Research Article

An External-Validated Prediction Model to Predict Lung Metastasis among Osteosarcoma: A Multicenter Analysis Based on Machine Learning

Table 1

Baseline data table of the training group and the validation group.

LevelOverall (N = 1201)Multicenter (validation group, N = 107)SEER (training group, N = 1094)

Race (%)Black163 (13.57)0 (0.00)163 (14.90)<0.0001
Other216 (17.99)107 (100.00)109 (9.96)
White822 (68.44)0 (0.00)822 (75.14)

Age (median [IQR])NA21.000 [13.000, 53.000]18.000 [13.000, 48.500]22.000 [13.000, 53.750]0.3895

Sex (%)Female549 (45.71)50 (46.73)499 (45.61)0.9048
Male652 (54.29)57 (53.27)595 (54.39)

Primary site (%)Axis bone313 (26.06)25 (23.36)288 (26.33)0.374
Limb bone787 (65.53)76 (71.03)711 (64.99)
Other101 (8.41)6 (5.61)95 (8.68)

Grade (%)Moderately differentiated41 (3.41)0 (0.00)41 (3.75)0.0934
Poorly differentiated294 (24.48)22 (20.56)272 (24.86)
Undifferentiated; anaplastic553 (46.04)49 (45.79)504 (46.07)
Unknown286 (23.81)34 (31.78)252 (23.03)
Well differentiated27 (2.25)2 (1.87)25 (2.29)

Laterality (%)Left514 (42.80)40 (37.38)474 (43.33)0.0524
Not a paired site161 (13.41)9 (8.41)152 (13.89)
Right526 (43.80)58 (54.21)468 (42.78)

T (%)T1420 (34.97)38 (35.51)382 (34.92)0.1914
T2562 (46.79)44 (41.12)518 (47.35)
T340 (3.33)7 (6.54)33 (3.02)
TX179 (14.90)18 (16.82)161 (14.72)

N (%)N01088 (90.59)91 (85.05)997 (91.13)0.12
N136 (3.00)5 (4.67)31 (2.83)
NX77 (6.41)11 (10.28)66 (6.03)

Surgery (%)No225 (18.73)23 (21.50)202 (18.46)0.5241
Yes976 (81.27)84 (78.50)892 (81.54)

Radiation (%)No1054 (87.76)100 (93.46)954 (87.20)0.0837
Yes147 (12.24)7 (6.54)140 (12.80)

Chemotherapy (%)No236 (19.65)10 (9.35)226 (20.66)0.0073
Yes965 (80.35)97 (90.65)868 (79.34)

Bone metastases (%)No1144 (95.25)102 (95.33)1042 (95.25)1
Yes57 (4.75)5 (4.67)52 (4.75)

Lung metastases (%)No987 (82.18)87 (81.31)900 (82.27)0.9085
Yes214 (17.82)20 (18.69)194 (17.73)

Times (median [IQR])NA24.000 [12.000, 48.000]23.000 [11.000, 51.500]24.000 [12.000, 47.000]0.9754

Abbreviation: SEER, Surveillance Epidemiology and End Results; IQR, interquartile range; T, tumor; N, lymph node.