Does subchondral bone of the equine proximal phalanx adapt to race training?

Abstract

Sagittal fractures of the first phalanx are a common, potentially catastrophic injury in racehorses. These fractures are often linked to an acute, one time, biomechanical event; however, recent evidence implies that chronic exposure to stress can lead to the accumulation of bony changes that affect the structural integrity of the bone and increase the likelihood of fracture. The aim of the study was to compare variations of two common metrics of bone adaptation - subchondral bone density and thickness across the proximal articular surface of the first phalanx in Thoroughbred horses that (1) raced but never experienced a first phalanx fracture (Raced Control); (2) raced and had experienced fracture of the contralateral first phalanx (Contralateral to Fracture); (3) had never raced or experienced a first phalanx fracture (Unraced Control). A total of 22 first phalangeal bones were sampled post-mortem and imaged using micro-computed tomography calibrated for mineral density measures. Measurements of volumetric subchondral bone mineral density and thickness were taken from images at five sites from medial to lateral, in three coronal planes (25, 50 and 75% dorsal-palmar). At each of the 15 sites, measurements were repeated and averaged across 10 adjacent micro-computed tomography slices of bone, spanning 0.75 mm. The magnitude and variance of these measurements were compared between sites and between cohorts with non-parametric statistical tests. Across the proximal osteochondral surface of the first phalanx, the pattern of subchondral bone volumetric bone mineral density and thickness varied with each coronal section studied. The subchondral bone thickness was greater for the central and dorsal coronal sections, compared with the palmar section. For the race-fit groups (Raced Control and Contralateral to Fracture), the highest volumetric bone mineral density was in the central sagittal groove. The volumetric bone mineral density was significantly greater in the sagittal groove in the central coronal section in the raced than the unraced group. The Contralateral to Fracture group demonstrated significantly greater variance of volumetric bone mineral density compared with the Raced Control and Unraced Control (P < 0.0001), with no difference in variance noted between the Raced Control and Unraced Control groups. There was a small (R rank = 0.3) but significant correlation between subchondral bone volumetric bone mineral density and thickness in the Contralateral to Fracture group (P = 0.005). The findings demonstrate that differences exist in subchondral bone volumetric bone mineral density and thickness across the proximal osteochondral surface of the equine first phalanx in horses with different training histories. The findings also demonstrate that the subchondral bone of the sagittal groove of the equine first phalanx adapts to race-training in the race-fit groups (Raced Control and Contralateral to Fracture) with an increase in volumetric bone mineral density relative to unraced controls. Within the race-trained groups, the Contralateral to Fracture bones had a greater variance of volumetric bone mineral density, suggesting that stress-induced bone adaptation had become more erratic, potentially contributing to the aetiology of sagittal fractures of the first phalanx in the Thoroughbred racehorse.

Keywords: bone adaptation; equine; subchondral bone density.

Figure 1

(A) Superimposed on this pilot scan image of the equine first phalanx are the longitudinal axis of the bone and the coronal slice locations used for the analysis at 25% (dorsal), 50% (central) and 75% (palmar) of the distance from dorsal to palmar along the proximal articular surface of the first phalanx. (B) A microcomputed tomography image through the entire central (50%) slice of P1 at 75‐μm isotropic voxel resolution. Only the proximal portion of this slice was used for analysis.

Figure 2

Three‐dimensional reconstruction of the proximal surface of the equine first phalanx illustrating the 15 sites at which the subchondral bone was analysed. First, the area was divided into dorsal (sites 1–5), central (sites 6–10) and palmer (sites 11–15) coronal planes, from dorsal to palmar. Then five experimental sites were designated from lateral to medial within these zones: lateral fovea (sites 1, 6, 11), lateral ridge of the sagittal groove (2, 7, 12), sagittal groove (3, 8, 13), medial ridge of the sagittal groove (4, 9, 14), medial fovea (5, 10, 15).

Figure 3

Box plots of subchondral bone density (mgHA cm⁻³) at the 15 sites in each cohort. Boxes represent the first and third quartiles separated by the median (horizontal line). Whiskers show the minimum and maximum values. Annotations above each box indicate significant differences of the coefficient of variations between two cohorts (Fligner–Kileen). Annotations below each box indicate significant differences of the medians between two cohorts (Mann–Whitney). Schematics show the positions of the sample sites on the P1 bone, oriented with medial to the right and lateral to the left (as in Fig. 2). UC, unraced; CF, raced with contralateral fracture; RC, raced.

Figure 4

Box plots of subchondral bone thickness (mm) at the 15 sites in each cohort. Boxes represent the first and third quartiles separated by the median (horizontal line). Whiskers show the minimum and maximum values. Annotations above each box indicate significant differences of the coefficient of variations between two cohorts (Fligner–Kileen). Annotations below each box indicate significant differences of the medians between two cohorts (Mann–Whitney). Schematics show the positions of the sample sites on the P1 bone, oriented with medial to the right and lateral to the left (as in Fig. 2). UC, unraced; CF, raced with contralateral fracture; RC, raced.