A predictive radiological analysis of short stems versus both shortened and long stems in primary hip replacement: A case-control study of 100 cases of Metha versus ABG II and Omnifit HA at 2-8years' follow-up.

INTRODUCTION: Short hip stems, intended to conserve bone stock and ensure a more physiological distribution of stress in the femur under loading, are meeting with renewed interest. Radiologic semiology is not known exactly, particularly in relation to conventional implants; we therefore conducted a case-control study of 3 types of implant differing only in stem length: short, shortened or long. The aim was: (1) to compare radiographic aspects, (2) to attempt to systematize medium-term radiologic status for the 3 types, and (3) to assess the impact of radiographic aspect on loosening and revision rates. HYPOTHESIS: The short Metha stem is better adapted to the recipient bone than longer stems, without sacrificing stability. MATERIAL AND METHODS: A prospective series comprising the first 100 selected cases of hip replacement using the short Metha stem was compared to two other series of 100 "long" (Omnifit HA) and 100 "shortened" (ABG II) stems at comparable follow-up: 4.05±1.44years (range: 2-8years) for Metha, 4.48±0.97years (range: 2-8years) for Omnifit, and 4.75±2.07years (range: 2-8years) for ABG II. Selection criteria in this initial phase were very strict: young age and/or high activity level, with good bone stock and femoral morphology suited to fitting a Metha stem (no "stovepipe" or "champagne-flute" femurs), for which 12.8% of primary hip replacements were selected. Matching was performed by sampling on criteria of age, gender, body-mass index and etiology. Radiographic parameters were compared between the short stem group and the two control groups and classified according to Engh-Massin score (10 points for fixation and 17 for stability). RESULTS: The short Metha stem provided excellent fixation scores: 7.65/10, versus 7.16 (P=0.003) and 5.92 (P=0.0001) for ABG II and Omnifit, respectively. Likewise, stability was scored 14.23/17 for Metha, vs. 14.51 (NS) and 11.83 (P=0.0001) respectively, and the total score was higher for Metha (21.88/27) than ABG II (21.67; P=0.03) or, more particularly, Omnifit (17.83; P=0.0001). The Metha stem was never associated with thigh pain or periprosthetic fracture. 8-year survival was 100%, without significant difference with respect to ABG II (100%; NS) or Omnifit (98.8%; 95% CI: 0.964-1; NS). DISCUSSION: The apparent radiologic superiority of the short Metha stem requires long-term confirmation in non-selected series. Meanwhile, Metha can be asserted to have demonstrated optimal compromise between lasting bone anchorage and respect of bone physiology under loading at medium term. LEVEL OF EVIDENCE: III, case-control study.

[Results of lumbar and lumbosacral fusion: clinical and radiological correlations in 113 cases reviewed at 3.8 years]. / Résultats des ostéosynthèses-arthrodèses lombaires et lombosacrées: corrélation radio-clinique à propos de 113 cas revus avec 3,8 années de recul moyen.

PURPOSE OF THE STUDY: Spinal fusion requires the use of hardware for reduction and stabilization. We present the clinical and radiological behavior of a population of patients with lumbar and lumbosacral spinal fusion. MATERIALS AND METHODS: Between 1990 and 1992, 113 patients were operated for lumbar and lumbosacral fusion. Mean age of the population was 43 years and mean follow-up was 3.8 years. Most of the fusions were L4-S1 fusions. 56% of the patients had a previous surgery. Thirteen patients in the series were reoperated and analyzed separately. In the majority of the cases, the indication for surgery was back pain with or without leg pain. Diagnoses were: spondylolisthesis, discopathy, scoliosis, and pseudoarthrosis. The spine was fused and reduced using two lordotic rods. Peroperative and postoperative lordosis were calculated on X-rays. Clinical results were analyzed with the Beaujon-Lassale score. RESULTS: Mean improvement was significantly better for spondylolisthesis than for other pathologies (85.6% versus 77.1%). Returning to work was possible for 85.5% of those with improvement and was not possible for 69.8%. The gain achieved in lordosis at surgery was lost at last follow-up. The lordosis of the construct appeared to protect against the development of discopathies above and below the construct. Discopathis led to a poor score. The rate of non-union was 7.9%, the rate of repeated surgery 6.1% and the rate of hardware removal 23.8%. At last follow-up, improvement was achieved in 45.6% of the 13 patients of the series who had repeat surgery. DISCUSSION: The results in our series are similar to those reported by others. Lumbar lordosis is an important factor: if lost, more interbody fusions may be subsequently required. Diagnosis of non-union is difficult and reoperation is the only sure manner to prove it by applying distraction-compression manoeuvres on the screws. All non-unions presented were symptomatic; incidence in the series was thus probably higher. Non-union and reoperation with a longer fusion are perhaps correlated with insufficient elasticity in the osteosynthesis. Optimal rod elasticity is a factor which remains to be defined. CONCLUSION: Clinical results of lumbar and lumbosacral fusions are not unsatisfactory, but in our series almost one patient out of three had to be reoperated. One of the reasons for so many reoperations is certainly hardware rigidity. Hardware was not removed without testing the fusion as this is the only means of sure diagnosis of non-union. Reoperation should not be considered a failure in this difficult surgery of back pain which requires long-term surgical follow-up.

[Our experience with medial unicondylar knee prosthesis]. / Notre expérience de la prothèse unicompartimentaire médiale du genou.

The indications for unicondylar knee replacement for osteoarthritis are discussed. Its theoretical advantages against total knee replacement are the small size of the implants, an easier operative technique and faster rehabilitation. But some surgeons assume that tibial osteotomy has the same indications and gives the same long-term results while preserving the natural joint.52 medial resurfacing unicondylar knee prostheses have been followed for a mean time of 4 years, with a minimum of 1 year. There were 16 males and 31 females, with a mean age of 70 years (range 65 to 83 years). The indication was primary (37 cases) or post-traumatic (2 cases) osteoarthritis and necrosis of the medial femoral condyle (13 cases). The Cartier-Mansat prosthesis was most frequently used.Four prostheses were revised because of unexplained pain (1 case), aseptic loosening (2 cases) or osteoarthritic change in the opposite compartment (1 case). 48 prostheses were still in place at the time of this study. The results were studied with Hungerford's classification, giving the knee joint a maximum score of 100 points.Mean pain score was 42/50, and only 4 patients had significant pain. No abnormal laxity was seen. 45 patients had complete extension, and 3 a deficit between 5 and 10°. 43 patients had over 90° flexion. Quadriceps force was always normal. Only 23 patients had a normal mechanical axis measured on a monopedal stance view, while 19 had a remaining varus deformation and 4 an overcorrection. The mean global score was 85/100, and 33 of the 39 patients with no missing item had a global score over 80/100.Objective and subjective results were very satisfactory. But we were not so pleased with the radiological appearance of the prosthesis. Malposition of both prosthetic components, including global limb malalignment, malalignment of one or both implants, incongruency of both implants or lowering of the joint space with reference to lateral compartment, were very common. The actual influence of these malpositions are not known, but is likely to be significant and the long term survival remains then questionable. A more precise operative technique could decrease the incidence of this disadvantage.