Identifying Eligibility for Axillary LND in Breast Cancer
Identifying Eligibility for Axillary LND in Breast Cancer
On the whole, SLNs of 840 patients (77.8%) were negative with the OSNA assay and 240 (22.2%) were positive. Among these, 123 (51.2%) had micrometastasis and 117 (48.8%) macrometastasis; 194 patients underwent ALND, including all the patients with macrometastasis and 77 (77/123; 62.6%) patients with micrometastasis. In patients undergoing ALND, 125 (125/194; 64.4%) did not have any additional metastatic lymph nodes in ALN, whereas one or more positive ALNs were detected in 69 patients (69/194; 35.6%) (figure 1).
First, we calculated the mean value of CK19 mRNA copy number in all the positive SLNs of the patients who underwent ALND, independently to the distinction between micrometastasis and macrometastasis. The CK19 mRNA copy number ranged from 250 to 7 400 000; the mean value of 320 000 CK19 mRNA copies was used as cut-off in order to distinguish patients with a high and low risk of metastasis in ALN. In the group of patients with a CK19 mRNA copy number higher than 320 000 in SLN, 55.2% had one or more positive ALN (true-positive), whereas in patients with a CK19 copy number less than 320 000 in SLN, 32.2% had one or more positive ALN (false-negative). Although the difference was statistically significant (p=0.021, Table 2), the number of false-negative cases (53/165, 32%) was too high, probably related to the wide range of CK19 mRNA copy numbers and therefore not useful in clinical practice. In order to overcome the limitation of this result, we used the ROC curve analysis to search for an optimal cut-off value with the highest sensitivity and specificity (figure 2). Across various cut-off points, Youden's index maximised the difference between sensitivity and specificity and between real-positive and false-positive subjects; thus, the optimal cut-off value was calculated. The total sum of sensitivity and 1-specificity of the single copy number cut-offs was represented by the ROC analysis, with a higher area under the curve (AUC) indicating the best cut-off for our objective. On the basis of ROC analysis, with an AUC of 0.69, the value for a cut-off was calculated to 7700 CK19 mRNA copies. Therefore, two groups were identified. The first group included 86 patients having ≤7700 of CK19 mRNA copies: 71 (82.6%) were ALN-negative and 15 (17.4%) ALN-positive, with one or more positive lymph nodes (false-negative cases, Table 2). The main pathological findings of these 15 patients are reported in Table 3. Except for patient 2, the CK19 mRNA copy number was always under 5000 copies, meaning that these cases would have been considered false-negative also using the classical cut-off of Tsujimoto et al. Interestingly, in four cases (n# 2, 7, 8 and 12), more than two additional ALNs were positive (pN2, Table 3). The second group included 108 patients with more than 7700 CK19 mRNA copies; 54 (true-positive, 50%) were ALN-positive (p=0.001, Table 2). The new cut-off showed 78% sensitivity and 57% specificity in differentiating patients with negative ALN or with one or more metastatic lymph nodes (Table 4). Positive and negative predictive values of this new cut-off were 50% and 83%, respectively.
(Enlarge Image)
Figure 2.
The new best cut-off of cytokeratin 19 mRNA copy number obtained by receiver operative characteristics curve analysis.
Finally, our case series was further analysed by using the classical cut-off of 5000 CK19 mRNA copy number proposed by Tsujimoto et al. As shown in Table 4, only one patient with CK19 mRNA copy number between 5000 and 7700 had one or more positive ALN, whereas eight patients had negative ALN.
Results
OSNA Assay Results and Management of the Axilla
On the whole, SLNs of 840 patients (77.8%) were negative with the OSNA assay and 240 (22.2%) were positive. Among these, 123 (51.2%) had micrometastasis and 117 (48.8%) macrometastasis; 194 patients underwent ALND, including all the patients with macrometastasis and 77 (77/123; 62.6%) patients with micrometastasis. In patients undergoing ALND, 125 (125/194; 64.4%) did not have any additional metastatic lymph nodes in ALN, whereas one or more positive ALNs were detected in 69 patients (69/194; 35.6%) (figure 1).
Correlation Between CK19 mRNA Copy Number in Positive SLN and the Risk of Axillary Involvement
First, we calculated the mean value of CK19 mRNA copy number in all the positive SLNs of the patients who underwent ALND, independently to the distinction between micrometastasis and macrometastasis. The CK19 mRNA copy number ranged from 250 to 7 400 000; the mean value of 320 000 CK19 mRNA copies was used as cut-off in order to distinguish patients with a high and low risk of metastasis in ALN. In the group of patients with a CK19 mRNA copy number higher than 320 000 in SLN, 55.2% had one or more positive ALN (true-positive), whereas in patients with a CK19 copy number less than 320 000 in SLN, 32.2% had one or more positive ALN (false-negative). Although the difference was statistically significant (p=0.021, Table 2), the number of false-negative cases (53/165, 32%) was too high, probably related to the wide range of CK19 mRNA copy numbers and therefore not useful in clinical practice. In order to overcome the limitation of this result, we used the ROC curve analysis to search for an optimal cut-off value with the highest sensitivity and specificity (figure 2). Across various cut-off points, Youden's index maximised the difference between sensitivity and specificity and between real-positive and false-positive subjects; thus, the optimal cut-off value was calculated. The total sum of sensitivity and 1-specificity of the single copy number cut-offs was represented by the ROC analysis, with a higher area under the curve (AUC) indicating the best cut-off for our objective. On the basis of ROC analysis, with an AUC of 0.69, the value for a cut-off was calculated to 7700 CK19 mRNA copies. Therefore, two groups were identified. The first group included 86 patients having ≤7700 of CK19 mRNA copies: 71 (82.6%) were ALN-negative and 15 (17.4%) ALN-positive, with one or more positive lymph nodes (false-negative cases, Table 2). The main pathological findings of these 15 patients are reported in Table 3. Except for patient 2, the CK19 mRNA copy number was always under 5000 copies, meaning that these cases would have been considered false-negative also using the classical cut-off of Tsujimoto et al. Interestingly, in four cases (n# 2, 7, 8 and 12), more than two additional ALNs were positive (pN2, Table 3). The second group included 108 patients with more than 7700 CK19 mRNA copies; 54 (true-positive, 50%) were ALN-positive (p=0.001, Table 2). The new cut-off showed 78% sensitivity and 57% specificity in differentiating patients with negative ALN or with one or more metastatic lymph nodes (Table 4). Positive and negative predictive values of this new cut-off were 50% and 83%, respectively.
(Enlarge Image)
Figure 2.
The new best cut-off of cytokeratin 19 mRNA copy number obtained by receiver operative characteristics curve analysis.
Evaluation of the Comparison Between the Traditional Cut-off and the New Cut-off Value
Finally, our case series was further analysed by using the classical cut-off of 5000 CK19 mRNA copy number proposed by Tsujimoto et al. As shown in Table 4, only one patient with CK19 mRNA copy number between 5000 and 7700 had one or more positive ALN, whereas eight patients had negative ALN.
Source...