Case of the Month #21: Spinal Stenosis by Dr Helen Laycock

Incidental asymptomatic lumbar spinal stenosis does not require treatment.  

Evidence for most treatments is limited, NICE has not produced guidelines on lumbar spinal stenosis management and therefore management plans should involve shared decisions making with the patient, taking into account the symptoms, their severity and impact on the patients life, their overall health and how this impacts on risks and benefits of different options and also patient preference.

As lumbar spinal stenosis is a degenerative condition, treatment should not focus on cure, rather to stabilise or improve symptoms.  Initial management should start with information on the relationship between posture and symptoms and advice on activities that involve flexion of the lumbar spine (such as swimming side stroke or cycling).  Randomised clinical trials have in general demonstrated benefits of structured, supervised exercise programs and manual therapy, to improve both functional status and pain when compared to self directed or group exercise.

Utility of medications in lower back pain has reasonable evidence, however studies are lacking in those with spinal stenosis and evidence from other spinal disorders should only be applied cautiously.  Whilst paracetamol is associated with little harm, overall there is minimal efficacy when used in lower back pain.  Non-steroidal anti-inflammatory medications do reduce chronic lower back pain from numerous causes, however their utility is limited in older populations due to their side effects profile impact on cardiovascular, renal and gastroenterology systems.  Small studies show some benefits from gabapentin in spinal stenosis but were associated with increased dizziness and drowsiness.  There have been no studies of duloxetine in spinal stenosis and those for lower back pain show modest efficacy but adverse events were common.  A recent Cochrane review on systemic steroids in spinal stenosis indicated they are probably no more effective than placebo for short-term pain or function. In general pain medications should be considered sparingly for short term pain relief or as a bridge to more definitive management if planned, and should involve a discussion of side effects with patients.

Overall, there appears to be no benefit of epidural steroids compared to local anaesthetic alone.  Their effects were short lived (up to 3 weeks) with no sustained benefit over 6 weeks. Epidural injections are associated with small risks of important adverse events including infection and haematomas. Whilst a review of interventional pain medicine in 2018 gave weak recommendation for epidural local anaesthesia without steroids based on low quality evidence, current NICE guidance is to not use epidural injections for neurogenic claudication in people who have central spinal canal stenosis. 

A small proportion of patients require operative treatment for persistent symptoms and functional limitation that is not responsive to conservative management, especially where these symptoms are worsening or there is neurological deficit.  This involves increasing the space around the neural structures and is achieved commonly by a decompressive laminectomy, where all or part of the vertebral lamina are removed.  This may also require lumbar fusion if there is instability and surgical choice will be determined by the spinal surgery team and patient based on risks and benefits of each technique.  Results from randomised controlled trials are mixed, especially when compared with rigorous physiotherapy as opposed to less structured non-operative care.