Evolution of the stem

By 1974, it had become clear that a stronger stem was needed with more sizes. A new range of stems was introduced at the beginning of 1976, slightly heavier in section than the original polished stems but incorporating the same basic geometry. A minor change was that the anterior and posterior surfaces of the upper fourth of these stems were parallel to each other and not tapered up to the base of the neck, as with the original polished stems. The new stem range was manufactured from 316L stainless steel for which there was no standard that required a polished surface. Thus, since polishing a stem is expensive, the new stems had a matt-surface that, in surface roughness terms, was two orders of magnitude rougher than the original polished stems (Figs.7a & b). No significance was attached to this change at the time. From 1983 onwards, all Exeter stems were manufactured from the high nitrogen, high fatigue strength, low corrosion stainless steel REX 73413 (a trademark of Heymark metals Ltd) under the proprietary name ‘Orthinox’ (a trademark of Howmedica Inc.).

Differences in the Behaviour of the Matt and Polished Stems:
Whilst the incidence of stem fracture with the matt-surfaced stems was virtually abolished, there gradually emerged the paradoxical finding of a significant increase in the incidence of focal femoral lysis and aseptic loosening, by comparison with the original polished stems. Of 180 matt stems inserted at the Princess Elizabeth in 1980, 10% had been revised for aseptic loosening by 10 years. This was a revision rate for aseptic stem loosening  almost four times higher than found with the original polished stems over 20 years14,15.

Focal femoral lysis gradually emerged as a problem with the matt surfaced stems and a study of the stems retrieved in these cases revealed the probable cause of the discrepancy in the results of the matt-surfaced and polished stems16. Every matt surfaced stem retrieved from cases with focal lysis showed evidence of abrasive wear on its surface (Fig. 8) that was associated with wear of the internal surface of the cement mantle, so leading to an enlargement of the inside dimensions of the mantle and thus potential instability of the stem with associated hydrostatic effects. It was shown that these focal lytic lesions, at or distal to the level of the lesser trochanter, invariably occurred in association with defects in the cement mantle (usually iatrogenic) so that fluid, debris and pressure changes in the enlarged stem-cement interface could obtain access to the endosteal surface of the femur at the site of the defect. Subsequently, this mechanism was shown to be valid for a number of different non-polished cemented stems17. An extensive study of stem wear on a wide variety of stem designs18 showed that the wear processes at the stem-cement interface were entirely different as between polished and matt-surfaced stems. Polished surfaces wear against the cement by a fretting mechanism that produces sub-surface pits on the stem and does not damage the cement. Matt surfaces wear by an abrasive mechanism that sacrifices the inner surface of the mantle and can, as a consequence, open the stem-cement interface to fluid flow and sometimes later destabilise the stem, especially in torsion. Furthermore, experimental studies19subsequently suggested that subsidence of the polished taper was able to seal off the stem-cement interface and so prevent fluid flow, a phenomenon that did not occur experimentally with non-polished tapers. These experimental studies provided a basis for understanding why punch-out fractures20 (Fig. 5) in the cement mantle surrounding the polished Exeter stems are so benign21.

Indifferent results with the matt-surfaced Exeter stems were reported by a number of authors outside Exeter 22 and also emerged in the Swedish Hip Registry23.

The Re-introduction of the Polished Surface:
By 1985, it had become clear that the polished surface should be re-introduced and henceforward, the  monblock Orthinox matt-surfaced stems were manufactured with a polished surface, entering clinical practice  at the beginning of 1986. With the steadily increasing length of follow-up of the original polished Exeter stems, in which persistence of the benign X-ray appearance was the rule, it eventually  became apparent that the phenomenon of subsidence of the polished stems within the cement was of importance functionally and biomechanically. It was therefore necessary to ensure that the new polished stems could not be ‘end-bearing’ in the cement and thus inhibit such subsidence. The old metal centralisers (Fig.9a), that sometimes ‘held up’ the original polished stems, were abandoned and replaced by an hollow centraliser, initially manufactured from ultra-high molecular weight polyethylene and later from pre-polymerised PMMA (Fig. 9b), that ensured that the stem tip could not be end-bearing in the cement.

The monoblock polished stems were used during 1986 & 1987 and then succeeded in 1988 by the Exeter Universal series that has been in use ever since (Figs. 6 & 10). Figs. 10a & 10b emphasise the differences between the Universal stems and the 1976-88 monoblock stems. In the lateral view, the parallel anterior & posterior surfaces of the monoblock 316L & Orthinox stems were changed to restore the taper up to the base of the neck, as with the original EN58J polished stems, and the upper end of the stem was stiffened slightly by adding metal to the shoulder. Modularity was introduced with interchangeable head sizes, the 26mm diam. head becoming the size most used in Exeter. Rasps were introduced for femoral canal preparation, together with an improved femoral introducer.

Since 1988, the available range of stem sizes and offsets has been gradually extended, though the basic double taper geometry has remained unchanged. To address certain problems of loss of femoral diaphyseal bone stock in revision surgery, longer stems were introduced. In order to allow the use of a ceramic bearing combination, the spigot at the head-neck junction was changed in 2001 to the V40 design24,25  of which there was already extensive clinical experience with other stem geometries.

Clinical experience with the Exeter Universal stem, using ‘contemporary’ cementing techniques, has been very satisfactory26,27,28. With the end point revision for aseptic stem loosening, stem survivorship is 100% at 10-12 years of follow-up29 of the first 325 of these implants to be inserted by the Exeter surgeons and their trainees at the Princess Elizabeth Orthopaedic Hospital. None of these hips have been lost to follow-up & the fate of every implant is known. Benign X-ray appearances are the rule (Fig.11). It is clear that the device functions in a similar way to the original polished Exeter. After 14 years of use, no evidence has yet emerged of any drawback to the modular head-neck connection.