- Email: [email protected]
Sentinel node biopsy compared with complete axillary dissection for staging early breast cancer with clinically negative lymph nodes: results of randomized trial
original article
Annals of Oncology 20: 1001–1007, 2009 doi:10.1093/annonc/mdn746 Published online 27 January 2009
Sentinel node biopsy compared with complete axillary dissection for staging early breast cancer with clinically negative lymph nodes: results of randomized trial G. Canavese1, A. Catturich1, C. Vecchio1, D. Tomei1, M. Gipponi2, G. Villa3, F. Carli4, P. Bruzzi5 & B. Dozin5* 1 5
Surgical Senology Unit, National Cancer Research Institute; Departments of 2General Surgery; 3Nuclear Medicine, A.O.U San Martino; 4Anatomy and Cytohistology Unit; Clinical Epidemiology Unit, National Cancer Research Institute, Genoa, Italy
Received 4 August 2008; revised 5 November 2008; accepted 6 November 2008
cancer patients with negative SLNs. Evidence of a similar efficacy, in terms of survival and regional control, of this strategy as compared with axillary resection is based on few clinical trials. In 1998, we started a randomized study comparing the two strategies, and we present here its results. Materials and methods: Patients were randomly assigned to sentinel lymph node biopsy (SLNB) and axillary dissection [axillary lymph node dissection (ALND arm)] or to SLNB plus axillary resection if SLNs contained metastases (SLNB arm). Main end points were overall survival (OS) and axillary recurrence. Results: One hundred and fifteen patients were assigned to the ALND arm and 110 to the SLNB arm. A positive SLN was found in 27 patients in the ALND arm and in 31 in the SLNB arm. Overall accuracy of SLNB was 93.0%. Sensitivity and negative predictive values were 77.1% and 91.1%, respectively. At a median follow-up of 5.5 years, no axillary recurrence was observed in the SLNB arm. OS and event-free survival were not statistically different between the two arms. Conclusions: The SLNB procedure does not appear inferior to conventional ALND for the subset of patients here considered. Key words: axillary lymph node dissection, breast cancer, overall survival, randomized clinical trial, regional node recurrence, sentinel lymph node biopsy
introduction Until few years ago, axillary lymph node dissection (ALND), together with resection of the primary tumor, was the standard procedure used in the surgical management of breast cancer patients, as well as the routine method for axillary node staging. The pathological stage of the axillary nodes represents a major prognostic factor in early breast cancer and is crucial in the decision on the need and type of adjuvant therapy. However, complete axillary dissection often results in significant physical morbidities and psychological distress [1]. A decade ago, the technique of visualization, identification and intraoperative analysis of the sentinel lymph node (SLN) opened new perspectives in the management of early breast cancer patients [2, 3]. From these early studies, it appeared that SLN staging could predict the status of the axillary nodes and that patients with negative SLNs may be spared total axillary dissection [4, 5]. At the time the present study was started, the evidence in support of the use of the SNL procedure in place of *Correspondence to: Dr B. Dozin, Clinical Epidemiology Unit, National Cancer Research Institute, Largo Rosanna Bensi 10, 16132 Genoa, Italy. Tel: +39-010-5737492; Fax: +39-010-354103; E-mail: [email protected]
standard axillary dissection was inadequate for the standards currently used to introduce a new technology in routine clinical practice. We thus carried out, in a population of early breast cancer patients, a randomized comparison of sentinel lymph node biopsy (SLNB) followed by routine axillary dissection with SLNB followed by axillary dissection when the SLN was found metastatic. The study was designed as a noninferiority trial. Although it is underpowered, the results support the evidence of the efficacy of SLNB for axillary staging in early breast cancer and indicate that this procedure is equivalent to routine axillary dissection, in terms both of overall survival (OS) and regional recurrence. Noteworthy, even today, while the beneficial short-term effects of SNL on quality of life have been clearly demonstrated [6], reliable information on longterm morbidity and mortality associated with its use is still lacking [1].
materials and methods patient population The present single-center randomized study was conducted at the National Cancer Research Institute of Genoa, Italy. Patients were enrolled according
ª The Author 2009. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected]
original article
Background: Sentinel lymph node (SLN) staging is currently used to avoid complete axillary dissection in breast
original article to the following eligibility criteria: age 18–75 years, primary invasive breast carcinoma as revealed by mammography and cytohistology, clinically negative axillary lymph nodes and unifocal tumor sized £3 cm as estimated by ecography. At the time we started this study, on the basis of the SLN validation study of Krag et al. [4], we considered this size as the cutoff dimension above which patients were to undergo standard axillary dissection because of their higher risk of bearing metastatic lymph nodes. Patients were also excluded from the study if they had undergone previous surgery on the same breast or on the ipsilateral axilla or if they were suffering from chronic life-threatening disease possibly preventing adjuvant therapy.
study design, randomization procedure and end points Patients were randomly assigned to undergo, together with breast surgery, either SLNB and ALND or SLNB followed by ALND only if the SLN was found positive at the intraoperative evaluation. Randomization was carried out at the time of surgery by phoning the data-coordinating center at the Epidemiology and Clinical Trials Unit of the Institute where the data manager verified patient eligibility and communicated the allocated treatment. Treatment allocation was carried out in a 1 : 1 ratio according to a computer-generated randomization list. The study had been approved by the Ethical Committee of the Institute and was conducted in accordance with the International Good Clinical Practice Guidelines. All patients were aware of the aims and procedures of the study and signed an informed consent. The main objective of the study was the comparison of event-free survival (EFS) and OS between the two arms at 5 years from surgery. An additional endpoint was the evaluation of the frequency of axillary recurrences in patients of the SLNB arm and did not undergo axillary dissection. Secondary analyses, conducted in the ALND arm, focused on the evaluation of the sensitivity and predictive value of the SLN procedure with respect to the presence of axillary metastases.
SLN identification and intraoperative evaluation All patients underwent SLN dissection and intraoperative evaluation. The SLN was identified both by breast lymphoscintigraphy and lymphatic dye mapping. A peritumoral injection of 0.2 mCi of 99mTechnetium (Nanocol, Amersham-Sorin Biomedica, Saluggia, Italy) was carried out the day before surgery; an additional peritumoral s.c. injection of blue dye (Patent Blue-V) was carried out at the time of surgery. Through a small axillary incision, the radioactive SLN was localized with a c-ray detecting probe and removed for immediate intraoperative search for metastases. The SLN was bisected along its major axis and five pairs of frozen sections, each 4-lm thick, were cut every 10 lm in each half of the node. The first, third and fifth sections were stained with hematoxylin–eosin. If this histological evaluation resulted negative or ambiguous, the second and the fourth sections were tested by immunohistochemistry for the presence of cytokeratins [EPOS (Enhanced Peroxidase One-Step) method with cytokeratin MNF116 mAb and horseradish peroxidase (Dako, Milan, Italy)]. The remaining tissue was paraffin embedded for definitive postoperative evaluation. Any patient whose SLNs could not be identified was subjected to axillary dissection, independently of the treatment assigned.
surgery, adjuvant therapy and follow-up Mastectomy or conservative quadrantectomy was carried out according to standard criteria. Only patients who underwent conservative surgery received radiotherapy to the ipsilateral breast (50-Gy dose over 8 weeks). No radiotherapy to the axilla was carried out. The choice of adjuvant chemo- and/or hormone therapy regimen, when given, was based on the main prognostic factors of the primary tumor (nodal status, tumor size, tumor grading, hormonal receptor status). Follow-up included clinical examinations every 4 months for the first 3 years, every 6 months for the next 2 years and once a year thereafter. A
1002 | Canavese et al.
Annals of Oncology
mammography was scheduled annually. In case of clinically suspected lymph nodes, an axillary ecotomography was carried out, followed by a needle biopsy aspiration for cytological examination when deemed necessary.
statistics The main objective of the study was to evaluate whether the choice of not carrying out axillary dissection in patients with intraoperative negative SLN could ensure to those patients a prognosis equivalent to that expected for patients undergoing routine axillary dissection. The primary end point was EFS, and sample size estimates were based on an expected 5 years of 80% in the control ALND arm based on published data and on our own previous experience. The study was designed as a noninferiority trial, with the aim of rejecting the null hypothesis that the experimental treatment was associated with an absolute decrease in disease-free survival of 4% (hazard ratio = 1.21) due to the reduction in the use of adjuvant chemotherapy in falsenegative patients in the SNLB arm. Setting a study power at 90% for a onesided a error of 0.025 and considering both accrual and follow-up periods of 5 years, together with an estimated annual lost-to-follow-up rate of 2%, we established that 2570 patients were to be enrolled in the study. EFS was calculated from the date of surgery to the date of occurrence of any of local, controlateral, regional node or distant relapse or death from any cause, whichever came first. OS was estimated as the time interval between the date of surgery and the date of death from any cause. In both analyses, survival times of patients free from relapse and/or alive at last examination were censored at that date. EFS and OS curves were obtained using the Kaplan–Meier method, and differences between arms were assessed by the log-rank test. Differences between the two arms in baseline prognostic factors were assessed by the Pearson chi-square test or the Fisher’s exact test. Two-sided P values and/or 95% confidence intervals (CIs) are reported. All statistical analyses were based on the intention-to-treat approach, in that all patients, after the exclusion of ineligible patients, were included in the treatment arm assigned at randomization, regardless of the treatment actually received. Analyses were carried out using the SPSS package (version 14.0 for Windows)
results patients’ characteristics From November 1998 to October 2001, 248 patients were consecutively enrolled in the study and 124 were randomly assigned to each of the study arms (Figure 1). This number is far below the projections indicated above; indeed, enrollment was prematurely interrupted when patients became aware of the early promising results of the SLNB technique [5, 7] as they were increasingly refusing randomization in a study where routine axillary dissection was one of the alternative treatments. In each arm, nine patients were excluded from evaluation because of an inadequate presurgery cytology that should have precluded their eligibility. In the SLNB arm, three more patients were excluded because the intraoperative examination revealed the presence of a multifocal lesion. In the same arm, two more patients who were eligible could not be evaluated because postsurgical pathology revealed that the lesion was benign. The final number of assessable patients was 225, among whom 115 (51.1%) were allocated to the ALND arm and 110 (48.9%) to the SLNB arm. Among these 225 patients, only two underwent total mastectomy; all the others received conservative breast surgery. The main patient and tumor characteristics are summarized in Table 1. It should be noted that two patients, one in each
Volume 20 | No. 6 | June 2009
Annals of Oncology
Figure 1. Trial flow chart. ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy.
arm, were staged pTis upon definitive diagnosis; these two patients were maintained for final evaluation because at randomization they were meeting the eligibility criteria. Patient distribution in the two arms was similar with respect to all the prognostic factors (P values: 0.112–0.978).
identification of the SLN and evaluation of the axillary status A total of 405 SLNs were removed from the 225 patients (mean 1.80 per patient), 211 nodes in the ALND arm (mean 1.83 per patient) and 194 nodes in the SLNB arm (mean 1.76 per patient). A mean of 12.7 nonsentinel nodes per patient were removed from the 148 patients who underwent axillary dissection. Through intraoperative scintigraphy, no radioactive SLN was revealed in three patients (Table 2). All three patients, including two allocated to the SLNB arm, underwent routine axillary dissection, and macrometastases were found in the axilla in all of them. Upon intraoperative evaluation, no isolated tumor cells (groups of cells £ 0.2 mm) were found in any of the SLNs analyzed. Instead, micrometastasis (foci > 0.2 mm but £2 mm) and macrometastases (foci > 2 mm) in the SLN were found in 9 and 40 patients, respectively. In all these cases, the intraoperative diagnosis was confirmed at the definitive histological examination. In the ALND arm, none of the six patients with micrometastatic SLN and 10 (55.6%) of the 18 patients with macrometastatic SLN presented metastases in the axilla. As for negative patients, in the ALND arm, we have distinguished three classes of negative patients with respect to the SLN status: (i) 79 true-negative patients with a negative SLN both at intra- and postoperative examination and no
Volume 20 | No. 6 | June 2009
original article metastasis in the axillary nodes; (ii) three patients whose SLN was negative at intraoperative evaluation but was found positive with micrometastases at the definitive examination and (iii) eight true false-negative patients with a negative SLN both at intra- and postoperative diagnosis, but with metastases in the axilla: Out of them, two patients had axillary micrometastases, two patients had one macrometastatic axillary node and four patients had more than one node involved. In the SLNB arm, two (66.7%) of the three patients with micrometastatic SLN and 10 (45.5%) of the 22 patients with macrometastatic SLN had also metastatic axillary nodes. In the SLNB arm, in six patients, SLN was negative at intraoperative evaluation but was found positive with micrometastases at the definitive examination. Among them, four and two patients had micro- and macrometastases, respectively. To summarize the number of axillary dissections carried out in the whole population, all 115 patients in the ALND arm and 33 patients (30%) of the 110 patients in the SLNB arm underwent the procedure (two patients with SLN not visualized, six patients with false-negative SLN and 25 patients with a SLN found positive). As scheduled, the remaining 77 (70%) patients in the SLNB arm with a negative SLN at definitive examination were spared complete axillary dissection. Overall, in the ALND arm, the procedure of intraoperative evaluation of the SLN had a sensitivity of 68.6% and a specificity of 100%. Taking into account the three SLNs that resulted positive only at the definitive evaluation, sensitivity was 77.1% (27 of 35 patients with positive nodes were identified). The negative and positive predictive values were 91.1% and 100%, respectively. The overall accuracy of the SLNB procedure was 93.0% (correct in 106 of the 114 assessable patients).
EFS and OS At a median follow-up of 5.5 6 1.4 years, 22 unfavorable events occurred, 12 in the ALND arm and 10 in the SLNB arm (Table 3). One ipsilateral breast recurrence was observed in the SLNB arm. It was treated with surgery alone. Contralateral primary tumors developed in four patients, three in the ALND arm and one in the SLNB arm. All these five patients were still alive at the end of the observation period. No overt axillary metastasis was observed in the SLNB arm. In the ALND arm, axillary recurrence occurred in only one patient 3.1 years after surgery; this patient died 161 days after diagnosis of the recurrence. Six patients in each arm developed distant metastases. Four of them (one in the ALND arm and three in the SLNB arm) died at 153 days, 148 days, 10 months and 4.7 years from relapse, respectively. Four more patients, two in each arm, died without disease recurrence. The overall annual rate of events was 18.6 per 1000 patients in the ALND arm and 16.2 per 1000 patients in the SLNB arm (risk ratio of SLNB versus ALND = 0.87, 95% CI 0.38–2.01, P = 0.741). Accordingly, 5-year EFS was not statistically different between the two arms (Figure 2A; ALND arm: 89.8%, 95% CI 86.9% to 92.7%; SLNB arm: 94.5%, 95% CI 90.9% to 98.1%; log-rank P = 0.715).
doi:10.1093/annonc/mdn746 | 1003
original article
Annals of Oncology
Table 1. Patient characteristics by treatment arm: 115 patients in the ALND arm and 110 patients in the SLNB arma Characteristic Age (years) Mean Median Range (minimum– maximum) Age by category £45 years 46–55 years 56–65 years >65 years Tumor site Right Left Quadrant site Outer Inner or central >1 Quadrant Histology Ductal NOS Lobular Other In situ Pathological T pTis pT1mic pT1a pT1b pT1c pT2 Pathological N pN0 pN1mic pN1a pN2a pN3a Grading G1 G2 G3 Unknown Hormonal receptor status ER+ /PgR+ ER2 /PgR+ ER+ /PgR2 ER2 /PgR2 Unknown Ki67 £15% >15% Unknown c-erB-2 Negative 1+
Total, n (% of total)
ALND, n (% by arm)
SLNB, n (% by arm)
58.3 6 10.0 59.0 28–75
58.0 6 10.2 59.0 28–75
58.6 6 9.9 60.0 35–75
P
(13.8) (20.4) (37.4) (28.4)
15 24 44 32
(13.0) (20.9) (38.3) (27.8)
16 22 40 32
Characteristic
Total, n (% of total)
ALND, n (% by arm)
SLNB, n (% by arm)
2+ 3+ Unknownc
2 (0.9) 11 (4.9) 62 (27.5)
1 (0.9) 7 (6.9) 32 (27.8)
1 (0.9) 4 (3.6) 30 (27.3)
b
Pb
a
0.978 31 46 84 64
Table 1. (Continued)
(14.5) (20.0) (36.4) (29.1)
ALND = sentinel lymph node (SLN) biopsy followed by axillary dissection; SLNB = sentinel lymph node biopsy followed by axillary dissection, if SLN found positive. b Pearson chi-square test for heterogeneity. c In these patients, c-erB-2 status was not determined because the assay was not yet routinely carried out at the time of the study. ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy; ER, estrogen receptor; PgR, progesterone receptor.
0.622 121 (53.8) 104 (46.2)
60 (52.2) 55 (47.8)
61 (55.5) 49 (44.5) 0.146
109 (48.4) 55 (24.4) 61 (27.2)
63 (54.8) 24 (20.9) 28 (24.3)
46 (41.8) 31 (28.2) 33 (30.0) 0.897
217 3 3 2
(96.5) (1.3) (1.3) (0.9)
110 2 2 1
(95.7) (1.7) (1.7) (0.9)
107 1 1 1
(97.3) (0.9) (0.9) (0.9)
2 2 21 42 116 42
(0.9) (0.9) (9.3) (18.7) (51.5) (18.7)
1 0 10 18 57 29
(0.9) (0) (8.6) (15.7) (49.6) (25.2)
1 2 11 24 59 13
(0.9) (1.8) (10.0) (21.8) (53.6) (11.8)
156 16 38 11 4
(69.3) (7.1) (16.9) (4.9) (1.8)
79 11 17 5 3
(68.7) (9.6) (14.8) (4.3) (2.6)
77 5 21 6 1
(70.0) (4.5) (19.1) (5.5) (0.9)
12 90 119 4
(5.3) (40.0) (52.9) (1.8)
7 46 60 2
(6.1) (40.0) (52.2) (1.7)
5 44 59 2
(4.6) (40.0) (53.6) (1.8)
0.112
discussion
0.451
0.965
0.427 135 2 51 29 8
(60.0) (0.9) (22.6) (12.9) (3.6)
65 0 29 16 5
(56.6) (0) (25.2) (13.9) (4.3)
70 2 22 13 3
(63.7) (1.8) (20.0) (11.8) (2.7) 0.430
99 (44.0) 110 (48.9) 16 (7.1)
47 (40.9) 58 (50.4) 10 (8.7)
52 (47.3) 52 (47.3) 6 (5.4)
143 (63.6) 7 (3.1)
71 (61.7) 4 (3.5)
72 (65.5) 3 (2.7)
0.921
1004 | Canavese et al.
Patients with negative or micrometastatic SLN showed a similar incidence of events (Figure 2B, 5-year EFS of 93.5% and 93.8%, respectively, P = 1.0) whereas this rate was significantly higher in patients with macrometastatic SLN (5year EFS of 85.9%, P = 0.012). Over a mean follow-up period of 5.6 6 1.3 years (median of 5.6 and 5.7 years in the ALND arm and SLNB arm, respectively), nine patients died among whom four in the ALND arm and five in the SLNB arm. Five-year OS was 97.2% (95% CI 95.4% to 98.9%) in both arms (Figure 3, log-rank P = 0.697).
The accuracy of the SLNB procedure in predicting the axillary status in patients with early-stage breast cancer and clinically negative lymph nodes has been demonstrated in at least four studies [3, 4, 7, 8]. Based on their results, in these patients, a negative SLN is considered sufficient to rule out metastases in other axillary nodes and to avoid further axillary dissection. This leads to reduced short-term morbidity and improved quality of life [6, 9, 10]. Today most clinical guidelines [e.g. 11] include appropriate SLN identification and sampling procedure in their recommendations. Yet, information on the long-term effects of a surgical strategy based on SNLB, as compared with routine ALND, is still inadequate. The present study was designed at a time when this information was not available at all. It was designed as a noninferiority trial analyzing theoretically 2750 enrolled patients. Despite an accrual much lower than what had been prospected, thus rendering the study underpowered, the results provide further support to the hypothesis that SLNB not only reduces the need for complete axillary dissection without compromising long-term outcome but also confirms that crucial questions still need to be answered. The sensititvity of the SLN procedure (77.1%) and the negative predictive value (91.1%) we observed were lower than those reported in two other randomized trials (>96.0% and >95%, respectively) [12, 13]. However, a sensitivity as low as 70% has also been reported [14, 15]. These differences may be due to chance because of the small size of our own study. Other
Volume 20 | No. 6 | June 2009
original article
Annals of Oncology
Table 2. Status of sentinel and axillary lymph nodes by treatment arm Results of SLN evaluation during surgery
ALND armb (n = 115) Postd Intrac
Total SLN (n = 225)a
Not visualized
3 (1.3%)
1
–
Micrometastasis
9 (4.0%)
6
neg 0 pos 6
Macrometastasis
Negative
40 (17.8%)
18
173 (76.9%)
neg 0 pos 18
90
neg 87 pos 3h
Test
Sensitivity Specificity Negative predictive value Positive predictive value
68.6% 100% 87.8%
Metastasis in other axillary nodes neg pos – neg pos – neg pos neg pos neg pos
0 1
SLNB armb (n = 110) Intrac Postd 2
–
3
neg 0 pos 3
22
neg 0 pos 22
83
neg 77
6 0 8 10 79e 8g 3 0
pos 6h
77.1%
– – –
91.1%
100%
Metastasis in other axillary nodes neg 0 pos 2 – neg 1 pos 2 – neg 12 pos 10 n. e.f neg 5 pos 1
– –
–
a
In three patients, the SLN was not visualized and was thus not staged. ALND = sentinel lymph node biopsy followed by axillary dissection; SLNB = sentinel lymph node biopsy followed by axillary dissection, if SLN found positive. c Intrasurgery diagnosis: evaluation on frozen sections, when available. d Postsurgery diagnosis: evaluation on paraffin-embedded sections. e True negative: patients with SLN negative both at intra- and postsurgery, without metastases in the axillary nodes. f Not evaluable. g True false negative: patients with SLN negative both at intra- and postsurgery evaluation, but with metastasis in other axillary nodes. h False negative: patients with SLN negative at intrasurgery evaluation but positive at postsurgery diagnosis. SLN, sentinel lymph node; ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy; neg, negative; pos, positive. b
Table 3. Unfavorable events by treatment arm: 115 patients in the ALND arm and 110 patients in the SLNB arma Type of recurrence
Ipsilateral breast Controlateral breast Ipsilateral axillary MTS Distant metastasesd Death without disease recurrence Total
Event, n ALND
SLNB
Annual rate per 1000 patients, n ALND
0 3 1 6 2 12
1 1 0 6 2 10
– 4.8 1.6 9.6 3.2 18.6
Risk ratio (95% CI)b SLNB 1.7 1.7 – 9.9 3.3 16.2
– 0.35 – 1.04 1.04 0.87
Pc
(0.04–3.33)
0.336
(0.34–3.22) (0.15–7.38) (0.38–2.01)
0.947 0.969 0.741
a
ALND = sentinel lymph node (SLN) biopsy followed by axillary dissection; SLNB = sentinel lymph node biopsy followed by axillary dissection, if SLN found positive. b Risk ratio: SLNB arm versus ALND arm. c Pearson chi-square test. d One patient in the ALND group had distant metastases together with a controlateral relapse.
factors influencing the sensitivity are higher patient age (>50 years) often associated with more fatty nodes, which prevents retention of the radioactive tracer [4], low number of SLNs resected [13, 16, 17], location of the tumor in the outer quadrant of the breast [4, 13, 17], presence of heavily metastatic
Volume 20 | No. 6 | June 2009
nodes causing a lymphatic blockage and deviating the tracer to nodes free of disease [4, 13]. In the ALND arm, all the eight patients with a negative SLN, as confirmed at postoperative examination, but with metastases in other axillary nodes had one or more of these characteristics.
doi:10.1093/annonc/mdn746 | 1005
original article
Annals of Oncology
Figure 2. Panel (A) Breast cancer-related events. ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy; P value from log-rank test (two sided), 0.715. Panel (B): Breast cancer-related events in combined ALND and SLNB arms according to the status of the sentinel lymph node. MicroMTS, micrometastasis; MacroMTS, macrometastasis; SLN, sentinel lymph node; N, number of patients; EFS, event-free survival; 95% CI, 95% confidence intervals. P values: MacroMTS versus MicroMTS, Fisher’s exact test P = 0.192; MicroMTS versus negative, Fisher’s exact test P = 1.0; MacroMTS versus negative, log-rank test P = 0.012.
Yet, it must be underlined that this finding of eight falsenegative patients (8.9%) of the 90 patients with a negative SLN in the ALND arm, together with the findings from other trials, indicates that between 1 of 10 and 1 of 20 patients not undergoing axillary dissection because of a negative SLN is indeed node positive. In terms of sensitivity, this implies that more than one-fifth of the patients with axillary metastases in the ALND arm are not recognized by SLN. In many of these cases, adjuvant treatments will either not be used or modified because of the falsely negative result. It has to be considered that the relative efficacy of adjuvant treatments appears to be independent of nodal status and that the indication to limit their use to node-positive (and high-risk node negative) patients is dictated only by the low risk of relapse of most nodenegative patients and by the consequent imbalance between the benefits and the toxicity of these treatments. Should it be confirmed that SLN-negative, node-positive women carry a risk of relapse comparable to that of all node-positive women, it could be speculated that these women are denied an effective treatment based on a false-negative SLN result. This consideration has three implications: (i) All case series and trials in which both SLNB and ALND were carried out systematically in all patients should be reviewed, to assess if the absolute risk of relapse of SLN-negative, ALND-positive patients resembles that of node-negative patients or approaches that of node-positive patients. (ii) It should be clear that the overall results of trials comparing groups of women undergoing either procedure, though reassuring, did not have the power to detect relapse-free-survival differences arising from a different use of adjuvant therapies in false-negative patients. In fact, if we assume that adjuvant treatments are associated with a risk reduction of 25% and that 7% of all randomized patients with a false-negative SLN do not receive adjuvant therapy in the SLNB arm and do receive it in the ALND arm, we can estimate a relative risk reduction in the overall study population of <2%. Differences of this order of magnitude require trials of a size
1006 | Canavese et al.
Figure 3. Kaplan–Meier curves for overall survival. ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy; N, number of patients at risk; OS, 5-year overall survival with 95% confidence intervals (CIs) in parentheses; P value from log-rank test (two sided), 0.679.
much larger than that of those supporting the safety of SLN procedure. (iii) Based on these considerations, predictors of axillary node positivity should be developed, in order to limit the use of the SLN procedure to women at low risk of bearing metastases in the axillary nodes. This would not affect the sensitivity of the procedure, but would increase its negative predictive value and the unfavorable consequences of its use. A main issue still debated is the clinical significance of micrometastases in the SLN and their prognostic value on
Volume 20 | No. 6 | June 2009
Annals of Oncology
recurrence and survival. Of the 58 patients with a positive SLN, 16 had micrometastases and 14 of them having a tumor >1 cm. All 16 patients underwent axillary dissection and at least one additional lymph node involved was found in 11 cases (68.8%). This finding strengthens the recommendation to carry out routine axillary dissection in this subgroup of patients even if, in the present setting, they showed a rate of breast-related events similar to that of SLN negative (Figure 2B). The overall annual rate of unfavorable events (Table 3), the cumulative incidence of events (Figure 2A) and the OS (Figure 3) were not statistically different between the two arms. These results go in line with those from previous randomized trials [6, 15, 18] or nonrandomized studies [19, 20], but in light of the above considerations, are not fully reassuring since they might conceal a nonnegligible increase in risk in a tiny proportion of women who do not receive full adjuvant therapies because of a falsely negative SNL procedure. Also the assessment of the true diagnostic accuracy of the SLN procedure outside clinical trials would be desirable, but it may prove impossible to obtain, for ethical reasons, as it would require systematic axillary dissection in large series of SLN-negative patients. In conclusion, the results of our settings, in terms of EFS and OS, go in line with the concept that SLNB is an efficient tool for staging assessment of early breast cancer in this subset of patients with a tumor dimension of <3 cm at presurgery diagnosis and with clinically node-negative disease. In particular, the low recurrence rate in the SLNB arm, comparable to that following conventional ALND, supports the emerging consensus that complete axillary dissection could be safely omitted in SLN-negative patients, thus sparing them the known immediate sequelae of axillary surgery. However, pooled long-term results of randomized studies are awaited to definitely confirm the comparability of SLNB and ALND in terms of OS and incidence of nodal failure and the improved quality of life associated with SLNB.
references 1. Pater J, Parulekar W. Sentinel lymph node biopsy in early breast cancer: has its time come? J Natl Cancer Inst 2006; 98: 568–569. 2. Krag DN, Weaver DL, Alex JC, Fairbank JT. Surgical resection and radiolocalization of the sentinel lymph node in breast cancer using a gamma probe. Surg Oncol 1993; 2: 335–339. 3. Giuliano AE, Kirgan DM, Guenther JM, Morton DL. Lymphatic mapping and sentinel lymphadenectomy for breast cancer. Ann Surg 1994; 220: 391–401.
Volume 20 | No. 6 | June 2009
original article 4. Krag D, Weaver D, Ashikaga T et al. The sentinel node in breast cancer—a multicenter validation study. N Engl J Med 1998; 339: 941–946. 5. Krag DN. The sentinel node for staging breast cancer: current review. Breast Cancer 1999; 6: 233–236. 6. Mansel RE, Fallowfield L, Kissin M et al. Randomized multicenter trial of sentinel node biopsy versus standard axillary treatment in operable breast cancer: the ALMANAC Trial. J Natl Cancer Inst 2006; 98: 599–609. 7. Veronesi U, Paganelli G, Viale G et al. Sentinel lymph node biopsy and axillary dissection in breast cancer: results in a large series. J Natl Cancer Inst 1999; 91: 368–373. 8. Albertini JJ, Lyman GH, Cox C et al. Lymphatic mapping and sentinel node biopsy in the patient with breast cancer. JAMA 1996; 276: 1818–1822. 9. Giuliano AE, Haigh PI, Brennan MB et al. Prospective observational study of sentinel lymphadenectomy without further axillary dissection in patients with sentinel node-negative breast cancer. J Clin Oncol 2000; 18: 2553–2559. 10. Peintinger F, Reitsamer R, Stranzl H, Ralph G. Comparison of quality of life and arm complaints after axillary lymph node dissection vs sentinel lymph node biopsy in breast cancer patients. Br J Cancer 2003; 89: 648–652. 11. Lyman GH, Giuliano AE, Somerfield MR et al. American Society of Clinical Oncology guideline recommendations for sentinel lymph node biopsy in earlystage breast cancer. J Clin Oncol 2005; 23: 7703–7720. 12. Veronesi U, Paganelli G, Viale G et al. A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer. N Engl J Med 2003; 349: 546–553. 13. Krag DN, Anderson SJ, Julian TB et al. Technical outcomes of sentinel-lymphnode resection and conventional axillary-lymph-node dissection in patients with clinically node-negative breast cancer: results from the NSABP B-32 randomised phase III trial. Lancet Oncol 2007; 8: 881–888. 14. Kim T, Giuliano AE, Lyman GH. Lymphatic mapping and sentinel lymph node biopsy in early-stage breast carcinoma: a metaanalysis [review]. Cancer 2006; 106: 4–16. 15. Zavagno G, De Salvo GL, Scalco G et al. A randomized clinical trial on sentinel lymph node biopsy versus axillary lymph node dissection in breast cancer: results of the Sentinella/GIVOM trial. Ann Surg 2008; 247: 207–213. 16. Wong SL, Edwards MJ, Chao C et al. Sentinel lymph node biopsy for breast cancer: impact of the number of sentinel nodes removed on the false-negative rate. J Am Coll Surg 2001; 192: 684–689; discussion 689–691. 17. Martin RC II, Chagpar A, Scoggins CR et al. Clinicopathologic factors associated with false-negative sentinel lymph-node biopsy in breast cancer. Ann Surg 2005; 241: 1005–1012; discussion 1012–1015. 18. Veronesi U, Paganelli G, Viale G et al. Sentinel-lymph-node biopsy as a staging procedure in breast cancer: update of a randomised controlled study. Lancet Oncol 2006; 7: 983–990. 19. Takei H, Suemasu K, Kurosumi M et al. Sentinel lymph node biopsy alone has no adverse impact on the survival of patients with breast cancer. Breast J 2006; 12 (5 Suppl 2): S157–S164. 20. Cox C, White L, Allred N et al. Survival outcomes in node-negative breast cancer patients evaluated with complete axillary node dissection versus sentinel lymph node biopsy. Ann Surg Oncol 2006; 13: 708–711.
doi:10.1093/annonc/mdn746 | 1007
Annals of Oncology 20: 1001–1007, 2009 doi:10.1093/annonc/mdn746 Published online 27 January 2009
Sentinel node biopsy compared with complete axillary dissection for staging early breast cancer with clinically negative lymph nodes: results of randomized trial G. Canavese1, A. Catturich1, C. Vecchio1, D. Tomei1, M. Gipponi2, G. Villa3, F. Carli4, P. Bruzzi5 & B. Dozin5* 1 5
Surgical Senology Unit, National Cancer Research Institute; Departments of 2General Surgery; 3Nuclear Medicine, A.O.U San Martino; 4Anatomy and Cytohistology Unit; Clinical Epidemiology Unit, National Cancer Research Institute, Genoa, Italy
Received 4 August 2008; revised 5 November 2008; accepted 6 November 2008
cancer patients with negative SLNs. Evidence of a similar efficacy, in terms of survival and regional control, of this strategy as compared with axillary resection is based on few clinical trials. In 1998, we started a randomized study comparing the two strategies, and we present here its results. Materials and methods: Patients were randomly assigned to sentinel lymph node biopsy (SLNB) and axillary dissection [axillary lymph node dissection (ALND arm)] or to SLNB plus axillary resection if SLNs contained metastases (SLNB arm). Main end points were overall survival (OS) and axillary recurrence. Results: One hundred and fifteen patients were assigned to the ALND arm and 110 to the SLNB arm. A positive SLN was found in 27 patients in the ALND arm and in 31 in the SLNB arm. Overall accuracy of SLNB was 93.0%. Sensitivity and negative predictive values were 77.1% and 91.1%, respectively. At a median follow-up of 5.5 years, no axillary recurrence was observed in the SLNB arm. OS and event-free survival were not statistically different between the two arms. Conclusions: The SLNB procedure does not appear inferior to conventional ALND for the subset of patients here considered. Key words: axillary lymph node dissection, breast cancer, overall survival, randomized clinical trial, regional node recurrence, sentinel lymph node biopsy
introduction Until few years ago, axillary lymph node dissection (ALND), together with resection of the primary tumor, was the standard procedure used in the surgical management of breast cancer patients, as well as the routine method for axillary node staging. The pathological stage of the axillary nodes represents a major prognostic factor in early breast cancer and is crucial in the decision on the need and type of adjuvant therapy. However, complete axillary dissection often results in significant physical morbidities and psychological distress [1]. A decade ago, the technique of visualization, identification and intraoperative analysis of the sentinel lymph node (SLN) opened new perspectives in the management of early breast cancer patients [2, 3]. From these early studies, it appeared that SLN staging could predict the status of the axillary nodes and that patients with negative SLNs may be spared total axillary dissection [4, 5]. At the time the present study was started, the evidence in support of the use of the SNL procedure in place of *Correspondence to: Dr B. Dozin, Clinical Epidemiology Unit, National Cancer Research Institute, Largo Rosanna Bensi 10, 16132 Genoa, Italy. Tel: +39-010-5737492; Fax: +39-010-354103; E-mail: [email protected]
standard axillary dissection was inadequate for the standards currently used to introduce a new technology in routine clinical practice. We thus carried out, in a population of early breast cancer patients, a randomized comparison of sentinel lymph node biopsy (SLNB) followed by routine axillary dissection with SLNB followed by axillary dissection when the SLN was found metastatic. The study was designed as a noninferiority trial. Although it is underpowered, the results support the evidence of the efficacy of SLNB for axillary staging in early breast cancer and indicate that this procedure is equivalent to routine axillary dissection, in terms both of overall survival (OS) and regional recurrence. Noteworthy, even today, while the beneficial short-term effects of SNL on quality of life have been clearly demonstrated [6], reliable information on longterm morbidity and mortality associated with its use is still lacking [1].
materials and methods patient population The present single-center randomized study was conducted at the National Cancer Research Institute of Genoa, Italy. Patients were enrolled according
ª The Author 2009. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected]
original article
Background: Sentinel lymph node (SLN) staging is currently used to avoid complete axillary dissection in breast
original article to the following eligibility criteria: age 18–75 years, primary invasive breast carcinoma as revealed by mammography and cytohistology, clinically negative axillary lymph nodes and unifocal tumor sized £3 cm as estimated by ecography. At the time we started this study, on the basis of the SLN validation study of Krag et al. [4], we considered this size as the cutoff dimension above which patients were to undergo standard axillary dissection because of their higher risk of bearing metastatic lymph nodes. Patients were also excluded from the study if they had undergone previous surgery on the same breast or on the ipsilateral axilla or if they were suffering from chronic life-threatening disease possibly preventing adjuvant therapy.
study design, randomization procedure and end points Patients were randomly assigned to undergo, together with breast surgery, either SLNB and ALND or SLNB followed by ALND only if the SLN was found positive at the intraoperative evaluation. Randomization was carried out at the time of surgery by phoning the data-coordinating center at the Epidemiology and Clinical Trials Unit of the Institute where the data manager verified patient eligibility and communicated the allocated treatment. Treatment allocation was carried out in a 1 : 1 ratio according to a computer-generated randomization list. The study had been approved by the Ethical Committee of the Institute and was conducted in accordance with the International Good Clinical Practice Guidelines. All patients were aware of the aims and procedures of the study and signed an informed consent. The main objective of the study was the comparison of event-free survival (EFS) and OS between the two arms at 5 years from surgery. An additional endpoint was the evaluation of the frequency of axillary recurrences in patients of the SLNB arm and did not undergo axillary dissection. Secondary analyses, conducted in the ALND arm, focused on the evaluation of the sensitivity and predictive value of the SLN procedure with respect to the presence of axillary metastases.
SLN identification and intraoperative evaluation All patients underwent SLN dissection and intraoperative evaluation. The SLN was identified both by breast lymphoscintigraphy and lymphatic dye mapping. A peritumoral injection of 0.2 mCi of 99mTechnetium (Nanocol, Amersham-Sorin Biomedica, Saluggia, Italy) was carried out the day before surgery; an additional peritumoral s.c. injection of blue dye (Patent Blue-V) was carried out at the time of surgery. Through a small axillary incision, the radioactive SLN was localized with a c-ray detecting probe and removed for immediate intraoperative search for metastases. The SLN was bisected along its major axis and five pairs of frozen sections, each 4-lm thick, were cut every 10 lm in each half of the node. The first, third and fifth sections were stained with hematoxylin–eosin. If this histological evaluation resulted negative or ambiguous, the second and the fourth sections were tested by immunohistochemistry for the presence of cytokeratins [EPOS (Enhanced Peroxidase One-Step) method with cytokeratin MNF116 mAb and horseradish peroxidase (Dako, Milan, Italy)]. The remaining tissue was paraffin embedded for definitive postoperative evaluation. Any patient whose SLNs could not be identified was subjected to axillary dissection, independently of the treatment assigned.
surgery, adjuvant therapy and follow-up Mastectomy or conservative quadrantectomy was carried out according to standard criteria. Only patients who underwent conservative surgery received radiotherapy to the ipsilateral breast (50-Gy dose over 8 weeks). No radiotherapy to the axilla was carried out. The choice of adjuvant chemo- and/or hormone therapy regimen, when given, was based on the main prognostic factors of the primary tumor (nodal status, tumor size, tumor grading, hormonal receptor status). Follow-up included clinical examinations every 4 months for the first 3 years, every 6 months for the next 2 years and once a year thereafter. A
1002 | Canavese et al.
Annals of Oncology
mammography was scheduled annually. In case of clinically suspected lymph nodes, an axillary ecotomography was carried out, followed by a needle biopsy aspiration for cytological examination when deemed necessary.
statistics The main objective of the study was to evaluate whether the choice of not carrying out axillary dissection in patients with intraoperative negative SLN could ensure to those patients a prognosis equivalent to that expected for patients undergoing routine axillary dissection. The primary end point was EFS, and sample size estimates were based on an expected 5 years of 80% in the control ALND arm based on published data and on our own previous experience. The study was designed as a noninferiority trial, with the aim of rejecting the null hypothesis that the experimental treatment was associated with an absolute decrease in disease-free survival of 4% (hazard ratio = 1.21) due to the reduction in the use of adjuvant chemotherapy in falsenegative patients in the SNLB arm. Setting a study power at 90% for a onesided a error of 0.025 and considering both accrual and follow-up periods of 5 years, together with an estimated annual lost-to-follow-up rate of 2%, we established that 2570 patients were to be enrolled in the study. EFS was calculated from the date of surgery to the date of occurrence of any of local, controlateral, regional node or distant relapse or death from any cause, whichever came first. OS was estimated as the time interval between the date of surgery and the date of death from any cause. In both analyses, survival times of patients free from relapse and/or alive at last examination were censored at that date. EFS and OS curves were obtained using the Kaplan–Meier method, and differences between arms were assessed by the log-rank test. Differences between the two arms in baseline prognostic factors were assessed by the Pearson chi-square test or the Fisher’s exact test. Two-sided P values and/or 95% confidence intervals (CIs) are reported. All statistical analyses were based on the intention-to-treat approach, in that all patients, after the exclusion of ineligible patients, were included in the treatment arm assigned at randomization, regardless of the treatment actually received. Analyses were carried out using the SPSS package (version 14.0 for Windows)
results patients’ characteristics From November 1998 to October 2001, 248 patients were consecutively enrolled in the study and 124 were randomly assigned to each of the study arms (Figure 1). This number is far below the projections indicated above; indeed, enrollment was prematurely interrupted when patients became aware of the early promising results of the SLNB technique [5, 7] as they were increasingly refusing randomization in a study where routine axillary dissection was one of the alternative treatments. In each arm, nine patients were excluded from evaluation because of an inadequate presurgery cytology that should have precluded their eligibility. In the SLNB arm, three more patients were excluded because the intraoperative examination revealed the presence of a multifocal lesion. In the same arm, two more patients who were eligible could not be evaluated because postsurgical pathology revealed that the lesion was benign. The final number of assessable patients was 225, among whom 115 (51.1%) were allocated to the ALND arm and 110 (48.9%) to the SLNB arm. Among these 225 patients, only two underwent total mastectomy; all the others received conservative breast surgery. The main patient and tumor characteristics are summarized in Table 1. It should be noted that two patients, one in each
Volume 20 | No. 6 | June 2009
Annals of Oncology
Figure 1. Trial flow chart. ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy.
arm, were staged pTis upon definitive diagnosis; these two patients were maintained for final evaluation because at randomization they were meeting the eligibility criteria. Patient distribution in the two arms was similar with respect to all the prognostic factors (P values: 0.112–0.978).
identification of the SLN and evaluation of the axillary status A total of 405 SLNs were removed from the 225 patients (mean 1.80 per patient), 211 nodes in the ALND arm (mean 1.83 per patient) and 194 nodes in the SLNB arm (mean 1.76 per patient). A mean of 12.7 nonsentinel nodes per patient were removed from the 148 patients who underwent axillary dissection. Through intraoperative scintigraphy, no radioactive SLN was revealed in three patients (Table 2). All three patients, including two allocated to the SLNB arm, underwent routine axillary dissection, and macrometastases were found in the axilla in all of them. Upon intraoperative evaluation, no isolated tumor cells (groups of cells £ 0.2 mm) were found in any of the SLNs analyzed. Instead, micrometastasis (foci > 0.2 mm but £2 mm) and macrometastases (foci > 2 mm) in the SLN were found in 9 and 40 patients, respectively. In all these cases, the intraoperative diagnosis was confirmed at the definitive histological examination. In the ALND arm, none of the six patients with micrometastatic SLN and 10 (55.6%) of the 18 patients with macrometastatic SLN presented metastases in the axilla. As for negative patients, in the ALND arm, we have distinguished three classes of negative patients with respect to the SLN status: (i) 79 true-negative patients with a negative SLN both at intra- and postoperative examination and no
Volume 20 | No. 6 | June 2009
original article metastasis in the axillary nodes; (ii) three patients whose SLN was negative at intraoperative evaluation but was found positive with micrometastases at the definitive examination and (iii) eight true false-negative patients with a negative SLN both at intra- and postoperative diagnosis, but with metastases in the axilla: Out of them, two patients had axillary micrometastases, two patients had one macrometastatic axillary node and four patients had more than one node involved. In the SLNB arm, two (66.7%) of the three patients with micrometastatic SLN and 10 (45.5%) of the 22 patients with macrometastatic SLN had also metastatic axillary nodes. In the SLNB arm, in six patients, SLN was negative at intraoperative evaluation but was found positive with micrometastases at the definitive examination. Among them, four and two patients had micro- and macrometastases, respectively. To summarize the number of axillary dissections carried out in the whole population, all 115 patients in the ALND arm and 33 patients (30%) of the 110 patients in the SLNB arm underwent the procedure (two patients with SLN not visualized, six patients with false-negative SLN and 25 patients with a SLN found positive). As scheduled, the remaining 77 (70%) patients in the SLNB arm with a negative SLN at definitive examination were spared complete axillary dissection. Overall, in the ALND arm, the procedure of intraoperative evaluation of the SLN had a sensitivity of 68.6% and a specificity of 100%. Taking into account the three SLNs that resulted positive only at the definitive evaluation, sensitivity was 77.1% (27 of 35 patients with positive nodes were identified). The negative and positive predictive values were 91.1% and 100%, respectively. The overall accuracy of the SLNB procedure was 93.0% (correct in 106 of the 114 assessable patients).
EFS and OS At a median follow-up of 5.5 6 1.4 years, 22 unfavorable events occurred, 12 in the ALND arm and 10 in the SLNB arm (Table 3). One ipsilateral breast recurrence was observed in the SLNB arm. It was treated with surgery alone. Contralateral primary tumors developed in four patients, three in the ALND arm and one in the SLNB arm. All these five patients were still alive at the end of the observation period. No overt axillary metastasis was observed in the SLNB arm. In the ALND arm, axillary recurrence occurred in only one patient 3.1 years after surgery; this patient died 161 days after diagnosis of the recurrence. Six patients in each arm developed distant metastases. Four of them (one in the ALND arm and three in the SLNB arm) died at 153 days, 148 days, 10 months and 4.7 years from relapse, respectively. Four more patients, two in each arm, died without disease recurrence. The overall annual rate of events was 18.6 per 1000 patients in the ALND arm and 16.2 per 1000 patients in the SLNB arm (risk ratio of SLNB versus ALND = 0.87, 95% CI 0.38–2.01, P = 0.741). Accordingly, 5-year EFS was not statistically different between the two arms (Figure 2A; ALND arm: 89.8%, 95% CI 86.9% to 92.7%; SLNB arm: 94.5%, 95% CI 90.9% to 98.1%; log-rank P = 0.715).
doi:10.1093/annonc/mdn746 | 1003
original article
Annals of Oncology
Table 1. Patient characteristics by treatment arm: 115 patients in the ALND arm and 110 patients in the SLNB arma Characteristic Age (years) Mean Median Range (minimum– maximum) Age by category £45 years 46–55 years 56–65 years >65 years Tumor site Right Left Quadrant site Outer Inner or central >1 Quadrant Histology Ductal NOS Lobular Other In situ Pathological T pTis pT1mic pT1a pT1b pT1c pT2 Pathological N pN0 pN1mic pN1a pN2a pN3a Grading G1 G2 G3 Unknown Hormonal receptor status ER+ /PgR+ ER2 /PgR+ ER+ /PgR2 ER2 /PgR2 Unknown Ki67 £15% >15% Unknown c-erB-2 Negative 1+
Total, n (% of total)
ALND, n (% by arm)
SLNB, n (% by arm)
58.3 6 10.0 59.0 28–75
58.0 6 10.2 59.0 28–75
58.6 6 9.9 60.0 35–75
P
(13.8) (20.4) (37.4) (28.4)
15 24 44 32
(13.0) (20.9) (38.3) (27.8)
16 22 40 32
Characteristic
Total, n (% of total)
ALND, n (% by arm)
SLNB, n (% by arm)
2+ 3+ Unknownc
2 (0.9) 11 (4.9) 62 (27.5)
1 (0.9) 7 (6.9) 32 (27.8)
1 (0.9) 4 (3.6) 30 (27.3)
b
Pb
a
0.978 31 46 84 64
Table 1. (Continued)
(14.5) (20.0) (36.4) (29.1)
ALND = sentinel lymph node (SLN) biopsy followed by axillary dissection; SLNB = sentinel lymph node biopsy followed by axillary dissection, if SLN found positive. b Pearson chi-square test for heterogeneity. c In these patients, c-erB-2 status was not determined because the assay was not yet routinely carried out at the time of the study. ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy; ER, estrogen receptor; PgR, progesterone receptor.
0.622 121 (53.8) 104 (46.2)
60 (52.2) 55 (47.8)
61 (55.5) 49 (44.5) 0.146
109 (48.4) 55 (24.4) 61 (27.2)
63 (54.8) 24 (20.9) 28 (24.3)
46 (41.8) 31 (28.2) 33 (30.0) 0.897
217 3 3 2
(96.5) (1.3) (1.3) (0.9)
110 2 2 1
(95.7) (1.7) (1.7) (0.9)
107 1 1 1
(97.3) (0.9) (0.9) (0.9)
2 2 21 42 116 42
(0.9) (0.9) (9.3) (18.7) (51.5) (18.7)
1 0 10 18 57 29
(0.9) (0) (8.6) (15.7) (49.6) (25.2)
1 2 11 24 59 13
(0.9) (1.8) (10.0) (21.8) (53.6) (11.8)
156 16 38 11 4
(69.3) (7.1) (16.9) (4.9) (1.8)
79 11 17 5 3
(68.7) (9.6) (14.8) (4.3) (2.6)
77 5 21 6 1
(70.0) (4.5) (19.1) (5.5) (0.9)
12 90 119 4
(5.3) (40.0) (52.9) (1.8)
7 46 60 2
(6.1) (40.0) (52.2) (1.7)
5 44 59 2
(4.6) (40.0) (53.6) (1.8)
0.112
discussion
0.451
0.965
0.427 135 2 51 29 8
(60.0) (0.9) (22.6) (12.9) (3.6)
65 0 29 16 5
(56.6) (0) (25.2) (13.9) (4.3)
70 2 22 13 3
(63.7) (1.8) (20.0) (11.8) (2.7) 0.430
99 (44.0) 110 (48.9) 16 (7.1)
47 (40.9) 58 (50.4) 10 (8.7)
52 (47.3) 52 (47.3) 6 (5.4)
143 (63.6) 7 (3.1)
71 (61.7) 4 (3.5)
72 (65.5) 3 (2.7)
0.921
1004 | Canavese et al.
Patients with negative or micrometastatic SLN showed a similar incidence of events (Figure 2B, 5-year EFS of 93.5% and 93.8%, respectively, P = 1.0) whereas this rate was significantly higher in patients with macrometastatic SLN (5year EFS of 85.9%, P = 0.012). Over a mean follow-up period of 5.6 6 1.3 years (median of 5.6 and 5.7 years in the ALND arm and SLNB arm, respectively), nine patients died among whom four in the ALND arm and five in the SLNB arm. Five-year OS was 97.2% (95% CI 95.4% to 98.9%) in both arms (Figure 3, log-rank P = 0.697).
The accuracy of the SLNB procedure in predicting the axillary status in patients with early-stage breast cancer and clinically negative lymph nodes has been demonstrated in at least four studies [3, 4, 7, 8]. Based on their results, in these patients, a negative SLN is considered sufficient to rule out metastases in other axillary nodes and to avoid further axillary dissection. This leads to reduced short-term morbidity and improved quality of life [6, 9, 10]. Today most clinical guidelines [e.g. 11] include appropriate SLN identification and sampling procedure in their recommendations. Yet, information on the long-term effects of a surgical strategy based on SNLB, as compared with routine ALND, is still inadequate. The present study was designed at a time when this information was not available at all. It was designed as a noninferiority trial analyzing theoretically 2750 enrolled patients. Despite an accrual much lower than what had been prospected, thus rendering the study underpowered, the results provide further support to the hypothesis that SLNB not only reduces the need for complete axillary dissection without compromising long-term outcome but also confirms that crucial questions still need to be answered. The sensititvity of the SLN procedure (77.1%) and the negative predictive value (91.1%) we observed were lower than those reported in two other randomized trials (>96.0% and >95%, respectively) [12, 13]. However, a sensitivity as low as 70% has also been reported [14, 15]. These differences may be due to chance because of the small size of our own study. Other
Volume 20 | No. 6 | June 2009
original article
Annals of Oncology
Table 2. Status of sentinel and axillary lymph nodes by treatment arm Results of SLN evaluation during surgery
ALND armb (n = 115) Postd Intrac
Total SLN (n = 225)a
Not visualized
3 (1.3%)
1
–
Micrometastasis
9 (4.0%)
6
neg 0 pos 6
Macrometastasis
Negative
40 (17.8%)
18
173 (76.9%)
neg 0 pos 18
90
neg 87 pos 3h
Test
Sensitivity Specificity Negative predictive value Positive predictive value
68.6% 100% 87.8%
Metastasis in other axillary nodes neg pos – neg pos – neg pos neg pos neg pos
0 1
SLNB armb (n = 110) Intrac Postd 2
–
3
neg 0 pos 3
22
neg 0 pos 22
83
neg 77
6 0 8 10 79e 8g 3 0
pos 6h
77.1%
– – –
91.1%
100%
Metastasis in other axillary nodes neg 0 pos 2 – neg 1 pos 2 – neg 12 pos 10 n. e.f neg 5 pos 1
– –
–
a
In three patients, the SLN was not visualized and was thus not staged. ALND = sentinel lymph node biopsy followed by axillary dissection; SLNB = sentinel lymph node biopsy followed by axillary dissection, if SLN found positive. c Intrasurgery diagnosis: evaluation on frozen sections, when available. d Postsurgery diagnosis: evaluation on paraffin-embedded sections. e True negative: patients with SLN negative both at intra- and postsurgery, without metastases in the axillary nodes. f Not evaluable. g True false negative: patients with SLN negative both at intra- and postsurgery evaluation, but with metastasis in other axillary nodes. h False negative: patients with SLN negative at intrasurgery evaluation but positive at postsurgery diagnosis. SLN, sentinel lymph node; ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy; neg, negative; pos, positive. b
Table 3. Unfavorable events by treatment arm: 115 patients in the ALND arm and 110 patients in the SLNB arma Type of recurrence
Ipsilateral breast Controlateral breast Ipsilateral axillary MTS Distant metastasesd Death without disease recurrence Total
Event, n ALND
SLNB
Annual rate per 1000 patients, n ALND
0 3 1 6 2 12
1 1 0 6 2 10
– 4.8 1.6 9.6 3.2 18.6
Risk ratio (95% CI)b SLNB 1.7 1.7 – 9.9 3.3 16.2
– 0.35 – 1.04 1.04 0.87
Pc
(0.04–3.33)
0.336
(0.34–3.22) (0.15–7.38) (0.38–2.01)
0.947 0.969 0.741
a
ALND = sentinel lymph node (SLN) biopsy followed by axillary dissection; SLNB = sentinel lymph node biopsy followed by axillary dissection, if SLN found positive. b Risk ratio: SLNB arm versus ALND arm. c Pearson chi-square test. d One patient in the ALND group had distant metastases together with a controlateral relapse.
factors influencing the sensitivity are higher patient age (>50 years) often associated with more fatty nodes, which prevents retention of the radioactive tracer [4], low number of SLNs resected [13, 16, 17], location of the tumor in the outer quadrant of the breast [4, 13, 17], presence of heavily metastatic
Volume 20 | No. 6 | June 2009
nodes causing a lymphatic blockage and deviating the tracer to nodes free of disease [4, 13]. In the ALND arm, all the eight patients with a negative SLN, as confirmed at postoperative examination, but with metastases in other axillary nodes had one or more of these characteristics.
doi:10.1093/annonc/mdn746 | 1005
original article
Annals of Oncology
Figure 2. Panel (A) Breast cancer-related events. ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy; P value from log-rank test (two sided), 0.715. Panel (B): Breast cancer-related events in combined ALND and SLNB arms according to the status of the sentinel lymph node. MicroMTS, micrometastasis; MacroMTS, macrometastasis; SLN, sentinel lymph node; N, number of patients; EFS, event-free survival; 95% CI, 95% confidence intervals. P values: MacroMTS versus MicroMTS, Fisher’s exact test P = 0.192; MicroMTS versus negative, Fisher’s exact test P = 1.0; MacroMTS versus negative, log-rank test P = 0.012.
Yet, it must be underlined that this finding of eight falsenegative patients (8.9%) of the 90 patients with a negative SLN in the ALND arm, together with the findings from other trials, indicates that between 1 of 10 and 1 of 20 patients not undergoing axillary dissection because of a negative SLN is indeed node positive. In terms of sensitivity, this implies that more than one-fifth of the patients with axillary metastases in the ALND arm are not recognized by SLN. In many of these cases, adjuvant treatments will either not be used or modified because of the falsely negative result. It has to be considered that the relative efficacy of adjuvant treatments appears to be independent of nodal status and that the indication to limit their use to node-positive (and high-risk node negative) patients is dictated only by the low risk of relapse of most nodenegative patients and by the consequent imbalance between the benefits and the toxicity of these treatments. Should it be confirmed that SLN-negative, node-positive women carry a risk of relapse comparable to that of all node-positive women, it could be speculated that these women are denied an effective treatment based on a false-negative SLN result. This consideration has three implications: (i) All case series and trials in which both SLNB and ALND were carried out systematically in all patients should be reviewed, to assess if the absolute risk of relapse of SLN-negative, ALND-positive patients resembles that of node-negative patients or approaches that of node-positive patients. (ii) It should be clear that the overall results of trials comparing groups of women undergoing either procedure, though reassuring, did not have the power to detect relapse-free-survival differences arising from a different use of adjuvant therapies in false-negative patients. In fact, if we assume that adjuvant treatments are associated with a risk reduction of 25% and that 7% of all randomized patients with a false-negative SLN do not receive adjuvant therapy in the SLNB arm and do receive it in the ALND arm, we can estimate a relative risk reduction in the overall study population of <2%. Differences of this order of magnitude require trials of a size
1006 | Canavese et al.
Figure 3. Kaplan–Meier curves for overall survival. ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy; N, number of patients at risk; OS, 5-year overall survival with 95% confidence intervals (CIs) in parentheses; P value from log-rank test (two sided), 0.679.
much larger than that of those supporting the safety of SLN procedure. (iii) Based on these considerations, predictors of axillary node positivity should be developed, in order to limit the use of the SLN procedure to women at low risk of bearing metastases in the axillary nodes. This would not affect the sensitivity of the procedure, but would increase its negative predictive value and the unfavorable consequences of its use. A main issue still debated is the clinical significance of micrometastases in the SLN and their prognostic value on
Volume 20 | No. 6 | June 2009
Annals of Oncology
recurrence and survival. Of the 58 patients with a positive SLN, 16 had micrometastases and 14 of them having a tumor >1 cm. All 16 patients underwent axillary dissection and at least one additional lymph node involved was found in 11 cases (68.8%). This finding strengthens the recommendation to carry out routine axillary dissection in this subgroup of patients even if, in the present setting, they showed a rate of breast-related events similar to that of SLN negative (Figure 2B). The overall annual rate of unfavorable events (Table 3), the cumulative incidence of events (Figure 2A) and the OS (Figure 3) were not statistically different between the two arms. These results go in line with those from previous randomized trials [6, 15, 18] or nonrandomized studies [19, 20], but in light of the above considerations, are not fully reassuring since they might conceal a nonnegligible increase in risk in a tiny proportion of women who do not receive full adjuvant therapies because of a falsely negative SNL procedure. Also the assessment of the true diagnostic accuracy of the SLN procedure outside clinical trials would be desirable, but it may prove impossible to obtain, for ethical reasons, as it would require systematic axillary dissection in large series of SLN-negative patients. In conclusion, the results of our settings, in terms of EFS and OS, go in line with the concept that SLNB is an efficient tool for staging assessment of early breast cancer in this subset of patients with a tumor dimension of <3 cm at presurgery diagnosis and with clinically node-negative disease. In particular, the low recurrence rate in the SLNB arm, comparable to that following conventional ALND, supports the emerging consensus that complete axillary dissection could be safely omitted in SLN-negative patients, thus sparing them the known immediate sequelae of axillary surgery. However, pooled long-term results of randomized studies are awaited to definitely confirm the comparability of SLNB and ALND in terms of OS and incidence of nodal failure and the improved quality of life associated with SLNB.
references 1. Pater J, Parulekar W. Sentinel lymph node biopsy in early breast cancer: has its time come? J Natl Cancer Inst 2006; 98: 568–569. 2. Krag DN, Weaver DL, Alex JC, Fairbank JT. Surgical resection and radiolocalization of the sentinel lymph node in breast cancer using a gamma probe. Surg Oncol 1993; 2: 335–339. 3. Giuliano AE, Kirgan DM, Guenther JM, Morton DL. Lymphatic mapping and sentinel lymphadenectomy for breast cancer. Ann Surg 1994; 220: 391–401.
Volume 20 | No. 6 | June 2009
original article 4. Krag D, Weaver D, Ashikaga T et al. The sentinel node in breast cancer—a multicenter validation study. N Engl J Med 1998; 339: 941–946. 5. Krag DN. The sentinel node for staging breast cancer: current review. Breast Cancer 1999; 6: 233–236. 6. Mansel RE, Fallowfield L, Kissin M et al. Randomized multicenter trial of sentinel node biopsy versus standard axillary treatment in operable breast cancer: the ALMANAC Trial. J Natl Cancer Inst 2006; 98: 599–609. 7. Veronesi U, Paganelli G, Viale G et al. Sentinel lymph node biopsy and axillary dissection in breast cancer: results in a large series. J Natl Cancer Inst 1999; 91: 368–373. 8. Albertini JJ, Lyman GH, Cox C et al. Lymphatic mapping and sentinel node biopsy in the patient with breast cancer. JAMA 1996; 276: 1818–1822. 9. Giuliano AE, Haigh PI, Brennan MB et al. Prospective observational study of sentinel lymphadenectomy without further axillary dissection in patients with sentinel node-negative breast cancer. J Clin Oncol 2000; 18: 2553–2559. 10. Peintinger F, Reitsamer R, Stranzl H, Ralph G. Comparison of quality of life and arm complaints after axillary lymph node dissection vs sentinel lymph node biopsy in breast cancer patients. Br J Cancer 2003; 89: 648–652. 11. Lyman GH, Giuliano AE, Somerfield MR et al. American Society of Clinical Oncology guideline recommendations for sentinel lymph node biopsy in earlystage breast cancer. J Clin Oncol 2005; 23: 7703–7720. 12. Veronesi U, Paganelli G, Viale G et al. A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer. N Engl J Med 2003; 349: 546–553. 13. Krag DN, Anderson SJ, Julian TB et al. Technical outcomes of sentinel-lymphnode resection and conventional axillary-lymph-node dissection in patients with clinically node-negative breast cancer: results from the NSABP B-32 randomised phase III trial. Lancet Oncol 2007; 8: 881–888. 14. Kim T, Giuliano AE, Lyman GH. Lymphatic mapping and sentinel lymph node biopsy in early-stage breast carcinoma: a metaanalysis [review]. Cancer 2006; 106: 4–16. 15. Zavagno G, De Salvo GL, Scalco G et al. A randomized clinical trial on sentinel lymph node biopsy versus axillary lymph node dissection in breast cancer: results of the Sentinella/GIVOM trial. Ann Surg 2008; 247: 207–213. 16. Wong SL, Edwards MJ, Chao C et al. Sentinel lymph node biopsy for breast cancer: impact of the number of sentinel nodes removed on the false-negative rate. J Am Coll Surg 2001; 192: 684–689; discussion 689–691. 17. Martin RC II, Chagpar A, Scoggins CR et al. Clinicopathologic factors associated with false-negative sentinel lymph-node biopsy in breast cancer. Ann Surg 2005; 241: 1005–1012; discussion 1012–1015. 18. Veronesi U, Paganelli G, Viale G et al. Sentinel-lymph-node biopsy as a staging procedure in breast cancer: update of a randomised controlled study. Lancet Oncol 2006; 7: 983–990. 19. Takei H, Suemasu K, Kurosumi M et al. Sentinel lymph node biopsy alone has no adverse impact on the survival of patients with breast cancer. Breast J 2006; 12 (5 Suppl 2): S157–S164. 20. Cox C, White L, Allred N et al. Survival outcomes in node-negative breast cancer patients evaluated with complete axillary node dissection versus sentinel lymph node biopsy. Ann Surg Oncol 2006; 13: 708–711.
doi:10.1093/annonc/mdn746 | 1007