The exeter cemented stem provides outstanding long-term fixation and bone load at 15 years follow-up: A systematic review and meta-analysis

Abstract

Purpose

Since its introduction in 1988, the double-tapered polished Exeter cemented stem has been widely adopted in primary total hip arthroplasty (THA). Despite the results coming from the arthroplasty registries have proven great survivorship, the aim of this study was to dig deeper and describe the modes of failure of the Exeter stem at 15 years follow-up while reporting the clinical and radiographic outcomes.

Methods

A search of PubMed, MEDLINE, and Embase was performed using the Preferred Reporting Items for Systematic Review and Meta-Analyses since inception of database to January 2022. A meta-analysis was performed on stem’s failure rates and clinical outcomes using random effects models. Publication bias was assessed with funnel plots.

Results

Overall, ten studies met the inclusion criteria with 2167 hips at mean 14.8 ± 4.1 years follow-up. The meta-effect estimate for revision rate for stem-related reasons was 3.8% (CI 95% 2.1–5.6, p < 0.01). The meta-effect for revision rate for stem aseptic loosening (AL) was 0.22% (CI 95% 0–0.4, p = 0.048) and for periprosthetic fracture was 0.6% (CI95% 0.3–0.9, p < 0.001). The meta effect estimate for Oxford Hip Score (OHS) at final follow-up was 32.4 (moderate; CI 95% 23.2–41.6, p <0.001) with and heterogeneity among the studies of I² 0%. Radiolucent lines were reported in 5.5% of cases, with 1.0% of cases (21 hips) reported to be progressive.

Conclusion

Current evidence suggests that the Exeter cemented stem not only has proven long-term outstanding reliability with a revision rate of 3.8%, but also incredibly low revision rates for AL (0.22%) and periprosthetic fracture (0.6%). It is suitable for a variety of indications, and the consistent radiological appearances indicate durable fixation and load transmission while being associated with a remarkably low stem-related complication rate.

Keywords

aseptic loosening cemented hip arthroplasty exeter stem hip replacement periprosthetic fracture total hip arthroplasty

Introduction

Total Hip Arthroplasty (THA) is the gold standard in case of end-stage hip osteoarthritis.¹ Despite uncemented being the preferred method of fixation,² cemented stems have been reported to have excellent survivorship and clinical outcomes.^3,4 Since its introduction, cementing technique has evolved from hand mixing and manual insertion with finger packing, to the modern generation that includes centrifugation, pulsatile lavage, cement plug with pressurization, stem preheating and centralizers.^5–8

Cemented stems are associated with decreased risk of periprosthetic fractures, better control of the femoral version, and minimization of thigh pain.⁹ Moreover, several limitations described in the past were based on outdated cementation technique that compromised the mechanical strength.^1,10–12

The polished taper-slip Exeter Universal stem (Stryker, Kalamazoo, Michigan) was introduced in 1988 and it has been widely used in THA with outstanding results.¹³ The Exeter V40 stem was then introduced in 2001, but apart from a minor change to the trunnion to the current 5°40′, the design has remained the same, and continues to perform well at medium-to-long term follow-up.¹⁴

The long-term performance of this specific stem has been widely reported by studies and registries,^13,15–25 however, registry reports do not clarify details on the modalities of failure of this stem. Therefore, we sought to evaluate via meta-analysis, the long-term failure rates (15 years) of the Exeter stem focusing on revision for stem related complications, revision for aseptic loosening (AL) of the femoral component, and revision for periprosthetic fracture. In addition, reporting on clinical outcomes, rate of radiological subsidence, and rate of radiolucencies of the cemented Exeter stem.

Materials and methods

Search criteria

This search was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (PRISMA).²⁶ The US National Library of Medicine (PubMed/MEDLINE), EMBASE, and the Cochrane Database of Systematic Reviews were queried for publications utilizing the following keywords: “Exeter femoral stem” OR “cemented femoral stem” OR “cemented Exeter femoral component” OR “Exeter cemented system” AND “total” AND “hip” AND “arthroplasty” AND “primary” since inception of database to January 2022. No limit was set regarding the year of publication. Two authors (F.M. and H.T.) independently conducted all the searches and screened the titles and abstracts to identify relevant studies. Differences were resolved by consulting a third senior reviewer (C.W.J.). Only abstracts that evaluated the clinical outcomes of patients with Exeter femoral stem in primary THA with a follow-up greater than 10 years were reviewed. If the title and abstract of each study contained insufficient information, the full manuscript was reviewed. An additional search was conducted by screening the references list of each selected article, as well as the available grey literature at our institution. When studies were reporting outcomes on the same cohort of patients, the one with the longest follow-up was selected.

Inclusion and exclusion criteria

The inclusion criteria were: (1) clinical trials investigating the long-term (>10 years) outcome of the cemented Exeter femoral stem in primary THA, (2) primary outcomes recorded and assessed including (3) implant survivorship free of component revision, (4) revisions due to stem related complications (5) perioperative rates of mortality and complications, (6) clinical outcomes, and (7) radiographic outcomes.

The exclusion criteria were: (1) non-english language studies, (2) studies reporting on less than 10 hips, (3) biomechanical or preclinical studies, (4) studies exclusively using cementless primary THA or cemented THA other than the Exeter stem or hybrid THA including cemented acetabular component and cementless femoral stem, (5) studies on revision THA, (6) studies without clinical, functional, or radiological outcomes, (7) general review and systematic review, (8) registry-based studies, (9) studies with results including mixed types of femoral fixation without stratification per type of femoral fixation, (10) non full text articles, and (11) studies with less than 10 years mean follow-up.

Data collection

Two independent reviewers (F.M. and H.T.) separately examined all the identified studies and extracted data. During initial review of the data, the following information was collected for each study: title, first author, year of publication, study design, number of patients, patients died and lost at follow-up, age of patients, length of follow-up, indication for index surgery, stem revision rate, complication types, reoperations for any reason, implant loosening, dislocations, deep infections, periprosthetic fractures, patient-reported outcomes, and radiographic outcomes.

The level of evidence in the included studies was determined using the Oxford Centre for Evidence-Based-Medicine Level of Evidence (LoE).²⁷ 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 (MCMS). The methodological quality of each study and the different types of detected bias were assessed independently by each reviewer, and then they were combined synthetically. Finally, a comprehensive analysis of the eligible studies was performed, focusing on specific questions that were relative to the topic. When more than one studies were reporting outcomes on the same cohort of patients at different follow-up times, the one with the longest follow-up was included in the meta-analysis to avoid reporting twice the results of the same cohort.

When reported, the quality of the cementation technique was typically assessed using Barrack’s grading system²⁸: grade A is defined as a white-out with complete filling of the medullary canal, grade B as radiolucent lines (RLLs) covering up to 50% of the cement-bone interface (slight defects), grade C as a defective cement mantle or RLLs covering between 50% and 99% of the cement-bone interface, grade D by having 100% of RLLs at the cement-bone interface or complete absence of cement distally and failure to cover the tip of the stem. RLLs were defined as per Kobayashi et al.²⁹ while osteolysis was described as a progressive non-linear lucency of >2 mm resulting in endosteal erosion of cortical or cancellous bone.³⁰ The position of lucent lines in the femur was assessed using the zones described by Gruen et al.³¹ and extended by Johnston et al.⁴ Subsidence of the femoral stem was measured using the method described by Fowler et al.³²

Statistical analysis

For survivorship, the overall number of stems revised, the number of stems revised for AL, the number of stems revised for periprosthetic fracture, and the number of stable stems revised in the context of an acetabular component revision, were the most reported outcomes and used for meta-analysis. For functional outcomes, the Oxford Hip Score (OHS)³³ was the most reported outcomes and used for meta-analysis.

Due to methodological differences of the included studies, random effects models were used to combine reported outcomes. This decision was made, having considered a fixed-effects model not feasible due to outcomes in which the variation in study effects is assumed to occur from sampling error alone as opposed to between-study differences. I² was reported for each outcome, representing the proportion of variation due to between-studies differences or heterogeneity.

Confidence intervals (CI) at 95% for individual studies are based on normal distribution for clinical scores and binomial distribution for survivorship outcomes. Proportions were combined using a raw proportion model.

Funnel plots were used for each measure meta-analyzed to explore for evidence of publication bias and results were considered statistically significant with an α error of 0.05. All analyses were performed using Jamovi V2.3.9.

Results

The literature search initially yielded 570 relevant citations (Figure 1). After removal of duplicate studies, 392 were subject to application of the predetermined inclusion and exclusion criteria. Following application of these criteria, 28 studies were subject to a full-text screening process. Ultimately, ten studies were included for analysis^13,15–25 (Table 1).

Figure 1.

PRISMA flow diagram outlining the systematic review process.

Table 1.

Baseline characterics of the studies.

Author (year)	Type of study (LoE)	Modified coleman score	No. Of hips [initial]	No. Of hips [final]	Age [yrs] average (range)	Sex m/F	Follow-up [yrs] average (range)
Mahon et al. (2020)	Retrospective (IV)	48	829	664	68 (25–89)	297/448	12.3 (10.0–15.1)
Keeling et al. (2019)	Retrospective (IV)	65	130	130	42 (17–50)	N/A	22.0 (20.0–26.1)
Westerman et al. (2018)	Retrospective (IV)	58	540	395	68 (16–93)	201/305	12.3 (10.1–14.1)
Shmitz et al. (2017)	Retrospective (IV)	58	104	100	31 (16–39)	34/44	13.6 (10.0–20.0)
Peterham et al. (2016)	Retrospective (IV)	65	382	146	66 (17–94)	147/203	22.4 (20–26.6)
Carrington et al. (2009)	Rerospective (IV)	65	325	325	68 (24–87)	133/192	15.7 (14.7–17.3)
Young et al. (2009)	Retrospective (IV)	65	230	121	69 (44–89)	88/142	11.2 (10.0–13.0)
Lewthwaite et al. (2008)	Rerospective (IV)	66	130	123	42	N/A	12.7 (10.0–17.0)
Hook et al. (2006)	Rerospective (IV)	60	142	88	61 (18–84)	26/48	12.7 (10.0–17.0)
Chiu et al. (2005)	Retrospective (IV)	51	112	75	64 (41–85)	24/43	12.8 (10.0–16.5)
Total			2924	2167	Mean 57.9		Mean 14.8

N/A: Not Available; BMI: Body Mass Index; LoE: Level of Evidence; yrs: Years; M: Male; F: Female.

All ten studies were a retrospective cohort design with a level IV LoE. The mean Modified Coleman Score was 61.1 ± 6.4, with scores ranging from 66 to 48, indicating a moderate methodological quality. The mean follow-up was 14.8 ± 4.1 years. A total of 2924 hips were initially included in the studies, while at long-term follow-up 2167 were available (74.1%), in line with the percentage of patients lost to follow-up for long-term studies.^15,21,22,34 Mean age of the patients at time of index surgery was 57.9 years (range, 31–69 years) (Table 1).

Nine studies^{15–19,21–23,25} reported the indication for surgery: primary osteoarthritis (OA) was the most common, followed by avascular necrosis, hip dysplasia, post-traumatic OA, and others (Table 2).

Table 2.

Indications and characteristics exeter stem total hip arthroplasty.

Author (year)	Indication for surgery	Type of implant	Cementation technique
Mahon et al. (2020)	N/A	Exeter V40 (stryker orthopaedics, Mahwah, NJ)	N/A
Keeling et al. (2019)	OA 43 (33.1%), hip displasia 40 (30.7%), IA 24 (18.5%), legg-calve -perthes 6 (4.6%), AVN	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	N/A
Keeling et al. (2019)	4 (3.1%), slipped upper femoral epiphysis 3 (2.3%), septic arthritis 3 (2.3%), other 7 (5.4%)	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	N/A
Westerman et al. (2018)	437 OA, 30 postraumatic, 58 congenital/developmental abnormality, 60 AVN	Exeter V40 (stryker orthopaedics, Mahwah, NJ)	N/A
Shmitz et al. (2017)	Congenital hip dysplasia 30 (28.8%); corticosteroid induced AVN 23 (22.1%); RA 13 (12.5%); idiopathic AVN 5 (4.8%); perthes’ disease 5 (4.8%); post-traumatic arthritis 5 (4.8%); slipped capital femoral epiphysis 4 (3.9%); Morquio’s disease 4 (3.9%); other diagnoses 15 (14.4%)	Exeter Universal (stryker ltd., Newbury, UK)	Third generation
Shmitz et al. (2017)		Exeter Universal (stryker ltd., Newbury, UK)	Simplex bone cement (stryker orthopaedics, Mahwah, NJ) [antibiotic loaded]
Peterham et al. (2016)	Primary (OA) 280 (73.30%), protrusio 25 (6.54%), RA 16 (4.19%), dysplasia 15 (3.92%), AVN 9 (2.36%), post-traumatic OA 8 (2.09%), failed osteotomy 5 (1.31%), fracture of the proximal femur 5 (1.31%), perthes’ disease 4 (1.05%), epiphysiolysis 4 (1.05%), Still’s disease 3 (0.79%), revision of hip fusion 3 (0.79%), septic arthritis 1 (0.26%), other 1 (0.26%)	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	Third generation
Peterham et al. (2016)		Exeter Universal (stryker orthopaedics, Mahwah, NJ)	Simplex RO (stryker orthopaedics, Mahwah, NJ)
Carrington et al. (2009)	Primary OA 245; RA 17; protrusio acetabuli 15; others 48	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	Third generation
Carrington et al. (2009)	Primary OA 245; RA 17; protrusio acetabuli 15; others 48	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	Simplex RO (stryker orthopaedics, Mahwah, NJ)
Young et al. (2009)	Primary OA 195; RA 14; AVN 6; previous fracture 13; ankylosing spondylitis 2	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	Second generation
Young et al. (2009)		Exeter Universal (stryker orthopaedics, Mahwah, NJ)	228 LV CMW 3 cement (CMW laboratories dentsply, exeter, UK); 2 palacos (schering-plough ltd, Welwyn Garden city, UK)
Lewthwaite et al. (2008)	Primary OA 43, dysplasia 40, IA 24, legg-calve´-perthes 6, AVN 4, slipped upper femoral epiphysis 3, septic arthritis 3, other 7	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	Third generation
Hook et al. (2006)	OA 55 (63%) inflammatory arthritis 5 (6%) developmental dysplasia of the hip 18 (20%) avascular necrosis 8 (9%) other 2 (2%)	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	Second generation
Chiu et al. (2005)	The most common diagnosis was avascular necrosis of the femoral head in 60% and the other diagnoses included previous fracture, acetabular dysplasia, ankylosing spondylitis, rheumatoid arthritis, psoriatic arthropathy, and old tuberculosis infection	Exeter Universal (stryker orthopaedics, Mahwah, NJ)	Second generation

N/A: Not Available; LV: Low Viscosity; MV: Medium Viscosity; HV: High Viscosity; SS: Stainless Steel; FH: Femoral head; IA: Inflammatory Arthritis; AVN: AVASCULAR necrosis.

Survivorship

All ten studies reported on stem revisions (146 revised stems of 2167). The meta-effect estimate for overall revision rate at 15 years follow-up was 8.3% (CI 95% 2.7–13.9, p = 0.004), however, when excluding the number of well-fixed stems revised during cup revision to provide better exposure, the meta effect estimate for long-term revision rate for stem-related reasons was 3.4% (CI 95% 1.7–5.2, p < 0.01) (Figure 2). If considering a mean follow-up >20 years (mean, 20.3 years), the meta-effect estimate for the overall revision rate was 18.7% (CI 95% 4.3–33, p = 0.011) (Figure 3).

Figure 2.

Forest plot for overall stem revision rate and revision rate for stem-related reasons.

Figure 3.

Forest plot for >20 years stem revision rate and revision rate for stem aseptic loosening.

All studies reported on the number of stems revised for AL (13 stems of 2167), demonstrating a meta-effect estimate of 0.22% (CI 0–0.4, p = 0.048) with heterogeneity among the studies of I² 9.23% (Figure 3). When considering the number of stems revised for periprosthetic fracture (16 stems of 2167), the meta-effect estimate was 0.6% (CI95% 0.3–0.9, p < 0.001) with heterogeneity among the studies of I² 0% (Figure 4).

Figure 4.

Forest plot for stem revision rate for periprosthetic fracture and Oxford Hip Score.

Clinical outcomes and perioperative morbidity

All studies reported on clinical outcomes, however, the most commonly used clinical scoring system was the OHS in six studies.^{13,15,17–19,21,23,24} The meta effect estimate for OHS at final follow-up was 30.7 (CI 95% 20.8–40.5, p < 0.001) with and heterogeneity among the studies of I² 0% (Figure 4).

Young et al.²⁵ reported that of 215 patients, 2 deaths occurred within the acute postoperative period; 1 pulmonary embolism-related death (PE) 14 days post-op and 1 myocardial infarction-related (MI) death 12 days post-op. Although all studies reported mortality during the period of follow-up, no death was attributed to arthroplasty. No other studies reported on perioperative PEs or cardiovascular (CV) related events (Table 3).

Table 3.

Survivorship and complications of exeter hip stems.

Author (year)	No. of hips	Overall stem revisions	Stem revision for femoral AL/lysis	Stem revision for PJI	Periprosthetic fracture	Stem revision for dislocation	Stem revision for pain/Noise	Revision for broken stem	Stable stem revisions (rate)	CV events no.	DVT	PE	Other complications
Mahon et al. (2020)	664	16	1	6	4 [4 rev, 1 ORIF and subsequent rev]	1	3	0	2	N/A	N/A	N/A	18 cup rev
Keeling et al. (2019)	130	35	3	2	1 [rev]	0	0	1	26 during acetabular revision	N/A	N/A	N/A	29 cup revision
Westerman et al. (2018)	395	32	0	3	10 [6 rev, 2 conservative, 2 ORIF]	1	0	1 [neck]	21	0	0	0	N/A
Shmitz et al. (2017)	100	4	0	3	1 [rev]	0	0	1	0	0	0	0	0
Peterham et al. (2016)	146	37	1	4	3 [rev]	9	0	0	20 CIC during acetabular revision	N/A	N/A	N/A	7 isolated cup revision
Carrington et al. (2009)	325	5	0	4	4 [3 ORIF, 1 rev]	0	0	0	0	N/A	N/A	N/A	16 [13 AL cup, 2 heterotopic bone, 1 heterotopic cement]
Young et al. (2009)	121	3	0	3	0	0	0	0	0	1	0	1	0
Lewthwaite et al. (2008)	123	0	0	0	0	0	0	0	N/A	N/A	N/A	N/A	13 cups revised [9 AL]
Hook et al. (2006)	88	8	2	1	0	0	0	0	5	N/A	N/A	N/A	19 cups
Chiu et al. (2005)	75	6	6	0	0	0	0	0	N/A	N/A	0	0 (0.0)	4 [2 AL cup, 2 dislocation [conservative]]

LLD: Leg Length Discrepancy; DVT: Deep Venous Thrombosis; HO: Heterotopic Ossifications; PJI: Periprosthetic Joint Infection; Rev: Revision; AL: Aseptic Loosening; CIC: Cement-in-Cement; ORIF: Open Reduction Internal Fication; N/A: Not Available; No: Number; CV: Cardio Vascular; PE: Pulmonary Embolism.

Radiographic outcomes

Radiographic outcomes were available at a mean follow-up of 15.2 ± 4.41 years, and focused on implant alignment, stem subsidence, presence and progression of peri-implant RLLs. Five studies (1172 hips) reported Barrack Grade classifications^{17–19,23,25} with 745 hips available for review at the time of follow up reporting grade A, B, C, and D in 406 (36.6%), 224 (19.1%), 87 (7.4%), and 8 (0.68%) hips, respectively. Ten cases (0.85%) were not classified due to poor-quality radiographs, and the remaining cases were not classified due to failure to report grades, lost to follow-up, or death during the follow-up period.

Three of the studies reported details on stem subsidence. Hook et al.¹⁷ reported that no stems had measurable migration at the cement-bone interface. The mean vertical subsidence at the stem-cement interface was 1.52 mm (0–8.3) over 12.7 years. The mean subsidence was 0.12 mm (0–0.65) per year, slowing from 0.5 mm (0–0.9) in the first year to 0.03 mm (0–0.9) per year after 4 years. Chiu et al.¹⁶ reported that stem subsidence within the cement mantle occurred in 7 out of the 75 hips, however, further details were missing. Westerman et al.²³ reported subsidence on 307 hips (out of 506), 83 of them <1 mm, 162 between 1 and 1.9 mm, 44 between 2 and 2.9 mm, 16 between 3 and 3.9 mm, and 2 between 4 and 4.9 mm. Overall, RLLs were visible in 105 (5.5%) of cases, with 21 (1.0%) of cases reported to have progressive lines (Table 4).

Table 4.

Clinical and radiographic outcomes of exeter hips stems.

Author (year)	Clinical outcomes pre	Clinical outcomes post	Radiographic outcomes
Mahon et al. (2020)	HHS 48.8	WOMAC 17.5	N/A
Mahon et al. (2020)	WOMAC 61	WOMAC 17.5	N/A
Keeling et al. (2019)	HHS pain 0 (96)	HHS pain 42 [82 hips]	[85 hips]
	HHS function 0 (96)	HHS function 38 [82 hips]	RLL <2 mm across zones 1–14
		OHS 35 [83 hips]	No radiologically loose stems
			Barrack: 40 A, 51 B, 27 C, 2 D
			Mean stem subsidence 1.6 mm (0.25–4.8)
Westerman et al. (2018)	HHS pain 16.6 (0–44)	[312 hips]	24 RLL ≥1 gruen zone
	HHS function 22.1 (0–42)	HHS pain 38.7 (0–44)	14 localized femoral lysis
	OHS 18.3 (0–42)	HHS function 35.3 (0–47)	Barrack: 203 A, 69 B, 4 C
	OHS 18.3 (0–42)	OHS 38.9 (7–48)	Mean subsidence 1.2 mm (0–4 mm) [n.s. Between Barrack grades]
Shmitz et al. (2017)	HHS 51 (15–77)	[99 hips]	11 progressive RLL, 5 stable RLL
		HHS 85 (36–100)	1 radiographically loose [not revised]
			5 osteolytic areas
			34 subsidence >2 mm
Peterham et al. (2016)	N/A	[101 hips]	[91 hips]
		HHS pain 44 (0–44)	4 RLL
		HHS function 34 (0–47)	10 cortical hypertrophy
		OHS 39.5 (10–48)	2 localized lysis
		OHS 39.5 (10–48)	80 subsidence <1 mm
Carrington et al. (2009)	N/A	[26 hips]	[106 hips]
		HHS 82.1	9 stems RLL <14% of surface of the stem
		OHS 43.5	1 radiographic lysis
		OHS 43.5	Mean subsidence 1.82 (0.4–3.4)
Young et al. (2009)	HHS pain 5.2 (0–40)	HHS pain 38.2 (0–44)	15 RLL (<40% cement-bone interface)
	HHS function 14.7 (10–44)	HHS function 34.1 (8–47)	14 distal cortical hypertrophy
			Barrack: 60 A, 53 B, 8 C
			121 subsidence <3 mm [n.s. Between Barrack grades]
Lewthwaite et al. (2008)	HHS 36 (1–68)	HHS 81 (18–100)	RLL 6 in zone 7, 5 in two proximal zones, 1 in four zones
		OHS 21 (12–47)	Barrack: 40 A, 51 B, 27 C, 2 D
		OHS 21 (12–47)	Mean subsidence 1.29 (0–2.9)
Hook et al. (2006)	N/A	OHS 24 (12–49)	9 RLL at one or more zones (8 progressive)
			Barrack: 63 A, 21 C, 4 D
			Mean subsidence 1.52 (0–8.3) (1 failure not revised yet)
Chiu et al. (2005)	HHS 39.8 (20–62)	HHS 82.3 (42–92)	4 RLL >2 mm + subsidence
			6 RLL <2mmin 3 zones (suspect loosening)
			7 subsidence <2 mm

RLL: Radiolucent Lines; HHS: Harris Hip Score; OHS: Oxford Hip Score; JOA: Japanese Orthopaedic Association Hip Score; N/A: Not Available; WOMAC: Western Ontario and McMaster Universities; SAPS: Self-Administered Patient Satisfaction Scale; UCLA: University of California at Los Angeles; VAS: Visual Analog Scale; RLL: Radiolucent Lines.

Publication bias

The reported funnel plots show the level of publication bias; funnel plots are based on symmetry and when asymmetry is shown the plot is suggestive of publication bias.³⁵ Regarding overall stem revision rate, the funnel plot visually suggests publication bias with smaller studies reporting higher revision rates (p < 0.001). Regarding revision rate due to stem AL, the funnel plot visually suggests modest publication bias with smaller studies reporting higher revision rates (p = 0.003). For revision rate for periprosthetic fracture and for OHS, the funnel plots did not suggest any obvious publication bias (p = 0.369 and p = 0.55). The plots show that for larger studies the error associated with their estimate effect gets smaller, meaning that larger studies are associated with less variation around the true effect while smaller studies associated with greater error may have greater variation (Figure 5).

Figure 5.

Funnel plot for measured outcomes.

Discussion

In this meta-analysis on long-term outcomes of the Exeter cemented stem in primary THA, we confirmed the well-established excellent overall 15 years survivorship. In addition, we quantified the long-term revision rate for stem-related complications, for AL of the femoral component and for periprosthetic fracture, reporting outstanding results. Difficulties with loss to follow-up can be an issue with long-term studies and the way in which the data is reported, including reporting of numbers remaining at risk at the tail, therefore results must be interpreted with caution. The meta-effect for >20 years stem revision rate was 18.7% (CI 95% 4.3–33, p = 0.011), comparing favorably with studies reporting failure rates up to 35–40% at 20 years follow-up.^34,36

Registry reports are extremely important in total joint arthroplasty to evaluate the overall performance of a specific implant design. However, one big limitation is the lack of details regarding the different modes of failure of an implant. If we consider the national registries reporting the highest number of Exeter cemented stems, the overall revision rate at 15 years follow-up of this meta-analysis (8.3%) is in line with the most updated data. The National Joint Registry of England, Wales and Northern Ireland (NJR) reported a revision rate between 2.91 and 7.11% at 15 years.³⁷ Similarly, the AOA reported a revision rate between 5.6 and 9.8% at the same time mark.¹⁴

Nevertheless, when digging deeper into the long-term performance of the Exeter stem, registries come short in providing additional details. Despite an overall revision rate of 8.3%, the 15 years meta effect estimate for revision rate for stem-related reasons was 3.4% (CI 95% 1.7–5.2, p < 0.01). This because when analyzing the reasons of revision, the authors noted that often a well-fixed Exeter stem was revised with the cement-in-cement (CIC) technique during cup revisions,^15–21 highlighting the ease with which this stem can be removed from the cement mantle and exchanged for another one.³⁸ In addition, after thoroughly analyzing implant failures, the meta-effect for revision rate due to stem AL was 0.22% (CI 0–0.4, p = 0.048). This extremely low rate of AL suggests that the polished tapered Exeter stem provides outstanding reliability and durable mode of fixation after primary THA.

To support the exceptional outcomes of the Exeter stem, our results showed that the femoral component was considered directly related with hip dislocation/instability in 0.005% of the cases (11 stems of 2167 hips), and therefore revised. It is worth noticing that this result compares favorably with data reported on the NJR for either overall cemented (patient time incidence rate [PTIR] 0.8), uncemented (PTIR 0.8) or hybrid fixation (PTIR 0.93). This finding supports the advantage of a cemented stem to allow the surgeon to freely decide the femoral anteversion despite the native bony anatomy and therefore to achieve a more accurate combined version.³⁹ Moreover, the meta-effect estimate revision rate for periprosthetic fracture was 0.6% (CI 95% 0.3–0.9, p < 0.001) suggesting favorable loading of the femoral bone comparing favorably with the current overall incidence of periprosthetic femur fracture of approximately 4.1%, with higher rates reported for uncemented and revision THA.^40,41 It has been reported that late periprosthetic fracture usually account for approximately 6% of revisions and is the third most common reason, after AL and infection.^42,43 This can be explained because of the more forceful procedure of trying to achieve a tight interference fit of the uncemented prosthesis compared to cemented devices where a space is left for the cement to act as a grout, thereby reducing the iatrogenic femoral fracture that occur more commonly with cementless prosthesis.

Cementation quality among the studies was reported to be either Barrack type A or B, suggesting high quality of the technique with only 0.7% of the cases being type D. This information relates to the low incidence of AL of the implants. Moreover, radiological review showed that the number of hips with radiolucent lines was small (5.5%), most of them <2 mm and stable, and that progressive ones were reported only in 1% of the cases. The mean subsidence was 1.45 ± 0.27 mm, ranging from 0.00 mm to 8.30 mm, with an average stem subsidence among the studies of <2 mm, in line with what reported by Murray et al.^44,45 with a 10-year migration of 1.91 mm, 1.09 of which occurred between years two and ten. This shows once again the strong stability and great reliability of the Exeter stem, proving that progressive migration of this type of stem is not associated with subsequent increased rate of AL. These radiological findings reinforce the taper-slip theory of functioning of the stem where a small amount of subsidence in the cement mantle is a physiological behavior of this design and is not indicative of loosening and failure.^46–48 Loudon and Older⁴⁹ described femoral component subsidence of more than 2 mm as abnormal. According to our data, only one study²³ reported on cases of subsidence of the femoral component >3 mm (3.6% of that cohort), however, none of them were revised for AL.

On the other hand, the use of cement may relate to disadvantages including a rare but severe risk of cardiovascular complications due to embolism and longer operative (OR) time compared to the use of uncemented prosthesis.⁵⁰ However, the infection rate seemed to not be increased by the longer OR time with a stem revision rate for PJI of 1.3%, in line with the current literature, and only one study reported 2 cases of death in the immediate postoperative period (of 2167 patients, 0.0009%), 1 PE-related and 1 MI-related.²⁵

There were a variety of limitations to this study, including the ones inherent with the approach to conducting a systematic review and meta-analysis. First, due to our strict inclusion and exclusion criteria, only 10 studies were selected for evaluation. Not all studies reported all the same data point and there were studies missing one or more of the variables we used in our analysis. As a result, there was variability in the overall pool of patients and hips that were evaluated for each outcome. This variation of total patients evaluated for each outcome must be considered when interpreting these results. However, achieving a follow-up of more than 20 years is not easy, especially considering as THA is usually undertaken in elderly patients. Second, we were limited by the quality of the original studies, the variability of the preoperative diagnosis, the different cementation techniques, and number of patients analyzed. Third, our methodology did not allow for identification of unpublished literature and is limited by potential publication bias. Fourth, several different outcome scores were used across the included studies to assess overall hip function and the long-term follow-up might have been adversely affected by significant numbers of patients lost and dead during at follow-up. Finally, we deliberately excluded registry-based studies to keep relative homogeneity of number of patients, surgeons performing the operation, and cementation technique.

Conclusion

To sum up, this meta-analysis confirmed the excellent long-term survivorship of the Exeter cemented stem after primary THA. In addition, it showed the principal long-term reasons of failure proving extreme reliability regarding implant fixation with a revision rate for AL of 0.22% and exceptional bone loading of the taper-slip design with a revision rate for periprosthetic fracture of 0.6% at 15 year follow-up. Moreover, this stem was associated with outstanding clinical and radiological outcomes, and it was frequently and easily revised with the CIC technique to provide better surgical exposure in case of cup revision.

Footnotes

Author contributions

FM and MM: Conceptualization. FM and HT: Data curation, formal analysis, investigation. MM, TB, and CJ: Supervision and validation. FM: Writing the original draft. MM, TB, and CJ: review and editing. FM, HT, CJ, TB, and MM: Final approval of the version to be published.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial or material support for the research, authorship, and/or publication of this article.

ORCID iDs

Fabio Mancino

Christopher W Jones

References

Gwynne-Jones

Gray

. Cemented or uncemented acetabular fixation in combination with the exeter universal cemented stem: long-term survival to 18 years. Bone Jt J 2020; 102-B(4): 414–422.

Troelsen

Malchau

Sillesen

, et al. A review of current fixation use and registry outcomes in total hip arthroplasty: the uncemented paradox. Clin Orthop 2013; 471(7): 2052–2059.

Berry

Harmsen

Cabanela

, et al. Twenty-five-year survivorship of two thousand consecutive primary charnley total hip replacements: factors affecting survivorship of acetabular and femoral components. J Bone Jt Surg-Am 2002; 84(2): 171–177.

Callaghan

Bracha

Liu

, et al. Survivorship of a charnley total hip arthroplasty: a concise follow-up, at a minimum of thirty-five years, of previous reports. J Bone Jt Surg 2009; 91(11): 2617–2621.

Trumm

Callaghan

George

, et al. Minimum 20 year follow-up results of revision total hip arthroplasty with improved cementing technique. J Arthroplasty 2014; 29(1): 236–241.

Mulroy

Harris

. The effect of improved cementing techniques on component loosening in total hip replacement. an 11-year radiographic review. J Bone Joint Surg Br 1990; 72-B(5): 757–760.

Halai

Gupta

Gilmour

, et al. The Exeter technique can lead to a lower incidence of leg-length discrepancy after total hip arthroplasty. Bone Jt J 2015; 97-B(2): 154–159.

Vaishya

Chauhan

Vaish

. Bone cement. J Clin Orthop Trauma 2013; 4(4): 157–163.

Cassar-Gheiti

McColgan

Kelly

, et al. Current concepts and outcomes in cemented femoral stem design and cementation techniques: the argument for a new classification system. EFORT Open Rev 2020; 5(4): 241–252.

10.

Scanelli

Reiser

Sloboda

, et al. Cemented femoral component use in hip arthroplasty. J Am Acad Orthop Surg 2019; 27(4): 119–127.

11.

Lindberg-Larsen

Petersen

Jørgensen

, et al. Lundbeck foundation center for fast-track hip and knee arthroplasty collaborating group. Postoperative 30 day complications after cemented/hybrid versus cementless total hip arthroplasty in osteoarthritis patients > 70 years: a multicenter study from the lundbeck foundation centre for fast-track hip and knee replacement database and the danish hip arthroplasty register. Acta Orthop 2020; 91(3): 286–292.

12.

Mudgalkar

Ramesh

. Bone cement implantation syndrome: a rare catastrophe. Anesth Essays Res 2011; 5(2): 240.

13.

Franklin

Robertsson

Gestsson

, et al. Revision and complication rates in 654 exeter total hip replacements, with a maximum follow-up of 20 years. BMC Musculoskelet Disord 2003; 4(1): 6.

14.

No authors listed . Hip, Knee & Shoulder Arthroplasty: Annual Report 2022. Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR).

15.

Carrington

Sierra

Gie

, et al. The exeter universal cemented femoral component at 15 to 17 years: an update on the first 325 hips. J Bone Joint Surg Br 2009; 91-B(6): 730–737.

16.

Chiu

Shen

Cheung

, et al. Primary exeter total hip arthroplasty in patients with small femurs. J Arthroplasty 2005; 20(3): 275–281.

17.

Hook

Moulder

Yates

, et al. The exeter universal stem: a minimum ten-year review from an independent centre. J Bone Joint Surg Br 2006; 88-B(12): 1584–1590.

18.

Keeling

Howell

Kassam

AAM

, et al. Long-term survival of the cemented exeter universal stem in patients 50 years and younger: an update on 130 hips. J Arthroplasty 2020; 35(4): 1042–1047.

19.

Lewthwaite

Squires

Gie

, et al. The exeter^TM universal hip in patients 50 years or younger at 10-17 years’ followup. Clin Orthop 2008; 466(2): 324–331.

20.

Mahon

McCarthy

Sheridan

, et al. Outcomes of the exeter v40 cemented femoral stem at a minimum of ten years in a non-designer centre. Bone Jt Open 2020; 1(12): 743–748.

21.

Petheram

Whitehouse

Kazi

, et al. The exeter universal cemented femoral stem at 20 to 25 years: a report of 382 hips. Bone Jt J 2016; 98-B(11): 1441–1449.

22.

Schmitz

MWJL

Bronsema

de Kam

DCJ

, et al. Results of the cemented exeter femoral component in patients under the age of 40: an update at ten to 20 years’ follow-up. Bone Jt J 2017; 99-B(2): 192–198.

23.

Westerman

Whitehouse

Hubble

MJW

, et al. The exeter v40 cemented femoral component at a minimum 10 year follow-up: the first 540 cases. Bone Jt J 2018; 100-B(8): 1002–1009.

24.

Williams

HDW

Browne

Gie

, et al. The exeter universal cemented femoral component at 8 to 12 years: a study of the first 325 hips. J Bone Joint Surg Br 2002; 84-B(3): 324–334.

25.

Young

Duckett

Dunn

. The use of the cemented exeter universal femoral stem in a district general hospital: a minimum ten-year follow-up. J Bone Joint Surg Br 2009; 91-B(2): 170–175.

26.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021; 372: n71.

27.

OCEBM Levels of Evidence Working Group* . The Oxford levels of evidence 2”. Oxford Centre for Evidence-Based Medicine. https://www.cebm.net/index.aspx?o=5653 (accessed 16 January 2022).

28.

Barrack

Mulroy

Harris

. Improved cementing techniques and femoral component loosening in young patients with hip arthroplasty. A 12 year radiographic review. J Bone Joint Surg Br 1992; 74-B(3): 385–389.

29.

Kobayashi

Donnelly

Scott

, et al. Early radiological observations may predict the long-term survival of femoral hip prosthesis. J Bone Joint Surg Br 1997; 79-B(4): 583–589.

30.

Kwong

Jasty

Mulroy

, et al. The histology of the radiolucent line. J Bone Joint Surg Br 1992; 74-B(1): 67–73.

31.

Dorr

Faugere

Mackel

, et al. Structural and cellular assessment of bone quality of proximal femur. Bone 1993; 14(3): 231–242.

32.

Fowler

Gie

Lee

, et al. Experience with the exeter total hip replacement since 1970. Orthop Clin North Am 1988; 19(3): 477–489.

33.

Wylde

Learmonth

Cavendish

. The Oxford hip score: the patient’s perspective. Health Qual Life Outcomes 2005; 3(1): 66.

34.

Swarup

Leeyu

Chiufen

, et al. Implant survival and patient-reported outcomes after total hip arthroplasty in young patients. J Arthroplasty 2018; 33(9): 2893–2898.

35.

Sterne

JAC

Sutton

Ioannidis

JPA

, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ 2011; 343(1): d4002–d4002.

36.

Sochart

Porter

. The long-term results of charnley low-friction arthroplasty in young patients who have congenital dislocation, degenerative osteoarthrosis, or rheumatoid arthritis. J Bone Joint Surg Am 1997; 79(11): 1599–1617.

37.

No authors listed . National joint registry of England, Wales and northern Ireland (NJR) 19th edition. https://reports.njrcentre.org.uk/Downloads

38.

Malahias

Mancino

Agarwal

, et al. Cement-in-cement technique of the femoral component in aseptic total hip arthroplasty revision: a systematic review of the contemporary literature. J Orthop 2021; 26: 14–22.

39.

Mancino

Jones

Sculco

, et al. Survivorship and clinical outcomes of constrained acetabular liners in primary and revision total hip arthroplasty: a systematic review. J Arthroplasty 2021; 36(8): 3028–3041.

40.

Berry

. Epidemiology. Orthop Clin North Am 1999; 30(2): 183–190.

41.

Palan

Smith

Gregg

, et al. The influence of cemented femoral stem choice on the incidence of revision for periprosthetic fracture after primary total hip arthroplasty: an analysis of national joint registry data. Bone Jt J 2016; 98-B(10): 1347–1354.

42.

Lindahl

Malchau

Herberts

, et al. Periprosthetic femoral fractures. J Arthroplasty 2005; 20(7): 857–865.

43.

Lindahl

Garellick

Regnér

, et al. Three hundred and twenty-one periprosthetic femoral fractures. J Bone Jt Surg 2006; 88(6): 1215–1222.

44.

Murray

Gulati

Gill

. Ten-year RSA-measured migration of the exeter femoral stem. Bone Jt J 2013; 95-B(5): 605–608.

45.

Alfaro-Adrián

Gill

Murray

. Should total hip arthroplasty femoral components be designed to subside? J Arthroplasty 2001; 16(5): 598–606.

46.

Lee

AJC

. The time-dependent properties of polymethylmethacrylate bone cement: the interaction of shape of femoral stems, surface finish and bone cement. In: Learmonth

(ed). Interfaces in Total Hip Arthroplasty [Internet]. London: Springer London, 2000.

47.

Wheeler

JPG

Miles

Clift

. The influence of the time-dependent properties of bone cement on stress in the femoral cement mantle of total hip arthroplasty. J Mater Sci Mater Med 1999; 10(8): 497–501.

48.

Ebramzadeh

Sangiorgio

Longjohn

, et al. Effects of total hip arthroplasty cemented femoral stem surface finish, collar and cement thickness on load transfer to the femur. J Appl Biomater Biomech JABB 2003; 1(1): 76–83.

49.

Loudon

. Subsidence of the femoral component related to long-term outcome of hip replacement. J Bone Joint Surg Br 1989; 71-B(4): 624–628.

50.

Ahn

Man

Park

, et al. Systematic review of cemented and uncemented hemiarthroplasty outcomes for femoral neck fractures. Clin Orthop 2008; 466(10): 2513–2518.