Background and purpose — Modular-neck hip stems have been identified with corrosion-related problems from the neck–stem junction. We report an ongoing varus deformity of modular-neck hip stems with simultaneous metal ion release observed during a study comparing the migration of modular vs. standard hip stems. Patients and methods — We followed 50 patients with modular and 25 with standard neck stems using radiostereometry (RSA). At 5-year follow-up, we noted a compromised integrity of the modular stem with varus deformity in the neck–stem interface. Changes in head–tip distance as well as whole-blood ion concentration and MRI findings were analyzed. The modular stems were followed further up to 8 years. Results — The head–tip distance decreased continuously by 0.15 mm per year resulting in 1.2 (95% CI 1.0–1.4) mm at 8 years for modular stems, while for the standard stems at 5 years, the decrease was 0.09 (CI 0.0–0.2) mm or 0.02 mm/year. For the modular stems, the reduction in head–tip distance correlated to the increase in whole-blood cobalt concentration at 8 years but not to the MRI grading of tissue reactions. At 5 years, cobalt levels were 4.9 µg/L for modular stems and at 8 years 4.8 µg/L, whereas for standard stems this was 1.0 µg/L. After 8 years, 9 of 72 stems had been revised for different reasons, but only 1 with obvious adverse local tissue reaction (ALTR). Interpretation — We present a surprisingly large progressive deformation at the modular neck–stem junction, but so far without a definite clinical problem. Even the femoral head seems to show slight compression onto the taper over time. A high rate of revisions for the modular type of this stem has raised general concerns, and it has been recalled from the market.

Background: There is an increasing interest in shorter, proximally loading, uncemented femoral stems. The aim of this study was to evaluate a new short stem type, its design based on a well-proven precursor, but with several new design features to facilitate easier insertion and possible preservation of proximal bone stock. It is available with or without a collar. Methods: The study includes 50 patients with primary osteoarthritis, mean age 59 (range 36–75) years, randomised to receive either the collar-fitted or the collarless stem. The patients underwent repeated radiostereometry (RSA) examinations (0, 0.5, 3, 12, 24 months), conventional radiography and filled out both hip-specific (HOOS) and general health (EQ-5D) questionnaires. Results: There was no statistically significant difference between the collared and collarless stems regarding migration and clinical outcome. Both stem types exhibit a similar pattern of initial subsidence (collared 0.63 mm vs. collarless 0.75 mm [ p = 0.50]) and retroversion (collared 0.71° vs. collarless 0.97° [ p = 0.36]) up to 3 months followed by stabilisation, in similarity with its precursors, suggesting good osseointegration. No stem has been revised or considered loose. Conclusions: The new design features, including shortening the stem, do not compromise migration pattern or osseointegration, regardless collar or not.