Table 2.
Parameter | Cancers* | Benign nodules * | Total* | P value** |
---|---|---|---|---|
Number | 65 | 675 | 740 | |
Participants | 64 | 549 | 613 | |
Age in years | 61 (52–75) | 58 (50–71) | 58 (50–75) | 0.655 |
Sex Male Female |
35 (55%) 29 (45%) |
292 (53%) 257 (47%) |
327 (53%) 286 (47%) |
0.820 |
Family history of lung cancer | 17 (27%) | 90 (16%) | 107 (17%) | 0.043 |
Emphysema | 47 (73%) | 367 (67%) | 414 (68%) | 0.287 |
Nodule size in mm: Dlongest-C |
Median: 13.9 16.5 (3.3–124.8) |
Median: 6.1 7.6 (1.6–104.4) |
8.4 (1.6–124.8) | <0.001 |
Nodule size in mm: Dmean3D |
Median: 10.2 12.5 (2.7–84.8) |
Median: 4.9 5.8 (1.3–62.3) |
6.4 (1.3–84.8) | <0.001 |
Nodule size in mm: DmeanAxial |
Median: 12.1 13.8 (2.9–95.0) |
Median: 5.4 6.4 (1.2–95.5) |
7.1 (1.2–95.5) | <0.001 |
Nodule type Solid Part-solid Non-solid |
43 (66%) 15 (23%) 7 (11%) |
590 (87%) 19 (3%) 66 (10%) |
633 (85%) 34 (5%) 73 (10%) |
<0.001 |
Perifissural | 0 (0%) | 55 (8%) | 55 (7%) | 0.017 |
Calcified | 0 (0%) | 96 (14%) | 96 (13%) | 0.001 |
Nodule count | 0.3 (0–4) | 0.5 (0–5) | 0.5 (0–5) | 0.017 |
Nodule location Upper lobe |
38 (58%) | 327 (48%) | 365 (49%) | 0.123 |
Spiculation | 18 (28%) | 10 (1%) | 28 (4%) | <0.001 |
* Percentages or ranges are in parentheses
** P-value for benign nodules versus cancers. A p-value < 0.05 indicates a significant difference. Family history of lung cancer pertained to parents or siblings. Presence of emphysema was dichotomous and not corresponding to the degree of emphysema. Nodule size was measured as the longest diameter. Nodule count pertained to the number of additional nodules in the scan. Spiculation was defined as reticular markings of tissue density with elements of circular symmetry centered around a nodule
D longest-C longest diameter on axial sections, derived from computerized semi-automated segmentations, D mean3D mean diameter based on volumetric information, derived from computerized semi-automated segmentations, D meanAxial mean of longest and perpendicular diameter on axial sections, derived from computerized semi-automated segmentations