Abstracts 4005-8 Jordan Berlin, MD Ingram Professor of Clinical Research Co-Director, GI Oncology Program Vanderbilt-Ingram Cancer Center
Jan 14, 2016
Abstracts 4005-8
Jordan Berlin, MD
Ingram Professor of Clinical Research
Co-Director, GI Oncology Program
Vanderbilt-Ingram Cancer Center
Disclosures
• With regards to these presentations:– I have done an advisory board and have had
multiple trials with Amgen,– Vanderbilt has trials open from Daiichi and
Arqule
The obvious groupings
•Abstracts 4005-7
–These abstracts address the use of Cmet inhibition as a therapeutic target in GI cancers
•I will be reversing the order of these abstracts
•Abstract 4008
–This abstract addresses the oft ignored issue of trying to reduce the discomforts caused by one of our therapeutic options
First, and Foremost
• All 4 abstracts represent excellent work from dedicated individuals who want to make a difference in the lives of our patients– All the authors should be congratulated for their
efforts.– This does not mean I won’t be critical
cMet
• You have seen the pathway• Key factor: Expression has been a prognostic marker• Frequently dysregulated in a number of tumor types
including HCC and gastric cancers– Gene amplification in the primary has been reported in a limited
number of diseases including colon and gastric primaries
• Overexpression has been found to be a potential resistance mechanism to HER pathway inhibition, VEGF inhibition and possibly IGF1R inhibition
• It does bad things: proliferation, survival, etc, etc– We have all seen most of this rationale for almost any targeted
agent
Activity of Cabozantinib (XL184) in Hepatocellular Carcinoma:
Results From a Phase 2 Randomized Discontinuation Trial (RDT)
Chris Verslype1, Allen Lee Cohn2, Robin Katie Kelley3, Tsai-Shen Yang4, Wu-Chou Su5, David A. Ramies6, Yihua Lee6, Xiaodong Shen6,
Eric Van Cutsem1
University Hospitals Gasthuisberg, Leuven, Belgium1; Rocky Mountain Cancer Center, LLP, Denver, CO2; Helen Diller Family Comprehensive Cancer Center, University of California, San
Francisco, San Francisco, CA3; Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan, Taiwan4; National Cheng Kung University Hospital, Tainan, Taiwan5;
Exelixis, South San Francisco, CA6
Tivantinib (ARQ 197) vs Placebo in Patients (Pts) with Hepatocellular Carcinoma (HCC) Who Failed One
Systemic Therapy:Results of a Randomized Controlled Phase 2 Trial (RCT)
L Rimassa, C Porta, I Borbath, B Daniele, S Salvagni, JL Van Laethem, H Van Vlierberghe,J Trojan, F Kolligs, A Weiss, N Barahona, A Gasbarrini, M Lencioni, A Pande, M Lamar, Y Chen, G Abbadessa, B Schwartz, A Santoro
Similarities• Both are randomized phase II studies of -nibs that
are primarily cMet inhibitors studied in patients with HCC
• Both study groups clearly understood the difficulties of this disease and incorporated it into study design– Frequent stable disease, heterogeneous etiologies, etc
• Both studies had designs that encouraged enrollment
• Both groups felt further study is warranted based on their results
Significant differences• Cabozantinib
– Has inhibitory effects on two key pathways: cMet and VEGFR2
– Had a novel trial design (randomized discontinuation)
• Recognized the frequent stability of this disease in Child’s A patients and low platelets despite good PS
– Focused on PFS as well as tumor shrinkage and AFP reduction
– Had a more heterogeneous patient population• Number and type of prior therapies
Significant Differences
• Tivantinib– Selective for cMet pathway inhibition– Had a predefined biomarker analysis
incorporated into the design– Clearly defined prior therapies
• (1 systemic therapy taken for at least 21 days)
– Primary endpoint was TTP, but looked at OS and PFS
– Had two different doses of drug due to toxicities in the first 57 patients randomized
Best Target Lesion Regression of Patients with ≥1 Post-Baseline Tumor Assessment (N=36)*
*per Original RECIST 1.0
‡ ‡
Prior sorafenib
No prior sorafenib
Confirmed PRs‡
Best AFP Time Point Response of Patients with ≥1 Post-Baseline Measurement
(N=26)*
+ Increase >100% from Baseline
* Only Patients with AFP ≥ 20ng/mL at baseline are shown.
Prior sorafenib
No prior sorafenib
+ + +
Cabozantinib Results
• There is reduction in AFP, change in tumor measurements
• PFS over 4 months suggests some effect and while 15 month survival is encouraging, this is a highly selected patient group in a small study
• Does it warrant further study?– Further phase II planned is appropriate.– Biomarkers should be considered based on the
other two abstracts I am discussing
Tivantinib Results
• First, cMet expression was prognostic– OS 9 months for cMet low and 3.8 months for
cMet high (based on 28 patients)
• While tivantinib had no apparent effect in cMet low patients,
• cMet high patients appeared to get benefit
Primary Endpoint: TTP (ITT Population)
• Study powered to detect a treatment difference with a 1-sided type I error α = 0.05• PFS consistent with TTP: HR 0.67 (95% CI: 0.44-1.04) Log Rank: P=0.06• 1 PR was observed in the 240mg BID group. Disease control rate: 44% on tivantinib (32-56) vs 31% (16-48) • Of 23 crossed-over patients, 11 showed best response of SD (3 ongoing at time of data cut-off), 8 PD, 4 non evaluable
ARQ 197-215 PRESENTED BY: LORENZA RIMASSA
HR: 0.64 (90% CI: 0.43-0.94) Log Rank: P=0.04
Median TTP Patients Events
Tivantinib 6.9 wks 71 46
Placebo 6.0 wks 36 30
Improved TTP in MET Diagnostic High Group
PFS: HR 0.45 (95% CI: 0.21-0.95) Log Rank: P=0.02DCR: 50% on tivantinib (28-72) vs 20% (4-48)
ARQ 197-215 PRESENTED BY: LORENZA RIMASSA
HR: 0.43 (95% CI: 0.19-0.97) Log Rank: P=0.03
Median TTP Patients Events
Tivantinib 11.7 wks 22 14
Placebo 6.1 wks 15 13
Improved OS in MET Diagnostic High Group
ARQ 197-215 PRESENTED BY: LORENZA RIMASSA
HR: 0.38 (95% CI: 0.18-0.81) Log Rank: P=0.01
Median OS Patients Events
Tivantinib 7.2 mos 22 17
Placebo* 3.8 mos 15 15
*8 MET Dx High patients crossed-over, 5 remained on open-label tivantinib for at least 6 weeks (1 non-evaluable at cut-off date)
cMet in HCC: Conclusions• Both studies demonstrated some benefit• Both studies warrant further analysis• A randomized phase II study for crizontanib is
appropriate• A randomized phase III study for tivantinib seems
appropriate, and I laud the conviction to follow the biomarker hypothesis, but– A large, randomized trial is now being built on a subset
analysis of 37 patients out of a 107 patient study
– This scenario has been repeated multiple times in pancreatic cancer and look where it has gotten us
– Do the randomized phase II in cMet high patients first
INTRIGUINGLY, SUPPORT FOR A BIOMARKER APPROACH COMES FROM ANOTHER DISEASE SITE
Evaluation of MET Pathway Biomarkers in a Phase 2 Study of Rilotumumab or Placebo in
Combination With Epirubicin/Cisplatin/Capecitabine in Patients With
Locally Advanced or Metastatic Gastric or Esophagogastric Junction Cancer
Kelly S. Oliner,1 Rui Tang,1 Abraham Anderson,1 Yun Lan,1 Timothy Iveson,2 Ross C. Donehower,3 Yizhou Jiang,1 Sarita Dubey,4 Elwyn Loh4
1Amgen Inc., Thousand Oaks, CA, USA; 2Southampton General Hospital, Southampton, UK; 3Johns Hopkins Cancer Center, Baltimore, MD, USA; 4Amgen Inc., South San Francisco, CA, USA
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Pro
gre
ssio
n-F
ree
Su
rviv
al (
%)
Time (months)
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17Time (months)
Ove
rall
Su
rviv
al (
%)
Clinical Efficacy in the Intent-to-Treat Population*
*Results based on the updated analysis with data cutoff of April 1, 2011. †Adjusted for baseline randomization stratification variables (ECOG status [0 or 1] and disease extent [locally advanced or metastatic]).Iveson T, et al. European Multidisciplinary Cancer Congress, September 23-27, 2011,Stockholm, Sweden; abstract #6504.
Median Months(80% CI)
HR†
(80% CI) P Value
Rilotumumab + ECX (n = 82) 5.6 (4.9–6.9) 0.64 (0.48–0.85) 0.045
Placebo + ECX (n = 39) 4.2 (3.7–4.6)
Progression-Free Survival
Overall Survival Median Months(80% CI)
HR†
(80% CI) P Value
Rilotumumab + ECX (n = 82) 11.1 (9.5–12.1) 0.73 (0.53–1.01) 0.215
Placebo + ECX (n = 39) 8.9 (5.7–10.6)
HR 0.64
HR 0.73
Selected Biomarkers to find a better population for further study– Biomarkers selected
– Tumor MET protein expression HER2 status MET gene copy number
– Circulating plasma total HGF/SF and soluble MET
– For Met protein expression, used similar methods to the Brivantinib study– Criteria chosen for clinical trial assay performed by a central lab:
– METHigh: > 50% of tumor cells with ≥ 1+ cytoplasmic staining
– METLow: ≤ 50% of tumor cells with ≥ 1+ cytoplasmic staining
– For Met expression ~75% of patients in each arm had samples
Improved PFS and OS in METHigh PatientsProgression-Free Survival
Overall Survival
*HR adjusted for baseline disease extent and ECOG PS.
Time (Months)0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
100
80
60
40
20
0
Pro
gre
ssio
n-F
ree
Su
rviv
al (
%)
Median Months(80% CI)
HR*
(95% CI) P Value
Rilotumumab + ECX (n = 27) 6.9 (5.1–7.5) 0.51 (0.24–1.10) 0.085
Placebo + ECX (n = 11) 4.6 (3.7–5.2)
Median Months(80% CI)
HR+
(95% CI) P Value
Rilotumumab + ECX (n = 27) 11.1 (9.2–13.3) 0.29 (0.11–0.76) 0.012
Placebo + ECX (n = 11) 5.7 (4.5–10.4)
0
20
40
60
80
100
Time (Months)0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Ove
rall
Su
rviv
al (
%)
HR 0.51
HR 0.29
High Levels of Tumor MET May Be Predictive and Prognostic
• Patients with gastric tumors with high MET expression may have a poorer prognosis but may receive more benefit from rilotumumab
Ove
rall
Su
rviv
al (
%)
0
10
20
30
40
50
60
70
80
90
100
Time (Months)0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Median Months(80% CI)
Logrank P value
Placebo + ECX (METLow, n = 17) NE (8.5–NE)0.023
Placebo + ECX (METHigh, n = 11) 5.7 (4.5–10.4)
Rilotumumab in Gastric Cancer Conclusions• Once again, Met expression was prognostic
• And predictive of drug effect
• It is good when these things are consistent across tumor types
• Once again, the company is moving forward with a phase III trial based on the data from a subset analysis
• Once again, I would urge caution, and prefer a less dramatic step
Last thing on cMet trials
• Thank you to the authors for calling the toxicity profiles manageable.
• The most common phrase concluding ASCO abstracts– “This (drug, regimen, etc) was well-tolerated
and further study is warranted”• Often, neither is true
Symptom Management
• We have spent millions of dollars, man/woman-hours, and gigabytes of memory on global quality of life– This has been of limited utility
• Difficult to interpret– Worsening disease negatively affects QOL, as can side
effects of drug
• Heterogeneity as to how we as individuals define QOL• We mostly use it to justify that it is okay to use
borderline effective agents– Desperation will make patients “tolerate” a lot more side effect
Safety events from first SHARP presentation
• Drug-related adverse events (%) All Grade 3/4 All Grade 3/4
Diarrhea 39 8 11 2Pain (abdomen) 8 2 3 <1Weight loss 9 2 <1 0Anorexia 14 <1 3 <1Nausea 11 <1 8 1Hand-foot skin reaction 21 8 3 <1Vomiting 5 1 3 <1Alpecia 14 0 2 0Liver dysfunction <1 <1 0 0Bleeding 7 <1 4 <1/<1
Sorafenib(n=297)
Placebo(n=302)
• Treatment-emergent serious adverse events (SAE, %) 52 54
• Drug-related adverse events (%) 13 9
CTC Version 3.0:
– Grade 2: Skin changes (e.g., peeling, blisters, bleeding, edema) or pain, not interfering with function
– Grade 3: Ulcerative dermatitis or skin changes with pain interfering with function
– While we might not know how to measure global quality of life in a meaningful way, we can choose a symptom/side effect and address it
A Randomized Controlled Phase II Study of the Prophylactic Effect of Urea-Based Cream on the Hand-Foot Skin Reaction Associated with Sorafenib in Advanced Hepatocellular Carcinoma
1 Zhongshan Hospital, Fudan University, Shanghai, China; 2 The Third Affiliated Hospital of Sun Yat-sen University, Guangdong, China; 3 301 Military Hospital, Beijing, China; 4 Eastern
Hepatobiliary Surgery Hospital of the Second Military Medical University, Shanghai, China; 5 Guangdong Provincial People's Hospital, Guangdong, China; 6 Tianjin Cancer Hospital , Tianjin,
China; 7 Jilin Provincial Tumor Hospital, Jilin, China; 8 Union Hospital of Fujian Medical University, Fujian, China; 9 The 81 Hospital of the Chinese People's Liberation Army, Nanjing, China; 10 302 Military Hospital, Beijing, China; 11 Heilongjiang Provincial Cancer Hospital, Heilongjiang, China
Zhenggang Ren1 , Kangshun Zhu2, Haiyan Kang3, Minqiang Lu2, Zengqiang Qu4, Ligong Lu5, Tianqiang Song6, Weiping Zhou4, Hui Wang7, Weizhu Yang8, Xuan Wang9, Yongping Yang10, Lehua Shi4, Yuxian Bai11, Sheng-Long Ye1*
Incidence of Worst-grade HFSR
Grade of HFSR*
Prophylactic Urea-based Cream + BSC
(Arm A) N=439 (%)
BSC(Arm B)
N=432 (%)
TotalN=871
(%)P Value
0 193(44.0) 114(26.4) 307(35.3)
All Grade (%) 246(56.0%) 318(73.6%) 564(64.8%) <.0001
1 155(35.3) 192(44.4) 347(39.8)
2 72(16.4) 98(22.7) 170(19.5)
3 19(4.3) 28(6.5) 47(5.4)
2/3 91 (20.7) 126(29.2) 217(24.9) 0.004
• Primary Endpoint: The incidence of all-grade HFSR was significantly lower in Arm A (p<0.0001)
• Secondary Endpoint: The incidence of grade ≥ 2 HFSR was significantly lower in Arm A
(p=0.004)
• The incidence of HFSR by grade was lower in Arm A
• More patients in Arm A did not develop HFSR
Prevalence of All-grade HFSR at Each Visit
%
*EOS: 2 weeks after end of study
Time to the first HFSR eventSecondary Endpoint
• The median time to the first HFSR event was 2.5 fold longer in Arm A compared to Arm B (84 days vs. 34 days, p<0.0001)
HFSR Conclusions
• Urea-based hand creams helped– Less severity– Delay of onset
• This reduces the number of people who will not get benefit from sorafenib who have to endure this side effect
• It is worth a consideration for your patients• More studies like this should be done to
alleviate the suffering brought on by our treatments
Overall Conclusions
• We all have little tricks we do to help ameliorate side effects of our drugs– But Ren, et al showed that we can
systematically analyze what we are doing– I would argue these types of studies that focus
on a symptom or side effect are more valuable than all the QOL analyses we have ever done
• I cannot name one incidence where we changed our therapies based on global QOL
• We have solely used it to justify what we wanted to do anyhow