1 Introduction

The “French Paradox” technique for total hip replacement inevitably prompts discussion regarding the thickness of the cement mantle around the femoral stem. The French Paradox technique involves clearing all cancellous bone from the medullary canal in order to position a large canal filling implant, with an associated thin and sometimes incomplete cement mantle.¹ This method has been traditionally used with the Charnley-Kerboull and Ceraver Osteal stems, whose rectangular cross-sections are thought to contribute to intrinsic stability.¹ This is opposed to the method of cementation more commonly used in the USA and UK in which the canal is over-reamed by 2 mm with a view to creating at least a 2 mm cement mantle.² Theoretical, laboratory, clinical and post-mortem studies have recommended a minimum of a 2–4 mm thick, complete cement mantle.¹ Focusing on clinical studies, Anthony et al. present local cement defects as a route for the occurrence of osteolysis, Star et al. reviewed a series of total hip replacements (THRs) and found that a thin cement mantle at the medial diaphysis contributed to femoral component mechanical loosening and Ebramzadeh et al. found that a 2–5 mm cement mantle was optimal and THRs with either a thicker or thinner cement mantle had worse outcomes.^3–5 However, Ebramzadeh et al. also found less than 2 mm of cancellous bone to be advantageous as well as stems that filled more than half of the medullary canal.⁵

The “French Paradox” technique adopts a canal filling approach in which the cancellous bone is removed during femoral preparation and the largest possible stem is inserted with line-to-line cementation with a thin and potentially incomplete cement mantle.⁶ This technique is not limited to a triple taper stem. Despite being at odds with the findings of the above studies, excellent results were demonstrated with the canal filling technique with the Charnley-Kerboull and Ceraver Osteal stems.^7–10 The Ceraver Osteal stem is a smooth collared stem, the Charnley-Kerboull stem is a polished, double-taper stem with a quadrangular cross-section also with a collar. Skinner et al. directly compared the 10-year survival and radiological outcome of THR using two different techniques, one provided a thicker cement mantle and the other carried out line to line reaming leading to a thinner cement mantle. They found that a thinner cement mantle was not worse and may produce better long-term outcomes than current teaching suggests.²

Femoral stems are designed to function as either composite beam or taper slip prostheses.¹¹ The Kerboull and Osteal stems in which good results have been demonstrated have a collar leaning them toward the composite beam design, while Skinner et al. used the Freeman femoral component (Corin), a neck retaining implant felt to improve stability.¹² This study looks at a large single-surgeon consecutive series of cemented total hip replacements using the canal filling technique with a triple tapered stem. We aim to evaluate the incidence of aseptic loosening of the C-Stem total hip replacement leading to revision when using the canal filling cemented stem technique, as well as the survivorship including all cause revision. We hypothesise that the use of the canal filling technique with a triple tapered stem does not negatively impact on survivorship and does not lead to increased revision for aseptic loosening of the stem.

2 Patients and methods

This is a retrospective study of a consecutive series of cemented total hip replacements (THR) using the C-Stem 9/10 taper stem (DePuy Synthes) implanted using the canal filling technique, from July 2007 to March 2018. Regarding inclusion criteria, all patients undergoing primary total hip replacement during the study period under the care of the study consultant, where a C-Stem implant was used with canal filling cementation technique, were included. Patients not suitable for a cemented stem, where the C-Stem implant was not used or who were undergoing revision surgery were excluded. All THRs were under the care of the same consultant with 37% performed by trainees under that consultant's supervision. They were performed within a single NHS hospital and one local private hospital. All patients undergoing primary total hip replacement did so due to primary osteoarthritis (89%), avascular necrosis (2%), acute fracture (6%), non-union of hip fracture (2%) or protrusio (1%) and were considered suitable for a cemented total hip replacement. The acetabular components were the Charnley Elite Plus cup (DePuy Synthes) for the earlier cases (32 THRs) and Marathon cup (DePuy Synthes) for all remaining cases.

Data were obtained from local electronic databases including clinic letters, operation notes and x-rays. Information regarding age, gender, diagnosis, stem size, postoperative follow-up, complications, surgeon grade and any revision surgeries. NJR records were reviewed to confirm the exact number of revisions, including some performed in other hospitals. The local ethics committee was consulted, and approval was not needed.

The total hip replacements were performed with the patient supine using a modified Hardinge approach, without trochanteric osteotomy with an omega incision to detach the abductors and expose the capsule. Antibiotics were given preoperatively according to hospital policy. The femoral neck cut was preformed using a Gigli saw. Acetabular exposure was optimised with Charnley pins and retractors. Sequential reaming was performed, subchondral cysts curetted and drill holes made appropriately. An acetabular component was selected, the flange trimmed, and implant cemented and positioned appropriately. Femoral preparation involved initial use posterolaterally of the box chisel followed by two pencil reamers. A Charnley spoon was used to clear as much cancellous bone as possible from the medullary canal and the largest trial stem possible was identified and inserted. A trial reduction was done, from which offset was also judged. A size 28 metal head was used for all patients. An intramedullary plug was inserted, the femoral canal was washed with pulse lavage and dried. Palacos R + G (Heraeus Medical, Wehrheim, Germany) bone cement was prepared by vacuum mixing and inserted in a retrograde manner using a cement gun. The selected C-stem implant with void centraliser was inserted and cement allowed to set. A final trial reduction was done to confirm femoral head selection, which was then positioned, and the hip reduced. Local infiltration was done according to local enhanced recovery protocol and closure was done in layers.

Post-operatively, patients were mobilised fully weight bearing with hip precautions. A postoperative x-ray was completed and they were followed up 6 weeks, 6 months and 1 year postoperatively. Follow-up beyond 18 months was patient specific. An example is seen in Fig. 1.

Close

Fig. 1 X-ray demonstrating the C-stem total hip replacement using the canal filling techinque.

Export to PPT

3 Statistical analysis

Statistical analysis was performed with the use of percentages, mean values and Kaplan-Meier survivorship curves with all-cause revision as the endpoint.

4 Results

321 cemented total hip replacements in 307 patients were identified, between July 2007 and March 2018. Patients were followed up for a mean of 68 months (9–137). The mean age was 70 (30–91), with 10% (33) of patients less than 60 years old. 59% (190) of patients were female, while 41% (131) were male and 37% (119) were performed by a trainee under direct supervision of the consultant. The mean stem size was 3 (1-8), with the most commonly used head being 28, 0 (76%). A summary of patient characteristics, indication for surgery and implant size is demonstrated in Table 1. At the time of data collection, 283 patients were alive. 8 patients (2.5%) were lost to follow-up. Survival analysis was performed for all-cause revision. The secondary outcome was revision for aseptic loosening of the stem.

Revision rate for all reasons was 2%, of which 2 were for acetabular loosening, 2 for recurrent dislocation, 1 for inadequate offset, 1 for periprosthetic fracture and one for infection (Table 2). Mean time to revision for these patients was 53 months (29–106). Ten-year implant survival (91 hips at risk) was 95% (95% confidence interval 94.9–95.0%) (Table 3) with the Kaplan-Meier survival curve seen in Fig. 2. Other complications included trochanteric bursitis (3%), superficial infection (1%), ongoing pain (1%), dislocation (1%) and deep infection (0.3%). There were no cases of revision for aseptic loosening of the femoral stem.

Close

Fig. 2 Kaplan-Meier survival analysis.

Export to PPT

5 Discussion

Our results demonstrated a ten-year implant survival (91 hips at risk) of 95% (95% CI 94.9–95.0%). A total of 7 hips were revised. In the context of the literature, with the Charnley Kerboull implant, there was a survival rate of 90.5% (95% CI 84.2 to 96.8) at 17 years,⁷ and with Ceraver Osteal there was a survival rate of 87.3% at survival at 20 years. In Skinner's comparative study, survival at 10 years with the thicker cement mantle was 97.2% (95% CI 90.6 to 99.2) and with the thin cement mantle, 98.8% (95% CI 92.9 to 99.8). The difference between these was not found to be statistically significant.² Purbach et al. obtained a survival analysis only assessing for those revised for aseptic stem loosening or stem fracture, there were no such cases in 10 years.¹³ The survivorship of this study was therefore lower than that of the literature. Reasons for revision were acetabular loosening (0.6%), recurrent dislocation (0.6%), inadequate offset (0.3%), periprosthetic fracture (0.3%) and infection (0.3%). Purbach had a revision rate of 5% in 10 years which is in fact higher than the 2% in this study. Their reasons for revision were similar: dislocation, infection and aseptic cup loosening.¹³ The most common complications experienced in this study were trochanteric bursitis (3%), superficial infection (1%), ongoing pain (1%), dislocation (1%) and deep infection (0.3%). These are in keeping with Purbach et al. who found infection rates of 1.7% and dislocation 1.9%. A further study comparing results of the C-Stem with the Exeter Universal stem found similar complication rates, although trochanteric bursitis was not mentioned.¹⁴

Our results demonstrate that in 321 THRs, there were no cases of revision for aseptic loosening of the femoral stem. This confirms that using the canal filling technique with a triple tapered cemented stem did not increase the incidence of revision for aseptic loosening. In the context of the literature, looking back to the Charnley-Kerboull stem, three stems (1.8%) were revised for femoral loosening.⁷ These results are in keeping with the 20 year follow up for the Ceraver Osteal stem.⁹ In Skinner's comparative study, 2 of the 92 cases with the thicker cement mantle were revised for aseptic loosening, while 1 of the 97 cases with the thinner cement mantle was revised for the same reason. In the context of the C-Stem femoral implant, a large study with long term follow-up of the C-stem THR using broaches to produce a minimum 2 mm cement mantle, found no revisions for aseptic loosening and no cases of clinical or radiological loosening with at 10 years.¹¹ In summary, regarding the secondary outcome of this study, our results did not demonstrate any cases of revision for aseptic loosening and this is in keeping with pre-existing literature for the canal filling cemented stem technique as well as for the C-Stem implant. Despite the use of a triple tapered stem of a taper slip design, the thinner cement mantle has not led to any revisions for aseptic loosening within the follow-up remits of this study.

This study analyses data from a large consecutive series of total hip replacements using the same femoral implant, the same surgical technique with all cases supervised or performed by the same consultant. There is medium range follow up and data were cross referenced with the NJR to maximise accuracy.

In contrast with the pre-existing literature, this study to our knowledge is the first of its kind to demonstrate the use of the canal filling cemented stem technique in combination with a taper slip design femoral prosthesis. Kerboull's theory was to ensure implant alignment, remove cancellous bone that may not carry load with aging and to ensure the cement “was not subjected to forces that it could not withstand”. He believed that a simple way to achieve this was to remove all of the cancellous bone and force in the largest possible stem to fill the medullary canal.⁶ Kerboull made alterations to the original Charnley stem, widening the proximal stem in order to increase loading of the proximal femur and reduce axial forces at the stem tip, simultaneously making it double tapered.^11,16 This provides the advantages of a good initial stability with direct load transfer to the cortex bypassing the mechanically weak trabecular bone as well as increased rotational stability.¹⁷ Shen et al. state that from an engineering point of view, in a composite beam scenario secure bonding between metal and cement is essential.¹⁵ By using the canal filling technique, using the largest possible stem will create higher cement pressures, which could improve cement interdigitation and penetration up to the cortex.^9,11 The C-Stem implant is most definitely a taper-slip design implant.¹¹ However, there is one feature that the C-Stem has in common with the Charnley-Kerboull stem which may explain its unlikely successful outcomes with this technique. It is a polished, triple tapered stem, the third taper being medio-lateral, proximally, to promote axial loading of the calcar.¹⁸ The broad lateral surface and narrow medial face has been shown in finite element analysis to reduce stress concentration at the edges in contrast to the double tapered stems, thereby reducing risk of fracture of the cement mantle, which in the cases analysed in this study, is thin19.

Similarities between the C-Stem and Charnley-Kerboull stem are the polished surface, rounded edges and the proximally wider stem.^16,18 The key relevant feature is that unlike other tapered stems, the C-Stem is designed as a wedge in order to specifically load the proximal medial cortex.^20,21 By positioning the largest possible stem, this effect would be expected to be accentuated. In addition to this, Numata et al. evaluated the biomechanical behaviours of the CMK stem and surrounded cement and concluded that the stem produced hoop stresses, as are found in the taper slip stems, without excessive subsidence.²² This may further explain why the C-Stem was successful in a thinner cement environment.

It has been suggested that our understanding of the behaviour of the cement mantle in vivo is incomplete, with Langlais et al. suggesting that either a thin cement mantle behaves differently mechanically, or that with a canal filling technique the cement is protected in the way it is loaded.¹ Kerboull states that a canal filling stem provides good stability proximally, thereby reducing shear stresses and micromotion at the cement bone interface and decreasing stresses on the cement.¹⁶ Jansen et al. carried out used finite element analysis to evaluate the mechanical consequences of differing stem size and cementation technique. They found that canal-filling stems produced fewer cement cracks and less rotation than smaller stems and that where the cement mantle was supported in trabecular bone three were more cement cracks and more rotation.²³ This confirms the idea that a thinner cement mantle behaves differently mechanically. Regarding cement thickness, a radiological study by Scheerlinck et al. used CT to evaluate the cement mantle around Charnley-Kerboull replica stems and found that the cement mantle actually averages 3 mm, with few areas with less than 1 mm of cement mantle. They suggest that this is due to the pressurisation of the cement into cancellous bone and that the thinnest areas of cement tended to be supported by cortical bone.²⁴ This implies that in spite of the difference in technique, the cement mantle is not so thin as believed and due to removal of cancellous bone and the use of a canal-filling prosthesis, is well supported by cortical bone.

6 Limitations

The key limitation of the study is the lack of patient reported outcome measures. There was also a wide variety as to follow-up time period and some patients were lost to follow up. We have accurate figures regarding revision rates, but it is not known how many patients may have asymptomatic aseptic loosening. No routine follow-up imaging was carried out to look for aseptic loosening. We also did not investigate subsidence of the femoral prosthesis.

7 Conclusion

This study makes a valuable addition to current literature, demonstrating good medium-term results of a taper slip design prosthesis with a thin cement mantle. There were no cases of revision for aseptic loosening of the stem. Regarding the primary outcome of survivorship, ten-year implant survival was 95%, which is lower than in the literature. Long term survivorship analysis would be valuable to further assess this.

Author conflict of interests

None.

Contributions not meeting authorship criteria

None.

Funding sources

None.

Author financial support

None.

CRediT authorship contribution statement

Anna S. Walsh: Methodology, Formal analysis, Data curation, Writing – original draft, Writing – review & editing. Muni Pinjala: Conceptualization, Methodology, Formal analysis, Data curation, Writing – original draft. Siddharth Lokanathan: Conceptualization, Methodology, Data curation. Saqif Hossain: Conceptualization, Methodology, Validation, Data curation, Supervision.