Almost any kind of bacteria can be involved in prosthetic joint infections (PJI).

In 2014 we collected data on 3 patients presenting with hip or knee pain and radiological signs of prosthetic joint mobilization without any other clinical or biochemical suspect of infection.

Interestingly, in all cases we isolated bacteria very rarely involved in human infections: Cupriavidus pauculus, Ralstonia paucula and Neisseria flavescens. Both Cupriavidus pauculus and Ralstonia paucula are Gram negative environmental bacteria which can contaminate water (tap, pool and mineral) and have been isolated from a variety of human clinical sources including blood, wounds, sputum, urine, eye, throat and peritoneal fluid, as non pathogenic flora. In such two patients we were able to cure the infection with quinolones (both strains were sensible). Neisseria species are Gram-negative bacteria included among the proteobacteria. Neisseria flavescens is often found in the upper respiratory tract and oropharynx of humans, rarely associated with infectious process (necrotizing pneumonia and empyema). In this last patient we successfully used piperacillin-tazobactam i.v. during hospitalization and shifted to amoxicillin-clavulanate per os at discharge.

In conclusion, any bacteria can induce PJI: in some cases pre-operative blood tests are normal and therefore useless. We need new tests to overcome this diagnostic problem (such as alpha-defensin) and we must check on rare bacteria prolonging incubation to a more extended period of time (as in Propionebacterium acnes or Mycobacterium spp) and/or sending samples to specialized laboratories.

We thank all the medical and nurse staff of Prosthetic Joint Replacement Unit of Orthopaedic Center, Humanitas Research Hospital

Pedicle screws fixation to stabilise lumbar spinal fusion is the gold standard for posterior stabilisation. Pedicle screws are today positioned in free hand or under fluoroscopic guidance with an error from 20% up to 40–50%, which can determine the inefficacy of treatment or severe damages to close neurologic structures. Surgical navigation drastically increases screws placement accuracy. However its clinical application is limited due to cost reasons and troubles related to the presence of a localiser in the OR and the need to perform a registration procedure before surgery. An alternative image guided approach is the use of patient specific templates similar to the ones used for dental implants or knee prosthesis. Until now, the proposed solutions allow to guide the drill, and in some cases, as templates fit completely around vertebra, they require the complete removal of soft tissues on a large portion of the spine, so increasing intervention invasiveness. To reduce the soft tissue demolition, some authors proposed a fitting based on small “V shape” contact points, but these solutions can determine instability of the template and the reacting of wrong stable positions.

In our solution, after spine CT acquisition, each vertebra is segmented using a modified version of ITK-SNAP software, on which the surgeon plans screws positioning and finally the template is designed around the chosen trajectories, using a tool which allows to insert cylinders (full or empty) in the segmented images. Each template, printed in ABS, contains two hollow cylinders, to guide the screws, and multiple contact points on the bone surface, for template stabilisation.

We made an in-vitro evaluation on synthetic spine models (by Sawbones) to study different template designs. During this first step an ongoing redesign allowed to obtain an optimal template stability and an easy template positioning to minimise the intervention invasiveness. A first contact point, which fits on the sides of the spinous process, is used to simplify template alignment. The other 4 contact points, which consists of cylinders (diameter 5 mm), fit exactly on spine surface in correspondence to the vertebra's lamina and articular processes to stabilise the template in an unique position. Templates can be used to guide not only the drill, but also Kirschner wires, to guide cannulated screws. After the Kirschner wires insertion the template can be dismounted for its removal (the direction of the kirschner wires are not parallel).

After the definitive template design an ex-vivo animal test on 2 porcine specimens has been conducted to evaluate template performance in presence of soft-tissue in place. The specimens have been scanned with CT, we realised a total of 14 templates and we performed the insertion of 28 Kirschner wires. We evaluated that after the soft tissue dissection and the bone exposure, the template can be easily positioned in the right unique position, with no additional tissue removal compared to the traditional approach, requiring just removal of the soft tissue under the small contact points using an electric cutter. The surgeon evaluated (and corrected) some wrong stable template positions when not all the contact points were in contact with the bone surface. The post-op evaluation was made with a CT scan that showed 1 cortical pedicle violation (3.5%) (grade II according to the FU classification).