Keywords:
Multicentre study, Observational, Retrospective, Transplantation, Surgery, CT, Thorax, Bones, Chest
Authors:
F. Nistri1, A. Dell'Amore2, G. Dolci3, P. Ferrigno2, R. Stramare2, E. Quaia4, F. Rea4, C. Giraudo4; 1Treviso/IT, 2Padua/IT, 3Bologna/IT, 4Padova/IT
DOI:
10.26044/ecr2020/C-07654
Methods and materials
Study design
CT scans of all patients who underwent STCA in two tertiary centers (i.e., Thoracic Surgery of the Padova University Hospital and Thoracic Unit of the University Hospital of Bologna) between 2009 and 2017 were collected. The following criteria were also applied: i) availability of at least two CTs after the STCA (CT1 and CT2, respectively); ii) CT examinations without contrast injection; iii) slice thickness of 1-3 mm. In case of multiple post-surgical CT examinations, the first and the last available scans were selected.
Quantitative and qualitative CT assessment
Seven standardized 30mm2 regions of interest (ROIs) were bilaterally drawn on CT1 and CT2 on as many cancellous bone areas along the transplanted sternum, collecting the density values (Hounsfield Unit (Hu)). All measurements were performed on the axial plane applying a constant interval of 15 mm between subsequent ROIs.
A ROI of the same size was also applied on the cancellous bone of the forth dorsal vertebra (D4), at both time points, for each patient, and used as reference.
To avoid artifacts due to surgical metal implants, the regions close to the fixation devices were excluded from the quantitative measurements and then assessed by a modified version of the Stacy score (0=no healing; 1=low-; 2=moderate-; 3=complete healing).
Statistical analysis
Descriptive statistics were applied for demographics and clinical categorical data.
To evaluate bone mineralization, the analysis of the variance with post-hoc Bonferroni correction was applied for the quantitative measurements while the Wilcoxon test was used for the qualitative score (p<0.05).
To evaluate the reliability of the qualitative and the quantitative analyses, the same rater repeated the measurements after two months and the Cohen's kappa (k) and the intraclass correlation coefficient (ICC) were computed. ICC values >.750 were considered excellent [10].