The use of uncemented stems in femoral neck fractures in elderly patients: A comparative study of 671 cases

Abstract

Introduction

The use of cemented stems in elderly patients is associated with the risk of intraoperative embolic complications. Uncemented stems eliminate this risk, but their use is controversial because of the risk of mechanical complications, with estimated subsidence between 3.6 and 30% and periprosthetic fractures between 2.1 and 11% at 6 weeks. A retrospective multicenter comparative study was conducted using a propensity score analysis to evaluate mechanical performances of uncemented stems in femoral neck fractures in elderly patients to (1) compare the risk of mechanical complications and (2) assess the use of metaphyseal-anchored stems for this indication.

Hypothesis

There is no difference in the risk of mechanical complications, intraoperatively or postoperatively, between these stems.

Materials and methods

We conducted a multicenter retrospective comparative study including 358 uncemented and 313 cemented stems. The mean age was 84.5 years [83.9–85.1]. The inclusion criteria were patients 70 years and older and a follow-up of at least 6 weeks. The primary endpoint was a composite endpoint comprised of stem subsidence ≥ 2 mm or periprosthetic fracture (up to 3 months postoperatively). The secondary endpoints were infection, stem subsidence ≥ 2 mm, and operative time. These endpoints were analyzed using a propensity score to control confounding factors. A secondary analysis used the same endpoints to compare metaphyseal-anchored (short stems) versus Corail-like stems.

Results

After adjusting for the propensity score, we found 11.17% mechanical complications in the uncemented group (n = 40, 5.59% subsidence, and 5.59% fractures) versus 13.42% for the cemented group (n = 42, 7.99% subsidence, and 5.43% fractures). There was no statistically significant difference between the 2 values (Odds Ratio [OR] = 0.64 [95% Confidence Interval [CI]: 0.14–2.85] [p = 0.7]). The mortality rate due to cement embolism was 1%. There was no difference in the rate of subsidence (OR = 0.55 [95% CI: 0.02–12.5] [p = 0.7]), periprosthetic fracture (OR = 0.65 [95% CI: 0.13–3.12] [p = 0.7]) or infection (OR = 0.71 [95% CI: 0.32–1.55] [p = 0.4]). However, the operative times were longer in the cemented group (p = 0.03 for hemiarthroplasties [mean additional time 16 minutes] and p = 0.02 for total hip arthroplasties [mean additional time 22 minutes]). No difference was observed between the metaphyseal-anchored (short stems) and Corail-like stems regarding operative time, rate of infection, and rate of stem subsidence or periprosthetic fractures.

Discussion

This is one of the first studies to highlight cemented stem subsidence when used for femoral neck fractures in elderly patients. Using uncemented stems in this indication is still warranted, especially since they do not bring about more mechanical complications in the first few months. Metaphyseal-anchored short stems seem to give the same results as “standard” stems. However, these findings need to be assessed in the longer term.

Level of evidence

III; retrospective comparative study.