What is the Role of Stem Size and Offset in the Risk of Nonseptic Revision of the Exeter® 150-mm Stem? A Study From the Swedish Arthroplasty Register

Abstract

Background: The Exeter® stem is used worldwide, often in older patients, and it is the second most commonly used cemented stem design in Sweden. Previous studies have shown that for cemented stems with a composite beam, the smallest sizes result in an increased risk of revision for mechanical failure. However, little is known about whether the survivorship of the polished Exeter stem, which generally has been shown to be good, might be associated with design parameters such as stem dimensions or offset at extreme implant sizes.

Question/purpose: Are differences in (1) stem size or (2) offset of the standard Exeter V40 150-mm stem associated with differences in the risk of stem revision for aseptic loosening?

Methods: Between 2001 and 2020, 47,161 Exeter stems were reported to the Swedish Arthroplasty Register, with very high coverage and completeness documented during the period studied. In this cohort, we included patients with primary osteoarthritis who underwent surgery with a standard Exeter stem length of 150 mm and V40 cone with any type of cemented cups that had had at least 1000 reported insertions. This selection resulted in a study cohort representing 79% (37,619 of 47,161) of the total number of Exeter stems in the registry during that time. The primary study outcome was stem revision for aseptic indications such as loosening, periprosthetic fracture, dislocation, and implant fracture. A Cox regression was performed, with adjustment for age, gender, surgical approach, year of surgery, use of highly crosslinked polyethylene cups, and femoral head size and length dictated by the shape of the head trunnion. Adjusted hazard ratios are presented with 95% confidence intervals. Two separate analyses were performed. The first analysis excluded stems with the highest offsets (50 mm and 56 mm, which were not available for stem size 0). The second analysis excluded stem size 0 to include all offsets. Because stem survival was not proportional over time, we divided the analyses into two insertion periods, 0 to 8 years and beyond 8 years.

Results: Stem size 0 (compared to size 1) was associated with an increased risk of revision up to 8 years when all stem sizes were included (first analysis 0 to 8 years, HR 1.7 [95% CI 1.2 to 2.3]; p = 0.002). Forty-four percent (63 of 144) of revisions of size 0 stems were for periprosthetic fracture. There was no consistent association between stem size and risk of aseptic stem revision when size 0 was excluded in the second analysis beyond 8 years. The most common offset (44 mm) was associated with an increased risk of revision (compared with 37.5 mm) up to 8 years when all sizes were included (first analysis, HR 1.6 [95% CI 1.1 to 2.1]; p = 0.01). In the second analysis (beyond 8 years, all offsets included), offset of 44 mm was compared with offset of 37.5 mm; compared with the first period, this offset was associated with a reduced risk (HR 0.6 [95% CI 0.4 to 0.9]; p = 0.005).

Conclusion: We found overall high survival of the Exeter stem, with generally little or no influence of stem variations on the risk of aseptic revision. However, stem size 0 was associated with an increased risk of revision mainly for periprosthetic fractures. If the femoral anatomy offers a choice between sizes 0 and 1 in patients with poor bone quality who are at risk of periprosthetic fracture, our data speak in favor of choosing the larger stem if the surgeon believes it is safe to insert the larger size, or, if available, another stem design that has a documented lower risk of this complication. For patients with good cortical bone quality but very narrow canals, a cementless stem may also be a good alternative.

Level of evidence: Level III, therapeutic study.

Fig. 1

This flowchart shows how the number of stems included in this cohort was determined.

Fig. 2

This graph demonstrates the distribution of stem sizes included during the study period from 2001 to 2020. We clustered stem sizes 4 and 5 into one group. A color image accompanies the online version of this article.

Fig. 3

In this graph, we show the distribution of stem offsets between 2001 and 2020. We clustered 50-mm and 56-mm stem offsets into one group. A color image accompanies the online version of this article.

Fig. 4

This figure demonstrates the distribution of polyethylene types. There was a shift toward using highly crosslinked polyethylene after its introduction in 2008. A color image accompanies the online version of this article.