Istanbul Samatya Education and Research Hospital, Istanbul, Turkey.
Axillary lymph node status is one of the most powerful prognostic indicators in patients with breast cancer and has implications for adjuvant treatment. It has been demonstrated that enhanced histologic evaluation of axillary lymph nodes, including serial sectioning of paraffin tissue blocks and immunohistochemical (IHC) staining, increases the rate of detection of occult metastases. The clinical significance of occult lymph node metastases has been the subject of debate.
In the current study, the authors identified 267 patients who underwent axillary lymph node dissection (ALND) between 1987 and 1995 and were lymph node negative according to a routine pathologic evaluation, which included the complete submission of all lymph nodes and an examination of 1 hematoxylin and eosin (H&E)-stained section per paraffin block. Patients did not receive systemic chemotherapy or hormone therapy. All of the dissected lymph nodes from these patients were re-evaluated by intensified pathologic methods (serial sectioning with H&E levels plus IHC). Occult metastases were categorized by detection method and size. The clinical significance of the occult metastases was determined.
Thirty-nine patients (15%) who had lymph node-negative results on routine evaluation of their ALND specimens had occult metastases identified. Eight of these patients (20%) had macrometastases >2.0 mm, 15 (40%) had micrometastases (range, >0.2 mm to ≤2 mm), and 16 (40%) had isolated tumor cells (≤0.2 mm). The presence of occult metastases and the size of metastases did not affect recurrence-free or overall survival.
It is well accepted that axillary lymph node status is an important prognostic factor in patients with breast cancer. Patients without axillary lymph node metastases have better outcomes than patients with lymph node-positive disease. Before sentinel lymph node (SLN) biopsy became standard practice for patients with clinically lymph node-negative disease, all patients with breast cancer routinely underwent axillary lymph node dissection (ALND). During the era of routine ALND, at The University of Texas MD Anderson Cancer Center (MD Anderson), ALND specimens were dissected to identify all lymph nodes. Large lymph nodes were bivalved or serially sectioned, and small lymph nodes were submitted as a whole specimen. All lymph node material was embedded in paraffin blocks, and 1 hematoxylin and eosin (H&E) section from each block was obtained for histologic evaluation. The pathology report included the total number of lymph nodes examined, the total number of metastatic lymph nodes, and the largest size of metastatic focus.
The term “occult metastasis” has been defined as a metastasis that was not identified in the original H&E section but was detected by additional sampling of paraffin blocks, including serial sectioning and/or immunohistochemistry (IHC) staining for cytokeratin. Published studies have reported that occult metastases can be identified in 9% to 33% of patients who have negative lymph node status determined initially using standard pathologic processing of ALND specimens.1 However, the prognostic significance of these occult metastases has been controversial, because some studies have demonstrated reduced survival when occult metastases were identified,2-18 and others have demonstrated no survival difference.19-36 The current study was performed to determine the impact of occult metastases on recurrence-free survival (RFS) and overall survival (OS) in patients who underwent complete ALND and were determined to be lymph node negative by standard pathologic evaluation. None of these patients received systemic chemotherapy or hormone therapy, which was the standard practice for lymph node-negative patients at that time; therefore, this cohort allowed for determination of the natural history of occult lymph node disease.
MATERIALS AND METHODS
A prospectively maintained database of breast cancer patients who were treated at MD Anderson was queried, and 311 consecutive patients with T1-T3N0M0 breast cancer who were treated between 1987 and 1995 were identified. Patients who did not receive systemic chemotherapy or hormone therapy were identified and made up the study population (n = 267). All patients underwent either segmental mastectomy or total mastectomy and complete ALND as part of the surgical management of their disease. Two hundred four patients underwent segmental mastectomy, and 200 of those patients received adjuvant radiation using tangential photon fields to treat the breast. Sixty-three patients underwent mastectomy, and only 17 of those patients received adjuvant radiation. Clinicopathologic data, including age, tumor size, histologic subtype, nuclear grade, and estrogen receptor (ER) and progesterone receptor (PR) status, were recorded. This study was approved by the MD Anderson Institutional Review Board, and the consent of individual patients was waived.
Pathology slides for each patient were retrieved and rereviewed, and the diagnosis of lymph node-negative disease was confirmed. For all 267 study patients, a total of 2059 paraffin blocks from ALND specimens were reanalyzed using intensified method as follows: All blocks were cut at 4 different levels at 10-micron to 20-micron intervals with the intent to cut though the paraffin block. Levels 1 and 4 were stained with H&E, and level 3 was stained with pancytokeratin IHC. This intensified method is similar to our current SLN evaluation protocol. All slides were evaluated independently by 2 pathologists (C.K. and A.A.S.).
Tumor cells identified by either H&E and/or IHC in the parenchyma and/or sinusoid of the lymph nodes were considered to be occult metastases. The size of the metastasis was measured by ocular micrometer and categorized according to the sixth edition of the American Joint Committee on Cancer staging criteria as macrometastases (>2.0 mm), micrometastases (>0.2 mm, ≤2 mm), and isolated tumor cells (ITCs) (≤0.2 mm).37
RFS was measured from the date of diagnosis to the date of first local, regional, or distant metastasis. Patients who did not experience any of these events were censored at last follow-up or at the time of their death. OS was measured from the date of diagnosis to the date of death from any cause or was censored at last follow-up. RFS and OS curves were calculated using the Kaplan-Meier method, and differences between patient subgroups were compared using the log-rank test. Differences between patient subgroups also were investigated using a Cox proportional hazards model stratified on type of surgical treatment and receipt of adjuvant radiation therapy as a way to correct for the potential confounding effects of these 2 variables. All calculations were performed with Stata software (Stata/SE 11; Stata Corp., College Station, TX). Two-tailed P values ≤ .05 were considered statistically significant.
Detection of Occult Lymph Node Metastasis
An average of 16 lymph nodes per patient (median, 15 lymph nodes; range, 4-50 lymph nodes) was examined. Of the 267 patients evaluated, serial sectioning and IHC staining identified occult metastases in 39 patients (15%), including 8 patients (20%) with macrometastases, 15 patients (40%) with micrometastases, and 16 patients (40%) with ITCs. Metastases were detected in 16 patients (41%) by H&E and IHC and in 23 patients (59%) by IHC only. All occult metastases identified by H&E also were identified by IHC. Of the 16 patients who had ITCs, 14 had ITCs that were detected by IHC only and were not detectable by H&E staining. The median size of metastases identified by serial sectioning and IHC was 1.2 mm (range, 0.02-7.5 mm). In 33 patients, only 1 lymph node was involved; whereas in 4 patients, 2 lymph nodes were positive, and 2 patients had 3 and 4 lymph nodes involved, respectively.
Occult metastases were identified in 11 of 31 patients (35%) who had pure invasive lobular carcinoma (ILC) (n = 22) and/or mixed invasive ductal carcinoma (IDC) and ILC (n = 9) compared with 26 of 225 patients (12%) who had IDC (P < .001). No occult metastases were identified in 11 other types of invasive carcinoma, including mucinous (n = 4), papillary (n = 1), and medullary (n = 6) carcinomas.
The median follow-up for all study patients was 15.7 years (range, 0.3-29.7 years). The median follow-up for patients who remained alive at the time of analysis was 17.8 years (range, 2.6-29.7 years). There were 30 patients (11%) who died of disease, 77 patients (29%) who died of other causes, 157 patients (59 %) who were alive and disease free at last follow-up, and 3 patients (1%) who were alive with disease at last follow-up. There was no difference in RFS or OS between lymph node-negative patients and patients with occult metastases (P = .438 and P = .771, respectively). RFS and OS according to the size of occult metastases are illustrated in Figures 1A and 2A. In addition, the detection method/pattern was not associated with RFS or OS (P = .702 and P = .677, respectively). RFS and OS are illustrated according to the methods of detection in Figures 1B and 2B.
Because it is well known that metastases from ILC and mixed IDC and ILC often present as dispersed single cells that are not readily identified on H&E sections because of a discohesive growth feature, we evaluated this histology separately. Although a greater proportion of occult metastasis was identified in the ILC and mixed IDC/ILC group, RFS and OS were not impacted by the presence of occult metastasis compared with patients who did not have occult metastasis (Fig. 3A,B).
In addition to the presence of occult lymph node metastasis, other clinicopathologic factors, including patient age, tumor type (IDC, ILC, mixed IDC/ILC, or other types), tumor size, nuclear grade, and ER/PR status, were analyzed for their impact on OS and RFS (Tables 1 and 2). With the exception of age, none of these clinicopathologic factors had a significant association with either RFS or OS. Analyses were repeated stratified on type of surgery and administration of adjuvant radiation, and the results were qualitatively the same (ie, P values remained > .05 for all factors except age).
Table 1. Summary Statistics for Recurrence-Free Survival by Patient Groupa
aStatistical analysis excludes patients with unknown values.
Age ≥50 y
Tumor size, cm
No. of positive lymph nodes
The clinical significance of occult lymph node metastasis in patients with breast cancer remains controversial. Although some studies have demonstrated that occult metastases have prognostic significance,1-18 others have not.19-36 A recent meta-analysis of 27 studies demonstrated a small increased risk for disease recurrence and death in patients with breast cancer who have occult lymph node metastasis compared with lymph node-negative patients at 10 years of follow-up (recurrence: relative risk, 1.58; 95% confidence interval, 1.22-2.05; death: relative risk, 1.31; 95% confidence interval, 1.05-1.63).38 However, it is difficult to compare these studies, because the methodologies used to detect occult metastases varied, and treatments, including the use of systemic therapy, were heterogeneous. In the current study, we used an intensified pathologic method to identify the presence of occult lymph node metastasis in patients who predominantly had early stage disease (T1 or T2) and received no systemic chemotherapy or hormone therapy. Our study indicated that occult metastases were present in 15% of patients; however, in this homogeneous group of patients, the presence of occult metastases did not have prognostic significance with respect to either RFS or OS.
In contrast to the current study, a recent report from the Memorial Sloan-Kettering Cancer Center indicated that occult axillary lymph node metastases detected by intensified pathologic evaluation were prognostically significant.2 The Memorial study and our current study are comparable in several ways: 1) similar pathologic evaluation was used to detect occult metastasis (H&E levels plus IHC), 2) none of the patients received systemic treatment, and 3) both studies had long follow-up. However, in the Memorial study, patients were treated in the 1970s; they were mostly symptomatic, and they had larger primary tumors (66% T1, 34% T2). In contrast, we evaluated patients who were treated between 1987 and 1995, and their tumor sizes were smaller (82% T1, 18% T2-T3, and only 2 patients with T3 disease). In the Memorial study, there was a higher rate (23%) of occult metastases compared with the current study (15%), and the Memorial study also included more patients who had more than 1 lymph node involved (6%) compared with the current study (2%). Therefore, we conclude that there is a greater likelihood of occult lymph node metastasis in patients with clinically symptomatic breast cancer who have larger tumors than in patients who have smaller tumors. This is consistent with previously published literature demonstrating that the presence of lymph node metastases increases with increasing primary tumor size.39
The clinical significance of occult lymph node metastasis has recently been investigated in 2 prospective studies evaluating clinically lymph node-negative patients undergoing SLN biopsy: the American College of Surgeons Oncology Group (ACOSOG) Z0010 trial40 and the National Surgical Adjuvant Breast and Bowel Project (NSABP) B32 trial.41 An objective of both trials was to evaluate the prognostic significance of occult metastases in the SLN(s) that were not identified by standard H&E evaluation. The ACOSOG Z0010 trial enrolled patients with clinical T1/T2N0 breast cancer who underwent breast-conserving surgery to include segmental mastectomy and SLN biopsy. SLNs were evaluated by standard pathologic techniques using H&E staining. SLNs that were negative by H&E were sent to a central laboratory for IHC staining (the results were blinded to the treating physician). In that study, of 5119 evaluable patients, 3904 (76%) had H&E-negative SLNs and, of those, 349 (10.5%) were had positive IHC results. The OS rate at 5 years was 95.7% for lymph node-negative patients and 95.1% for patients who had occult lymph node metastases detected by IHC. There were no differences in OS between patients who were lymph node negative (H&E-negative/IHC-negative) and those that had occult lymph node metastasis (H&E-negative/IHC-positive) (P = .64).40 Similar to the ACOSOG Z0010 trial, the NSABP B-32 trial enrolled clinically lymph node-negative patients (n = 5611), including 3887 patients who were lymph node negative by standard H&E evaluation and had SLNs evaluated at a central laboratory for occult metastasis. Compared with the Z0010 study, the NSABP B-32 trial detected a higher prevalence of occult metastasis (15.9%).41 However, the NSABP B-32 trial did not limit enrollment to patients who had T1 and T2 cancers. Patients were stratified by tumor size (≤2.0 cm, 2.0-4.0 cm, or ≥4.1 cm), and the size of the primary tumor was an important predictor of identifying occult lymph node metastases. Approximately 22% of patients who had primary tumors ≥4.1 cm had occult metastases identified versus 15% of patients who had tumors ≤2.0 cm. In the NSABP B-32 study, the OS rate at 5 years was 95.8% for lymph node-negative patients and 94.6% for those with occult lymph node metastasis (P = .03). Despite the statistically significant difference, the magnitude of the difference in outcome was small, which led the NSABP investigators to conclude that there was no clinical benefit of enhanced pathologic evaluation of SLNs that initially were deemed negative using standard pathologic techniques.41 Data from the ACOSOG Z0010 and NSABP B-32 trials are not directly comparable with data from the current study. Most notably, in the current study, we evaluated for the presence of occult metastases in entire ALND specimens; whereas, in the cooperative group trials, the investigators were evaluating SLNs. In addition, in the current study, no patient received systemic therapy; whereas, in the cooperative group trials, the majority of patients did receive systemic therapy. In contemporary practice, decisions regarding the use of systemic therapy frequently are made based on primary tumor characteristics and not on the presence of lymph node metastases. Therefore, we believe that the administration of chemotherapy would have further reduced the clinical impact of occult metastasis on the survival of the patients in our study.
Although the clinical significance of occult metastases in axillary lymph nodes is still debated, there is strong evidence that not all pathologically detected axillary lymph node metastases are clinically significant and that not all occult metastases will progress to clinically evident regional or distant disease. The recently published ACOSOG Z0011 trial enrolled patients with clinical T1/T2N0 invasive breast cancer who had 1 or 2 positive SLNs. Patients were randomized to completion ALND or no further surgery, and additional lymph node disease was identified in 27.3% of patients who underwent additional surgery. At a median follow-up of >6 years, there were no differences between the 2 groups with respect to locoregional recurrences or OS.42, 43 It should be noted that all patients in the Z0011 trial underwent breast-conserving therapy to include whole-breast radiation. In the current study, even when patients were stratified according to receipt of radiation, the presence of occult metastasis did not affect RFS or OS. In addition, in contrast to the current study, in which no patients received adjuvant systemic therapy, such treatment was administered to >96% of patients in the Z011 trial.
Before to the Z0011 trial, the NSABP B-04 study provided evidence that not all axillary lymph node metastases will become clinically significant. Conducted in the 1970s, the B-04 study demonstrated that the regional recurrence rate in clinically lymph node-negative patients who underwent mastectomy without ALND or axillary irradiation was 16.8%; in contrast, 38% of patients who were randomized to undergo axillary dissection at initial presentation had lymph node metastases identified.44 The patients who did not undergo ALND did not receive adjuvant systemic therapy or axillary radiation; therefore, the low axillary recurrence rate is solely because of its biologic factors, not treatment effect. Although axillary lymph node metastasis is highly correlated with disease-free survival and OS, the development of distant recurrence may not originate from a regional lymph node source. It is possible that tumor metastases are organ-specific, as indicated by the “seed and soil” theory put forth by Fidler and colleagues.45 Lymph node-specific metastatic tumor cells may not spread and grow to other vital organs and, thus, would not affect the survival of patients with breast cancer.
In summary, for the current study, we reviewed a cohort of patients with early stage breast cancer who had ALND specimens that were lymph node negative on initial evaluation and did not receive systemic therapy. Enhanced pathologic processing identified occult axillary lymph node metastases in a small subset of those patients (15%). Although these patients did not receive systemic chemotherapy or hormone therapy, there were very low recurrence rates in both the occult lymph node metastasis group and in the group without occult metastasis. It is noteworthy that there was no difference in RFS or OS between patients with or without occult lymph node metastasis among those with early stage disease. The results from this study contribute further to the growing body of literature indicating that occult lymph node metastasis identified by intensified pathologic methods in patients with early stage breast cancer may not be biologically or clinically significant.
Dr. Canan Kelten was sponsored by The Scientific and Technological Research Council of Turkey.