Professional Documents
Culture Documents
7827
The latest version is at http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2014.55.7827
(N = 153)
The American Society of Clinical Oncology Clinical Practice Exclusion criteria included clinical T4 or N3 breast cancer, prior axillary
Guideline Update Committee considers the current reported FNR of surgery (including SNB before NAC), and neoadjuvant radiotherapy to the
SNB after NAC in patients who present with involved axillary nodes breast or axilla.
(from 10% to 30%) unacceptable and will take into consideration any
new information regarding this specific patient population, for which Clinical Examination and Axillary Ultrasound
Clinical staging and axillary ultrasound before and after NAC were
there is insufficient data.9 Currently, for most patients with biopsy- obtained. Physicians performing clinical examination and radiologists re-
proven node-positive breast cancer who receive NAC, completion sponsible for the axillary node assessment after NAC were asked to pro-
node dissection (CND) is recommended, regardless of their response spectively record the size and presence of residual nodal involvement.
to treatment; 22% to 35% of these patients will have no residual Timelines for the clinical examination, ultrasound evaluation, and surgery
disease in the axilla (axillary pathologic complete response [pCR]) are shown in Figure 1.
after NAC.5,10 With better selection of patients for NAC and improved
effectiveness of targeted therapy, these rates are likely to increase. CND Surgical Technique
could potentially be avoided in a large number of node-positive pa- All patients accrued to this trial were required to undergo SNB followed
by CND. The SNB technique was left to the discretion of the operating sur-
tients if SNB were shown to be accurate at identifying those who have
geon, but the use of radioisotope was mandatory. The use of blue dye (lymp-
achieved an axillary pCR after NAC. hazurin, patent blue, or methylene blue) and the sites of blue dye and
In the absence of NAC, the size of sentinel node (SN) metastases radioisotope injection were recorded prospectively. An SN was defined as the
is correlated with the probability of non-SN involvement, and low- hottest node (ie, node with highest count using gamma probe), any node with
volume disease in SNs does not always mandate CND when adjuvant counts of ⬎ 10% of those of the hottest node, any blue node, or any palpable
radiotherapy and systemic therapy are planned.11 Therefore, most suspicious node.
guidelines do not support, in the absence of NAC, the routine use of
immunohistochemistry (IHC) and recommend against CND when Pathologic Evaluation
only SN isolated tumor cells (pN0[i⫹]) or micrometastases (pN1mi) The number of SN specimens and the total number of SNs were
recorded. SNs were sliced at intervals of ⱕ 2 mm. Hematoxylin and eosin
are identified.1 After NAC, the relevance of SN ypN0(i⫹) or ypN1mi (HE) stain was used, and if negative, IHC was mandatory. The size of the
and the value of IHC are not well established. largest nodal metastasis was recorded. The nodes retrieved from CND were
The primary objective of the SN FNAC (Sentinel Node Biopsy analyzed by standard institutional guidelines, usually bivalved and stained
Following Neoadjuvant Chemotherapy) trial was to evaluate the ac- by HE. Pathology (SNB plus CND slides) were prepared, evaluated on site,
curacy of SNB after NAC in patients presenting with biopsy-proven and shipped for central review, which was performed by a single patholo-
node-positive breast cancer. It was hypothesized that the FNR would gist at the trial coordinating site. SNs with metastases of any size were
considered positive, including IHC-only– detected ypN1mi (⬎ 0.2 to 2
be ⱕ 10%, similar to the rate reported in the NSABP B-32 trial.12 The
mm) and ypN0(i⫹) [ⱕ 0.2 mm].
other main objective of this study was to evaluate the IR of SNB in this
setting (estimated at ⱖ 90%). The secondary objective was to evaluate
Sample-Size Calculation and Expected Accrual Period
and compare the accuracy of clinical examination, axillary ultrasound, With an expected FNR of 10%, 95% CI, and expected axillary pCR rate of
and SNB in identifying patients with axillary pCR after NAC. 40% to 50%, sample size was calculated to be 230 to 276 patients. The target
accrual was set at 300 patients for an expected accrual period of 3 years. A
higher rate of axillary pCR than previously published10 was used to take into
PATIENTS AND METHODS account the fact that patients with a higher probability of axillary pCR (oper-
able human epidermal growth factor receptor 2 (HER2) –positive and triple-
The SN FNAC study was a prospective multicentric phase II trial aimed at negative breast cancers) were more likely to participate in this trial.
evaluating the accuracy of SNB after NAC in patients with biopsy-proven
node-positive disease at presentation. Ten academic university-affiliated cen- Unplanned Interim Analysis and Early Closure
ters in Canada and the United States participated in the study. In December 2012, results from the ACOSOG (American College of
The protocol and consent forms were approved by the scientific com- Surgeons Oncology Group) Z1071 trial,13 which had a similar study design,
mittee or research ethics board of each participating center. Patients accrued to were reported at the San Antonio Breast Cancer Symposium. At that time, the
the trial were required to sign the study consent before their surgery date. SN FNAC study had reached 51% of its targeted accrual in a period of 3.75
years. It was estimated, using the most recent accrual rates, that a minimum of
Eligibility Criteria 2 years would be needed to reach our accrual target. An unplanned interim
Patients with stage II to IIIa (T0-T3, N1-2, M0-X) biopsy-proven node- analysis was performed, which yielded results similar to those presented by the
positive breast cancer selected to receive NAC were eligible for this study. ALLIANCE group. Because of the similarities between the two trials and the
of SNB in this setting was 87.6% (127 of 145; 95% CI, 82.2% to
Table 1. Patient Demographic, Clinicopathologic, and Treatment
Characteristics (N ⫽ 153) 93.0%). In the presence of a technical failure, the rate of positive nodal
involvement was 66.7% (12 of 18). The axillary pCR rate was 34.5%
Characteristic No. %
(50 of 145). A total of 30.3% (44 of 145) of patients eligible to partic-
Age, years
ipate in the SN FNAC study could have potentially avoided node
Median 50
Range 26-75
dissection if SNB only had been used after NAC.
Clinical T stage at presentation The average number of SNs removed was 2.7. The impact of the
T0 5 3 size of SN metastases on FNR, NPV, and accuracy of post-NAC SNB is
T1 8 5 summarized in Table 2. If ypN0(i⫹) SNs had been considered nega-
T2 76 50 tive, the FNR would have increased to 13.3% (11 of 83; 95% CI, 6.0%
T3 61 40
to 20.6%). There was no direct correlation between the size of SN
NA 3 2
Clinical N stage at presentation
metastases and the rate of positive non-SNs (Table 3). The average
N0 26 17 number of positive nodes in the CND after a positive ypN0(i⫹),
N1 113 74 ypN1mi, or ypN1 SN after neoadjuvant therapy was 0.7, 0.5, and 2.8,
N2 10 6 respectively (Appendix Table A1, online only). A total of 20.5% (17 of
NA 4 3 83) of patients had ⱖ four positive nodes in the CND, all in the
Clinical tumor subtype
presence of ypN1 SNs. The use of IHC increased the detection of
ER positive and/or PR positive, HER2 negative 87 57
Triple negative 23 15
small SN metastases in this setting; 63% of micrometastases and
HER2 positive 43 28 100% of isolated tumor cells were detected by IHC (Appendix
Neoadjuvant chemotherapy regimen Table A2, online only).
Anthracycline plus taxane 147 96 If only one SN was obtained, the FNR increased to 18.2%, and
Anthracycline based 1 1 accuracy decreased to 87.5%. In the presence of ⱖ two SNs, the FNR
Taxane based 3 2
of SNB decreased to 4.9%, and accuracy increased to 96.8% (FNR P ⫽
NA 2 1
.076; Table 2). SNB mapping with dual tracers (isotope and blue dye)
Abbreviations: ER, estrogen receptor; HER2, human epidermal growth factor
receptor 2; NA, not available; PR, progesterone receptor.
was associated with an FNR of 5.2% (three of 58), whereas single-
agent isotope use had an FNR of 16.0% (four of 25; P ⫽ .190; Table 2).
In the presence of a false-negative SNB, the average number of positive
non-SNs was 1.3 (Appendix Table A1, online only). The impact of
comparable methodology and results, it was decided to close the trial to accrual clinical stage at presentation on the FNR and NPV is summarized in
on December 5, 2012.
Table 2. Accuracy of clinical examination, axillary ultrasound, and
SNB at identifying residual axillary disease after NAC was 45%, 62%,
Statistical Analysis
The two-tailed Fisher’s exact test was used to compare frequencies be- and 95% respectively (Table 4). Positive predictive value of clinical
tween two groups. The Freeman-Halton extension of Fisher’s exact test was examination and axillary ultrasound was 89% (17 of 19) and 81% (44
used to compare three groups. Two-sided 95% CIs were calculated for the FNR of 54), respectively.
and IR.
RESULTS DISCUSSION
From March 2009 to December 2012, 153 patients were accrued to the In the SN FNAC study, patients who presented with biopsy-proven
SN FNAC study. Median age was 50 years (range, 26 to 75 years). A node-positive breast cancer had a post-NAC SNB FNR of 8.4%, meet-
total of 40% of patients had tumors ⬎ 5 cm in size; 74% had clinical ing the study prespecified end point of ⱕ 10%. The ACOSOG Z1071
N1 disease at presentation; 15% had triple-negative breast cancer, and trial, results of which were recently published,13 followed a similar
28% had HER2-positive breast cancer; 96% of patients in the study methodology, except that the use of IHC was not mandatory, and SNs
had received anthracycline and taxane– based chemotherapy regi- with metastases ⱕ 0.2 mm (ypN0[i⫹]) were considered node nega-
mens, and all patients with HER2-positive breast cancer had received tive. The reported FNR of SNB when ⱖ two SNs were examined was
neoadjuvant trastuzumab. Demographic, clinicopathologic, and 12.6%, which exceeded their prespecified threshold of 10%. The au-
treatment characteristics are listed in Table 1. thors concluded that changes in approach or patient selection would
The results for the primary objectives of the study are summa- be necessary to support the use of SNB in this setting.
rized in Figure 2. Using the information obtained from the central The SENTINA (Sentinel Neoadjuvant) study evaluated SNB in
review of pathology, the FNR of SNB after NAC for patients with 592 patients presenting with clinically node-positive breast cancer (by
biopsy-proven node-positive breast cancer was 8.4% (seven of 83; physical and ultrasound evaluations) that converted to clinically
95% CI, 2.4% to 14.4%). Using the data provided from the participat- node-negative disease after NAC.14 IHC was not mandatory, and the
ing institutions, the FNR was 9.6% (eight of 83; 95% CI, 3.3% to FNR was 14.2%. The design of the trial required patients who re-
15.9%). A patient with ypN0(i⫹) found on HE staining in a single mained clinically node positive after NAC to proceed directly to axil-
non-SN node, in the presence of a negative SN, was reclassified from lary node dissection. This is supported by the high positive predictive
false negative to true negative (ypT0N0[i⫹]) after central review. The values of physical examination (89%) and axillary ultrasound (81%)
negative predictive value (NPV) of SNB was 86.3% (44 of 51). The IR obtained in the SN FNAC study.
Study eligible
(n = 145)
Identification rate
(n = 127 of 145) Axillary pCR (n = 44) CND potentially avoided
(87.6%; 95% CI, 82.2% to 93.0%) (n = 44 of 145; 30.3%)
Fig 2. SN FNAC (Sentinel Node Biopsy Following Neoadjuvant Chemotherapy) trial: Results of central pathology review. CND, completion node dissection; NAC,
neoadjuvant chemotherapy; pCR, pathologic compete response; SNB, sentinel node biopsy.
On the basis of the three trials, it is clear that harvesting only one with an increased FNR in the absence of NAC.15 Patients with only one
SN is associated with a high FNR (SN FNAC, 18.2%; ACOSOG Z1071, SN represented 26.5% (22 of 83) of patients in the SN FNAC study,
31.5%; SENTINA, 24.3%). This finding is not specific to the neoadju- 20.4% (78 of 382) in the ACOSOG Z1071 trial, and 31.0% (70 of 226)
vant setting; obtaining a single SN is a well known factor associated in the SENTINA study. The limitations of this finding is that if we
Table 2. Impact of No. of SNs Removed, Method of SN Identification, Definition of Positive SN, and Clinical Stage at Presentation on FNR, NPV, and Accuracy
of SNB After Neoadjuvant Chemotherapy
FNR NPV Accuracy
Abbreviations: FNR, false-negative rate; NPV, negative predictive value; SN, sentinel node; SNB, sentinel node biopsy.
tions of this trial is the relatively small sample size, which explains the
Table 3. Size of Postneoadjuvant SN Metastases and Rate of
Positive Non-SNs large CI around the FNR that overlaps the 10% value. To confirm the
importance of IHC in the evaluation of SNs in this setting, we recom-
ypN1mi
ypN0(i⫹) (⬎ 0.2 to ypN1 mend reanalyzing the results of the ACOSOG Z1071 trial using IHC
[ⱕ 0.2 mm] 2 mm) (⬎ 2 mm) and considering SN metastases of any size as node positive.
Positive
Non-SNs No. % No. % No. % In the SN FNAC trial, the use of dual tracers (isotope and blue
Rate 4 of 7 57 3 of 8 38 34 of 61 56
dye) was associated with a three-fold decrease in the FNR, when
compared with isotope alone (5.2% v 16.0%; P ⫽ .190). This was not
NOTE. P ⫽ .637.
Abbreviation: SN, sentinel node. statistically significant, possibly because of the small sample size. Two-
fold decreases in the FNR have been previously reported with the use
of dual tracers in similar settings.13,14 The use of a single tracer using
blue dye alone was not permitted in the this study.
follow the standard definition of what constitutes an SN,2 it is not The IR (87.6%) was lower than the prespecified rate (ⱖ 90%). In
possible to create more SNs if only one is present. Harvesting more the presence of a technical failure, CND is warranted, because 66.7%
nodes (ie, blind node sampling) to increase the number of SNs to of these patients will have positive nodal metastases. A lower IR is a
ⱖ two was not evaluated in any of the trials and should not common finding when SNB is performed after NAC,3,4,7,8,13,14 but
be recommended. this should not preclude its use in patients with biopsy-proven node-
An unplanned retrospective analysis from the ACOSOG Z1071 positive disease, because the default treatment would be an axillary
trial reported that placing clips in suspicious nodes at the time of node dissection that could potentially be spared. SNB is more accurate
biopsy could possibly decrease the FNR of SNB.16 At the time of node than clinical examination and axillary ultrasound at identifying pa-
biopsy, 32.8% (172 of 525) of patients had clip placement. The clips tients who have achieved an axillary pCR.
were found in the SNs at the time of surgery in 55.8% (96 of 172) of In the presence of a false-negative SNB, the disease burden in the
cases. Patients in whom the clips were found had an FNR of 7.4% (v non-SNs was minimal (average, 1.3 nodes positive). Although this was
14.0% if clips were not found and 13.6% if clips were not placed). It is not recorded prospectively, in most of the participating centers, pa-
difficult to recommend this technique as a solution to increase the tients with biopsy-proven node-positive disease undergoing neoadju-
accuracy of SNB after neoadjuvant therapy, because it would only vant therapy received regional node irradiation (RNI) regardless of
benefit approximately half of patients. This should be considered as an SNB result. The AMAROS (After Mapping of the Axilla: Radiotherapy
exploratory analysis. Nodal clip placement was not evaluated in the SN or Surgery?) trial compared CND versus RNI in patients with a posi-
FNAC trial. tive SNB and found comparable adequate regional control rates be-
One of the most important findings of the SN FNAC trial is that tween the two modalities, but the trial excluded patients with tumors
after NAC, SN metastases of any size are significant and should be ⬎ 5 cm and patients who had received neoadjuvant therapy.18 The
considered positive. At the time of pathologic evaluation, SNs that significance of leaving even a minimal amount of residual disease in
were negative on HE staining were submitted to mandatory IHC the axilla after neoadjuvant therapy is not known, especially if RNI is
staining. ypN1mi and ypN0(i⫹) SNs were more likely to be identified administered. Patients in the SN FNAC study are not being observed
primarily by IHC staining. In the absence of NAC, the rate of positive beyond their surgery and are not an appropriate cohort for answering
non-SNs increased with the size of SN metastases.17 After NAC, at least this question, because all patients in the trial underwent CND. The
for patients with biopsy-proven node-positive disease at presentation, relevance of residual axillary disease after neoadjuvant therapy, the
the rate of positive non-SNs was high and independent of the size of role of RNI in the presence of axillary pCR, and the need for CND and
SN metastases. RNI versus RNI alone in the presence of a positive SNB after NAC are
The FNR for patients with ⱖ two SNs and SN metastases ⬎ 0.2 questions being addressed in ongoing clinical trials.19
mm was 11.5% (seven of 61), comparable to the FNR of 12.6% (39 of In conclusion, a low FNR (8.4%) of SNB after NAC in biopsy-
310) reported in the ACOSOG Z1071 trial in similar patients. By proven node-positive breast cancer can be achieved with mandatory
including IHC in the pathologic evaluation of the SNs and considering use of IHC for SN evaluation. SN metastases of any size, including
SN metastases of any size as positive, an FNR of ⬍ 10% can be micrometastases and isolated tumor cells, should be considered posi-
achieved (8.4% in this trial) without excluding patients. Using this tive and are associated with a significant rate of involved non-SNs.
method, 30.3% of patients who present with biopsy-proven node- This is the most convenient and inclusive method to decrease the FNR
positive disease could potentially be spared CND. One of the limita- of SNB below the threshold of 10% and should be further evaluated
Table 4. Evaluation of Residual Nodal Disease After Neoadjuvant Therapy: FNR, NPV, PPV, and Accuracy of Clinical Examination, Axillary Ultrasound, and SNB
FNR NPV PPV
Abbreviations: FNR, false-negative rate; NPV, negative predictive value; PPV, positive predictive value; SNB, sentinel node biopsy.
Support
Supported by the Quebec Breast Cancer Foundation, Cancer Research Society, Week-end to End Women’s Cancers, and Montreal Jewish
General Segal Cancer Centre.
■ ■ ■
GLOSSARY TERMS
immunohistochemistry: the application of antigen- neoadjuvant therapy: the administration of chemotherapy prior
antibody interactions to histochemical techniques. Typically, a to surgery. Induction chemotherapy is generally designed to decrease
tissue section is mounted on a slide and incubated with antibod- the size of the tumor prior to resection and to increase the rate of com-
ies (polyclonal or monoclonal) specific to the antigen (primary plete (R0) resections.
reaction). The antigen-antibody signal is then amplified using a
second antibody conjugated to a complex of peroxidase-
pathologic complete response: the absence of any residual tu-
mor cells in a histologic evaluation of a tumor specimen.
antiperoxidase, avidin-biotin-peroxidase, or avidin-biotin alka-
line phosphatase. In the presence of substrate and chromogen, sentinel lymph node: the lymph node that is anatomically located such
the enzyme forms a colored deposit at the sites of antibody- that it is the first site of lymph drainage from the location of the primary tumor.
antigen binding. Immunofluorescence is an alternate approach to It is suspected and assumed that if a malignancy is going to disseminate via the
visualize antigens. In this technique, the primary antigen- lymphatic system, metastases will first be evident in the sentinel lymph node. In
antibody signal is amplified using a second antibody conjugated this manner, this lymph node is said to stand guard or sentinel over the meta-
to a fluorochrome. On ultraviolet light absorption, the fluoro- static state of the tumor. For many cancers, the sentinel lymph node is biopsied
chrome emits its own light at a longer wavelength (fluorescence), as part of the staging process and presence of macro- or micrometastases in the
thus allowing localization of antibody-antigen complexes. sentinel lymph node is a negative prognostic factor.
Sentinel Node Biopsy After Neoadjuvant Chemotherapy in Biopsy-Proven Node-Positive Breast Cancer: The SN FNAC Study
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are
self-held unless noted. I ⫽ Immediate Family Member, Inst ⫽ My Institution. Relationships may not relate to the subject matter of this manuscript. For more
information about ASCO’s conflict of interest policy, please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.
Jean-Francois Boileau Stephen E. Karp
No relationship to disclose No relationship to disclose
Brigitte Poirier Isabelle Trop
No relationship to disclose No relationship to disclose
Mark Basik Andre Lisbona
No relationship to disclose No relationship to disclose
Claire M.B. Holloway Frances C. Wright
No relationship to disclose No relationship to disclose
Louis Gaboury Rami J. Younan
No relationship to disclose
No relationship to disclose
Lucas Sideris
Louise Provencher
No relationship to disclose
No relationship to disclose
Sarkis Meterissian
No relationship to disclose Erica Patocskai
No relationship to disclose
Angel Arnaout
No relationship to disclose Atilla Omeroglu
No relationship to disclose
Muriel Brackstone
No relationship to disclose Andre Robidoux
No relationship to disclose
David R. McCready
No relationship to disclose
Acknowledgment
Presented at the 49th Annual Meeting of the American Society of Clinical Oncology (ASCO), Chicago, IL, May 31-June 4, 2013, and ASCO
Breast Cancer Symposium, San Francisco, CA, September 7-9, 2013.
Appendix
Table A1. Impact of Post-NAC Size of SN Metastases and False-Negative SNs on Average No. of Positive SNs and Non-SNs in CND
Status of SN After NAC Mean No. of Positive SNs Mean No. of Positive Nodes in CND
ypN0(i⫹) [n ⫽ 7] 1.3 0.7
ypN1mi (n ⫽ 8) 1.4 0.5
ypN1 (n ⫽ 61) 2.3 2.8
False negative (n ⫽ 7) 0 1.3
Abbreviations: CND, completion node dissection; NAC, neoadjuvant chemotherapy; SN, sentinel node.
Table A2. Impact of Post-NAC Size of Largest SN Metastasis on Rate of Positive SNBs Detected by IHC
NOTE. P ⫽ .037.
Abbreviations: IHC, immunohistochemistry; NAC, neoadjuvant chemotherapy; SN, sentinel node; SNB, sentinel node biopsy.