Skip to main content

Table 4 Comparison of diagnostic performance for detecting macrometastasis

From: Prediction of macrometastasis in axillary lymph nodes of patients with invasive breast cancer and the utility of the SUV lymph node/tumor ratio using FDG-PET/CT

 

NT ratio

SUV-LN

Combination

All cases (n = 171)

Cutoff 0.288

Cutoff : 1.00

 

 Sensitivity

58.2 (46.4-69.2)

78.2 (67.3-86.7)

85.5 (74.7-92.8)

 Specificity

59.5 (50.0-68.5)

80.2 (75.0-84.2)

50.0 (44.9-53.5)

 PPV

40.5 (29.6-52.2)

65.2 (56.1-72.2)

44.8 ( 39.1-48.6)

 NPV

75.0 (64.9-83.4)

88.3 (83.0-92.5)

87.9 (78.9-94.0)

 Accuracy

59.1 (51.5-66.2)

81.9 (75–87.3)

61.4 (54.5-66.1)

SUV-T ≥ 2.5 (n = 118)

Cutoff 0.199

Cutoff : 1.00

 

 Sensitivity

68.9 (56.4-79.6)

75.6 (63.5-85.1)

84.4 (72.7-92.6)

 Specificity

65.8 (68.1-72.4)

76.7 (69.3-82.6)

54.8 (47.5-59.8)

 PPV

55.4 (45.3-64.0)

65.4 (55.0-73.7)

53.5 (46.1-58.7)

 NPV

77.4 (68.4-85.2)

83.6 (75.5-90.0)

85.1 (73.8-92.9)

 Accuracy

66.9 (57.4-75.1)

76.3 (67.1-83.5)

66.1 (57.7-72.3)

  1. The diagnostic performance of the NT ratio, SUV-LN, and both techniques in combination is indicated. Cutoff values for each modality were determined based on the sensitivity and specificity. In the whole patient population (n = 171), the cutoffs for the NT ratio and SUV-LN were 0.288 and 1.00, respectively; these values were compared using the McNemar test (P = 0.013). In patients with an SUV-T ≥2.5 (n = 118), the cutoffs for the NT ratio and SUV-LN were 0.199 and 1.00, respectively, with no significant difference noted.