1. San Antonio Breast Cancer Symposium 2006 Highlights – Breast Surgery Frederick M. Dirbas, M.D. Assistant Professor of Surgery Stanford Cancer Center

2. Breast Conservation +/- RT Abstract 29, Intergroup Study E5194: Local Excision Without Radiation for Selected Patients with DCIS Abstract 11, CALGB Study: Tamoxifen +/- RT in Women ≥70 with Breast Cancer

3. Abstract 29, Intergroup* Study E5194 711 women enrolled 1997 to 2002 Grade 1 or 2 DCIS, ≤ 2.5 cm:580 pts Grade 3 DCIS, ≤ 1 cm:102 pts Ineligible:29 pts Mean age of eligible patients is 60 years 31% declared intent to take Tam Median f/u 4.96 years Post excision mammogram for residual microcalcifications, central path review *ECOG, NCCTG

4. Abstract 29, Intergroup* Study E5194 Principle Outcome Measures Conclusions – Low to intermediate grade DCIS has low recurrence rate w/o RT – High grade DCIS has a high recurrence rate suggesting surgery alone is inadequate Low or intermediate grade DCISHigh grade DCIS N, mean age 60 (range 28 – 88)580102 Median size6 mm (only 18% > 1 cm)7 mm Median margin width5 to 10 mm IBTR at 5 years6.8%13.7% IBTR DCIS50% IBTR Invasive53%47% Absolute IBTR Invasive≈ 3.6%≈ 6.4% (B-17/RT 3.9% 8 years) Contralateral breast events at 5 years3.5%4.2%

5. Abstract 29, Intergroup* Study E5194 Is this real? – Dana Farber data demonstrates higher recurrence rates without RT, IBTR 12% at 5 years, with low/intermediate grade and margins > 1 cm (TAM not allowed) – Van Nuys data demonstrates low recurrence rates overall if margins > 1 cm, IBTR 13.9%, invasive IBTR 3.4% at 12 years higher recurrence rates with grade 3 DCIS J Clin Oncol (United States), Mar 1 2006, 24(7) p1031-6 Am J Surg (United States), Oct 2006, 192(4) p420-2

6. Abstract 29, Intergroup* Study E5194 Where does one go from here? – RTOG 98-04 randomized study closed to accrual, results pending Favorable DCIS +/- Tam +/- RT –Low to intermediate grade –3-9 mm margins vs 1 cm or greater –< 1 cm lesion vs 1 to 2.5 cm –Age 50 –TAM yes vs no

7. Sentinel Node Biopsy ITC (Nanomets) and Micromets Abstract 25, Axillary Lymph Node ITC (Nanometastases) are Prognostic Factors for Metastatic Relapse. “These results support the inclusion of procedures for nanometastasis detection in TNM pathologic staging.” – National Cancer Institute, Milan, Italy Plenary session, 3, Micrometastases in the Sentinel Node “One should not look too hard for micrometastases in the sentinel node.” – The Netherlands Cancer Institute, Amsterdam, Netherlands

8. Abstract 25, Sentinel Node Biopsy ITC (Nanomets) Compared with standard H&E staining – step sectioning increases sensitivity 10% – IHC increases sensitivity 10% further – RT-PCT increases sensitivity 10 further still What is the value of these observations? – Are these metastatic “cells” viable? – Are these simply displaced cells?

9. Abstract 25, Sentinel Node Biopsy ITC (Nanomets) Abstract 25, Axillary Lymph Node ITC (Nanometastases) are Prognostic Factors 702 consecutive patients at the National Cancer Institute, Milan – pN0 – Completion axillary dissections 8 years median f/u Outcomes – Risk of first adverse event Crude cumulative incidence curves generated to estimate cumulative probability of occurrence of adverse events – Distant relapse pN0(i+) is a strong risk factor for event free survival (p<.0005) and for metastatic relapse in both univariate and multivariate analysis accounting for grading, T stage, and age

10. Plenary Session 3, Sentinel Node Biopsy Micromets Meta-analysis of 25 studies, > 8,687 published patients who had neg SN bx – 3 years mean f/u 31 relapses, axillary recurrence only.36% (.8% to 2.3% reported for ALND) Compare this with false negative rate after upfront ALND, which ranges from 2 to 11% in the literature Conclusion: not all histologic findings of residual disease represent viable tumor –Iatrogenic tumor cell displacement recognized, papillary lesions, DCIS (Bleiweiss JCO, 2006) What does this mean for micrometastases? – Should they be ignored? Can micrometastases predict non-SN metastasis, whether additional micrometastases or macrometastases? Prognostically, may not add much information above and beyond tumor size and grade Conclusions – ITC can be ignored – Micrometastases should be treated with completion ALND or systemic therapy – If completion ALND, and other nodes are negative, treatment should be based on the characteristics of the primary tumor, not the presence of the micrometastasis

11. Sentinel Node Biopsy Micromets Are these findings real? – Are nanometastases prognostic? – Should micrometastases be ignored?

12. Sentinel Node Biopsy Significance of ITC and Micromets Where does one go from here? Data from randomized studies pending – NSABP B-32 – ACOSOG Z10 – IBCSG 230-1: focused specifically on micromets New randomized studies incorporate gene signature patterns as prognostic tools – Trial Assigning IndividuaLized Options for Treatment (Rx), or TAILORx – MINDACT (MIcroarray for Node negative Disease may Avoid ChemoTherapy). – Will micrometastases be as/more/less prognostic than a gene signature?

13. Intraoperative SN Evaluation Abstract 26, Sentinel Node Biopsy in Invasive Ductal Carcinoma, Invasive Lobular Carcinoma, Favorable Histologic Subtypes Abstract 28, Multiplex Molecular Assay Has Improved Sensitivity over Histologic Intraoperative Nodal Metastases Tests

14. Abstract 26 Sentinel Node Biopsy in IDC, ILC, and Favorable Subtypes Single institution, 5,298 consecutive patients with T1-3 invasive carcinoma, 1996 to 2004 SLN bx with frozen section (FS) For IDC and ILC, but not Fav, yield increased with Tumor Size Sensitivity and yield of FS were higher with patients < 50, increasing tumor size, and AL invasion. Yield of FS was < 10% for all patients with ID/IL tumors < 1 cm in size who were older than age 60. IDCILCFavorable Frozen Section (FS) Sensitivity62%52%p=.00646% Yield (% pos SN Bx)22%21%p=.4873%p=<0.001

15. Abstract 26 Sentinel Node Biopsy in IDC, ILC, and Favorable Subtypes Conclusion: for ID and IL, overall sensitivity is > 50% For any individual with age > 60, T1a or b tumor of any histology, yield < 10% – Intraoperative FS is not worthwhile for this low yield subset.

16. Abstract 28, Multiplex Molecular Assay Standard H&E processing of axillary node samples 2 to 5%, and will miss 10 to 15% of nodal metastases Intraoperative evaluation even less sensitive

17. Abstract 28, Multiplex Molecular Assay RT-PCR kit for intraoperative SN evaluation – GeneSearch™ BLN Assay Markers –Mammoglobin (breast) –Cytokeratin 19 (epithelial) Closed tube system If either or both are “positive”, node is “positive” – Approximately ½ hour to perform test – Does not require a pathologist – Price not set

18. Abstract 28, Multiplex Molecular Assay RT-PCR kit for intraoperative SN evaluation – Approximately 1,000 cells in a.2 mm micromet – Assay designed to report < 1,000 cells as “negative” – Therefore, will detect metastases >.2 mm and give a “positive” result Clinical trial design – Half of lymph node sent for standard SN processing – Half of lymph node sent for RT-PCR

19. Abstract 28, Multiplex Molecular Assay FS used at 11 sites testing 319 patients TP used for 29 subjects PtsOverall FN rate Overall Sensitivity Sensitivi ty for ILC Sensiviti ty for tumor > 2 mm in size Sensitivi ty for microme ts Specifici cty FS31914.4%85.60%65.2%90.8%54.5%97.8% BLN Assay4.4%95.6%91.3%100%81.8%94.3% TP2954.5%45.5%57.1%25.0%100% BLN Asay36.4%63.6%100%0%100%

20. Abstract 28, Multiplex Molecular Assay Are these results real? – RT-PCR has been used for evaluation of axillary nodes previously, not novel – Prior difficulty has been false positive findings and complexity of performing assay in “real time” – Addition of “real time” evaluation of nodal material, and quantitative assessment of RT-PCR findings is novel – No other publications for direct comparison

21. Abstract 28, Multiplex Molecular Assay Where do we go from here? – Additional data forthcoming from company regarding cost Weigh cost of assay versus cost of return trip to OR for completion ALND Long term – Where will SNB fit compared to gene signature assays, such as Mammaprint, Oncotype DX? Will SNB become obsolete?

22. Abstract 27, The RACS/SNAC Trial 32 Sites in Australia/New Zealand – SNB + ALND (RAC), 544 patients – SNB +/- Delayed ALND if SN + (SNBM), 544 patients Surgeon accreditation required – Lymphoscintigraphy and blue dye Outcomes measures – To assess performance of SNB – To assess morbidity of SNB vs ALND in first 12 mos Subjective symptoms Objective findings Complication rates

23. Abstract 27, The RACS/SNAC Trial Patient characteristics – 1,088 patients – Mean age 60 – Identification Screening 58% – Mean tumor size 1.6 cm – Breast conservation 87% Mapping technique – Tracer 89% – Blue dye 99% Blue dye alone 11% – Mean number of SN, 1.7 nodes – Mean number of SN in RAC 14.6 nodes +/- 7

24. Abstract 27, The RACS/SNAC Trial Results – SNB performance Conclusions – SNB is accurate – SNB has lower morbidity Should dysfunction in SNBM group decreased over time Arm swelling in RCA group increased over time SN Not found SN +/-SN only + node Sensitivity, FN rate Proportion with > 15% inc arm vol (p=.05) All subj sx 0 to 100 (worst) (p <.001) Complications Seroma SNBM4%29%/67%63%4.2%4.393 RAC7%25%/69%92%, 8%7.1%7195

25. Abstract 27, The RACS/SNAC Trial Are these findings real? – SB accuracy rate comparable to that seen in other randomized trials – Lymphedema rate for SNB alone arm higher than originally expected, but lower than that seen in other trials NSABP B32 Z10 Almanac

26. Abstract 27, The RACS/SNAC Trial Where does one go from here? – SNB alone remains attractive from a quality of life perspective Still SNB alone is not without significant symptoms – Is SNB oncologically safe? Survival data lacking from this study Question remains unanswered –NSABP B32 pending –Z10 pending

27. CALGB Study C9343 Update Clinical T1N0, ER +, age ≥ 70 636 women enrolled 1994 to 1999 – Tam RT 317 – Tam No RT 319 Previous report 5 year f/u Current update 7.9 year f/u

28. CALGB Study C9343 Update Principle Outcome Measures – IBTR – Frequency of mastectomy – Time to distant metastases – Breast cancer specific mortality – All cause mortality

29. CALGB Study C9343 Update Principle Outcome Measures Conclusions – WB-XRT reduces IBTR 5.3% for women ≥ 70, clinical T1N0, ER pos – WB-XRT reduction in mastectomy 1% vs 3%, p=NS at 7.9 years f/u – Breast cancer specific mortality identical at 2% at 7.9 years f/u – All cause mortality ≈ 26% for both w or w/o RT at 7.9 years f/u Tam + RTTam BreastAxillaBreastAxilla IBTR3 (1%)020 (6.3%)4(p <.001) Mastectomy3 (1%)9 (3%)p =.07 Time to DM9 (3%)11 (3%)p =.59 Breast Cancer Specific Mortality5 (2%) p =.92 All Cause Mortality82 (26%)86 (27%)p =.84

30. CALGB Study C9343 Update Is this real? Probably so. – NSABP B-21 – Fyles subset analysis – Milan III This study just proves point – IBTR is less likely as patients age

31. CALGB Study C9343 Update Where does one go from here? Arimidex replacing Tamoxifen – Arim reduces IBTR compared with Tam – Arim alone even more compelling in women ≥ 70 However, APBI will replace WB-XRT – APBI easier on patients, better tolerated, than WB-XRT – Will effectiveness of lower morbidity of APBI diminish arguments against WB-XRT in terms of time, cost, complications? NSABP B39 in progress