Retrotilt of the Pelvis During Periacetabular Osteotomy: How to Avoid a Systematic Error Resulting in Acetabular Retroversion and Possible Femoroacetabular Impingement.

BACKGROUND: Pelvic tilt directly influences acetabular version on radiographs. Changes of pelvic tilt potentially affect acetabular reorientation after periacetabular osteotomy (PAO).

PURPOSE: (1) To compare the ratio of the pubic symphysis height to the sacroiliac width (PS-SI) between hips with dysplasia and acetabular retroversion, uni- and bilateral PAO, and male and female patients. (2) To evaluate pelvic tilt (quantified using the PS-SI ratio) in patients after PAO by tracking it from preoperative to intra- and postoperative and short- and middle-term follow-up.

STUDY DESIGN: Case series; Level of evidence, 4.

METHODS: A retrospective and radiographic study was conducted evaluating pelvic tilt in 124 patients (139 hips) with dysplasia and 46 patients (57 hips) with acetabular retroversion who were undergoing PAO (January 2005-December 2019). Patients were excluded if they had insufficient radiographic data, previous or concomitant hip surgery, posttraumatic or pediatric deformities, or combined dysplasia and retroversion (90 patients, 95 hips). Dysplasia was defined as a lateral center-edge angle <23°; retroversion was defined by simultaneous appearance of a retroversion index 30% and positive ischial spine and posterior wall signs. Anteroposterior pelvic radiographs were taken in the supine position preoperatively, during PAO, postoperatively, and at short- and middle-term follow-up (mean ± SD [range]; 9 ± 3 weeks [5-23 weeks] and 21 ± 21 weeks [6-125 months]). The PS-SI ratio was calculated at 5 observation periods (preoperatively to middle-term follow-up) for different subgroups (dysplasia vs retroversion, uni- vs bilateral surgery, male vs female) and validated with intra- and interobserver agreement (intraclass correlation coefficients, 0.984 (95%CI, 0.976-0.989) and 0.991 (95% CI, 0.987-0.994), respectively).

RESULTS: The PS-SI ratio differed between dysplasia and retroversion at all observation periods ( P = .041 to P < .001). Male dysplastic hips had a lower PS-SI ratio when compared with female dysplastic hips at all observation periods ( P < .001 to P = .005). In hips with acetabular retroversion, the PS-SI ratio was lower in men than women at short- and middle-term follow-up ( P = .024 and .003). No difference was found between uni- and bilateral surgery ( P = .306 to P = .905) except for short-term follow-up in dysplasia ( P = .040). The PS-SI ratio decreased in all subgroups preoperatively to intra- or postoperatively ( P < .001 to P = .031). At short- and middle-term follow-up, the PS-SI ratio increased as compared with intraoperatively ( P < .001 to P = .044) and did not differ from preoperatively in all subgroups ( P = .370 to P = .795).

CONCLUSION: A lower PS-SI ratio was found for male or dysplastic hips. In all subgroups, the PS-SI ratio decreased during surgery, indicating retrotilt of the pelvis. Correct pelvic orientation during surgery is crucial for accurate acetabular reorientation. Retrotilt during surgery results in underestimation of acetabular version and iatrogenic retroversion of the acetabulum at follow-up, with the pelvis in the correct and more forward-tilted orientation. Not taking into account retrotilt during PAO potentially results in femoroacetabular impingement. Therefore, we changed our intraoperative setting with adjustment of the central beam to compensate for retrotilt of the pelvis.