CT-measurement predicts shortening of stable intertrochanteric hip fractures

Abstract

Purpose

Intertrochanteric (IT) hip fractures can be treated with sliding hip screws (SHS) or cephalomedullary nails (CMN) based on the stability of the fracture. This stability is affected by the initial impaction of the fracture which can be difficult to assess. The aim of this paper is to develop specific pre-operative computed tomography (CT) measurements of IT fractures which are predictive of post-operative shortening.

Methods

A retrospective review was performed of 141 patients with AO/OTA 31A1 or 31A2 fracture patterns, who had pre-operative radiographs and CT scans, and who were treated with a SHS or a CMN. Pre-operative and post-operative imaging of IT fractures were analyzed for those fractures that shortened ≥15 mm post-fixation.

Results

11 fractures shortened ≥15 mm with CMN being protective of shortening (6/36 SHS versus 5/105 CMN, p = 0.0268). A novel measurement made on the pre-operative CT scan called the cortical thin point (CTP) detected differences between patients with <15 mm and ≥15 mm of post-operative shortening for the SHS group (p = 0.0375). CTP was found to be a reliable predictor for post-operative shortening of ≥15 mm when a cutoff threshold of 9 mm was used in the SHS group (p = 0.0161).

Conclusions

Measuring the CTP is predictive of post-operative shortening after fixation of an IT fracture with a SHS. CMN fixation may be protective of shortening. Patients with a CTP of ≤9 mm are at risk for fracture site shortening of more than 15 mm when treated with a SHS.

Introduction

Intertrochanteric (IT) hip fractures can be classified as stable or unstable based on the fracture pattern which affects implant selection.^1,2 Sliding hip screws (SHS) have been shown to have equivalent outcomes at lower cost compared to cephalomedullary nails (CMN) when used for stable fracture patterns.^1,3 Stable fracture patterns have a fracture line obliquity that allows compression at the fracture site under physiologic loading. For this to occur a buttress for the proximal fragment to settle against must be present on the distal fragment.^4,5 This buttress has been described as consisting of the posteromedial calcar and the lateral wall of distal fracture segment.^2,6

More recently, patterns that affect the stability of the buttress – including coronal plane fractures of the proximal segment and fractures with thin but intact lateral walls on the distal fragment – have been identified as markers of unstable patterns that have higher all-cause failure rates when treated with a SHS.^7,8 These variants may occur as part of what has been thought of as stable AO/OTA 31A1 and 31A2 pattern fractures.⁹ They can be difficult to detect on injury plain film x-rays taken in the emergency department, and therefore CT scans may be useful to better asses the fracture pattern.¹⁰

The ability to detect patterns predictive of fracture settling that causes substantial limb shortening after fixation of otherwise-stable fracture patterns remain poorly elucidated. SHS and CMN devices were designed to allow fracture sliding to improve union rates, and even substantial settling of the fracture may not be considered a mal-union.¹¹ However, too much settling may cause a diminished abductor level arm and a limb length discrepancy (LLD) which are associated with worse functional outcomes, albeit no decrease in survival.^12,13 A symptomatic threshold of shortening has not been defined for the ambulatory hip fracture population, but LLDs have been extensively studied for the similarly-aged hip arthroplasty population. A post-operative LLD of greater than 10 mm has been shown to lead to lower Oxford Hip Scores and physical function three years after total hip arthroplasty.¹⁴ Although controversial, an LLD of 1 cm or less is commonly used as an acceptable threshold that is well-tolerated after arthroplasty.^14,15

One characteristic of a stable IT fractures that might predict substantial settling of a fracture is the impaction of the fracture present on the initial imaging. Fractures that have impaction of the lateral edge of the proximal segment against the lateral wall of the shaft will have a large void present after reduction. These impacted fracture patterns may be at risk for sliding back to the injury position if there is no implant, such as a CMN, to fill that void.¹⁶ This impaction of the fracture site at injury cannot be reliably seen on plain film x-rays but is easily seen and measurable on CT scans. The purpose of this study was to determine if measuring impaction could be predictive of an IT fracture that is at risk of substantial shortening. Specifically, we hypothesized that: 1) impaction is easily and reliably measured on CT scans, and 2) the amount of impaction measured on the coronal reformats of the pre-operative CT scan will predict the risk of post-operative shortening for stable fracture patterns when a SHS is used but not when a CMN is used.