Abstract
Purpose:
Although previous studies have reported encouraging results of cementless Spotorno (CLS) stem, studies with more than 15 years of follow-up are rare. The objective of this study is to investigate the long-term results of CLS stem and the factors potentially influencing the outcomes.
Methods:
The clinical and radiographic data of 79 hips (64 patients) were reviewed. Clinical outcome was determined using the Japanese Orthopedic Association’s hip scoring system (JOA hip score). Survival rate was assessed by Kaplan–Meier survival analysis. The main end point for survival analysis was revision of stem. The correlations between patient demographics, radiographic factors, and stem survival rates were analyzed.
Results:
At a mean follow-up period of 20.1 years, the mean JOA hip score at final follow-up was 84.7 points. Stem survival rate for all revisions was 97.5% at 20 years, and stem survival for aseptic loosening was 98.9%. Varus alignment had a significant negative influence on the survival of the femoral stem.
Conclusion:
This study demonstrates acceptable long-term clinical and radiographic results of the CLS stem in Japanese patients. Caution should be exercised to avoid varus stem alignment.
Introduction
Although encouraging results of cementless Spotorno (CLS) stem, initially established by Prof. Lorenzo Spotorno in 1984, have been reported in previous literature, 1,2 studies on this stem with more than 15 years of follow-up have been rarely reported. We believe it is necessary to evaluate the performance of this cementless stem at more than 15 years after the initial implantation. Furthermore, there is evidence that the proximal femoral morphology and bone mineral content of Asians and Westerners are different; 3 thus, it is essential to investigate the long-term outcome of the CLS stem in Asian patients and assess factors potentially influencing the outcomes.
We present the results of a cohort of Japanese patients at a mean follow-up of 20.1 years. The two objectives of this study are to assess the long-term clinical and radiographic results of the CLS stem and to identify whether patient demographics and radiographic factors influence the stem survival rate.
Methods
Patients
From 1990 to 1997, a consecutive series of 123 patients (142 hips) underwent primary total hip arthroplasty (THA) using the CLS stem (Zimmer Inc., Warsaw, IN, USA). Patient demographics and diagnoses are reported in Table 1. As 24 patients (25 hips) who died and 27 patients (30 hips) who lost to follow-up were excluded, 72 patients (87 hips) were available for review at a mean follow-up of 20.1 years (range, 15.0–24.9 years). Patient status at the latest follow-up is shown in Figure 1. Nine patients (11 hips) had undergone previous osteotomies on the femoral side, and 11 patients (14 hips) had undergone previous osteotomies on the acetabular side. This consecutive series of surgeries were performed by two surgeons of the same department, and the follow-up data were directly obtained by two surgeons who were not involved with the operation.
Patient demographics and diagnoses.
BMI: body mass index; y: years; no.: number.
Patient status at latest follow-up.
Before 1997, only one version of the CLS stem with a neck–shaft angle of 145° was available. Bipolar acetabular components (Zimmer Inc., Warsaw, IN, USA) were implanted in 34 hips for patients with osteonecrosis and neck fracture, and Saint Nabor total hip acetabular components (Zimmer Inc., Warsaw, IN, USA) were implanted in 108 hips for patients with osteonecrosis (5 hips), osteoarthritis, and rheumatoid arthritis.
Operations were performed via the anterior approach (52 hips), lateral approach (48 hips), and posterior approach (42 hips). Bone graft was performed on the acetabular side in 42 hips to obtain better coverage of the acetabular components. Full weight bearing was encouraged within 6 weeks for patients after THA, and 3 weeks for patients after bipolar hip arthroplasty. This study was approved by the institutional review board of our hospital.
Clinical evaluation
As the Japanese Orthopedic Association’s hip scoring system (JOA hip score) 4 was used to evaluate the preoperative hip function, from 1990 to 1997 in our institution, we used the same scoring system at the latest follow-up, and the clinical results were classified as excellent (90–100), good (80–89), fair (70–79), or poor (<70). Similar to the Harris hip score, 5 the JOA hip score includes analysis of pain (0–40), range of motion (0–20), ability to walk (0–20), and daily activities (0–20). Since thigh pain was included in the JOA hip score, we did not evaluate thigh pain separately. Complications such as dislocation, intraoperative fracture, and infection were reviewed from the medical records.
Radiographic evaluation
The standard anteroposterior and lateral hip radiographs at the latest follow-up were compared with immediate postoperative images, and the radiographs in the intervening period were also reviewed. Parameters including stem loosening, 6,7 subsidence, 6 radiolucent lines, 8 stress shielding, 7 pedestal formation, 9 spot-welds, 8 cortical hypertrophy, 10 osteolysis, 8 heterotopic ossification, 11 malalignment, 12 canal fill index (CFI), 13 and canal flare index 14 were separately examined by two surgeons.
Radiolucent lines were defined as low-density areas more than 2 mm thick between the stem and diaphyseal bone. The distribution of radiolucent lines and osteolysis was described according the method of Gruen et al. 8 Subsidence was defined as the downward migration of stem more than 5 mm along the axis compared with the stem position in the immediate postoperative radiograph. The position of stem was measured from the tip of the greater trochanter to the stem shoulder. 6 Stress shielding was classified into four degrees according to the criteria of Engh et al., reflecting the periprosthetic bone density. 7 Pedestal formation was defined as a shelf or endosteal new bone at the tip of the stem that partially or completely bridged the intramedullary canal. 9 Spot-welds were defined as the focal bone formation adjacent to femoral stem and were evaluated according to the method of Gruen et al. 8 Cortical hypertrophy was defined as thickening of the distal periprosthetic diaphyseal bone >2 mm on the most recent anteroposterior radiograph as compared with the first postoperative radiograph. 10 Heterotopic ossification was classified into four stages according to the criteria of Brooker et al. 11 Varus and valgus malalignment of the stem 12 was defined as deviation of the longitudinal femoral axis by more than 2°. CFI was measured on the first postoperative radiograph. If the CFI is ≤80%, the stem was defined as an undersized stem. 13 The canal flare index was the ratio of the intracortical width of the femur at a point 20 mm proximal to the lesser trochanter and at the canal isthmus. 14 According to the canal flare index, the canal shapes were classified as normal (canal flare index ≥3.0–<4.7), stovepipe (canal flare index <3.0), and champagne flute (canal flare index ≥4.7–<6.5). 14,15
Radiographic loosening of the femoral stem was identified if any of the following symptoms was presented: (1) radiolucent lines more than 2 mm in thickness were detected around the entire implant on the anteroposterior radiograph; 6 (2) serial radiographs showed a progressive stem subsidence of >5 mm; 6 (3) position of the stem progressively changed with varus or valgus tilt of >5°. 7,16
Statistical analysis
Kaplan–Meier survivorship analysis with 95% confidence intervals (CIs) was performed to evaluate the survival rates for stem revision for aseptic loosening and stem revision for any other reason as end points. Cox regression analysis, χ 2 test, and Fisher exact test were used to analyze the relationship between demographic factors, radiographic factors, and the survival rate. A difference of p < 0.05 was considered significant. Statistical analyses were performed using SPSS Statistics 21 (SPSS Inc., Chicago, IL, USA).
Results
The 20-year Kaplan–Meier survival rates were 98.9% (95% CI 96.7–100) with “stem revision for aseptic loosening” as the end point, 97.5% (95% CI 94.2–100) with “stem revision for all reasons” as the end point, and 92.6% (95% CI 86.9–98.3) with “any component revision for any reason” as the end point (Figure 2).
Kaplan–Meier survivorship curve with (a) “stem revision for aseptic loosening” as end point, (b) “stem revision for any reason” as end point, and (c) “any component revision for any reason” as end point.
Revision was performed in eight hips (eight patients). One hip underwent revision surgery for aseptic loosening of the femoral component after 9.4 years. Two hips underwent revision surgery for periprosthetic fracture of the femoral components after 16.4 and 20.8 years, respectively. One hip underwent revision surgery for deep infection after 24.1 years; this patient was interviewed via telephone. Four hips underwent revision surgery for aseptic loosening of acetabular components without stem revision.
In addition to revisions, three intraoperative longitudinal fractures of the proximal femur occurred when stems were impacted into the medullary canal during the index surgery. Postoperative complications included one greater trochanter fracture due to trauma that was treated by internal fixation. One dislocation of the bipolar hip arthroplasty was treated by closed reduction followed by bed rest with a brace for 3 weeks.
Clinical results
The mean JOA hip score increased from a preoperative value of 47.7 (range, 17–82) to 84.7 (range, 57–99) at the latest follow-up. Twenty-seven hips (36.5%) had excellent results, 25 hips (33.8%) had good results, 16 hips (21.6%) had fair results, and 6 hips (8.1%) had poor results.
Radiographic results
We evaluated the radiographic parameters of the 67 patients (81 hips) who had preoperative and immediate postoperative radiographic data available. Of these, 60 patients (74 hips) came to our hospital for clinical and radiographic follow-up and 7 patients (7 hips) had radiographic data before revision. Radiographic results are shown in Table 2. The mean canal flare index of all available cases was 3.57 ± 0.63 (range, 2.66–5.97). There were 66 stems (81.5%) implanted in neutral alignment, 5 stems (6.2%) in varus alignment, and 10 stems (12.3%) in valgus alignment. No change in stem alignment >5° was detected in the series of postoperative radiographs in all cases. Correct sized stems were used in 53 hips (65.4%), and undersized stems were used in 28 hips (34.6%).
Radiographic results.
CFI: canal fill index; n: number.
In the 78 hips without stem revision, there was no radiographic evidence of stem loosening during the mean follow-up of 20.1 years (Table 2). Subsidence was found in three hips (3.8%); mean subsidence was 5.3 mm, without any progressive stem subsidence. All subsidence occurred in the first year after the initial procedure. Stress shielding was present in 53 hips (67.9%). No cortical hypertrophy was found. Heterotopic ossification was found in 21 hips (26.9%). Spot-welds were found in three hips (3.8%) located in Gruen zones 3, 5, and 6. Femoral osteolysis was present in 22 hips (28.2%). Radiolucent lines were present in 31 hips (39.7%). Distribution of osteolysis and radiolucent lines are shown in Figure 3.
Distribution of osteolysis and radiolucent lines according to Gruen zone for the 78 hips without stem revision at the latest follow-up in (a) anteroposterior and (b) lateral radiographs.
Cox regression analysis showed a significant negative influence of varus alignment on the survival of the femoral stem (p = 0.004). There was no influence of age (p = 0.458), gender (p = 0.116), diagnosis (p = 0.987), femoral osteotomy (p = 0.906), bipolar hip components (p = 0.991), joint interface (p = 0.782), body mass index (BMI; p = 0.08), CFI (p = 0.618), bone graft (p = 0.129), or canal flare index (p = 0.709) on the survival of the femoral stem. χ 2 test showed no influence of bipolar hip components on osteolysis (p = 0.915) and radiolucent lines (p = 0.206). There was no influence of CFI (p = 0.763), alignment (p = 0.933), or canal flare index (p = 0.882) on stem subsidence. There was also no influence of CFI (p = 0.963), alignment (p = 0.123), or canal flare index (p = 0.661) on intraoperative fracture. There was a significant negative influence of bone graft on heterotopic ossification (p = 0.016). Fisher exact test showed no influence of bipolar hip components on stem subsidence (p = 0.473).
Discussion
With improvements in the design of cementless stems, clinical results and survival rates have constantly improved. 17,18 However, to our knowledge, studies with a mean follow-up duration of more than 20 years are rare. As the proximal femoral morphology and bone mineral content of Asians and Westerners are different, potentially influencing the outcomes of hip components, 3 we investigated the long-term performance of the CLS stem with a mean follow-up period of 20.1 years. We found excellent stem survival rates, which were slightly better than those reported in previous studies 1,6,19 and similar to those in a minimum 10-year follow-up study. 9
Although opinion is divided, undersized stems, 20 varus stem alignments, 21 and previous osteotomies 1 were reported to have negative influence on stem survival rates. In our study, we found that varus stem alignment had a statistically significant negative influence on stem survival rate, similar to the results of Mallory et al. and Vresilovic et al. 21,22 However, our study confirmed that undersized stems and previous osteotomies did not compromise stem survival rate. In our experience, CLS stems with a tapered proximal morphology could achieve stability following a mild subsidence after the initial implantation. Thus, we did not find a negative influence on the incidence of stem loosening, although undersized stems were used in 28 hips. We believe that varus stem alignment could cause abnormal stress distribution and poor prosthesis seating in the proximal femur, which might be associated with periprosthetic fracture and stem loosening. Two stems were revised in the varus alignment group. One, which had been stable before revision, was revised because of periprosthetic fracture caused by trauma and the other was revised because of loosening. This patient had developed proximal femoral fracture before bipolar hip arthroplasty with the CLS stem, which was performed to improve the pain caused by traumatic femoral head necrosis. Loosening was caused by progressive osteolysis, which was detected radiographically before revision, and debris around the stem, which was identified by pathology at revision surgery. The bipolar hip component produced a great deal of debris as a result of impingement between the liner and neck, and debris led to the failure of self-centering mechanism with this type of component, whereas there was no CLS stem loosening in that cohort, which had been followed for 10 years. 23 In the present cohort, which was followed for 20.1 years, we also evaluated the result of bipolar hip components and expected a high risk of revision and compromised radiographic outcome, because of the severe circumferential wear and debris. However, most debris were located in the proximal femur, and there was no problem associated with bipolar hip components, such as revision, osteolysis, radiolucent lines, or subsidence of the CLS stem.
The clinical results in our cohort were satisfactory: 70.3% of patients without revision had a good or excellent JOA hip score, which is comparable to previous study. 6 Six patients had a poor JOA hip score because of pain and compromised activity associated with increasing age and osteoarthritis of other joints.
Evola et al. 6 reported that varus alignment showed a significantly higher risk of cortical hypertrophy, whereas no cortical hypertrophy was detected in our study, similar to the result reported by Aldinger et al. 20 The CLS stem rarely causes cortical hypertrophy even in cases of malalignment because it has a tapered geometry and is made of flexible titanium alloy, resulting in low filling of the distal medullary cavity and decreased abnormal stress distribution at the tip of the stem. 10 In our study, osteolysis and radiolucent lines were observed mainly at the proximal femur as a result of wear particles, which did not compromise stem stability, though spot-welds formed in three cases. Metal and ceramic femoral heads were used in our cohort with no significant influences on the long-term outcome of CLS stem.
Three cases (2.1%) of exact intraoperative longitudinal fracture of the proximal femur were recorded in medical histories. Although the cementless stem intraoperative fracture rate has been reported to be approximately 1%, 24 our result is higher than previous reports. When the CLS stem was inserted in a neutral alignment, it is necessary to rasp the medullary cavity of the greater trochanter and insert the stem. When inserting the stem, care must be taken not to fracture the medial bone cortex of the femoral neck. However, similar to the result of the study by Hwang et al. 9 , we found no significant influence of femoral morphology or CFI on intraoperative fracture.
Patient demographic factors such as age, gender, diagnosis, or BMI had no significant influence on the clinical and radiographic results. A high rate of heterotopic ossification was detected at the latest follow-up, which was attributable to the use of bone graft (p = 0.016). Foruria et al. 25 reported that the development of heterotopic ossification was significantly associated with the use of bone graft, and the relationship between bone graft and heterotopic ossification in our study confirmed this result.
Conclusion
This study demonstrates acceptable long-term clinical and radiographic results of the CLS stem in Japanese patients. To improve the stem survival rate, caution should be exercised to avoid varus stem alignment.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
