早期乳腺癌腋窝淋巴结转移的预测及其临床适用性
Predictors of axillary lymph node metastases in early breast cancer and their applicability in clinical practice
- a Multidisciplinary Breast Centre, University Hospitals Leuven, Leuven, Belgium
- b BIOSTAT, Catholic University of Leuven, Leuven, Belgium
- c Department of Radiology, University Hospitals Leuven, Leuven, Belgium
- d Department of Pathology, University Hospitals Leuven, Leuven, Belgium
- e Department of Medical Oncology, University Hospitals Leuven, Leuven, Belgium
- f Department of Gynecology, University Hospitals Leuven, Leuven, Belgium
- http://dx.doi.org/10.1016/j.breast.2012.09.003, How to Cite or Link Using DOI
Abstract
Purpose
Lymph node involvement is the most important prognostic factor in breast cancer. It is a multifactorial event determined by patient and tumour characteristics. The purpose of this study was to determine clinical and pathological factors predictive for axillary lymph node metastasis (ALNM) in patients with early breast cancer and to build a model to portend lymph node involvement.
Methods
We evaluated 1300 consecutive patients surgically treated in our institution (2007–2009) for cT1-T2 invasive breast cancer. The patient and tumour characteristics evaluated included: age at diagnosis, number of foci, histologic grade, location, tumour size, histologic subtype, lymphovascular invasion (LVI), estrogen-receptor (ER), progesterone-receptor (PR) and Her-2 status. Univariate and multivariate analyses were performed. Factors significantly associated with ALNM by univariate analysis plus histologic subtype were included in the multivariate analysis.
Results
By univariate analysis, the incidence of ALNM was significantly associated with the presence of LVI (P < 0.0001), larger tumour size (P < 0.0001), higher histologic grade (P < 0.0001), retroareolar or lateral location in the breast (P < 0.0001), multiple foci (P = 0.0002) and in patients who underwent an axillary lymph node dissection. We found no effect of age, ER⁄PR nor HER-2 status. By multivariate analysis, ALNM was significantly associated with the presence of LVI (P < 0.0001), larger tumour size (P < 0.0001), axillary lymph node dissection (P = 0.0003), retroareolar and lateral tumour location in the breast (P = 0.0019) and the presence of multiple foci (P = 0.0155).
Conclusions
LVI and tumour size emerged as the most powerful independent predictors of ALNM, followed by the location of the tumour in the breast and the presence of multiple foci.
Keywords
- Breast cancer;
- Predictors;
- Lymph node metastases
1. Introduction
The axillary lymph node (ALN) status is the most important prognostic factor in patients with early breast cancer. It is a multifactorial event determined by patient and tumour characteristics. However, 20–30% of node positive patients remain free of distant metastasis whereas 20–30% of lymph node negative patients will eventually develop metastasis.1
To determine the ALN status, axillary lymph node dissection (ALND) has been the standard of care in patients with invasive breast cancer in order to provide correct staging of the patient and to obtain good local control. In recent years however, sentinel lymph node biopsy (SLNB) has emerged as an alternative to ALND and has become the standard treatment for axillary staging in all clinically node negative patients (by clinical examination, on ultrasound and/or fine needle aspiration cytology) except for T4 tumours.2 and 3 The SLN procedure has been proven to be a feasible, accurate and suitable method for the staging of the axilla, while avoiding the morbidity of an ALND.4 and 5
The underlying pathways of lymph node involvement remain unclear. Several predictors of lymph node metastasis have been described such as multiple foci, higher histologic grade, lateral and retroareolar location of the tumour, larger tumour size and the presence of LVI.6 and 7 The predictive role of ER/PR and HER-2 status,6 and 8 histologic subtype9 and age10 remains controversial.
The goal of our study is to determine factors associated with axillary lymph node metastasis (ALNM) in our series of patients, to compare these findings with the literature and to build a model to predict lymph node involvement.
2. Materials and methods
2.1. Patients
Data were obtained from retrospective review of the Multidisciplinary Breast Centre (MBC) database from the University Hospitals Leuven (Leuven, Belgium). From 1 January 2007 till 31 December 2009, 1300 patients with a primary operable cT1-cT2 invasive breast carcinoma underwent resection of the primary tumour and axillary staging by SLNB and/or ALND. The local surgical treatment consisted of a wide excision followed by radiotherapy or a mastectomy with or without radiotherapy. We excluded patients1 treated for a local recurrence,2 with a carcinoma in situ,3 who received neo-adjuvant therapy and4 with primary metastatic disease.
In order to determinate factors that are associated with ALNM, we evaluated each of the following variables: age at diagnosis, number of foci, histologic tumour grade, tumour location, tumour size, histologic subtype LVI, ER, PR, HER-2 status and the procedure used for axillary staging (SLNB or ALND). Univariate and multivariate analyses were performed. Factors significantly associated with lymph node metastases in univariate analysis as well as histologic subtype were included in the multivariate model.
2.2. Sentinel lymph node localization
The sentinel lymph node (SLN) procedure was performed by injection of a radioactive (99mTc-labelled nanocolloid) tracer at the level of the tumour and Patent Blue® retroareolar. The SLNs were removed surgically using a hand-held gamma-ray detection probe.
2.3. Examination of the lymph nodes
SLNs were routinely examined by serial sectioning. Every 300 μm 2 coupes were stained, 1 with routine haematoxylin and eosin (H&E) and 1 stained immunohistochemically using cytokeratin. Lymph nodes in an ALND were examined by H&E staining using 3 sections per node. According to published guidelines, lymph nodes from lobular breast cancers, classified as lymph node negative on H&E, were additionally stained with epithelial markers.
2.4. Statistical methodology
Logistic regression models were used to analyse the effect of various predictors on the presence or absence of positive lymph nodes. Given that females with bilateral breast tumour appear twice in the data set, generalized estimating equations (GEE) were used to account for the association between the two responses coming from the same person.
First we fit univariate models for a number of predictors of which a relationship with the response was to be expected. The presence of a non-linear relationship was checked for the continuous variables: tumour size, age and number of foci using restricted cubic splines. In a second stage, we fit a multivariate model including these predictors that had shown to be related to the response in the univariate analyses. Moreover, histologic subtype was included as we previously had shown that lobular and non-lobular breast cancers differ regarding axillary lymph node metastases.9 Odds ratios and P-values were calculated. The 95% confidence intervals (CI) were calculated and pair-wise comparison between groups was made to determine which results were statistically significant. All P-values smaller than 0.05 were used for statistical significance. All analyses have been performed using SAS software 9.2 (SAS Institute, Cary, NC).
3. Results
A total of 1300 patients were found to meet the inclusion criteria. Thirty-eight patients had bilateral invasive tumours, providing 1338 breasts for analysis. In the cases where patients had multiple foci, the size and location of the largest focus were used for the analyses.
Table 1 shows the clinical and pathological characteristics of the patients. The mean age was 58.59 years (range 22–94 years). Eight hundred forty-three patients underwent a SLNB, followed by an ALND only when the sentinel lymph node was positive. The remaining 495 patients underwent an ALND. Four hundred ninety-seven patients (37.14%) were lymph node positive.
-
- a
-
Other: Tubular, Medullary, Mucinous (colloid), Metaplastic, Encapsulated papillary, Cribriform, Invasive micropapillary, Apocrien and Lymfoepithelioma-like
- b
-
ER, estrogen-receptor.
- c
-
PR, progesterone-receptor.
Results of the univariate analyses are presented in Table 2. For binary (two levels) or continuous predictors we present the odds ratio (OR) with its 95% confidence interval and the P-value. For categorical predictors with more than two levels we present an overall test with the difference between the various levels and, in case of significance, odds ratios for the pair-wise comparisons between the levels.
-
- a
-
ER, estrogen-receptor.
- b
-
PR, progesterone-receptor.
According to tumour size, the risk for ALNM steadily increases from the smallest to the largest tumours (OR = 1.0567, 95% CI: 1.0457–1.0679). The larger the tumour size, the higher the probability of positive lymph nodes. No evidence was found for a non-linear relationship of tumour size (data not shown).
Second, we found a positive effect of the number of foci (OR = 1.3980, 95% CI: 1.1738–1.6651). Patients with multiple foci have a higher probability of ALNM.
Furthermore, there is a highly significant effect of LVI. Patients with LVI have a much higher probability of positive lymph nodes compared to patients without LVI (OR = 8.6468, 95% CI: 6.4149–11.6552).
As shown in Table 2 there is an overall effect of the tumour grade. The higher the tumour grade, the higher the probability of ALNM. Patients with grade 2 tumours have a lower probability compared to patients with grade 3 tumours (OR = 0.7463, 95% CI: 0.5850–0.9520) and patients with grade 1 tumours have the lowest probability of ALNM (grade 1 vs. grade 3 OR = 0.6600; 95% CI: 0.4770–0.9133).
The tumour location was also found to have a statistically significant effect on nodal involvement. Patients with lateral tumours have a higher probability of positive lymph nodes compared to patients with medial tumours (OR = 1.4814, 95% CI: 1.1191–1.9612). However, patients with a retroareolar tumour have the highest probability of positive lymph nodes (OR = 0.3937, 95% CI: 0.2162–0.7170).
Furthermore, we analysed the difference between the two procedures for axillary staging (ALND and SLNB) on the presence or absence of positive lymph nodes. We found a significant difference: the probability of finding positive lymph nodes is higher with the ALND (OR = 2.9405; 95% CI: 2.3311–3.7093). This is not surprising as patients undergoing ALND usually have larger primary tumours and/or clinically positive lymph nodes.
Histologic subtype, age at diagnosis and the molecular markers ER, PR and HER-2 were found not to be significantly correlated with ALNM.
Subsequently, multivariate analyses were carried out on those variables found to be statistically significant on univariate analyses. Moreover histologic subtype was included as we have previously shown that the histologic subtype is an independent predictive factor for lymph node involvement 9. The results are presented in Table 3.
-
- a
-
Invasive ductal carcinoma.
- b
-
Invasive lobular carcinoma.
All variables of the univariate analyses were found to be independently significant in the multivariate analyses, except for tumour grade and histologic subtype. Lymphovascular invasion and tumour size were found to be the most powerful predictors of nodal status (P value < 0.0001) followed by the procedure, tumour location in the breast and the number of foci.
We validated our model on the same group of 1300 patients, using a correction factor. The predictive power of our model, measured by the area under the curve (AUC), is 0.778.
4. Discussion
The presence of lymph node metastasis observed in our series of 1338 breast tumours (37.14%) is in accordance with the incidence range reported in literature (33.2%–39%).7 and 8
Four parameters emerged as independent predictors of the lymph node metastasis status in multivariate analyses, namely lymphovascular invasion, primary tumour size, multiple foci and tumour location. The age of the patient at diagnosis and histologic grade did not retain in multivariate analyses. These results are largely in concordance with other series published in the literature.
Lymphovascular invasion has been described as the strongest independent predictor of nodal involvement.11, 12 and 13 This finding was confirmed in our present study. Of our patients with LVI, 75.5% had positive lymph nodes. Woo et al found LVI was associated with a significant decrease in survival at 12 year follow-up despite absence of nodal disease and LVI portended an even worse outcome in patients with nodal disease.14
In our series, tumour size was the second strongest predictive factor. This has also been shown consistently in various studies.11, 12 and 15 For example, in tumours smaller than 10 mm, 15% of patients were found to be lymph node positive, while 25% of the patients with tumours smaller than 20 mm.16
However, the seemingly simple relationship between tumour size and lymph node status is not always clinically reliable. Cases of small primary tumours with extensive lymph node metastases have been reported to have a particularly aggressive clinical course.17
A third predictor is the tumour location in the breast. We found a lower frequency of LNM in the medial quadrant located tumours. The small number of published studies that evaluated the predictive value of tumour location in the process of lymph node involvement showed similar findings to our study. The lower frequency of lymph node involvement in the medial quadrant located tumours observed in this study, suggests the possibility of alternative routes of lymphatic flow, particularly to the internal mammary chain. In an earlier comprehensive review of literature on internal mammary nodes, Bevilacqua et al18 showed that in tumours of the upper inner quadrant (UIQ) ALNM is less frequent than tumours in any other quadrant. However, these patients have a worse prognosis in terms of overall survival. Because UIQ tumours are associated with a lower risk of axillary metastases, we might underestimate the risk of internal mammary lymph node metastases in these UIQ tumours. Other recent studies also have demonstrated a poorer prognosis associated with medial quadrant tumours in general than with lateral-quadrant tumours in terms of disease-free and overall survival.19 The authors of these studies explained that these differences may be due to the greater likelihood of occult spread to internal mammary lymph nodes seen in medial tumours.
In our population only 49 patients had a retroareolar tumour. Nevertheless, 62.1% of these patients had at least one positive lymph node. In another study of Bevilacqua et al20 the authors documented that retroareolar tumours have the highest probability of LNM. This could be explained by the anatomy of the lymphatic system. Because carcinoma of the breast originates in the perilobular segment of the acini, metastases progress in a centrifugal way from the periacinar and duct region into larger lymphatic vessels that then flow into two main lymphatic trunks, the medial and the lateral. In their courses to the areolar region, the two trunks receive secondary tributary vessels and eventually converge in the retroareolar area.21 As a result all the main lymphatic vessels pass retroareolarly.
The number of foci has also rarely been evaluated as a potential predictive factor. We found that the number of foci of the primary tumour per se was a significant independent predictor. This could be explained by the occurrence of multiple foci suggesting the presence of additional (occult) microscopic foci of malignancy; the higher the number of foci, the larger the tumour size.
There is some disagreement about the molecular markers in literature: some authors described an association,8, 11, 22 and 23 some showed no correlation,20, 24 and 25 while Viale et al26 actually demonstrated an inverse relationship. Our study suggests no significant association between ER, PR nor HER-2 status and lymph node involvement.
In our analyses, the probability of finding positive lymph nodes was higher with ALND versus SLNB. This result could be expected as patients undergoing ALND usually have larger primary tumours (mean 26.65 mm versus 18.43 mm), are more frequently multifocal (24.44% versus 8.19%) and/or have clinically positive lymph nodes. Therefore, we performed the same analysis only with the lymph node negative patients. Here we found no longer a significant difference concerning the procedure (data not shown).
There is an ongoing debate about the relation between axillary lymph node involvement and age.10, 26,27 and 28 By univariate analysis we found no effect of age on the nodal status.
Finally, we validated our model with the same 1300 patients on the basis of a multivariable logistic regression model containing the 7 variables, using a correction factor. The discrimination of the model, calculated by the area under the curve (AUC), was 0.778.
The results of our study are useful in clinical practice as they allow rationalized decision making in specific clinical situations. First, in patients with a contra-indication for an intervention under general anaesthesia and a tumour that can be treated by breast conservative surgery. A wide excision of the tumour under local anaesthesia can be considered, with omission of axillary surgery, provided that the risk of lymph node involvement is low. Next, in patients with an unexpected diagnosis of an invasive tumour, our results can be helpful to determine the further surgical management of the axilla. Finally, the risk of lymph node involvement may influence the timing of an autologous breast reconstruction. In patients at low risk for lymph node involvement, a primary reconstruction can be considered while in patients at high risk for lymph node involvement, a reconstruction in a second time may be preferred.
Our results will not alter the indications for a SLNB. Even patients with a high probability of lymph node involvement are candidates for a SLNB as a subgroup of these patients can safely avoid an ALND.
Finally, this study cannot be used to predict the tumour load in the axilla. There is an ongoing debate on the prognostic implications of minimal lymph node involvement. By the introduction of the SLNB there was an upstaging of a subgroup of patients. These patients have mainly isolated tumour cells or micrometastasis in the SLN due to more detailed histologic examination. However, our model to predict lymph node involvement cannot differentiate between micro- or macrometastatic disease in the axilla (data not shown).
5. Conclusion
Lymphovascular invasion and tumour size emerged as the most powerful independent predictors of ALNM, followed by the location of the tumour in the breast and the presence of multiple foci. These results can be used for decision making in clinical practice.
Conflict of interest statement
None declared.
References
-
- 1
-
Breast cancer metastasis: markers and models
-
Nat Rev Cancer, 5 (8) (2005), pp. 591–602
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (407)
-
- 2
-
American Society of Clinical Oncology guideline recommendations for sentinel lymph node biopsy in early-stage breast cancer
-
J Clin Oncol, 23 (2005), pp. 7703–7720
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (714)
-
- 3
- Cardoso F, Stordeur S, Vlayen J, Bourgain C, Carly B, Christiaens MR, et al. Scientific support of the college of oncology: update of the national guidelines on breast cancer, KCE reports 143C, study nr: 2008-52.
-
- 4
-
Short-term morbidity of the upper limb after sentinel lymph node biopsy or axillary lymph node dissection for Stage I or II breast carcinoma
-
Cancer, 98 (2003), pp. 690–696
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (49)
-
- 5
-
Physical and psychological morbidity after axillary lymph node dissection for breast cancer
-
J Clin Oncol, 17 (1999), pp. 143–149
-
View Record in Scopus| Cited By in Scopus (186)
-
- 6
-
Predictors of axillary lymph node metastasis in breast cancer a systematic review
-
Eur J Surg Oncol, 33 (4) (2007 May), pp. 409–419
-
Article|PDF (433 K)|View Record in Scopus| Cited By in Scopus (19)
-
- 7
-
Predicting the status of axillary sentinel lymph nodes in 4351 patients with invasive breast carcinoma treated in a single institution
-
Cancer, 103 (3) (2005), pp. 492–500
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (63)
-
- 8
-
Frequency and predictors of axillary lymph node metastases in invasive breast cancer
-
ANZ J Surg, 71 (2001), pp. 723–728
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (46)
-
- 9
-
Lobular and non-lobular breast cancers differ regarding axillary lymph node metastasis: a cross-sectional study on 4292 consecutive patients
-
Breast Cancer Res Treat, 128 (2) (2011 Jul), pp. 429–435
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (2)
-
- 10
-
Relationschip between age and axillary lymph node involvement in women with breast cancer
-
J Clin Oncol, 27 (18) (2009 Jun 20), pp. 2931–2937
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (20)
-
- 11
-
Predictors of axillary lymph node metastases in patients with T1 breast carcinoma
-
Cancer, 79 (1997), pp. 1918–1922
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (168)
-
- 12
-
Predictors of axillary lymph node metastases in patients with T1 breast cancer
-
Cancer, 73 (1994), pp. 350–353
-
- 13
-
Predicting axillary lymph node metastases in patients with T1 infiltrating ductal carcinoma of the breast
-
The Breast, 10 (2001), pp. 155–159
-
- 14
-
Lymph node status combined with lymphovascular invasion creates a more powerful too for predicting outcome in patients with invasive breast cancer
-
Am J Surg, 184 (4) (2002 Oct), pp. 337–340
-
Article|PDF (52 K)|View Record in Scopus| Cited By in Scopus (47)
-
- 15
-
Relation of tumor size, lymph node status, and survival in 24,740 breast cancer cases
-
Cancer, 63 (1989), pp. 181–187
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (1141)
-
- 16
-
Prediction of axillary lymph node metastases in a screened breast cancer population. South-east Sweden Breast Cancer Group
-
Acta Oncol, 33 (1994), pp. 603–608
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (38)
-
- 17
-
The quotient of number of nodes and tumour size (N/T) from primary breast cancer predicts the clinical course after diagnosis of distant relapse
-
Eur J Surg Oncol, 30 (3) (2004), pp. 346–351
-
- 18
-
A selection algorithm for internal mammary sentinel lymph node biopsy in breast cancer
-
Eur J Surg Oncol, 28 (2002), pp. 603–614
-
Article|PDF (258 K)|View Record in Scopus| Cited By in Scopus (42)
-
- 19
-
Relationship between tumour location and relapse in 6781 women with early invasive breast cancer
-
J Clin Oncol, 18 (15) (2000), pp. 2828–2835
-
View Record in Scopus| Cited By in Scopus (70)
-
- 20
-
A model for predicting axillary node metastases on 2000 sentinel node procedures and tumour position
-
EJSO, 28 (2002), pp. 490–500
-
Article|PDF (197 K)|View Record in Scopus| Cited By in Scopus (28)
-
- 21
-
Common breast lesions: a photographic guide to diagnosis and treatment
-
Cambridge University press (edition 2003) p. 210
-
- 22
-
Predictors of axillary lymph node metastases in women with early breast cancer in Singapore
-
Singapore Med J, 46 (2005), pp. 693–697
-
View Record in Scopus| Cited By in Scopus (12)
-
- 23
-
Identification of women with T1-T2 breast cancer at low risk of positive axillary nodes
-
J Surg Oncol, 65 (1997), pp. 34–39
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (78)
-
- 24
-
Factors affecting nodal status in invasive breast cancer: a retrospective analysis of 623 patients
-
Breast J, 16 (3) (2010 May-Jun), pp. 271–278
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (7)
-
- 25
-
Predictive factors associated with axillary lymph node metastases in T1a and T1b breast carcinomas
-
J Am Coll Surg, 191 (2000), pp. 1–6
-
- 26
-
Predicting the status of axillary sentinel lymph nodes in 4351 patients with invasive breast carcinoma treated in a single institution
-
Cancer, 103 (2005), pp. 492–500
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (63)
-
- 27
-
Prediction of axillary lymph node involvement of women with invasive breast carcinoma: a multivariate analysis
-
Cancer, 83 (5) (1998), pp. 948–955
-
View Record in Scopus|Full Text via CrossRef| Cited By in Scopus (80)
-
- 28
-
Lymphatic invasion, tumor size, and age are independent predictors of axillary lymph node metastases in women with T1 breast cancers
-
Ann Surg, 230 (1999), pp. 692–696
- 您可能感兴趣的文章
-