INTRODUCCIN A LA INMUNOTERAPIA Guillermo Lerzo Especialista en

INTRODUCCIÓN A LA INMUNOTERAPIA Guillermo Lerzo Especialista en Oncología

Tumor Immunology: Overview perforin granzyme cytokines Resting T cell Activated T cell TUMOR Tumor antigen LYMPH NODE T cell clonal expansion TCR MHC B 7 Dendritic cell CD 28

Tumor-Derived Immune Suppression • Tumors go to great lengths to evade the immune response. • Systematic studies have identified multiple mechanisms cancers employ to defeat the immune response: – Immunosuppressive cytokines: TGF-β, IL-4, -6, -10. – Immunosuppressive immune cells: T-regs, macrophage. – Disruption of immune activation signaling: loss of MHC receptor, IDO production. • Goal: therapy strategies that “liberate” underlying anticancer immune responses. Weiner LM. N Engl J Med. 2008; 358: 2664 -2665.

Immunotherapeutic approaches to breast cancer Mary L. (Nora) Disis University of Washington Fred Hutchinson Cancer Research Center Seattle, WA ndisis@uw. edu

Major categories of the immune system Non-specific No antigens No memory Immediate Transient Specific Antigens Memory Slowly developing Lifelong drrajivdesaimd. com

Clinically effective anti-tumor immunity Fridman et al, Nat Rev Ca, 2012 Bindea et al, Curr Opin Immunol, 2010 High magnitude Type I CD 4 (Tbet+), CD 8 (GZB+) Memory Low levels of regulatory cells

Approaches to optimizing a therapeutic immune response Butt et al Oncogene, 2013 Increase effector T-cells Trastuzumab Vaccines Adoptive T-cell Therapy Enhance existing immunity Checkpoint inhibitors Cytokine Therapy (IL-15, IL-7) Depletion Tregs, MDSC Mo. AB (X-IL-10, TGFb) Modulate the tumor microenvironment

What is the optimal receptor-ligand interaction to target? Activate Stimulatory signals Suppress Inhibitory signals **** Topalian et al, JCO, Use early in treatment course in a subset of breast cancer: mutation status, high levels of TIL?

Approaches to optimizing a therapeutic immune response Butt et al Oncogene, 2013 Increase effector T-cells Trastuzumab Vaccines Adoptive T-cell Therapy Enhance existing immunity Checkpoint inhibitors Cytokine Therapy (IL-15, IL-7) Depletion Tregs, MDSC Mo. AB (X-IL-10, TGFb) Modulate the tumor microenvironment

Trastuzumab induced Type I immunity IFN-g secretion Ferris et al, JCO, 2010 n=97, Stage III/IV HER 2+ Stanton et al, 2014

Immunotherapeutic approaches to breast cancer Tumor immune environment • Level of TIL • Phenotype of TIL (Type I, II and regulatory) Provide Type I immunity Elicit Type I immunity Release Type I immunity Disis et al, CCR Focus, 2013 Propagate immune response

San Antonio Breast Cancer Symposium, December 9 -13, 2014 Tumor Infiltrating Lymphocytes (TILs) in Breast Cancer Associate Professor Sherene Loi, MD, Ph. D Consultant Medical Oncologist Head, Translational Breast Cancer Genomics and Therapeutics lab Peter Mac. Callum Cancer Centre, Melbourne, Victoria, Australia This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute.

San Antonio Breast Cancer Symposium, December 9 -13, 2014 What is the evidence that immunity is important in breast cancer? • Breast cancer incidence increases in age • Breast cancer in young women is more aggressive • Immunosuppressed patients have worse outcomes from breast cancer • TILs and immune-related gene signatures have been shown to have associations with prognosis in some breast cancer subtypes • Objective responses to T cell checkpoint inhibitors have observed in breast cancer (data this meeting) This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute.

Tumor infiltrating lymphocytes (TILs)- why evaluate TILs? • First publication in EJC in 1992 Aaltomaa et al 1992 • Immune gene signatures are associated with prognosis in ER-negative breast cancer Desmedt et al, 2008; Teschendorff et al 2007; Alexe et al, 2007; Rody et al 2009 • TILs represented a feasible way of evaluating the prognostic and predictive role of immunity in large cohorts of well annotated breast cancer samples (ultimate marker will depend on clinical utility) This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute.

Predefined parameters for TILs evaluation intratumoral TILs = direct contact to tumor cells stromal TILs = between the tumor cells LPBC = Lymphocytepredominant breast cancer „more lymphocytes than tumor cells“ (≥ 60% TILs /≥ 50% TILs ) TLS (tertiary lymphoid structures)= follicular aggregates outside of the tumor Salgado, Denkert et al, Guidelines for TILs Evaluation in Breast Cancer- Annals of Oncology 2014 This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute. Courtesy C Denkert

Predefined parameters for TILs evaluation intratumoral TILs = direct contact to tumor cells stromal TILs = between the tumor cells LPBC = Lymphocytepredominant breast cancer „more lymphocytes than tumor cells“ Salgado, Denkert et al, Guidelines for TILs Evaluation in Breast Cancer- Annals of Oncology 2014 This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute. Courtesy C Denkert

Higher levels in HER 2+ and TNBC This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute. Loi et al, JCO 2013; Ann Oncol 2014

Higher TILs=better survival in primary TNBC Primary TNBC, prior to Chemo P=0. 01 MFS OS Post-neoadjuvant setting in TNBC Dieci et al, Ao. O 2014; Loi et al, JCO 2013; Ao. O 2014

TILs prognostic in HER 2+ treated with anti. HER 2 agents and CT For every 1% increase in TILs, 3% decrease in risk of an event, independent of treatment arm (trastuzmab, lapatinib and combination). Unpublished data- Neo. ALTTO study This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute.

Clinical implication of TILs • Pre-existing host anti-tumor immune responses • The more you have, the better outcome from 1. 2. 3. Primary HER 2+ breast cancer treated with anti-HER 2 agents and chemo Primary TNBC treated with adjuvant anthracycline-based chemo Probably also in metastatic disease • Role in clinical decision making? • Role in predicting response to T cell checkpoint inhibitors and other immunotherapies This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute.

Converting tumors from low TILs into high TILs • Immunogenic chemotherapyanthracyclines, metronomic chemo, gemcitabine • Radiotherapy can drive a T cell response. – Dose and schedule could be critical – Combinations with immunotherapies could be beneficial BOSTON-II study- NCT 02303366 Verbrugge et al 2012; Dewan et al, 2009; Klug et al, 2013

Conclusions • TILs represent functional pre-existing Th 1 immunity • Why some breast cancers do and do not have varying levels of TILs remains to be elucidated • Role of TILs in clinical decision making – Analytical validity of TILs biomarker ongoing – Clinical utility of TILs remains to be determined • However prognostic associations of TILs supports the concept that immune approaches may improve outcome in HER 2+ BC and TNBC – TNBC- combination therapy This presentation is the intellectual property of the author/presenter. Contact sherene. loi@petermac. org for permission to reprint and/or distribute.

Abstract S 1 -06: Edith A. Perez et al. • Stromal tumor-infiltrating lymphocytes (Str-TILs): In the Alliance N 9831 trial Str-TILs are associated with chemothetapy benefit but not associated with trastuzumab benefit.

N 9831 Trial Incorporating Trastuzumab in Adjuvant Therapy. • 945 patients with HER 2 positive breast cancer. • 3 Arms: A. AC – T B. AC – T – H C. AC – T + H - Str-TILs defined as % tumor stromal that contains lymphocytic infiltrate (LI). - Str-TILs measurements: > 60% classified as “lymphocyte predominant breast cancer (LPBC).

Univariable Str-TILs Results (1) • Tumors with high Str-TILs were more likely to be hormone receptro negative (p< 0. 0001) • In Arm A (chemotherapy): - LPBC patients: 10 ys RFS = 90. 9%. - non-LPBC patients: 10 ys RFS = 64. 3%. - HR = 0. 22; 95% CI 0. 07 to 0. 68, p=0. 009. In Arm C (chemptherapy + trastuzumab): - LPBC patients: 10 ys RFS = 80. 0% - non-LPBC patients: 10 ys RFS = 79. 6%. - HR = 1. 13; 95% CI 0. 45 to 2. 84, p=0. 79.

Univariable Str-TILs Results (2) • In LPBC patients group (Str-TILs > 60%): - Arm A: 10 ys RFS = 90. 9%. - Arm C: 10 ys FRS = 80. 0%. - HR = 2. 43%; 95% CI 0. 58 to 10. 22, p = 0. 22. . In non-LPBC patients group (Str-TILs < 60%): - Arm A: 10 ys RFS = 64. 3%. - Arm C: 10 ys RFS = 79. 6%. - HR = 0. 49; 95% CI 0. 35 to 0. 60, p<0. 0001.

Str-TILs: Multivariable Results. • Variables: nodal status, HR status, tumor size, tumor grade, age. . Dichotomous cutoff of Str-TILs: LPBC status - associated with RFS in Arm A HR= 0. 19; 95% CI 0. 66 to 0. 61, p=0. 005 - not associated with RFS in Arm C HR= 1. 01; 0. 95% CI 0. 39 to 2. 6, p=0. 98. Increasing Str-TILs deciles - associated with RFS in Arm A (p<0. 0001) - not associated with RFS in Arm C (p=0. 13)

CONCLUSIONS • Provocative results: - increasing % Str-TILs correlates with benefit of chemotherapy in early stage HER 2+ BC. - impact of adding adjuvant trastuzumab not as clear in patients with LPBC. . Plans: corroborate in a separate cohort. . Identify subtypes of Str-TILs. . Correlate Str-TILs with inmune gene profiles. . Determine whether changing the amount and type of Str-TILs will improve patients outcome.

Str-TILs in Early Stage HER 2+ BC: Conclusions. • Increasing Str-TILs associated with increased RFS in pts treated with chemotherapy. - not foun to be associated with increased RFS in pts treated with chemotherapy plus trastuzumab. . Patients with non-LPBC had better RFS when treated with chemotherapy + trastuzumab compared to chemotherapy alone. - but pts with LPBC did not have better RFS when treated with chemotherapy + trastuzumab than chemotherapy alone.

Checkpoint Protein Inhibition 2014 San Antonio Breast Cancer Symposium San Antonio, TX December 9 th, 2014 Jeffrey Weber M. D. Ph. D. Moffitt Cancer Center

Immune Checkpoint Pathways CTLA-4 = cytotoxic T-lymphocyte-associated antigen 4 ; MHC = major histocompatibility complex; PD-1 = programmed death-1; PD-L 1 = programmed death ligand 1; TCR = T-cell receptor. 31

Immunotherapy for breast cancer: Myth or Fact? • Tumor infiltrating lymphocytes (TIL) in primary triple negative breast cancer after neo-adjuvant chemotherapy are associated with better RFS, DMFS, OS 1 • TIL in stroma and tumor tissue are associated with RFS and OS after adjuvant chemotherapy 2 • PD-L 1 expression and TIL in primary breast cancer are associated with a better outcome 3 • Myeloid derived suppressor cells are associated with high likelihood of nodal metastases in breast cancer 4 1: Adams, S et al JCO 2014 2: Loi, Y et al JCO 2013 3: Schalper, K et al CCR 2014 4:

PD-1/PD-L 1 blocking agents in development • Pembrolizumab - humanized Ig. G 4 anti PD 1 antibody, approved for second line therapy of melanoma • Nivolumab, human Ig. G 4 anti PD-1 antibody, approval for melanoma pending • MPDL-3280 A, humanized PD-L 1 antibody • MEDI 4736, human Ig. G 1 PD-L 1 antibody • AMP 224, fusion of Fc and anti-PD-L 1

Immunotherapy for breast cancer: Myth or Fact? • Tumor infiltrating lymphocytes (TIL) in primary triple negative breast cancer after neo-adjuvant chemotherapy are associated with better RFS, DMFS, OS 1 • TIL in stroma and tumor tissue are associated with RFS and OS after adjuvant chemotherapy 2 • PD-L 1 expression and TIL in primary breast cancer are associated with a better outcome 3 • Myeloid derived suppressor cells are associated with high likelihood of nodal metastases in breast cancer 4 1: Adams, S et al JCO 2014 2: Loi, Y et al JCO 2013 3: Schalper, K et al CCR 2014 4:

PD-1 blockade: Myth or Fact? • PD-L 1 staining is a predictive marker useful for choosing melanoma patients for PD 1/PD-L 1 blockade • There appears to be an association between ORR and tumor PD-L 1 positivity by IHC in most trials • Patients may still respond even if tumor PD-L 1 staining is negative • Equivocal data on association of PD-L 1 staining with overall survival.

PD-L 1 expression and response rate N PDL 1 + Positive PDL 1 Negative Nivolumab (Topalian, NEJM, 2012) 42 9/25 (36%) 0/17 (0%) Nivolumab (Weber #9011) 44 8/12 (67%) 6/32 (19%) MPDL 3280 A (Hamid #9010) 30 4/15 (27%) 3/15 (20%) Nivolumab/ Ipilimumab (Callahan #3003) 27 4/10 (40%) 8/17 (47%) Nivolumab (Grosso #3016) 34 7/16 (44%) 3/18 (17%) Urba, W ASCO 2014

Efficacy Based on Tumor PD-L 1 Expression (Central Review, RECIST v 1. 1) 100 P = 0. 0007 a PFS, % 80 60 PD-L 1+ 40 20 P = 0. 0051 PD-L 1– 0 0 20 40 Time, weeks 100 80 PD-L 1+ 80 OS, % 60 60 PD-L 1– 40 P = 0. 3165 20 0 a 1 -sided P values calculated by logistic regression, adjusting for dose/schedule. PD-L 1 positivity defined as staining in ≥ 1% of tumor cells. Analysis cut-off date: 18 October 2013. Daud A et al. Presented at: 2014 Annual AACR Meeting; April 5 -9, 2014; San Diego, CA. Presented by: Richard Kefford, ASCO 2014 0 20 40 60 Time, weeks 80

Overall survival based on tumor PD-L 1 expression by IHC does appear to favor PD-L 1+ tumors Ribas, A et al SMR 2014

Conclusions: Checkpoint protein inhibition for breast cancer • Many different histologies now respond to checkpoint protein inhibitory drugs, including breast cancer! • Slow regression, progression prior to regression are common in immuno-oncology and require new response criteria to accommodate ir. RC responses. • Immune related adverse events are a new field for toxicity management and require a learning curve. • Prolonged duration of response and plateauing of survival curves suggest that cures are possible. • The Law of Unintended Consequences suggests that new and unexpected toxicities will occur.

Clinical Development of Inhibitors of PD-1 Immune Checkpoint Target PD-1 PD-L 1 Antibody Molecule Development stage Nivolumab (BMS-936558) Fully human Ig. G 4 Phase III multiple tumors (melanoma, RCC, NSCLCa, HNSCC) Pembrolizumab (MK-3475) Humanized Ig. G 4 Phase I-II multiple tumors Phase III NSCLC/melanoma Pidilizumab (CT-011) Humanized Ig. G 1 Phase II multiple tumors MEDI-4736 Engineered human Ig. G 1 Phase I-II multiple tumors MPDL-3280 A Engineered human Ig. G 1 Phase I-II multiple tumors Phase III NSCLC MSB 0010718 C Fully human Ig. G 1 Phase I solid tumors

Abastract S 1 -09: Rita Nanda et al. • A phase IB study of pembrolizumab (MK-3475) in patients with advanced triple-negative breast cancer. University of Chicago, IL. • Pembrolizumab (MK-3475) is a humanized Ig. G 4, High-Affinity, Anti-PD-1 Antibody: - High affinity for the PD-1 receptor. - Dual ligand blockade of PD-L 1 and PD-L 2. - No cytotoxic activity. - PK supports dosingevery 2 weeks or every 3 weeks. - Demostrated clinical activity in multiple tumor types.

KEYNOTE-012: triple-negative breast cancer cohort. • • Recurrent or metastatic triple-negative BC. ECOG PS 0 – 1. PD-L 1 tumor expression was assesed: 58% of all pts. No systemic steroid therapy. No autoimmune disease. No active brain metastasis. Response assesment every 8 weeks. Treatment: Pembrolizumab 10 mg/kg iv Q 2 W.

Treatment-Related Adverse Events with Incidence >5%. • Any grade: arthralgia fatigue myalgia nausea ALT/AST increased diarrhea erythema headhache. Grade 3 -4: headhache 18. 8% 15. 6% 6. 3% 3. 1%

Best Overall Response (RECIST 1. 1) • Overall response rate: 5 (18. 5%) • Best overall response: - complete response 1 (3. 7%) - partial response 4 (14. 8%) - stable disease 7 (25. 9%) - progressive disease 12 (44. 4%) - no assesment 3 (11. 1%) * 66% pts > 4 lines for metastatic disease.

Pembrolizumab: summary. • Pembrolizumab showed an acceptable safety and tolerability profile in pts with heavily pretreated, PD-L 1 -positive. Advanced triple-negative breast cancer. • Pembrolizumab was associated with an OOR of 18. 5%. • Response was durable, with the median response duration not reached (range, 15 to 40+ weeks) and 3 of 5 responders on treatment for >11 months. • The acceptable safety and tolerability profile and promising antitumor activity support the further development of pembrolizumab in patients with advanced triple-negative breast cancer.