Clinical and Radiological Outcome of Acetabular Reconstruction Rings in Complex Primary and Revision Total Hip Arthroplasty

Abstract

Background

Acetabular reconstruction in complex primary and revision total hip arthroplasty (THA) with bone loss poses a great challenge. We aim to evaluate the medium-term clinical and radiological outcome of reconstruction rings used in these difficult situations.

Methods

We retrospectively reviewed a consecutive series of acetabular reconstructions with Muller ring or Bursh-Schneider cage from January 2009 to December 2016. The reconstruction rings were used in 66 hips (65 patients). There were 41 complex primary THA and 25 revision THA. The mean follow-up period was 76 months (range, 37–167 months). Clinical evaluation includes the assessment of Harris hip score, visual analogue scale (VAS) score, limb length discrepancy, and activities of daily living. The radiographs were analyzed for any signs of loosening, osteolysis, acetabular migration, and heterotopic ossification.

Results

The overall survival rate was 95% for revision in aseptic loosening and 87% for any reason at an average follow-up of 6.3 years. Twelve patients died and 9 patients were lost to follow-up leaving 45 patients for final functional analysis. Among the 45 patients, excellent to good results were seen in 33 patients, fair results were seen in 5 patients, poor results were seen in seven patients. Two patients had aseptic loosening and another two patients developed deep chronic infection awaiting two-stage revision.

Conclusion

Reconstruction rings still place a role in the armamentarium for complex acetabular reconstruction. It helps to reliably restore the bone stock, have an acceptable survival rate and satisfactory functional outcome at medium to long-term follow-up.

Introduction

Acetabular reconstruction in complex primary and revision total hip arthroplasty (THA) with bone loss poses a challenge even to experienced surgeons. Restoring the anatomy by building the lost bone stock, reconstructing the original center of rotation and stable fixation of the prosthesis forms the important goals of acetabular reconstruction [1–3]. There are various options to address the deficient acetabular bone stock which includes a spectrum ranging from the high hip center [4], reinforcement rings and cages [1–3], jumbo cup [5], bilobed cup [6], trabecular metal cup with augments [7] to more recently cup-cage construct [8]. The long-term outcomes for these reconstructive options are varied in the literature and need to be individualised according to the particular clinical situation.

Reconstruction rings can be classified into roof reinforcement ring or anti-protrusio cage. The roof ring spans superiorly from the ilium to the inferomedial aspect of the acetabulum and protects the dome of the acetabulum. The cage extends from the ilium superiorly to the ischium inferiorly protecting the whole medial aspect of the acetabulum [9]. The rings and cages span the bone defect, protect the underlying morselized or structural grafts, and distribute the load to the peripheral host bone [1]. The main advantage of reconstruction rings is to improve future bone stock by protecting the underlying bone graft. The disadvantages are the nonporous outer surface which has no potential for biologic ingrowth, increase the chance of fixation failure, and technical difficulty in implantation compared to hemispherical cups.

The literature regarding the survival rate and outcome regarding the reconstruction rings are varied. Some studies [2, 3, 10] have shown excellent and acceptable medium to long-term survival but other studies [11–13] contradicted with adverse results in the short to medium term. Some studies claim a high complication rate when using these rings and cages [14–16]. Our study aims to evaluate the medium-term clinical and radiological outcome of reconstruction rings (Burch-Schneider anti-protrusio cage /Muller roof reinforcement rings) used in patients with acetabular bone deficiencies encountered during the complex primary and revision THA.

Materials and Methods

After getting the approval from the institutional review board, we retrospectively reviewed the consecutive acetabular reconstruction with Muller ring or Bursh-Schneider (BS) cage from January 2009 to December 2016 in complex primary and revision THA from our institution’s database. Patients with other implants used for acetabular reconstruction, patients with tumour or metastasis were excluded from the study. The muller roof ring was selected when there is adequate column support and defects were superior or medial, mainly cavitary defects (Paprosky type 2). The rings were selected after ensuring the segmental integrity of the acetabulum [2]. For more extensive bony defects (Paprosky type 3) which involves anterior or posterior columns, a BS cage was chosen [17]. The study was conducted in compliance with the Helsinki declaration.

The reconstruction rings were used in 66 hips (65 patients). There were 41 complex primary THA and 25 revision THA (Table 1). Of the 25 revision THA, only the acetabular component changed in the 6 hips. The mean follow-up period was 76 months (range, 37–167 months). The preoperative diagnosis in the complex primary group and revision group are given in Fig. 1.

Table 1.

Patient demographics, etiology, bearing surface and stem fixation type used in the patients (N = 66)

Sl. no	Variable	Frequency	Percentage
1	Charnley grade (N = 65)
	A	38	58.5
	B	9	13.8
	C	18	27.7
2	Etiology
	Complex primary	41	62.1
	Revision	25	37.9
3	Approach
	Anterolateral	58	87.9
	Posterior	8	12.1
4	Bearing
	Ceramic on poly	27	40.9
	Metal on Poly	39	59.1
5	Stem fixation type (N = 60)
	Cemented	20	33.3
	Uncemented	40	66.7

Fig. 1.

Schematic graph showing the distribution of patients in complex primary and revision group

Surgical Technique

All the patients were operated in the lateral position by modified Hardinge approach or Southern Moore posterior approach (Table 1). After exposing the acetabulum, fibrous tissues were removed and the state of roof, floor, and walls was assessed and the bone defects were classified according to Paprosky classification. Proper preparation of acetabular bed done by reaming the acetabulum gently and the defects were curetted and packed with morselized autograft in primary and frozen morselized allograft in revision THA. Either reinforcement ring or cage was chosen depending upon the underlying bone defect.

Muller reinforcement ring (Zimmer, Warsaw, IN) covers four-fifths of a hemisphere with a titanium surface [10, 18]. The BS cage also has a titanium surface, the superior flange fixed with multiple 6.5 mm cancellous screws, and the inferior flange slotted into the ischium [15]. The technical difficulty with the BS cage is that the minimum size is also very large for our patient population, especially in females [19]. Wide exposure for the placement of flanges is another problem with the BS cage that risks damage to the superior gluteal nerve [15].

The reconstruction ring was impacted to attain primary press fit on the underlying host bone and augmented with multiple titanium cancellous screws. On average, three screws (range, 2–6) for Muller ring and five screws (range, 4–8) were used to improve the stability. Every care is taken to place the ring in the anatomical position and if the bone defect is large, then the ring was placed at an acceptable high hip center. After ensuring stability, a polyethylene cup was cemented in the correct version independent of the ring placement [2, 10, 18]. Postoperatively, the patients were mobilized with partial weight-bearing walking with walker frame support for 6 weeks and elbow crutch walking for another 6 weeks. Progressive increase in the amount of weight-bearing increased after 3 months. Patients followed up in 3 months, 6 months, and thereupon yearly once.

Patients and Follow-Up

Twelve of 65 patients died during the final follow-up. Nine patients were lost to follow-up leaving 45 patients for final functional analysis. There were 37 hips in the Muller ring group and 8 hips in the BS cage group at the final follow-up. Thirty-six patients had scheduled follow-up and the remaining nine patients were contacted through telephone. The follow-up period was calculated from the date of surgery to the date of final follow-up or date of death with the implant in situ. In the patients who were lost to follow-up, the last available radiographs were taken for the radiographic measurement.

Clinical Evaluation

Clinical evaluation includes the assessment of postoperative visual analogue scale (VAS) score at final follow-up, limb length discrepancy, activities of daily living, and Charnley’s patient grade [2]. All the patients were evaluated using the Harris hip score at the final follow-up and for those who cannot able to visit the hospital, a modified Harris hip score (HHS) was used [20]. Patients also asked about subjective contentment following the surgery with options very satisfied, predominantly satisfied, less satisfied, and not satisfied [3].

Radiographic Evaluation

At each follow-up, a standardized anteroposterior (AP) view of the pelvis, centering the pubic symphysis were taken. For radiographic analysis, the immediate postoperative radiograph was compared with the latest follow-up pelvis radiograph. All radiographic measurements were made using the Picture Archiving and Communication Systems (PACS) 3.0 software and the magnification corrections were digitally performed and corrected to the true head size.

Preoperative acetabular defects were classified according to the Paprosky classification. Postoperatively, the radiographs were analyzed for any signs of loosening, osteolysis, and acetabular migration. The osteolysis around the cup was assessed by noting any progressive radiolucent lines more than 2 mm in DeLee and Charnley acetabular zones compared to the radiographs obtained immediately after surgery [18]. The inclination angle of the cup and the ring was measured. The migration of the ring is assessed according to Nunn et al. criteria [21]. Vertical displacement measured from the center of the cup relative to the inter-teardrop line and the horizontal displacement measured from the center of the cup to the vertical line at the ipsilateral teardrop perpendicular to the inter-teardrop line.

The stability of the cage is classified according to Gill et al. [22]. Migration more than 5 mm or progressive radiolucent lines medially and superiorly or around the screw or screw breakage is classified as ‘definitely loose’. Progressive radiolucent lines around the medial and superior part of the cage are classified as ‘probably loose’. If there are non-progressive radiolucent lines present medial and superior to the cage without involving the screws are defined as ‘possibly loose’. The change in the inclination of the cup more than four degrees with more than 3 mm of vertical or horizontal position change also taken as the parameters for loose cage [10, 23].

The bone graft incorporation was assessed using Gross grading in the pelvis AP view with less than one-third of resorption as minor and more than half as severe and in between the two as moderate resorption [24]. The grafts were also analysed for evidence for trabecular bridging at the host bone-graft interface and no evidence of fragmentation or radiolucent lines [10, 24]. The presence of any heterotopic ossification is graded according to Brooker et al. [25] (Fig. 2).

Fig. 2.

Radiographs showing (a) posttraumatic arthritis left hip with heterotopic ossification (b) CT scan showing posterior wall defect with heterotopic ossification (c) 8-year follow-up radiograph showing acetabular reconstruction with Muller ring

Statistical Analysis

Descriptive statistics were reported as mean (SD) for continuous variables, frequencies (percentage) for categorical variables. The paired t test was used to find the significant difference between preoperative and postoperative HHS Scores. Kaplan–Meier survival analysis is done for all the hips to determine the cumulative revision rate of reinforcement rings. The endpoints noted are revision for aseptic loosening, revision of ring for any reason, and worst-case scenario assuming the lost to follow-up as aseptic loosening at the last date of visit. The patients without any revision were censored at the date of the last follow-up. The 95% confidence intervals were calculated for 5-year survival. p < 0.05 was considered significant. Data were statistically evaluated with IBM SPSS Statistics for Windows, Version 20.0., IBM Corp., Chicago, IL.

Results

The mean follow-up period was 76 months (range, 37–167 months). The average age of the patients was 55.3 years. There were 41 males and 24 females. Eighteen patients belong to Charnley preoperative class 3 (Table 1). The bearing surface used and stem fixation type are given in Table 1. The Muller rings were used in 54 hips and the BS cage was used in 12 hips (Fig. 3). The bone defect was stratified according to the Paprosky classification (Table 2). At an average of 6.3 years, the overall survival rate was 95% for revision for aseptic loosening and 87% for revision for any reason (Fig. 4A, B). In the worst-case scenario assuming the loss to follow-up as aseptic loosening at the last date of visit, the survival rate was 75% at 6.3 years (Fig. 4C).

Fig. 3.

Radiographs showing (a) post-tubercular arthritis right hip with Paprosky type 3b bone defect and post-Girdlestone arthroplasty left hip (b) coronal and axial section of CT scan (c) 6.5-year follow-up radiograph showing acetabular reconstruction with BS cage on the right side and Muller ring on the left side

Table 2.

Type of reinforcement ring and graft used in various Paprosky defects (N = 66)

Sl. no	Variable	Frequency	Percentage
1	Cage type
	Muller	54	81.8
	Bursh Schneider (BS)	12	18.2
2	Muller (n = 54)
	Paprosky classification
	2A	15	27.8
	2B	22	40.7
	2C	12	22.2
	3A	5	9.3
3	BS (n = 12)
	Paprosky classification
	3A	4	33.3
	3B	7	58.3
	Pelvic discontinuity	1	8.3
4	Graft
	Autograft	28	42.4
	Allograft	27	40.9
	Both	11	16.7
5	Heterotopic ossification
	Grade 0	57	86.4
	Grade 1	5	7.6
	Grade 2	1	1.5
	Grade 3	3	4.5
	Grade 4	0	0

Fig. 4.

The graph showing cumulative revision rate of the reconstruction rings for (A) aseptic loosening as the endpoint (B) any revision as endpoint and (C) worst-case scenario assuming all patients lost to follow-up revised for aseptic loosening

Among the 45 patients, functional outcome was evaluated with Harris hip score, excellent to good results were seen in 33 patients, fair results were seen in 5 patients, poor results were seen in seven patients. There was no statistically significant difference (p value = 0.569) between the preoperative Harris hip score among the primary group (19.11 ± 8.47), posttraumatic group (23.36 ± 10.86) and revision group (22.75 ± 9.99). There was no statistically significant difference (p value = 0.507) between the postoperative Harris hip score among the primary group (73.78 ± 23.39), posttraumatic group (81.63 ± 19.23) and revision group (82.88 ± 17.15). However, there was a statistically significant improvement in the postoperative Harris hip score compared to the preoperative scores among the three groups (p < 0.05) (Fig. 5).

Fig. 5.

Radiographs showing (a) aseptic loosening of cemented total hip arthroplasty-right side (b) 10-year follow-up radiograph showing acetabular reconstruction with Muller ring and medial side mesh

The subjective contentment was very and predominantly satisfied in 80%, less satisfied in 11%, and not satisfied in 9%. The VAS score at the final follow-up shows many patients were pain-free or with very minimal pain with VAS of less than three in 87.7% of patients. The activities of the daily living assessment show many patients were housebound or doing light work as in Table 3. Limb length discrepancy less than 2 cm in 89.4% of patients (Table 4).

Table 3.

Various postoperative radiological parameters like offset, limb length discrepancy, implant stability and graft incorporation

Sl. no	Variable	Frequency	Percentage
1	Average abduction angle of the ring	41.19 ± 7.47	Range (26.20–65.00)
2	Horizontal offset
	1–10	50	75.8
	10–20	13	19.7
	20–30	2	3.0
	> 30	1	1.5
3	Vertical offset
	1–10	39	59.1
	10–20	23	34.8
	20–30	3	4.5
	> 30	1	1.5
4	Limb length discrepancy (lengthening) (n = 26)
	1–10	20	76.92
	10–20	4	15.38
	20–30	2	7.69
	Limb length discrepancy (shortening)
	1–10	23	57.5
	10–20	12	30
	20–30	5	12.5
5	Gill classification
	No	55	83.3
	Possibly	5	7.6
	Probably	3	4.5
	Definitely	3	4.5
6	Graft incorporation (gross grading)
	Good	55	83.3
	Minor	5	7.6
	Moderate	3	4.5
	Severe	3	4.5

Table 4.

Postoperative follow-up activities, pain score and level of satisfaction (n = 45)

Sl. no	Variable	Frequency	Percentage
1	Activities of daily living
	Heavy	4	8.9
	Light	20	44.4
	Office	8	17.8
	Sitting	13	28.9
2	Postoperative visual analogue score (VAS)	1.64 ± 1.58	Range (0–6)
3	Satisfaction
	Very much	13	28.9
	Predominantly	23	51.1
	Less	5	11.1
	Not satisfied	4	8.9

In the radiological analysis, out of 66 hips, 55 hips had good graft incorporation, 8 had reasonable graft incorporation but no loosening and only 3 hips had evidence of definite loosening. The average inclination angle of the ring was 41.2 ± 7.5° (Table 4). The mean center of rotation was restored within 30 mm in 65 hips (Table 4). No loosening was noted in 55 hips (83.3%). Three cups were definitely loose, three were probably loose and five rings were deemed as possibly loose according to Gill’s classification. No heterotopic ossification was noted in 86.4% of the hips at the final follow-up. Most patients with heterotopic ossification had a previous history of trauma. 7.6% had grade 1, 1.5% had grade 2 and 4.5% had grade 3 heterotopic ossification. None of the patients had ankylosis.

Complications

Two patients had aseptic loosening of the ring and revised after three years postoperatively. Two patients developed a superficial infection which settled with debridement and antibiotics. One patient developed proximal deep vein thrombosis after 3 months from surgery which recanalized after anticoagulant therapy. Poor results (HHS < 70) were seen in seven patients at the final follow-up. Out of the seven patients, two patients had persisting discharging sinus with chronic infection with loosening and awaiting two-stage revision. Two patients had pre-existing neurological deficits after hip fracture-dislocation. One patient had cervical myelopathy and the other had degenerative scoliosis for which she underwent decompression and instrumented fixation and fusion. One patient sustained a periprosthetic fracture of the femur after 4 years of surgery for which she underwent open reduction and plate fixation.

Discussion

The key findings of our study are acetabular rings and cages still have a role in complex acetabular reconstruction. The acetabular rings have an acceptable survival rate at medium-term follow-up. The correct selection of either ring or cage for the particular bone loss will lead to satisfactory functional outcome in the medium to long-term follow-up. The reinforcement rings help to protect the graft and help to improve the future bone stock.

Complex acetabular reconstruction with bone graft alone without mechanical support led to a high failure rate as high as 60% at 16 years [11, 26]. Hence, it is essential to protect the graft with reconstruction rings which will stabilise and facilitate the integration of the bone graft to the host bone. Modern acetabular reconstruction has gadgets of many implants, but the reconstruction rings have distinct advantages. The acetabular rings are flexible and malleable during impaction which leads to press-fit within the acetabulum without the need for more host bone removal, unlike to accommodate the larger hemispherical cup [10]. The morselised graft can be packed to fill the underlying defect which acts like an impaction bone grafting to improve the bone stock. It is flexible in placement based on the available host bone and the polyethylene cup can be positioned independently on the desired version and inclination [27]. The multiple screws in different direction give inherent stability and screwheads can be locked by the cement used for fixing the polyethylene which increases the stability of the whole implant [28]. The titanium rings avoid stress shielding phenomenon since it is less rigid than the stainless steel implants [19, 29]. Moreover, small acetabular size is encountered in the Asian population which is not amenable for jumbo cups or cup-cage construct.

The outer surface has no potential for biologic ingrowth, hence the survivorship of the implant in the younger patients is a concern. The long-term results of some studies show poor survival rate with disappointing results on using reinforcement rings. Philippe et al. [11] showed restoring the center of rotation leads to better functional outcomes. They showed a survival rate of 77.9% at 13 years, but the caveat is 87% of the hips where the rings used were of Paprosky type 3 defects. Massin et al. [12] showed that the Muller rings could not maintain reliable hip function and start failing after the first decade, but about 69% of hips were AAOS type 3 and 4 defects. But, Kosters et al. [3] selectively used Muller rings in Paprosky type 2 defects and showed 86.7% survival rate after 10-year follow-up and advocate the use of rings in acetabular revision surgery. Hence, our selection criteria for using the muller rings should be with intact columnar support and only Paprosky type 2 cavitary defects and BS cage for Paprosky type 3 segmental defects.

The survivorship of using rings in complex primary and revision in recent long-term studies are encouraging. Sirka et al. [18] analysed 321 hips where muller rings used for primary THA and found the cumulative revision rate of 3.4% for aseptic loosening and 7% for revision due to any reason at 20 years. They have suggested excellent long-term survival with a low revision rate. Greber et al. [10] showed excellent long-term results in 259 revision THA treated by Muller rings with only 6.2% cumulative revision rate for aseptic loosening. Schlegel et al. [2] reported an acceptable outcome with 90% survival at 8 years. In a recent study, the outcome of porous tantalum cups compared to the reinforcement rings and concluded no clinical or statistical difference between both groups at 6-year follow-up [27]. Our study also shows a comparable survival rate with 95% for revision for aseptic loosening and 87% for revision for any reason at an average of 6.3 years.

Bursh-Schneider cage is indicated in extensive defects like Paprosky type 3 defects and for pelvic discontinuity. Perka et al. [19] warned about using the BS cage when the posterior columnar support is deficient which will compromise on the long-term stability. The long-term follow-up of 16 years showed a survival rate of 72% when using BS cage for pelvic discontinuity [15]. Given the complexity of the procedure, it is advocated as an acceptable treatment [16, 19]. We have used BS cage in Paprosky type 3 defects and pelvic discontinuity and found no revision for aseptic loosening at an average of 6.3-year follow-up.

The clinical outcome using HHS shown fair to poor outcomes in some studies [2, 3, 11, 16], but the patient satisfaction scores were better [3]. Interpretation of HHS would be difficult since many patients have multiple comorbidities. Our study also shows a fair outcome in the final follow-up, but the satisfaction rates are better when compared to the preoperative morbidity. The mortality rate in these patients is reported as high as 49% at 10-year follow-up [13] and our study also shows a higher mortality rate of 30% due to the complex procedures at the elderly age group and other associated comorbidities.

Posttraumatic acetabular fractures pose a challenge in reconstruction because of the distorted anatomy, underling non-union, hindrance by underlying protruding metal implants, and other associated injuries. This subset of patients has an increased chance of infection, dislocation rates and inferior outcomes when compared with THA for primary osteoarthritis [30]. In our study, we find that infection and heterotopic ossification are higher in these groups of patients but there is no statistically significant difference in the functional outcome in the final follow-up.

The other important concern in complex reconstruction is dislocation. Philippe et al. [11] reported a dislocation rate of 7.4%, but we had no dislocation at our final follow-up. Since the polyethylene cup and rings can be independently placed, it offers the advantage of optimal stability thereby mitigating the chances of dislocation when compared to hemispherical cups [27]. Periprosthetic joint infection after complex procedures is high because the operating time for placement of rings is higher than normal. There were two patients with deep infection in our study, which was comparable to the other studies [2, 10, 18].

The limitation of our study is its retrospective nature of the study without any comparison group. There were nine patients (13%) who were lost to follow-up. However, the dropout rates were comparable or low when compared to the other series [11]. We have analysed only the Muller ring and BS cage in our study. Hence the results are not applicable for the other types of reconstruction rings. Nevertheless, this is one of the largest consecutive single centre series evaluating the medium to long-term outcome of reinforcement rings.

Conclusion

The reconstruction rings even though has a non-biological fixation, still got a role in complex primary or revision acetabular reconstruction, particularly in a smaller acetabulum encountered in the Asian population. The correct selection of reconstruction rings for the particular bone loss is crucial for a satisfactory functional outcome. We emphasise that muller rings should be used in intact columnar support and Paprosky type 2 cavitary defects and Bursh-Schneider (BS) cage for Paprosky type 3 segmental defects. The reinforcement rings help to protect the graft and help to improve the future bone stock. The acetabular rings are cost-effective and have an acceptable survival rate at medium to long-term follow-up.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declarations

Conflict of interest

Natesan Rajkumar, Dhanasekaran Soundarrajan, Panda Chandan kumar, Palanisami Dhanasekararaja and Shanmuganathan Rajasekaran declare that they have no conflict of interest.

Ethical standards

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008.

Informed consent

Informed consent was obtained from all patients for being included in the study.