Risk of ceramic fracture is the single most important reason why orthopaedic surgeons are reluctant to embrace contemporary alumina-on- alumina total hip arthroplasty. One design modification to address this problem is use of a polyethylene-alumina composite liner. Theoretically, the polyethylene backing may improve the toughness of the alumina bearing couples, leading to reduction of the risk of chipping or actual fracture of the alumina liner during insertion or in vivo use. Although these benefits were demonstrated in one early clinical report(1), our experience revealed a worrisome frequency of ceramic failure with the same implant design.

As described by Benazzo, neck-liner impingement may occur more commonly in Asian patients who frequently squat or sit in a cross-legged position. However, evidence of femoral neck-to-acetabular rim contact has also been recognized as a common occurrence in Western populations following total hip replacement, with impingement seen in more than half (56%) of retrieved polyethylene acetabular liners(2). Under impingement conditions and over an extended period of time, an all polyethylene acetabular liner is more likely to deform rather than fracture (leading to excessive wear particle generation and consequent osteolysis and aseptic loosening of components). In contrast, the same impingement in a hip with an alumina liner may not be clinically evident until sudden fracture occurs. Therefore, there is a possibility of under-estimation of impingement in alumina-on-alumina total hip arthroplasties, and we are concerned that the rates of alumina liner fracture may increase with time in Western populations.

We agree that the proposed mode of failure of polyethylene-alumina composite liner may not sufficiently explain the findings observed in our series, but the failure mechanism proposed by Benazzo is equally speculative. We believe that the high contact pressure generated by repetitive impingements between the femoral neck and acetabular liner predominantly causes rim fractures of the liner rather than a subluxation of the head on the opposite site. As loading conditions are more complex and variable in vivo, it is difficult to demonstrate the exact mechanism of failure, and a prospective study in which a large group of patients with contemporary alumina-on-alumina bearings are evaluated with fluoroscopic observations is needed to better explain the mechanism of failure of alumina-on-alumina total hip arthroplasty in vivo.

Dr. Benazzo suggests that the reduced thickness of the alumina liner does not adversely influence its mechanical resistance, by citing the data came from the researchers involved with the design process. Unfortunately, Hasegawa et al., one of early adopter groups of this type of alumina bearing, present another cautionary report that a modular layered acetabular component incorporating a thin alumina insert (Kyocera, Kyoto, Japan) has poor durability because of unexpected mechanical failures including alumina liner fracture and component dissociation (3). Therefore, we would like to emphasize the fact that even the third- generation and proof-tested alumina ceramics may fracture if used in poorly conceived constructs.

References:

1. Ravasi F, Sansone V. Five-year follow-up with a ceramic sandwich cup in total hip replacement. Arch Orthop Trauma Surg. 2002;122:350-3.

2. Shon WY, Baldini T, Peterson MG, Wright TM, Salvati EA. Impingement in total hip arthroplasty a study of retrieved acetabular components. J Arthroplasty. 2005;20:427-35.

3. Hasegawa M, Sudo A, Uchida A. Alumina ceramic-on-ceramic total hip replacement with a layered acetabular component. J Bone Joint Surg Br. 2006;88:877-82.

To The Editor:

What appears most evident in the paper by Park et al.(1 is the high percentage of fractures of ceramic components (4 liners and 2 heads out of a total of 357 implants, or 1.7%). The same figures were presented in a poster by Park et al. at the 2006 AAOS annual meeting in Chicago(1) and on the same occasion other Korean surgeons presented a similar poster(2) in which 5 cases of sandwich ceramic liner fractures were reported out of 157 implants, giving a 3.5% fracture rate. Summing up the experiences of the two groups of surgeons, we have a percentage of fractures of 2.1%. The sandwich type of acetabular liner considered in these works has been in use since 1994, and to date more than 20,000 liners have been implanted in Europe, Asia, and Oceania. Excluding Korea, to our knowledge 28 fractures have occurred (about 0.14%). In all the cases examined, the cause of the failure was a sub- dislocation of the head, that can be often traced back to mal- positioning of the acetabular cup.

It is therefore plausible, as Park et al. affirm, that the high rate of fracture reported in Korea (15 times greater than reports from elsewhere may be ascribed to the particular posture habits of Asian populations (squatting), due to which, in conditions of hyperflexion, there is an impingement between neck and cup, with consequent subluxations of the head. The fractures occurred with 28 mm couplings, so the range of motion is of limited amplitude. We therefore agree fully with Park when he affirms that the subluxation and neck-cup impingement could be avoided, or at least reduced, by using large diameter (36 mm) ceramic-on-ceramic couplings. There are no known cases of breakage regarding the 36 mm ceramic-on- ceramic coupling (maximum follow-up 6 years on about 5,000 implants), excluding those due to incorrect positioning of the liner in the metal shell. Furthermore, as underlined by Park, it is certainly an advantage if the necks of the prosthetic stems are adequately shaped to increase the ROM. All cases of fracture reported in the article refer to old versions of the prosthetic stems, that nowadays present antero-posteriorly lowered necks to reduce the risk of impingement.

I would also point out that the the paper by Park et al. contains several inaccuracies; I believe it might be of use to clarify some issues.

With regards to the use of ceramic-polyethylene (sandwich) liners, the authors state that the only benefit lies in the reduction of the risk of chipping during introduction of the liner in the cup, and cites, in this connection, Ravasi and Sansone(3). Curiously,Park does not quote himself, despite having written about the same concepts a year before(4). As for the risk of fracture for liners with a 28 mm diameter, it must be borne in mind that the main difference between sandwich liners and ceramic- metalback liners with direct connection lies in the fact that, after fracture, sandwich liners come out of the polyethylene shell (Fig. 4 in Y.S. Park's article), while in those with direct connection the breakage of the rim is invisible to x-rays, often asymptomatic, and can be perceived only from the presence of articular noise. In other words, while with sandwich liners the breakage of the rim is always visible and can therefore be diagnosed, the same is not true for directly fixed liners. Finally, it must be remembered that the presence of the polyethylene shell protects the internal metal-back seat, and a new liner can therefore always be used during revision.

With respect to the way the ceramic liner comes out of the polyethylene shell, the above-mentioned Fig. 4 and the caption to the same in the article by Park describe an action that is not possible. According to the explanation given by Park, the neck of the stem produces the fracture of the ceramic by impingement (Fig. 4-I e 4-II). Subsequently, the neck comes in contact with the ceramic rim opposite it and the thrust of the neck itself (Fig. 4-III) causes the liner to come out of the polyethylene shell on the side where the liner is fractured (Fig. 4-IV). I believe the process that leads to the liner's breaking out of the shell is completely different. With reference to Fig. 4-I, the contact between neck and ceramic does not produce a severe breakage of the liner, instead it causes a subluxation of the head on the opposite side. The high contact pressure that is generated in the contact between the head and the rim of the liner causes some ceramic grains to break off. This mechanism has also been explained by Willmann(5). What is known as an ‘avalanche effect’ then takes place, so the fracture advances ever more rapidly until the head manages to wedge itself between the fractured rim and the polyethylene shell, generating the eccentric motion that allows the liner to come out.

Park et al. believe that in sandwich liners the reduction of the thickness of the ceramic articular core favours the fracture of the rim. The thickness of ceramic liners is, according to many surgeons, a potential cause of fracture, and this problem is particularly felt in Korea. Undoubtedly, the thickness of the ceramic liner has an impact on the mechanical resistance, but it has no influence whatsoever in the cases of fracture of the rim. Although it may seem strange, in the event of a sub-dislocation of the head it has been proven by Lima-Lto with the Finite Elements Method (FEM) that the greater the thickness of the ceramic, the greater the contact stress. In the past, cases of rim fracture were reported even in liners thicker than those mentioned by Y.S. Park. The thickness of the ceramic, on the other hand, might have more importance in axial load conditions, although, to the best of our knowledge and that of CeramTec, no cases of fracture for axial load have ever occurred.

Park et al. report two cases of fracture of the femoral head in the first year of implant. The authors' comment on these failures, questioning the data in the literature and the ‘proof-test’ that is carried out by CeramTec in 100% of the femoral heads during production. It is odd that Park and colleagues did not realise that the fracture of the two femoral heads is a consequence of the failure of the rim of the ceramic liner. In figures E1-A and E1-B published in the ‘Supplementary Material’, it can be observed that the non-fractured head of one of the explants is coarsely abraded and riddled with microfractures. These deteriorations of the head are due to the presence of ceramic particles that have detached themselves from the fractured rim and interposed themselves between the head and the liner. Since these particles are made of the same material and therefore have the same hardness as the head, they have produced on it the abrasions and the formation of cracks. In fact, if one observes Fig. E1-A, it is possible to see that the polar abrasion of the head has a circular profile (these are not traces of the processing of the piece, as the paper affirms), precisely because it is caused during the rotational movement in the area close to the coupling taper hole. The fracture of the heads, therefore, took place because of the presence of cracks. This allows us to understand Fig. 2, as well, where at the centre of the image it is possible to observe the polar fragment which is similar in shape and size to the deteriorated polar area in Fig. E1-A. This fracture mechanism has been known to have occurred with non-sandwich and thicker couplings. With reference again to Fig. 2, we agree with the authors that the curvilinear shape of the fracture of the liner is due to the explosion of the head, but the liner (as can be seen on the opposite side of the curvilinear fracture) was already fractured, and this was the cause of the fracture of the head.

The author(s) of this letter to the editor did not receive payment or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author(s) are affiliated or associated.

References:

1. Park Y.S., Moon Y.W, Lim S.J., Hwang S.K., Choy W.S., Kim Y.S.. Ceramic Failure After Alumina-On-Alumina Total Hip Arthroplasty. Poster Board Number: P041. AAOS Annual Meeting, Chicago, 2006.

2. Ha Y.C., Hwang S.C. Cementless Alumina-On-Alumina Total Hip Arthroplasty Using A Sandwich-Type Acetabular Component. Poster Board Number: P034. AAOS Annual Meeting, Chicago, 2006.

3. Ravasi F., Sansone V. Five-year follow-up with a ceramic sandwich cup in total hip replacement. Arch Orthop Trauma Surg. 2002;122:350-3.

4. Park Y. S., Han K. Y., Fenollosa Gomez J., Benazzo F.. Clinical Results of Sandwich type Ceramic-on-ceramic Couplings in Primary Cementless Total Hip Arthroplasty. 6th Biolox® Symposium, Stuttgart 2001.

5. Willmann G.. Retrieved Ceramic Wear Couple: Unexpected Findings. 6th BIOLOX® Symposium, Stuttgart 2001.

We thank Dr. Reis for his comprehensive considerations and encouraging comments, and his forthright presentation of his experiences of ceramic failure.

We agree that our paper emphasizes the failure of third-generation alumina ceramic bearings and does not comment on the overall favorable results of this bearing design. Indeed, our results for a press-fit metal-backed socket were comparable to those of other studies (1-3), which reported a substantial improvement in socket fixation versus earlier alumina-on-alumina total hip arthroplasties. However, in our series, five acetabular components (1.4%) had to be revised because of aseptic loosening, and all of these occurred in patients older than 65 years of age. This result differs somewhat from those reported by Dr.Ries in his letter and of Ravasi and Sansone (4), who reported excellent 5-year follow-up results in a cohort with an extended age indication for the use of the same cup design.

Recently, one more case of alumina liner fracture has been recorded at one of the four centers that participated in our study, which poses a concern for both patients and surgeons. In our experience with alumina liner fracture, none of the alumina inserts or heads were malpositioned (we have thoroughly re-examined all postoperative radiographs taken from seven failed hips (including a recently fractured one). Moreover, we do not consider that revision surgery after ceramic component fracture is a straightforward or safe procedure, because catastrophic abrasive wear can be caused by undetectable microscopic ceramic debris (5), and there is a potential for fracture of a new ceramic head if implanted on the existing taper(6). Therefore, we carefully follow patients who have undergone revision surgery because of ceramic fracture.

Finally, we agree entirely with Dr. Ries that new ceramics, such as, zirconia-toughened alumina (BIOLOX Delta; CeramTec AG, Plochingen, Germany), which allow the use of a larger femoral head, appear promising. However, we do not believe that the relatively short follow-up of four years in our series sufficiently warrants any claim concerning the clinical success of alumina-on-alumina total hip replacements.

References:

1. Bizot P, Hannouche D, Nizard R, Witvoet J, Sedel L. Hybrid alumina total hip arthroplasty using a press-fit metal-backed socket in patients younger than 55 years. A six- to 11-year evaluation. J Bone Joint Surg Br. 2004;86:190-4.

2. Yoo JJ, Kim YM, Yoon KS, Koo KH, Song WS, Kim HJ. Alumina-on- alumina total hip arthroplasty. A five-year minimum follow-up study. J Bone Joint Surg Am. 2005;87:530-5.

3. D’Antonio J, Capello W, Manley M, Naughton M, Sutton K. Alumina ceramic bearings for total hip arthroplasty: five-year results of a prospective randomized study. Clin Orthop. 2005;436:164?1.

4. Ravasi F, Sansone V. Five-year follow-up with a ceramic sandwich cup in total hip replacement. Arch Orthop Trauma Surg. 2002;122:350-3.

5. M. Hasegawa, A. Sudo, and A. Uchida. Alumina ceramic-on-ceramic total hip replacement with a layered acetabular component. J Bone Joint Surg Br. 2006;88:877-882.

We thank Dr. Satpathy for his interest in our article (1)and for the opportunity to respond.

Dr. Satpathy is correct when he states that lateralization of the hip center increases the body weight lever arm, and thereby increases loads on the hip. It is also true that the alumina inserts used in our study had identical internal dimensions and were of the same thickness (4.8 mm), whereas mating polyethylene shells varied in thickness to fit titanium alloy shells. However, we believe that the center of the prosthetic head is not altered by modifying the thickness of the interposed polyethylene; unlike modular offset (lateralized) liners that translate the center of rotation both laterally and inferiorly (2). In addition, all six patients who sustained ceramic failure did not have measurable limb length discrepancies.

Therefore, we do not believe that unfavorable hip biomechanics, e.g.,lateralization of the hip center or a low femoral offset, were a cause of the ceramic failures observed in our series. Rather, we would like to stress that impingement associated with patient characteristics (e.g., a high range of hip motion in Asian patients) or suboptimal component design (e.g., a thin alumina insert or a circular cross-section stem neck), were probably more important contributors to the risk of fracture of this type of alumina bearing design.

References:

1. Park YS, Hwang SK, Choy WK, Kim YS, Moon YW, Lim SJ. Ceramic failure after total hip arthroplasty with an alumina-on-alumina bearing. J Bone Joint Surg Am. 2006;88:780-787.

2. Berry DJ. Unstable total hip arthroplasty: detailed overview. Instr Course Lect. 2001;50:265-274.

To The Editor:

For the last 7 years, I have been performing precisely the same operation and using the identical implants as recently reported in the Journal by Park et al (1). In 140 operations performed by me, I have had 2 ceramic insert fractures, but no fractures of the ceramic head. In both cases of insert fracture the fault was mine.

The first case occurred 2 months after the operation in an active woman. The finding was of a tiny third body between the insert and the shell. In retrospect the post -operative radiograph, on very careful inspection, showed an incomplete faulty position of the insert which I had missed. The second case occurred after 2 years in an active woman who had recurrent sub- luxations which began immediately after surgery. Clearly my operation had failed to produce a stable joint (with the use of ceramic on ceramic couplings recurrent subluxations are an indication for immediate revision). I have not had any case in which I could blame the implant for the failure. I was. therefore, surprised that the authors found no "surgeon fault" in any of their 6 hips in which the ceramic failed.

Out of the 357 hips evaluated by the authors, 351 are presumably doing very well but this was not stressed by the authors. In any case, their use of the phrase "catastrophic failure" is unjustified. Yes, it is upsetting to have 1.7% breakages in an early follow-up, but all the revisions they did were small operations. None of the stems or cup-shells was loose. All they had to do was renew the coupling, and their revisions are apparently doing well.

I have been doing Charnley arthroplasty for the last 40 years, and the excellent results in elderly patients are well known. However, in comparison, the promising aspect of this ceramic-on-ceramic cementless design in younger active people is that at 5-7 years I find no evidence, whatever, of impending aseptic loosening: no migration of cups and stems and no signs of wear or osteolysis.

I, therefore, hope that the authors have not abandoned ceramic couplings. The new ceramics (Delta) are less brittle than alumina and allow the use of larger head diameters also looks promising.

When the vast majority of hips are doing so well, I do not think that a 1.7% rate of breakages which are easily revised should be regarded as very alarming, particularly when the worry of wear particle osteolysis seems not to exist with the use of ceramic on ceramic couplings. The authors' success rate is 98.3%!! Why change course when you are winning?

A much longer follow-up is needed to find out the long term rate of ceramic failure and, in particular, whether the failure rate will rise with time. Meanwhile, the authors are to be complimented on their splendid results.

Reference:

1. Park YS, Hwang SK, Choy WK, Kim YS, Moon YW, Lim SJ. Ceramic failure after total hip arthroplasty with an alumina-on-alumina bearing. J Bone Joint Surg Am. 2006;88:780-787.

To The Editor,

I read with interest the article by Park et al,"Ceramic Failure after Total Hip Arthroplasty with an Alumina-on-Alumina Bearing. I would like to comment on the causes of ceramic failure in this study.

The ceramic liner used in this study had a constant thickness and diameter but polyethylene thickness changed according to cup diameter. All the failures in this study were in large cups where polyethylene thickness would be large as well. Theoretically, a large polyethylene thickness would lateralize the center of head thereby shortening the abductor lever arm and resulting in abnormally increased loads on the hip. It would be interesting to know the offset of the failed prosthetic hips.

Shortening of leg can be an indicator of low offset. It would also be of interest to know if there were leg length discrepancies in these patients.

Reference:

1. Greenwald AS. Biomechanical factors in THR offset restoration. Presented as an instructional course lecture at the annual meeting of the American Academy of Orthopaedic surgeons;2003 Feb5-9;New Orleans, LA.