A.Carfagni, R .Giacomi, C.F.De Biase
Department of Orthopaedic and Traumatology
San Carlo di Nancy
Director : Prof. A.Carfagni
Keywords: hip prosthesis; coating; outcome study
The encouraging results of earlier studies on animals carried out with implants covered with hydroxyaepatite demonstrating a better bone ongrowth compared to press-fit porous coated titanium implant explain the growing interest for the clinical use of hydroxyaepatite showed in the last decade (Geesink 1987, 1988, 1989). The first clinical application have confirmed that implants covered with hydroxyaepatite, which allow to reach primary stability, could be a valid alternative to traditional cementless implants (Collis 1984, Goldring 1986, Callaghan 1988, Harris 1989.). The purpose of this study is to evaluate the results obtained by our group fare capo at the European A.G.O.R.A. group (Apatite Group of Orthopaedic Research on Arthroplasties). Supported by computerised system for data processing we have examined 171 cases with two years minimum follow up and maximum 5 years.
From January 1990 to December !994 we have examined 171 patients who had had total hip arthroplasty with the Omnifit stem whose proximal one third is covered with hydroxyaepatite. All surgeries were performed in the same centre by the same surgeon.
The Omnifit-Ha stem (Osteonics) has in its proximal one third 50 micron thick, Ha coating with a porosity less than 3%, a purity of 96% and a cristallinity of 65%. Hydroxyaepatite is plasma sprayed on the titanium substrate. The acetabular component used was the Dual radius Omnifit cup (Osteonics) in 131 cases, the dual geometry Omnifit cup in 21 cases and the Omnifit threared cup in 19 cases. All femoral heads were Co-Cr with a 28 or 32 mm diameter 6,5 mm titanium acetabular screws with a length of 35 mm and 20 mm were used in 31 cases.
Of the 171 patients examined 64% were female and 36% males with an average of 58 years (28 min - 79 max) ; The pathologies considered were : Osteoarthrosis, aseptic necrosis, rheumatoid arthritis, cdh and traumatic fractures. (Tab.1) Implant revisions have been excluded from our study.
Data have been processed with the help of the software O.soft II already used by the European Agora group. Each patient has undergone clinical and radiografical evaluation pre-operatively and post-operatively at 1, 3, 6, 12 months and yearly thereafter using the same clinical and radiological form. Clinical parameters aimed at assessing the pain and functional aspects such as range of motion, ability to climb stairs, distance walked without support, the use of walking aids, limp, ability to tie shoes, presence of deformity with functional disturbance and working ability. Thus, at each examination, we could calculate the Harris and Postel-Merle d'Aubigné scores and observe the variations with respect to previous examination. Pre-operative prophylaxys of patient included antibiotic therapy for three days with intravenous caphalosporin and pre and post-operative prophylaxys with sodic and low lateral in 41 cases and postero-lateral in 130 cases. The post-operative treatment was aimed at the functional rehabilitation of the patient we allowed the partial weight bearing with two canadian canes up to the 30th day after surgery and full bearing weight thereafter. No prevention treatment against heterotopic calcification was done. Post-operative blood collection was used in the 70% of the cases.
An accurate post-operative assessment of the radiographic parameters proposed by Engh for non-cemented implants (Engh 1990, 1994) allowed us to follow the evolution of signs at the bone - HA interface through the years (Epinette 1994). Pre-operative radiographic images were made in the anteroposterior and lateral views at the same focus distance so as to allow a precise and reliable pre-op planning. Post-operatively images were always obtained in the two standard projections at 3, 6, 12 months and yearly thereafter. The stem was divided in the zones proposed by Gruen and the cup in the zones proposed by Charnley ; we than evaluated endosteal bone formation, presence of reactive lines, calcar remodelling, cortical thickening, heterotopic ossification, stability, acetabular bone growth and eventually presence of osteolysis (Gruen 1979, Charnley 1961, 1972). Engh's score was used to quantify the evolution and the importance of the radiographic signs considered in relation to their future behaviour in the long term.
One of the complications frequently encountered was the intraoperative crack of the calcar due both to the excessive research for an optimal pressfit and the massive reaming of the medullary canal. Actually after the initial learning curve we have no longer had such inconvenience. No metallic wires were used after the crack though full weight bearing was postpone two - three weeks later without any clinical consequence for the patient. We have had two death for post-op cardiac complications ; one at 2nd month and the other after 18 months after surgery. We have had no cases of spontaneous hip dislocation but only two traumatic dislocations, one of which due to a fall and the other to a road accident. Only one patient was operated on for revision because of traumatic loosening of the acetabular component.
Clinical results were evaluated using the modified Harris and Postel-Merle d'Aubigné hip score. The table clearly shows that the average four years follow-up of the Harris score is largely satisfying (Tab. 2) . The incidence of thigh pain from the immediate post-op period to one years after surgery has been extremely low (less than 4%) after four years from surgery. Only 6% of the cases showed significant pain which diminished after one year decreasing the daily working activity of the patients.
The good clinical results have been enhanced by the excellent radiographic results achieved with this type of implants. Already by the 3rd month from surgery we have notice a remarkable variation in the femoral bone structure specially an augmentation of density around the zone coated with hydroxyaepatite in particular in the medial part of the femur. From the 6th to the 12th after surgery this bone remodelling was more evident extending proximally and distally contributing to the improvement of stem's stability through time.
It appears from the 3rd to the 6th month post-op and is always present in the Gruen's zone 6A and involves over 90% of the cases within one year post-op. Even in the lateral side of the zone 2A we have noticed endosteal bone ongrowth even though not as remarkable as in zone 6A . (Fig.1) At two years post-op the proximal part of the stem coated with hydroxyaepatite shows the typical signs of bone ongrowth and a good fixation (Fig.2).
Radiolucency in the proximal one third of the stem coated with hydroxyaepatite was completely absent in the patients we examined.
Reactive lines were initially observed around the distal part of the stem extending through time even proximally (Fig.3,Fig.4). After the third year, they gradually disappeared contributing to an augmentation of bone density around the part coated with hydroxyaepatite.
It is a typical reaction seen with this implant which increases as time goes by. In our study we have observed two types : One symmetric and represents an increase in thickness of the femoral cortex in Gruen's zones 2B and 6B after the 1st year post-op extending into zones 3 and 5 after the 3rd and 4th year directly proportional to the endosteal bone ongrowth in the part coated with hydroxyaepatite (Fig.5). The other is due to an error in the surgical technique with localised periosteal reaction due to an incorrect positioning of the prostheses in varus or valgus, though without any symptom felt by the patient.
It is present and can be seen radiographically from the 3rd to the 6th month post-op. It tends to stabilise with time and involves over 60% of the cases studied.
We have never noticed areas of osteolysis in our study. Only in one case we had a progressive calcar remodelling followed by 1 cm of osteolysis near the lesser trocanter (Fig.6). We have never noticed phenomenons of reaction in the bone-hydroxyaepatite interface.
Only one case, a 54 years old woman without any sort of general decease we observed a progressive decalcification of the bone surrounding the entire surface of the prostheses. Medullary necrosis due to an excessive speed during reaming of the femoral canal was supposed to be the cause (Fig.7).
We noticed an incidence of heterotopic calcification of degree grade 1 and according to Brooker's classification in 45% of the cases with asyntomatic patients. 2 cases of graded 3 and 4 have underwent surgical intervention followed by radiotherapy with a single dose of 600 rad.
The first clinical and radiographic results obtained by our group with the Omnifit-HA stem are definitively positive. The stem geometry favours the immediate stability that aims at reaching the best fit and fill, especially in the metaphyseal portion of the femur with transformation of the tortional forces in compression forces thanks to the normalisation of the implant while hydroxyaepatite speeds up the ongrowth of the new bone in the proximal one third. We have never noticed osteolytic areas as described by other authors (Capello et al 1994) due to the release of hydroxyaepatite particles. On the contrary Polietilene debris , in the long run , may determine , as already reported by Geesink (1995) and by J.A. D'Antonio (1995) the formation of small osteolytic areas in the calcar region . The presence of osteolytic areas in the distal part of the stem due to particles debris (Harris 1992) is definitely more frequent in the non-cemented (15%) rather than cemented prostheses (2%). D'antonio et al (1994) noticed the appearance of osteolysis in the distal part of the femoral stem of about 2% supposing that the coating of the proximal one third oh hydroxyaepatite form a sort of barrier to the passage of particles of polietilele similar to pmma. Also D'antonio et al (1995), report good clinical results in 92% of the cases, and recurrent hip pain in 8% of the patients with only 1,1% of the patients suffering from thigh pain. The results of our study are basically as satisfying as these , although with a slightly higher incidence of thigh pain (4%) , an outcome similar to the result reported with cemented prostheses (Charley 1972). The optimal clinical results are confirmed by good Harris score (Gustilo et al 1989, Kim et Kim 1992,1993 ; Martell 1993 ), which rates the outcome as good with a score between 91 to 100 points. Numerous papers report a score inferior to 90 points for porous-coated implants. The result of our study , though in accordance with American authors experience with the Omnifit-HA stem demonstrate an Harris scores close to the maximum score at 2 and 4 years follow-up.. From the radiographic point of view , the ongrowth of the new bone that appears prematurely in the HA-coated proximal one third of the stem follows the formation of reactive lines in the non-coated part of the stem. After the 6th month we noticed a progressive endosteal bone apposition at the level of the proximal one third together with distal reactive lines. In accordance with the other authors (Geesink 1990) distal reactive lines are due to the different modulus of elasticity of implants and bone and are completely asymptomatic for the patient . The third year reactive lines gradually disappeared beginning from the part immediately below the HA--zone with progressive improvement of implant stability. Calcar remodelling can be detected about 6 months after surgery and it increases progressively up to1 year post-op. We have never noticed the presence of osteolytic area at the bone-hydroxyaepatite interface. At the present time we have never had any loosening or subsidence of the femoral stem despite our relatively short follow-up. Bone is responsive especially in the HA-coated proximal one third, and this gives good hope for the stability of the implant throughout time .We have implanted Omnifit-HA stem even in patients of 79 year of age with good bone stock without any loosening or deficiency in the ongrowth of the new bone. In conclusion booth clinical and radiographical results, though in the short term, are definitively encouraging and at least up to the present time they can be overlapped to those obtained with the cemented series. The considerable decrease of thigh pain typically seen with other implants, is optimal for an early functional recovery and for the recovery of the patient's daily activity. Nevertheless we must wait several years to be able to compare the Omnifit-HA prosthesis to the current generation of hip prosthesis with a 10-15 year follow-up.