Real world HER2 testing - are they reliable? Result of the planned analysis of the initial 104 cases enrolled by IHC in NSABP B-31 Soonmyung Paik, MD Division.

Real world HER2 testing - are they reliable?

Result of the planned analysis of the initial 104 cases enrolled by IHC in

NSABP B-31

Soonmyung Paik, MDDivision of Pathology

National Surgical Adjuvant Breast and Bowel Project

NSABP B-31

AC x 4

Taxol x 4TAM x 5 yrs.*

AC x 4

Taxol x 4 + Herceptin

weekly for 1 yearTAM x 5 yrs.*

Operable Breast CancerHER-2 Positive Tumor

Path Positive Axillary Nodes

Randomization

* ER+ or PgR+, optional for > 50 yrs.whose tumors are ER– and PgR–

Eligibility determination for B-31 trial for adjuvant Herceptin

• Herceptest 3+ by any laboratory

• other antibodies: more than 33% of cells show strong membrane staining

• gene amplification by FISH

Planned central review of the initial 100 cases in B-

31• Objective: To make sure that we are

capturing the real HER2 positive patients for B-31

• If more than 20% of the cases are negative by both Herceptest and FISH, consider changing the eligibility criteria

Method• Central Herceptest and PathVysion FISH

test performed by a designated third party central laboratory (Laboratory Corporation of America)

• Results are available for both assays from 104 cases

• Results were reconfirmed by NSABP Pathology lab using tissue array of 81 cases

Central Herceptest versus FISH (n=104)

FI SHpositive

FI SHnegative

Herceptest3+

79 3

Herceptest0 to 2+

3 (all 2+) 19 (18%)

* 94% agreement

B-31 Tissue Microarray

FISH versus FISH (N=81)

FI SHPositive

FI SHnegative

My FI SHPositive

62 1

My FI SHnegative

3 15

* 95% agreement

Test usedforeligibility

Type ofLab used

CentralHerceptest0- 2+

CentralPathVysionNotamplified

Negative byboth assays

Smallvolume lab

10/52 12/52 10/52 (19%)Hereceptest3+(n=80) Larger

volume lab1/28 1/28 1/28 (4%)

Smallvolume lab

11/23 9/23 8/23 (35%)Other IHCassays(n=24) Larger

volume lab0/1 0/1 0/1 (0%)

B-31 central review of the first 104 cases

Small volume lab: process no more than 99 cases per monthLarger volume lab: process at least 100 cases per month

Test usedforeligibility

Lab CentralHerceptest0- 2+

CentralPathVysionNot amplified

Negative byboth assays

Small volumelab

10/52 12/52 10/52 (18%)Hereceptest

Largervolume lab

1/28 1/28 1/28 (4%)

Small volumelab

3/10 2/10 2/10 (20%)CB- 11

Largervolume lab

0 0 0/0

Small volumelab

7/12 7/12 6/12 (50%)Other IHCassays

Largervolume lab

1/1 0/1 0/1

B-31 central review of the first 104 cases

One publication showed CISH is a reliable alternative to FISH

FISH versus CISH (B-31)

* 92% agreement

FI SHNegative

FI SHPositive

CI SHNegative

47 2

CI SHPositive

3 8

CISH vs. FISH (NSABP B15, n=123)

FI SHNegative

FI SHPositive

CI SHNegative

30 5

CI SHPositive

5 83

92% agreement with FISH

Central review of cases entered by FISH (N=27)

CentralPathVysionNot-amplifi ed

Number ofcontributinglabs

Small volumelab

0/ 17 15

Larger volumelab

0/ 10 6

* 26/27 cases by PathVysion, 1/27 case by Inform test

Conclusions from central review of B-31

• Herceptest provided by larger volume reference labs was reliable in determining 3+ cases (about 95% agreement with FISH)– According to survey, all larger volume labs

perform both IHC and FISH, this may contribute to better QC

• IHC performed by small volume labs tends to be less reliable

• PathVysion FISH test is reliable

• CISH in the current format is not a good alternative to FISH

Eligibility change for B-31

• If FISH positive (amplification - as determined by PathVysion or HER2 Inform test) - eligible regardless of IHC result

• If 3+ IHC by an approved lab - eligible

• If 3+ IHC by a non-approved lab - needs to be confirmed by an approved lab

Lab Approval for B-31

• Automatic approval if the lab process more than 100 cases per month with Herceptest

• If a lab process less than 100 cases with Herceptest or using different antibody method, we will determine the quality of the lab on individual evaluation (based on concordance rate with FISH, track record)

• Automatic approval if using FISH

Plan• Reliability of real world FISH remains an open

question (can pathology labs that cannot do IHC do FISH well??)

• We will conduct another 100 cases review after implementation of the protocol amendment

• All cases are being arrayed in triplicate before return of the blocks