Results of Revision of Total HIP Arthroplasty for Alumina Ceramic-on-Ceramic Bearing Fracture

Abstract

Purpose

The purpose of this study was to compare the results of revision total hip arthroplasty after fracture of primary ceramic components using different type of revision bearing surfaces.

Methods

We analysed the results of 16 patients with a follow-up more than 3 years after first revision. 6 were revised to ceramic-on-ceramic (CoC) bearing, 9 to metal-on-polyethylene (MoP) and 1 to ceramic-on-cross-linked polyethylene (CoXLP) bearing.

Results

The mean follow-up was 87 months. Patients with revision to CoC had higher Harris Hip Score (HHS) of 89 points in comparison to the patients with revision to MoP with 84 points. Radiographic examinations revealed visible eccentric polyethylene wear with osteolysis in 3 out of 9 patients revised to MoP. There were no detrimental x-ray changes in patients revised to CoC components.

Conclusions

We consider CoC as the best option at revision operation for ceramic component fracture.

Keywords

Ceramic fracture Revision Total hip arthroplasty

Introduction

Alumina (Al₂O₃) ceramic-on-ceramic (CoC) bearing couple has been introduced in the 1970s and has been extensively used in total hip arthroplasty (THA) surgery (1). Several randomised and observational studies have confirmed superior wear characteristics of alumina-on-alumina bearings in comparison with traditional bearings with less osteolyses and aseptic loosenings (2–3–4–5–6). Large mid- and long-term series demonstrated excellent survivorships with revision rate due to aseptic loosening ranging from 99% at 7 years (7), 96% at 9 years (8), and 95.6% at 10 years (9).

The favourable wear properties of CoC bearings and consequent low aseptic loosening rates are partially off-set by failures due to squeaking and fracture of bearing components. Therefore, the clear advantage of CoC against traditional bearings has not yet been proven (10).

According to the reports of the manufacturers of ceramic components, femoral head fractures decreased from 13.4% for those produced before 1990 to 0.004% for Biolox® femoral heads produced after 1994 as the production quality improved (11). In the largest series published a 0.5% (40 out of 8,022 CoC THRs) incidence of alumina ceramic component fractures was reported (12).

Despite several reports (12–13–14) stating that the results after revision for ceramic component fracture are acceptable, the conclusive evidence to support these statements are still missing. Either the series were small (14, 15), or clinical follow-up scores were incomplete (12–13–14–15). Long-term clinical results for the revised patients are still scarce as there are only few larger studies (13, 16, 17) presenting clinical data regarding the outcome after revision for ceramic component fracture (12, 13, 18). Moreover, it is usually stated that ceramic component fracture presents a potential catastrophic failure (15, 18) but strong evidence to support this notion is missing.

The purpose of our study was to evaluate the clinical and radiological outcomes of THAs revised for fracture of third-generation alumina ceramic components with the addition of histological analysis of the surrounding tissue and to discuss the optimal revision procedure in these patients.

Patients and methods

Patient data

From January 2000 to December 2010, 1,239 primary consecutive THAs with alumina CoC bearings were performed at our institution. The selection of the bearing surface was decided by the surgeon, with a clear preference for CoC in case of young and active patients.

From the hospital Valdoltra Arthoplasty Register (19) we analysed the results of 16 patients that were revised for alumina CoC component fracture. The inclusion criteria was more than 3 years of follow-up time after revision operation. Among these, 6 revisions were due to fractured ball head, 6 due to fractured liner and 4 due to both head and liner fracture. The mean time from implantation to fracture was 35 months (0-102 months). In 4 patients a traumatic event (fall) was reported prior to the fracture. From the hospital register the data about age, gender, side, initial diagnosis (Tab. I), primary and revision implant type (Tab. II), and complications associated with revision (Tab. III) were abstracted. In 8 cases the periprosthetic tissue samples were collected during revision operation for histological analysis. In the remaining 8 cases tissue samples were not taken during surgery. The inability to fully control the sampling protocol is a disadvantage of retrospective study as ours. The patients were invited for an additional follow-up visit and all accepted the invitation. The follow-up comprised a clinical examination including range of motion measurements, pain visual analogue score (VAS) ranging from 1 to 10, Harris Hip Score (HHS), and radiographic evaluation.

TABLE I

Demographic data

Number of hips	16
Left	7
Right	9
Number of patients	16
Gender
Male	11
Female	5
Mean BMI (kg/m²)	30 (range 24.5-38.6)
Mean age at revision (years)	63 (range 45-74)
Initial diagnosis
Osteoarthritis	11
Posttraumatic osteoarthritis	4
Osteoarthritis secondary to dysplasia	1

TABLE II

Data on implants used at primary and revision operations

	Stem	Bearing: head-on-cup
Primary implant	12 SL-Plus (Smith&Nephew)	16 CoC: alumina-on-alumina
	1 Profemur Z (Wright)
	1 Corail (DePuy)
	1 Anca (Wright)
	1 Ribbed (Link)
Revision implant	1 from Profemur Z (Wright) to Profemur Z (Wright)	9 MoP: 5 stainless steel-on-PE 4 CoCr-on-PE
	1 from Anca (Wright) to Profemur Z (Wright)	6 CoC: alumina-on-alumina 1 CoXLP: delta-on-XLPE

The study received Institutional Ethical Review Board approval.

Implant data

Biolox Forte® (CeramTec GmBh, Plochingen, Germany) components were selected as a bearing material at primary operation in all patients. Primary cups and femoral stems differed among patients (Tab. II). 3 different bearing materials were used at revision surgeries: CoC, metal-on-polyethylene (MoP) and ceramic-on-cross-linked polyethylene (CoXLP). 6 patients were revised to CoC, 9 to MoP, and 1 patient to CoXLP (Durasul®, Zimmer) (Tab. II). The choice of revision components was at the discretion of the surgeon. We changed the cups in all cases and in 2 patients the femoral component was revised because of a damaged taper. In revision operations Watson-Jones and Harding (transgluteal) approaches were used. In 2 patients extra bone fixation with screws was needed for the cup.

1 patient revised to MoP sustained a second revision due to early postoperative dislocation. The standard postoperative regime included patient assessment 3, 6, and 12 months after the index surgery, and at 1- or 2-year intervals thereafter. No patient was lost to follow-up.

Radiographic analysis

At the latest follow-up visit anteroposterior radiographs of the pelvis, centred over the pubic symphysis, were taken and analysed according to the method of DeLee and Charnley on the acetabular side (20) and that of Gruen et al (21) on the femoral side. Radiographic acetabular loosening was defined as the presence of a complete radiolucent line in all 3 zones >1 mm, >5° change in cup inclination, or cup migration of >3 mm (22). The stability of the femoral component fixation was assessed using the classification described by Engh et al (23, 24). The femoral component was considered loose if serial radiographs demonstrated a change in the position of the femoral component (i.e., subsidence ≥2 mm, or varus/valgus tilting). Osteolysis was classified as linear or expansile, periarticular, or remote from the joint, according to the criteria proposed by Zicat et al (25). For this particular series osteolysis was defined as minor for radiolucent lines more than 5 mm long and 2-5 mm wide, and major in case of expansile lesions more than 5 mm wide.

Histological analysis

In 8 cases during revision operation periprosthetic tissue samples (2-4 tissue samples per patient) have been collected. For the study purposes all histological slides were re-evaluated by an experienced pathologist. The histological types of periprosthetic membrane (26) as well as morphological characteristics of periprosthetic tissue according to modified Mirra classification (27) were determined.

Results

Macroscopic images, radiographic and histological results after fracture

Examples of fractured ceramic heads and alumina ceramic inlay that occurred in 3 different patients are presented in Figure 1. X-ray images prior to the revision surgery show the position of fractured components (Fig. 2). No trauma was involved in any of these events. In the first patient the ceramic head fractured spontaneously and the hip dislocated 3 weeks following surgery (Fig. 2A). The head may have broken from the mounth region; the fracture was probably fast. The revision followed 2 days after fracture. In the second patient the ceramic inlay of the sandwich SPH combination fractured 11 months following surgery (Figs. 1A and 2B). No trauma was involved and fracture was preceded by cracking noises and pain for 5 days. It seems that the alumina ceramic liner was not appropriately supported by taper-fit in the socket. The revision was carried out 11 days following the fracture event.

Fig. 1

Images of fractured ceramic components of 3 different patients: (A) fractured ceramic inlay SPH Lima after 11 months in situ; and (B) fractured ceramic head and ceramic inlay of the Trilogy AB Zimmer after 24 months in situ.

Fig. 2

X-ray images of fractured ceramic components: (A) fractured ceramic head within Duraloc cup after 3 weeks in situ; (B) fractured ceramic inlay SPH Lima after 11 months in situ; and (C) fractured ceramic head and ceramic inlay of the Trilogy AB Zimmer after 24 months in situ.

In the third patient, both head and inlay of nonsandwich Trilogy AB combination fractured after 2 years in situ (Figs. 1B and 2C). It seems that fractures were rather flat and smooth. Head probably burst first, as cup was practically undamaged except for the metal traces which formed during the contact with metal neck after the fracture, as the revision was carried out 15 days following the fracture event. Detailed description of the fractographic evaluation of some cases have been reported earlier (28) and further analysis of all the cases is beyond the scope of the present study.

The mean time from implantation to fracture was 35 months (0-102 months) for the whole cohort. There was no statistically significant difference between the mean time from implantation to fracture for the patients later revised to MoP (34 months, from 0.1 to 76) or CoC groups 23 months (11-52 months) (p = 0.75) (Tab. III). Following the fracture, the time span from the event to revision operation differed between the groups, i.e. MoP (15 days [1-45 days]) or CoC groups (22 days [2-90 days]) but the difference was not statistically significant (p = 0.71) (Tab. III).

TABLE III

Results of revision operations for the 2 revision bearing types

Revision bearing surface	Follow-up (months)	HHS	Time to fracture (months)	Time from fracture to operation (days)	Radiographic changes	Complications
CoC	94(60-134)	89(81-96)	23(11-52)	22(2-90)	1 linear radiolucent lines in femoral area in zones 1 and 7	no: 2mild pain: 3 squeaking: 1
CoC	94(60-134)	89(81-96)	23(11-52)	22(2-90)	1 minor linear osteolysis around the femoral component in zone 1	no: 2mild pain: 3 squeaking: 1
MoP	87(62-130)	84(61-96)	34(0.1-76)	15(1-45)	2 extensively worn and deformed metal head with only slightly worn polyethylene liner	no: 5 mild pain: 4
					1 major expansile osteolysis in acetabular area in zone II
					1 minor linear osteolysis around the femoral component in zone 1
					1 minor osteolysis in acetabular area in zone II

Data are given as mean values.

Histological slides from all 8 cases show characteristics of the indeterminate type periprosthetic membrane (type IV) with connective tissue low in macrophages and giant cells and rich in collagen fibres (26) (Fig. 3). In order to find possible pattern of histological response to ceramic fractures, the samples were analysed in terms of time in situ (prior fracture), type of implant, cause of fracture and time between fracture and revision operation. Except for the latter, no significant effect of other factors on the histological response was observed. Thus, collected histological samples were divided into 3 groups: group I – patients operated 1 or 2 days after fracture (3 cases); group II – patients operated between 3 and 20 days after fracture (2 cases); group III – patients operated more than 1 month after fracture (3 cases). Histological slides from group I show connective tissue with some bone fragments without ceramic particles. In periprosthetic tissue from groups II and III, ceramic particles of different sizes (some even larger than 30 µm) were found.

Fig. 3

Periprosthetic tissue from patient operated 2 days after fracture shows connective tissue and several bone fragments (A). Histological section from patients operated 3-20 days after the fracture showing ceramic particles mostly extracellulary (B). In periprosthetic tissue samples collected from patients operated more than 1 month after fracture beside small ceramic particles (C) also some large particles (>30 µm) were found (D).

Clinical and radiographic results following revision

The mean follow-up of the revised THRs ranged from 46 to 134 months, on average 87.3 months. There was 1 second revision in a patient revised to MoP due to early postoperative dislocation. At the latest follow-up visit all patients were satisfied with the final result including the patient that had second revision due to dislocation. No other medical or surgical complications were observed in all groups.

7 patients (5 out of 9 revised to MoP and 2 out of 6 revised to CoC) were asymptomatic and capable of performing heavy labour or sport activity. 8 patients (4 out of 9 revised to MoP, 3 out of 6 revised to CoC and 1 revised to CoXLP) were complaining of mild pain and 1 patient (revised to CoC) reported squeaking when walking (Tab. III). Mild pain was defined as 0-3 out of 10 points on VAS score.

The mean HHS was 86 (range 61-96). Data of patients revised to CoC and MoP bearings are given in Table III. The average flexion for all patients was 117.5° (range 90°-135°), external rotation 42.9° (range 35°-70°), internal rotation 17.7° (range 0°-40°), abduction 42.1° (range 20°-60°), and adduction 30.4° (range 15°-40°). Patients revised to CoC bearing achieved a slightly higher HHS of mean 86.2 points in comparison to mean 83.3 points for the patients revised to MoP bearing but the difference was not statistically significant. The patient who suffered a dislocation had a polyethylene liner with an elevated lip and a new metal femoral ball leaving the socket and the stem in situ. At the final follow-up visit he achieved a HHS score of 93.

Radiographic results

According to the criteria in radiographic analysis we observed no loosening of acetabular or femoral components. In patients revised to MoP eccentric polyethylene wear with osteolyses was visible in 3 out of 9 patients. There was 1 major expansile osteolysis in acetabular area in zone II in 1 patient, minor osteolysis in acetabular area in zone II in 1 patient and minor linear osteolysis around the femoral component in zone I in 1 patient. In 2 patients with MoP bearing we observed extensively worn and deformed metal head. In contrast, polyethylene liner was only slightly worn. These patients were asymptomatic. In patients revised to CoC linear radiolucent lines in femoral area in zones 1 and 7 were observed in 1 patient, and minor linear osteolysis around the femoral component in zone 1 in another. Heterotopic ossifications Brooker type I were observed in 1 patient with MoP and in 1 patient with CoC bearing, Brooker type II in 2 patients with MoP and Brooker type III in 1 patients with MoP. No detrimental radiographic changes were noted in patients revised with CoC or CoXLP bearings.

Discussion

There is limited information on the long-term results of revision for fracture of ceramic components (12, 13). Component fracture is a serious complication and the results of the revision are unfavourable if the revision bearing couple was not CoC or CoP (12, 13). Among reported studies only few comprised larger group of patients with longer follow-up (13, 16); the others were case reports (15).

Traina et al (12) analysed 40 patients with fractured ceramic bearing (16 heads and 24 liners). There were 5 Biolox®, 33 BioloxForte®, and 2 BioloxDelta® components. Among the 30 patients revised to CoC, 93.3% achieved good results with no revision or osteolysis. In contrast, among 8 patients revised to MoP 87.5% had poor result with 1 revision and 6 patients developing osteolysis and polyethylene wear.

Koo et al (17) reported the 5-year results on 24 revisions for fractured ceramic head and noted poor result for those with stem left in situ, but good results if the stem was exchanged. The latter was probably due to taper failure, as new heads were placed on damaged tapers. High failure rate in the group where the stem was not revised could be also related to the fact that in this group mainly the MoP liners were used, in contrast to CoC bearings which were mainly used in the group where the stem was revised. We have not found a similar association in our study.

Allain et al (13) reported on a mean 3.5 year result of a meta-analysis of 104 patients revised due to a ceramic head fracture. The 63% survival rate at 5 years was poor and it was significantly worse if the cup had not been revised, and if the new femoral head was made of stainless steel. Radical synovectomy had a positive influence on the final result.

Similarly, Sharma et al (14) published 10-year follow-up results for 8 patients revised to a MoP bearing after a fractured ceramic bearing. No increased wear rate compared to matched controls was observed. The extensive synovectomy with pulsatile lavage was considered the reason for the optimal outcome.

There was only 1 revision to CoXLP in our study therefore we cannot objectively judge the results of this revision combination. In contrast to the results of other reports (13, 18), in our series excellent clinical results were achieved with either revision to MoP or CoC. Revisions to CoC gave slightly better but statistically not significantly different clinical results. However, patients revised to MoP bearings achieved worse radiographic results, as shown in other reports (12, 13) as well. Among 9 patients revised to MoP, worse radiological results with 3 cases of osteolysis were observed; moreover, 2 patients showed increased wear of the metal head. At the same time polyethylene liner remained virtually unworn. The incidence of this extreme behaviour was probably related to the extent of surgical debridement during the revision surgery, as it was shown that even if the fractured ceramic bearings were revised to MoP, it was not detrimental if only the surgical debridement and lavage was meticulous and radical enough (14). The ceramic debris released during ceramic fracture with time got trapped in the bearing, imbedded in the polyethylene liner, and, being harder than the metal head, damaged the metal head while protecting the liner.

The principal factors responsible for fracture are inappropriate geometry of the head/neck design, especially for patients of large mass or performing energetic movements, off-axis loading, and damage due to inappropriate handling during surgery, and in the case of acetabular components, by wrong socket positioning - implantation at inappropriate angles (28).

Although the revisions to MoP and CoC both gave excellent clinical results (Tab. III), radiographic analysis indicates that the CoC bearing is the optimal bearing couple to be used after ceramic fracture. This is valid for fracture of the head, as well as of the liner. The explanation lies in better tribological behaviour of the ceramics, especially high resistance of ceramics to third body wear. More wear and osteolyses were observed in patients revised to MoP, similarly as noted in other studies (12, 13, 16).

Based on the available data the revision operation after a ceramic component fracture should be performed as soon as possible to avoid ceramic particles spread, as shown by histologic analysis. It is important to achieve optimal alignment of revision components to avoid impingement related damage.

Extensive synovectomy and thorough pulsatile lavage are helpful to remove as much of ceramic debris as possible to reduce the risk of third body wear (12–13–14, 16). In the present study the revision to CoC is confirmed to be a preferred bearing choice with better radiographic outcomes. We would need more patients revised to CoP or CoXLP to properly evaluate the results of revisions to these bearing combination. Revision to MoP may be considered a salvage option only if thorough synovectomy and pulsatile lavage is adopted (14). In the case of a fractured ceramic head, the taper must be carefully inspected during the revision. If damaged, its exchange must be considered to avoid subsequent taper failure due to a damaged femoral taper surface. Due to worse radiological results at a mean follow-up of 87.3 months we expect that for the patients revised to MoP bearings more symptoms will evolve with time than in patients revised to CoC bearings.

According to our knowledge, this is the first study that directly compares the results of revision operations after fracture of primary ceramic components for different type of revision bearing surfaces. As ceramic fracture is a relatively rare event, the study cohort is relatively small. Nevertheless, based on the results obtained most probably CoC bearings should be preferentially used at revision of fractured ceramic component.

Footnotes

Financial support: None.

Conflict of interest: None.

References

Rieger

Ceramics in orthopedics -30 years of evolution and experience. In: Rieker C, Oberholzer S, Wyss U, eds. World Tribology forum in Arthroplasty. Bern

Verlag Hans Huber

2001;283–94.

D'Antonio

Capello

Manley

Naughton

Sutton

A titanium-encased alumina ceramic bearing for total hip arthroplasty: 3- to 5-year results.

Clin Orthop Relat Res. 2005;441(441):151–158.

D'Antonio

Capello

Manley

Naughton

Sutton

Alumina ceramic bearings for total hip arthroplasty: five-year results of a prospective randomized study.

Clin Orthop Relat Res. 2005;(436):164–171.

Capello

Dantonio

Feinberg

Manley

Alternative bearing surfaces: alumina ceramic bearings for total hip arthroplasty.

Instr Course Lect. 2005;54:171–176.

Capello

D'Antonio

Feinberg

Manley

Naughton

Ceramic-on-ceramic total hip arthroplasty: update.

J Arthroplasty. 2008;23(7) (Suppl):39–43.

Milošev

Kovač

Trebše

Levašič

Pišot

Comparison of ten-year survivorship of hip prostheses with use of conventional polyethylene, metal-on-metal, or ceramic-on-ceramic bearings.

J Bone Joint Surg Am. 2012;94(19):1756–1763.

Lusty

Tai

Sew-Hoy

Walter

Zicat

Third-generation alumina-on-alumina ceramic bearings in cementless total hip arthroplasty.

J Bone Joint Surg Am. 2007;89(12):2676–2683.

Murphy

Ecker

Tannast

Two- to 9-year clinical results of alumina ceramic-on-ceramic THA.

Clin Orthop Relat Res. 2006;453(453):97–102.

Topolovec

Milošev

A comparative study of four bearing couples of the same acetabular and femoral component: a mean follow-up of 11.5 years.

J Arthroplasty. 2014;29(1):176–180.

10.

Sedrakyan

Normand

Dabic

Jacobs

Graves

Marinac-Dabic

Comparative assessment of implantable hip devices with different bearing surfaces: systematic appraisal of evidence.

BMJ. 2011;343:d7434.

11.

Willmann

Ceramic femoral head retrieval data.

Clin Orthop Relat Res. 2000;379:22–28.

12.

Traina

Tassinari

De Fine

Bordini

Toni

Revision of a ceramic hip for fractured ceramic components: AAOS exhibit selection.

J Bone Joint Surg Am. 2011;93(24):e147.

13.

Allain

Roudot-Thoraval

Delecrin

Anract

Migaud

Goutallier

Revision total hip arthroplasty performed after fracture of a ceramic femoral head.

A multicenter survivorship study. J Bone Joint Surg Am. 2003;85-A(5):825–830.

14.

Sharma

Ranawat

Rasquinha

Weiskopf

Howard

Ranawat

Revision total hip arthroplasty for ceramic head fracture: a long-term follow-up.

J Arthroplasty. 2010;25(3):342–347.

15.

Pulliam

Trousdale

Fracture of a ceramic femoral head after a revision operation.

A case report. J Bone Joint Surg Am. 1997;79(1):118–121.

16.

Traina

De Fine

Di Martino

Faldini

Fracture of ceramic bearing surfaces following total hip replacement: a systematic review.

Biomed Res Int. 2013;2013:157247.

17.

Koo

Kim

Yoon

Min

Kim

Revision of ceramic head fracture after third generation ceramic-on-ceramic total hip arthroplasty.

J Arthroplasty. 2014;29(1):214–218.

18.

Fritsch

Gleitz

Ceramic femoral head fractures in total hip arthroplasty.

Clin Orthop Relat Res. 1996;328:129–136.

19.

Levašič

Milošev

Valdoltra Hip Arthroplasty Register.

http://www.ob-valdoltra.si/sl/international. Accessed January 20, 2016.

20.

DeLee

Charnley

Radiological demarcation of cemented sockets in total hip replacement.

Clin Orthop Relat Res. 1976;(121):20–32.

21.

Gruen

McNeice

Amstutz

“Modes of failure” of cemented stem-type femoral components: a radiographic analysis of loosening.

Clin Orthop Relat Res. 1979;(141):17–27.

22.

Trebše

Milošev

Kovač

Mikek

Pišot

Poor results from the isoelastic total hip replacement: 14-17-year follow-up of 149 cementless prostheses.

Acta Orthop. 2005;76(2): 169–176.

23.

Engh

Bobyn

Glassman

Porous-coated hip replacement.

The factors governing bone ingrowth, stress shielding, and clinical results. J Bone Joint Surg Br. 1987;69(1):45–55.

24.

Engh

Massin

Suthers

Roentgenographic assessment of the biologic fixation of porous-surfaced femoral components.

Clin Orthop Relat Res. 1990;(257):107-128. Erratum in: Clin Orthop Relat Res. 1992;284:310–312.

25.

Zicat

Engh

Gokcen

Patterns of osteolysis around total hip components inserted with and without cement.

J Bone Joint Surg Am. 1995;77(3):432–439.

26.

Morawietz

Classen

Schröder

et al.

Proposal for a histopathological consensus classification of the periprosthetic interface membrane.

J Clin Pathol. 2006;59(6):591–597.

27.

Doorn

Mirra

Campbell

Amstutz

Tissue reaction to metal on metal total hip prostheses.

Clin Orthop Relat Res. 1996;329(Suppl):S187–S205.

28.

Morrell

Danzer

Milošev

Trebše

An assessment of in-vivo failures of alumina ceramic total hip joint replacements.

J Eur Ceram Soc. 2012;32(12):3073–3084.