilitate the interpretation of study results and the synthesis of the results of different studies.

For clinicians it is important to know when to recommend a certain intervention to their patient. The lack of high quality RCTs that addresses these important clinical questions impedes clinicians from an evidence-based treatment algorithm in patients with lumbar spinal stenosis. Indications for conservative or surgical intervention are oftentimes based on clinical judgment of the treating physician and depend on personal beliefs and experiences. Future research is needed to provide more robust evidence for such treatment choices.