The Journal of Arthroplasty 28 (2013) 1160–1166

Contents lists available at SciVerse ScienceDirect

The Journal of Arthroplasty

j ourna l homepage: www.ar throp lasty journa l .o rg

Long-Term Outcomes of a New Model of Anatomical Hydroxyapatite-CoatedHip Prosthesis

Antonio Herrera MD, PhD a,b, Jesús Mateo MD, PhD a,b, Antonio Lobo-Escolar MD, PhD a,b,Juan José Panisello MD, PhD a,b, Elena Ibarz PhD c, Luis Gracia PhD c

a Department of Orthopaedic and Trauma Surgery, Miguel Servet, University Hospital, Zaragoza, Spainb Orthopaedic Surgery, Medicine School. University of Zaragozac Department of Mechanical Engineering, University of Zaragoza, Spain

The Conflict of Interest statement associated withhttp://dx.doi.org/10.1016/j.arth.2012.06.033.

Reprint request: Prof. Antonio Herrera. DepartmenTraumatology, Miguel Servet University Hospital, AvdZaragoza, Spain.

0883-5403/2807-0022$36.00/0 – see front matter © 20http://dx.doi.org/10.1016/j.arth.2012.06.033

a b s t r a c t

a r t i c l e i n f o

Article history:Received 1 June 2012Accepted 13 June 2012

Keywords:anatomical-THAhydroxyapatite coatinglong term outcomesfemur remodelingpolyethylene wear

This prospective study was designed to evaluate 196 Anatomique Benoist Giraud (ABG II) total hiparthroplasties which were implanted between September 1999 and December 2000. A minimum 11 yearsfollow up was completed in 183 cases. The bearing surfaces were polyethylene–zirconia in 84 cases,polyethylene–metal in 42 and ceramic–ceramic in 57. Changes in the femoral stem design, in relation to theprevious ABG I model, have led to a significant improvement in stress-shielding. Polyethylene wear rate waslower by more than 50% compared with non-crosslinked polyethylene. Excellent and good results wereobtained in 90.32% of cases, and implant survival was 98.39% at the end of follow-up.

this article can be found at

t of Orthopaedic Surgery anda Isabel la Católica, 1, 50009-

13 Elsevier Inc. All rights reserved.

© 2013 Elsevier Inc. All rights reserved.

Uncemented hip arthroplasties emerged in the 70's as analternative to cemented prostheses in order to achieve longer survivalrates. The first models had similar results to those obtained withcemented implants. In the late 80's, new designs were improved withthe introduction of coated surfaces, new alloys, hemispheric acetab-ular cups, etc.

In our Department, several types of uncemented hip implantswere used in that period looking for better clinical outcomes andlonger rates of survival. The use of hemispheric acetabular cups andanatomic femoral stems in patients treated in our Hospital started in1990. Both implants were designed for an uncemented press-fitfixation with hydroxyapatite coating (proximal coating in thefemoral stems) (prosthesis Anatomique Benoist Girard (ABG-I)).More than 1600 patients were treated with these arthroplasties from1990 to 1999. High rates of implant survival and excellent clinicaloutcomes were found in the 10 year follow-up [1,2]. Nevertheless,image studies showed high rates of femoral stress-shielding andpolyethylene wear, associated to granulomatous lesions in theproximal femur and acetabulum.

In 1999, significant changes in both components were introducedto the newmodel ABG-II, but the initial design – hemispheric cups and

anatomic femoral stems with hydroxyapatite coating – was main-tained. From September-1999 to December-2000, 196 ABG-II pros-theses were implanted in 168 patients in our Department. The aim ofthis survey was to assess, in a minimum 11 year follow-up (12.25yearsmaximum,mean of 11.3 years), the clinical outcomes in patientstreated with an ABG-II total hip arthroplasty.

Material and Methods

Implants

ABG-II implants are made of a Titanium TMZF alloy, consisting oftitanium molybdenum, zirconium and ferrous (iron). The Young'smodulus (74÷85 GPa) is slightly lower than in the ABG-I (WroughtTitanium (Ti 6Al-4V) alloy, 110 GPa). The hemispheric acetabular cupis provided with 5 holes allowing fixation with spikes or screws forrotational stability and can be sealed with obturator screws if theyremain vacant.

The anatomic ABG-II femoral stem was redesigned reducing boththe anteroposterior diameter and the total length. The “shoulder” ofthe stem is higher for a better contact and fixation with the cancellousmetaphyseal bone. The diaphyseal region of the stem has a smallerdiameter and is highly polished to avoid diaphyseal fixation. It has 7ºof anteversion in the metaphyseal area and 5º of anteversion in thefemoral neck.

The hemispheric acetabular cup and the metaphyseal area of thestem are coated with 99.99% pure hydroxyapatite with high

Fig. 1. Distribution of bearing surfaces in the prostheses.

1161A. Herrera et al. / The Journal of Arthroplasty 28 (2013) 1160–1166

crystallinity (98%–99%). A patented plasma spraying process is usedby the manufacturer (Stryker) for the implants coating. High cross-linked polyethylene (nitrogen irradiation, annealing) or ceramicinserts in the acetabular cups and metallic or ceramic femoral headswere used.

From September 1999 to December 2000, 196 arthroplasties wereimplanted in 168 patients in our Department. By the end of 2011, 156of these patients with 183 total hip prostheses had completed thefollow-up period. A total of 12 individuals (7.14%) were lost for thefollow-up. Nine patients died because of medical morbidities notrelated to the surgery and three changed their address and could notbe located.

For the clinical assessment, the Harris Scale [3] and the EuroQolGroup EQ-5D questionnaire [4] were used, as well as a subjective selfevaluation of the patients. Clinical, functional and radiologic measure-ments were recorded in the preoperative evaluation and during thefollow-up period (1st - 3rdmonth, 1st - 3rd - 5th – 7th – 11/12th years).

Technique and postoperative protocol

All patients received preoperative intravenous antibiotic prophy-laxis with 2nd generation cephalosporins. Low molecular weightheparin was also administered as antithrombotic prophylaxis. Thesurgical approach used for the total hip replacement was posterolateralin all cases. All the polyethylene inserts were placed with a 10º anti-dislocation lip placed in the postero-superior area. The acetabularreaming was performed to obtain subchondral bone in 120 cases(61.22%), cancellous bone in 21 (10.71%) and both in 55 cases (28.07%).In 21 patients, cancellous bone autografting from the femoral head wasadded. All the intraoperative and postoperative complications wererecorded including dislocation, loosening and infection.

All patients in this study were included in the standardpostoperative protocol used in our Department: 48 postoperativehours of bedrest with hip in abduction and external rotation, drainageremoval at 48 h, sitting up in a chair at 48–72 h and walking withpartial weight-bearing 5 days after the intervention.

Anteroposterior and axial radiographs with neutral rotation ofboth hips were taken and the following data were recorded:acetabular angle, rotation centre of the hip, rotation centre heightwith respect to the greater trochanter tip, femoral cortical thickness,cortico-medullary index at three areas (lesser trochanter (LT), 8 and12 mm distal to LT), and osteoporosis severity according to the Singhindex [5].

In the postoperative radiographs the acetabular angle and therotation centre of the prosthetic hip were measured. They wereconsidered to be adequate with a difference of 3 mm or less whencompared to the healthy side.With differences of more than 3mm therotation centre was considered to be high, low, medial or lateraldepending on the direction of displacement. A digital caliper was usedfor the radiographic assessment. A varus or valgus malposition of thefemoral stem was considered to exist when a difference of 4º or morewith the femoral diaphyseal axis was found.

A classification of the stem size in relation to the femoral canal wasmade. When the stem remained at 2.75 to 3.25 mm from the internalfemoral cortex it was classified as “adequate”. With more than 3.25mm it was considered to be a “small” stem andwith less than 2.75mmit was classified as a “big” stem.

To study the remodeling changes that occurred after the prostheticimplantation preoperative X-rays were compared with thoseobtained in the follow-up. The proximal femur and the acetabulumwere respectively divided into the seven zones described by Gruen [6]and the three zones described by De Lee and Charnley [7]. Heterotopicossifications were classified according to Broker [8]. The polyethylenewear was evaluated using the method described by Livermore [9]. Inour Hospital, all image studies are digitalized and the radiographiccomparisons were made using a computerized program. The femoral

stem sinking into the canal was assessed by measuring the distancefrom the greater trochanter tip to the stem based on the initialpostoperative X-rays.

The presence of osteolytic lesions in the proximal femur wasassessed according to the scale proposed by Goetz [10] in theradiographs obtained at the end of the follow-up. A “limited”osteolysis was defined as that involving two or less zones of Gruenwith less than 2.5 cm2. “Mild” osteolysis was that involving from 3 to 5zones of Gruen with 2.5 to 10 cm2 and “severe” osteolysis thatinvolving 6 or more zones of Gruen with more than 10 cm2.

Six experienced surgeons carried out the total hip replacements inall patients included in the study. The clinical and radiologicalassessment was made by the first author.

Statistical Analysis

Statistical analysis was performed using χ2 analysis for categoricaldata and percentages comparison; Student t test was used for thecomparison of means of isolated data or between pairs of related dataas appropriate, with Pearson correlation. The level of significance wasset at P b .05.

Results

From September 1999 to December 2000, 196 total hip ABG-IIarthroplasties were implanted in 168 patients (bilateral replacementin 28 cases). In the unilateral total replacements, 84 were of the rightand 56 of the left hip joint. In 122 cases the etiology was primaryosteoarthritis (72.62%), in 22 (13.09%) rheumatoid arthritis (RA), in 4(2.38%) psoriatic arthritis, in 5 (2.97%) postraumatic osteoarthritisand in 3 (1.80%) femoral neck fractures.

The gender distribution was 119 males (70.83%) and 49 females(29.17%). At the end of the study period (11–12 years), 156 patientswith 183 arthroplasties (bilateral in 27 cases) completed the follow-up. The mean age of the sample was 58.84 years ± 11.26 (SD) (from23 to 77 years).

The bearing surfaces used in the implants were distributedaccording to Fig. 1. All ceramic on ceramic arthroplasties wereimplanted in patients under 60 years of age with good physical andfunctional status.

The sizes of the implanted acetabular cups were distributedaccording to Fig. 2. The prosthetic acetabular cups were uncementedwith press-fit fixation in all cases. Anti-rotation spikes were added in163 cases. In 145 of them 2 spikes of 7 mm and one of 9 mm wereplaced in the 11, 12 and 1 o'clock positions; in 18 cases only 2 spikes of7 mm were used. Additional screws for acetabular fixation werenecessary in 18 patients (two screws in 15 cases and three in the other3 patients). The holes that remained vacant were sealed with

Fig. 2. Distribution of acetabular cup sizes in the prostheses.

Table 1Ranges of Motion Average Values at the Hip Joint, After Arthroplasty.

Movement Maximum amplitude (º)

Flexion 96.2Extension 2.1Abduction 31.3Adduction 16.4Internal rotation 9.3External rotation 25.1

1162 A. Herrera et al. / The Journal of Arthroplasty 28 (2013) 1160–1166

obturator screws. In 2 cases, no spikes or screws were placed foradditional fixation.

The sizes of the femoral stems were distributed according to Fig. 3.All prosthetic femoral heads implanted were 28 mm in diameter,except in 7 cases, where a 22 mm head was used in acetabular cupssized 44 to 46.

Clinical Outcomes

As we have mentioned before, 156 patients with 183 hiparthroplasties (bilateral in 27 cases) completed the follow-up period.Two intraoperative complications occurred during the study period.One was an arterial injury that had to be repaired during the surgeryand the other was a periprosthetic non-displaced metaphysealfemoral fracture, which needed osteosynthesis with cerclages, hencedelaying the normal weight-bearing protocol in this patient. Twopostoperative dislocations were reported. One case at the 3rdpostoperative week, secondary to a forced movement of the hipjoint and solved with conservative treatment after closed reductionwas reported. And one case at the 8th postoperative month, related tovertical malposition of the acetabular cup at 54º, which neededacetabular revision surgery. Another case of instability with sublux-ation, but no true dislocation, was reported in a patient withmalpositioned vertical and retroverted acetabular cup.

The global prevalence of infection (acute and chronic) in oursample was 1.53% (3 implants). Two cases of early postoperativeprosthetic infection were found (1.02 %), with good evolution afterdebridement and intravenous antibiotic treatment. One case ofchronic prosthetic infection was reported in a patient with sepsis

Fig. 3. Distribution of stem sizes in the prostheses.

secondary to pyelonephritis five years after surgery. A two-stageprosthetic exchange to a cemented implant was performed (thepatient was lost at the 7th year of follow-up).

Patients have shown a significant improvement in their clinicalstatus according to the Harris Hip Score, from the reference meanvalue of 32.55 in the preoperative evaluation to 85.80 (26.05-95.82) atthe end of the follow-up. 107 patients (69.03%) reported absence ofpain and 36 (23.22%) had slight occasional pain. 105 (67.74%)individuals showed complete range of motion of the hip joint. 102subjects were able to walk without aids (65.80%) and 110 did notshow any restriction in their daily life activities (DLA). Only 4 patients(2.58%) reported some need for help in DLA. 76 subjects (49.03%) didnot show any clinically relevant leg length discrepancy (defined asmore than 10 mm).

The ranges of motion average values at the hip joint, afterarthroplasty, are shown in Table 1.

The subjective patient self-evaluation of the clinical and functionaloutcomes showed the results included in Table 2.

Of all the individuals in this study with a ceramic on ceramic hipreplacement, 5 of them (8.62%) were found to have a painless hipclunking with some movements of the joint. This squeaking wasperceived as annoying, but no significant differences in the globalclinical outcomes were found in this group of patients.

Radiological Results

Singh score has averaged 4.78 (SD 1.13). Males score higher, 5.20(SD: 0.88), than females, 4.07 (SD: 1.13). Statistically significantdifferences (P b .001) were observed depending on age and sex. Alower Singh score was found in women and older patients.

Postoperative inclination angle of the cup ranged from 36° to 56°,with an average of 43.98°. The centre of rotation of the prosthesis,relative to that of the hip, was considered right in 95 cases (51.91%),high in 26 cases (14.21%), low in 27 cases (14.75%), lateral in 13 cases(7.10%) and medial in 22 cases (12.02%).

Femoral stem alignment was neutral in 116 cases (63.39%), valgusin 39 cases (21.31%) and varus in 28 cases (15.30%).

The stem size was regarded as adequate in 148 cases (80.87%),large in 23 (12.57) and small in 12 (6.56%).

Heterotopic ossifications were detected in 16 cases (8.74%) thelast revision. These were Brooker stage I or II in 11 cases (6.01%),which impaired neither joint mobility nor clinical outcome. Brookerstage III ossifications were found in 3 cases (1.64%), and stage IV in 2cases (1.09%).

In the radiographic study of femoral bone remodeling all stemshave been found to be stable, and no radiolucent lines have been

Table 2Subjective Patient Self-Evaluation.

Appreciation Number Percentage (%)

Excellent 94 60.64Good 46 29.68Unsatisfactory 9 5.80Bad 6 3.88

Fig. 4. X-rays showing densification of cancellous bone in right hip, zones 2 y 6, andbilateral cortical porosity (11 years follow-up).

Fig. 6. X-rays showing small distal pedestal in right hip, stem subsidence in left hip forsmall stem (12 years follow-up).

1163A. Herrera et al. / The Journal of Arthroplasty 28 (2013) 1160–1166

detected. We have seen minimal remodeling changes in 67 hips(36.61%), which showed minor cancellous bone densification mainlyin zones 2 and 6 (Fig. 4), without cortical hypertrophy or osteoporosis.More intense densification of cancellous bone, with cortical hyper-trophy, was observed in 63 hips (34.43%), affecting zones 2, 3, 5 and 6.Cortical osteoporosis in all zones, in addition to densification ofcancellous bone and cortical hypertrophy, was found in 54 cases(29.51%). Cortical osteoporosis is significantly more common inwomen, older people (P b .001) and lower values of Singh score(P b .001). Intense cancellous densification is significantly related tolarge stems (P b .002), but there is no relationship to age, gender orstem alignment.

Fig. 5. X-rays showing devitalization in zones 1 and 7 (12 years follow-up).

A decreased bone density in zones 1 and 7 (Fig. 5), secondary tostress-shielding, has been evident in 77 hips (42.07%). We found asignificant relationship to female sex (P b .001), older age (P b .001)and lower preoperative Singh score (P b .001).

Fig. 7. Subsidence values according stem size: A) initial (1st year): B) final.

Fig. 8. Polyethylene wear values according acetabular cup position.

1164 A. Herrera et al. / The Journal of Arthroplasty 28 (2013) 1160–1166

All cups were stable in the last radiographic evaluation. Neitherradiolucencies nor reactive lines were detected. We found cancellousbone densification in zone I in 39 cases (21.31%), in zone II in 11 cases(6.01%) and in zone III in 12 cases (6.56%). Those cups placed atinclination angles greater than 45° did not show statisticallysignificant differences in cancellous bone densification in zones IIand III.

We also found small distal pedestals in 21 cases (11.47%) (Fig. 6).The average stem subsidence, for stems in an adequate position,

was 0.514 ± 0.150 (SD) mm in the first year. But that value hadincreased slightly, by 0.638 ± 0.180 (SD) mm, in the latestradiographic revision. There are significant differences (P b .001) inthe subsidence measured, both in the first year and in the finalrevision, according to stem size classification: the mean subsidencewas 0.514 mm in adequate size stems, 1.075 mm in small stems(Fig. 6) and 2.435 mm in large stems, in the first year, and 0.638 mm,1.317 mm and 2.830 mm, respectively in the final revision (Fig. 7).

The Livermore method was used to measure polyethylene wear,with the results depicted in Fig. 8. 125 hips, with two types of bearingsurfaces, were studied: 40 were polyethylene–metal and 85 polyeth-ylene–zirconium. Wear was similar in both types of bearing surfaces,ranging from 0.093 mm/year to 0.087 mm/year, with an average wearof 0.090 ± 0.001 (SD) mm/year. Wear was significantly greater (pvalue 0.015) if rotation center of the cup was placed low, with a meanof 0.091 mm/year. We have also found a significantly greater wear (pvalue 0.001) in cups with an inclination angle less than 44º. There hasbeen no statistically significant difference in polyethylene wear incups placed either in high, lateral or medial positions, or in cups withan inclination angle greater than 45º. We detected no wear amongceramic–ceramic bearing surfaces.

Osteolytic lesions were restricted to polyethylene bearing surfaces.On the acetabular side, small osteolytic lesions in zone I have beenobserved in only 12 cases (9.60%), which in no way compromised thestability of the implant. Osteolytic lesions on the femoral side weredetected in 16 cases (12.80%), of which 11 cases were polyethylene–zirconium surfaces and 5 polyethylene–metal type. There was nosignificant difference in incidence between the two groups. Theselesions were rated as limited according to the Goetz scale, and affectedzones 1 and 7 of Gruen. Implant stability was not threatened by theosteolytic lesions, moreover they were asymptomatic.

Discussion

Our results depict an excellent clinical outcome, with animprovement in the Harris Hip Score from 32.55 to 85.80 points. Inaddition, 90.50% of our patients show a high degree of satisfactionwith their outcome.

Implant survival was 98.30% at over 11-years follow-up. Since theABG II implant is HA coated, we don't agree with some reportedresults stating that HA coating is a risk factor for implant loosening inthe long-term [11,12]. Our results are in line with other authors’reports which state that HA coating contributes to implant osseointe-gration and long-term stability [13–16]. Other studies find nodifference between HA coating and titanium plasma-spray or othercoating solutions in the long-term survival of implants [17–19]. In anycase, it seems clear to us that there is no evidence that HA coating is arisk factor for implant loosening. For over 20 years, our experiencewith the HA coated ABG system has been very satisfactory in terms ofosseointegration and implant survival, as is shown in this study, otherstudies by our group and other authors, and multinational studies[1,2,20–23].

Remodeling changes affecting the proximal femoral bone are dueto an altered load transmission pattern induced by the femoral stem.Adaptative remodeling, ruled by Wolf's law [24,25], is a multifactorialprocess in which influence the quality and stiffness of bone, theimplant's design, stiffness and type of fixation, and the forces actingon the hip all have an influence [24,26–28].

Densification of cancellous bone is present in virtually all cases to agreater or lesser extent. Cortical sclerosis in zones 2 and 6, and moredistally, delineates the area where the load is transmitted from thestem to the femur.

HA coating of the metaphyseal portion allows optimum osseointe-gration of the proximal stem, though loads are not transmittedthrough the metaphyseal area but metaphyseal–diaphyseal transi-tion, corresponding to the distal end of the stem coating [29,30].Therefore, remodeling is greater in zones 2 and 6, resulting in highercancellous bone densification, and cortical sclerosis onset in all distalzones of Gruen. This way of load transfer is common in metaphysealsupport designs [31,32]. Our biomechanical studies on ABG I and IIstems, by finite elements simulation, allowed us to confirm theseissues [33–35]. The higher density of cancellous bone and corticalsclerosis we found in oversized stems is also explained by thesemechanical factors, because the area immediately distal to the coatingcomes into contact with both cortical zones 2 and 6, and produces agreater transfer of load at those levels. The cortical porosity thatappears in 29.51% of our cases, associated with cancellous bonedensification and cortical sclerosis, is strongly associated with femalesex, older age and lower rates of Singh score, with age and sex beingthe main factors [32,36,37]. Our results are similar to those publishedby Capello, in a study on adaptive remodeling of a cementless stem 15years after the implantation [32].

42.07% of our cases showed bone devitalization in zones 1 and 7 inlong term radiographic studies. It is caused by the stress-shieldingphenomenon, common to all designs of uncemented stems[27,28,38,39]. Anatomical stems were designed to prevent stress-shielding in the proximal femoral bone. This objective has not beenachieved, as McAuley demonstrated [40]. But the changes in the designof the ABG II stem in relation to the previous model ABG I (decreasingthe volume of the metaphyseal portion of the stem ) have clearlyminimized the stress-shielding phenomenon, based on the incidence ofdevitalization in zones 1 and 7 previously published with the ABG Imodel [1,2,21]. Plain radiography does not allow a precise quantifica-tion of the problem, but studies with the use of DEXA, published by ourgroup, show that the new design induces less bone loss, estimated at33.66% in zone I and 36.2% in zone 7. These figures are clearly lowerthan the bone loss caused by the ABG I model [41]. These comparativestudies have been carried out at 5 years of follow-up, and we believethat these results are conclusive because adaptive remodeling of theproximal femur can be considered practically completed one year aftersurgery [41]. In our opinion, this lower bone atrophy caused by the newmodel of stem (ABG II) is due to a greater preservation of metaphysealcancellous bone, a fact which is essential for the remodeling of theproximal femoral bone, as was stated by Huiskes [24]. In our

1165A. Herrera et al. / The Journal of Arthroplasty 28 (2013) 1160–1166

biomechanical studies, we found no influence of the new metal alloy(with a lower modulus of elasticity) had no influence on theimprovement of stress-shielding [35]. It is clear that preoperativebone density of the proximal femur has an important role, since theproximal bone loss is significantly greater in those patients with lowerinitial bone density. None of these remodeling changes in the femoralbone have threatened implant stability over time.

Remodeling changes on the acetabular side are less pronouncedthan those ones reported in studies by us on ABG I [1,2,22], and donot affect the stability of the implant. The higher density ofcancellous bone occurs in zone I of DeLee, which is the higher loadbearing. These changes have no significant relationship to theinclination angle of the cup.

Stem subsidence has been less than that found in the ABG I model[1,2], and similar to that reported with other hydroxyapatite coateddesigns [42]. The increased subsidence from the first post-operativeyear to the latest revision was minimal in adequately sized stems. Incontrast, subsidence was significantly higher in oversized stems,whichmakes us believe that failure to preserve a thick enough layer ofcancellous bone makes proximal stem osseointegration difficult.Subsidence was lower in small stems; therefore primary stabilitywas not enough, leading to a small, but significant, distal migration.

A great improvement has been found in polyethylene wear ascompared to the results published by us on the ABG I model [1,2].The wear per year of the new polyethylene, Duration, has beendecreased by 54.55%. This finding agrees with published papersusing the same model [43], and studies on other models that usecross-linked polyethylene [44]. The method used to measurepolyethylene wear was Livermore's classic, performed by anindependent observer. Currently our center has digital radiography,so we have used methods of image magnification to facilitate themeasurement, combined with software for an accurate determina-tion of wear. Methods based on 3D CT are currently being applied formeasuring polyethylene wear [45], which theoretically may be morereliable. However, such methods are expensive and involve highdoses of radiation, so we believe they are not suitable for repeatedscans required during follow-up. We have found no difference inwear between the polyethylene–metal and polyethylene–zirconiumbearing surfaces, which is consistent with the results reported byStilling [46].

Osteolytic lesions have only been detected in polyethylenebearing surfaces. The incidence has been low, affecting theacetabulum (zone I) in 9.52% of cases, and the femur (zones 1 and7) in 12.70% of cases. We believe that these results are related to thelimited polyethylene wear and an excellent osseointegration of thestem in the metaphyseal area. The decreased number of cup holes, inrelation to the ABG I model, and the use of spikes or plugs to sealthem, may also have helped to avoid migration of wear debris to thebone area close to the cup.

We have found no wear in ceramic–ceramic bearing surfaces.Although ceramics have been used in younger and more activepatients, we have found no osteolytic lesions either in the acetabulumor in the femur. Our findings are in agreement with those reported byHernigou [47], on first-generation ceramics with a 20 years follow up.He suggests it is impossible to measure, by plain radiographs, theminimumwear that may occur during that period, and the inability todetect minimal granulomatous lesions, so he recommends the use ofCT scans.We don't believe it is suitable in asymptomatic patients withnormal radiographs.

But we did detect occasional squeaking noises in patients withceramic–ceramic bearing surfaces (8.47% of cases). The inclinationangle of the cup was greater or lesser than 45° in all cases. Althoughclinical outcomes in terms of pain, mobility and daily living activitieswere satisfactory, these patients had a sense of discomfort. This is adifficult problem to explain, which is not always related tomalposition of the prosthetic components [48,49].

It is our understanding that the ABG II design has been a majoradvance in the field of cementless total hip arthroplasty, with a highimplant survival after more than 11 years of follow-up. The changes instem design have minimized the stress-shielding phenomenon. Theintroduction of the Duration polyethylene has led to a lowwear rate inthe long-term and thus, a lower incidence of osteolytic lesions.

In addition, long-term clinical results are excellent, despite the factthat the hip arthroplasties included in this study have been implantedby a large group of surgeons.

We agree with Harris [50] that research will improve designs andmaterials in the future, but we believe the ABG II design alreadyprovides a good alternative in cementless hip arthroplasty.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.arth.2012.06.033.

References

1. Canales V, Panisello JJ, Herrera A, et al. “Ten years followup of an anatomichydroxyapatite-coated total hip prosthesis“. International Orthop 2006;30(2):84.

2. Canales V, Panisello JJ, Herrera A, et al. "Extensive osteolysis by polyethyleneparticle migration after ten years follow up of an anatomic hydroxyapatite coatedhip prosthesis". J Arthroplasty 2010;25(7):1115.

3. Harris WH. Traumatic arthritis of the hip after dislocation and acetabular fractures:treatment by mold arthroplasty. An end-result study using a new method of resultevaluation. J Bone Joint Surg 1969;51A(4):737.

4. Available in: http://www.euroqol.org/fileadmin/user_upload/Documenten/PDF/Languages/Sample_UK__English__EQ-5D-3L.pdf.

5. Singh M, Nagrath AR, Maini PS. “Changes in trabecular pattern of the upper end ofthe femur as an index of osteoporosis“. J Bone Joint Surg 1970;52A(3):457.

6. Grüen TA, McNeice GM, Amstutz HC. Modes of failure of cemented stem-typefemoral components: a radiographic analysis of loosening. Clin Orthop 1979;141:17.

7. DeLee JG, Charnley J. Radiological demarcation of cemented sockets in total hipreplacement. Clin Orthop 1976;121:20.

8. Brooker AF, Bowermann JW, Robinson RA, et al. Ectopic ossification following totalhip replacement: incidence and a method of classification. J Bone Joint Surg1973;55A:1629.

9. Livermore J, Ilstrup D, Morrey B. Effect of femoral head size on wear of thepolyethylene acetabular component. J Bone Joint Surg 1990;72A:518.

10. Goetz DD, Smith EJ, Harris WH. The prevalence of femoral osteolysis associatedwith components inserted with or without cement in total hip replacements. Aretrospective matched-pair series. J Bone Joint Surg Am 1994;76:1121.

11. Paulsen A, Pedersen AB, Johnsen SP, et al. Effect of hydroxyapatite coating on risk ofrevision after primary total hip arthroplasty in younger patients. Acta Orthop2007;78(5):622.

12. Lazarinis S, Kärrholm J, Hailer NP. Increased risk of revision of acetabular cupscoated with hydroxyapatite. A Swedish Hip Arthroplasty Register study involving8,043 total hip replacements. Acta Orthop 2010;81(1):53.

13. Capello WN, D'Antonio JA, Jaffe WL, et al. Hydroxyapatite-coated femoralcomponents: 15-year minimum followup. Clin Orthop Relat Res 2006;453:75.

14. Rajaratnam SS, Jack C, Tavakkolizadeh A, et al. Long-term results of a hydroxyap-atite-coated femoral component in total hip replacement: a 15- to 21-year follow-up study. J Bone Joint Surg Br 2008;90(1):27.

15. Epinette JA, Manley MT. Uncemented stems in hip replacement–hydroxyapatite orplain porous: does it matter? Based on a prospective study of HA Omnifit stems at15-years minimum follow-up. Hip Int 2008;18(2):69.

16. Camazzola D, Hammond T, Gandhi R, et al. A randomized trial of hydroxyapatite-coated femoral stems in total hip arthroplasty: a 13-year follow-up. J Arthroplasty2009;24(1):33.

17. Lombardi Jr AV, Berend KR, Mallory TH. Hydroxyapatite-coated titanium porousplasma spray tapered stem: experience at 15 to 18 years. Clin Orthop Relat Res2006;453:81.

18. Yoon KS, Kim HJ, Lee JH, et al. A randomized clinical trial of cementless femoralstems with and without hydroxyapatite/tricalcium-phosphate coating: an 8 to 12-year follow-up study. J Arthroplasty 2007;22(4):504.

19. Lee JM, Lee CW. Comparison of hydroxyapatite-coated and non-hydroxyapatite-coated noncemented total hip arthroplasty in same patients. J Arthroplasty2007;22(7):1019.

20. Tonino A, Romanini L, Rossi P, et al. Hydroxyapatite-coated hip prostheses. Earlyresults from an international study. Clin Orthop 1995;312:211.

21. Nourissat Ch, Adrey J, Berteaux D, et al. ABG Scientific Group: ABG cement-free hipreplacement with hydroxyapatite: clinical results at 10 years. Sydney: SICOTMeeting; 1999.

22. Herrera A, Canales V, Anderson J, et al. Seven to 10 years follow-up of an anatomichip prosthesis. an international study. Clin Orthop 2004;423:129.

23. Baker PN, McMurtry IA, Chuter G, et al. THA with the ABG I prosthesis at 15 yearsexcellent survival with minimal osteolysis. Clin Orthop Relat Res 2010;468:1855.

24. Huiskes R, Weinans H, Dalstra M. Adaptative bone remodeling and biomechanicaldesign considerations for no cemented total hip arthroplasty. Orthopedics1989;12:1255.

1166 A. Herrera et al. / The Journal of Arthroplasty 28 (2013) 1160–1166

25. Bobyn JD, Mortimer ES, Glassman AH, et al. Producing and avoiding stressshielding: laboratory and clinical observations of non-cemented total hiparthroplasty. Clin Orthop Relat Res 1992;274:79.

26. Huiskes R, van Rietbergen B. Preclinical testing of total hip stems: the effects ofcoating placement. Clin Orthop Relat Res 1995;319:64.

27. D'Antonio JA, CapelloWN,ManleyMT. Remodeling of bone around hydroxyapatite-coated femoral stems. J Bone Joint Surg 1996;78A:1226.

28. Sychterz CJ, Claus AM, Engh CA. What we have learned about long-term cementlessfixation from autopsy retrievals. Clin Orthop Relat Res 2002;405:79.

29. D'Antonio JA, Capello WN, Crothers OD, et al. Early clinical experience withhydroxyapatite-coated femoral implants. J Bone Joint Surg Am 1992;74:995.

30. D'Antonio JA, Capello WN, Jaffe WL. Hydroxylapatite-coated hip implants.Multicenter three-year clinical and roentgenographic results. Clin Orthop RelatRes 1992 Dec;285:102.

31. Cornelis JM, Oosterbos CJ, Rahmy AI, et al. High survival rate of hydroxyapatite-coated hip prostheses: 100 consecutive hips followed for 10 years. Acta OrthopScand 2004;75:127.

32. Capello WN, D' Antonio JA, Geesink RG, et al. Late remodeling around a proximallyHA-coated tapered titanium femoral component. Clin Orthop Relat Res 2009;467:155.

33. Herrera A, Panisello JJ, Ibarz E, et al. Long term study of bone remodelling afterfemoral stem : a comparison between Dexa and finite element simulation. JBiomech 2007;40(16):3615.

34. Herrera A, Panisello JJ, Ibarz E, et al. “Comparison between DEXA and Finite Elementstudies in the long term bone remodelling o an anatomical femoral stem”. JBiomech Eng 2009;31(4):041013.

35. Gracia L, Ibarz E, Puértolas S, et al. Study of bone remodeling of two models offemoral cementless stems by means of DEXA and finite elements. Biomed Eng On-line 2010;9:22.

36. McCalden RW, McGeough JA, Barker MB, et al. Age-related changes in tensileproperties of cortical bone: the relative importance of changes in porosity,mineralization, and microstructure. J Bone Joint Surg Am 1993;75:1193.

37. Bousson V, Bergot C, Meunier A, et al. CT of the middiaphyseal femur: cortical bonemineral density and relation to porosity. Radiology 2000;217:179.

38. D'Antonio JA, Capello WN, Manley MT, et al. Hydroxyapatite coated implants: totalhip arthroplasty in the young patient and patients with avascular necrosis. ClinOrthop 1997;344:124.

39. Brodner W, Bitzan P, Lomoschitz F, et al. Changes in bone mineral density in theproximal femur after cementless total hip arthroplasty. a five-year longitudinalstudy. J Bone Joint Surg 2004;86B:20.

40. McAuley JP, Sychterz CJ, Engh Sr CA. Influence of porous coating level on proximalfemoral remodeling. A postmortem analysis. Clin Orthop Relat Res 2000;371:146.

41. Panisello JJ, Canales V, Herrero L, et al. Changes in periprosthetic bone remodellingalter redesigning an anatomic cementless stem. Int Orthopaedic 2009;33(2):373.

42. Callary SA, Campbell DG, Mercer GE, et al. The 6-year migration characteristics of ahydroxyapatite-coated femoral stem: a radiostereometric analysis study. JArthroplasty 2012 [Epub ahead of print].

43. Geerdink CH, Grimm B, Vencken W, et al. Cross-linked compared with historicalpolyethylene in THA. An 8-year clinical study. Clin Orthop Relat Res 2009;467:979.

44. Beksac B, Salas A, Gonzalez Della Valle A, et al. Wear is reduced in THA performedwith highly cross-linked polyethylene. Clin Orthop Relat Res 2009;467:1765.

45. Vandenbussche E, Saffarini M, Hansen U, et al. Measurement of femoral headpenetration in polyethylene using a 3-dimensional CT-scan technique. Acta Orthop2010;81(5):563.

46. Stilling M, Nielsen KA, Soballe K, Rahbek O. Clinical comparison of polyethylenewear with zirconia or cobalt–chromium femoral heads. Clin Orthop Relat Res2009;467:2644.

47. Hernigou P, Zilber S, Filippini P, Poignard A. Ceramic–ceramic bearing decreasesosteolysis. A 20-year study versus ceramic–polyethylene on the contralateral hip.Clin Orthop Relat Res 2009;467:2274.

48. Mesko JW, D'Antonio JA, CapelloWN, et al. Ceramic-on-ceramic hip outcome at a 5-to 10-year interval: has it lived up to its expectations? J Arthroplasty 2011;26(2):172.

49. Cogan A, Nizard R, Sedel L. Occurrence of noise in alumina-on-alumina total hiparthroplasty. A survey on 284 consecutive hips. Orthop Traumatol Surg Res 2011[Epub ahead of print].

50. Harris WH. The first 50 years of total hip arthroplasty: lessons learned. Clin OrthopRelat Res 2009;467(1):28.