Quantitative ultrasonic evaluation of bone growth in a rabbit tibial distraction model by comparing computed radiography and micro-computed tomography

Abstract

Distraction osteogenesis (DO) has become the most popular orthopaedic treatment for correcting human limb discrepancies and deformities. However, a high post-operational complication rate has been reported during DO treatment. In these cases, delayed union or non-union of the bone callus is commonly due to the distraction protocol with the improper distraction rate or rhythm. Undesirable treatment outcomes can end up with the further remedial surgery such as the internal fixation of the distraction site. This may significantly delay the rehabilitation of DO patients and affect their quality of life. Thus, a closely monitoring of the bone callus growth condition during DO treatment, especially at the early treatment stage, is paramount important for informing the treatment progress and the prognosis of the treatment outcome. Optimization of the clinical distraction protocol or remedial measure can then be initiated at the early DO stage to alleviate the undesirable outcomes. Plain-film radiography is the conventional imaging modality to assess the bone callus growth during DO treatment. However, the detection of the subtle callus bone density changes during distraction period is a challenge to the plain-film radiography because of its sigmoid-shaped detector response. Computed radiography (CR) can provide a more linear detector response with the digital processing capability; however, its efficacy to assess the new bone formation at the early DO stage had not been well-studied yet. Previous in-vitro studies found that acoustic parameters could reflect the bone density, micro-structural and biomechanical properties of trabeculae. On the other hand, clinical studies had shown that ultrasonography could detect the new bone formation during fracture healing. The results suggest that combining ultrasonography and acoustic parameters measurement has the potential to determine the callus growth condition in a qualitative and quantitative manner, respectively during DO treatment. In the present study, CR, ultrasound (US) and micro-computed tomography (uCT) were applied to an animal DO model. A comparative evaluation of the efficacies in using CR and the combined US approach as the callus monitoring tool during DO treatment was conducted. The objectives of this study were: (i) to temporally and regionally compare among CR images, 2D and 3D US images acquired during DO treatment; (ii) to determine the reliability of the regional acoustic parameters measurement and to study the temporal callus growth changes by regional radiographic and acoustic parameters measured during DO treatment; (iii) to predict the bone density, micro-structural and biomechanical properties of the consolidated bone callus by radiographic and acoustic parameters measured during DO treatment. Seven male New Zealand white rabbits were operated with the tibial DO surgical procedures. After 7-day latency, right tibiae were distracted at 1.0 mm/day, up to 12.0 mm, and followed with a four-week consolidation. Callus growth condition was evaluated by CR and US at a 3-day and weekly interval during distraction and consolidation periods, respectively. Besides the images taken by CR and US during the treatment, the calibrated aluminum step-wedge thickness (AlTh) , as a surrogate measurement of the bone mineral density (BMD), and acoustic parameters including speed of sound (SoScal), integrated reflection coefficient (IRC), integrated backscattering (IBScorr) and integrated attenuation (IA) were acquired from CR and US images, respectively. These parameters were measured at the distal, middle and proximal regions of the callus. At the end of the four-week consolidation, distracted tibiae were dissected and scanned by uCT for evaluating the bone density, micro-structural and biomechanical properties of the consolidated callus. Findings showed that the callus density changes were not demonstrated by CR during distraction period. However, the callus echogenicity shown in the US images increased regionally and temporally at the same period. 2D US showed that echogenic foci in the distraction gap initially occurred as the speckled appearance. Later, they developed as an echogenic boundary at the anterior superficial surface of the distal and proximal callus regions but not at the middle one. 3D US even exhibited an undulating pattern in the callus echogenicity across the distracting bone callus. Such observation was further supported by the similar patterns of SoScal, IRC and IA measured across the distraction gap during distraction period. These findings indicated that the continuous callus mineralization and its differential pattern had been initiated during distraction period and could be detected by US, but not by CR. During consolidation period, however, only a hyper-echogenic line or surface could be observed at all three callus regions by 2D and 3D US, respectively. Instead, CR could provide the important radiographic characteristics of the bone callus growth such as the callus mineralization, bridging and corticalization at the late DO stage. Quantitatively, regional AlTh did not demonstrate significant trend changes (p>0.05) at the early DO stage. Instead, there were significant trend increases of regional SoScal, IRC and IA at this stage (p<0.05), and the measurement of these regional parameters were reliable (most ICC values> 0.7). This indicated that acoustic parameters were more sensitive to detect the growth changes of the distracting callus than AlTh. They might further help in studying the effectiveness of the pharmaceutical and biophysical intervention on the bone healing at the early DO stage. Furthermore, these acoustic parameters measured at this stage could significantly contribute up to 65%, 66% and 43% variances of the bone density, micro-structural and biomechanical properties of the consolidated callus, respectively (p<0.05). However, these consolidated callus properties were weakly predicted by AlTh measured at the same stage (r²<0.09, p>0.05). This implied that the measurement of the callus acoustic parameters measured at the early DO stage could provide the prognosis of the treatment outcome, which was evident in their predictions to the uCT-derived consolidated callus properties. But this was not the case in AlTh measured during the same period. The predictive values of the acoustic parameters measured at the late DO stage showed relatively weak predictions to the consolidated callus properties (adjusted R²<0.42, p<0.05). At the same period, AlTh not just showed the significant increases at all three callus regions (p<0.05), but their values measured at the end of the consolidation period also predicted strongly to the bone density (r²>0.70, p<0.01), micro-structural (r²>0.70, p<0.01) and biomechanical properties (r²>0.75, p<0.01) of the consolidated callus. In summary, the present study was the first imaging study to compare CR and US in evaluating the bone callus growth in a rabbit tibial distraction model. Results indicated that the combined US approach was not just sensitive to the early callus growth changes, but it also provided the early prediction of the consolidated callus properties. This might inform clinicians about the appropriateness of the patient's distraction protocol. Such US approach could act as an effective qualitative and quantitative assessment of the early callus growth condition. Also, it could be practiced frequently to reduce the use of X-ray at the early DO stage. Nevertheless, clinical study should be contemplated so as to verify the potential value of measuring the callus acoustic properties at the early DO stage as evident in the present study.