The Association between Sentinel Lymph Node Metastasis and Ki-67 Labeling Index

Background: The purpose of this study was to elucidate the association between sentinel lymph node (SLN) metastasis and Ki67 labeling index and to elucidate whether Ki-67 was useful or not for prediction of SLN metastasis in breast cancer. Methods: We identified 343 invasive breast cancer patients with sentinel lymph node biopsy (SLNB) from 2003 to 2012. The association between SLN status and clinicopathological features, molecular subtypes and Ki-67 labeling index were evaluated. Results: SLN metastasis was detected in 79 patients (23.0%). SLN metastasis was significantly associated with clinical T-stage (p = 0.0003), lymphovascular involvement (LVI) (p < 0.0001). Ki-67 labeling index of primary tumor was significantly lower in SLN positive patients (p = 0.0331), and Ki-67 cutoff point of 7.5% was useful for dividing SLN positive from negative (p = 0.0197). Conclusion: Low value of Ki-67 labeling index, in addition to progression of clinical T-stage and presence of LVI, is significantly associated with SLN metastasis, and it seems to be useful to consider Ki-67 labeling index for SLN metastasis prediction.

Recent clinical trials have suggested no outcome difference in patients with positive SLN between ALND versus no further axillary surgery, raising doubts on the role of SLNB itself [23,24].More recently, a new trial comparing SLNB versus observation when axillary ultra-sound is negative in patients with small breast cancer has been ongoing [25].Therefore, it seems to be very important to predict axillary node status before SLNB.
On the other hand, Ki-67, a nucleus protein, is an immunohistochemical proliferation marker in many types of cancer and has been widely studied including breast cancer, and independently improved the prediction of treatment response and prognosis [26][27][28].Ki-67 has been used to divide luminal A(ER/PR positive, HER negative and Ki-67 < 14%) and HER2 negative luminal B (ER/PR positive, HER negative and Ki-67  14%) in molecular subtype classification of St Gallen Consensus [29].However, the association between Ki-67 expression and SLN metastasis in breast cancer has not been clarified yet.
The purpose of this study was to examine the association between SLN metastasis and Ki-67, and to elucidate whether Ki-67 be useful or not for precidction of SLN metastasis in breast cancer.
SNB at Niigata University Hospital between January 2003 and April 2012 were entered in to the present study.This study was a retrospective chart review, and a total of 343 patients were enrolled into the analysis.On the other hand, Ki-67 labeling index was clinically used in our hospital since 2010 because Ki-67 has been recognized as a useful factor at St. Gallen 2009.Therefore, our data of Ki-67 labeling index in the present study was obtained from a total of 117 patients since 2010.

Clinicopathological Assessment
Immunohistichemical (IHC) ER and PR status was assessed, respectively.Tumors were deemed positive for each receptor if at least 10% of the invasive tumor cells in a section exhibited nuclear staining.HER 2 expression was also examined by IHC, however, gene amplification assay with fluorescence in situ hybridaization (FISH) method was introduced in case with difficult to decide positive or negative by IHC.And Ki-67 leveling index was also examined by IHC, and the results were expressed as percentage of tumor cells stained by the antibody as described prevuoisly [30].Lymphovascular involvement (LVI) and Histologic grading were assessed with hematoxylin eosin staining, and histlogic grading was defined according to Scarff-Bloom-Richardson system [31].SLN metastasis was judged by intraoperative frozen section, however, re-examened with fixed section and postopreratively re-judged.The staging of breast cancer was defined by TNM classification as proposed by the American Joint Committee on Cancer (AJCC).All theses IHC judgement was performed by several well-trained pathologists.According to the results of ER, PR, HER 2 status, patients were grouped into 4 subgroup: ER positive or PR positive, HER 2 negative was categorized as luminal A; ER positive or PR positive, HER 2 positive was categorized as luminal B; ER negative, PR negative and HER 2 positive was categorized as HER 2; ER negative, PR negative and HER 2 negative was categorized as triple negative.

Statistical Analysis
The relationship between SLN metastasis and clinicopahological factors including ER and/or PgR status, Her 2 status, molecular subtypes and Ki-67 leveling index were examined.Statistical analyses were performed using Mann Whitney's U test and Chi-square test, and multivariate analysis were performed using the logistic regression model.The diagnostic accuracy of Ki-67 labeling index was assessed by receiving operating characteristic (ROC) analysis.The area under the ROC curve (AUC) was to measure model disclimination.The AUC can range from 0.5 (which indicates a test with no information) to 1.0 (which indicates a perfect test).The statistical signifi-cance was defined as P < 0.05.

Patients Characteristics and Molecular Subtypes
A total of 343 patients were entered during the period.All patients were female and SLN metastasis was detected in 79 patients (23.0%).The mean age of all patients was 55.9 years old, and there was no significant association between age and SLN metastasis (P = 0.0621) (Table 1).There were some deficits of data in histological grade (20 patients), LVI (4 patients), HER 2 status (2 patients) and molecular subtypes (2 patients) because of no description in the reports of pathological diagnosis.Clinical T-stage was associated with SN metastasis, and SLN metastasis was more frequent in larger tumor (P = 0.0003).LVI was also strongly associated with SLN metastasis (P < 0.0001).
There was no significant correlation between SLN metastasis and ER and/orPgR status, and HER 2 status.The percentage distribution of molecular subtypes among 341 patients in this study was as follows: luminal A in 59.2%, luminal B in 23.5%, HER 2 in 7.0%, and triple negative in 10.3%.There was also no significant association between SLN metastasis and molecular subtype classification.

Ki-67 Expression and SLN Metastasis
We compared Ki-67 expression (%) between SLN positive and negative patients, and the Ki67 expression was significantly lower in SLN positive patients compared with SN negative patients (mean  SE; 14.4  2.5 versus 22.2  1.9, P = 0.0331) (Figure 1).
The result showed no significant association between Ki-67 expression and SLN metastasis (Table 2).There- fore, we investigated the threshold value of Ki-67 labeling index that differentiated patients who have SLN metastasis by ROC analysis.The ROC analysis identified a cut-off point at 7.5% (AUC, 0.646; Sensitivity, 86.3; Specificity, 36.4,P = 0.033) (Figure 2).Next, we used cut-off point of Ki-67 7.5% for reassessment.The result showed no significant association between Ki-67 expression was significantly associated with SLN metastasis by using the cut-off point at Ki-67 7.5% (Table 2).Multivariate analysis also showed that Ki-67 labeling index was one of the significant factors for predicting SLN metastasis in addition to clinical T-stage and LVI (P = 0.016; relative risk, 4.051, 95% confidence interval, 1.302 -12.602) (Table 3).

Discussion
Recent clinical trials have showed no outcome difference in patients with positive SLN between ALND versus no further axillary surgery, raising doubts on the role of SLNB itself.Therefore, it will be becoming important to  predict pathologically node negative before surgery.In our study, clinical T-stage and LVI were significantly associated with SLN metastasis: large tumor and prominent lymphatic and/or vascular invasion was shown to have a higher likelihood of being SLN positive.These findings were in accordance with previous reports [7][8][9][10][11][12][13].
Our study could not find out the association between molecular subtype and SLN metastasis, in contrast with previous reports [3,6,[14][15][16][17][18][19][20][21][22].Possible explanation to the discrepancy between previous reports and our results is that the distribution in among subtype might affect the results: HER 2 type was very few (7.0%), and this seems to be the reason why low frequency of SLN metastasis in HER 2 subtype was shown in our study.
Ki-67 was also a possible predictor for SLN metastasis in our results: Ki-67 labeling index of primary tumor was significantly lower in SLN positive patients compared with SN negative patients, and this finding has not been reported previously.First, we used Ki-67 cut-off point at 14% according to St Gallen consensus [29], however, our result showed that 14% was not appropriate cut-off point for dividing SLN positive from negative.Next, we performed ROC analysis to obtain the most effective Ki-67 cut-off point for dividing SLN positive from negative, and our result showed the validity of Ki-67 cut-off point at 7.5% for dividing SLN positive from negative.
In conclusion, Ki-67 labeling index, in addition to clinical T-stage and LVI, is significantly associated with SLN metastasis, and it seems to be useful to consider Ki-67 labeling index for SLN metastasis prediction.

Figure 2 .
Figure 2. A receiving operating characteristic (ROC) analysis of the accuracy of Ki-67 labeling index is illustrated.Accuracy is measured by calculating the area under the ROC curve (AUC).CI: confidence interval.