ASCO 2017 updates in Colorectal and Gastric Cancers May Cho, M.D.
ASCO 2017 updates in Colorectal and Gastric Cancers
May Cho, M.D.
Relevant financial relationships in the past twelve months by presenter or spouse/partner:
None
The speaker will directly disclosure the use of products for which are not labeled (e.g., off label use) or if the product is still investigational.
Outline
Colorectal Cancer– IDEA trial (adjuvant therapy)
– CALGB/SWOG 80405 Prognostic Factors
– SWOG S1406 (VIC vs IC in BRAF mCRC) Gastric Cancer
– FLOT
– KEYNOTE-059
IDEA: Pooled Analysis of Adjuvant Oxaliplatinfor 3 vs 6 Months in Stage III Colon Cancer
IDEA: Trials, Treatment
Prospectively pooled analysis of data from 6 concurrent randomized phase III trials in pts with stage III CC (mITT population: N ≥ 12,834)
– Pts randomized 1:1 to 3 vs 6 mos tx with oxaliplatin-based tx (investigator’s choice of FOLFOX or CAPOX)
Slide credit: clinicaloptions.comShi Q, et al. ASCO 2017. Abstract LBA1.
Trial Stage III CC Pts, N Treatment Country Median F/u,
MosPts on
CAPOX, %TOSCA 2402 CAPOX or FOLFOX4 Italy 62 35
SCOT 3983 CAPOX or mFOLFOX6 Australia, Denmark, New Zealand, Spain, Sweden, UK 37 67
IDEA France 2010 CAPOX or mFOLFOX6 France 51 10
C80702 2440 mFOLFOX6 Canada, US 35 0
HORG 708 CAPOX or FOLFOX4 Greece 48 58
ACHIEVE 1291 CAPOX or mFOLFOX6 Japan 37 75
IDEA: Statistical Plan
Primary endpoint: DFS in mITT population*– DFS: time from randomization to earliest date of relapse, secondary colorectal
primary tumor, or death
– Preplanned subgroup analyses by regimen, risk groups
Statistical analyses– DFS HR for 3 vs 6 mos (2-sided 95% CI) estimated with Cox model stratified by trial
– Predefined noninferiority margin for HR < 1.12 (12% increase in relative risk)
– Requires 3390 DFS events for 90% power with 1-sided α = 0.025
– Predefined noninferiority margin for 3-yr DFS rate difference (3 vs 6 mos): -2.7%
Additional endpoints: treatment compliance, safety
Shi Q, et al. ASCO 2017. Abstract LBA1. Slide credit: clinicaloptions.com
*Received ≥ 1 dose of study drug.
IDEA: Pt Characteristics by Tx Duration, Regimen
Slide credit: clinicaloptions.comShi Q, et al. ASCO 2017. Abstract LBA1.
CharacteristicFOLFOX CAPOX
3 Mos(n = 3870)
6 Mos(n = 3893)
3 Mos(n = 2554)
6 Mos(n = 2517)
Median age, yrs 64 64 65 65ECOG PS 0/1,* % 77/22 77/22 82/18 81/19T stage, % T1-2 T3 T4
136819
146719
136324
126325
N stage, % N1 N2
7228
7327
7129
7129
Reached final planned cycle, % 90 71 86 65*1% of FOLFOX-treated pts had ECOG PS 2.
Shi Q, et al. ASCO 2017. Abstract LBA1. Reproduced with permission.
IDEA: DFS in mITT Population, Risk Subgroups
Slide credit: clinicaloptions.com
Noninferiority of oxaliplatin-based tx for 3 vs 6 mos not proven
– DFS HR: 1.07 (95% CI: 1.00-1.15)
– Difference in 3-yr DFS rates: -0.9% (95% CI: -2.4% to 0.6%)
3-yr DFS rate difference of 20% between low risk (T1-3, N1) vs high risk (T4 or N2) subgroups
T1-3, N1T4 or N2
HR
Interaction P Value
Favors 3 mos Favors 6 mosRisk Group
1.011.12
0.5 1 1.12 1.5
.11
HR
Shi Q, et al. ASCO 2017. Abstract LBA1. Reproduced with permission.
IDEA: DFS by Risk Subgroups and Regimen
Slide credit: clinicaloptions.com
Favors 3 mos Favors 6 mos HR (95% CI)
Noninferiority margin
Risk Group
Regimen
T1-3, N1
T4 or N2
FOLFOXCAPOX
FOLFOXCAPOX
Not provenNoninferior
InferiorNot proven
1.10 (0.96-1.26)0.85 (0.71-1.01)
1.20 (1.07-1.35)1.02 (0.89-1.17)
HR 1.0 1.12
IDEA: Safety
Slide credit: clinicaloptions.comShi Q, et al. ASCO 2017. Abstract LBA1.
AE, %FOLFOX CAPOX
3 Mos 6 Mos P Value* 3 Mos 6 Mos P Value*Any event† Grade 2 Grade 3/4
3238
3257
< .0001 4124
4837
< .0001
Neurotoxicity Grade 2 Grade ¾
143
3216
< .0001 123
369
< .0001
Diarrhea Grade 2 Grade 3/4
115
137
< .0001 107
139
.0117
*For Chi-squared test for trend.†19 grade 5 events reported.
IDEA: Investigator Conclusions DFS noninferiority of adjuvant oxaliplatin-based tx for 3 vs 6 mos not established in overall
pts with stage III CC
Shorter tx associated with greater treatment compliance, much less grade ≥ 2 neuropathy
Investigators recommend risk-vs-benefit approach when selecting duration of adjuvant oxaliplatin-based tx
– Low-risk pts (T1-3, N1): 3 mos
– High-risk pts (T4, N2, or other high-risk variables): 3-6 mos depending on tolerability, ptpreferences, recurrence risk, regimen (FOLFOX vs CAPOX)
Slide credit: clinicaloptions.comShi Q, et al. ASCO 2017. Abstract LBA1.
Prognostic Factors for First-line Chemotherapy + Bevacizumab or Cetuximab in Metastatic
Colorectal Cancer
CALGB/SWOG 80405: Study Design and Survival Randomized, open-label phase III trial
No difference between cetuximab and bevacizumab when combined with FOLFIRI or FOLFOX with respect to OS or PFS[1]
Pts with KRAS wild-type (Codons 12, 13) metastatic/advanced CRC and no previous therapy for advanced
disease (N = 1137)
Cetuximab + Chemotherapy*(n = 578)
Bevacizumab + Chemotherapy*(n = 559)
Slide credit: clinicaloptions.com
OS, Mos PFS, Mos
29.9 10.4
29.0 10.8
*Physician choice of FOLFIRI or FOLFOX.
1. Venook A, et al. JAMA. 2017 (in press). 2. Venook A, et al. ASCO 2017. Abstract 3503. 3. Innocenti F, et al. ASCO 2017. Abstract 3504.
Pts with right-sided tumor, n = 293 (27%); left-sided, n = 732 (68%)
– Pts with transverse colon tumors excluded from analysis (n = 66)
CALGB/SWOG 80405 Prognostic Factors: OS by Side of Tumor and Biologic Treatment
Venook A, et al. ASCO 2016. Abstract 2016. Reproduced with permission. Slide credit: clinicaloptions.com
OS by Sidedness OS by Side and Treatment
OS
(%)
Mos MosO
S (%
)
0
20
40
60
100
80
0 12 24 36 48 60 8472 10896
Side
Left
Right
mOS(95% CI)
33.3(31.4-35.7)
19.4(16.7-23.6)
N (Events)
732 (550)
293 (242)
HR(95% CI)
1.55
(1.32-1.82)
P Value
< .0001
Left
Right
0
20
40
60
100
80
0 12 24 36 48 60 8472 10896
Left/BevacizumabmOS 31.4 (95% CI: 28.3-33.6)Left/CetuximabmOS 36.0 (95% CI: 32.6-40.3)Right/BevacizumabmOS 24.2 (95% CI: 17.9-30.3)Right/CetuximabmOS 16.7 (95% CI: 13.1-19.4)
CALGB/SWOG 80405 Prognostic Factors: Current Analyses Multivariate analysis of prognostic
utility of tumor sidedness (independent of other molecular features)[1]
– Evaluated in subset of pts with left-/ right-sided tumors (no transverse) and available molecular data (n = 728)
– Used Cox proportional hazard models*
– Evaluated potential biomarkers of sidedness/tumor burden
Analysis of mutational profile and prognostic, clinical value of DNA alterations[2]
– DNA from 504 tumor specimens
– Mutation profile determined by PCR genotyping in 12 genes, MSI-high status by microsatellite mutation analysis, mutational load in 395 genes by NGS
– Primary endpoint: OS
– Used Cox proportional hazard models* to test association with molecular alterations†
Slide credit: clinicaloptions.com1. Venook AP, et al. ASCO 2017. Abstract 3503. 2. Innocenti F, et al. ASCO 2017. Abstract 3504.
*Models stratified by adjuvant CT, prior RT. Models adjusted for age, race, sex, synchronous vs metachronous, BRAF. Sidedness analysis also adjusted for consensus molecular subtypes, MSI, NRAS, KRAS, HRAS; mutational analysis for liver metastases only, sidedness. †No adjustment for multiple comparisons.
CALGB/SWOG 80405 Prognostic Factors: Sidedness Independent of Molecular Features Multivariate analysis found that sidedness
a significant prognostic factor independent of other molecular features
– HR: 1.392 (95% CI: 1.032-1.878; P = .031*)
Clinical characteristics compared between pts with right- vs left-sided tumors to determine whether sidedness potentially a surrogate for tumor burden
– Pattern of metastases significantly differed by side (Chi-squared P = .0136)
– No significant differences with other tumor burden indicators
Slide credit: clinicaloptions.comVenook AP, et al. ASCO 2017. Abstract 3503.
Tumor Burden Indicator, %
Right-Sided Tumors(n = 167)
Left-Sided Tumors(n = 330)
Median LDH, IU/L 195.5 196.5Metastatic sites, n 1 2 ≥ 3
53.933.911.5
55.930.113.1
Prior adjuvant tx 12.0 18.81⁰ in place at tx initiation 4.8 1.8Intent, palliative/curative 86.4/13.6 83.1/16.9Pattern of metastases Liver only Liver + extrahepatic Extrahepatic only
30.362.437.0
38.373.325.8
*Per stratified, adjusted Cox proportional hazard model.
CALGB/SWOG 80405 Prognostic Factors: OS by BRAF Genotype
Patient Genotype Group Median OS, Mos (95% CI) HRadj (95% CI) P for Interaction,
BRAF/biologics
BRAF genotype Wild type (n = 432) Mutant (n = 72)
34.2 (31.0-36.4)12.9 (11.1-19.0)
1.82 (1.37-2.44) P = .0001*
BRAF wild type Treated with cetuximab (n = 225) Treated with bevacizumab (n = 207)
33.4 (29.1-39.2)34.4 (30.3-37.6)
0.97 (0.77-1.23) P = .75
.13BRAF mutant Treated with cetuximab (n = 31) Treated with bevacizumab (n = 41)
11.7 (8.6-19.7)15.0 (11.8-23.7)
0.61 (0.35-1.06) P = .51
Innocenti F, et al. ASCO 2017. Abstract 3504.
*If further adjusted for sidedness of tumor, HR: 1.67 (95% CI: 1.20-2.33; P = .0035).
Slide credit: clinicaloptions.com
CALGB/SWOG 80405 Prognostic Factors: OS by MSI Status
Patient Subgroup Median OS, Mos (95% CI) HRadj (95% CI)P for
Interaction, MSI/biologics
Microsatellite instability High level of instability (n = 29; 52%
BRAF mutant) Stable (n = 389; 11% BRAF mutant)
30.3 (22.6-NE)31.8 (29.0-35.2)
0.84 (0.51-1.39) P = .50
Microsatellite stable Treated with cetuximab (n = 224) Treated with bevacizumab (n = 220)
33.4 (29.1-39.3)32.8 (29.0-36.0)
1.03 (0.82-1.30) P = .48
.0002Microsatellite instability-high Treated with cetuximab (n = 13, 46%
BRAF mutant) Treated with bevacizumab (n = 18,
61% BRAF mutant)
11.5 (10.3-NE)30.3 (23.6-NE)
0.17 (0.07-0.41)P = .002
Innocenti F, et al. ASCO 2017. Abstract 3504. Slide credit: clinicaloptions.com
CALGB/SWOG 80405 Prognostic Factors: OS by Mutational Load
Patient Subgroup Median OS, Mos (95% CI) HRadj (95% CI)
P for Interaction, Mutational
Load/BiologicsMutational Load in MSS mCRC ≤ 8 (n = 228) > 8 (n = 65)
30.1 (25.6-34.3)35.7 (31.2-45.5)
0.67 (0.47-0.95) P = .02
Mutational Load ≤ 8 in MSS mCRC Treated with cetuximab (n = 103) Treated with bevacizumab (n = 125)
29.1 (24.3-38.5)30.3 (25.3-34.5)
1.00 (0.72-1.39) P = .97
.56Mutational Load > 8 in MSS mCRC Treated with cetuximab (n = 27) Treated with bevacizumab (n = 38)
35.8 (30.3-63.7)35.1 (31.2-53.6)
0.81 (0.43-1.52) P = .79
Innocenti F, et al. ASCO 2017. Abstract 3504. Slide credit: clinicaloptions.com
CALGB/SWOG 80405 Prognostic Factors: Investigator Conclusions In pts with mCRC and no previous therapy for advanced disease:
– Left-sided tumors significantly associated with improved OS, regardless of biologic tx[1]
– Sidedness a prognostic factor independent of other molecular features (HR: 1.392; P = .031)
– Investigators concluded that tumor sidedness differences not due to tumor burden; further study warranted to identify mechanisms behind sidedness as independent prognostic factor
– BRAF status strong prognostic marker, even when adjusted for tumor sidedness, but has no predictive interaction with biologic therapy[2]
– MSI status not prognostic but may be predictive for biologic therapy[2]
– Mutational load in pts with MSS CRC has prognostic potential but does not seem to be predictive for biologic therapy[2]
According to investigators, these data indicate that pt and tumor characteristics may help predict response to biologic therapy but larger numbers of pts and more markers need to be explored[1,2]
1. Venook AP, et al. ASCO 2017. Abstract 3503. 2. Innocenti F, et al. ASCO 2017. Abstract 3504. Slide credit: clinicaloptions.com
SWOG S1406: Vemurafenib in Combination With Irinotecan and Cetuximab in BRAF V600E
Metastatic CRC
SWOG S1406: Background
BRAF V600E: mutation leading to constitutive activation of BRAF kinase and MAPK pathway[1]
– BRAF mutated in ~ 8% of mCRC cases
– Associated with aggressive disease, poor prognosis, minimal benefit from standard chemotherapy[2]
Vemurafenib: kinase inhibitor selective for mutated BRAF protein[3]
– Vemurafenib monotherapy associated with limited activity in mCRC
– Phase I study suggested improved survival and response in combination with irinotecan + cetuximab in refractory mCRC with BRAF V600E[4]
Current study evaluated efficacy and safety of vemurafenib addition to irinotecan + cetuximab in pts with BRAF V600E mCRC[5]
Slide credit: clinicaloptions.comReferences in slidenotes.
SWOG S1406: Study Design
Randomized, multicenter, open-label phase II trial
Slide credit: clinicaloptions.comKopetz S, et al. ASCO 2017. Abstract 3505. ClinicalTrials.gov. NCT02164916.
Pts with mCRC, extended RAS WT and BRAF V600E, previously treated with 1-2
systemic CT lines for metastatic or advanced unresectable disease, no prior
panitumumab or cetuximab, no prior BRAF or MEK inhibitors, ECOG PS 0-1
(N = 106)*
Until PD
Until PD; crossover allowed to
vemurafenib arm after PD
Cetuximab 500 mg/m2 IV Q2W + Irinotecan 180 mg/m2 IV Q2W
(n = 50)
Vemurafenib 960 mg PO BID + Cetuximab 500 mg/m2 IV Q2W +
Irinotecan 180 mg/m2 IV Q2W(n = 49)
Primary endpoint: PFS
Secondary endpoints: OS, ORR, treatment-related AEs
Stratified by prior irinotecan treatment (yes vs no)Randomized 1:1
*Only 99 pts eligible to receive treatment after randomization.
SWOG S1406: Baseline Characteristics
Slide credit: clinicaloptions.comKopetz S, et al. ASCO 2017. Abstract 3505.
Characteristic, n (%) VIC(n = 49)
IC(n = 50)
Median age, yrs (range) 60 (34-83) 62 (31-83)Female 21 (43) 37 (74)Race White Black Asian
43 (88)1 (2)4 (8)
49 (98)0 (0)1 (2)
Hispanic ethnicity 2 (4) 2 (4)ECOG PS 1 25 (51) 27 (54)Prior irinotecan 20 (41) 19 (38)Prior regimens for mCRC 1 2 Failed adjuvant within 6 mos
27 (55)19 (39)3 (6)
26 (52)17 (34)7 (14)
SWOG S1406: PFS (Primary Endpoint)
Slide credit: clinicaloptions.com
PFS
(%)
HR: 0.48 (95% CI: 0.31-0.75; P = .001)
Events, n mPFS (95% CI)VIC 40 4.3 (3.6-5.7)
IC 48 2.0 (1.8-2.1)
Kopetz S, et al. ASCO 2017. Abstract 3505. Reproduced with permission.
Median follow-up: 7.3 mos
Among 48% of pts who crossed over to VIC after PD on IC, mPFS was 5.8 mos
PFS significantly prolonged with vemurafenib in most subgroups
– Significant benefit in pts with tumors on right vs left/rectum, no prior irinotecan, MSS, or mutated PIK3CA
0
20
40
60
80
100
0 3 6 8 10 12 14Mos
SWOG S1406: OS, Response In crossover pts, mOS: 12.1 mos (95%
CI: 4.5-12.5)
Distribution of responses significantly different with addition of vemurafenib vs IC alone (Chi-squared P = .001)
Slide credit: clinicaloptions.com
Endpoint, % VIC(n = 44)*
IC(n = 47)*
Crossover(n = 24)†
PR‡ 16 4 17SD 50 17 55PD§ 18 66 NRDCR 67 22 72
*Measurable disease in 93 pts overall. †Excludes 6 pts (2 pts without PD before crossover, 4 pts without measurable disease). ‡Includes confirmed and unconfirmed. §Includes symptomatic deterioration.
mOS, Mos 95% CIVIC 9.6 (7.5-13.1)IC 5.9 (3.0-9.9)
HR: 0.73 (95% CI: 0.45-1.17; P = .19)
OS
(%)
0
20
40
60
80
100
0 3 6 9 12 15 18Mos
Kopetz S, et al. ASCO 2017. Abstract 3505. Reproduced with permission.
SWOG S1406: Safety
AE-related discontinuations more common with vemurafenib combination (16%) vs cetuximab + irinotecan alone (6%)
Median duration of treatment imbalanced between arms
– 88 days for VIC
– 47 days for IC
Slide credit: clinicaloptions.comKopetz S, et al. ASCO 2017. Abstract 3505.
Grade 3/4 AE, n (%) VIC(n = 46)
IC(n = 46)
Anemia 6 (13) 0 (0)Dehydration 5 (11) 3 (7)Diarrhea 11 (24) 6 (13)Febrile neutropenia 5 (11) 2 (4)Fatigue 7 (15) 7 (15)Neutropenia 15 (33) 3 (7)Rash 2 (4) 3 (7)Hypomagnesemia 0 (0) 2 (4)Nausea 9 (20) 1 (2)Arthralgia 3 (7) 0 (0)
SWOG S1406: Investigator Conclusions
Addition of vemurafenib to cetuximab + irinotecan associated with significantly prolonged PFS in mCRC pts with BRAF V600E
– mPFS: 4.3 vs 2.0 mos, respectively (HR: 0.48; P = .001)
– PFS benefit observed across most subgroups
Addition of vemurafenib associated with benefit in pts crossing over after PD on IC alone
– mPFS: 5.8 mos; mOS: 12.1 mos
mOS numerically higher with VIC (9.6 mos) vs IC alone (5.9 mos), but did not reach statistical significance (HR: 0.73; P = .19)
– Analysis limited by 48% crossover to VIC after PD on IC alone
Investigators concluded that VIC is a new treatment for BRAF V600E mCRCSlide credit: clinicaloptions.comKopetz S, et al. ASCO 2017. Abstract 3505.
Perioperative chemotherapy with docetaxel, oxaliplatin, and fluorouracil/leucovorin(FLOT)
versus epirubicin, cisplatin, and fluorouracil or capecitabine (ECF/ECX) for resectable gastric or
gastroesophageal junction (GEJ) adenocarcinoma (FLOT4): A multi center,
randomized phase III trial
FLOT4 Study Design
Presented By Salah-Eddin Al-Batran at 2017 ASCO Annual Meeting
Endpoints and Populations
Presented By Salah-Eddin Al-Batran at 2017 ASCO Annual Meeting
Baseline 1
Presented By Salah-Eddin Al-Batran at 2017 ASCO Annual Meeting
Surgery 1
Presented By Salah-Eddin Al-Batran at 2017 ASCO Annual Meeting
Histopathology (ypTN)
Presented By Salah-Eddin Al-Batran at 2017 ASCO Annual Meeting
FLOT4: Progression-Free Survival
Presented By Salah-Eddin Al-Batran at 2017 ASCO Annual Meeting
FLOT4: Overall Survival
Presented By Salah-Eddin Al-Batran at 2017 ASCO Annual Meeting
Chemo Related Toxicity 1
Presented By Salah-Eddin Al-Batran at 2017 ASCO Annual Meeting
<>>><<<<
DiscussionStrengths Well designed and executed
prospective randomized phase III study
Clear benefit to FLOT over ECF – Increased rates of curative surgery, PFS
and OS
Limitations Overlap with patients treated in
CROSS trial – similar mOS (48.6m)
Does not answer the question – do patients need adjuvant therapy?
Practice Changing?• Yes, for perioperative chemotherapy approach
KEYNOTE-059 (Cohort 1): PembrolizumabMonotherapy in Previously Treated Advanced
Gastric or GEJ Adenocarcinoma
Pembrolizumab for Pretreated Advanced Gastric or GEJ Adenocarcinoma: Background PD-L1 overexpression a feature of gastric cancers[1-3]
Pembrolizumab: humanized anti–PD-1 IgG4-κ mAb; blocks PD-1 interaction with its ligands PD-L1/2[4]
– Phase Ib KEYNOTE-012: manageable toxicity, 22% ORR in pts with advanced PD-L1+ gastric cancer[5]
Current analysis reports data from phase II KEYNOTE-059 of pembrolizumab with focus on cohort 1 (pts with advanced gastric/GEJ cancer and ≥ 2 prior lines of therapy who received pembrolizumab monotherapy)[6]
1. Kim JW, et al. Gastric Cancer. 2016;19:42-52. 2. Qing Y, et al. Drug Des Devel Ther. 2015;9:901-909. 3. Dong M, et al. Hum Pathol. 2016;53:25-34. 4. Pembrolizumab [package insert]. 5. Muro K, et al. Lancet Oncol. 2016;17:717-726. 6. Fuchs CS, et al. ASCO 2017. Abstract 4003. Slide credit: clinicaloptions.com
KEYNOTE-059: Study Design
Open-label, multicohort phase II study
Primary endpoints: ORR, safety; secondary endpoints: DoR, PFS, OS
Exploratory biomarker endpoints: efficacy by MSI, GEP
Cohort 1≥ 2 prior
lines of CTPts with recurrent or metastatic gastric or
GEJ adenocarcinoma; ECOG PS 0/1;
HER2/neu negative*; no prior PD-1/PD-L1 tx, systemic steroids, autoimmune disease, ascites, or CNS mets
(N = 259)
Pembrolizumab200 mg Q3W
Pembrolizumab 200 mg Q3W +Cisplatin 80 mg/m2 Q3W +
5-FU 800 mg/m2 Q3W orCapecitabine 1000 mg/m2 BID Q3W
Tx continued for 24 mos or until PD, intolerable toxicity,
or withdrawal of consent; survival
follow-up until study end, death, or
withdrawal
Slide credit: clinicaloptions.com
Pembrolizumab200 mg Q3W
Cohort 2No prior tx
Cohort 3No prior tx,
PD-L1+
Fuchs CS, et al. ASCO 2017. Abstract 4003.
*HER2/neu positive allowed in cohort 1 if prior trastuzumab administered.
KEYNOTE-059 (Cohort 1): Baseline Characteristics
Slide credit: clinicaloptions.comFuchs CS, et al. ASCO 2017. Abstract 4003.
Characteristic All Pts (N = 259)
Median age, yrs (range) 62 (24-89)Male, n (%) 198 (76.4)Geographic region, n (%)United States East AsiaOther
124 (47.9)34 (13.1)
101 (39.0)ECOG PS, n (%) 0 1
107 (41.3)151 (58.3)
Primary tumor location, n (%)GastricGEJ
125 (48.3)133 (51.4)
Characteristic, n (%) All Pts (N = 259)
Prior therapies 2 3 ≥ 4
134 (51.7)75 (29.0)50 (19.3)
Prior surgery for gastric cancer 66 (25.5)
HER2 positive 63 (24.3)PD-L1 expression Positive*Negative
148 (57.1)109 (42.1)
*CPS ≥ 1% where CPS is (PD-L1 staining cells/total tumor cells) x 100.
KEYNOTE-059 (Cohort 1): Response
Median follow-up: 5.8 mos (range: 0.5-21.6 mos)
Slide credit: clinicaloptions.comFuchs CS, et al. ASCO 2017. Abstract 4003.
*CR + PR + SD ≥ 2 mos.
Confirmed Response, % (95% CI)
All Pts (N = 259)
ORR 11.6 (8.0-16.1)CR 2.3 (0.9-5.0)PR 9.3 (6.0-13.5)SD 16.2 (11.9-21.3)PD 56.0 (49.7-62.1)DCR* 27.0 (21.7-32.9)
KEYNOTE-059 (Cohort 1): Safety
Slide credit: clinicaloptions.comFuchs CS, et al. ASCO 2017. Abstract 4003.
D/c for TRAEs: abnormal hepatic function, bile duct stenosis, n = 1 each.Grade 5 TRAEs: acute kidney injury, pleural effusion, n = 1 each.
irAE Occurring in > 1% of Pts, %
All Pts (N = 259)Any Grade Grade 3/4
Any 17.8 4.6Hypothyroidism 8.9 0.4Hyperthyroidism 3.5 0Colitis 2.3 1.2Pneumonitis 1.9 0.8Thyroiditis 1.5 0.4Infusion reaction 1.5 0Severe skin reaction* 1.5 1.5
TRAE Occurring in > 5% of Pts, %
All Pts (N = 259)Any Grade Grade 3/4
Fatigue 18.9 2.3Pruritus 8.9 0Rash 8.5 0.8Hypothyroidism 7.7 0.4Decreased appetite 7.3 0Anemia 6.9 2.7Nausea 6.9 0.8Diarrhea 6.6 1.2Arthralgia 5.8 0.4 *Includes erythema multiforme, jaundice, rash, maculopapular rash.
Systemic corticosteroids for irAEs: n = 13. Treatment interruption due to irAEs: n = 10.
KEYNOTE-059 (Cohort 1): Response by PD-L1 Expression and Line of Therapy
Slide credit: clinicaloptions.comFuchs CS, et al. ASCO 2017. Abstract 4003.
*CR + PR + SD ≥ 2 mos.
Confirmed Response, % (95% CI)
PD-L1 Line of Therapy PD-L1 and Third Line of Therapy
Positive(n = 148)
Negative(n = 109)
Third (n = 134)
≥ Fourth(n = 125)
Positive(n = 75)
Negative(n = 58)
ORR 15.5 (10.1-22.4)
6.4 (2.6-12.8)
16.4 (10.6-23.8)
6.4 (2.8-12.2)
22.7 (13.8-33.8)
8.6 (2.9-19.0)
CR 2.0 (0.4-5.8)
2.8 (0.6-7.8)
3.0 (0.8-7.5)
1.6 (0.2-5.7)
2.7 (0.3-9.3)
3.4 (0.4-11.9)
PR 13.5 (8.5-20.1)
3.7 (1.0-9.1)
13.4 (8.2-20.4)
4.8 (1.8-10.2)
20.0 (11.6-30.8)
5.2 (1.1-14.4)
DCR* 33.1 (25.6-41.3)
19.3 (12.3-27.9)
31.3 (23.6-39.9)
22.4 (15.4-30.7)
38.7 (27.6-50.6)
22.4 (12.5-35.3)
KEYNOTE-059 (Cohort 1): Maximum Change From Baseline in Target Lesion Size
Slide credit: clinicaloptions.comFuchs CS, et al. ASCO 2017. Abstract 4003. Reproduced with permission.
*Included pts with measurable disease at BL and ≥ 1 post-BL assessment (n = 223).
Pts With Reduction, %All pts* 42.4PD-L1 positive 47.3PD-L1 negative 36.3
120
Cha
nge
From
BL
(%)
10080
6040
200
-20-40
-60-80
-100
PD-L1 positivePD-L1 negativePD-L1 expression unknown
12
KEYNOTE-059 (Cohort 1): Depth and Duration of Response
Slide credit: clinicaloptions.comFuchs CS, et al. ASCO 2017. Abstract 4003. Reproduced with permission.*Included pts with measurable disease at BL and ≥ 1 post-BL assessment (n = 30). †No PD at last disease assessment.
Outcome All Pts* PD-L1+ PD-L1-Median DoR, mos (95% CI) 8.4 (1.6+† to 17.3+) 16.3 (1.6+ to 17.3+) 6.9 (2.4 to 7.0+)
Treatment Exposure and Duration of Response Longitudinal Change From BL in Tumor Size Among Responders (n = 30)
240 2 4 6 8 10 1614 18 20 22Mos Since First Dose
Con
firm
ed R
espo
nder
s (n
= 3
0)
CRPRPDDeathOngoing pembrolizumab treatment
Treatment discontinuedTreatment ongoing
Cha
nge
From
BL
(%)
20
0
-20
-40
-60
-80
-10024
Mos Since Treatment Initiation0 2 4 6 8 10 12 14 16 18 20 22
KEYNOTE-059 (Cohort 1): Survival
Slide credit: clinicaloptions.comFuchs CS, et al. ASCO 2017. Abstract 4003. Reproduced with permission.
All Pts (N = 259)
Median OS, mos (95% CI) 5.6 (4.3-6.9)12-mo OS rate, % 23.4
All Pts (N = 259)
Median PFS, mos (95% CI) 2.0 (2.0-2.1)100
80
60
40
20
0
OS
(%)
Mos220 2 4 6 8 10 12 14 16 18 20
Pts at risk, n259 199 144 112 51 27 22 12 7 287 0
100
80
60
40
20
0
PFS
(%)
Mos220 2 4 6 8 10 12 14 16 18 20
Pts at risk, n259 136 51 34 17 4 2 2 2 022 0
OS PFS
KEYNOTE-059 (Cohort 1): Exploratory Analyses
MSI assessed in 174 pts
– MSI-high: 4.0%
18-gene T-cell–inflamed GEP score, assessed in 144 pts, associated with significantly improved response to pembrolizumab (P = .014)
Slide credit: clinicaloptions.com
Confirmed Response, % (95% CI)
MSI-High(n = 7)
Non–MSI-High(n = 167)
ORR 57.1 (18.4-90.1)
9.0 (5.1-14.4)
CR 14.3 (0.4-57.9) 2.4 (0.7-6.0)PR 42.9 (9.9-81.6) 6.6 (3.3-11.5)DCR* 71.4 (29.0-96.3) 22.2 (16.1-29.2)*CR + PR + SD ≥ 2 mos.
0.0
0.5
-0.5
18-G
ene
T-C
ell−
Infla
med
G
EP S
core
Nonresponder Responder
Fuchs CS, et al. ASCO 2017. Abstract 4003. Reproduced with permission.
KEYNOTE-059 (Cohort 1): Conclusions
In pts with advanced gastric/GEJ cancer and ≥ 2 prior therapies, pembrolizumab well tolerated with promising antitumor activity, durable responses– ORR: 11.6%; higher in PD-L1+ vs PD-L1- tumors (15.5% vs 6.4%)
and MSI-high vs non–MSI-high tumors (57.1% vs 9.0%)
Study investigators suggest pembrolizumab as potential therapeutic option for this pt population
Pembrolizumab in earlier-line therapy and in chemotherapy combinations under investigation for advanced gastric/GEJ cancer in ongoing randomized trials
Slide credit: clinicaloptions.comFuchs CS, et al. ASCO 2017. Abstract 4003.
Thank You