Breast Cancer Immunotherapy. Leisha A. Emens, MD PhD Johns Hopkins University Bloomberg Kimmel Institute for Cancer Immunotherapy

Transcription

1 Breast Cancer Immunotherapy Leisha A. Emens, MD PhD Johns Hopkins University Bloomberg Kimmel Institute for Cancer Immunotherapy

2 Conflict of Interest I have the following financial relationships to disclose: Consultant for: Vaccinex, Celgene, Bristol Meyers Squibb, AstraZeneca, Amgen, Syndax, Molecuvax, etherna, Peregrine, Bayer Grant/Research support from: Genentech/Roche, EMD Serono, Maxcyte, Merck, AstraZeneca, Aduro, Corvus Under a licensing agreement between Aduro Biotech, and the Johns Hopkins University, the University and Dr. Emens are entitled to milestone payments and royalty on sales of a GM- CSF-secreting breast cancer vaccine. The terms of these arrangements are being managed by the Johns Hopkins University in accordance with its conflict of interest policies I will discuss the following off-label use/investigational use: Avelumab, Pembrolizumab, Atezolizumab for breast cancer

3 Breast Cancer Can Induce Adaptive Immune Responses Poor prognostic factors (ER neg, PR neg, high grade, LN + ) are associated with higher T cell infiltrates at diagnosis Higher numbers of CD8 + TILs predict longer DFS and OS Total CD3 + T cells, untreated breast cancer Cimino-Mathews/Taube/Emens Ratio of CD8 + T cells/foxp3 + Tregs can predict cpr, PFS, and OS Emens et al, 2012, Expert Rev Anticancer Therapy 12: Cimino-Mathews/Emens et al, 2016, Human Pathol 47:

4 Targeting the PD-1 Pathway in Breast Cancer Ribas A. N Engl J Med 2012;366: Cimino-Mathews/Taube/Emens et al Human Pathol 2016; 47: 52-83; Cimino-Mathews/Foote/ Emens Oncology 2015; 29: Of the breast cancer subtypes, TNBC is a particularly attractive candidate for cancer immunotherapy: Higher rate of mutational complexity Presence of PD-1 + TIL Higher rates of PD-L1 expression by tumor cells and immune cells No current targeted therapy options

5 AvelumabActivity in Breast Cancer Patients 168 unselected patients/all BC subtypes ORR in breast cancer = 4.8% 1 CR, 7 PRs, 39 patients with SD; DCR = 28% DirixL et al SABCS 2015

6 Keynote 012: Pembrolizumabin TNBC Change From Baseline, % Tumor Burden Over Time ORR: 18.5%; PFS rate at 24 weeks: 23.3% (n= 27) 1CR, 4 PRs, 7 SD, 3 of 5 responses ongoing Median PFS = 1.9 months (95% CI ) Time, weeks On treatment Discontinued treatment 58% of all patients screened had PD-L1+ tumors Analysis cut-off date: November 10, Time in Weeks Nanda R et al. JCO 2016;34:2460.

7 Keynote 028: Pembrolizumabin ER+ HER-2- Breast Cancer Tumor Burden Over Time % of 248 tumors screened were PD-L1+ ORR: 12%; PFS rate at 24 weeks: 20% (n=20) No CRs, 3 PRs, evidence of pseudoprogression 75 Responder Change From Baseline, % new bone lesion Nonresponder new liver lesions Time, weeks Investigator Assessment Data cutoff date: July 1, 2015 Rugo, SABCS, 2015

8 Atezolizumabin Metastatic TNBC: Patient Population Baseline Characteristics Patients (N = 115) Medianage (range) 53 y (29to 82) ECOG PS, % 52% 2% Safety-Evaluable Patients Received 1 dose of atezolizumab (N = 115) Visceral metastaticsites a 65% Bone metastaticsites b 30% PD-L1 status on IC c IC0/1 (< 5%) 33% IC2/3 ( 5%) 63% Median prior systemic therapies (range) d 7 (0to 21) Anthracycline taxane 85% 94% Platinum bevacizumab 58% 21% Currentline of therapy, e 1L 2L 3L+ 17% 24% 58% Prior to receiving atezolizumab, most patients were heavily pretreated 1L, first line; 2L, second line; 3L, third line. a Includes lung, liver, adrenal and pelvis metastatic sites. b Includes bone and other sites. c Four patients (4%) had unknown IC status. d Refers to all treatment settings. e Refers to treatment in metastatic setting only. Data cutoff: March 31, Efficacy-Evaluable Patients Had 12 weeks of follow-up (n = 113) Objective Response Evaluable Patients (n = 112) At data cutoff, median treatment duration was 2.1 mo (range, ) Median of 4 cycles (range, 1-45) Patients without RECIST measurable disease at baseline were excluded Schmid P, et al. AACR 2017 Phase Ia Atezolizumab in TNBC

9 Change in Tumor Burden On All Response-Evaluable Patients Atezolizumab 1 Year RECIST v1.1 Response Clinical benefit was observed in some patients with RECIST v1.1 SD or PD status irpr 2 Years Overall TNBC cohort irpr a irpr Criteria MedianDOR (range) Median PFS (95% CI) irpr a RECIST v mo (2.8 to 26.5+) 1.4 mo (1.3, 1.6) irrc irpr, PR per irrc; SLD, sum of target lesion longest diameter. a Re-treatment period not plotted. Confirmed, investigator-assessed RECIST responses are included for patients with post-baseline tumor measurements. Data cutoff: March 31, mo (2.8 to 33.9+) 1.9 mo (1.4, 2.6) Schmid P, et al. AACR 2017 Phase Ia Atezolizumab in TNBC

10 Change in Tumor Burden On Atezolizumab Patients With RECIST v1.1 Response or Stable Disease or irrc Response 1 Year RECIST v1.1 Response Clinical benefit was observed in some patients with RECIST v1.1 SD or PD status Overall TNBC cohort irpr 2 Years Criteria MedianDOR (range) Median PFS (95% CI) irpr a irpr a irpr RECIST v mo (2.8 to 26.5+) 1.4 mo (1.3, 1.6) irrc 21.1 mo (2.8 to 33.9+) 1.9 mo (1.4, 2.6) irpr, PR per irrc; SLD, sum of target lesion longest diameter. a Re-treatment period not plotted. Confirmed, investigator-assessed RECIST responses are included for patients with post-baseline tumor measurements. Data cutoff: March 31, Schmid P, et al. AACR 2017 Phase Ia Atezolizumab in TNBC

11 Objective Response Rate to Atezolizumabby Subgroups ORR 60% 50% 40% 30% 20% RECIST v1.1 ORR 26% Numerically higher ORRs were observed in IC2/3 and 1L subgroups 10% 0% ORR 95% CI, % b 10% Overall n = 112 a 13% IC2/3 n = 71 5% IC0/1 n = 37 1L n = 19 2L n = 28 3L+ n = 65 5, 17 6, 23 1, 18 9, 51 0, 18 3, 17 4% 8% a Objective response evaluable patients. Four patients had unknown PD-L1 status. Confirmed, investigator-assessed responses are plotted. Patients with missing or unevaluable responses are included (16 per RECIST v1.1 and 23 per irrc). b ORR 95% CI was estimated using Clopper- Pearson method. Data cutoff: March 31, Schmid P, et al. AACR 2017 Phase Ia Atezolizumab in TNBC

12 Overall Survival by Response Status (RECISTv1.1 and irrecist) RECIST v1.1 Criteria 1-y OS: 100% irrc Criteria 2-y OS: 100% 3-y OS: 100% 1-y OS: 100% 2-y OS: 100% Overall Survival 1-y OS: 33% 1-y OS: 69% RECIST Response b CR/PR (n = 11) SD (n = 15) PD (n = 70) Overall Survival 1-y OS: 33% 1-y OS: 51% irrcresponse b CR/PR (n = 15) SD (n = 19) PD (n = 55) 11% 2-y OS: 11% 3-y OS: 11% Time (months) No. At Risk: CR/PR SD PD Time (months) No. At Risk: CR/PR SD PD Median OS was 9.3 mo(95% CI: 7.0, 12.6) in all patients a Landmark OS rates (95% CI) were: 41% (31, 51) at 1 year, and 22% (12, 32) at both 2 and 3 years Pseudo-progression was observed in patients with RECIST PD and long-term OS Schmid P, et al. AACR 2017 Phase Ia Atezolizumab in TNBC a Median survival follow-up (range) was 15.2 mo (0.4+ to 36.7) in all patients, 17.0 mo (0.43+ to 36.7) in IC2/3 patients and 12.8 mo (0.8+ to 16.9) in IC0/1 patients. b Patients included in the Kaplan-Meier plots were alive for 6 weeks. Data cutoff: March 31, 2016.

13 Overall Survival by PD-L1 and TIL Status OS Based on PD-L1 Status PD-L1 Status IC2/3 (n = 71) IC0/1 (n = 38) OS Based on TIL Status TIL Levels a > 10% (n = 53) 10% (n = 56) Overall Survival 1-y OS: 37% 1-y OS: 45% 2-y OS: 28% 3-y OS: 28% Overall Survival P= Time (months) No. At Risk: IC2/ IC0/ Time (months) No. At Risk: > 10% % a Four patients had unknown PD-L1 status. Median survival follow-up (range) was 15.2 mo (0.4+ to 36.7) in all patients, 17.0 mo (0.43+ to 36.7) in IC2/3 patients and 12.8 mo (0.8+ to 16.9) in IC0/1 patients. Median TIL level based on median TIL. a Samples unevaluable for TIL assessments (6 per RECIST v1.1 and 5 per irrc) are not included. Objective response evaluable population includes patients with unevaluable response assessments (16 per RECIST v1.1 and 23 per irrc). Log-rank (Mantel-Cox) P value is exploratory. Data cutoff: March 31, Schmid P, et al. AACR 2017 Phase Ia Atezolizumab in TNBC

14 PD-1/PD-L1 Blockade in Breast Cancer Antibody Target Subtype Patients ORR Avelumab PD-L1 All % PD-L1+ All % TNBC % PD-L1+ TNBC % Pembrolizumab PD-1 PD-L1+ TNBC % PD-L1+ ER+HER % Atezolizumab PD-L1 TNBC % PD-L1+ TNBC 71 13% DirixL et al SABCS 2015 Nanda R et al JCO 2016;34:2460 Rugoet al SABCS 2015 Emens LA et al AACR 2015 SchmidP et al AACR 2017

15 The Major Challenge of Breast Cancer Immunotherapy Today Converting nonrespondersto responders

16 One Framework for Personalizing Breast Cancer Immunotherapy Non-inflamed Patterns of T Cell Infiltration Inflamed Chemotherapy, XRT HER-2-directed antibodies Vaccines, STING agonists Anti-PD-1/PD-L1 IDO inhibition A2AR inhibition Gajewski TF Semin Oncol : HerbstRS et al Nature : Chen DS MellmanI Immunity : Cimino-Mathews A/Emens LA, unpublished images.

17 Atezolizumaband Nab-Paclitaxel Have Activity in Metastatic TNBC Change in Tumor Burden Over Time with Line of Therapy PD-L1 unselected patients Atezolizumab 840 mg every 2W; Nab-paclitaxel 100 mg/m 2 weekly Confirmed ORR = 41.7%; 3 pseudoprogressors n = 9 (ORR ~ 67%) n = 15 (ORR ~ 25-28%) Adams S, et al SABCS 2015

18 Summary of Responses to Atezolizumabwith Best Overall Response & Confirmed ORR* Nab-Paclitaxel by RECIST 1.1 First Line (n = 9) SecondLine (n = 8) > Third Line (n = 7) All Patients (n = 24) 66.7% 25% 28.6% 41.7% ORR # 88.9% 75% 42.9% 70% CR 11.1% % PR 77.8% 75% 42.9% 66.7% SD 11.1% 25% 28.6% 20.8% PD % 8.3% *Confirmed responses based on at least 2 consecutive assessments of CR or PR. # Includes unconfirmed investigator assessment. & Three additional patients appeared to have pseudoprogression Ȧdams S, et al, SABCS 2015

19 ImPassion 130: A Phase III randomized trial of atezolizumab in combination with nab-paclitaxel as first line therapy for patients with metastatic triple-negative breast cancer 1L mtnbc n=900 (1:1 double-blinded) R Atezolizumab 840 mg q2wk + Nab-paclitaxel 100 mg/m 2 qwk 3/4 wks Placebo q2wk + Nab-paclitaxel 100 mg/m 2 qwk 3/4 wks Co-Primary Endpoints Progression Free Survival (RECIST 1.1) Overall Survival Secondary Endpoints Progression Free Survival according to immunemodified RECIST, Stratification: Tumour tissue PD-L1 expression (IHC 0 vs IHC 1,2,3) Liver metastases (Yes vs No) Prior taxane treatment (Yes vs No) Overall Response Rate, Duration of Response, Time to Deterioration in Global Health Status and Health Related Quality of Life

20 Combination Immunotherapy for HER-2+ Breast Cancer Stromal TILs increase about 7% on therapy CD4+ and CD8+ T cells increase on therapy Müller P et al, Science Translation Medicine, 2015; 315:315ra188

21 KATE2: A Global Study of Atezolizumab with T-DM1 DESIGN: PHASE II DOUBLE-BLIND MULTICENTER RANDOMIZED PLACEBO-CONTROLLED Patients with HER2+ LABC or mbc - Prior taxane and trastuzumab - Progression on metastatic tx or within 6 months of adjuvant tx - Measurable disease (n=200) Stratification Factors: Tumor PD-L1 Status World Region Liver Metastases 1:2 n=67 n=133 Study Treatment Phase T-DM1, 3.6 mg/kg q3w + Placebo, 1200 mg q3w T-DM1, 3.6 mg/kg q3w + Atezolizumab, 1200 mg q3w

22 Conclusions Breast cancer can be immunogenic, most breast tumors are not Multiple layers of regulation within the TME shut down tumor immunity Standard cancer therapies can augment the activity of immunotherapies The future is in combination immunotherapies which should have synergistic clinical activity but may come at a toxicity cost We need to do smart trials elucidating immunologic mechanisms of response and resistance in patients

23 A special thanks to my colleagues and especially all of the patients and families who participated in these studies!