Secondary Logo

Journal Logo

Horizontal Platform Supported Hip Replacement

Pritchett, James, W*; Townley, Charles, O

Clinical Orthopaedics and Related Research: June 2004 - Volume 423 - Issue - p 123-128
doi: 10.1097/01.blo.0000127585.10230.17

The cementless horizontal platform supported femoral component for hip replacement was designed to provide physiologic stress loading of the proximal femur after hip replacement. The goal is to extend pain-free implant survival by reducing periprosthetic bone resorption and to provide improved hip function and many years of relief from hip pain. A total hip replacement using this prosthesis was done in 334 consecutive nonselective hips in 322 patients. The mean age of the patients at the time of total hip replacement was 69 years (range, 29–89 years). Thirty-two patients (34 hips) were lost to followup and 81 patients (83 hips) died less than 10 years postoperatively. The remaining 209 patients (217 hips) were followed up for a minimum of 10 years (maximum, 14 years). There were six revisions of the femoral prosthesis. Three (1.8%) were revised for loosening, two were revised for femoral fracture, and one was revised for unexplained pain. The rate of stem survival was 0.97 ± 0.02 (mean and standard error). There were 209 (96%) hips with excellent or good results. Seven (3%) patients had osteolytic changes develop. Radiographic evidence of diminished bone density in the proximal femur was seen in 34 hips (16%). The clinical results are comparable with other reported results for cementless hip replacements at 10 years.

From the *Department of Orthopaedics and Sports Medicine, University of Washington Seattle, WA; and the †Townley Orthopedic Clinic, Port Huron, MI.

Received: July 29, 2002

Revised: January 17, 2003; January 16, 2004; February 11, 2004

Accepted: March 5, 2004

Correspondence to: James W. Pritchett, MD, 1101 Madison Medical Tower #400, Seattle, WA 98104. Phone: 206-505-1043; Fax: 206-505-1148; E-mail:

The horizontal platform supported femoral component for THR (DePuy Co, Warsaw, IN) was introduced by Townley in 1976.13 Initially the prosthesis was cemented with great care to place cement around only the proximal 5 cm of the stem to avoid diaphyseal fixation and stress relief of the proximal femur. In 1982, porous coating was applied to the proximal 5 cm and the prosthesis has been used without cement since. The primary objective of this stem design was to reduce periprosthetic bone resorption caused by abnormalities in stress loading, which ultimately may lead to osteolysis, loosening, and clinical failure. Consequently, the goal of this implant design is to simulate as closely as possible the normal transmission of weightbearing forces through the prosthesis to supporting bone.11,12 A clinical goal of reducing thigh pain was addressed by designing and implanting the prosthesis to load the entire cortical arc of the proximal femur rather than the medullary endosteal surface. The design features include a circumferentially expanded and horizontally inclined platform; porous coating that is limited to the undersurface of the platform and a small area of the proximal stem; and a straight stem that is tapered distally. The implant is made of CoCr substrate with sintered CoCr beads and a median pore size of 200 mm. The stem has been available with a shorter stem length and modular femoral heads since 1987 (PSL, Biopro Inc, Port Huron, MI). This implant was used by the first author in 334 consecutive primary arthroplasties in 322 patients. The patients were followed up by the same author for a minimum of 10 years. We describe the clinical and radiographic results of these procedures.

Back to Top | Article Outline


Of the 334 procedures done, 239 (72%) were done because of osteoarthrosis or postraumatic arthrosis, 31 (9%) were done because of ON, 24 (7%) were done because of failed femoral neck fracture fixation, 23 (7%) were done because of hip dysplasia, and 17 (5%) were done because of inflammatory arthritis. The mean age of the patients at the time of THR was 69 years (range, 29–89 years). The mean preoperative weight of the patients was 91 kg (range, 46–168 kg).

The operations were done with the hip exposed through an anterolateral approach. The femoral head was removed at a right angle to the hip resultant forces (assuming single leg stance) preserving a majority of the neck. Hand broaches that compress rather than remove medullary cancellous bone were used. The broaches were inserted to the depth determined by the use of preoperative templates and the correct size that will fill the medullary canal of the proximal femur. With the broach in place, the neck of the femur was planed using the power planer. The prosthetic stem then was inserted and a femoral head of the appropriate neck length was attached and the hip was reduced for a trial reduction. Because much of the femoral neck is preserved, the most commonly used femoral head is of the shortest neck design. Also, the offset of the femoral prosthesis usually is the same as the natural femur for the same reason. The platform of the implant had complete contact with the neck of the femur in 313 patients (94%). Each patient received a two-piece cementless acetabular prosthesis. The PE was at least 8 mm thick and the metal shell was porous coated.

Each patient was prospectively assessed clinically by the first author using the Harris hip score.5 The patients were questioned and examined preoperatively and postoperatively by a physician not involved in the study. The study protocol for human subjects was approved by the institutional review board at the University of Washington. Patients signed consent forms to participate. The annual postoperative evaluations included questions about pain and function, a physical examination, and analysis of radiographs. The seven-zone method of Gruen et al4 was used for assistance in describing the radiographs. Kappa statistics were used to assess interobserver reliability in reading radiographs. Kappa values of 0.8 to 1.0 indicate near-perfect agreement.7

Areas of localized loss of trabecular bone or localized cortical erosions were considered signs of osteolysis. These lesions usually had discrete borders and could be measured. There were small focal areas of radiolucency, which did not progress, seen under the platform in five patients. These were attributed to incomplete seating of the platform. Radiolucent lines around the distal stem (Zones 3 and 4) were seen occasionally, indicating inadvertent shear loading of the stem. Hypertrophy of the cortical bone was seen in Zones 3 and 4 around stems that were loaded distally (Fig 1). Bone loss was assessed by comparison of the bone in Zone 7 on sequential radiographs with time. The stability of the femoral prosthesis was determined by change in position of the implant with time. An implant that had tilted more than 5° or subsided more than 5 mm was considered loose. The reaction of the host bone in the area of the porous coating was examined to determine whether the implant has osseointegrated. If there was circumferential radiolucency around the porous coating, the implant was considered not osseointegrated. The absence of reactive lines and the presence of spot welds were considered signs of osseointegration.

Fig 1.

Fig 1.

Back to Top | Article Outline


Patients Who Were Lost to Followup

Thirty-two patients (34 hips, 10%) were lost to followup less than 10 years postoperatively. The mean age at the time of surgery for these patients was 60 years (range, 32–78 years). The mean duration of followup was 3 years (range, 1–8 years). None of these patients had required a reoperation before they were lost to followup. Among the patients lost to followup, there were 25 patients (27 hips) with excellent results, six patients (six hips) with good results, and one patient (one hip) with a fair result when last evaluated. At these early dates there were no radiographic failures.

Back to Top | Article Outline

Patients Who Died Before Followup

Eighty-one (83 hips, 24%) patients died before followup. Five patients died within the first 6 months after surgery. Two patients died of cardiac events, one patient died of a pulmonary embolus, one patient died of gastrointestinal bleeding aggravated by anticoagulation, and one patient died of a stroke. One of the patients who died of a cardiac event, 4 months postoperatively, was treated for a dislocation with a closed reduction under general anesthesia 3 weeks after surgery. The remainder of the patients died 6 months to 9 years postoperatively from causes unrelated to their hip replacement.

The mean age at the time of the operation for the patients who died was 74 years (range, 59–89 years). The mean interval from the operation until death was 6 years. The mean duration of followup was 5 years (range, 0–9 years). One patient had a fair result because of unexplained pain and one patient had osteolysis in Zone 7. No prostheses in this group were revised but one patient had an ORIF of a periprosthetic fracture. The remaining implants remained stable radiographically. Twenty-two patients, however, were chronically ill and had not been active.

Back to Top | Article Outline


There were 17 reoperations during the 10 years after the index arthroplasty. Five acetabular PE liners were exchanged. Four liners were exchanged for wear and one liner was exchanged during debridement for a late infection. One porous-coated shell was revised for loosening. Three patients had a reoperation for internal fixation of a femoral fracture that did not involve revision of the prosthesis. One patient had excision of heterotopic ossification and one patient had closed reduction of a dislocation. There only was one dislocation in the entire series.

There were six revisions of the femoral prosthesis. Two patients had a femoral fracture that was treated by revision to a long stem version of the primary prosthesis. One of the fractures occurred in the immediate postoperative period and one occurred 7 years later. Three patients had revision of the femoral prosthesis for painful loosening. One was at 1.5 years, one was at 2 years, and one was at 5 years (Fig 2). One revision was done for pain but the implant was not loose and the pain did not improve after the revision. The Kaplan-Meier survival rate with any femoral revision as the end point was 97.2 ± 1.7% at 10 years.

Fig 2.

Fig 2.

For one revision (Fig 2), a fully porous-coated AML implant was used (DePuy Co, Warsaw, IN). The SROM prosthesis (DePuy Co) was used in one revision, and the PSL prosthesis was used in two revisions. Except for the patient in whom the implant was not loose, all patients who had revision surgery achieved excellent or good results.

Back to Top | Article Outline

Other Complications

Three infections occurred that resolved with antibiotic treatment alone. There was one nonfatal pulmonary embolus. There was one temporary sciatic nerve palsy and two fractures of the greater trochanter for which the patients did not require surgery. There were two limb lengthenings of 1 cm that were not apparent to the patients. Two patients with limb lengthening less than 1 cm were not satisfied with the excess length. Two patients had shortening of 1 cm but neither patient complained.

Back to Top | Article Outline

Ten-Year Clinical Results

Two hundred nine patients (217 hips) were questioned and examined every year for at least 10 years postoperatively. The mean age of these patients at the time of the arthroplasty was 68 years (range, 29–82 years).

The average preoperative Harris hip score was 55 points (range, 30–62 points).5 At final followup, the average Harris hip score was 92 points and at 5 years, the average score was 94 points. There were 167 (77%) excellent results, (94 points); 40 good (19%) results, (86 points); five fair (2%) results, (73 points); and five (2%) poor results, (65 points).

When asked specifically about pain, 189 patients (87%) said they had no pain or pain that did not limit their activity. Twenty-six patients (12%) had pain that occasionally limited activity, and two patients (1%) had pain that always or almost always limited their activity. None of these patients rated their pain more severe than it had been preoperatively. The two patients who always had pain said their preoperative pain had not been relieved. In one of these patients, the preoperative radiograph showed only moderate arthritic changes but the MRI scan showed large degenerative cysts on the acetabular side. In the other patient, the preoperative radiograph also showed moderate changes but the bone scan showed intense uptake in the joint.

When specifically asked about limitations caused by their hip, 137 patients (63%) stated they had no limitations. Sixty-three patients (29%) said they had only slight limitations, 15 patients (7%) stated they had moderate limitations, and two patients (1%) said they had severe limitations.

No patient in the current series experienced thigh pain. The average abduction power was 4.5 of 5 when tested with the patients in the side-lying position. Thirteen patients had a decrease in their hip score between 5 and 10 years after their hip replacement. All of these patients were older than 80 years at final followup and had other debilitating health conditions. All patients were asked if they were satisfied with their hip replacement. Five patients were dissatisfied. Two elderly patients thought their ambulatory abilities were not improved (although their pain was improved). Their poor ambulatory skills seemed to be related to poor general balance and weakness rather than to a problem with their hip replacement. Two patients were dissatisfied because of poor pain relief from the surgery. One patient had revision of her prosthesis elsewhere but remained in pain without explanation. The other patient declined further evaluation. One patient did not give a reason.

Back to Top | Article Outline

Ten-Year Radiographic Evaluation

At a minimum of 10 years after implantation, the 211 hip replacements that had not been revised were evaluated radiographically. Osteolysis was evident in the femur on seven radiographs (3%). The lesions all were in Zones 1 and 7. Three lesions were larger than 1.5 cm. In 34 radiographs (16%), there was diminished bone density in Zone 7 at 10 years when compared with preoperative radiographs. The remainder of the radiographs (81%) showed excellent maintenance of bone in Zone 7 (Fig 3). Kappa values for reading the radiographs were 0.9 (unweighted) (p < 0.0001).

Fig 3.

Fig 3.

In 13 radiographs, there was cortical hypertrophy around the distal stem in Zones 3 and 4 (Fig 1). In nine radiographs, there were radiolucent lines around the distal stem with osseointegration seen around the proximal porous-coated area. This reaction is thought to be caused by loading around the distal stem in shear. The femur under the platform was well maintained with time even in the patients with osteoporosis (Fig 4).

Fig 4.

Fig 4.

The platform did not achieve contact medially on the AP radiograph in 15 hips but on the lateral view, anterior and posterior contact was evident. In addition to three hip replacements that were revised for loosening, two hip replacements were radiographically loose but the patients did not have symptoms.

Back to Top | Article Outline


Implantation of a stem-supported prosthesis results in a reduction of the strain measured in the proximal medial femur. The degree of strain reduction depends on the shape of the stem and method of attachment to bone.2 Strain is reduced 50% to 60% for the AML (DePuy Co) prosthesis and 30% to 40% for a cemented prosthesis. Using strain gauge testing, the PSL femoral prosthesis showed only a 10% reduction of strain in the proximal medial femur compared with the matched intact femur.2

Oh and Harris8 measured the strain in the proximomedial femur using identical cemented stems except for the size of the collar. They reported that 30% to 40% of the normal strain could be restored by the use of an extended collar.

In a previous study, a cemented prosthesis with a collar, the uncemented AML prosthesis, and the PSL prosthesis were compared.9 Dual energy xray absorptiometry (DEXA) scans of the surgically treated side were compared with DEXA scans of the side that was not treated surgically after 3 years. They showed bone loss of 43% for the cemented prosthesis, 34% for the AML prosthesis, and 8% for the PSL prosthesis.

For conventionally designed and stabilized femoral components, the implant is supported predominantly on the inner wall of the femur. For cemented and extensively porous-coated cementless designs, the femur is substantially loaded from distal to proximal. This pattern of loading results in stress-related changes in the proximal femur with time.

The major difference between our series and other reported series is the design of the femoral prosthesis. The large, horizontally positioned platform is unique. The platform is perpendicular to the resultant forces brought to bear on the hip, rather than oriented 90° to the rotational axis of the hip like most prostheses with small, medial-only collars.8,9,13

Cementless femoral stems can be described as tapered, anatomic, distal fixation, and proximal fixation (Table 1). Stress shielding in Zones 1 and 7 occurs in 50% or more of patients with tapered, anatomic, and distally fixed implants.1,3 The clinical relevance of stress shielding is not clear. Stress shielding with the stem described in our study occurred in 16% of patients at 10 years.

Table 1

Table 1

The literature results shown in Table 1 serve as controls for patients in this series. The studies listed were comparable in technique of insertion, type of acetabular prosthesis used, and method of followup. The patients with horizontal platform supported stems had the most favorable experience with osteolysis and thigh pain. We chose to compare the PSL femoral prosthesis with other cementless prostheses.1,2,6,10,14 There are numerous reported studies using cemented femoral prostheses showing equivalent results at 10 years.

The PSL prosthesis has been used extensively by Townley as a primary and revision implant. He reported excellent or good results in 94% of primary and 84% of revision cases.13 Important technical factors for inserting a PSL prosthesis include: full circumferential seating of the platform on the prepared supporting femur; precise stem sizing that provides secure press fit; and no distal loading.

The PSL femoral prosthesis evolved gradually starting with the original design for an endoprosthesis in 1951. The initial implants were placed press-fit because cement and porous coating were not available.12 Cement limited to the proximal 5 cm was used with essentially the current design starting in 1976. The horizontal platform prosthesis has been used cementless with porous coating since 1982.

During the developmental stages acetabular prostheses progressed from none (hemiarthroplasty), to metal-on-metal, to metal-on-polyurethane, to metal-on-polyethylene, and to ceramic-on-polyethylene as each technology became available. The PSL prosthesis is a successful implant. It achieved the goals of limiting thigh pain, bone loss, and osteolysis. The clinical results reported here are as good as the results of other reported series. Because of the high degree of patient satisfaction and low rate of revision, this prosthesis is recommended for most patients neeeding hip replacement.

Back to Top | Article Outline


1. Bourne RB, Rorabeck CH, Patterson JJ, et al: Tapered titanium cementless total hip replacements: A 10- to 13-year followup study. Clin Orthop 393:112–120, 2001.
2. Collier J, Kennedy F, Mayor M, et al: The importance of stem geometry, porous coating and collar angle of femoral hip prostheses on the strain distribution in the normal femur. Transactions for the Society for Biomaterials 9:96, 1983. Abstract.
3. Engh Jr CA, Culpepper II WJ, Engh CA: Long-term results of use of the anatomic medullary locking prosthesis in total hip arthroplasty. J Bone Joint Surg 79A:177–184, 1997.
4. Gruen TA, McNeice GM, Amstutz HC: “Modes of failure” of cemented stem-type femoral components: A radiographic analysis of loosening. Clin Orthop 141:17–27, 1979.
5. 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 result evaluation. J Bone Joint Surg. 51A:737–755, 1969.
6. Kawamura H, Dunbar MJ, Murray P, et al: The porous coated anatomic total hip replacement: A ten to fourteen-year follow-up study of a cementless total hip arthroplasty. J Bone Joint Surg. 83A:1333–1338, 2001.
7. Norman G, Streiner D: Biostatistics: The Bare Essentials. St Louis, Mosby 1994.
8. Oh I, Harris WH: Proximal strain distribution in the loaded femur: An in vitro comparison of the distributions in the intact femur and after insertion of different hip-replacement femoral components. J Bone Joint Surg 60A:75–85, 1978.
9. Pritchett JW: Femoral bone loss following hip replacement: A comparative study. Clin Orthop 314:156–161, 1995.
10. Purtill JJ, Rothman RH, Hozack WJ, et al: Total hip arthroplasty using two different cementless tapered stems. Clin Orthop 393:121–127, 2001.
11. Townley C: Physiological stress loading THR: The relevance of Wolff’s law. Orthop Today 10:17–18, 1990.
12. Townley C: Complications in Total Hip Replacement: Etiology, Prevention and the Role of a Ceramic Articulation. In Fishman G, Clare A, Hench L (eds). Bioceramics: Materials and Applications. Westerville, OH, American Ceramic Society Ceramic Transactions l48:23–34, 1995.
13. Townley CO: The platform supported total hip. Clin Orthop 119:279–280, 1976. Abstract.
14. Xenos JS, Callaghan JJ, Heekin RD, et al: The porous-coated anatomic total hip prosthesis, inserted without cement: A propective study with a minimum of ten years of follow-up. J Bone Joint Surg 81A:74–82, 1999.
© 2004 Lippincott Williams & Wilkins, Inc.