Skip to main content
. 2021 Mar 5;7:25. doi: 10.1038/s41523-021-00233-9

Table 3.

Statistical tests evaluating the predictive value of MRI in detecting nodal status after neoadjuvant therapy overall and by clinical node status.

Overall Cohort
Neoadjuvant chemotherapy (n = 46) Neoadjuvant endocrine therapy (n = 33) P-Value
Sensitivity 37.9% 66.7% 0.0116
Specificity 70.6% 66.7% 0.7117
Negative predictive value 40.0% 70.6% 0.0072
Positive predictive value 68.8% 62.5% 0.5594
Overall accuracy 50.0% 66.7% 0.1393
Neoadjuvant chemotherapy Cohort
Clinically node negative (n = 24) Clinically node positive (n = 22) P-Value
Sensitivity 20.0% 47.4% 0.0485
Specificity 78.6% 33.3% 0.0019
Negative predictive value 57.9% 9.1% 0.0005
Positive predictive value 40.0% 81.2% 0.0044
Overall accuracy 54.2% 45.5% 0.5555
Neoadjuvant endocrine therapy cohort
Clinically node negative (n = 25) Clinically node positive (n = 8) P-Value
Sensitivity 42.9% 87.5% 0.0277
Specificity 66.7% N/A1
Negative predictive value 75.0% 0.0% 0.0002
Positive predictive value 33.3% 100.0% 0.001
Overall accuracy 60.0% N/Aa

aThere were no clinically node positive patients who received endocrine therapy and were pathologically node negative; as such, specificity and accuracy of post-treatment MRI cannot be calculated for this subset.