Esophageal Cancer: Incorporating Molecular Markers Into Clinical Practice

Esophageal Cancer: Incorporating Molecular Markers into Clinical Practice

Nov 2019

Jill Lacy, MD Professor of Medicine (Medical Oncology) Disclosures

• Merck • Celgene • Astra Zeneca Agenda

• Review of impact anatomy, histology, molecular profiling on clinical practice • Current molecular molecular markers in clinical practice – HER2 – PD-L1 – MMR deficiency / microsatellite instability • Future directions and trials in progress Esophageal Cancer: Many Diseases Histology Squamous cell carcinoma Adenocarcinoma Anatomy Proximal and mid esophagus Distal esophagus GE junction Molecular features HER2 PDL1 MMR AJCC Stage Correlate Siewert Classification of Adenocarcinoma of the Esophagus Esophagogastric Junction Adenocarcinoma

Esophagus (EGJ) Adenocarcinoma

Gastric Cancer

Copyrights apply Esophageal Cancer: Anatomic / Histologic “Buckets” in Practice and Trials

Esophagus SCC Esophagus AC (Siewert type I)

GE Junction AC (Siewert type II) Esophageal Cancer: Trial Design and Eligibility

Esophagus SCC Esophagus AC

CROSS Preop chemoRT v surgery GE Junction AC CALGB 9781 Preop chemoRT v surgery Esophageal Cancer: Trial Design and Eligibility

Esophagus SCC Esophagus AC

KEYNOTE 181 2nd line chemotherapy v pembrolizumab Esophageal Cancer: Trial Design and Eligibility

Herskovic RTOG 85-01 (90% SCC) Definitive chemoRT v RT Esophagus SCC Stahl (German) Definitive chemoRT v preop chemoRT

PRODIGE5/ACCORD17 (85% SCC) Definitive chemoRT (FU/cisplatin vs FOLFOX

Chinese NEOCRTEC5010 Preop chemoRT v surgery

Attraction-3 2nd line Nivolumab vs taxane chemo Esophageal Cancer: Trial Design and Eligibility

MAGIC Trial (+ gastric) Peri-operative chemo v surgery

CALGB 80803 Esophagus AC Preop FOLFOX/RT v Preop carbo/taxol/RT

German POET Preop chemo v preop CRT GE Junction AC (included Siewert’s III) Esophageal Cancer: Trial Design and Eligibility

Many studies in metastatic disease excluded squamous and Siewert’s I adenocarcinoma

GE Junction AC Esophageal Cancer: Trial Design and Eligibility

TOGA: Chemo v Chemo + Herceptin (HER2+) REGARD: Ramucirumab vs BSC RAINBOW: Taxol v Taxol + Ramucirumab KEYNOTE 059: 3rd line Pembrolizumab KEYNOTE 061: 2nd line Chemo v Pembrolizumab JAVELIN: 3rd line Chemo vs Avelumab KEYNOTE 062: 1st line Pembro vs Chemo + Pembro vs Chemo

GE Junction AC Gastric Cancer

15-30% 70-85% Molecular Heterogeneity:

The Cancer Genome Atlas (TCGA) Integrated Genomic Characterization of Oesphageal Cancer

The Cancer Genome Atlas Research Network. Integrated genomic characterization of oesophageal carcinoma. 1 2 JA N UA RY 2 0 1 7 | VO L 5 4 1 | N AT U R E | 1 6 9 TCGA Key Features of Gastric Cancer Subtypes

Chromosomal instability (CIN)

Genomic stability (GS) Nature. 2014 September 11; 513(7517): 202–20 Molecular Heterogeneity: TCGA Analysis of Esophageal Caner Esophageal SCC vs AC

• Esophageal AC and SCC are distinct in their molecular characteristics • Esophageal SCC emerges as a disease reminiscent of other SCCs

The Cancer Genome Atlas Research Network. Integrated genomic characterization of oesophageal carcinoma. 1 2 JA N UA RY 2 0 1 7 | VO L 5 4 1 | N AT U R E | 1 6 9 Molecular Heterogeneity of Esophageal Cancer: TCGA

• 98.6%Esophageal of esophageal ACsAdenocarcinoma characterized by chromosomal (vs Gastric instability AC)(CIN), and strongly resemble CIN gastric cancer. • No esophageal ACs were positive for MSI or EBV. • No consistent separation of esophageal AC and CIN gastric cancers, arguing against classifying these as distinct diseases • Among GEJ ACs, MSI-positive, EBV-positive and GS tumors were identified (low incidence) • Gastric, GEJ, and esophageal ACs show a progressive gradation of subtypes, with increasing prevalence of the CIN proximally, to the point that esophageal ACs appear to represent a disease of chromosomal instability. Molecular Heterogeneity: TCGA Key Features of Esophageal Subtypes

SCC

Gradations of molecular subtypes of gastro esophageal cancer. Shifting proportions of subtypes from proximal esophagus to distal stomach. Widths of the color bands represent the proportion of subtypes at each anatomic region. Molecular Heterogeneity: TCGA Analysis of Esophageal Caner SCC vs AC • Molecular data argue against combining esophageal AC and SCC for clinical trials • Molecular data argue against excluding esophageal AC from clinical trials for gastric/GEJ cancer

The Cancer Genome Atlas Research Network. Integrated genomic characterization of oesophageal carcinoma. 1 2 JA N UA RY 2 0 1 7 | VO L 5 4 1 | N AT U R E | 1 6 9 Localized Esophageal Cancer: Incorporating Molecular Markers into Clinical Practice: Not yet!

• No biomarkers currently in clinical practice for localized potentially curable esophageal cancer

• Studies with trastuzumab and pembrolizumab in progress Trastuzumab in Localized HER2-positive Esophageal/GEJ Adenocarcinoma Pembrolizumab in Localized Gastric/GE Adenocarcinoma KEYNOTE 585

Phase III Randomized, Double-blind Trial of Peri-operative Pembrolizumab plus Chemotherapy (XP, FP, FLOT) vs Placebo plus Chemotherapy (XP, FP, FLOT) for Patients with Gastric and Gastroesophageal Junction* Adenocarcinoma (KEYNOTE 585)

*Siewert type 1 tumors limited to those for whom the planned treatment is perioperative chemotherapy and resection. Pembrolizumab in Localized Gastric or GE Junction Adenocarcinoma

or FLOT + Pembro or FLOT + Pembro

or FLOT + Placebo or FLOT + Placebo

KEYNOTE 585 Metastatic Esophageal Cancer: Incorporating Molecular Markers into Clinical Practice in 2019

HER2 PD-L1 MMR deficient/MSI high HER2/ERBB2

• First biomarker approved for use in clinical practice in gastroesophageal cancer HER2/ERBB2 Incidence and Testing in GE Adenocarcinoma • Overall incidence approx 20% • Incidence varies with histologic subtype . Intestinal 30% . Diffuse 5% . Mixed 20% • Incidence varies with anatomic site . GEJ/esophagus 20-30% . Lower in Gastric 10-20% • Positive result defined as: • IHC 3+ • IHC 2+ and HER2 gene amplified (by FISH HER2/CEP17 ratio ≥2) HER2 HER2/Neu: ToGA Trial Design Randomized Phase III, open-label, multi-center

Stratification: • Advanced v Metastatic • Gastric v GEJ • Meas v non-meas • ECOG PS 0-1, v 2 • 5FU v Xeloda* 18% GEJ 82% Gastric *Chosen at investigator’s discretion

Treatment q 3 wks X 6 cycles in both arms; Herceptin until disease progression Lancet. 2010;376(9742):687 HER2 HER2

Response rate: 47% v 35%

Favorable treatment effect in all predefined subsets including GEJ (HR 0.67) HER2

Addition of trastuzumab to chemotherapy significantlyResponse rate: improved overall 47% v 35% survival compared with chemotherapy alone in pts with advanced HER2-positive gastric or GEJ cancer (13.8 mo v 11.1). HER2

ToGA Analysis: Longer overall survival in pts with high HER2 expression vs in low HER2 expression HER2

ToGA Analysis: Longer overall survival in pts with high HER2 expression vs in low HER2 expression

Longer OS in pts w/ high vs low HER2/ERBB2 copy number.

HER2/CEP17 ratio 4.7 optimal cutoff to discriminate Herceptin- sensitive vs -refractory disease

Gomez-Martin C, et al. J Clin Oncol 2013;31:4445–52. HER2 Does chemo matter? Oxaliplatin preferred to cisplatin

Fifteen studies (N = 557 pts)

Overall survival: oxaliplatin + FU/Cape 20.7 mo cisplatin + FU/Cape/S1 16.0 mo

Oxaliplatin less toxic.

Ter Veer et al. Int J Cancer. 2018 Jul 15;143(2):438-44 HER2 Post-ToGA Observations

• Evidence does NOT support continuing Herceptin after progression on 1st line Herceptin-containing regimen

• Resistance mechanisms defined – loss of expression of HER2 (16-60%) – emergence of subclone lacking HER2/ERBB2 amplification – deletion of ERBB2 exon 16 – comutations in the receptor tyrosine kinase, RAS, and PI3K pathways

Janjigan YY. Genetic Predictors of Response to Systemic Therapy in Esophagogastric Cancer. Cancer Discovery 2018 Jan;8(1):49-58. HER2 Post-ToGA Era

• No other anti-HER2 agents have been approved in gastroesophageal cancer (in contrast to breast cancer) • Several notable negative trials HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma

27% GEJ 73% Gastric HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma

JACOB

Addition of pertuzumab to first line chemotherapy + trastuzumub did not improve survival in HER2-positive GE adenocarcinoma.

There were modest improvements in PFS and RR. HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma

13% esophageal / GEJ HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma

LOGIC

Addition of lapatinib to first line chemotherapy with 13% esophageal capecitabine/ GEJ plus oxaliplatin did not improve survival in HER2-positive GE adenocarcinoma. HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma

32% GEJ 77% prior anti-HER2 therapy HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma GATSBY

TDM-1 did not improve survival compared to a taxane in 32% GEJ previously treated HER-positive GE adenocarcinoma.

Repeat biopsy and HER2 testing not required HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma HER2 Notable Negative Studies in HER2-positive GE Adenocarcinoma

TyTan

Addition of lapatinib to paclitaxel did not improve survival compared to paclitaxel in the 2nd line setting in HER2-positive GE adenocarcinoma.

Repeat biopsy and HER2 testing not required. HER2 Challenges in HER2-positive GE Adenocarcinoma

• CIN subtype – Amplification and activation of multiple RTK pathways • Intra-tumoral heterogeneity of HER2 • High rate of loss of HER2 expression during treatment • Multiple resistance mechanisms HER2 On the Horizon

• Next generation anti-HER2 antibodies • Next generation anti-HER2 antibody-drug conjugates • Anti-HER2 antibodies + immune checkpoint inhibitor combinations HER2 On the Horizon: Better anti-HER2 mAbs

Margetuximab • Anti-HER2 mAb with optimized Fc domain to enhance Fc-dependent functions, esp ADCC • Evaluated with pembro in 2nd line post-Herceptin, with promising preliminary results (NCT02689284) • Repeat biopsy for HER2 testing was not required • HER2 expression was lost in 23/56 (41.1%) of pts tested post-Herceptin (measured by HER2/ERBB2 ctDNA) HER2 On the Horizon: Better anti-HER2 mAb/drug conjugates Trastuzumab deruxtecan • Anti-HER2 antibody, cleavable peptide-based linker, and potent topoisomerase I inhibitor payload. • 44 pts with HER2-positive gastric/GEJ cancer treated in Phase I* • 43% response rate • Phase II study underway in 2nd line post-Herceptin (NCT04014075)

*Shitara et al. Lancet Oncology 20:827, 2019 HER2 On the Horizon: Add IO to anti-HER2 mAb HER2 On the Horizon: Add IO to anti-HER2 mAb

First Line: Combination margetuximab, immune checkpoint inhibitor, and chemotherapy Phase 2/3 Trial in HER2+ Gastric/GEJ Cancer (MAHOGANY, NCT04082364) PD-L1

• Emerging as a predictive biomarker for immuno-oncology therapy • Many challenges in use of PD-L1 as a biomarker generally and in esophageal/GEJ cancers specifically • Many challenges in interpreting the trial data incorporating PD-L1 in esophageal/GEJ cancers. PD-L1 as a Biomarker in Esophageal Cancer

• Several trials of PD1 and PD-L1-targeted agents in gastroesophageal cancers reported • Different histologies, prior therapies, geographic areas • Variable if not confusing/conflicting results • Predictive value of PD-L1 uncertain • Two FDA approvals of PD1-targeted agents in advanced PDL-1 positive gastroesophageal cancers KEYNOTE 059: 3rd line Pembrolizumab in Gastroesophageal Adenocarcinoma

JAMA Oncol. 2018;4(5):e180013. KEYNOTE 059: 3rd line Pembrolizumab

GEJ Siewert’s II (and Gastric) Adenocarcinoma

JAMA Oncol. 2018;4(5):e180013. doi:10.1001/jamaoncol.2018.0013 Published online March 15, 2018.Corrected on March 7, 2019. KEYNOTE 059: 3rd line Pembrolizumab

Key Inclusion: • Gastric or GEJ AC • ≥2 lines of therapy • PDL1 pos or negative • ECOG PS 0,1

Endpoints: • 1o Response rate • 2o PFS, OS, DOR

Patients: N = 259 • GEJ 51% • 2 prior therapies 52% • PDL1 pos 57% • HER2 pos 24% • MSI high 4% KEYNOTE 059: 3rd line Pembrolizumab KEYNOTE 059: 3rd line Pembrolizumab KEYNOTE 059: 3rd line Pembrolizumab

Response Rate: Overall 11.6% PDL1+ 15.5% PDL1- 6.4% 3rd Line 16.4% 4th Line 6.4% MSI high 4 of 7

Disease Control Rate: Overall 27% PDL1+ 33.1% PDL1- 19.3% KEYNOTE 059: 3rd line Pembrolizumab KEYNOTE 059: 3rd line Pembrolizumab Conclusions

• Pembro is active with durable responses in a small subset of patients with advanced gastric/GEJ adenocarcinoma progressing after ≥2 lines of prior therapy

• Response rate higher in PD-L1 positive (CPS ≥1) tumors, but responses observed in PD-L1 negative tumors FDA grants accelerated approval to pembrolizumab for advanced gastroesophageal adenocarcinoma

• On Sept 22, 2017, FDA granted approval to pembrolizumab for pts:  With advanced gastric or GE junction adenocarcinoma (recurrent locally or metastatic)  Whose tumors are PD-L1 positive (CPS ≥ 1%) determined by FDA-approved test.  With disease progression on/after ≥2 prior systemic therapies, including prior FU- and platinum-containing chemo and HER2-targeted therapy if appropriate. • Approval based on the results of KEYNOTE 059 (NCT02335411) • FDA also approved the PD-L1 IHC 22C3 pharmDx (Dako), to select patients with gastric cancer for treatment with pembrolizumab. FDA grants accelerated approval to pembrolizumab for advanced gastroesophageal adenocarcinoma

• On Sept 22, 2017, FDA granted approval to pembrolizumab for pts:  With advanced gastric or GE junction adenocarcinoma (recurrent locally or metastatic)Questionable exclusions from approval:  Whose? Pts tumors with PDL1are PD-negative-L1 positive tumors (CPS ≥ 1%) determined by FDA-approved test.  With? diseasePts with progression esophageal on/after adenocarcinoma ≥2 prior systemic (Siewert’s therapies I) , including prior FU- and platinum-containing chemo and HER2-targeted therapy if appropriate. • ApprovalNotable based evidence on the- basedresults exclusion of KEYNOTE from approval: 059 (NCT02335411) Pts who have had progression after just one line of therapy • FDA also approved the PD-L1 IHC 22C3 pharmDx (Dako), to select patients with gastric cancer for treatment with pembrolizumab. Contrast KEYNOTE 059 with JAVELIN

Bang at al. Ann Oncol 29:2052, 2018 30% GEJ 27% PD-L1 positive CPS ≥ 1% Contrast KEYNOTE 059 with JAVELIN

RR <5% in both arms

Compared to chemotherapy, Avelumab did not improve survival, progression-free survival, or response rate, regardless of PD-L1 status Contrast KEYNOTE 059 with Attraction 2 ATTRACTION 2: ≥3rd line Nivolumab vs BSC in Gastric/GEJ AC

Primary endpoint: -Overall survival

Secondary endpoints: 2:1 -Response rate -Duration of response -PFS

 GEJ 8.5%  ≥3 prior therapies: 80%  PD-L1 data on 39% Contrast KEYNOTE 059 with Attraction 2 Contrast KEYNOTE 059 with Attraction 2

Response Rate: Nivo 11% Placebo 0% Contrast KEYNOTE 059 with Attraction 2 Contrast KEYNOTE 059 with Attraction 2

ATTRACTION 2

Nivolumab improved survival compared to placebo in heavily pre-treated PD-L1-positive and PDL1-negative cohorts.

Survival was comparable in PD-L1-positive and PD-L1-neg cohorts.

Nivolumab is approved in Asia irrespective of PD-L1 status. KEYNOTE 061: 2nd line Pembrolizumab in Gastroesophageal Adenocarcinoma KEYNOTE 061: 2nd line Pembrolizumab

GE Junction 34%

PD-L1 CPS ≥1% 67%

PD-L1 CPS ≥10% 18%

MSI high 5% KEYNOTE 061: 2nd line Pembrolizumab

Current standard of care in 2nd line setting: GE Junction 34% Paclitaxel + Ramucirumab PD-L1 CPS ≥1% 67%

PD-L1 CPS ≥10% 18%

MSI high 5% KEYNOTE 061: 2nd line Pembrolizumab KEYNOTE 061: 2nd line Pembrolizumab KEYNOTE 061: 2nd line Pembrolizumab

2nd line pembrolizumab did not significantly improve overall survival or progression free survival compared to paclitaxel GEJ/gastric adenocarcinoma KEYNOTE 061: 2nd line Pembrolizumab KEYNOTE 061: 2nd line Pembrolizumab

*Response rate w/ Ram/Taxol 28% KEYNOTE 061: 2nd line Pembrolizumab

2nd line pembro did not improve response rate compared to paclitaxel in PD-L1-positive GEJ/gastric adenocarcinoma KEYNOTE 061: 2nd line Pembrolizumab

• Pembro did not improve survival, PFS, or response rate compared to paclitaxel as 2nd line therapy for advanced PD-L1+ gastric/GEJ adenocarcinoma • PFS was notably shorter with pembro vs paclitaxel • Pembro treatment effect appeared greater for PD-L1 CPS ≥ 10 (BUT small numbers, not statistically validated) • Responders to pembro had more durable responses compared to paclitaxel (BUT small numbers, not statistically validated) KEYNOTE 061: 2nd line Pembrolizumab

• Pembro did not improve survival, PFS, or response rate compared to paclitaxel as Take2nd line home therapy message: for advanced KEYNOTE PD-L1+ 061 gastric/GEJ does not adenocarcinoma support routine • PFSuse was of notablypembro shorter in 2nd withline pembrosetting forvs paclitaxeladvanced GEJ/gastric • Pembroadenocarcinomatreatment effect including for survival those appeared with high greater PD-L1 for CPS PD-L1 score* CPS ≥ 10 (BUT small numbers, not statistically validated) • Responders*except in ptsto pembrowith MSI hadhigh moretumors durable responses compared to paclitaxel (BUT small numbers, not statistically validated) KEYNOTE 062: 1st Line Pembrolizumab in PD-L1 positive GE Adenocarcinoma

Pembrolizumab vs Chemotherapy Pembro + Chemotherapy vs Chemotherapy KEYNOTE 062: 1st Line Pembrolizumab in PD-L1 positive GE Adenocarcinoma KEYNOTE 062: 1st Line Pembrolizumab

• Bottom line: Interpretation of Keynote-062 is complicated!

• Opinion: 1st line pembro in any subset not ready for prime time 31% GEJ

37% CPS ≥10%

7% MSI high KEYNOTE 062: 1st Line Pembrolizumab KEYNOTE-062: 1st Line Pembrolizumab (P) vs Chemotherapy (C)

KEYNOTE-062: P vs C

Presented By Josep Tabernero at 2019 ASCO Annual Meeting Overall Survival: P vs C (CPS ≥1)

Presented By Josep Tabernero at 2019 ASCO Annual Meeting Overall Survival: P vs C (CPS ≥1) Primary endpoint met:

1st line pembro is not inferior to chemotherapy in PD-L1 positive gastroesophageal adenocarcinoma

Presented By Josep Tabernero at 2019 ASCO Annual Meeting 6-mo rate 62% 78%

Overall Survival: P vs C (CPS ≥1)

Presented By Josep Tabernero at 2019 ASCO Annual Meeting 6-mo rate 62% 78%

Overall Survival: P vs C (CPS ≥1) During first 12 months of treatment, pts are dying at higher rate with pembro vs chemo

Presented By Josep Tabernero at 2019 ASCO Annual Meeting Overall Survival: P vs C (CPS ≥10)

Presented By Josep Tabernero at 2019 ASCO Annual Meeting In PD-L1-positiveOverall Survival: P vswith C (CPS ≥10) CPS ≥10%

Appeared to be a survival benefit for 1st line pembro compared to chemo.

However, statistical test was not applied to the differences because the statisticians had designed this as a subsidiary question.

Presented By Josep Tabernero at 2019 ASCO Annual Meeting Progression-Free Survival: P vs C

Presented By Josep Tabernero at 2019 ASCO Annual Meeting CautionaryProgression-Free Survival:Notes P vs C

Early progression occurs at higher rate with pembrolizumab compared to chemo in both CPS ≥1 and CPS ≥10 cohorts

PFS favors chemo in both subsets

Presented By Josep Tabernero at 2019 ASCO Annual Meeting Keynote-062 Summary: 1st line Pembro vs Chemo in Advanced GEJ/gastric adenocarcinoma

Summary: P vs C

Presented By Josep Tabernero at 2019 ASCO Annual Meeting KEYNOTE-062: 1st Line Pembro with Chemo vs Chemo Alone

KEYNOTE-062: P+C vs C

Presented By Josep Tabernero at 2019 ASCO Annual Meeting Overall Survival: P+C vs C (CPS ≥1)

Presented By Josep Tabernero at 2019 ASCO Annual Meeting Overall Survival: P+C vs C (CPS ≥10)

No improvement in overall survival with addition of pembro to first line chemo compared chemotherapy alone in PD-L1-positive (CPS ≥1 or CPS ≥10) gastric/GEJ adenocarcinoma.

Presented By Josep Tabernero at 2019 ASCO Annual Meeting Take Away Messages From KEYNOTE-062 1st Line Pembro vs Chemo

• Primary endpoint of non-inferiority of pembro alone versus chemo alone for first line therapy in pts with PD-L1 CPS >1 was met.  For CPS >10 subset, notable survival benefit for pembro, though statistical test not applied as this was a subsidiary question. • PFS curves in both CPS >1 and CPS >10 are revealing and important.  Chemotherapy does much better than pembro for the first 9-12 mo. • Is there an option for the use of immunotherapy in the first line for some patients? Those with positive CPS scores and relatively low tumor burden? Take Away Messages From KEYNOTE-062 1st Line Pembro with Chemo vs Chemo

• No benefit with addition of pembro to chemo vs chemo alone in previously untreated patients with PD-L1 positive advanced gastric/GEJ AC • Why was this a negative result, considering positive findings in lung cancer, head/neck cancer, and breast cancer?? • Biomarker? • Wrong hemo backbone? • Low activity of pembro in general • Biostatistical design KEYNOTE 181: 2nd line Pembrolizumab vs Chemotherapy in Advanced Esophageal Cancer SCC 64% AC 36%

PD-L1 CPS ≥10 35% KEYNOTE 181: 2nd line P vs C in Esophageal Ca KEYNOTE 181: 2nd line P vs C in Esophageal Ca KEYNOTE 181: 2nd line P vs C in Esophageal Ca KEYNOTE 181: 2nd line P vs C in Esophageal Ca KEYNOTE 181: 2nd line P vs C in Esophageal Ca KEYNOTE 181: 2nd line P vs C in Esophageal Ca KEYNOTE 181: 2nd line P vs C in Esophageal Ca FDA approves pembrolizumab for advanced esophageal squamous cell cancer

• On July 30, 2019, the FDA approved pembrolizumab for patients – with recurrent, locally advanced or metastatic, squamous cell carcinoma of the esophagus – whose tumors express PD-L1 CPS ≥10, as determined by an FDA-approved test – with disease progression after one or more prior lines of systemic therapy

• FDA also approved a new use for the PD-L1 IHC 22C3 pharmDx kit as a companion diagnostic device for selecting patients for the above indication. ATTRACTION 3: 2nd line Nivolumab vs Chemo in Esophageal SCC ATTRACTION 3: 2nd line Nivolumab vs Chemo in Esophageal SCC

Stratified: -Geographical region (Japan vs other) -No. of organs w/ mets (≤1 vs ≥2) -Expression of PD-L1 (<1% vs ≥1%)

Primary endpoint: -Overall survival Secondary endpoints: -Response rate -Duration of response -PFS ATTRACTION 3: 2nd line Nivo vs Chemo Esophageal SCC

96% Asian 51% PD-L1 negative 29% PD-L1 ≥10% ATTRACTION 3: 2nd line Nivo vs Chemo Esophageal SCC

Secondary Endpoints:

Response rate: Nivo 19% Chemo 22%

Median response duration: Nivo 6.9 mo Chemo 3.9 mo ATTRACTION 3: 2nd line Nivo vs Chemo Esophageal SCC ATTRACTION 3: 2nd line Nivo vs Chemo Esophageal SCC

2nd line treatment with nivolumab was associated with significant improvement in survival compared to chemotherapy in previously treated pts with advanced esophageal squamous cell carcinoma.

Survival benefit with nivolumab occurred regardless of PD-L1 expression. Summary: PD-L1 Biomarker and Immune Checkpoint Inhibitor Use in Esophageal Cancer • Some pts with esophageal/EGJ cancer derive benefit from anti-PD1 mAbs • FDA approvals for pembro in: – PD-L1 positive esophageal SCC (2nd line, CPS ≥10%) – PD-L1 positive GEJ adenocarcinoma (3rd line, CPS ≥1%) • Anti-PD1 mAbs clearly have greater efficacy in SCC compared to AC in second line setting and beyond • Data are conflicting re predictive value of PD-L1: histology matters, line of therapy matters • BUT benefit seen in subset of pts with PD-L1 negative tumors in all studies Summary: PD-L1 Biomarker and Immune Checkpoint Inhibitor Use in Esophageal Cancer • Anti-PD1 MoAbs should not be combined w/ chemo outside of a clinical trial • Anti-PD1 MoAbs not yet approved in first line setting for any sub-type • Rapid progression/early death observed in pts with PD-L1+ GEJ/gastric adenocarcinoma treated with 1st line anti-PD1 MoAbs • Compelling need for better predictive biomarkers for anti-PD1/PD-L1 MoAb use in gastroesophageal cancer • Majority of pts with esophageal/EGJ cancer do not derive benefit, highlighting need for better biomarkers and better I-O drugs MMR Deficiency/MSI High in Gastroesophageal Cancer MMR Deficiency/MSI High in Gastroesophageal Cancer

• MMRd/MSI-H advanced disease: – Benefit of immune checkpoint inhibitors established – FDA approval for gastroesophageal ca included in tissue agnostic approval

• MMRd/MSI-H localized disease: – Does MMR deficiency/MSI predict lack of benefit of chemotherapy? Patient survival and clinical response to pembrolizumab across 12 different tumor types with mismatch repair deficiency.

86 pts, 12 tumor types: Response rate 53% Disease control rate 77%

Dung T. Le et al. Science 2017;357:409-413

Copyright © 2017 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works MMR-D/MSI-H Incidence in Gastroesophageal Cancers

Mismatch repair deficiency across 12,019 tumors *

* MMR-D/MSI-H Incidence in Gastroesophageal Cancers Higher in Low Stage Disease

• Esophageal carcinoma <2%

• Metastatic EGJ AC <2% • Metastatic Gastric 5%

• Localized EGJ AC <2% • Localized Gastric 5-19% FDA grants accelerated approval to pembrolizumab for first tissue/site agnostic indication

• On May 23, 2017, the FDA granted approval to pembrolizumab for adult and pediatric patients with unresectable or metastatic, microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR) solid tumors that have progressed following prior treatment and who have no satisfactory alternative treatment options. • This is the FDA’s first tissue/site-agnostic approval. • Based on data from 149 patients with MSI-H or dMMR cancers enrolled across five uncontrolled, multi-cohort, multi-center, single-arm clinical trials. 59 patients were diagnosed with one of 14 non-CRC cancer types, including gastroesophageal. • Efficacy outcome measures were response rate (39.6%) and response duration (>6 mo for 78% of responders). • Response rate was similar irrespective of whether patients were diagnosed with CRC (36%) or a different cancer type (46% across the 14 other cancer types). Pembro for MMR-D/MSI-H Advanced GE Adenocarcinoma

Keynote 061: 2nd line Pembrolizumab v Paclitaxel 25 of 592 pts MSI high (4.5%) MMR-D/MSI-H Localized GE Adenocarcinoma Does it predict for lack of benefit of chemotherapy? • MAGIC trial: surgery vs surgery and peri-op chemotherapy (ECF) • CLASSIC trial: surgery vs surgery and adjuvant chemo (XELOX)

Cunningham et al. MAGIC. N Engl J Med 2006; 355:11-20 Adjuvant XELOX

Surgery

Noh et al. CLASSIC.. Lancet Oncol. 2004 15:1389-96 MMR-D/MSI-H Localized GE Adenocarcinoma

MAGIC Trial: Overall Survival by MSI Status and Treatment (Surgery vs Surgery plus Peri-op Chemotherapy)

Total group (N = 303): 7.2% MSI High and/or MMR Deficient

GEJ and Esophagus cohort (N =69): 0% MSI High and/or MMR Deficient

Gastric cohort (N = 234): 9.4% MSI High and/or MMR Deficient

Smyth et al. JAMA Oncol 2017 MMR-D/MSI-H Localized GE Adenocarcinoma MAGIC and CLASSIC Trials: Overall Survival by MSI Status and Treatment (Surgery Alone vs Surgery plus Peri-op or Adjuvant Chemotherapy)

MSI-H, Surgery alone (N = 29)

MSI-H, Surgery + Chemo (N = 23)

MSI-H Surgery + Chemo,

Pietrantonio et al. JCO. 2019 MMR-D/MSI-H Localized GE Adenocarcinoma MAGIC and CLASSIC Trials: Overall Survival by MSI Status and Treatment (Surgery Alone vs Surgery plus Peri-op or Adjuvant Chemotherapy)

MSI-H, Surgery alone (N = 29) Surgery alone: Pts w/ MSI-H tumors appear to have superior survival vs pts w/ MSS tumors MSI-H, Surgery + Chemo (N = 23) Surgery and chemotherapy: Pts w/ MSI-H tumors appear to have inferior survival with addition of chemotherapy to surgery vs surgery alone.

Conclusion: Pts w/ MSI-H tumors may not benefit (or experience detrimental effect) from MSI-H Surgery + Chemo, addition of chemotherapy (peri-operative or adjuvant) MMR-D/MSI-H Localized GE Adenocarcinoma MAGIC and CLASSIC Trials: Overall Survival by MSI Status and Treatment (Surgery Alone vs Surgery plus Peri-op or Adjuvant Chemotherapy)

MSI-H, Surgery alone (N = 29) Small numbers in both studies preclude statistical analysis MSI-H, Surgery + Chemo (N = 23) Is this observation practice changing?

Pts with operable MSI-H gastric/GE adenocarcinoma should undergo multidisciplinary discussion re role of neoadjuvant or adjuvant chemotherapy MSI-H Surgery + Chemo, Future Directions

• Better biomarkers • Better drugs for biomarker targets • Biomarker-driven trial design Many Negative Studies with Targeted Agents: Not Biomarker Driven! • EGFR inhibitors (not biomarker driven) – REAL3 (chemo +/- panitumumab) – EXPAND (chemo +/- cetuximab – RTOG 0436 (chemoRT +/- cetuximab) • mTOR inhibitors (not biomarker driven) – GRANITE-1: ( BSC +- everolimus) • Met inhibitors – FOLFOX +/- onartuzumumab (Met positive by IHC) FGFR2

FPA144 – An ADCC Enhanced Monoclonal Antibody Against FGFR2b

Amplified FGFR2b In 5% of GE ACs

Presented By Jeeyun Lee at 2016 ASCO Annual Meeting Anti-FGFR2b (FPA144) in FGFR2+ Gastric/GEJ Adenocarcinoma (“FIGHT” NCT03694522) Claudin 18.2

• Transmembrane protein expressed at mucosal tight junctions • Epitopes exposed with malignant transformation • Expressed in 50% of GE AC • Zolbetuximab (IMAB362): chimeric anti-claudin mAb which mediates ADCC and CDC cell death. • Promising activity in combination with EOX in claudin-positive GE adenocarcinoma ((NCT01630083) Claudin 18.2

A Phase 3 Efficacy, Safety and Tolerability Study of Zolbetuximab (Experimental Drug) Plus mFOLFOX6 Chemotherapy Compared to Placebo Plus mFOLFOX6 as Treatment for Gastric and Gastroesophageal Junction (GEJ) Cancer (Spotlight) (NCT03504397)

Eligibility: Claudin 18.2 positive, HER2-negative Incorporating Molecular Markers into Clinical Practice: 2019 in the US • HER2: – Highly validated biomarker for clinical practice – Herceptin 1st line in HER2+ (IHC 3+ or IHC2+/FISH+) gastroesophageal AC • PD-L1 – Data conflicting; histology and line of therapy matter – Pembro 2nd line in PD-L1+ CPS ≥10% esophageal SCC – Pembro 3rd line in PD-L1+ CPS ≥1% gastroesophageal AC • MMR-D/MSI-H – Highly validated biomarker for clinical practice – Pembro 2nd line and beyond in any MMR-D/MSI-H cancer including gastroesophageal cancer Thank You