Implant-associated osteomyelitis is caused by persistent bacterial infections, predominantly by staphylococci species forming biofilms on implants or osteosynthesis – materials. In the majority of patients the systemic immune response appears to be inconspicous with only minor upregulation of activation-associated receptors on the polymorphonuclear neutrophils (PMN). We found, however, evidence the activation of T cells, apparent as the expansion of CD4+ and CD8+ T cells bearing the activation-associated receptor CD11b. These cells also lacked the co-stimulatory molecule CD28, which is a further indicator for T cell activation. Moreover, small populations of T cells expressing Toll-like receptors (TLR)1, TLR2 or TLR4 were detected in the patients, while in healthy donors less than 1 % of T cells express TLR. A preferential association of TLR1- and TLR2-expression with CD28-CD11b+ cells was seen, compatible with the fact these cells represent an activated phenotype. In addition to the peripheral blood we also analysed leukocytes recovered from the infected site during surgery for removal of the implant. Predominantly PMN were found, highly activated as judged from their surace recpetor pattern, but also CD4+ and CD8+ T cells. As expected, these T cells represented an activated phenotype, and particularly the CD8+ cells expressed CD57, a receptor identifying end-differentiated T cells. The T cells recovered from the infected site, but not the peripheral blood T cells, produced interferon gamma, a cytokine known to support the function of phagocytic cells. In conclusion our data provide evidence that in response to local, persistent bacterial infections T cells are activated to acquire – among others – receptors selective for bacterial products, which in turn might modulate the T cell function and hence the host defence.

We evaluated the potential of a vastus lateralis muscle flap in controlling infection after resection arthroplasty of the hip. We retrospectively reviewed 119 patients with 120 chronic infections after resection arthroplasty treated with this procedure. The flap was fixed with Mitek anchors in the acetabular cavity. The mean duration of infection after resection before the muscle flap procedure was 6.5 months (2 to 13). The patients had previously undergone a mean of 4.9 operations (2 to 25). In all patients the infected cavity was the origin of the persistent infection. The mean follow-up was for 2.6 years (1.0 to 4.7).

No patient had recurrent infection post-operatively and all had an improvement in the pain and better quality of life.

The formation of bacterial biofilms is increasingly recognised as the leading cause of chronic infections. It limits the application of implant materials including catheters, heart valves, or orthopaedic prostheses. It is generally assumed that the infection persists because bacteria organised as biofilms escape the host defence mechanisms. Nevertheless, when studying patients with infected implants, we found a massive infiltration of leukocytes particularly polymorphonuclear neutrophils, PMN, into the site of infection, which led to the question, whether the PMN interact with the bacterial biofilm or not.

The interaction of human PMN with Staphylococcus aureus biofilms was studied in vitro.

S.aureus was cultivated on glass cover slips for various times under conditions allowing formation of biofilms. Adherence of PMN to biofilms and phagocytosis of the bacteria were observed by confocal laser scan microscopy and time lapse video microscopy.

Migration of PMN on and into the biofilm was identified as being phagocytosis, apparent as uptake of bacteria into the cell. Concominantly, in the wake of migrating PMN bacteria depleted zones appeared, which increased in size with time. In addition to phagocytosis, release from PMN of DNA and also of elastase was seen, suggesting the formation of neutrophil extracellular traps (NETs). So far, the signal for DNA release and NET formation has not been identified; of note is, however, that they occurred preferentially on established “old” biofilms and in the absence of the opsonising human serum, while phagocytosis was most efficient with developing “young” biofilms.

Taken together, our data provide evidence that bacteria in biofilms are not entirely protected against host defence but that phagocytosis is still possible, especially when the biofilm is opsonised with human serum. Whether NET formation also contributes to bacteria killing in biofilms cannot be decided as yet but remains an attractive alternative.

P.aeruginosa causes acute and chronic-destructive infections, particularly wound infections, or device-associated infections by colonising respiratory tubes, catheters, or implants. The pathogenicity of P.aeruginosa is largely attributed to the relative resistance towards host defence. Especially when organised as biofilms, the bacteria evade phagocytosis and killing by polymorphonuclear neutrophils (PMN).

To elucidate the evasion mechanisms, the migration of PMN towards and through P.aeruginosa biofilms was studied. Migration of PMN towards P.aeruginosa biofilms was tested using various in vitro techniques.

We found that PMN migrated towards developing P.aeruginosa biofilms, attracted by the quorum-sensing molecule N-3-oxododecanoyl homoserine lactone (3OC12-HSL). Mature biofilms which no longer produced 3OC12-HSL did not attract PMN. Addition of interleukin 8, a potent chemokine, restored the migratory capacity. Once arrived at the biofilms, PMN readily attached with no important difference between developing and mature biofilms. Migration into and penetration of the films, however, was only seen with developing films. By mass spectroscopy it became obvious that a major difference between developing and mature biofilms was the composition of the extracellular polymer substance, of which alginate is a prominent component. A series of experiments with isolated alginate showed that PMN did not migrate on or into alginate-containing matrices, but remained affixed at the contact site just as they did on mature biofilms. The mechanism of this firm attachment is still under investigation; prominent up-regulation of various adhesion molecules was seen, which could provide possible explanation.

Mature biofilms, most probably due to the composition of the extracellular polymer substance, do not allow the penetration of PMN. Consequently, bacteria embedded in deeper layers of the biofilm are protected against the host response. Due to the restricted movement of PMN, the bactericidal activity of PMN is only efficient against bacteria in the immediate vicinity, explaining the inefficient host defence.

In selected patients, knee arthrodesis is a well-recognised salvage procedure after infected total knee arthroplasty (TKA). Several procedures of arthrodesis have been introduced and should be adapted to the individual situation of the patient. In our center we regularly treat elderly patients after multiple revision operations; in 36% defects of the bone, soft tissue or the extensor mechanisms are present. In these cases we prefer arthrodesis to reimplantation. Because of the high rate of non-unions when using an external fixator, we perform arthrodesis by use of an intramedullary rod system.

The objective of this study was to compare the results of different rod systems for knee arthrodesis after TKA infection.

We reviewed the results of 3 rod systems in 34 patients: cementless system (Brehm; n=9), cement rod usually used in tumor patients (Mutars; n=7) and a regular cement rod system (Link; n=18).

In the group of cementless rods we had to explantate 3 rods because of a relapse of the infection. This is most propably due to the technical design of the system: in poor soft tissue situation the tissue is compressed by the voluminary docking part which causes continuous necrosis. This problem can be avoided by an early tissue flap. Of the Mutars rod system we had to explantate 2 systems; one because of an infection, the other one due to telescoping, which can be avoided by use of a longer stem with the option to interlock. In the group of the Links system no revision was necessary.

In our opinion, arthrodesis of the knee using a rod system is a satisfactory salvage procedure following an infected TKA, especially in elderly patients, and can provide reliable, painless extremity and satisfactory quality of life.

Introduction: Improved treatment of primary tumors leads to an increased life expectancy and thus to an increasing number of patients with bone metastases. Techniques like auto- or allogenic bone grafts, vascularised bone transfer and distraction osteogenesis often require multistage surgical procedures and inhibit full early limb function. Diaphyseal replacement using nails, plates and bone cement do not guarantee long term bone stability. Due to this experiences a new diaphyseal replacement device for humerus, femur and tibia has been developed.

Materials and Methods: The diaphyseal replacement implant OSTEOBRIDGE^™ consists of two semi-circular cylindrical shells. The spacer is clamped around two nails via eight screws. Different sizes of spacers and nails can be used to bridge the bone loss correctly.

The outer diameter of the spacer ranges from 20–25–30 mm and the length from 40 to 70 mm. Two or three spacers can be combined via a special connector. Nails in the length 60 to 200 mm and the Ø 7 to 18 mm with the possibility of static or dynamic interlocking complete the modular system entirely made from Ti-6Al-4V.

Biomechanical Testing: Static compression tests to determine the maximum longitudinal forces of the clamp connection spacer/nail were performed, 4-point bending tests with the complete system to investigate the fatigue resistance were undertaken and torsional test to evaluate the rotational stability.

Prospectiv clinical evaluation: In between 2004 and 2006 35 patients were treated with the OSTEO-BRIDGE^™ system.

The indication: Tumor: Humerus 8, Femur 16, Tibia 2, Postinfectious: Tibia 4, Posttramatic Femur 4, Tibia 2.

Results:

Biomechanical testing: the clamp connection spacer/ nail can neutralize axial loads which can not be expected in human beings. The clamp connection spacer/nail Ø 10 mm resisted an average axial load of 8,5 kN. This can be compared to a force of 850 kN (equivalent to 10 multiples of 85 kg body weight). The bending test with a nail Ø 10 mm shows that the spacer can resist long term loads from an occurring stress of 400 N/mm² in the nail.

One spacer was removed during amputation for recurrency of osteosarcoma. No infection, no loosing were reported.

Discussion: The OSTEOBRIDGE^™ spacer system allows to replace lost daiphyseal bone over long distances with proved biomechanical stability. An advantage seems to be the early bony bridging over the spacer within the first 9 months. There might be another advantage in using the spacer as a container for antitumor, bactericidal or bone growth stimulating drugs.

Introduction: If you loose the talus, especially in infection cases, you have a problem. Literature says, that in a lot of cases you should amputate. We analysed 17 cases, being transferred to our hospital with a non-union an 12 own cases. The question was: What ca you do, to get an acceptable result. An analysis of the mistakes was tried to achieve.

Material and Method: 29 cases were analysed in this retrospective study. In 27 cases an infection situation was found. In 2 cases a vascular disturbance of the talus, avascular necrosis, was found. We analysed the stabilization method, the length of non weight bearing and the personal compliance of the patient. We found patients being stabilized with a nail, an external fixator of AO or Ilisarov Fixator.

Results: In about 68% of the non-union cases too early weight bearing therapy according to mobilization of the joints occurred. In 32% a cancellous bone graft didn’t take place. Shortening of the leg > 4 cm happened in 69%.

Discussion: The problem of stabilization of this anatomical area is well known. The anatomy itself brings problems. But the surgeons opinion, to make a very early functional treatment possible, is contraproductive in these cases. Leave the implant much longer than you would think, start weight bearing only after 6 weeks and you can get reasonable results. Callus distraction solves the problem of leg length discrepany. During that time, lasting longer than the arthrodesis, the bone heals. Leave the fixator for the whole time in situ and you will get an acceptable result for the patient. More than that, make an additional arthrodesis between the tibia and the Naviculare bone. Our own results demonstrate a success rate of 80% in these desperate cases.

Introduction: Posttraumatic osteitis is a localised inflammatory process leading to tissue destruction and eventually osteolysis. The molecular mechanisms underlying the disease progress are not yet fully understood. In a previous study we demonstrated infiltration of polymorphonuclear neutrophils (PMN) into the site of infection; the PMN were highly activated as seen by upregulation of the activation-associated surface receptors CD14 and CD64. In this study we analysed the superoxide generation by the infiltrated PMN as possible pathomechanism of the local tissue destruction.

Material and Methods: Ten patients with device-associated osteomyelitis requiring surgery were recruited into the study. When removing the infected implant the site was rinsed intraoperatively. The leukocytes were recovered, then activation-associated surface receptors were determined by cytofluorometry as was superoxide generation by reduction of cytochrome C.

Results: 1–2 x 10⁷ leukocytes were recovered from the «lavage» fluid; 80 to 90% were identified as PMN. The PMN were highly activated as seen by an upregulation of CD14 and CD64, and a concomitant downregulation of the selectin CD62L. In response to phorbol ester (PMA) the superoxide production of the infiltrated PMN was enhanced when compared to peripheral PMN of the same patient. The infiltrated PMN, but not the PMN of the peripheral blood, responded to the bacterial peptide f-Met-Leu-Phe (f-MLP) with superoxide production, indicating an enhanced responsiveness of the cells. The underlying molecular mechanisms were analysed in vitro using PMN of healthy donors: only the induction of superoxide production by f-MLP, but not by PMA, required a «priming» of the cells, for example by low doses of lipoploysaccharide (LPS) or cytokines (e.g. TNFa, IL-8).

Conclusions: In posttraumatic osteomyelitis PMN infiltrate the infected site; they are locally activated as seen by an upregulation of the appropriate receptors and by “priming” for superoxide generation. Priming of local PMN could on one hand potentiate the bactericidal activity, on the other hand contribute to tissue destruction. The occurrence of viable bacteria and activated «armed» PMN at the same site points to an esacpe mechanism, possibly due to biofilm formation. Due to their cytotoxic and proteolytic potential PMN might participate in local tissue destruction and osteolysis.