La statique pelvienne et lombaire de 49 patients opérés pour changement de prothèse totale de hanche entre septembre 2000 et mars 2002 a fait l'objet d'une étude prospective. Cinq patients ont présenté une luxation postopératoire, malgré l'absence apparente de défaut de positionnement des implants. La valeur moyenne de la pente sacrée était significativement différente (p = 0,006) chez les patients s'étant luxés en comparaison avec ceux indemnes de luxation. Cette différence restait significative (p = 0,017) si on limitait l'étude aux 33 patients ne présentant aucun facteur de risque associé de luxation postopératoire (antécédents de luxation récidivante, ou d'interventions multiples, ou infectieux, pseudarthrose du grand trochanter).

Notre étude évoque le rôle de la statique pelvienne et lombaire dans la survenue de luxations postopératoires après reprise de prothèse totale de hanche qui doit être analysée une fois les causes classiques éliminées. Elle permet enfin, d'isoler des sujets à risque, chez qui des artifices techniques anti-luxation peuvent être discutés.

Hip dislocation after revision total hip arthroplasty (RTHA) is a frequent post-operative complication. Certain risk factors are well identified (nonunion of the greater trochanter, history of recurrent dislocation or infection, multiple procedures), the role of spine morphological remains to be fully examined. The purpose of this prospective analysis was to assess the role of spine morphology in post-operative dislocations.

Forty-nine patients who underwent RTHA between September 2002 and March 2002 were evaluated prospectively. A complete pre- and postoperative spinal work-up was available for all patients to evaluate the lumbopelvic static using the Legaye and Duval Beaupère morphology criteria and the pelvic-femoral angle to assess hip joint extension. Pre-, per- and post-operative data including the usual risk factors for dislocation related to the clinical situation and the technique used were recorded on a digital datasheet. Five patients developed postoperative dislocation despite the absence of defective implant position. There was no relation with access (p = 0.832) or pelvic-femoral angle (p = 0.515).

The mean value of the sacral slope was significantly different (p = 0.006) in patients who developed dislocation in comparison with the other patients. This difference remained significant (p = 0.017) for the cohort of 33 patients who had no associated risk factor for postoperative dislocation (history of recurrent dislocation or infection, multiple procedures, tight nonunion of the greater trochanter).

Our results suggest that the morphology of the lumbar spine can be involved in the risk of postoperative dislocation. The morphology of the lumbar spine affects the pelvic static and thus the landmarks usually used for implantation, but it can also limit the amplitude of pelvic movement when changing from the sitting to the standing position, which would be compensated for by greater hip movement, particularly extension. The method we used did not fully take into account the consequences of changes in spinal balance due to thoracic deformations nor to analgesic (or not) hip flexion and subsequent deformation of the lumbar spine.