Purpose: Serum cobalt and chromium levels after metal-on-metal implantations are not well known. There has been little data on the correlation with clinical and radiological surveillance. This prospective analysis followed the changes in serum cobalt, chromium, and titanium levels in order to ascertain the behaviour of the femoral implant and the bearing.

Material and methods: The series included 292 patients followed for 27–72 months who were free of kidney failure, were not given vitamin B12 supplementation, and did not have occupational exposure. The same cemented implants were used in all patients: titanium femoral stem, Metasuly heads (28 mm). Serum samples were drawn preoperatively and at three, six and twelve months then annually (systematic activity questionnaire). The limit for detection of serum cobalt and chromium was 1 nmol/l (0.05μg/l); for titanium it was 30 nmol/l (1.4m/l).

Results: The main problems encountered were two impingements and two femoral loosenings. Four patterns were identified in the time courses of serum cobalt and chromium. In decreasing order, they were: type 1, low initial level (< 50 nmol/l) then no change; type 2, high initial level (> 50 nmol/l) then decrease; type 3 low initial level then progression; type 4, high initial level then further elevation. Bilateral prostheses showed a particular pattern with elevation after the second implantation. The serum levels rapidly returned to normal after revision in the two impingement cases. The titanium levels were correlated with femoral problems which were not initially detected on the x-rays.

Discussion: Correlation between serum cobalt level and wear is difficult to establish. There was no specific pattern after dislocation. The kinetics showed certain patterns with possible prognostic significance: groups 1 (metallic silence) and 2 (breaking-in) would correspond to favourable evolution; type 3 would be difficult to interpret (missed impingement, foreign body, articular decoaptation or major change in activity level); type 4 is highly suggestive of premature wear or a biological problem (excepting bilateral implants). It is important to monitor serum cobalt and titanium simultaneously to detect an interface problem and or femoral loosening.

Purpose: A significant increase in serum cobalt level has been reported after metal-on-metal total hip arthroplasty with wide individual variability related to activity level, mechanical conditions of the implant, and urinary elimination of cobalt. We studied serum cobalt levels over time to further analyse these factors.

Material and methods: The Metazul® prosthesis was implanted in 119 patients (72 men and 47 women, 12 bilateral implantations) (131 implants). We selected 50 patients (27 men and 23 women, mean age 53 years) who had two blood samples after the procedure allowing an assessment of the serum cobalt kinetics. Other chromium-cobalt implants, vitamin B12 intake, renal failure, or haematological disorders were recorded. An activity questionnaire was filled out by the patients at the time of the blood sample. Samples were drawn with a special kit to avoid metal contamination. The detection limit was 1 nmol/L (0.06 μg/L) with direct electrothermic atomic spectrometric absorption.

Results: In the overall series, serum cobalt level was 44 nmol/L for a physiological level in a control population of 4.28 nmol/L. The difference was significant (p < 0.0001) between the levels observed before surgery and after 18 months implantation. There was no significant correlation with the indication for arthroplasty, presence of dislocation or subdislocation, functional outcome or radiographic findings. Activity level the week before sampling did not influence the results. For the 50 cases evaluated longitudinally, four groups of patients could be identified. The first group (29 patients) had a serum cobalt level below 50 nmol/L over the entire study period. The second group (nine patients) had a level greater than 50 nmol/L followed by a decline ending with a final level below 50 nmol/L. In the third group (six patients) serum cobalt was greater than 50 nmol/L with no trend to a decline. In the fourth group (six patients) the cobalt levels were very high (greater than 150 nmol/L).

Discussion: The six patients in the fourth group were very particular. There were three patients with secondary bilateral implants with a late peak in serum cobalt, one with an impingement on the acetabular rim, one with renal failure, and one who had a very high level of physical activity. The first group had what appears to be a favourable course, similar to the second group where a stabilisation phenomenon could be operating. An explanation in the third group is difficult but could involve a third segment abrasion phenomenon.

Conclusion: Longitudinal analysis of serum cobalt levels provides more information than point measures in patients with metal-on-metal arthroplasties. Intercurrent mechanical phenomena can be detected; unexpected behaviour of the metal-on-metal junction can be suspected in certain patients.