Aims

The aim of this study was to evaluate the ability of a machine-learning algorithm to diagnose prosthetic loosening from preoperative radiographs and to investigate the inputs that might improve its performance.

We present detailed information about early morbidity after aseptic revision knee replacement from a nationwide study. All aseptic revision knee replacements undertaken between 1st October 2009 and 30th September 2011 were analysed using the Danish National Patient Registry with additional information from the Danish Knee Arthroplasty Registry. The 1218 revisions involving 1165 patients were subdivided into total revisions, large partial revisions, partial revisions and revisions of unicondylar replacements (UKR revisions). The mean age was 65.0 years (27 to 94) and the median length of hospital stay was four days (interquartile range: 3 to 5), with a 90 days re-admission rate of 9.9%, re-operation rate of 3.5% and mortality rate of 0.2%. The age ranges of 51 to 55 years (p = 0.018), 76 to 80 years (p < 0.001) and ≥ 81 years (p < 0.001) were related to an increased risk of re-admission. The age ranges of 76 to 80 years (p = 0.018) and the large partial revision subgroup (p = 0.073) were related to an increased risk of re-operation. The ages from 76 to 80 years (p < 0.001), age ≥ 81 years (p < 0.001) and surgical time > 120 min (p <  0.001) were related to increased length of hospital stay, whereas the use of a tourniquet (p = 0.008) and surgery in a low volume centre (p = 0.013) were related to shorter length of stay.

In conclusion, we found a similar incidence of early post-operative morbidity after aseptic knee revisions as has been reported after primary procedures. This suggests that a length of hospital stay ≤ four days and discharge home at that time is safe following aseptic knee revision surgery in Denmark.

Cite this article: Bone Joint J 2014;96-B:1649–56.

Metallosis is a rare cause of failure after total knee replacement and has only previously been reported when there has been abnormal metal-on-metal contact. We describe 14 patients (15 knees) whose total knee replacement required revision for a new type of early failure caused by extensive metallosis. A modification of a cementless rotating platform implant, which had previously had excellent long-term survival, had been used in each case. The change was in the form of a new porous-beaded surface on the femoral component to induce cementless fixation, which had been used successfully in the fixation of acetabular and tibial components. This modification appeared to have resulted in metallosis due to abrasive two-body wear. The component has subsequently been recalled and is no longer in use. The presentation, investigation, and findings at revision are described and a possible aetiology and its implications are discussed.

We describe the survivorship of the Medial Rotation total knee replacement (TKR) at ten years in 228 cemented primary replacements implanted between October 1994 and October 2006, with their clinical and radiological outcome. This implant has a highly congruent medial compartment, with the femoral component represented by a portion of a sphere which articulates with a matched concave surface on the medial side of the tibial insert.

There were 78 men (17 bilateral TKRs) and 111 women (22 bilateral TKRs) with a mean age of 67.9 years (28 to 90). All the patients were assessed clinically and radiologically using the American Knee Society scoring systems. The mean follow-up was for six years (1 to 13) with only two patients lost to follow-up and 34 dying during the period of study, one of whom had required revision for infection.

There were 11 revisions performed in total, three for aseptic loosening, six for infection, one for a periprosthetic fracture and one for a painful but well-fixed replacement performed at another centre.

With revision for any cause as the endpoint, the survival at ten years was 94.5% (95% CI 85.1 to 100), and with aseptic loosening as the endpoint 98.4% (95% CI 93 to 100). The mean American Knee Society score improved from 47.6 (0 to 88) to 72.2 (26 to 100) and for function from 45.1 (0 to 100) to 93.1 (45 to 100). Radiological review failed to detect migration in any of the surviving knees.

The clinical and radiological results of the Medial Rotation TKR are satisfactory at ten years. The increased congruence of the medial compartment has not led to an increased rate of loosening and continued use can be supported.

We used single-photon emission computed tomography (SPECT) to determine the long-term risk of degenerative change after reconstruction of the anterior cruciate ligament (ACL). Our study population was a prospective series of 31 patients with a mean age at injury of 27.8 years (18 to 47) and a mean follow-up of ten years (9 to 13) after bone-patellar tendon-bone reconstruction of the ACL. The contralateral normal knee was used as a control. All knees were clinically stable with high clinical scores (mean Lysholm score, 93; mean Tegner activity score, 6). Fifteen patients had undergone a partial meniscectomy and ACL reconstruction at or before reconstruction of their ACL.

In the group with an intact meniscus, clinical symptoms of osteoarthritis (OA) were found in only one patient (7%), who was also the only patient with marked isotope uptake on the SPECT scan compatible with OA. In the group which underwent a partial meniscectomy, clinical symptoms of OA were found in two patients (13%), who were among five (31%) with isotope uptake compatible with OA. Only one patient (7%) in this group had evidence of advanced OA on plain radiographs.

The risk of developing OA after ACL reconstruction in this series is very low and lower than published figures for untreated ACL-deficient knees. There is a significant increase (p < 0.05) in degenerative change in patients who had a reconstruction of their ACL and a partial meniscectomy compared with those who had a reconstruction of their ACL alone.