Dear Editor, Spinal infections occur with increasing incidence. Conservative, nonoperative treatment, including external immobilisation with bed rest or orthesis and long-term antibiotic therapy, is still considered the first tier therapy by some authors, but should, in our view, be reserved for beginning slight spondylodiscitis, patients suffering from catecholaminedependent sepsis until improvement of general health, and patients with significant co-morbidities, which exclude surgery, e.g. vital indications for anticoagulation. Long bedrest implies complications like muscle atrophy, pneumonia, lung embolism and acceleration of osteoporosis. Above that, it leads to a “clinical success” in only 75% of patients. In cases of neurological deficits and/or segmental instability, surgical treatment, including decompression and stabilisation, is required. There is no significant evidence for the benefit of postoperative bracing in spondylodiscitis. In case of sparse bone damage, there is no significance for an additional external immobilisation after the surgery to be beneficial for the recovery. As yet, there is plenty of experience with autologous bone or titanium implants, but only scarce experience with the use of polyetheretherketone (PEEK) implants in the surgical treatment of spinal infections. Both Mondorf et al. [1] and Walter et al. [3] described their clinical experience with single-step surgery using PEEK implants for the cervical spine in five patients suffering from spondylodiscitis with good outcome and complete recovery from the infection in all cases [1, 3]. Pee at al. [2] described the use of PEEK implants in the surgical treatment of ten patients presenting with lumbar spondylodiscitis with comparable satisfactory findings. We present a consecutive retrospective series of nine patients [three cervical (monosegmental) and six lumbar (three monosegmental and three bisegmental spondylodiscitis)] treated in our institution between January 2006 and May 2010 consisting of clinical evaluation, standardised questionnaires [EQ-5D, SF-36 and oswestry disability index (ODI)], magnetic resonance imaging (MRI) and computerized tomography (CT) scans for assessment of instability, as well as C-reactive protein (CRP) levels collected during the perioperative period and the outpatient follow-ups (mean: 13 months). This series represents currently our complete experience with PEEK in this disease. The mean age was 61.7 years (min–max 47–81) for cervical spondylodiscitis and 65.8 years (min–max 60– 82) for lumbar spondylodiscitis. All patients had significant neck/back pain, one patient complained of brachialgia. Two cervical and two lumbar spondylodiscitis patients suffered from infection with significant epidural mass and sensorimotor deficits, one of the cervical patients even with tetraplegia and bowel/bladder dysfunction. Two lumbar patients presented a septic disease with significantly increased infection parameters, especially the CRP. One cervical and three lumbar patients had additional further A. Brase (*) : F. Ringel :C. Stuer : B. Meyer :M. Stoffel Department of Neurosurgery, Klinikum rechts der Isar, University of Technology Munich, Ismaninger Strasse 22, 81675 Munich, Germany e-mail: agnes.podlewski@lrz.tum.de