Research Article

A Genomic-Clinicopathologic Nomogram for the Prediction of Lymph Node Invasion in Prostate Cancer

Table 1

The clinicopathological characteristics of primary patients in SMOTE-balanced training and validation sets.

Number of patients (%) value
CharacteristicSMOTE-balanced training set (n = 291)SMOTE-balanced validation set (n = 135)

Age, years
 Mean [median]61.50 [62]61.15 [62]0.640a
 IQR57–6656–66

PSA level (most recent), ng/mL
 Mean [median]1.17 [0.10]1.00 [0.10]0.698a
 IQR0.03–0.180.03–0.17

Clinical stage
 cT193 (31.96)44 (32.59)0.294b
 cT2104 (35.74)55 (40.74)
 cT330 (10.31)19 (14.07)
 cT41 (0.34)0 (0)
 NA63 (21.65)17 (12.59)

Pathologic stage
 pT298 (33.68)47 (34.81)0.534b
 pT3181 (62.20)85 (62.96)
 pT410 (3.44)1 (0.74)
 NA2 (0.69)2 (1.48)

Gleason grade
 <717 (5.84)7 (5.19)0.739b
  = 7140 (48.11)69 (51.11)
 >7134 (46.05)59 (43.70)

Surgical margin resection status
 R0178 (61.17)89 (65.93)0.283b
 R197 (33.33)36 (26.67)
 R22 (0.69)2 (1.48)
 NA14 (4.81)8 (5.93)

Number of dissected lymph nodes
 Mean [median]12.15 [10]10.75 [9]0.170a
 IQR5–165–15

Number of positive lymph nodes
 Mean [median]0.51 [0]0.35 [0]0.279a
 IQR00

N stage
N0235 (80.76)112 (38.49)0.297c
N156 (19.24)23 (7.90)

IQR: interquartile range; NA: not available; PSA: prostate-specific antigen; SMOTE: synthetic minority oversampling technique; at-test; bMann–Whitney U test; cchi-square test.