ABSTRACT
The authors tried to determine the ways in which vertebral puncture biopsy and the direct surgical approach can helpin the etiological diagnosis of spondylodiscitis, when the clinical radiological, and laboratory examination leave the clinician in doubt. Out of 28 vertebral puncture biopsies, proof of tuberculosis was obtained in 4 cases, and proof of a non-tuberculous cause was obtained in 1 case as a result of isolating the responsible organism. Thirty-eight cases were submitted to the direct surgical approach. When the indication for surgery was solely in order to investigate the etiology, proof of tuberculosis was obtained in half of the cases, and only exceptionally was a non-tuberculous organism discovered. In almost half the cases, the histological characteristics were non-specific and no organisms were detected. Most of these latter cases were, in fact, non-tuberculous spondylodiscitis, although subsequently some of them exhibited evidence of a tuberculous character. In the present state of knowledge, it seems that, in cases of clearly non-tuberculous spondylodiscitis, the direct surgical approach is not justified if the objective is solely to isolate the organism, so that its sensitivity to antibiotics can be tested. The following reasons are given for this conclusion : the direct surgical approach only rarely leads to isolation of the causal organism; although treatment based on knowledge of antibiotic sensitivity may help to restrict evolution of the disease, it does not reduce significantly, or only rarely, the permanent partial incapacity. If the results published by Seignon and Gougeon are confirmed, early needle puncture of the diskovertebral centre of the disease should be practised more widely.