Epidemiology
24 (85.7%) patients were male and 4 (14.3%) were female. Average age at time of diagnosis was 66.3 years (standard deviation (SD) 8.2 years), varying from 51 to 78 years. In 16 patients (57.1%), prior exposure to asbestos was confirmed. For 1 case (3.6%), asbestos-exposure could be ruled out. In 11 cases (39.3%), no reliable information in this regard was available.
In 23 patients, data for the histological subtype of MPM was available showing epitheloid pleural mesothelioma in 16 patients (57.1%) and non-epitheloid pleural mesothelioma in 7 patients (25.0%).
Subjective analysis
In all 28 cases and for both readers, MPM was better visible on late phase compared with arterial phase images. An example for a case, in which the tumor lesion showed considerably better conspicuity in late phase compared to early phase is presented in Fig. 2.
Objective analysis
HUearlyLes was in average 56.7 HU (SD 18.1 HU), HUlateLes in average 75.4 HU (SD 22.9 HU). Hence, HUdiffLes was 18.6 HU (SD 10.1 HU) with the difference being statistically significant (p < 0.001).
Density of the surrounding soft tissue was − 17.4 HU (SD 47.5 HU) in arterial phase (HUearlyAdj) and − 17.3 HU (SD 12.1 HU) in late phase (HUlateAdj), resulting in an average difference (HUdiffAdj) of 0.1 HU (SD 5.6 HU) between arterial phase and late phase with p = 0.92.
Average difference of contrast between tumor and adjacent soft tissue in arterial phase and late phase (HUdiffContrast) was 18.5 HU (SD 10.6 HU) with p < 0.001. The results are summarized in Table 1.
Table 1
Results of the objective analysis
Parameter | Value (SD) | Significance (t-test) |
HUearlyLes | 56.7 HU (18.1 HU) | |
HUlateLes | 75.4 HU (22.9 HU) | |
HUearlyAdj | -17.4 HU (47.5 HU) | |
HUlateAdj | -17.3 HU (12.1 HU) | |
Parameter differences | | |
HUdiffLes | 18.6 HU (10.1 HU) | p < 0.001 |
HUdiffAdj | 0.1 HU (5.6 HU) | p = 0.92 |
HUdiffContrast | 18.5 HU (10.6 HU) | p < 0.001 |
Technical and epidemiological data were as follows:
Tube voltage varied between 100 kV and 140 kV in arterial phase (mean 121.8 kV, SD 8.6 kV) and 100 kV and 130 kV in late phase (mean 117.8 kV, SD 7.3 kV). In 22 cases (79%), tube voltage was identical in arterial phase and late phase. In 6 cases (21%), tube voltage was higher in arterial phase than in late phase.
For one patient, no conclusive data for the tube-current-time-product was available. Mean tube-current-time-product for arterial phase was 104.7 mAs (SD 54.6 mAs) with a range from 13 to 228 mAs. Mean tube-current-time-product for late phase was 148.8 mAs (SD 88.6 mAs), ranging from 13 to 385 mAs.
Slice thickness varied between 1.25 mm and 8.00 mm with an average of 5.30 mm (SD 1.80 mm) in arterial phase scans and 4.50 mm (SD 1.20 mm) in late phase scans. Slice thickness was 1.25 mm, 2.00 mm and 2.50 mm in one patient each in arterial as well as in late phase. In two cases, slice thickness was 3.75 mm in arterial and late phase. In 15 cases, slice thickness in arterial phase was 5.00 mm, in two cases 6.00 mm and in 6 cases 8.00 mm. In 22 cases, slice thickness in late phase was 5 mm, in one case 6 mm. In 20 cases (71.4 %), slice thickness was identical between arterial phase and late phase. In 8 cases (28.6 %), slice thickness was thinner in arterial phase.
In 23 patients (82.1 %), the histological subtype of MPM was available. Histology showed an epitheloid type MPM in 16 patients (69.6 %) and a biphasic, sarcomatoid or rare subtype of MPM in 7 patients (30.4 %).
For one patient, no information about the therapeutic regime was available. 14 Patients (51.9 %) had been treated with chemotherapy prior to CT scans. 13 patients (48.1 %) had not received any chemotherapy prior to imaging.
Multiple linear regression analysis using a general linear model with HUdiffContrast as dependant variable and contrast phase, tube voltage, tube-current-time-product and slice thickness as quantitative predictors was performed to test possible confounding because of technical parameters. Contrast phase (p < 0.001) and tube voltage (p < 0.001) proved to be highly significant independent predictors for tumor contrast. In a second and third general linear model confounding because of epidemiological parameters was tested with contrast phase as quantitative predictor and tumor entity as well as chemotherapy as two-level factors. Both epidemiological parameters were no significant predictor for tumor contrast. Results of the multiple regression analysis are summarized in Table 2.
Table 2
Results of the multiple regression analysis using general linear models.
| adjusted mean (95% CI) | significance |
contrast arterial phase (HU) | 76.6 (58.9, 94.2) | |
contrast late phase (HU) | 92.7 (75.1, 110.3) | |
difference in contrast (HU) | 16.1 (14.0, 18.2) | p < 0.001 |
| correlation coefficient (95% CI) | |
tube voltage (kV) | -1.05 (-1.30, -0.80) | p < 0.001 |
tube-current-time-product (mAs) | -0.014 (-0.045, 0.017) | p = 0.369 |
slice thickness (mm) | 1.35 (-0.06, 2.75) | p = 0.059 |
epitheloid MPM (n = 23) | -29.5 (-64.2, 5.16) | p = 0.091 |
chemotherapy prior to CT (n = 27) | 14.8 (-18.2, 47.7) | p = 0.366 |