Multi-center study of use of the Exeter stem in Japan: a 10-year follow-up report

Purpose Since the introduction of the Exeter stem for clinical use in Japan in 1996, the number of stems used has continued to rise owing to its favorable results. We investigated the outcomes of patients who had previously undergone total hip arthroplasty with the Exeter stem in Japan with a 10-year + follow-up period. Methods This retrospective cohort study used clinical and radiographic data of 682 cases of primary total hip arthroplasty performed using the Exeter stem. Results The mean postoperative follow-up period was 13.3 years. Femoral-side revision was required in 14 hips, with no cases of aseptic stem loosening-associated revision observed. Kaplan–Meier survival analysis predicted 97.3% 15-year survival when revision for any reason was used as the endpoint. Conclusion The obtained findings suggested the excellent long-term stability of the Exeter stem for primary total hip arthroplasty in Japan.


Introduction
The Exeter stem (Stryker Orthopedics, Mahwah, NJ, USA), a polished, double-tapered, colorless stem, has been delivering favorable postoperative clinical outcomes since its advent in 1969 [1].Over the years, the Exeter stem has undergone various modifications in terms of materials and surface processing.In 1988, the universal taper design was introduced for clinical use.In 2001, the trunnion shape was changed to a V40 taper.No other modifications have been made to its overall shape; however, the size has been adjusted over time.In 1997, small stems with 30-mm and 33-mm offsets were introduced for use in Asia-Pacific populations.In 1988, stems with 37.5-mm and 35.5-mm offsets were introduced for patients with acetabulum dysplasia or a small build.In 2014, shorter stems of 125 mm with 37.5-mm and 44-mm offsets were introduced for narrow medullary canals and highly curved femurs.
Clinical adoption of Exeter stems has been widespread in Japan since 1996, with their annual usage steadily increasing, establishing them as the preferred choice for cement stems.However, studies involving small Exeter stems with small offsets in Asian populations remain limited.In 2012, Fujita et al. [2] conducted a multi-center study involving 1000 joints with multiple surgeons who received direct surgical training from the hospital facility where the Exeter stem was developed.Interestingly, they reported excellent short-term radiographic results with a mean follow-up period of 4 years.Afterward, as some cases of revision or changes in the area around the femur were observed over time, a long-term follow-up study was conducted.Therefore, in this study, we aimed to investigate the radiographic data of patients who underwent primary total hip arthroplasty (THA) using the Exeter stem in four teaching hospitals in Japan over a mean follow-up period of 13.3 years.

Surgical technique
The posterior and lateral approaches were used in 878 and 122 hips, respectively, according to the modified Dall technique [3,4], at the surgeon's discretion.On the acetabular side, cementless and cement cups were used in 613 and 387 cases, respectively.On the femoral side, the Endurance bone cement (DePuy CMW, Blackpool, UK) and Surgical Simplex P radiopaque bone cement (Stryker Orthopedics, Mahwah, NJ, USA) were used in 44 and 956 cases, respectively.A stem developed for Asian populations (with a 30-mm, 33-mm, and 35.5-mm offset) was used in 280 hips (28.0%).Each surgeon selected the appropriate stem size and offset according to the patient's physique.All surgeons used modern cementing techniques.However, one surgeon combined the interface bioactive bone-cement technique (IBBC) [5][6][7].

Clinical assessment
We recorded the postoperative complications, such as dislocation, infection, deep vein thrombosis, nerve palsy, and fractures; associated factors; and surgical techniques used in THA revisions.

Radiological assessment
For radiographic evaluation, preoperative bilateral anteroposterior (AP) images were obtained in all patients who underwent follow-up examinations.Postoperative radiolucent lines (RLLs) at the bone-cement interface were assessed using the Barack classification.At the final examination, the quality of cementation was evaluated in the radiographic images based on the Gruen zone classification.Nonlinear localized balloon-like transparent image findings were carefully evaluated for osteolysis.Radiographic loosening was assessed using the Harris classification system.Stem slip was evaluated using the Fowler method.Cortical hypertrophy (CH) was evaluated using a modified Gruen classification with zone 4 divided into medial (4 M) and lateral (4L) to analyze stress distribution [2].Varus/valgus insertion angle (≥ 3° from neutral) was measured in postoperative radiographic bilateral AP images.All measurements were corrected for magnification using known dimensions of the femoral head.Radiographs were evaluated independently by four co-authors blinded to the patient's clinical status.Agreement was reached through consensus-based discussion in cases of discrepancies among the observers.

Statistical analysis
Primary endpoints were any revision and stem revision for any reason at 10 and 15 years.The chi-square test was used to determine the relationship between CH and the varus/ valgus insertion angle, as well as with stem offset size.Survival curves were estimated using the Kaplan-Meier method [8].EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan) [9], a graphical user interface for R 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria), Fig. 1 Among the 1000 primary THAs performed, 195 patients died, 101 were not available for follow-up, and 22 underwent stem revision.In total, this study included 682 patients over a minimum follow-up period of 10 years was used for all statistical analyses.The significance level was set at 5%.

Clinical assessment
Postoperative complications were as follows: dislocation (total n = 36; conservative, n = 25; revision for recurrent dislocation, n = 11), infection (n = 9, remission achieved after conservative treatment in one patient only), deep vein thromboses (n = 5), incomplete sciatic nerve palsy (n = 2, resolved with conservative treatment), and trauma-related femoral fractures (n = 1).Fractures were classified as Vancouver type B1, osteosynthesis combined with allogeneic bone grafting was performed, and implant revision was not required.Notably, no fractures occurred during the intraoperative period.No fatal pulmonary thromboembolisms were observed.
At the final follow-up examination, 22 hips had been revised for several reasons (Table 1).Of these, eight hips required revision only on the acetabular side, whereas 14 hips also required removal and revision of the femoral stem.Femoral revision techniques included cement-in-cement (n = 7) [10], in-cement (n = 2) [10], impaction bone grafting (n = 2) [11,12], IBBC (n = 1), combined strut allograft and cement-in-cement (n = 1), and cement long stem (n = 1).The cement-in-cement method was used for all recurrent dislocations.The standard size Exeter stem was used for revision in 13 hips, except for one case, which was revised with a long stem.
Stem slip < 2 mm occurred in 88.6% of hips (Table 2).Stem slip ≥ 2 mm occurred in 67 hips, with a mean of 2.7 mm. Excessive slip (≥ 5 mm) occurred in one case due to cement fracture, with no revision required (17 mm).

Discussion
In this follow-up study of 1,000 primary THAs, 14 (1.4%) required removal and revision of the femoral stem.Among the 682 primary THAs followed up for ≥ 10 years, radiographic findings indicated excellent overall results, with no cases of RLL at the bone-cement interface observed.These results were consistent with those of studies reporting the efficacy of the universal and V40 Exeter stems, even after modifications to its trunnion shape in 2001 [13,14].The long-term stability of the Exeter stem can be attributed to its design, which, in accordance with the tapered slip theory [1,15], allows for controlled subsidence within the cement that primarily transmits the load from the proximal femur to the bone [16].Typically, a 1-2-mm slip occurs in the early postoperative period and continues to progress, albeit slightly, without stopping.In this follow-up study, even after > 10 years, 88.6% of slips were < 2 mm, and only one patient exhibited an excessive slip of 17 mm due to a cement fracture.One case of excessive slip involved an inactive older adult patient with mild pain; therefore, revision was not performed.A substantial slip over time could predict the occurrence of periprosthetic or cement fracture; in cases of rapid progression or persistent pain, revision surgery may be required.
Focal osteolysis was observed in 10 hips, presenting as localized areas of transparent bone in the radiographic images.Although this phenomenon does not directly lead to stem loosening, widespread occurrence can contribute to loosening over time.Focal osteolysis may occur when the thin cement mantle ruptures and affects the bone-cement interface [17].Therefore, the stem must be inserted with sufficient care to avoid defects to the cement mantle.In most of our cases, revision was not required, with the exception of one hip.
The earliest case of CH was observed at 1 year postoperatively, with the number of cases gradually increasing thereafter.The incidence of CH increased from 92 out of 985 hips (9.5%) in the previous study by Fujita et al. to 200 out of 682 hips (29.3%) in the current study, with 170 hips (24.9%) representing new cases.While there were no cases in which the CH completely disappeared, some cases demonstrated changes in CH thickness (Fig. 4), and most of the CHs were located in zone 5. We inferred that this was the result of Exeter stem slipping within the cement mantle, which, owing to the differing taper shapes medial and lateral side, increased the load stress transmitted from the vicinity of the calcar to the distal medial portion.
Among the 200 cases showing CH, varus/valgus insertion was observed in four hips (0.5%) and Asian-sized stems were used in 67 hips (29.0%).While a significant difference was observed in the varus/valgus insertion, it resulted from a lower incidence of CH and was not the cause of CH.There was no significant difference between Asian-sized stems and CH appearance.
Fujita et al. [7] conducted a 10-year follow-up study comparing two cement stems with different designs and fixation styles.Their radiographic analysis revealed that CH was predominantly found in the lateral zones 3 and 4L of the composite beam design, and in the medial zones 4 M, 5, and 6 of the taper-slip Exeter stem design.
Although CH was observed in a large proportion of cases (29.3%) in the present study, it did not lead to loosening to the extent of periprosthetic fracture or necessitate revision.Numerous studies have explored the occurrence of CH and its impact on clinical scores and periprosthetic fractures.2 Kaplan-Meier survival analysis predicts 97.9% and 97.3% 10-year and 15-year survival, respectively, when revision for any reason is the endpoint Fig. 3 Kaplan-Meier survival analysis predicts 98.7% and 98.2% 10-year and 15-year survival, respectively, when femoral stem revision for any reason is the endpoint Iwase et al. [18], in their 10-year follow-up study of 100 hips with CH and hip pain, reported that patients with CH had better pain scores on the Japanese Orthopedic Association Hip Disease Evaluation Questionnaire [19].In a 15-year postoperative follow-up period of THAs using cobalt chromium CPT stems (Zimmer, Warsaw, IN, USA), Baryeh et al. [20] reported CH in six out of 500 hips (1.2%) but observed no cases of periprosthetic fracture.Conversely, Oe et al. [21] used a cobalt chromium SC-stem (Kyocera, Kyoto, Japan) with a curved triple taper shape and a stainless steel C-stem (DePuy International, Leeds, UK) featuring a straight triple taper shape in a cohort of 1,777 hips.After a mean follow-up period of 5.4 (range, 3.5-6.3)years, atypical periprosthetic fractures were observed in five SC-stem hips.In four of these cases, cortical bone changes were observed in radiographic images before the fractures occurred.Similarly, Wakama et al. [22] used SC-stem in 559 hips and reported atypical periprosthetic fractures in five hips (0.9%) at 3.5-6.4years postoperatively after local cortical bone changes occurred.
Hirata et al. [23] developed an experimental model consisting of four different metals and two types of cement with varying viscosities.Their findings demonstrated that metals with the same degree of surface roughness exhibited different friction and wettability coefficients.This suggests that the extent of femoral stem subsidence in vivo is influenced by factors such as surface roughness, metal composition, and cement viscosity.Thus, even slight variations in design shape and material properties may impact outcomes; hence, caution should be exercised.
In the present study, which involved the use of the Exeter stem in Asian patients, only one patient (0.1%) experienced a periprosthetic fracture due to trauma during the follow-up period.Palan et al. [24] compared four different brands of cemented stems for the revision of periprosthetic fractures and proposed that the rate of periprosthetic fractures varied depending on the fixation method and stem design.Kazi et al. [25] analyzed National Joint Registry data of 292,987 primary THAs performed over a follow-up period of approximately 10 years and compared the causes of revision in 5,492 hips in terms of fixation style and primary brand used, including the Exeter and Charnley stems (DePuy International).Their findings indicated that taper-slip stems were associated with a higher incidence of periprosthetic fractures, fewer cases of aseptic loosening, and lower rates of revision than those of composite beam stems.Additionally, a comparison between the Exeter stem and other tapered slip stems revealed that the Exeter stem exhibited a lower risk of periprosthetic fracture-related revision.Notably, the Exeter stem demonstrated the highest survival rate.
The present study had several limitations.First, longitudinal clinical outcomes were not included because these data were not obtained from one institution, which accounted for > 50% of the participants.Second, despite the large sample size, many patients died or dropped out, resulting in long-term follow-up being achieved in only 68.2% of participants.Third, this was an observational study that had no control series.
In conclusion, this was the first study to investigate the use of the Exeter stem in Japan with a follow-up period > 10 years and reported excellent midterm results when the Exeter stem and modern cementing techniques were used for primary THAs in Japan.Importantly, no cases of aseptic loosening were observed.To determine the performance of the Exeter stem among Asian populations, further studies from additional centers and longer follow-up periods are needed.Overall, the Exeter stem shows great

Fig.
Fig.2 Kaplan-Meier survival analysis predicts 97.9% and 97.3% 10-year and 15-year survival, respectively, when revision for any reason is the endpoint

Table 3
Relationship between stem position or offset size and cortical hypertrophy *1: Chi-square test result was p < 0.05 *2: Chi-square test result was not significant *3: Asia-Pacific stem size range