Sentinel Node Biopsy After Neoadjuvant Chemotherapy: The Pros

Sentinel Node Biopsy Sentinel Node Biopsy AfterAfterNeoadjuvant Chemotherapy:Neoadjuvant Chemotherapy:

The ProsThe Pros

Terry Mamounas, M.D., M.P.HTerry Mamounas, M.D., M.P.HAssociate Professor of SurgeryAssociate Professor of Surgery

Northeastern Ohio Universities College of MedicineNortheastern Ohio Universities College of MedicineMedical DirectorMedical Director

Aultman Cancer CenterAultman Cancer Center

Original Clinical RationaleOriginal Clinical Rationalefor Neoadjuvant Chemotherapyfor Neoadjuvant Chemotherapy

•• Convert inoperable BC to operable BCConvert inoperable BC to operable BC

•• Convert operable BC patients requiring Convert operable BC patients requiring mastectomy to candidates for BCS mastectomy to candidates for BCS

Axillary Node DownAxillary Node Down--Staging with NCStaging with NC

ACACNSABP BNSABP B--1818

5050

4040

3030

2020

1010

00

% % ConversionConversion

FromFromNode (+)Node (+)

ToToNode (Node (--))

ATAT CMFCMFECTOECTO

30 37

FECFECEORTCEORTC

19

ACAC TXTTXTNSABP BNSABP B--2727**

43

*Assuming 30% nodal down*Assuming 30% nodal down--staging with neoadjuvant ACstaging with neoadjuvant AC

Effect of NC on AxillaryEffect of NC on AxillaryNodal MetastasesNodal Metastases

•• NC downstages axillary nodes in NC downstages axillary nodes in about 20about 20--40% of the patients40% of the patients

•• This was of no particular clinical This was of no particular clinical significance when axillary dissection significance when axillary dissection was the sole method for staging the was the sole method for staging the axillaaxilla

Effect of NC on AxillaryEffect of NC on AxillaryNodal MetastasesNodal Metastases

•• The advent of sentinel node biopsy The advent of sentinel node biopsy introduced another potential benefit introduced another potential benefit from neoadjuvant chemotherapyfrom neoadjuvant chemotherapy

•• Potential for decreasing the extent of Potential for decreasing the extent of axillary surgery with SNB vs. AND if axillary surgery with SNB vs. AND if the axillary nodes are downthe axillary nodes are down--staged staged with NCwith NC

SNB After NCSNB After NCTwo Main Reasons GivenTwo Main Reasons Givenby Those Who Oppose Itby Those Who Oppose It

1.1. It does not work as well as it does It does not work as well as it does before systemic therapybefore systemic therapy

2.2. By doing SNB after NC, we lose By doing SNB after NC, we lose information that is important for information that is important for further patient managementfurther patient management

SNB After NCSNB After NCTwo Main Reasons for Opposing ItTwo Main Reasons for Opposing It



SNB After NCSNB After NC

•• Is SNB after NC as feasible and Is SNB after NC as feasible and accurate as before systemic therapy?accurate as before systemic therapy?•• Does response to NC cause scarring that Does response to NC cause scarring that

could affect the lymphatic drainage making could affect the lymphatic drainage making SN identification more difficult and/or less SN identification more difficult and/or less accurate?accurate?

•• Is NC equally effective in downIs NC equally effective in down--staging staging SNsSNsand nonand non--SNsSNs

SNB After NCSNB After NCFeasibility and AccuracyFeasibility and Accuracy

•• Information from:Information from:

•• Single institution trialsSingle institution trials

•• Multicenter TrialsMulticenter Trials

•• MetaMeta--AnalysesAnalyses

•• Limited early experience with SNB Limited early experience with SNB after NCafter NC

•• Initial small studies have shown Initial small studies have shown variability in:variability in:•• Rates of SN identification (72Rates of SN identification (72--100%) 100%) •• Rates of false negative SN (0%Rates of false negative SN (0%--33%)33%)

SNB After NCSNB After NCSingle Institution ExperienceSingle Institution Experience

YesYes55979732 (19)32 (19)Balch, 2003Balch, 2003

YesYes88858520 (12)20 (12)AiharaAihara, 2004, 2004

YesYes1717989842 (18)42 (18)PiatoPiato, 2003, 2003

YesYes0088883333(18)(18)HaidHaid, 2001, 2001

YesYes77878730 (15)30 (15)ReitsamerReitsamer, 2003, 2003

YesYes00939314 (11)14 (11)Brady, 2002Brady, 2002

YesYes9910010021 (11)21 (11)Schwartz, 2003Schwartz, 2003

10.810.889.189.1398 (182)398 (182)AllAll

YesYes00868635 (9)35 (9)Miller, 2002Miller, 2002

NoNo2525858540 (16)40 (16)Fernandez, 2001Fernandez, 2001

Yes*Yes*1414858534 (13)34 (13)Stearns, 2002Stearns, 2002

NoNo3333878715 (9)15 (9)NasonNason, 2000, 2000

YesYes1212848451 (25)51 (25)BreslinBreslin, 2000, 2000

AccurateAccurateFN RateFN Rate(%)(%)

SuccessSuccessRate ( %)Rate ( %)

# Pts # Pts (Node +)(Node +)

AuthorAuthorSNB After NC: Single Institution SeriesSNB After NC: Single Institution Series

*Not in IBC*Not in IBC

YesYes44949453 (24)53 (24)Lang, 2004Lang, 2004

8.18.188.788.7309 (160)309 (160)AllAll

YesYes00959542 (19)42 (19)Julian, 2004Julian, 2004

YesYes1212949447 (33)47 (33)ShimazuShimazu, 2004, 2004

YesYes1111949477 (27)77 (27)Kinoshita, 2006Kinoshita, 2006

NoNo1111818136 (18)36 (18)Jones, 2005Jones, 2005

YesYes1111727254 (27)54 (27)Kang, 2004Kang, 2004

AccurateAccurateFN RateFN Rate(%)(%)

SuccessSuccessRate ( %)Rate ( %)

# Pts # Pts (Node +)(Node +)

AuthorAuthor

SNB After NC: Single Institution SeriesSNB After NC: Single Institution Series

•• Identification Rate: Identification Rate: 85%85%•• With blue dye: 78%With blue dye: 78%•• With isotope With isotope ++ blue dye: blue dye: 8888--89%89%

•• False Negative Rate: False Negative Rate: 11%11%•• With blue dye: 14%With blue dye: 14%•• With isotope With isotope ++ blue dye: blue dye: 8.4%8.4%

SNB After NCSNB After NCMultiMulti--Center Studies: NSABP BCenter Studies: NSABP B--2727

(n=428)(n=428)

Mamounas EP: J Clin Oncol, 2005

SNB After NCSNB After NCMetaMeta--Analysis of SingleAnalysis of Single--Institution Institution

and Multiand Multi--Center StudiesCenter Studies

•• 21 studies21 studies•• 1273 patients1273 patients•• Identification rates: Identification rates: 7272--100%100%

––Pooled estimate: Pooled estimate: 90%90%•• False Negative Rates: False Negative Rates: 00--33%33%

––Pooled estimate:Pooled estimate: 12%12%

Conclusion:Conclusion:SNB is a reliable tool forSNB is a reliable tool for

planning treatment after NCplanning treatment after NC

Comparison of False Negative Rates Comparison of False Negative Rates Between SN Multicenter StudiesBetween SN Multicenter Studies

Multicenter SBMulticenter SB--2 Trial2 Trial 11%11% (13/114)(13/114)Italian Randomized Trial 9%Italian Randomized Trial 9% (8/91)(8/91)Ann Arundel 13% (25/19Ann Arundel 13% (25/193)3)University of LouisvilleUniversity of Louisville 7% (24/333)7% (24/333)NSABP BNSABP B--32 Randomized Trial 10%32 Randomized Trial 10% (75/766)(75/766)NSABP BNSABP B--27 (After NC) 11%27 (After NC) 11% (15/140)(15/140)MetaMeta--Analysis (After NC)Analysis (After NC) 12% 12% (65/540)(65/540)

StudyStudy FNRFNR (SN(SN--/N+)/N+)

Krag DN: Surg Oncol 1993 Veronesi U: N Engl J Med 2003 McMasters KM: J Clin Oncol 2000Mamounas EP: J Clin Oncol 2005 Tafra L: Am J Surg 2001 Xing Y:Br J Surg 2005 Julian JB: SABCS 2004

•• Optimal candidates should have low Optimal candidates should have low risk for a positive nonrisk for a positive non--SNSN

•• SNB inaccuracy rate is a function of:SNB inaccuracy rate is a function of:–– False Negative RateFalse Negative Rate

»»Anatomic variabilityAnatomic variability»»Surgeon’s performanceSurgeon’s performance

–– Rate of axillary node positivityRate of axillary node positivity

SNB After NC: Optimal CandidatesSNB After NC: Optimal Candidates

NSABP BNSABP B--27:Rate of Positive Nodes 27:Rate of Positive Nodes According to Tumor ResponseAccording to Tumor Response

%%

0

20

40

60

80

pCR cCR cPR cSD cPD

P < 0.0001P < 0.0001

16

35

5954 58

Bear HD: J Clin Oncol 2003

%

cCR/pCR(1/56)

NSABP BNSABP B--27: SN Inaccuracy Rate27: SN Inaccuracy RateAccording to Tumor ResponseAccording to Tumor Response

All PatientsAll Patients

0

2

4

6

8

10

cCR/pINV(4/99)

1.8

4.05.1

c(PR+SD+PD)(8/158)

P=0.36P=0.36

Mamounas EP: J Clin Oncol 2005

SNB After NCSNB After NCTwo Main Reasons for Opposing ItTwo Main Reasons for Opposing It



•• No imaging modality is of significant value in No imaging modality is of significant value in predicting subclinical involvement of axillary predicting subclinical involvement of axillary nodesnodes

•• Ultrasound of the axilla with FNA of Ultrasound of the axilla with FNA of indeterminate/suspicious nodes:indeterminate/suspicious nodes:–– Simple, minimally invasiveSimple, minimally invasive–– Decreased sensitivityDecreased sensitivity–– Can provide useful clinical information (avoid Can provide useful clinical information (avoid

SNB, demonstrate direct chemosensitivity)SNB, demonstrate direct chemosensitivity)

Clinical Assessment of Axillary Nodal Clinical Assessment of Axillary Nodal Status Before NCStatus Before NC

All this is fineAll this is fineBUTBUT

SNB SNB BeforeBefore NC is not!NC is not!

•• Information on the status of SN can be Information on the status of SN can be obtained without the confounding effects obtained without the confounding effects of NCof NC

•• This may provide an advantage regarding:This may provide an advantage regarding:

–– Further surgical management of the axillaFurther surgical management of the axilla

–– Selection of optimal NC or adjuvant chemo Selection of optimal NC or adjuvant chemo after NCafter NC

–– Selection of optimal locoSelection of optimal loco--regional XRTregional XRT

SNB SNB BeforeBefore NCNC: : Arguments in FavorArguments in Favor

SNB SNB BeforeBefore NCNC: : Two Surgical ProceduresTwo Surgical Procedures

(+) SN(+) SNANDAND NCNC

AND + BCT/MASTAND + BCT/MASTNCNC

BCT/MASTBCT/MAST

((--) SN) SN BCT/MASTBCT/MASTNCNC

•• Patients with large operable breast Patients with large operable breast cancer have cancer have high likelihood of positive high likelihood of positive nodes (50nodes (50--70%)70%)

•• This approach does not take advantage This approach does not take advantage of the downstaging effects of NC on of the downstaging effects of NC on nodes: nodes: 3030--40% conversion from (+) to (40% conversion from (+) to (--))

SNB SNB BeforeBefore NCNC: : Potential DisadvantagesPotential Disadvantages

•• This approach This approach assumesassumes surgeons are surgeons are comfortable performing SNB alone before comfortable performing SNB alone before NC but not after NCNC but not after NC

•• Outcome results from Outcome results from large randomized large randomized trialstrials comparing SNB alone with axillary comparing SNB alone with axillary dissection are dissection are pendingpending

SNB SNB BeforeBefore NC Rather than NC Rather than AfterAfter NC?NC?

SNB at Diagnosis vs. After NCSNB at Diagnosis vs. After NCConfidence Intervals Around FNRConfidence Intervals Around FNR

0 5 10 15 200 5 10 15 20I I I I II I I I I

Multicenter SBMulticenter SB--2 Trial2 TrialItalian Randomized TrialItalian Randomized TrialAnn Arundel Ann Arundel University of LouisvilleUniversity of LouisvilleNSABP BNSABP B--32 Randomized Trial32 Randomized TrialNSABP BNSABP B--27 (After NC)27 (After NC)MetaMeta--Analysis (After NC)Analysis (After NC)

Krag DN: Surg Oncol 1993 Veronesi U: N Engl J Med 2003 McMasters KM: J Clin Oncol 2000Mamounas EP: J Clin Oncol 2005 Tafra L: Am J Surg 2001 Xing Y:Br J Surg 2005 Julian JB: SABCS 2004

•• May be useful in patients who will not May be useful in patients who will not need chemotherapy if the SN is negative need chemotherapy if the SN is negative (uncommon situation among typical (uncommon situation among typical candidates for NC)candidates for NC)

•• Usually Usually original tumor size, age and original tumor size, age and primary tumor markersprimary tumor markers are good guides are good guides for appropriate NC for appropriate NC

SNB SNB BeforeBefore NCNC: : Selection of Optimal NC?Selection of Optimal NC?

•• Consideration for adjuvant chemo after Consideration for adjuvant chemo after NC depends on:NC depends on:–– What NC was usedWhat NC was used (anthracyclines only (anthracyclines only

or anthracyclines and taxanes)or anthracyclines and taxanes)–– Clinical and path breast tumor responseClinical and path breast tumor response–– Status of axillary nodes after NCStatus of axillary nodes after NC

•• Uncertain significance of negative nodes Uncertain significance of negative nodes after NC and prior SNB after NC and prior SNB (downstaging vs. (downstaging vs. prior removal of all (+) nodes)prior removal of all (+) nodes)

SNB SNB BeforeBefore NC: NC: Selection of Adjuvant Chemo?Selection of Adjuvant Chemo?

•• Breast XRT:Breast XRT: Should be always given after Should be always given after lumpectomylumpectomy

•• Chest Wall and Regional XRT:Chest Wall and Regional XRT: Consider Consider factors predicting localfactors predicting local--regional failure after regional failure after NCNC

•• These factors may predict LR failure more These factors may predict LR failure more accurately than the original pathologic nodal accurately than the original pathologic nodal status before NCstatus before NC

SNB SNB BeforeBefore NC:NC:Selection of LocoSelection of Loco--Regional XRT?Regional XRT?

Problem:Problem:Not much informationNot much informationexists on the subject!exists on the subject!

NSABP BNSABP B--18: Predictors of LRF after NC18: Predictors of LRF after NCMultivariate AnalysisMultivariate Analysis

Cox ModelCox ModelNeoadjuvant ChemotherapyNeoadjuvant Chemotherapy

•• Number of pathNumber of path--positive positive nodes (nodes (pp<0.0001) <0.0001)

•• Age (Age (pp=0.005) =0.005) •• Breast tumor response Breast tumor response

((pp=0.054)=0.054)6.6 8.3

11.4

21

0

5

10

15

20

25

Node (-)pCR

Node (+)pCR

Node (-)No pCR

Node (+)No pCR

n=76n=12

n=329

n=262

1010--year Cum. Incidence of LRF (%)year Cum. Incidence of LRF (%)

Mamounas EP: Br Ca Res Treat 2003

•• Univariate and multivariate analysis of predictors of Univariate and multivariate analysis of predictors of LR failureLR failure

•• Includes the Includes the preoppreop ACAC arms from arms from BB--18 and B18 and B--2727 and and the the preoppreop ACAC---->T>T arm from arm from BB--2727

•• Similar results were obtained by using only the two Similar results were obtained by using only the two preoppreop AC arms or by adding the third BAC arms or by adding the third B--27 arm (AC27 arm (AC---->S>S---->T) >T)

•• Analysis is based on Analysis is based on 2192 pts2192 pts and and 229 events (LRF)229 events (LRF)

•• Pathologic complete response Pathologic complete response (pCR)(pCR) was defined as was defined as no invasive disease in the breast no invasive disease in the breast andand negative negative axillary nodesaxillary nodes

Updated LRF Analysis: NSABP BUpdated LRF Analysis: NSABP B--18/B18/B--2727

LRF Update: NSABP BLRF Update: NSABP B--18/B18/B--2727MVA: Predictors of LRFMVA: Predictors of LRF

2.582.58Node(+) vs. Node(Node(+) vs. Node(--)/pCR )/pCR <0.0001<0.00011.421.42Node(Node(--)/No pCR vs. Node()/No pCR vs. Node(--)/pCR )/pCR

0.00070.00071.601.60Clin. Node (+) vs. Clin. Node (Clin. Node (+) vs. Clin. Node (--) )

1.361.36Clin. Tumor Size > 5 vs. 0Clin. Tumor Size > 5 vs. 0--2 cm 2 cm 0.010.010.860.86Clin. Tumor Size 2.1Clin. Tumor Size 2.1--5 vs. 05 vs. 0--2 cm 2 cm

PP--ValueValue

Hazard Hazard RatioRatio

VariableVariable

10.89.6

13.3

0

5

10

15

20

<= 2 cm 2.1-5 cm > 5 cm

n=433n=433n=1333n=1333

n=543n=543

LRF Update: NSABP BLRF Update: NSABP B--18/B18/B--272788--Year Cum. Incidence of LRF byYear Cum. Incidence of LRF by

Clinical Tumor SizeClinical Tumor Size

9.3

14.2

0

5

10

15

20

Clin. Node (-) Clin Node (+)

n=1642n=1642

n=667n=667


Clinical Nodal StatusClinical Nodal Status

6.68.4

14.6

0

5

10

15

20

Node (-) pCR Node (-) NopCR

Node (+)


Path Nodal Status and pCRPath Nodal Status and pCR

n=313n=313

n=914n=914

n=965n=965

5.6

1.2

5.9

1.1

6.4

3.4

0

5

10

15


Node (+)

IBTR Regional

88--Year Cum. Incidence of LRF byYear Cum. Incidence of LRF byPath Nodal Status and pCRPath Nodal Status and pCR

(Lumpectomy Pts)(Lumpectomy Pts)

n=245n=245 n=644n=644

n=518n=518

1.5

4.45

3 11.2

3.7

0

5

10

15

20


Node (+)

Chest Wall Regional

88--Year Cum. Incidence of LRF byYear Cum. Incidence of LRF byPath Nodal Status and pCRPath Nodal Status and pCR

(Mastectomy Pts)(Mastectomy Pts)

n=68n=68n=270n=270

n=447n=447

6.1 7.7 6.910.1

12

17.3

0

5

10

15

20

Node (-)/pCR Node(-)/No pCR Node (+)

Clin. Node (-) Clin Node (+)

n=196n=196 n=519n=519

n=395n=395n=117n=117

n=510n=510

n=455n=455

88--Year Cum. Incidence of LRF byYear Cum. Incidence of LRF byAccording to According to Path Nodal Status/pCRPath Nodal Status/pCR

and and Clinical Nodal StatusClinical Nodal Status

•• SNB SNB afterafter NC is feasible and accurate with NC is feasible and accurate with performance characteristics similar to those for performance characteristics similar to those for SNB before systemic therapySNB before systemic therapy

•• By performing SNB By performing SNB afterafter NC, up to 40 percent of NC, up to 40 percent of patients who present with involve axillary nodes patients who present with involve axillary nodes may be spared from axillary dissectionmay be spared from axillary dissection

•• SNB SNB beforebefore NC does not offer particular clinical NC does not offer particular clinical advantages and reduces the number of patients advantages and reduces the number of patients who could benefit from the downwho could benefit from the down--staging effect staging effect of NC in the axillary nodesof NC in the axillary nodes

ConclusionsConclusions

Documents

Sentinel Node Biopsy After Neoadjuvant Chemotherapy: The Pros