Validation of a Radiography-Based Quantification Designed to Longitudinally Monitor Soft Tissue Calcification in Skeletal Muscle

INTRODUCTION Soft tissue calcification, including both dystrophic calcification and heterotopic ossification, may occur following injury. These lesions have variable fates as they are either resorbed or persist. Persistent soft tissue calcification may result in chronic inflammation and/or loss of function of that soft tissue. The molecular mechanisms that result in the development and maturation of calcifications are uncertain. As a result, directed therapies that prevent or resorb soft tissue calcifications remain largely unsuccessful. Animal models of post-traumatic soft tissue calcification that allow for cost-effective, serial analysis of an individual animal over time are necessary to derive and test novel therapies. We have determined that a cardiotoxin-induced injury of the muscles in the posterior compartment of the lower extremity represents a useful model in which soft tissue calcification develops remote from adjacent bones, thereby allowing for serial analysis by plain radiography. The purpose of the study was to design and validate a method for quantifying soft tissue calcifications in mice longitudinally using plain radiographic techniques and an ordinal scoring system. METHODS Muscle injury was induced by injecting cardiotoxin into the posterior compartment of the lower extremity in mice susceptible to developing soft tissue calcification. Seven days following injury, radiographs were obtained under anesthesia. Multiple researchers applied methods designed to standardize post-image processing of digital radiographs (N = 4) and quantify soft tissue calcification (N = 6) in these images using an ordinal scoring system. Inter- and intra-observer agreement for both post-image processing and the scoring system used was assessed using weighted kappa statistics. Soft tissue calcification quantifications by the ordinal scale were compared to mineral volume measurements (threshold 450.7mgHA/cm3) determined by μCT. Finally, sample-size calculations necessary to discriminate between a 25%, 50%, 75%, and 100% difference in STiCSS score 7 days following burn/CTX induced muscle injury were determined. RESULTS Precision analysis demonstrated substantial to good agreement for both post-image processing (κ = 0.73 to 0.90) and scoring (κ = 0.88 to 0.93), with low inter- and intra-observer variability. Additionally, there was a strong correlation in quantification of soft tissue calcification between the ordinal system and by mineral volume quantification by μCT (Spearman r = 0.83 to 0.89). The ordinal scoring system reliably quantified soft tissue calcification in a burn/CTX-induced soft tissue calcification model compared to non-injured controls (Mann-Whitney rank test: P = 0.0002, ***). Sample size calculations revealed that 6 mice per group would be required to detect a 50% difference in STiCSS score with a power of 0.8. Finally, the STiCSS was demonstrated to reliably quantify soft tissue calcification [dystrophic calcification and heterotopic ossification] by radiographic analysis, independent of the histopathological state of the mineralization. CONCLUSIONS Radiographic analysis can discriminate muscle injury-induced soft tissue calcification from adjacent bone and follow its clinical course over time without requiring the sacrifice of the animal. While the STiCSS cannot identify the specific type of soft tissue calcification present, it is still a useful and valid method by which to quantify the degree of soft tissue calcification. This methodology allows for longitudinal measurements of soft tissue calcification in a single animal, which is relatively less expensive, less time-consuming, and exposes the animal to less radiation than in vivo μCT. Therefore, this high-throughput, longitudinal analytic method for quantifying soft tissue calcification is a viable alternative for the study of soft tissue calcification.