Time to reconsider steroid injections in the spine?

To the Editor: As surgeons with a combined experience of more than 80 years in managing lumbar disc prolapse, we would like to add to this debate on the place of corticosteroid injection, from the advantage of those who have both seen the problem during surgery and managed the condition non-operatively. The opinions expressed recently in the Journal1-4 suggest such experience is essential to the understanding of both the pathogenesis and natural history of the disorder.

In the United Kingdom in the 1950s, lumbar disc excision was commonly performed under local anaesthesia with patients lying on their sides and the sciatic side uppermost. Then, it was observed that retracting a normal, white nerve root produced no response whereas retracting an inflamed root taut over a prolapse reproduced the patient’s radicular pain. When the surgery is performed under general anaesthetic, the irritability of the inflamed nerve root is evidenced by twitching in the distal musculature served by the root as it is retracted; a rise in protein levels in cerebrospinal fluid (CSF) is seen in the acute phase.5 Further, and most importantly, when patients with disc prolapses proven on imaging are successfully managed non-operatively without corticosteroid injection, and the imaging is repeated when they have recovered clinically, the size of the lesion is essentially unchanged. In short, it is the focal inflammatory response that is responsible for the acute radicular pain (irritative radiculopathy), which may be accompanied by variable degrees of root function impairment (compressive radiculopathy).

The natural history of discogenic radicular pain in the vast majority of cases is for some degree of gradual resolution over 6–8 weeks. Nevertheless, there is a rational basis for the use of local corticosteroid injections under computed tomography guidance in the acute phase (up to 4 weeks) for relief of demanding symptoms; the higher the CSF protein levels, the better the response.5

Relying on data that do not relate to the time between the onset of symptoms and the intervention leads to spurious conclusions.6 Focal inflammatory reactions secondary to degenerative (degradative) conditions in connective tissues generally (eg, tennis elbow, supraspinatus tendinitis) respond well to local infiltrations of corticosteroids in the acute phase but far less well, if at all, when the condition is chronic.