The theme of this work is the application of the engineering theory of frictional torque to total hip replacement.

Can the benefit of the low-frictional torque be demonstrated in the long-term results of the Charnley LFA?

The statement, in Charnley’s final definitive volume: “Low friction arthroplasty of the hip. Theory and Practice” summarizes the concept and its application in clinical practice. The smallest, practically acceptable diameter head of the stainless steel femoral component, articulating with a thick-walled UHMWPE cup. This would encourage movement, preferentially at the level of the articulation rather than at the bone-cement interface, thus reducing the likelihood of cup loosening. Why has this basic principle not been generally recognised and accepted in clinical practice? Was the use of large diameter heads driven by fear of post-operative dislocation? Was the use of hard on hard articulations dictated by the belief that tissue reaction to UHMWPE wear particles is responsible for component loosening? It is probably correct to suggest that even Charnley had some doubts as to the importance of low-frictional torque principle as applied in clinical practice.

“The property of the low–frictional resistance in the artificial joint does not appear to reveal itself in the clinical results as a matter of such importance as I had originally imagined.” (1966) UHMWPE had already been routinely used for 4 years, the principles of the design, materials and the details of the surgical technique were established. A total of 1321 LFAs had been implanted and none have been revised for loosening of the cup, the stem or fracture of the stem. Clinical success, absence of revisions and lack of reports of other designs to allow any comparison, probably prompted Charnley’s statement. It was two further years and 1180 LFAs before the first revision, for the above mentioned complications, had to be carried out.

Anderson and colleagues (1972) questioned the significance of low-frictional torque as a contributor to the preservation of the bone-cement interface. They compared torsional moments needed to loosen cemented Charnley and metal on metal McKee design cups. They concluded that the torsional moments were from 4 to more than 25 times higher than the frictional moments measured and, therefore, unlikely to contribute top cup loosening. Charnley argued “…that if the demarcation of the cemented socket from the adjacent cancellous bone is present, the avoidance of high frictional torque might permit such a socket to function for many years than would be the case if the high frictional torque were present.” (1978)

In 1981/1982 Charnley reviewed the longest follow-up results available. He pointed out that “…gross demarcation, together with migration, affected a total of 25 % of sockets, though at the time of the review none of these patients showed any clinical defect. Given time and function 25 % of sockets were expected to fail by migration.” But “where severe demarcation was present, the fact that the migration of the socket had not progressed to clinical failure was attributed to the delaying action of the low–frictional torque using small diameter prosthetic heads.”

Charnley recognised aseptic loosening of the cemented UHMWPE cup as occurring at two levels: radiological and clinical. Radiological failure was described as: “… the operations judged on clinical grounds are still successful, but the radiological appearances indicate …that sooner or later the clinical failure must follow.” Clinical failure: “… pain returns, the early excellent function following the operation is lost and re–operation may be necessary.”

It is probably correct to suggest that misunderstandings and even confusion regarding interpretation of results of various total hip joint designs and methods, could have been avoided, if a strict distinction was made between clinical results and radiographic appearances.

Freedom from pain is due to the neuropathic nature of the arthroplasty – subject to correct patient selection and sound fixation of the components. Patients’ activity level is a reflection of patient selection for the operation. Clinical results do not reflect the mechanical state of the arthroplasty; loose and even migrating components are not necessarily symptomatic [1].

Ma and colleagues [2] studied frictional torque both in resurfacing and conventional hip replacements with cemented UHMWPE cups. They concluded that frictional torque was proportional to the diameter of the head of the femoral component; it was lowered as the thickness of the cup was increased.

Ritter and colleagues [3] compared the results of 67 Mueller arthroplasties (32 mm head, 50 mm cup diameter) with 300 Charnley arthroplasties (22 mm head, 44 mm cup diameter) at a minimum follow-up of 7 years. The incidence of cup loosening was 15 % compared with 4 %; survivorship at 5 years was 87 % compared with 94 %, and at 7 years 70 % compared with 86 % – were all in favour of the Charnley design.

When Morrey and colleagues [4] examined the correlation between the head size and acetabular revisions in 6128 arthroplasties, they concluded that “… although friction increases with the size of the femoral head theoretically, it remains significantly low that the bone-cement interface is not at risk”. (The author’s statement: “friction increase with the size of the femoral head” is not correct: frictional torque increases with the size of the femoral head, friction does not).

Radiolucent lines were more frequent with the 32 mm diameter head: 29 % as opposed to 15 % with the 22 mm diameter head. Even though the 32 mm head design was introduced after 5 years experience had been accumulated with the cementing of the 22 mm diameter heads, the incidence of aseptic cup loosening was still higher with the 32 mm diameter head: 6.8 % as compared with 2–3 % with the Charnley design. Similar experience was reported by Frankel et al. [5] “… three zone demarcation of the acetabular bone-cement was 56 % of the 32 mm group as compared with 5 % of the 22 mm group”. Their conclusion was that “… these results emphasise the adverse affects of large femoral head prostheses on the bone-cement interface and underline the need for alternative methods of fixation”. The authors make no suggestion as to what alternative methods of fixation could be used or how they would “circumvent the adverse effects of large head diameters.”

Mai and colleagues [6] examined the contribution of frictional torque to loosening in Tharies hip replacements. The conclusions, based on 1970 cases with a follow-up to 16 years, did not support their suggestion that larger heads performed better. With an overall revision rate of 47.10 % there was no difference in the results whether the head diameter was small, medium or large (23.5 %, 23.5 % and 26.5 % revision rate). Their explanation was that “polyethylene wear has a greater effect on the durability of fixation of the implant than frictional torque does.”