Research Article

Development and External Validation of Machine Learning-Based Models for Predicting Lung Metastasis in Kidney Cancer: A Large Population-Based Study

Table 1

Clinicopathologic characteristics.

VariablesOverall (N = 52,714)Non-LM (N = 50,096)LM (N = 2,618)

Age
Mean (SD)59.7 (12.3)59.6 (12.4)61.3 (11.0)

Sex
Female19,555 (37.1%)18,805 (37.5%)750 (28.6%)
Male33,159 (62.9%)31,291 (62.5%)1,868 (71.4%)

Laterality
Left25,973 (48.8%)24,395 (48.7%)1,330 (50.3%)
Right26,741 (50.3%)25,233 (50.4%)1,288 (48.7%)

Race
Asian3,625 (6.9%)3,408 (6.8%)217 (8.3%)
Black5,701 (10.8%)5,523 (11.0%)178 (6.8%)
Others1,031 (2.0%)992 (2.0%)39 (1.5%)
White42,357 (80.4%)40,173 (80.2%)2,184 (83.4%)

T_stage
T135,124 (66.6%)34,842 (69.6%)282 (10.8%)
T25,646 (10.7%)5,189 (10.4%)457 (17.5%)
T310,959 (20.8%)9,465 (18.9%)1,494 (57.1%)
T4985 (1.9%)600 (1.2%)385 (14.7%)

N_stage
N050,392 (95.6%)48,646 (97.1%)1,746 (66.7%)
N12,322 (4.4%)1,450 (2.9%)872 (33.3%)

Tumor_size
Mean (SD)53.6 (37.3)51.1 (34.7)101 (51.0)

Histology
8120182 (0.3%)147 (0.3%)35 (1.3%)
82551,344 (2.5%)1,219 (2.4%)125 (4.8%)
82606,029 (11.4%)5,931 (11.8%)98 (3.7%)
831035,878 (68.1%)34,176 (68.2%)1,702 (65.0%)
83125,736 (10.9%)5,351 (10.7%)385 (14.7%)
83172,210 (4.2%)2,190 (4.4%)20 (0.8%)
Others1,335 (2.5%)1,082 (2.2%)253 (9.7%)

Grade
I5,923 (11.2%)5,862 (11.7%)61 (2.3%)
II26,761 (50.8%)26,280 (52.5%)481 (18.4%)
III15,484 (29.4%)14,401 (28.7%)1,083 (41.4%)
IV4,546 (8.6%)3,553 (7.1%)993 (37.9%)

LM, lung metastasis; 8120, transitional cell carcinoma; 8255, adenocarcinoma with mixed subtypes; 8260, papillary adenocarcinoma; 8310, clear cell adenocarcinoma; 8312, renal cell carcinoma; 8317, chromophobe renal carcinoma.