The surgeon’s dilemma when faced with bone loss during hip replacement surgery is to try and leave more bone than he finds and risk the complications of bone grafting or use more cement or a bigger prosthesis and postpone and complicate later reconstructions. It is however a fact that good cement or prosthesis build up is better than a bad allograft. Types of allograft include bulk allograft, small fragment allograft and demineralized bone matrix. The author had in recent years done more and more mixed allografts in combination with bone graft substitutes and the present favourite is calcium sulphate pellets. Slooff believes that fresh frozen small fragments are the best, but in South Africa allografts are gamma radiated and although fears existed that gamma radiation could be detrimental to the biological response 2.5MRad dose seems to eliminate risk of infection and keep its biological properties. One of the arguments against bulk allograft is the slow and superficial incorporation and the risk of late collapse. More recently immunological response as a factor in a late failure has come to the fore. Clinical experience of up to 22 year follow up with these various types of bone grafts is discussed and representative cases shown. Where morsellized bone is used in combination with a supporting ring of cages or pressfit cups it is important that 50% of host bone contact with the metal is achieved and allograft filling up the rest as the prosthesis or cage resting on an allograft can easily fail when compression of the allograft occurring during weight bearing. Femoral struct grafts are used, where the concave side is filled with the mixed allograft and makes excellent biological plates when femoral shaft defects or peri-prosthetic fractures are treated, and full incorporation takes place. In conclusion allografts are very useful in hip revision surgery provided certain principles are adhered to.
This is a report of local experience with an elastic acetabular titanium cup, hydroxy-apatite coated. The cup was developed 14 years ago in France, where 4000 THRs are done annually and where more than one series reports a 98.9% survival rate at 10 years. From February 2001 to December 2003, 486 cups were implanted in 466 patients, 222 female and 244 male, in 379 primary and 107 revision procedures. Mostly we used HDP cups cross-linked by the South African Gammalink method but we also used 169 Alumina sandwich-type inserts (Ceramtec, Germany). There were 158 uncemented modular hydroxy-apatite Esop stems and 265 cemented stems. The mean age of patients in whom the uncemented cup and stem combination was used was 47.8 years. The mean age of the other patients was 60.8 years. The uncemented modular stems were used in younger patients because this design is intended to give proximal stress transfer and prevent anterior thigh pain. UCLA activity scores of the younger patients compared favourably with those of patients with other implants. Reoperation was necessary because of loose uncemented femoral stems, broken Alumina cups, recurrent dislocation in the absence of the functional abductor apparatus in revision surgery, and sepsis in two-stage revisions. Because of the ease of insertion and excellent results, in the author’s practice this press-fit cup has all but replaced cemented cups in primary THR. In revision THR, supporting rings are used only where bone stock is inadequate for this type of implant.
Between January 1998 and December 2002, 418 hip revisions were performed. Of these, 45 hips were diagnosed as infected and two-stage revisions were done six weeks apart. These were excluded from the study, leaving 373 revisions for mechanical failure. In 310 cases both components were exchanged, in 59 the acetabular prostheses only and in three the stem only. One permanent resection arthroplasty was done for bone loss. Where necessary bone graft was used liberally. In 83 patients (22%) specimens taken at surgery cultured positive. A first generation cephalosporin was given as systemic prophylaxis. Routine usage of Gentamycin was reinforced by Vancomycin or fucidic acid in the cement and bone grafts. Double lumen irrigation was inserted in only 5% of cases on the basis of operative findings. Gram stains done intraoperatively in suspected cases were non-contributory. Seventy-six percent of cultures were gram positive, with a preponderance of coagulase negative staphylococcus. Twelve percent were gram negative and 12% were mixed cultures of gram positive and gram negative organisms. One methicillin-resistant Staphylococcus aureus and one fungus were identified. As most of these patients were referred from elsewhere, we did not know whether or not Gentamycin had been used in the cement during previous surgery. Implant failure due to low-grade infection was under-diagnosed in this series. Because of the relatively low number of failures caused by infection in this group, we recommend, with some improvements, the revision protocol presented. The expense of more detailed preoperative evaluation should be weighed against the success of the protocol.
We conducted an engineering failure analysis of retrieved acetabular cups. From one centre, 37 properly-marked components were retrieved. The details of the patients were noted. Of the 37 components, 27 were brought to the laboratory for an engineering investigation of the cause of failure. A further 10 components were also taken to the biochemistry laboratory within an hour of retrieval, with tissue removed from the patients. The purpose of the investigation was to determine whether any proteins were deposited inside the cups and to see whether there was wear debris on either the retrieved component or in the tissue surrounding the prosthesis. We used visual inspection, colour-dye penetrant, stereo-microscopes, scanning electron microscopes, and mass-spectometric analysis to examine the cups. Debris was captured using 0.4-um filters. We found mechanical failure in vivo is mainly caused by plastic flow. Fretting is the second most likely cause of failure. Both of these are indicative of localised overheating between the acetabular component and the ceramic femoral head. The most likely cause of the overheating is lack of lubrication. With electrophoresis it became evident that at some time in their in-vivo service these cups had reached temperatures exceeding 60°C. These temperatures were confirmed on a five-poster hip simulator. We suggest that an in-depth study be undertaken to establish the method of in-vivo lubrication and the lubricity of the available lubricant.
The mechanical failure of ultra-high molecular weight polyethylene (UHMWPE) acetabular cups in vivo is due mainly to a combination of excessive plastic flow and fretting. Localised overheating of the bearing surface, due to insufficient lubrication, causes this. The purpose of this study was to determine the amount of creep in UHMWPE under various conditions. Test pieces were cut from a piece of raw material and tested according to ASTM D2990. In the first test, to determine the anisotropic behaviour of the material, test pieces of raw material were cut at various orientations. The material was then tested in the virgin state and the virgin state at different temperatures. It was also gamma sterilised under different conditions, namely 24 kGy in air, 25 kGy in a nitrogen atmosphere and 25 kGy in air, and heat treated at 80°C to get an annealing effect. Further tests were conducted to determine the effect of cross-linking on creep behaviour. These tests were administered at room temperature, at 50°C and at 60°C. The material showed extreme anisotropic behaviour. It was more sensitive to creep in the centre of the bar than on the outside (32%). Maximum creep, however, occurred at a 45°-angle. This is significant if we assume that maximum loading of an acetabular cup occurs at an angle of 70.7°. The difference in creep for the virgin material, measured at room temperature and at 60°C, was 87.3% or 0.716 mm. The variance in creep for the different methods of sterilisation was a maximum of 0.3 mm. Creep for the cross-linked material, however, was markedly less than for the virgin material. There was a decrease of 36% (0.58 mm) in creep at room temperature and almost 83% (0.84 mm) at 60°C. The test results show that the cross-linked material is much more stable. This may explain the good in-vivo service of these products.
Information acquired from retrieved polyethylene ace-tabular components is invaluable to the design engineer in achieving better in-vivo service. The failure criteria used, namely mode-1 to 4 wear, are vague and not user-friendly. Based on an in-depth evaluation of over 100 failed acetabular cups, this study investigated a more precise classification Although all the cups were obtained from one centre, they were not clearly marked, making an accurate assessment of the in-vivo life impossible. This, however, was not the aim of the study. We used visual inspections, magnifying glasses, colour-dye penetrant and stereo-microscopes to examine the cups. The most common defects identified were mechanical damage, cracks in the material, plastic flow, scratches, fretting, flaking and wear particles embedded in base material. These cups provided valuable data for compilation of a proposed set of failure criteria to be used in future. Visible defects should be used as a classification tool in future cup-failure analysis. They are explicit and can be used with confidence.
In high-demand situations, modern thinking and experience in total hip arthroplasty (THA) favours the uncemented press-fit cup over its cemented counterpart. Before its regular use in 1996, a high-demand cemented stem was designed for use as a short revision stem with a press-fit cup, with or without impaction bone grafting, in active people, especially those over 55 years. Conceptually, a collarless double-tapered highly polished design was preferred. The clip-on hollow centraliser was designed for 5-mm subsidence. The valgus stem, with cement superior to the shoulder, limited upward pistoning in the cement sleeve, creating less debris. The stiff upper and flexible distal part resulted in a decreased contribution from shear and an increased contribution from compression in load transfer from prosthesis to cement. Three sizes are available: G1, G2 and G3. A straight type is presently being developed for smaller patients with congenital dysplasia of the hip. The stem, made by (Thornton Heavy Engineering Sheffield, United Kingdom), from Rex 734 stainless steel to ISO-2002 standards, tapers 10 mm to 12 mm (6°). All tolerances are adequate to handle Inox or Ceramic heads. From April 1996 to December 2002, 278 stems were implanted in Dr Weber’s practice. The first 172 hip operations (168 patients) were studied. The mean age was 58.6 years. There were 137 primary hips and 25 revisions. The mean follow-up period was 4.5 years (3 to 7). Three patients died with the prosthesis in situ. Two reoperations were done: one cup was revised for recurrent dislocation and one fracture below the step was successfully plated. Only three cases of subsidence were documented, all of them less than 3 mm. To date there have been no stem revisions. The prosthesis, together with the stainless steel head and cross-linked cup, can be regarded as cost-effective and can be used routinely, as a high-demand prosthesis with press-fit cup, or as a short-revision prosthesis.
Using the EOL+ cup, 25 operations were performed between December 1999 and February 2003. Most of them were salvage procedures for recurrent dislocations following primary and revision hip surgery. The 18 women and seven men (mean age 65 years) had experienced a total of about 50 dislocations and 20 previous revision procedures. One patient had seven recorded dislocations, three had each had three previous revisions, and three cases had each had two previous revisions. The mean follow-up was 22 months. No redislocations have occurred. One patient was revised to another EOL+ cup. This cup presents an alternative salvage solution in problem cases, including those due to poor musculature, which do not respond to conventional solutions.
Because of the preconception that they are doomed to early failure, the use of rotating hinges in total knee arthroplasty (TKA) remains a controversial topic. We share our experience of more than 100 patients undergoing TKA using a rotating hinge. As we had in mind the single purpose of allaying fears of early failure, this was a fairly simple study, using revision as a concrete endpoint. We discuss indications, contra-indications and some examples. With our study showing a 93.6% survivorship rate at a mean of 4.8 years, we conclude that the rotating hinge has a definite place in TKA.
Using the EOL cup, we performed 15 operations between December 1999 and January 2001. Most of them were salvage procedures after recurrent dislocation of total hip arthroplasty and subsequent revision surgery. The six men and nine women (mean age 63 years) had experienced a total of 42 dislocations and 16 previous revision procedures. The mean follow-up was 10 months. No redislocation has occurred. We believe the EOL cup is an alternative salvage solution for problem cases.
Using the EOL cup, 15 operations were performed between December 1999 and January 2001. Most of them were salvage procedures after recurrent dislocations of total hip replacement and their revisions. The six men and nine women (mean age 63 years) had experienced a total of 42 dislocations and 16 previous revision procedures. One patient had seven recorded dislocations, two cases each had three previous revisions, and three cases each had two previous revisions. The mean follow-up was 10 months. No redislocations have occurred. This cup presents an alternative salvage solution for problem cases.
Between 1997 and 2000, internal arthrodiastasis procedures (endo-apparatus), using an internal skeletal distraction device, were performed on 33 young patients who had reached the point of total hip arthroplasty or arthrodesis. The mean age of the 20 males and 13 females was 19 years (range 11 to 51 years). We removed 19 implants, eight after completion of treatment or because they had outlived their usefulness, and 11 because no improvement in the hip disorder had been achieved. Good results were achieved in two thirds of the patients, including patients suffering from avascular necrosis of the femoral head, old Perthes’ disease and contained hip dysplasia with joint space narrowing and pain. Chondrolysis and stiffness of the hip appear to be contraindications for this type of treatment. The three post-traumatic hip disorders were probably also not ideal cases. In young patients, the results of total hip arthroplasty after trauma are poor, and the indications for internal arthrodiastasis should be redefined.
Wear of ultra-high molecular weight polyethylene (UHMWP) acetabular cups is a well-known cause of osteolysis and loosening of the components. Improvement of the wear resistance of UHMWP could extend the clinical life of total hip arthroplasty (THA). Chemical cross-linking in acetylene with gamma radiation is a cheap and effective way of increasing wear resistance of UHMWP. This study is a report on 263 patients (123 males and 140 females) on whom Dr Weber performed THA between 1977 and 1984, using the Pretoria (Grobbelaar) monobloc stainless steel hip with 30-mm metal head. There were 96 patients (107 prostheses) available for follow-up at a mean of 18.3 years, with 89 surviving prosthesis in 79 patients (83.2%). We collected complete sets of radiographs of 54 patients (mean age 71.4 years) for a radiological survey in 1999. In 41 patients (76%) we found no wear. The mean age of these patients was 72 years. Wear was noted in the other 13 patients (24%), whose mean age was 75 years. The mean follow-up time was 16 years (8 to 23). The mean magnification in the radiological study was 18%. Mean wear for the total group was 1.29 mm and mean annual wear 0.17 mm. A similar analysis performed on a group of 64 of Dr Grobbelaar’s patients at 15.5 year mean follow-up shows remarkable similarity, with mean wear of 0.172 mm for a group of 64 patients and annual wear 0.11 mm. Dr Oonishi of Japan, who has conducted the only other long-term follow-up, found similarly promising results.
A total of 138 revision hip arthroplasties in 134 patients, all operated upon by one surgeon, were followed for an average of 7.4 years. The overall results were reasonable, with good to excellent Mayo hip scores in 62% and little or no pain in 86%. These results were independent of body weight, age, primary diagnosis and type of arthroplasty used. The incidence of radiological loosening was alarming, though comparable to that found in other series. Our findings indicate that it may be wise to replace both components even when, after careful assessment before and during operation, one of them seems to be securely fixed.
