Does Previous Osteotomy Compromise Total Hip Arthroplasty? A Systematic Review

The Journal of Arthroplasty 30 (2015) 79–85

Contents lists available at ScienceDirect

The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org

Does Previous Osteotomy Compromise Total Hip Arthroplasty? A Systematic Review Stephen Duncan, MD a, Scott Wingerter, MD b, Angela Keith, MS b, Susan A. Fowler, MLIS b, John Clohisy, MD b a b

University of Kentucky, Lexington, Kentucky Washington University School of Medicine, St. Louis, Missouri

a r t i c l e

i n f o

Article history: Received 19 September 2013 Accepted 17 August 2014 Keywords: total hip arthroplasty femoral osteotomy pelvic osteotomy dislocation aseptic loosening

a b s t r a c t Hip osteotomy surgery has increased over the past several years, yet the impact of these procedures on subsequent total hip arthroplasty (THA) remains controversial. The purpose of this study was to perform a systematic review of the literature to determine the clinical results, procedure complications, and survivorship of THA following previous hip osteotomy. Ten studies met inclusion criteria. The operative time and estimated blood loss were higher in the post-osteotomy cohorts; while the clinical results and survivorship between groups were similar. THA following previous pelvic and femoral osteotomy provides pain relief and improved function with similar complication rates, clinical outcomes, and survivorship compared to hips undergoing routine primary THA. These procedures can be technically more demanding with increased operative times and intraoperative blood loss. © 2014 Elsevier Inc. All rights reserved.

The management of pre-arthritic and early arthritic hip conditions can be challenging due to young patient age at presentation, highly variable structural deformities, and an increased risk for secondary hip osteoarthritis over time. Acetabular osteotomies such as the Bernese periacetabular osteotomy (PAO) are performed to correct structural deformities, improve the biomechanical environment of the hip, improve pain and function, and potentially delay the need for total hip arthroplasty (THA) [1]. One long-term study indicates that conversion to total hip arthroplasty (THA) following PAO occurs in 38% of hips at twenty years [2]. Performing THA following previous pelvic osteotomy can be technically challenging with the altered anatomy but good results have been reported [3]. Performing total hip arthroplasty (THA) in patients with these complex deformities who did not have previous osteotomies can also be technically challenging and associated with increased complication rates compared to routine primary THA [4]. Proximal femoral osteotomies have also been performed to address a variety of deformities about the hip with decreased pain, improved function, and joint preservation for most patients [5–12]. Performing a THA following femoral osteotomy has been met with mixed results for long-term survival with some showing no difference and others showing poorer results [13–16]. Technically, performing the THA following the osteotomy is challenging, and various rates for intraoperative complications have been reported [13,16–20].

Funding: None of the authors received any funding to support the preparation and writing of this manuscript. The Conflict of Interest statement associated with this article can be found at http://dx. doi.org/10.1016/j.arth.2014.08.030. Reprint requests: Stephen T. Duncan, MD, University of Kentucky, 125 E. Maxwell Street, Suite 201, Lexington, KY 40508. http://dx.doi.org/10.1016/j.arth.2014.08.030 0883-5403/© 2014 Elsevier Inc. All rights reserved.

With the number of patients undergoing pelvic or femoral osteotomies as well as those with previous osteotomies aging and potentially requiring conversion to THA, there is a need to better define the efficacy of total hip arthroplasty after previous osteotomy. Such information will facilitate patient and surgeon informed decisionmaking when contemplating both joint preserving osteotomy surgery and THA after previous osteotomy. Therefore, we performed the present systematic review to provide objective analysis of THA following previous pelvic and/or femoral osteotomy procedures. Our goal was to compare these procedures with standard primary THA’s to determine if differences exist in: (1) clinical outcomes; (2) short-term and longterm complications; and (3) the survivorship of the procedures.

Search Strategy and Criteria We performed an electronic search of PubMed, Scopus, CINAHL®, Cochrane Central Register of Controlled Trials, and ClinicalTrials.gov for articles published between January 1823 and March 2013, and we searched EMBASE™ from 1947 through 2013 for papers appropriate to this study on March 6, 2013. Search strategies were created to cover the concepts of pelvis/femoral/hip osteotomy, time factors, and total hip arthroplasty as thoroughly as possible to optimize recall using terms harvested from standard term indices and on-topic articles. Case Reports and Reviews were excluded using filters constructed in a similar way as the concept hedges though with the goal of precision. To exclude animals, we used the Human filter for PubMed recommended in Cochrane Handbook for Systematic Reviews of Interventions and used that as a model to create similar filters for the other databases searched [21]. Some of the database search terms used were “hip arthroplasty,” and “pelvic osteotomy,” and “femoral osteotomy.” A comprehensive list of search terms is available in the appendix. Search

80

S. Duncan et al. / The Journal of Arthroplasty 30 (2015) 79–85

PUBMED 724 ARTICLES

EMBASE 908 ARTICLES

SCOPUS 715 ARTICLES

CINAHL 134 ARTICLES

COCHRANE 11 ARTICLES

Results sent to EndNote to delete duplicate articles 1 not published in English

1444 articles did not meet inclusion criteria

1474articles

Reviewed abstracts

30 articles met inclusion criteria

569 Lacked 2-year minimum follow-up

6 involved case reports

766 did not involve previous hip osteotomy

20 articles excluded

10 articles met inclusion and exclusion criteria Soballe et al. 1989 Peters et al. 2001 Hashemi-Nejad et al. 2002 Boos et al. 2005 Kawasaki et al. 2005 Haverkamp et al. 2006 Minoda et al. 2006 Lee et al. 2009 Rijnen et al. 2009 Tokunaga et al. 2011

16 Not Level III Evidence or Better

102 Involved Osteotomies for Fracture Nonunion

3 Reviews 1 Letter to Editor

Evidence

Fig. 1. A flow diagram illustrates the method of article selection for study inclusion.

results were entered in EndNote® to remove duplicate studies. The flowchart (Fig. 1) demonstrates the review process from the original search to those included in the final study. These searches identified 1474 potentially eligible studies. Abstracts of the identified studies were reviewed by two of the authors (STD, SAW) to include only studies that were peer-reviewed, published in English, reporting the clinical and radiographic outcomes of total hip arthroplasty following previous pelvic or femoral osteotomy, and had a minimum of 2 year follow-up (Fig. 1). Of the remaining 1474 articles, 569 were excluded for failure to have a minimum 2-year follow-up, 766 did not involve previous pelvic or femoral osteotomy, 102 involved osteotomies for femoral fracture nonunion, 6 were limited to case reports, and 1 article was published in French. 30 articles met inclusion criteria. The full text of these articles was reviewed independently by two authors (STD, SAW). We excluded articles that did not have a control group (minimum Level III evidence) or were a review or letter to the editor. Risk of potential bias was also performed. Limiting articles to those published in the English literature introduced potential publication bias; however, the resources to translate the single article not in the English language would be too great and it was felt that the effect of language bias minimally impacted the potential conclusions of this review. Potential for additional sources of bias in these articles was also assessed critically (performance, attritional). 10 articles met the inclusion criteria and were analyzed in this review [13,14,20,22–28]. The authors reviewed the study type, level of evidence, groups in the study, demographic information, prostheses used, surgical details, outcome measures, clinical and radiographic findings, survival of the prosthesis and complications, and statistical findings. Pooled results for the Harris Hip Score, the complications, and the survivorship were also performed.

Results The ten studies included in the review included 1167 patients (Table 1). Of the four studies involving previous pelvic osteotomy, 90 patients with a history of previous pelvic osteotomy that underwent THA were compared to 116 patients who underwent standard THA. Of the six studies involving previous femoral osteotomy, 338 patients with a history of previous femoral osteotomy that underwent THA were compared to 623 patients who underwent THA. The Level of Evidence was Level II for one study and Level III Evidence for nine studies. All of the included studies had a minimum follow-up of at least two years. The duration of follow-up of the individual patients ranged from two to nineteen years and the pooled mean duration of followup was 11.2 years. The percentage of patients in the individual reports with adequate follow-up ranged from 57% to 100%, with only one study having less than 70% patient follow-up and the pooled average follow-up was 94.2%. The duration of follow-up in any individual study did not impact our comparisons or conclusions regarding outcome measures. Potential bias in the included studies was kept to a minimum. In the included studies, performance bias was minimal as the control and study groups from each study were treated by the same surgeon, operative approach, and implants utilized. Attritional bias was minimized as well with few patients being lost to follow-up, helping to ensure the accuracy of the final survivorship and clinical outcomes data. Not all studies reported information on the patient age and gender distributions. On the basis of available information, the pooled mean age of the patients at the time of THA was 47.3 years and 60% were male. Cemented fixation was used exclusively in four studies, cementless implants alone were used in five studies, and one study reported on

1965–1982 OA, RA, congenital dislocation (3.4%) 96.5 94.9 4.7 (3.3–8) 4.5 (3.5–6.7) 63 (30–80) 44/72 64 (23–94) 276/243 116/116 125/118

1992–2004 AVN 93 100 6.4 (2.2–12.7) 4.2 (2.1–7.3) 35 (24–53) 39 (25–49) 4/9 3/12 16/13 17/14

1998–2005 AVN 100 100 5.1 4.8 (2–9.6) 40.0 6/18 40.1 3/10 14/13 28/24

1989–2003 AVN 100 100 5.0 (3.4–8.7) 4.9 (3.0–6.6) 43 (30–53) 2/12 47 (21–57) 4/11 15/14 16/15

III

III

III

Transtrochanteric Rotational Transtroch Anterior Rotational Transtroch Rotational (TRO) Sugioka Medial-Displacement Intertroch III

10.6 (2.5–20.6) 11.9 (2.4–19.6) 94.1 NR NR NR NR 290/253 121/108 Intertrochanteric III

Bold: P b 0.05.

Femoral

Femoral

Femoral

Femoral

81

both techniques (Table 2). The operative approach utilized was a mix of the posterolateral, direct lateral, and anterolateral. Three of the studies utilized the posterolateral approach only, two studies utilized the anterolateral or direct lateral only, two studies utilized both the posterolateral and anterolateral approach, and three of the studies failed to report the approach utilized. Three of the four studies [23–25] involving previous pelvic osteotomy reported on operating room time and average blood loss with mixed results being observed with two studies [22,24] reporting longer operative times and increased estimated blood loss in the previous pelvic osteotomy group versus one study [23] finding that the operative time was shorter and the corresponding estimated blood loss was less in the previous osteotomy group (Table 2). Three of the six studies involving THA following previous femoral osteotomy reported on the operative time and the estimated blood loss, both being significantly elevated compared to controls (P b 0.05) (Table 2). Reported clinical outcomes at final follow-up in each study included the Harris Hip score (HHS) in all but one study, which used the Merle d’Aubigne score [22]. In the THA following previous pelvic osteotomy, the final HHS ranged from 76 to 85 in the osteotomy cohort compared to 81 to 88 in the control group (Table 3). Only one of the three studies [24] found a statistically significant difference between the two groups with the control group having a higher HHS (P b 0.05). In the THA following previous femoral osteotomy, five of the six studies reported the post-operative HHS with the one study [20] reporting the Merle d’Aubigne score. In the THA following previous femoral osteotomy, the final HHS ranged from 87.7 to 93.7 in the osteotomy cohort compared to 87.7 to 95.3 in the control group (Table 3). Compared to the standard THA group, there was not a difference in the final HHS in any of the studies. Comparison between the groups for complications during the perioperative period failed to demonstrate any statistically significant differences. In the THA following previous pelvic or femoral osteotomy, there was not a difference in the incidence of intraoperative fracture or postoperative dislocation (Table 4). Survival of the THA following previous osteotomy was defined as revision THA for any reason. For THA following previous pelvic osteotomy, the survival rate ranged from 71.2% to 100%; which, compared to the comparison group range of 80.4% to 100% was not statistically significant in the studies (Table 5). The reasons for revision in this cohort involved aseptic loosening, recurrent dislocation, and infection. Aseptic loosening ranged from 3.6% to 23.1%, recurrent dislocation ranged from 3.8% to 7.7%, and infection ranged from 1.9% to 7.7% (Table 5). For THA following previous femoral osteotomy, the survival rate ranged from 75% to 100%; which, compared to the comparison group range of 94.1% to 100% was not statistically significant in the studies (Table 5). Rijnen et al found a difference in the survival rate to be 75.1% vs. 94.1%, but when using Kaplan Meier survivorship curve, the difference was not statistically significant (P = .19) [28]. For THA following previous femoral osteotomy, four of the six [14,20,26,27] studies had a survivorship greater than 90% while the comparison group had five of the six studies [14,20,26–28] having a survivorship greater than 90%, which wasn’t felt to be significantly different. The reasons for revision in this cohort involved aseptic loosening, recurrent dislocation, and infection. Aseptic loosening ranged from 0.9% to 7%, recurrent dislocation only occurred in one study [28] with an incidence of 12.5%, and infection ranged from 1% to 12.5% (Table 5). Compared to the comparison group, differences between the two groups were not statistically significant. To improve the power, pooled results for the Harris Hip Score, complications, and survivorship were collected (Table 6). The pooled results failed to show significant differences between the groups (P N 0.05).

Retrospective Comparative Retrospective Comparative Retrospective Comparative Retrospective Comparative Retrospective Comparative Femoral

Haverkamp et al (2006) Kawasaki et al (2005) Lee et al (2009) Rijnen et al (2009) Soballe et al (1989)

II Femoral

Pelvic

Pelvic

Pelvic

57.4

1980–1984 OA: 53% DDH: 34% AVN, Other: 13% 1974–1993 OA, Dysplasia 73.0 75.7 6.9 (4.75–9.7) 35/39 61.6 (33–82) 57.4 (24–79) 7.3 (4.7–12.2) 39/35 74/74

100 100 7.3 (3–15) 8.5 (3–17) 41 (25–58) 47 (27–66) 43/2 40/2 52/45 51/42 III

Salter (40), Chiari (9) Salter and Chiari (3) Intertrochanteric

74/74

DDH: 100% Control: OA 1987–2002 DDH: 100% Late 1970s–1997 100 100 3 2.3 37 (16–50) 41 (17–54) 9/2 9/2 13/11 13/11 Triple innominate III

1981–1997 DDH: 100% 100 100 60 (25–199) 61 (24–197) 45 (38–64) 39 (25–51) 24/2 42/4 28/26 50/46 Chiari III

1998–2003 DDH: 100% 100 100 2.7 (2–5) 3.1 (2–5) 55.4 56.3 10/0 20/0 10/8 20/17 Chiari III Pelvic

Minoda et al (2006) Hashemi-Nejad et al (2002) Peters et al (2001) Tokunaga et al (2011) Boos et al (1997)

Retrospective Comparative Retrospective Comparative Retrospective Comparative Retrospective Comparative Prospective Comparative

Control Study Study Control Study Control Study Control Study Control

Type of Pelvic Osteotomy Level of Evidence Study Design Pelvic or Femoral Osteotomy Study

Table 1 Demographics of Studies for THA Following Pelvic and/or Femoral Osteotomy.

Hips/Patients

Female/Male

Age (range)

Follow-Up (yrs)

% Follow-Up

Yrs of Investigation

Diagnosis

S. Duncan et al. / The Journal of Arthroplasty 30 (2015) 79–85

Discussion The goals of pelvic and/or femoral osteotomies in the symptomatic, pre-arthritic or early arthritic hip have been to decrease pain, improve

NR

116 0 116 0 NR

1386a NR 787a

276 0 276 0 NR 16 0 16 0 161a 17 0 17 0 123a

0 14 0 14 190.0a (±42.9) 1050.0a (±249.5) 0 28 0 28 117.9a (±19.6) 724.7a (±486.1)

a

NR NR NR NR NR NR

121 0 121 0 NR 290 0 290 0 NR 74 0 74 0 NR 74 0 74 0 NR 0 52 0 52 NR 0 51 0 52 NR

0 13 0 13 130 (100–167) 721a (300–1500) 0 13 0 13 118 (78–170) 448a (300–700) 4 24 3 25 97a (70–126) 662 a (285–1100) P b 0.05 a

1598a 1192a

6 44 3 47 111a (75–145) 1020a (200–2100) 0 10 0 10 127a 0 20 0 20 102a

Prosthesis Cemented Cup Type Cementless Cup Cemented Stem Cementless Stem Operating Room Time EBL

Control (13) Study (28) Control (50) Study (10) Control (20)

Previous Pelvic Osteotomy Minoda et al Hashemi-Nejad et al Peters et al Tokunaga et al Previous Femoral Osteotomy Boos Haverkamp Kawasaki Lee Rijnen Soballe

0 16 0 16 103a (±26) 314a (±112)

0 15 0 15 146a (±26) 508a (±271)

Study (16)

Rijnen

Control (17) Study (14)

Lee Control (28) Control (16) Study (13)

Control Study Control Study Control Study (51) (52) (74) (74) (290) (121)

Femoral Osteotomy

Kawasaki Haverkamp Boos Tokunaga Peters

Pelvic Osteotomy

Hashemi-Nejad Minoda

Table 2 Surgical Details of Performing THA Following Previous Pelvic and/or Femoral Osteotomy.

Table 3 Clinical Outcome Scores for THA Following Previous Pelvic and/or Femoral Osteotomy. Study

Study (15)

Soballe

S. Duncan et al. / The Journal of Arthroplasty 30 (2015) 79–85 Control Study (276) (116)

82

Harris Hip Score, Post THA Control

Study

NR 81 88a 87

NR 79 76a 85

90.1 87.7 88 95.3 91 NR

87.7 89.1 93 93.7 89 NR

P b 0.05.

function, and prolong the survivorship of the native hip joint [29]. Although most patients benefit from these procedures, many of these patients will eventually develop symptomatic end stage arthritis, requiring a THA after their previous hip preservation surgery [3,29,30]. Therefore, it is important to understand the impact of hip osteotomy surgery on subsequent THA. To date, there is controversy regarding this topic and the efficacy of THA after previous osteotomy continues to be debated. The current literature is limited relative to this topic and there are no large patient cohort studies to guide treatment decision-making. Therefore, this systematic review was performed to examine THA following previous pelvic and/or femoral osteotomy procedures to determine if differences exist in: (1) clinical outcomes; (2) short-term and long-term complications; and (3) the survivorship of the procedures. THAs in patients with complex primary hip disorders have achieved mixed clinical results in the past. Historically, the outcomes following THA in these hips compared to primary osteoarthritis were worse relative to clinical outcomes and survivorship [31,32]. In patients with DDH, anterolateral acetabular bone deficiency, soft tissue contractures, and variation in femoral anteversion can make the THA a difficult challenge [25]. Following a properly performed pelvic osteotomy, the placement of the acetabular component may be easier; however, previous authors have pointed out that the osteotomy may alter the anatomy and the version of the acetabulum such that retroversion of the acetabular component may occur, increasing the likelihood of post-operative dislocation [3]. Despite the theoretical risk of increased difficulty with performing the THA following pelvic osteotomy, the literature doesn’t support an increased complication risk or decreased survival of the surgery in matched controlled studies [23,25]. Our systematic review shows no difference in the dislocation rate between the study group and control groups. In our review, the average survivorship following THA in the post-osteotomy compared to controls was the same (90%), suggesting that previous pelvic osteotomy does not compromise the long term results of the THA. Similar improvements in functional outcomes (HHS) were also seen between groups, demonstrating that prior pelvic osteotomy did not compromise the success of the THA compared to patients without a prior pelvic osteotomy. Following previous femoral osteotomy, THA can be a technically demanding procedure depending on the distortion of the proximal femur [13]. Preparation for the femoral component can be difficult with potential displacement of the femoral shaft from the previous osteotomy, angulation of the femur, or retained hardware [14]. With the use of cemented versus cementless femoral components, different obstacles can be encountered. With cementless components, optimal placement of the components may be difficult with abnormal femoral version or displacement of the osteotomy fragments [13]. However, with cemented components, potential for extravasation of cement from empty screw holes must be prevented and pressurization at the time

Table 4 Complications Following THA After Previous Pelvic and/or Femoral Osteotomy. Pelvic Osteotomy Minoda

Complication

Fracture Dislocation

Hashemi-Nejad

Femoral Osteotomy

Peters

Tokunaga

Boos

Control (20)

Study (10)

Control (50)

Study (28)

Control (13)

Study (13)

Control (51)

Study (52)

NR NR

NR NR

3 0

0 0

1 1

3 2

3 (5.9%) 2 (3.9%)

1 (1.9%) 4 (7.7%)

Haverkamp

Kawasaki

Lee

Rijnen

Soballe

Control (74)

Study (74)

Control (290)

Study (121)

Control (16)

Study (15)

Control (28)

Study (14)

Control (17)

Study (16)

Control (276)

Study (116)

7 0

6 0

6 NR

6 NR

0 0

0 2

0 0

0 1

0 0

0 2

3 2

6 1 S. Duncan et al. / The Journal of Arthroplasty 30 (2015) 79–85

Table 5 Survivorship and Reasons for Revision THA Following Previous Pelvic and/or Femoral Osteotomy. Pelvic Osteotomy Minoda

Survival % Revision Reason

Loosening Dislocation Infection Other

Hashemi-Nejad

Peters

Femoral Osteotomy Tokunaga

Boos

Haverkamp

Kawasaki

Lee

Rijnen

Soballe

Control (20)

Study (10)

Control (50)

Study (28)

Control (13)

Study (13)

Control (51)

Study (52)

Control (74)

Study (74)

Control (290)

Study (121)

Control (16)

Study (15)

Control (28)

Study (14)

Control (17)

Study (16)

Control (262)

Study (112)

100% 0 0 0 0

100% 0 0 0 0

88% 6% 6% 0 0

96.4% 3.6% 0 0 0

92.3% 0 7.7% 0 0

92.3% 0 7.7% 7.7% 0

80.4% 13.7% 2% 0 3.9% (osteolysis)

71.2% 23.1% 3.8% 1.9% 0

89.9% 4.1% 0 2.7% 0

81.9% 4.1% 0 8.1% 0

90% 7% 0 3% 0

92% 7% 0 1% 0

100% 0% 0% 0 0

93.3% 6.7% 0 0 0

100% 0 0 0 0

100% 0 0 0 0

94.1% 5.9% 0 0 0

75% 0 12.5% 12.5% 0

98.5% 1.5% 0 0 0

99.1% 0.9% 0 0 0

83

84

S. Duncan et al. / The Journal of Arthroplasty 30 (2015) 79–85

Table 6 Pooled Outcomes for THA Following Previous Pelvic and/or Femoral Osteotomy. Outcome Previous Pelvic Osteotomy Harris Hip Score Complications (number) Survivorship (%) Previous Femoral Osteotomy Harris Hip Score Complications (number) Survivorship (%)

Control (Std. Dev.)

Study (Std. Dev.)

P Value

85.3 (±3.8) 3.3 (±1.5) 90.2 (±8.2)

80.0 (±4.6) 3.3 (±2.9) 90.0 (±12.9)

P N 0.05 P N 0.05 P N 0.05

90.4 (±3.1) 3.0 (±3.3) 95.4 (±4.8)

90.5 (±2.7) 4.0 (±2.6) 90.2 (±9.9)

P N 0.05 P N 0.05 P N 0.05

of THA may be jeopardized as well [13]. Initial reports on THA following femoral osteotomy demonstrate inferior outcomes and survival; however, these initial reports did not have control groups for proper comparison [15,16,33]. In the current study with a minimum of Level III evidence, the survival of THA following previous femoral osteotomy was similar (90.2% vs. 95.4%) and there wasn’t a difference between the two groups in terms of their clinical outcomes (P = 0.9). Nevertheless, it is important to note that three of the six studies did demonstrate increased operative time and blood loss in hips with a previous proximal femoral osteotomy suggestive of increased technical difficulty with these cases. This literature review is limited by our exclusion criteria and by the design and quality of the included studies. THA following previous osteotomy is an uncommon procedure; thus, finding quality level studies difficult to have a large volume of cases to compare to controls. The potential exists for introducing Type II error, especially for survivorship, as few studies were examined and there was heterogeneity within the studies. Post-hoc power analysis was performed using pooled data for survivorship. For THA following previous femoral osteotomy, the power was 91.4%; while the power analysis for THA following previous pelvic osteotomy was 2.9%. The potential for too few articles for THA following previous pelvic osteotomy may not be able to detect a difference in outcomes and could be the focus of further study. The exclusion of articles not published in the English literature may have eliminated some studies that could have contributed to the review, and introducing publication bias. However, none of the authors were fluent in another language other than English and only one study was eliminated, which the authors didn’t feel would have affected the conclusions in this review. The control group and the post-osteotomy group were both treated with similar prostheses, surgical approach, and treating surgeon at the time of THA, helping to decrease the risk of performance bias within the studies. The type of fixation utilized (Cementless vs cemented) did vary amongst the studies for the femoral and pelvic osteotomies, and this could contribute to bias on the survival rate; however, the choice of fixation within each study did not vary widely. With eight of the ten articles having greater than 90% follow-up, the risk of introducing attritional bias in the review was also kept to a minimum. The Quadas tool (Table 7) was utilized to help limit the potential for bias and the

conclusions reached. Limiting only those studies that had a minimum of Level III evidence may have also excluded quality studies and the retrospective nature of the included articles may not be as strong as prospective study design. However, by eliminating those studies without a minimum of Level III evidence, we believe this to be a strength of this review as direct comparison of those patients with a previous osteotomy to those who did not have a previous osteotomy was able to be performed, examining the clinical outcomes and survival of THA between the two cohorts. The lack of consistency of reporting of clinical outcome measures limited our ability to compare multiple outcome measures; instead, we were limited to report on a few that were consistently reported in the majority of the studies. The heterogeneity of the type of pelvic or femoral osteotomy reported in the studies made synthesizing the data into generalizable outcomes difficult but we did attempt to pool those outcomes that were consistently reported amongst all of the studies. Depending on the type of osteotomy and the distortion of the bony anatomy, the conversion to a THA may be more or less difficult as a result compared to the native hip. These limitations demonstrate the need for future prospective longitudinal cohort studies with THA as the main focus and examining the results after specific types of osteotomies. Such studies should be performed with a multicenter study design to provide large enrollment numbers and more generalizable data. In summary, our review of the available, yet limited, literature supports the general concept that following a pelvic and/or femoral osteotomy, the clinical results, complications, and survivorship of subsequent THA are not compromised. Following pelvic or femoral osteotomy, the clinical results are similar compared to controls. However, these conclusions must be accepted with caution due to potentially underpowered analysis given the paucity of available studies. Aside from increased operating room time and blood loss (associated with previous femoral osteotomy), there were no significant differences in the short or long term complications between the groups. Finally, the survivorship of the THA following pelvic or femoral osteotomy may not be compromised by the previous surgery with greater than 90% survival seen in this study, but further multicenter studies are needed to help confirm these results. Appendix. PubMed Search Strategy ((("Arthroplasty, Replacement, Hip"[Mesh] OR "Hip Prosthesis"[Mesh] OR "Hip Replacement Arthroplasties" OR "Hip Replacement Arthroplasty" OR "Hip Prosthesis Implantation" OR "Hip Prosthesis Implantations" OR "Total Hip Replacement" OR "Total Hip Replacements" OR "Hip Prosthesis" OR "Hip Prostheses" OR “Femoral Head Prosthesis” OR “Femoral Head Prostheses” OR "Hip Arthroplasty" OR "Hip Arthroplasties" OR “THA” OR "Hip Replacement Arthroplasties" OR "Hip Replacement Arthroplasty" OR "Hip Prosthesis Implantation" OR "Hip Prosthesis Implantations" OR "Total Hip Replacement" OR "Total Hip Replacements" OR "Hip Prosthesis" OR "Hip Prostheses" OR “Femoral Head Prosthesis” OR “Femoral Head Prostheses” OR "Hip Arthroplasty" OR

Table 7 QUADAS-2 Results. Study

Minoda et al (2006) Hashemi-Nejad et al (2002) Peters et al (2001) Tokunaga et al (2011) Boos et al (1997) Haverkamp et al (2006) Kawasaki et al (2005) Lee et al (2009) Rijnen et al (2009) Soballe et al (1989)

Risk of Bias

Applicability Concerns

Patient Selection

Index Test

Reference Standard

Flow and Timing

Patient Selection

Index Test

Reference Standard

Low Low Low Low Low Low Low Low Low Low

High High High High High High High High High High

Low Low Low Low Low Low Low Low Low Low

Low Low Low Low Low Low Low Low Low Low

Low Low Low Low Low Low Low Low Low Low

Low Low Low Low Low Low Low Low Low Low

Low Low Low Low Low Low Low Low Low Low

S. Duncan et al. / The Journal of Arthroplasty 30 (2015) 79–85

"Hip Arthroplasties" OR “THA” OR “hip endoprosthesis” OR “hip total arthroplasty” OR “hip total endoprosthesis” OR “hip total joint replacement” OR “hip total prosthesis” OR “hip total replacement” OR “total hip endoprosthesis” OR “total hip joint endoprosthesis” OR “total hip joint prosthesis” OR “total hip replacement” OR “total hip prosthesis” OR “Daubenspeck hip endoprosthesis” OR “dauberspeck hip endoprosthesis” OR “Mckee Farrar prosthesis” OR “mac kee farrar endoprosthesis” OR “mac kee farrar hip prosthesis” OR “mac kee prosthesis” OR “mackee farrar endoprosthesis” OR “mackee farrar hip prosthesis” OR “mackee farrar prosthesis” OR “mc kee farrar prosthesis” OR “mc kee prosthesis” OR “mckee-farrar prosthesis” OR “mckee-ferrar prosthesis” OR “mckee farrar hip endoprosthesis” OR “mckee farrar hip prosthesis” OR “mckee farrar total hip replacement” OR “mckee ferrar prosthesis” OR “Ring hip endoprosthesis” OR “Thompson hip endoprosthesis” OR (("Prostheses and Implants"[Mesh] OR "Prostheses” OR "Prosthesis” OR “artificial implant” OR “artificial implants” OR “prosthetic”) AND ("Hip"[Mesh] OR hip OR hips OR coxa OR coxas OR coxae))) AND ("Time Factors"[Mesh] OR "Postoperative Complications"[Mesh] OR “previous” OR “prior” OR “previously” OR “postoperative” OR “after” OR “past” OR “former” OR “earlier” OR “foregoing” OR “preexistent” OR “preexisting” OR “later” OR “after” OR “before”) AND (("Osteotomy"[Mesh] OR "Osteotomy” OR "Osteotomies”) AND ("Femur"[Mesh] OR Femur OR Femurs OR Trochanter OR Trochanters OR trochanteric OR intertrochanteric OR transtrochanteric OR subcapital OR “varus derotational” OR “VDRO” OR Imhauser[tiab] OR Southwick[tiab] OR Dunn[tiab] OR femoral OR “thigh bone” OR "Pelvis"[Mesh] OR "Pelvis" OR “Pelvic” OR chiari[tiab] OR salter[tiab] OR Pemberton[tiab] OR Dega[tiab] OR “Triple Innominate” OR Periacetabular OR Shelf OR Rotational OR Spherical OR Ganz[tiab] OR Dial[tiab] OR Steele[tiab] OR Sutherland[tiab] OR Greenfield[tiab] OR Staheli[tiab] OR “Double innominate” OR Bernese OR "Hip"[Mesh] OR hip OR hips OR coxa OR coxas OR coxae) OR (“pubiotomy”))) NOT (("Animals"[Mesh]) NOT ("Animals"[Mesh] AND "Humans"[Mesh])) NOT ("Case Reports" [Publication Type] OR "Review" [Publication Type] OR "Case Reports" OR "Case Report" OR “Case Studies” OR “Case Study” OR “Case Histories”) References 1. Leunig M, Siebenrock KA, Ganz R. Rationale of periacetabular osteotomy and background work. Instr Course Lect 2001;50:229. 2. Steppacher SD, Tannast M, Ganz R, et al. Mean 20-year followup of Bernese periacetabular osteotomy. Clin Orthop Relat Res 2008;466(7):1633. 3. Parvizi J, Burmeister H, Ganz R. Previous Bernese periacetabular osteotomy does not compromise the results of total hip arthroplasty. Clin Orthop Relat Res 2004(423):118. 4. Engesaeter LB, Furnes O, Havelin LI. Developmental dysplasia of the hip—good results of later total hip arthroplasty: 7135 primary total hip arthroplasties after developmental dysplasia of the hip compared with 59774 total hip arthroplasties in idiopathic coxarthrosis followed for 0 to 15 years in the Norwegian Arthroplasty Register. J Arthroplast 2008;23(2):235. 5. Pellicci PM, Hu S, Garvin KL, et al. Varus rotational femoral osteotomies in adults with hip dysplasia. Clin Orthop Relat Res 1991(272):162.

85

6. Iwase T, Hasegawa Y, Kawamoto K, et al. Twenty years' followup of intertrochanteric osteotomy for treatment of the dysplastic hip. Clin Orthop Relat Res 1996(331):245. 7. Albers CE, Steppacher SD, Ganz R, et al. Joint-preserving surgery improves pain, range of motion, and abductor strength after Legg–Calve–Perthes disease. Clin Orthop Relat Res 2012;470(9):2450. 8. Gotoh E, Inao S, Okamoto T, et al. Valgus-extension osteotomy for advanced osteoarthritis in dysplastic hips. Results at 12 to 18 years. J Bone Joint Surg Br 1997;79(4):609. 9. Mont MA, Fairbank AC, Krackow KA, et al. Corrective osteotomy for osteonecrosis of the femoral head. J Bone Joint Surg Am 1996;78(7):1032. 10. Scher MA, Jakim I. Intertrochanteric osteotomy and autogenous bone-grafting for avascular necrosis of the femoral head. J Bone Joint Surg Am 1993;75(8):1119. 11. Marti RK, Schuller HM, Raaymakers EL. Intertrochanteric osteotomy for non-union of the femoral neck. J Bone Joint Surg Br 1989;71(5):782. 12. D'Souza SR, Sadiq S, New AM, et al. Proximal femoral osteotomy as the primary operation for young adults who have osteoarthrosis of the hip. J Bone Joint Surg Am 1998; 80(10):1428. 13. Boos N, Krushell R, Ganz R, et al. Total hip arthroplasty after previous proximal femoral osteotomy. J Bone Joint Surg Br 1997;79(2):247. 14. Haverkamp D, de Jong PT, Marti RK. Intertrochanteric osteotomies do not impair long-term outcome of subsequent cemented total hip arthroplasties. Clin Orthop Relat Res 2006;444:154. 15. Iwase T, Hasegawa Y, Iwasada S, et al. Total hip arthroplasty after failed intertrochanteric valgus osteotomy for advanced osteoarthrosis. Clin Orthop Relat Res 1999(364):175. 16. Ferguson GM, Cabanela ME, Ilstrup DM. Total hip arthroplasty after failed intertrochanteric osteotomy. J Bone Joint Surg Br 1994;76(2):252. 17. Benke GJ, Baker AS, Dounis E. Total hip replacement after upper femoral osteotomy. A clinical review. J Bone Joint Surg Br 1982;64(5):570. 18. Gerundini M, Avai A, Taglioretti J. Total hip replacement after intertrochanteric osteotomy. Int Orthop 1995;19(2):84. 19. Lemaire R, Colinet J. Technical problems posed by total hip arthroplasty after failure of another surgical treatment. Acta Orthop Belg 1985;51(2–3):411. 20. Soballe K, Boll KL, Kofod S, et al. Total hip replacement after medial-displacement osteotomy of the proximal part of the femur. J Bone Joint Surg Am 1989;71(5):692. 21. Higgins J. Cochrane handbook for systematic reviews of interventions version 5.1.0. In: S G, editor. The Cochrane collaboration; 2011. 22. Minoda Y, Kadowaki T, Kim M. Total hip arthroplasty of dysplastic hip after previous Chiari pelvic osteotomy. Arch Orthop Trauma Surg 2006;126(6):394. 23. Hashemi-Nejad A, Haddad FS, Tong KM, et al. Does Chiari osteotomy compromise subsequent total hip arthroplasty? J Arthroplast 2002;17(6):731. 24. Peters CL, Beck M, Dunn HK. Total hip arthroplasty in young adults after failed triple innominate osteotomy. J Arthroplast 2001;16(2):188. 25. Tokunaga K, Aslam N, Zdero R, et al. Effect of prior Salter or Chiari osteotomy on THA with developmental hip dysplasia. Clin Orthop Relat Res 2011;469(1):237. 26. Kawasaki M, Hasegawa Y, Sakano S, et al. Total hip arthroplasty after failed transtrochanteric rotational osteotomy for avascular necrosis of the femoral head. J Arthroplast 2005;20(5):574. 27. Lee YK, Ha YC, Kim KC, et al. Total hip arthroplasty after previous transtrochanteric anterior rotational osteotomy for femoral head osteonecrosis. J Arthroplast 2009; 24(8):1205. 28. Rijnen WH, Lameijn N, Schreurs BW, et al. Total hip arthroplasty after failed treatment for osteonecrosis of the femoral head. Orthop Clin North Am 2009;40(2):291. 29. Millis MB, Murphy SB, Poss R. Osteotomies about the hip for the prevention and treatment of osteoarthrosis. Instr Course Lect 1996;45:209. 30. Siebenrock KA, Leunig M, Ganz R. Periacetabular osteotomy: the Bernese experience. Instr Course Lect 2001;50:239. 31. Crowe JF, Mani VJ, Ranawat CS. Total hip replacement in congenital dislocation and dysplasia of the hip. J Bone Joint Surg Am 1979;61(1):15. 32. Garvin KL, Bowen MK, Salvati EA, et al. Long-term results of total hip arthroplasty in congenital dislocation and dysplasia of the hip. A follow-up note. J Bone Joint Surg Am 1991;73(9):1348. 33. Suominen S, Antti-Poika I, Santavirta S, et al. Total hip replacement after intertrochanteric osteotomy. Orthopedics 1991;14(3):253.