Isolated Liner Exchange and Bone Grafting for the Management of Periacetabular Osteolysis in Well-Fixed Cups with an Intact Locking Mechanism at Short-Term to Medium-Term Follow-Up: A Systematic Review

Abstract

Background: Polyethylene liner exchange and bone grafting is an effective surgical option for the management of periacetabular osteolysis following total hip arthroplasty with well-fixed cups and intact liner locking mechanisms. Purpose: We aimed to evaluate the revision-free survivorship and radiographic lesion progression after polyethylene liner exchange and bone grafting is performed for periacetabular osteolysis. Methods: A systematic review of the literature was performed. We queried Medline, EMBASE, and Cochrane Library for articles published from January 1999 to January 2023 using the following keywords: “osteolysis” AND “well-fixed,” “osteolysis” AND “retro-acetabular,” “bone graft” AND (“retention” OR “retained” OR “stable”) AND “cup,” and “uncemented liner” AND “well-fixed.” Results: Of 596 articles found, 9 articles were selected for final inclusion (227 cases, mean follow-up time 43.6 months). The overall cup revision rate after liner exchange was 6.6% (15 hips) due to progressive osteolysis (5 hips), aseptic loosening of the acetabular component (5 hips), dislocation (4 hips), and periprosthetic infection (1 hip). There was either radiographic resolution or regression of periacetabular osteolysis in all reported cases that provided measurements (52 hips) except 1 (1.9%) requiring revision. All studies reporting clinical outcomes indicated improved pain and functional scores. Conclusion: This systematic review found that isolated liner exchange with bone grafting for the management of periacetabular osteolysis was associated with a high revision-free survival rate (93.4%) and minimal radiographic progression (1.9%) of osteolytic lesions at short-term to medium-term follow-up. Liner exchange with bone grafting is recommended for the management of large periacetabular osteolytic lesions (> 450 mm²) in well-fixed acetabular cups. We encourage future studies to develop a grading scale for lesions to guide clinical management and risk stratification for patients.

Introduction

While decreasing in incidence, polyethylene liner wear remains a problematic complication following total hip arthroplasty (THA). Aseptic periacetabular osteolysis has been linked to polyethylene wear, especially in first-generation uncemented cups with conventional polyethylene [6,9,17,28,29,32,48]. Patients with periacetabular osteolysis may present asymptomatically; however, the risk of long-term complications [20,21,32,36,39] may necessitate surgical intervention to prevent lesion progression, complete polyethylene wear, or even cup wear through [1,7,23].

Surgical interventions involve either complete acetabular cup revision or an isolated liner exchange with the retention of the well-fixed acetabular cup [20]. Acetabular cup revision allows for adequate visualization and access to lesions [23]. However, cup revision is more invasive than isolated liner exchange. Component revision carries significant risks, including increased blood loss, increased operating room time, and increased bone loss due to ingrown acetabular component removal, in addition to the surgical challenges of reconstruction in the setting of significant acetabular bone loss [39,40,49].

In contrast, isolated liner exchange with acetabular cup retention is an efficient surgery with low morbidity. Patients with a well-fixed and well-positioned acetabular component are candidates for this approach, with added clinical benefits of faster patient recovery and mobility after surgery [33,39,54]. In addition, isolated liner exchange is often recommended in conjunction with debridement and curettage of osteolytic material, followed by bone grafting. As reviewed by Hall et al, bone grafting can provide mechanical support in addition to osteoconductive, osteogenic and osteoinductive properties in the treatment of osteolytic lesions depending on the type of graft utilized [15,20,33]. Bone grafting is performed with specialized instruments through the screw holes present in the existing acetabular cup; however, screw holes are not essential. If lesions are difficult to access, artificial windows can be created in the ilium or periphery of the acetabular cup to enhance access as long as these options do not compromise stability or functionality of the cup [33].

Although a number of papers have been published on the clinical performance of isolated liner exchange and bone grafting for the management of well-fixed cups with periacetabular osteolysis, no systematic review of the literature on this topic has been published to date focusing on polyethylene exchanges when the liner locking mechanism is intact. This study’s aim was 3-fold in regards to isolated liner exchange and bone grafting in well-fixed cups with intact liner locking mechanism: (1) What is the acetabular revision rate after isolated liner exchange and, specifically, revision for aseptic loosening or liner-related revisions (dissociation or wear)? (2) What is the postoperative rate of radiographic progression of osteolytic lesions? and (3) What types of graft and bone grafting techniques are used to treat periacetabular osteolysis in cases of retention of well-fixed cups?

Methods and Search Criteria

The Medline, EMBASE, and Cochrane Library were searched for articles published from January 1999 to January 2023 using:

1. “osteolysis” AND “well-fixed”;

2. “osteolysis” AND “retro-acetabular”;

3. “bone graft” AND (“retention” OR “retained” OR “stable”) AND “cup”;

4. “uncemented liner” AND “well-fixed.”

All databases were last accessed by one author on January 26, 2023.

The inclusion criteria were clinical trials investigating the outcome of liner exchange combined with bone grafting for the treatment of periacetabular osteolysis in well-fixed cups. Primary outcomes were: (1) survivorship of the implant/THA, (2) complication rates, (3) radiological signs of component fixation, and (4) subjective functional scores.

The exclusion criteria were: (1) Non-English or nonfull-text, (2) liner exchange in a well-fixed cup without bone grafting, (3) cup re-implantation trials, (4) liner exchange without evidence of periacetabular osteolysis, (5) only liner cementation on a well-fixed cup with periacetabular osteolysis or no stratification of the liner fixation method, (6) lack of clinical, functional, or radiological outcomes, (7) general or systematic review, (8) mean follow-up <2 years, and (9) articles published after January 2023.

Three authors independently (Author 1, Author 2, and Author 3) conducted the search. Differences between reviewers were discussed until agreement was achieved, and the senior author had the final decision. The 3 reviewers independently extracted study data and assessed variable reporting of outcome data. Descriptive statistics were calculated, and parameters were analyzed. The level of evidence in the studies was determined using the Oxford Center for Evidence-Based Medicine-Level of Evidence. The “quality assessment” of the studies for methodological deficiencies, as a common alternative to “risk of bias,” was examined by the modified Coleman methodology Score. The methodological quality of each study and the different types of detected bias were assessed by each reviewer and were combined synthetically. Selective reporting bias (publication bias) was not included in the assessment. Finally, a comprehensive analysis of the eligible studies focused on topic questions.

During the initial review, the following information was collected for each study: title, author, year published, study design, patient numbers, demographic information, number of hips, sex, revision surgery indication, surgical approach, implant type, bone graft type, acetabular bone loss classification, mean preoperative clinical scores, postoperative clinical and functional outcomes, mean postoperative follow-up, re-revision rates, postoperative complications.

Results

A total of 596 articles were found. After duplicate removal, 353 article abstracts were subject to review. After the initial abstract screening, 35 articles underwent a full text-screening process based on inclusion and exclusion criteria, and 9 articles were selected for final inclusion in the study [10,18,24,25,27,37,41,44,51] (Supplemental Fig. 1).

Seven articles (77.8%) were retrospective case series (level of evidence IV) [10,18,24,25,37,41,51]. 1 article (11.1%) was a retrospective comparative study (level III) [27]. 1 article (11.1%) was a prognostic case study (level II) [44] (Table 1).

Table 1.

Type of study, level of evidence, quality assessment, and demographic information.

Author name (year)	Type of study	Level of evidence	Number of hips	Sex	Mean age (years)	Date of procedure	Modified Coleman score
Kang et al [24]	Retrospective case series	IV	22	N/R	46.6	N/R	50
Narkbunnam et al [41]	Retrospective case series	IV	33	N/R	60	Jul. 2002 to Dec. 2012	34
Deheshi et al [10]	Retrospective case series	IV	4	N/R	63	N/R	35
Maloney et al [37]	Retrospective case series	IV	26	N/R	49	Jan. 1991 to Jun. 1994	49
Koh et al [27]	Retrospective comparative study	III	6	N/R	51.3	Oct. 1997 to Dec. 2008	44
Restrepo et al [44]	Prognostic case study	II	32	N/R	65.5	2002 to 2004	48
Griffin et al [18]	Retrospective case series	IV	55	M: 32 F: 23	53.6	May 1986 to Mar. 2003	49
Smith et al [51]	Retrospective case series	IV	15	N/R	51	Jul. 1997 to Oct. 2001	47
Kim et al [25]	Retrospective case series	IV	34	N/R	49	Sept. 1995 to Aug. 2012	46

M Male, F Female.

The modified Coleman score was calculated as a quality assessment for each study by one author (Table 1). Modified Coleman scores range from 0 to 100 with various quality assessments of sample size, study type, standardization of surgical approach and technique, and assessments of bias in selection and outcomes. A higher score indicates a higher quality study. The mean modified Coleman score for the 9 included studies was 45.4, as the majority of included studies were retrospective case series with small sample sizes.

A total of 227 hips in the 9 articles satisfied our inclusion criteria. Only 1 article [18] (55 of 227 hips) provided the sex (58.2% male) of patients who specifically underwent isolated liner exchanges and bone grafts. The mean patient age varied between 46.6 years [24] and 65.5 years [44] (Table 1). The mean follow-up time ranged from 27.6 months [24] to 99.6 months [25], with a weighted mean of 46.9 months.

At least 1 periacetabular osteolytic lesion in each hip was packed with bone graft. However, no articles provided a specific number of grafted lesions per hip. Three articles [24,27,41] depicted the location of an osteolytic lesion on the DeLee and Charnley zone [11] (Supplemental Table 1). In these articles, there were 61 osteolytic lesions in 61 hips, of which 11 lesions (18%) were in Zone I, 40 lesions (65.6%) were in Zone II, and 10 lesions (16.4%) were in Zone III. One article [37] depicted the location of 46 osteolytic lesions in 26 hips. Thirty-five of 46 lesions (76.1%) were in the ilium, 8 (17.4%) in the pubis, and 3 (6.5%) in the ischium. Four articles [10,24,36,41] measured the size of the pelvic osteolytic lesion on the computerized tomography (CT) scan. The mean size per osteolytic lesion in 3 articles was 465.84 mm² [24], 564.25 mm² [41], and 682 mm² [37]. In the fourth article [10], lesion size ranged from 620 mm² to 4770 mm². The studies included in this analysis did not use any type of classification system to grade the amount of periacetabular osteolysis.

Three articles [10,24,37] reported that 81 hip femoral heads appeared eccentric within the acetabular component on the X-rays, indicating liner wear, whereas other articles did not provide this information.

Five articles [10,24,25,27,41] did not identify the primary acetabular cup type. One article [37] reported using a porous-coated metal acetabular component during the primary THA without providing the cup name. In the remaining 3 articles [18,44,51], 102 THAs used porous-coated or grit-blasted metal shells. No highly porous acetabular metal shells were reported in the articles.

Stable fixation of acetabular metal shell was confirmed by both preoperative radiographic control and intra-operative assessment. Subsequent radiographic loosening of the acetabular cup was defined as the presence of a complete radiolucent line greater than 2 mm in the DeLee and Charnley zones or as evidence of cup migration/cup position change [11].

Only 1 article [51] identified the type of revised liner. In this article, 12 of 15 hips (80%) used constrained liners, and 3 hips (20%) used neutral liners. One patient had a dislocation after liner exchange. The authors did not provide the type of liner used in this patient with a subsequent dislocation.

Three articles used a posterolateral approach in 69 of 227 hips (30.3%) [18,25,27]. In 1 article, [44] 32 hips (14.1% of 227 hips) were operated on with a direct lateral approach. Twenty-three hips (9.6% of 227 hips) in 1 article combined an ilioinguinal and a posterolateral approach; the former approach provided access to the inner ilium at the acetabular shell dome for an allograft and the latter approach revealed the acetabular cup for liner exchange [24]. An anterolateral approach was used in 3 articles [25,27,51], representing 23 of 227 hips (10.1%). For 18 of 227 hips (7.9%), a transtrochanteric approach was used [25]. The approach for the remaining 63 hips (27.8% of 227 hips) was not reported in 3 articles [10,37,41].

Two types of surgical techniques were used to debride the osteolytic lesion and pack the grafting material into the osteolytic cavity. The 1st method accessed the lesion by creating a cortical window adjacent to the ilium (30/227 hips or 13.2%) [24,25,27]. The second method was to debride and graft the lesion through the rim or screw holes in the acetabular component; 60 hips (26.4% of 227 hips) in 2 articles were subject to this technique [25,37]. Two articles [10,25] used either or both of these methods.

In 8 articles (223 hips or 98.2% of 227 hips), bone grafting was achieved with morselized bone allograft or bone chips [18,24,25,27,37,41,44,51]. The bone grafts to treat the osteolytic lesions in 73 hips (32.2% of 227 hips) were specified as cancellous allografts in 3 of the aforementioned 8 articles [18,24,25,27,37,41,44,51]. In 1 article [10], 4 hips (1.8% of 227 hips) were grafted with pellets or a calcium sulfate matrix as the bone substitute. A total of 213 of 227 liners (93.8%) were replaced with conventional polyethylene (PE) liners and locked into place using the intact liner locking mechanism in the cup [10,18,24,25,37,41,44,51]. One article [27] reported six liners (3.1% of 193 hips) that were replaced with ultrahigh molecular weight (UHMW) PE liners. In 8 hips (3.5% of 227) in 1 article [25], highly cross-linked (HXLPE) liners were used to replace worn PE liners. The exchange of femoral heads was performed in all cases with retained acetabular cups. Three articles [18,37,51] reported downsizing the femoral heads in certain cases to allow for a thicker polyethylene liner. One article reported upsizing 4 femoral heads during liner exchange [18]. Eight femoral stems (3.5% of 227 hips) were revised with liner exchange in 2 articles [41,51].

In 1 article, a 2nd liner exchange was performed in 2 revision cases (2 of 227 hips, 0.9%), and all component revision was performed for 4 of 227 (1.8%) hips [25]. The remaining revision cases in this review (15 of 227 hips, 6.6%) were treated with acetabular cup revision [18,25,41,44,51]. The overall cup revision rate was 6.6% after liner exchange (93.4% survival rate at a weighted mean follow-up of 46.9 months). Four articles [10,24,27,37] reported no cases of cup revision (100% survival rate at a mean follow-up ranging from 27.6 to 49.6 months). In contrast, 4 articles (135 of 227 hips, 59.5%) [18,41,44,51] reported a cup survival rate ranging from 90 to 95% at a mean follow-up ranging from 30.4 to 54 months. One article reported a cup survival rate of 70.6% at a mean follow-up ranging from 60 to 242.2 months [25].

The most common reasons for the failure of liner exchange requiring subsequent acetabular revision were progression of pelvic osteolytic lesion and/or PE liner wear (5 hips, 2.2%), aseptic loosening (5 hips, 2.2%), dislocation (4 hips, 1.8%), and periprosthetic infection (PJI) (1 hip, 0.4%).

The overall incidence rate of dislocation was 4.8% (11 out of 227 hips). Among the 11 dislocations, 4 hips required revision of an acetabular component, while 7 hips were treated with closed reduction and a brace [18,27,51]. One hip with PJI required a 2-stage revision [41]. Two hips (0.9% of 227 hips) showed evidence of liner wear progression but did not require revision [25]. Other postoperative complications included greater trochanter nonunion in 2 of 227 (0.9%) hips, sciatic nerve palsy in 1 of 227 (0.4%) hips, periprosthetic fracture in 1 of 227 (0.4%) hips, and heterotopic ossification in 1 of 227 (0.4%) hips [25]. One greater trochanter nonunion was treated with open reduction and internal fixation, while the other was treated conservatively. The periprosthetic fracture was a greater trochanter fracture treated with open reduction and internal fixation [25]. No hips were subject to revision due to femoral stem loosening (Supplemental Table 2).

Three articles [10,24,27,37] found no radiographic evidence of the progression of osteolytic lesions after liner exchange. One of these 3 articles [37] reported that 65% of osteolytic lesions filled with bone grafts appeared to have regressed. One article [18] reported 16 cases (27% of 55 hips in this study) that had asymptomatic progressive osteolysis; 1 case (1.8% of 55 hips) had the progression of a large pelvic osteolytic lesion which impacted the stability of the acetabular component 5 years after the procedure and required revision of the acetabular component. Another article reported radiographic evidence of PE liner wear or osteolysis in 8 of 34 (23.5%) hips in the study at a mean follow-up of 99.6 months but did not report on the size or progression of the lesions [25].

The overall Harris Hip score (HHS) was used in 4 articles [18,24,27,51]. The mean preoperative HHS score ranged from 64.1 [27] to 84 [24] whereas the mean postoperative HHS score ranged from 81 [51] to 93.8 [24] at the final follow-up. The HHS pain score was used in 2 articles [37,41]. In these 2 articles, the mean preoperative HSS pain score was 20.3 [41] and 18 [37], whereas the mean postoperative HSS pain score increased to 32.1 [41] and 38 [37] at the final follow-up. In summary, the mean postoperative clinical/functional subjective score to assess the surgical technique of replacing the old liner with a new liner locked to a well-fixed cup represented a significant improvement to the preoperative mean score (Table 2).

Table 2.

Clinical/functional outcomes.

Author name	Mean follow-up (months)	Preoperative clinical score	Postoperative clinical score	P value
Kang et al [24]	27.6	HHS: 84	HHS: 93.8	N/R
Narkbunnam et al [41]	54	HHS pain score: 20.3	HHS pain score: 32.1	P ≤ .001
Deheshi et al [10]	48	N/R	N/R	N/R
Maloney et al [37]	39.6	HHS pain score: 18	HHS pain score: 38	N/R
Koh et al [27]	49.6	HHS: 64.1	HHS: 92.7	P = .766
Restrepo et al [44]	33.6	N/R	N/R	N/R
Griffin et al [18]	30.4	HHS: 66.2	HHS: 84.2	N/R
Smith et al [51] Kim et al [25]	41 99.6	HHS: 70 N/R	HHS: 81 N/R	P = .008 N/R

Mean follow-up and clinical outcomes were assessed using studies with reported p-values for differences between preoperative and postoperative HHS or HHS pain score.

HHS Harris hip score.

Discussion

In this review, we aimed to characterize (1) the cup revision rate following ILE, (2) the progression of periacetabular osteolysis following ILE, and (3) the bone grafting techniques that are utilized in cases of ILE. The overall revision rate of 227 acetabular cups after liner exchange with bone grafting was 6.6% (15/227), predominantly due to progressive liner wear and/or osteolysis (5 hips), aseptic loosening (5 hips), and dislocation (4 hips). The rate of dislocation following isolated liner exchange was 4.8% (11/227). Bone grafting with allogenic bone grafts or calcium sulfate pellets resulted in minimal osteolytic lesion progression of 1.9%.

This review had several limitations. The majority of the articles included were retrospective studies without any control groups, and results are subject to biases in each article, reflected by the mean modified Coleman score of 45.4. No stark biases threatened the overall design of the included studies. There was some inherent selection bias based on the inclusion and exclusion criteria for each included study, for example, patients outside of specified age ranges or those lost to follow-up may have been excluded. Most of the included studies reported HHS as the predominant assessment of clinical outcome, which can be subject to observer bias. Only 3/9 (33.3%) reported on the significance of differences in preoperative and postoperative HHS and no articles referenced the achievement of minimal clinically important differences (MCID) [27,41,51]. Future studies should consider reporting on other patient-reported outcome measures and objective functionality assessments. The mean follow-up was short-term to medium-term, with only one study reporting a mean follow-up greater than 60 months. The study with the longest mean clinical follow-up reported the lowest survivorship [25]. Thus, future studies are encouraged to investigate outcomes of PE liner exchange at follow-up >60 months. Another limitation is that this review particularly investigated acetabular cup survivorship and did not account for all-cause revisions of other components. Furthermore, only 4 articles [10,24,25,37] reported on the radiographic evidence of progression of osteolytic lesions, and it was difficult to fully assess the impact of bone grafting with isolated liner exchange on osteolysis despite resolution and stabilization of most lesions. Finally, the majority of the articles reported the use of conventional polyethylene liners, and only 8 out of 227 hips in this review used a HXLPE liner. For the studies reporting the use of conventional PE, older cups were either incompatible with newer-generation liners, or the authors did not specify the reason for placing another conventional PE liner. Future studies are encouraged to investigate the survivorship of HXLPE liners particularly in ILE and bone grafting for periacetabular osteolysis.

Our findings are similar to those of a recent systematic review by Malahias et al that investigated polyethylene liner cementation into well-fixed cups for the management of periacetabular osteolysis [36]. However, they reported a higher overall cup revision rate (11.3%) compared to our reported cup revision rate of 6.6%. The majority of cup revisions in both reviews were due to aseptic loosening, yet there were more total cup revisions for aseptic loosening in the cemented liner cohort. One potential reason for this discrepancy is that the mean follow-up in the cemented liner cohort was 76.1 months, compared to 46.9 months in the intact liner-locking cohort. We recommend that liner exchange with or without cement should be performed based on surgeon preference and patient-specific factors. For example, patients who present with periacetabular osteolysis in the setting of a well-fixed cup and intact liner locking mechanisms would be a candidate for isolated liner exchange and bone grafting without cement. One caveat to this recommendation is that a surgeon may elect to use dual mobility (DM), constrained, or HXLPE liner. Older cups may not be compatible with these liners, in which case cementing the liner into place would be appropriate.

Furthermore, our reported outcomes are comparable with studies investigating complete acetabular revision for treating periacetabular osteolysis in well-fixed cups [14,16,52]. Three studies in this review also directly compared liner exchange against acetabular component revision, and all highlighted satisfactory, comparable outcomes with regard to fixation failure and complication rates [27,41,44]. However, a study by Lie et al found that in a cohort of 318 isolated acetabular liner exchanges, 52 (16.4%) required further revisions. In this study, there was no subanalysis reporting whether the indication for revision (cases with or without periacetabular osteolysis) or whether isolated liner exchange conducted with or without bone grafting affected their results [31]. Thus, the reported higher dislocation and revision rates in their study may be attributed to the inclusion of liner exchanges for THA failure due to reasons other than periacetabular osteolysis.

The cup revision and complication rates in this review are also favorable compared to studies investigating isolated liner exchange for indications other than acetabular osteolysis, including hip instability [54]. In particular, Earll et al reported that the use of liner exchange for hip instability in modular component exchange had a 55% dislocation rate [12], suggesting the limited utility of liner exchange for hip instability [54]. Likewise, Blom et al, also reported a complication rate of 23% after isolated liner exchange in cases with or without periacetabular osteolysis, with most complications due to dislocation [4]. Given our findings, the overall lower cup revision rate and dislocation rate of 4.8% suggests that the use of isolated liner exchange with bone grafting is an effective option with high revision-free survivorship, particularly in treating periacetabular osteolysis. The lower dislocation rate reported in our findings may be due to the fact that these liners were inserted into intact locking mechanisms, rather than cemented into a well-fixed shell. Since a cement mantle around the liner is not required, the liner can allow for a larger femoral head size which may have led to improved hip stability. Preoperatively, cup stability can be assessed radiographically over time to monitor for movement or by using CT scans to better visualize osteolysis [25]. Mechanical stability of the cup should be assessed intraoperatively by pulling on the cup with a rongeur and checking for fluid egress within the pelvis.

The articles included in the review utilized allogenic bone grafts with the exception of one which also reported favorable outcomes using calcium sulfate pellets for grafting [10]. No consensus exists on which type of bone graft provides the best clinical results in treating periacetabular osteolysis [33]. We find that there was either resolution or no evidence of progression of osteolytic lesions after isolated liner exchange with bone grafting in the majority of studies. However, the article with the longest mean clinical follow-up period reported radiographic evidence of liner wear or progressive osteolysis in 8 of 34 cases [25]. The authors of this article did not report preoperative or postoperative measurements of liner wear or osteolysis in these cases. Although the articles in this study all utilized bone grafting, previous studies have demonstrated similar results in regard to osteolytic lesion regression without the use of bone grafts. Schmalzried et al [49] compared 5 hips without bone grafting to 10 hips with bone grafting when the acetabular sockets were all left in situ. They found that 3 of 5 hips without bone grafting and all 10 hips with bone grafting had an increase in the density of osteolysis consistent with bone graft incorporation. Likewise, Maloney et al [39] compared 29 patients with allograft to 11 patients whose lesions were debrided but not grafted. At the mean follow-up of 3.5 years, approximately one-third of lesions had resolved completely, and two-thirds had decreased in size regardless of whether they were grafted. Consensus regarding the use of bone grafting with isolated liner exchange has also not been established. However, the results of this study with regard to the regression of osteolytic lesions with isolated liner exchange and bone grafting may extend to cases with isolated liner exchange without bone grafting as well.

Of the articles reporting on the location of osteolytic lesions based upon DeLee and Charney zones [24,27,41], the majority of the lesions (40/61) were located in Zone II. These lesions were debrided and packed through the rim or the screw holes of the acetabular component [24], and if necessary, a window was made for access to the back of the acetabular cup [27]. Additionally, Maloney et al located 76.1% (35/46) of osteolytic lesions in the ilium and accessed them through the acetabular component periphery instead of through the screw holes [37]. Based on these findings, it is recommended to debride and graft through the screw holes or rim for access for lesions depicted in Zone II, and it is recommended to debride and graft through the periphery for lesions located in the ilium. For both locations, a window may be necessary for adequate access to the posterior column/wall [10,37].

Four of the articles reported on the type of initial acetabular components in patients undergoing isolated exchange [18,37,44,51], and all initial cups were cementless porous-coated metal shells with conventional polyethylene liners. With the advent of highly porous coated cups with tantalum and titanium biomaterial, reconstructive surgeries can achieve improved mechanical support and stability through bone ingrowth compared to conventional cups [2,5,26]. Furthermore, with the use of HXLPE liners in THA implants, there are reduced rates of osteolysis and polyethylene wear [13,19,35,50]. Likewise, HXLPE also has the potential to increase femoral head diameter at the time of liner exchange [22], which further increases stability and decreases the risk for dislocation [3]. One study in our review indicated that no failure occurred in cases which a HXLPE liner was used [25]. Given that recent studies of HXLPE liners have been promising, we expect the survivorship of implants in both primary THA and isolated liner exchange with bone grafting to increase with the incorporation of highly porous cups and HXLPE due to their increased stability, decreased risk of osteolysis progression, and ability to increase femoral head diameter during liner exchange.

Current literature emphasizes that the posterolateral approach is associated with an increased risk of instability and dislocation following THA [3,45,46,55]. Three articles [18,25,27] in this review utilized the posterolateral approach in isolated liner exchange. Griffin et al cited the posterior approach as a possible reason for the increased dislocation rate in their cohort, and this study accounted for 4 out of 11 hips that had dislocations in this review and 3 out of 4 hips that required revision for dislocation. Interestingly, they also reported that of 55 patients in their cohort, 21 patients had decreased femoral head sizes after exchange, and 4 patients had increased femoral head sizes. Two other studies also reported decreasing the femoral head size at the time of revision to allow for a thicker polyethylene liner [37,51]. Given that smaller femoral head sizes are associated with increased dislocation rates [3], the uncontrolled femoral head size changes in these studies reporting the posterolateral approach may have contributed to the dislocation rates reported in this review. We recommend that ILE should be used with caution in patients with small retained acetabular cups that are incompatible with DM, constrained, or HXLPE liners. Small cups limit options for DM or constrained liners and older cups may be limited to the use of conventional PE liners, which are thicker than HXLPE. Our institutional experience suggests that these patients may be limited to smaller femoral head sizes, thus increasing the risk of dislocation following ILE. A more formal analysis is warranted to indicate whether the posterolateral approach, femoral head size, and cup size are significant risk factors for dislocation after isolated liner exchange with bone grafting.

Currently, several classification systems exist to guide the treatment of acetabular defects in revision THA. In particular, classification systems such as the Paprosky system grade based on cup migration in combination with other indicators of osteolysis including the teardrop stability, Kohler line, ischium, and bone loss [42,53]. Given that well-fixed cups exhibit no migration, all cups in this review would be classified under the Paprosky system as Type 1. Likewise, there was no significant loss of bone stock, and the Saleh and Gross classification system would also grade all well-fixed cups in this review as Type 1 [20,47]. Yet, these classification systems do not provide guidelines for the management of periacetabular osteolysis based on the amount of osteolysis present in a well-fixed cup. For example, does the amount of osteolysis in a cup without migration dictate whether revision THA should involve isolated liner exchange or the complete acetabular cup revision? Maloney et al utilized a classification system to divide well-fixed cups into 3 types to guide the decision between isolated liner exchange and cup revision. This system utilizes 6 criteria assessing cup positioning, locking mechanism, metal shell damage, polyethylene liner replacement, implant track record, and implant modularity [39]. However, this system also does not consider the amount of osteolysis in guiding the treatment plan [20]. Therefore, an accurate grading system for periacetabular osteolysis without cup migration is recommended as it would also aid studies in determining if higher grades of periacetabular osteolysis are associated with worse outcomes in isolated liner exchange and if such cases require cup revision or additional surgical interventions.

Furthermore, there is no guideline on how large osteolytic lesions should be in order to prompt intervention, even in asymptomatic patients. Lesions do not always require grafting [37,41], particularly in asymptomatic patients with no documented instability. Liner positioning is also critically important in order to avoid further progression of pain, osteolysis, and/or instability after ILE [18]. One study reported the average size of osteolytic lesions filled with graft and not filled with graft as 682 mm² vs. 96 mm² [37]. The mean size of osteolytic lesions in this review treated with isolated liner exchange and bone grafting ranged from 466 mm² to 682 mm² [24,37]. The lesions in this review may be larger than those reported in CT studies (mean dimension ~16 mm, mean size ~256 mm²) [43]. This suggests that the large-sized lesions in this review were successfully treated with isolated liner exchange. Given the low complication and revision rate, we suggest that lesions with a square size of 450 mm² or larger are particularly considered for isolated liner exchange with bone grafting. In patients that have a history of pain or osteolysis, we recommend that any accessible lesions smaller than 450 mm² at risk of further progression of osteolysis should be debrided and packed with bone graft. Current investigations are developing algorithms for the volumetric classification of osteolytic size via CT to guide the treatment of osteolysis [8,30]. The proper grading and classification of osteolytic lesions based on volumetric size and location will ultimately better inform the decision on when to intervene in patients with evidence of periacetabular osteolysis.

In conclusion, this review of 227 hips with periacetabular osteolysis showed that the retention of well-fixed cups and isolated liner exchange combined with bone grafting of the osteolytic lesions is associated with high survival rates (93.4%) and minimal lesion radiographic progression (1.9%) at short-term to medium-term follow-up. Liner exchange with bone grafting is recommended for the management of large-sized periacetabular osteolytic lesions (>450 mm²) in a well-fixed acetabular cup. The use of a cortical window hole created in the lateral wall is suggested for debriding and impaction grafting of iliac lesions located in DeLee and Charnley Zone I, whereas screw holes or the acetabular component rim should be preferred for lesions located in Zone II. The use of contemporary implants, such as highly porous coated cups and HXLPE liners, may increase femoral head size at the time of liner exchange, and the addition of osteoconductive agents preliner and postliner exchange may increase survivorship and decrease dislocation rates after liner exchange and bone grafting. However, further research is required to generate evidence-based conclusions on the effect of these factors on the clinical outcomes. We recommend that future studies include larger sample sizes (N ≥ 30) with sufficient follow-up (≥5 years) to better characterize the long-term outcomes of ILE with bone grafting, particularly in relation to HXLPE liners and highly porous cups. We also encourage the use of novel PROMs to compare preoperative and postoperative outcomes of ILE with bone grafting, and assessments of achievement of MCID. Finally, we acknowledge the lack of a systemic classification for quantifying size and location of periacetabular osteolysis in well-fixed cups, which is needed for better guidance of surgical management and risk stratification.