The Bernese periacetabular osteotomy has been performed for approximately 30 years. In that time, it has proven to be a technically complex procedure with the potential to considerably improve the natural history of the dysplastic hip. Significant refinements in the surgical technique combined with the recognition of femoroacetabular impingement, improvements in the understanding of hip biomechanics and acetabular orientation, and more discerning patient selection have improved the outcomes related to this procedure. Although the surgery is complex and the potential for complications exists, the results are reproducible and the technique has become the gold standard for acetabular reorientation.
- •
The Bernese periacetabular osteotomy (PAO) allows for improved position of the weight-bearing acetabular hyaline cartilage. It allows joint medialization and version reorientation while the posterior column is left intact.
- •
The procedure is indicated in patients who have minimal hip arthrosis and symptomatic hip dysplasia with a concentric hip joint.
- •
The procedure has a significant learning curve but has a reproducible technique with a low complication rate while prolonging the natural hip joint.
- •
Acetabular reorientation with a PAO can also improve symptoms from pathologic acetabular retroversion or anteversion.
Introduction
The goal of any acetabular osteotomy, regardless of the specific technique, is to change the pathologic mechanics of the hip that lead to intra-articular damage, pain, and osteoarthritis. This generally involves improving the coverage of the femoral head and/or changing the orientation of the acetabulum. Reorientation of a dysplastic acetabulum increases the load-bearing surface area of the joint while maintaining or improving joint stability. The Bernese PAO was developed by Reinhold Ganz and colleagues in 1983. Compared with other surgical techniques for acetabular reorientation in use at the time, it involves a series of reproducible hexagonal cuts around the acetabulum that leave the posterior column of the pelvis intact. Since the original description, the technique and understanding of the biomechanics have continued to evolve, such that Siebenrock and colleagues wrote in 2001, “Our understanding of what is an optimal correction has improved considerably over time. It is a balancing of the maloriented horseshoe-shaped acetabular cartilage over the femoral head, which leads to an optimal use of a limited area of hyaline cartilage for weight-bearing.”
The Bernese PAO has several advantages compared with other acetabular reorientation osteotomies. Specifically, the posterior column of the pelvis remains intact, which maintains the stability of the pelvis and allows early patient mobility. Before the development of the PAO, osteotomy techniques violated the posterior column, requiring either a period of spica cast immobilization or more-extensive pelvic fixation, with an inherent risk of nonunion at the osteotomy site. Because the osteotomy is close to the joint, there is no change in the dimensions of the true pelvis. As a result, vaginal childbirth is still safe for these patients, which is not the case for adult patients who have had double or triple pelvic osteotomies. The proximity of the osteotomy to the joint also allows for potentially powerful correction of the bony mechanics. In particular, because the joint is medialized, the lever arm of the abductors improves and the joint reactive forces decrease. Finally, because all of the osteotomy cuts are made from the inner aspect of the pelvis, the abductors can be preserved. The curved or spherical PAO, which is more common in Asia, has some of the same advantages. With this technique, however, the osteotomy fragment is smaller; it may be more susceptible to osteonecrosis and the medialization of the hip center may be more limited. Nonetheless, the Bernese PAO is technically complex and there is a substantial surgical learning curve to the procedure.
Anatomy
The original technique was developed in 1983 and published in 1988 with 1-year follow-up results from the first 75 patients. The bone cuts were developed based on knowledge of the blood supply to the acetabular fragment ( Figs. 1 and 2 ). Briefly, the pelvis is approached through either a direct anterior or modified Smith-Petersen approach, taking care to preserve the abductors. The ischium is cut just inferior to the infracotyloid notch; the pubis is cut adjacent to the acetabulum; and the supra-acetabular osteotomy begins in the region of the anterior superior iliac spine (ASIS). Just lateral to the acetabular brim, the cut is angled 120° to meet the ischial osteotomy. The endosteal blood supply to the fragment is disrupted by the osteotomy, but fragment perfusion is maintained by 2 branches of the superior gluteal artery, the acetabular branch of the inferior gluteal artery and the acetabular branch of the obturator artery. There is additional contribution of the capsular blood supply, unless the osteotomy is performed too close to the capsule.
Investigation of the effects of PAO in various patient populations has been reported in the literature. Many patients with dysplasia are young, female, and of childbearing age. The impact of the PAO on the dimensions of the true pelvis, pregnancy, and childbirth has been studied. Study results reveal no change in the anteroposterior (AP) inlet, midpelvis, bispinous diameter, or transverse inlet measurements after PAO. Clinically, women who have had prior PAOs report vaginal deliveries and cesarean sections for reasons unrelated to the osteotomy. In one series, the rates of cesarean section were increased above population averages. Although firm conclusions could not be drawn, the increase was attributed in part to obstetrician concerns for the patients’ prior osteotomies. Patients who had been sexually active after a PAO reported both increased and decreased sexual activity secondary to changes in pain and range of motion after the surgery. Patients who became pregnant after having a PAO reported back and hip pain during pregnancy that resolved after childbirth. The authors of this study of PAO patients reporting pain with pregnancy observed a correlation of patient pain and range of motion to acetabular retroversion and probable impingement.
Anatomy
The original technique was developed in 1983 and published in 1988 with 1-year follow-up results from the first 75 patients. The bone cuts were developed based on knowledge of the blood supply to the acetabular fragment ( Figs. 1 and 2 ). Briefly, the pelvis is approached through either a direct anterior or modified Smith-Petersen approach, taking care to preserve the abductors. The ischium is cut just inferior to the infracotyloid notch; the pubis is cut adjacent to the acetabulum; and the supra-acetabular osteotomy begins in the region of the anterior superior iliac spine (ASIS). Just lateral to the acetabular brim, the cut is angled 120° to meet the ischial osteotomy. The endosteal blood supply to the fragment is disrupted by the osteotomy, but fragment perfusion is maintained by 2 branches of the superior gluteal artery, the acetabular branch of the inferior gluteal artery and the acetabular branch of the obturator artery. There is additional contribution of the capsular blood supply, unless the osteotomy is performed too close to the capsule.
Investigation of the effects of PAO in various patient populations has been reported in the literature. Many patients with dysplasia are young, female, and of childbearing age. The impact of the PAO on the dimensions of the true pelvis, pregnancy, and childbirth has been studied. Study results reveal no change in the anteroposterior (AP) inlet, midpelvis, bispinous diameter, or transverse inlet measurements after PAO. Clinically, women who have had prior PAOs report vaginal deliveries and cesarean sections for reasons unrelated to the osteotomy. In one series, the rates of cesarean section were increased above population averages. Although firm conclusions could not be drawn, the increase was attributed in part to obstetrician concerns for the patients’ prior osteotomies. Patients who had been sexually active after a PAO reported both increased and decreased sexual activity secondary to changes in pain and range of motion after the surgery. Patients who became pregnant after having a PAO reported back and hip pain during pregnancy that resolved after childbirth. The authors of this study of PAO patients reporting pain with pregnancy observed a correlation of patient pain and range of motion to acetabular retroversion and probable impingement.
Evolution of the technique
Since its initial description, the surgical technique has undergone various modifications. Although a Smith-Petersen approach to the anterior aspect of the pelvis was part of the original technique, the abductors were stripped from the iliac wing to perform the supra-acetabular osteotomy. This has evolved such that the abductors are largely left intact when the osteotomy is performed. In addition to preserving the abductor function, protecting the abductors preserves the obturator, superior and inferior gluteal arteries, and the capsular contribution to acetabular perfusion, decreasing the risk of acetabular osteonecrosis. Initially, the bone cuts were performed from both sides of the iliac wing; however, to preserve the abductors, the bone cuts have been changed and are now mostly performed from the inner aspect of the pelvis. More recently, it has become apparent that hip flexion strength is decreased for up to 2 years postoperatively. Thus, a rectus-sparing approach, which leaves the direct and indirect heads of the rectus femoris attached to the ilium, has been advocated for preserving hip flexor strength.
The recognition that femoroacetabular impingement (FAI) could be responsible for continued pain after a PAO was an important development. The femoral head in a dysplastic hip is known to have an elliptical shape with decreased head-neck offset and lateral flattening from a hypertrophic gluteus minimus. When the acetabulum is reoriented such that there is excess lateral or anterior coverage, FAI can occur. As a result, an arthrotomy has been incorporated into the surgical technique for evaluation of impingement. When necessary, a femoral-head neck osteoplasty is performed through the arthrotomy. Better recognition of acetabular version during the correction also helps minimize FAI because the acetabulum can be excessively retroverted, resulting in pincer-type FAI or undercoverage of the posterior aspect of the hip.
Other modifications to the original surgical technique have been described and adopted to varying degrees. One modification was the description of a 2-incision technique so that the ischial osteotomy could be performed under direct visualization. The disadvantage of this technique, as it was described, is that it requires takedown of the external rotators posteriorly. This may endanger the medial femoral circumflex artery, which is the primary blood supply to the femoral head. There may also be difficulty with the correction and assessing fragment version because of patient positioning and loss of some of the soft tissue tension. Additional variations to the technique include intraoperative fluoroscopy to judge the position of the osteotomies and minimally invasive incisions.
Recently, some investigators have also begun performing arthroscopy at the time of PAO to evaluate the articular cartilage and refix or stabilize the labrum. The clinical outcome of labral refixation in this setting is, as yet, unknown.
Biomechanics
Computer modeling and finite element analysis haves allowed the biomechanical effects of PAO to be quantified. On average, a PAO increases the load-bearing surface of the joint by approximately 50%, with a range of 35% to 70% depending on the degree of correction. This is comparable to the load-bearing surface for normal control patients. The increased femoral head coverage causes improved load distribution, with near-normal contact stresses observed in 3-D finite element analysis. Contact pressure analysis has also revealed that the amount of correction required for an optimal decrease in load varies among patients. For patients with a typical pattern of dysplasia requiring lateral and anterior coverage, the largest reduction in cartilage contact pressures occurred when the fragment was adducted and extended.
Acetabular reorientation also improves hip stability. The dysplastic hip is both statically and dynamically unstable, with superior, posterior, and lateral movement of the femoral head with walking, compared with hips with normal morphology. In a recent study, the dynamic instability was restored to near-normal values after acetabular reorientation. In a computational model of hip dysplasia, lateral hip stability improved to normal values when the center-edge angle measured 25° to 30°. The same model observed that anterior and posterior femoral head stability correlated with the degree of acetabular version.
Improvement in hip biomechanics can also be observed clinically. For example, redistribution of the acetabular load may result in a change in the bone density. As such, acetabular bone density after PAO has been assessed with both CT and dual energy x-ray absorptiometry (DEXA) scanning. Anteromedial bone density was found increased 2 years postoperatively, but there was no change in lateral bone density. An underpowered study using DEXA scanning to assess acetabular bone density 2.5 years after PAO failed to observe a difference in density. The investigators concluded that DEXA scanning did not provide enough resolution to adequately assess bony changes after a PAO. Nonetheless, the redistribution of load has some effect on the subchondral bone, as evidenced by acetabular cysts that healed after acetabular reorientation. Furthermore, the degree of healing also correlated with the degree of correction and the clinical outcome score. Close examination of the articular surface revealed that cartilage thickness and the shape of the articular surface are both preserved in short-term follow-up (1–2.5 years). Finally, gait and strength comparative analysis preoperatively and postoperatively reveals that abductor strength was better 1 year postoperatively but flexion was still weak compared with healthy controls. This may be the reason that the increased knee flexion seen in preoperative gait analysis improved but did not entirely resolve.
Indications and contraindications
The most accepted indication for acetabular reorientation is mild to moderate symptomatic dysplasia. Initially, there was controversy about the degree of dysplasia and concomitant femoral head deformity that can be adequately addressed with a PAO ; it was believed that severely subluxed hips would require salvage osteotomies instead. Subsequently, good outcomes have been published for more severe deformities and the indications have been expanded. Thus, dysplasia secondary to Legg-Calvé-Perthes disease and dysplasia secondary to flaccid and spastic neuromuscular disorders are considered appropriate indications for PAO. Global acetabular retroversion causing impingement is also considered an indication for PAO, particularly if the retroversion is associated with posterior wall deficiency or posterior instability.
There may also be a role for acetabular reorientation in patients with borderline dysplasia (center-edge angle of Wiberg of 20°–25°) and clinically symptomatic instability. This occurs most often in the setting of increased femoral and/or acetabular anteversion ( Fig. 3 ) and primarily manifests as iliopsoas or abductor fatigue symptoms. Although there are biomechanical data demonstrating that hip stability and acetabular load are related to acetabular version, there are as yet no clinical outcomes data for this patient population.
Midterm and long-term follow-up have established that the most predictable outcomes are in patients under 35 years old at the time of surgery with little to no arthritis (Tönnis grade 0 or 1) on plain radiographs. Good results have, however, been reported for patients over age 35 or 40, even in early series, with the caveat that the hip was well preserved and concentric on preoperative imaging.
Study results of PAO in patients with preoperative arthrosis reveal that Tönnis grades 2 and 3 are greater predictors of failure. A cost-efficacy analysis found that total hip arthroplasty was preferable for Tönnis grade 3 arthrosis but that in grade 1 or 2 arthrosis, PAO was the more appropriate treatment. Although the presence of grade 2 or 3 arthrosis is a predictor of failure, some patients with this degree of arthrosis show improved outcomes scores and relatively preserved joint space in mid-term follow-up. Thus, although arthrosis is a relative contraindication to PAO, PAO may be preferable to total hip arthroplasty in certain younger patients.
Contraindications to PAO include incongruence on functional radiographs (abduction and internal rotation images or flexion false profile images), which indicates potentially worse congruency after acetabular reorientation and is thus a predictor of worse outcomes. This can occur in nonspherical femoral heads or when the acetabular radius is smaller than the femoral head radius. PAO is also contraindicated in young patients (<10 or 11 years of age) because of the risk of injury to the triradiate cartilage, through which the PAO bone cuts are made. In series of patients with post-traumatic acetabular dysplasia secondary to an acetabular fracture and subsequent growth disturbance, the oldest affected patients were approximately 10 years of age at the time of the fracture. Significant growth disturbances have not been observed in adolescents with open triradiate cartilage who sustain acetabular fractures. Thus, there seems little risk of injury to the triradiate cartilage after age 10 or 11, which is the lower age limit for a PAO.
Preoperative planning
The history and physical examination should focus on determining whether a patient’s symptoms are primarily from dysplasia and static overload, from FAI, or from both. Pain originating from the spine, pelvis, or muscle should be ruled out as part of the initial evaluation. Symptoms characteristic of dysplasia include pain independent of motion and peritrochanteric muscular fatigue after standing or prolonged periods of walking. Symptoms more characteristic of FAI include sharp anteromedial groin pain and pain with prolonged sitting. Patients with dysplasia and an acetabular rim lesion, however, may present with sharp anterior groin pain. On examination, patients with dysplasia and true abductor weakness may have a Trendelenburg gait and Trendelenburg sign. Full or increased hip range of motion should be expected for patients with normal acetabular cartilage. Also, when deciding between FAI and dysplasia of the hip, patients with instability have more hip flexion and internal rotation at 90° of flexion. They may also have a positive apprehension test (pain with extension and external rotation). Because the acetabulum is often anteriorly deficient, these patients may also have psoas irritation or snapping.
Initial radiographs should include a standing AP pelvis radiograph, 45° or 90° Dunn lateral view, a false profile view, and a functional view with the affected hip abducted and internally rotated. The AP pelvis radiograph is used to assess lateral coverage and acetabular version; it is important that the radiograph be obtained with the pelvis in neutral flexion and rotation. In the neutral position, the obturator foramen and teardrops appear symmetric and the coccyx is directly in line with the pubic symphysis. The distance between the coccyx and pubic symphysis is a marker of pelvic tilt and measures 1 cm to 3 cm in a correctly obtained radiograph. The Dunn lateral is used to assess the femoral head-neck offset. In patients with insufficient offset, there may be coexistent FAI and dysplasia; furthermore, these patients may have symptoms of FAI after acetabular reorientation. The false profile view is used to evaluate anterior acetabular coverage, whereas the abduction and internal rotation view is used to assess hip congruency after a potential correction.
CT scanning serves as a useful addition to plain radiographs for preoperative planning. 3-D reconstructions improve the assessment of acetabular version and socket depth. In addition, the prominence of the anterior inferior iliac spine (AIIS) can be assessed with a CT scan ( Fig. 4 ). After reorientation of the acetabulum, a prominent AIIS can contribute to extra-articular impingement. An MRI provides a better evaluation of the labrum and cartilage. For patients with mild or questionable pre-existing arthrosis, cartilage-specific sequences, such as delayed gadolinium-enhanced MRI of cartilage (dGEMRIC), T1rho, or T2 mapping, as described in a previous article, may play a role in surgical decision-making. Pre-existing intra-articular damage is a negative prognostic factor, which may have an impact on surgical decision-making and is important to take into account in managing patient expectations.
Authors’ preferred technique
PAO can be performed with either general anesthesia or under combined spinal-epidural anesthesia and sedation. One advantage of the spinal-epidural anesthesia is that the epidural can be left in place for postoperative pain relief. Patients may opt to predonate 1 unit of autologous blood and receive it back the morning after surgery. Intraoperatively, a cell saver is also used.
If adjuvant arthroscopy is performed for labral repair or débridement and cartilage assessment, the arthroscopy is performed first with the patient on a traction table using a standard technique. The combination of arthroscopy and PAO in the same setting has been previously described, but outcomes data are not yet available. Nonetheless, PAO patients with intact labra had better long-term outcomes scores and a lower risk of arthrosis progression.
For the PAO, patients are positioned supine on a standard radiolucent table. All bony prominences are well padded and the ipsilateral arm is positioned such that it does not impede the placement of the chisel or screws intraoperatively. The senior author’s preference is to place it in a 90°–90° position of abduction and external rotation. The entire operative leg is prepped and draped. The incision is slightly curved lateral to the ASIS and the tensor-sartorius interval ( Fig. 5 ). The fascia over the tensor fascia lata is opened to expose the medial aspect of the muscle to its insertion at the pelvis between the ASIS and the AIIS. The deep lateral rectus fascia is opened and the rectus is retracted medially. The lateral floor of the rectus fascia is then visualized. Often the transverse vessels of the lateral femoral artery are visualized deep to this fascia, which marks the distal extent of the exposure. Opening the floor of the rectus fascia then exposes the lateral aspect of the iliocapsularis muscle. The interval between the hip capsule and the iliocapsularis is developed as the iliocapsularis is retracted medially. Next the ASIS is osteotomized and reflected medially, preserving the sartorius attachment on the osteotomized fragment. Care is taken near the ASIS because proximally the lateral femoral cutaneous nerve emerges within 5 cm of and medial to the ASIS and can be injured during the approach. Proximally, the external oblique aponeurosis is sharply incised and the iliacus is elevated subperiosteally and medially off the iliac wing. At this point in the procedure, flexing the leg to 45° decreases the soft tissue tension and facilitates further dissection.
Related posts:
Surgery for Residual Femoral Deformity in Adolescents
Periacetabular Osteotomy for Hip Preservation
Femoral Considerations for Total Hip Replacement in Hip Dysplasia
Technical Considerations in Total Hip Arthroplasty After Femoral and Periacetabular Osteotomies
Surgery for Residual Femoral Deformity in Adolescents
Technical Considerations in Total Hip Arthroplasty After Femoral and Periacetabular Osteotomies
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
