1 Introduction

Total hip arthroplasty is the standard surgical treatment for advanced osteoarthritis (OA) of the hip. However, for early-to mid-stage OA in adolescents and middle-aged adults, joint-preserving procedures are a viable alternative. One of the most representative procedures in this category is periacetabular osteotomy (PAO), first reported by Nishio in 1956.¹ PAO involves spherical osteotomy and rotational displacement of the dysplastic acetabulum to improve femoral head coverage and slow the progression of OA. PAO has several variations, classified based on osteotomy techniques, such as spherical periacetabular osteotomy (SPO), which preserves the inner cortex of the acetabulum while allowing bone fragment displacement.² Alternatively, techniques that include osteotomy of the inner cortex, such as rotational acetabular osteotomy (RAO) and curved periacetabular osteotomy (CPO), have been developed.^3,4

SPO offers the advantage of maintaining pelvic ring stability since it preserves the continuity of the ilium, ischium, and pubis, thereby preventing stress concentration in specific areas. However, a notable drawback is the risk of unintentional chisel penetration into the joint space during osteotomy.²

Conversely, RAO and CPO, which involve osteotomy of the inner cortex, offer several advantages. These include a reduced risk of intra-articular chisel penetration due to complete pubic osteotomy, increased thickness of the displaced fragment, which reduces the likelihood of fragment fractures, and facilitated medialization of the femoral head. However, since pubic osteotomy disrupts the continuity of the anterior column, it may affect the mechanical balance of the entire pelvis. Among the postoperative complications of RAO, posterior column fractures and inferior pubic ramus fractures have been reported, with incidence rates ranging from 2.9 to 18.4%0.^5–8 In some cases, additional surgical fixation is required to alleviate pain.^5–8 The timing of these fractures varies, with reports indicating their occurrence between 1.8 and 3 months postoperatively; however, some cases have been observed as early as 2 weeks post-surgery.^5–8.

Although weight-bearing restriction is generally recommended after PAO, fractures can occur regardless of the initiation of weight-bearing. The specific movements or postures that contribute to these fractures remain unclear. However, based on the authors' experience, multiple cases of fractures have been encountered even before the start of weight-bearing. Despite the potential biomechanical implications of anterior column discontinuity following RAO, this aspect has not been thoroughly investigated.

Therefore, this study aimed to evaluate stress distribution across the pelvis after RAO using finite element analysis (FEA) to elucidate the mechanisms underlying fractures of the posterior column and inferior pubic ramus. We hypothesized that discontinuity at the pubic osteotomy site after RAO is a risk factor for these fractures, with an in-depth analysis presented to test this hypothesis.

2 Material and methods

3 Results

The maximum stress observed in the ischial ramus (white arrow) ranged from 23.6 MPa to 26.9 MPa, with no significant differences across the various models (Fig. 1a). In Fig. 1b, the maximum preoperative stress at the inferior pubic ramus was 11.9 MPa (a). However, in the Nonunion model, the stress increased to 43.0 MPa immediately postoperatively, approximately 3.6 times higher than the preoperative value. At 1 year postoperatively, the stress in the Nonunion model was 23.1 MPa, approximately 1.9 times the preoperative value (c). In the Union model, the immediate postoperative stress was 16.7 MPa, approximately 1.4 times the preoperative value, while the stress at 1 year postoperatively was 11.0 MPa, nearly equal to the preoperative level. Furthermore, a comparison of the immediate postoperative conditions between the Nonunion and Union models showed that the Nonunion model experienced approximately 2.6 times higher stress than the Union model (b vs. d).

Close

Fig. 1– Von Mises Stress Distribution in the Ischial Ramus (Anterior View). a. Preoperative; b. Immediate postoperative (Nonunion model); c. One year postoperative (Nonunion model); d. Immediate postoperative (Union model); e. One year postoperative (Union model).

Export to PPT

Close

Figs. 1–2 Von Mises Stress Distribution in the Inferior Pubic Ramus (Posterior View). a. Preoperative; b. Immediate postoperative (Nonunion model); c. One year postoperative (Nonunion model); d. Immediate postoperative (Union model); e. One year postoperative (Union model).

Export to PPT

In Fig. 1c, regarding the posterior column, the preoperative maximum stress was 10.6 MPa (a). In the Nonunion model, the immediate postoperative stress soared to 53.9 MPa (b, white arrow), approximately 5.1 times the preoperative level. At 1 year postoperatively, the stress in the Nonunion model decreased to 14.4 MPa (c), which was still 1.4 times the preoperative value. In the Union model, the immediate postoperative stress was 34.2 MPa, approximately 3.2 times the preoperative level (d), whereas, at 1 year postoperatively, the stress returned to 11.0 MPa (e), a value nearly equivalent to the preoperative state. The Nonunion model experienced approximately 1.6 times the stress of the Union model immediately postoperatively. No element fractures were observed in the preoperative state, the Nonunion model at 1 year postoperatively, or the Union model at 1 year postoperatively. However, in the Nonunion model immediately after surgery, the posterior column exhibited the highest number of tensile and compressive fractures among the five conditions, and a few tensile fractures were also observed in the inferior pubic ramus. In the immediate postoperative Union model, many elements in the posterior column showed tensile and compressive fractures, but no fractures were observed in the inferior pubic ramus. Overall, tensile fractures occurred more frequently than compressive fractures (Table 3, Supplementary Fig. 5.). Additionally, analysis of element vector directions revealed that loading caused posterior displacement of the posterior column (Supplementary Fig. 6).

Close

Figs. 1–3 Von Mises Stress Distribution in the Posterior Column (Medial View of the Affected Side). a. Preoperative; b. Immediate postoperative (Nonunion model); c. One year postoperative (Nonunion model); d. Immediate postoperative (Union model); e. One year postoperative (Union model).

Export to PPT

Additionally, in regions where element fractures occurred, stress levels exceeded 30 MPa, indicating a high risk of structural failure (Fig. 2).

Close

Fig. 2 Maximum von Mises Stress. At locations where element failure occurred, high stress levels exceeding 30 MPa were detected.

Export to PPT

4 Discussion

Fractures of the posterior column and inferior pubic ramus represent significant complications following RAO, potentially leading to delayed rehabilitation and prolonged pain. In this study, no substantial differences in von Mises stress were observed in the ischial ramus, regardless of the continuity of the superior pubic ramus. However, in the inferior pubic ramus, when the superior pubic ramus was discontinuous, the equivalent stress was approximately 3.6 times higher immediately after surgery than preoperatively, and 1.9 times higher at 1 year postoperatively. Conversely, when pubic continuity was preserved, stress levels immediately after surgery were only 1.4 times higher than preoperative values and returned to nearly identical levels at 1 year.

In the posterior column, stress levels immediately postoperatively were 5.1 times higher than preoperative values, decreasing to 1.4 times higher at 1 year. When pubic continuity was maintained, stress levels immediately postoperatively were 3.2 times higher than preoperative values and returned to preoperative levels by the first year. Furthermore, element failure was observed in areas where stress exceeded 30 MPa, suggesting a clinically relevant risk of fracture. Under normal conditions, load transmission through the obturator ring occurs primarily via the superior pubic ramus rather than the inferior pubic ramus. However, as RAO involves osteotomy of the superior pubic ramus, load transmission is forced to occur exclusively through the inferior pubic ramus and ischium. This increased mechanical burden may contribute to fractures of the inferior pubic ramus or ischium.¹³ Indeed, patients who sustain pubic and ischial fractures after PAO exhibit a significantly higher rate of pubic discontinuity than patients without fractures. Compared to patients with complete bone union, those with pubic pseudoarthrosis exhibit an 11.8 times higher risk of stress fractures in the pubis and ischium.^6,8

The incidence of ischial fractures after PAO has been reported to be approximately 0.9% to 1.0%, making it a relatively rare complication.^5,6,14 This is likely attributable to the strong bony structure of the ischium. However, iatrogenic intraoperative ischial fractures have been reported when the osteotomy extends to the sciatic notch, indicating a possible mechanism for PAO-related ischial fractures.^15–17 Although no previous finite element studies have specifically analyzed stress distribution in the ischial ramus after PAO, our study found that stress in the Nonunion model was 1.1 times higher than that in the Union model, suggesting minimal impact on ischial fractures. This finding aligns with clinical observations indicating that ischial fractures after PAO are uncommon. Conversely, the incidence of inferior pubic ramus fractures after PAO has been reported at 2–4%, although some studies have documented a higher rate (18.4%). An FEA by Imai et al. demonstrated that a pubic discontinuity model exhibited stress in the inferior pubic ramus 2.38 times higher in a standing half-weight-bearing position and 2.02 times higher in a seated position compared to a continuity-preserved model.¹⁸ Similarly, we found that RAO-induced discontinuity of the pubic osteotomy site yielded approximately a 2.6-fold increase in stress on the inferior pubic ramus, suggesting that pubic discontinuity increases fracture risk in both standing and seated postures.

The incidence of posterior column fractures following PAO ranges from 0.5% to 13.9%. However, posterior column stress fractures may be underrecognized as a postoperative complication.^5,8,19 Notably, to the best of our knowledge, no prior studies have conducted finite element analyses of stress distribution in the posterior column after PAO, making our study the first to report on this issue. The present findings indicate that immediately after PAO, stress temporarily increases regardless of pubic continuity. However, in discontinuity models, tensile stress is concentrated on the anterior elements of the posterior column-initiated bone failure, suggesting that posterior column fractures may develop due to posterior column extension under loading in a seated position. Nevertheless, by 1 year postoperatively, the bone union between the mobilized fragment and iliac bone restored preoperative stress levels, effectively eliminating the risk of fracture.

Preventing inferior pubic ramus and posterior column fractures after PAO requires measures to promote bone healing at the pubic osteotomy site. When performing RAO, avoiding delayed bone union at the osteotomy site should be a key consideration. Furthermore, alternative surgical techniques such as SPO, which preserves continuity between the ilium, ischium, and pubis, may be a viable option. In cases where delayed union of the pubic osteotomy site is present, additional precautions should be taken to mitigate fracture risk. Patients should be advised to lower themselves slowly while sitting and avoid sudden impact on the ischium, such as falling onto the buttocks, even during early postoperative non-weight-bearing periods.

To prevent the delayed union of the pubic osteotomy site, preoperative and postoperative changes in the center-edge angle have emerged as significant predictors of postoperative stress fracture risk, suggesting that great corrective displacement and extensive acetabular reorientation may contribute to pubic discontinuity.⁷ Specifically, a gap exceeding 5.1 mm at the pubic osteotomy site is a critical threshold.²⁰ However, in cases of moderate to severe hip dysplasia, achieving adequate acetabular coverage often necessitates extensive correction, making increased osteotomy site gaps unavoidable. To mitigate the risk of delayed union, osteotomy should be performed as close as possible to the pubic base, where cancellous bone is abundant. Conversely, osteotomies performed farther from the joint may result in larger bone defects, impairing bone healing.⁶ If a gap remains postoperatively, bone grafting should be considered to facilitate bone union.

This study has several limitations. First, direct mechanical experiments to validate the FEA results were not included. Second, only one patient's pelvic morphology was analyzed, limiting generalizability. Third, the loading conditions were restricted to static seating, without considering dynamic activities such as walking or stair climbing. Furthermore, physiological (i.e., anterior or posterior) pelvic tilt in the seated posture was not incorporated, which could affect stress distribution vectors. Additionally, factors such as posterior column width and bone quality were not examined, despite their potential impact on fracture risk. Naturally, patients with higher bone density and thicker posterior columns would likely exhibit reduced fracture susceptibility. Moreover, the Union model of the pubic osteotomy site was hypothetically reconstructed rather than derived from actual postoperative CT scans of patients. However, this study was designed in accordance with existing literature and clinical observations, and the locations of stress concentration and the patterns of potential fractures observed in the analysis closely correspond to findings reported in clinical cases after PAO, thereby supporting the validity of the model. Furthermore, we analyzed multiple conditions under consistent mechanical settings—such as the presence or absence of continuity at the pubic osteotomy site and the postoperative time point. This approach revealed consistent trends in stress distribution and fracture risk, indicating that the findings may be applicable beyond individual cases. By using a single, well-defined anatomical model, we were able to eliminate inter-individual variability and precisely control model parameters, thereby clarifying the causal relationships.

In conclusion, this study utilized finite element analysis to evaluate stress distribution in seated postures and assess the risk of fractures following RAO. The findings suggest that loss of continuity at the pubic osteotomy site impairs normal load transmission, thereby increasing mechanical stress on the posterior column and inferior pubic ramus, potentially increasing fracture risk. Particularly, when contact at the iliopubic eminence is lost, yielding pelvic ring discontinuity, fracture risk is elevated in standing and seated postures. Therefore, special attention should be given to the posterior column and inferior pubic ramus fractures following RAO, regardless of patient positioning.

Patient consent

This retrospective study did not obtain individual consent for publication from each patient; however, comprehensive consent for the academic use of clinical information was obtained from all patients upon hospital admission. Furthermore, no identifiable personal information is included in the manuscript.

Ethical approval

This study was conducted at Oita University (Yufu City, Japan) and in accordance with the principles of the Declaration of Helsinki. It was approved by the Oita University Ethics Committee (approval number: 2563; approval date: June 22, 2023). This retrospective study did not obtain individual consent for publication from each patient; however, comprehensive consent for the academic use of clinical information was obtained from all patients upon hospital admission. Furthermore, no identifiable personal information is included in the manuscript.

Data statement

The datasets generated and/or analyzed during the current study are not publicly available due to patient confidentiality and institutional regulations, but are available from the corresponding author on reasonable request.

Ethical approval statement

This study was conducted at Oita University (Yufu City, Japan) and in accordance with the principles of the Declaration of Helsinki. It was approved by the Oita University Ethics Committee (approval number: 2563; approval date: June 22, 2023).

Credit author statement

Tsuguaki Hosoyama: Conceptualization, Methodology, Formal analysis, Validation, Writing – Original Draft, Writing - Review & Editing, Funding acquisition.

Nobuhiro Kaku: Supervision, Project administration.

Yutaro Shibuta: Investigation, Resources.

Kensei Tanaka: Software, Investigation, Data Curation, Visualization.

Funding

This work was supported by Teijin Nakashima Medical Co., Ltd., including travel expenses related to the study and English proofreading services [grant number 2024-A18]. These funders had no role in the study design; in the collection, analysis or interpretation of data; in the writing of the report; or in the decision to submit the article for publication.