Transcript Phase I Study Design

Designing Phase I Trials with Targeted Agents:
Challenging Tradition
Alex A. Adjei, MD, PhD
Definition of a Phase I Study
• First evaluation of a new cancer therapy in humans
• First in-human single-agent study
• Combination of novel agents
• Combination of novel agent and approved agent
• Combination of novel agent and radiation therapy
• Eligible patients usually have refractory solid tumors of any type
The 3 Basic Tenets of Phase I Studies
• Define a recommended dose:
– SAFELY (minimum number of serious toxicities)
– EFFICIENTLY (smallest possible number of patients)
– RELIABLY (high statistical confidence)
– SAFETY TRUMPS EVERYTHING ELSE
Preclinical Toxicology
• Typically, a rodent (mouse or rat) and non-rodent (dog or non-human primate)
species
• Very few animal organ-specific toxicities predict human toxicity
– Bone marrow and GI toxicity more predictable
– Hepatic and renal toxicities – large false-positives
– Toxicologic parameters
• LD10 – lethal dose in 10% of animals
• TDL (toxic dose low) – lowest dose that causes any toxicity in animals
Phase I Trials: Starting Dose
• 1/10 of the LD10 in rodents
or
• 1/3 of the TDL in large animals
• Expressed as mg/m2
• These have historically been safe doses
Equivalent Surface Area Dosage Conversion Factors
Mouse
20 g
Rat
150 g
Monkey
3 kg
Dog
8 kg
Man
60 kg
Mouse
1
1/2
1/4
1/6
1/12
Rat
2
1
1/2
1/4
1/7
Donkey
4
2
1
3/5
1/3
Dog
6
4
5/3
1
1/2
Man
12
7
3
2
1
Freireich EJ et al. Cancer Chemother Rep. 1966;50:219-244.
Phase I Dose Escalation
The modified Fibonacci schedule
Cohort
Dose
Escalation (%)
1
n
First dose
2
2n
100
3
3.3 n
67
4
5n
50
5
7n
40
6 and higher
25-33
Phase I Study End Points
• Dose, toxicity, pharmacology (efficacy?)
• Classical goals
– Identify DLTs
– Identify the MTD
– Assess pharmacokinetics
• Evaluate target modulation
Defining Toxicities: NCI Common Toxicity Criteria
• Grade 1 = mild
• Grade 2 = moderate
• Grade 3 = severe
• Grade 4 = life-threatening
• Grade 5 = fatal
Maximum Tolerated Dose
• Inconsistently defined as either:
– Dose at which 33% of patients experience unacceptable toxicity (DLT in 2 of 3 or
2 of 6)
or
– 1 dose level below that
• MTD = level @ DLT (in Europe or Japan)
• MTD = level below DLT (in US)
• 6-10 patients treated at the recommended phase II dose (MTD or 1 dose level
below)
Recap: Transatlantic Differences in Terminology
• Important to note that “Maximum tolerated dose” (MTD):
– Usually means “recommended phase 2 dose (RP2D)” in US
– Usually means dose level above RP2D in Europe and some other countries
Dose-Limiting Toxicities
• Toxicities that are considered to be unacceptable and limit further dose
escalation
• Defined in advance of starting trial
• Classically based on cycle 1 toxicity
• Examples
– ANC <500 for 5 or 7 days
– ANC <500 of any duration with fever
– PLT <10,000 or 25,000
– Grade 3 or greater non-hematologic toxicity
– Inability to re-treat patient within 2 weeks of scheduled treatment
Definition of DLT Is Dynamic
• Examples: DLTs in 2014
– Diarrhea: ≥ Grade 3 in spite of adequate antidiarrheal therapy (loperamide)
– Nausea and vomiting: ≥ Grade 3 in spite of adequate
antiemetic prophylaxis and therapy (steroids, 5HT3 antagonists)
– Hypertension: ≥ Grade 3 in spite of adequate
antihypertensive therapy
– Rash : ≥ Grade 3 in spite of adequate therapy
– Hyperglycemia : ≥ Grade 3 in spite of adequate therapy
– Inability to take at least 90% of drug doses in a cycle (continuous oral meds)
– Grade 2 chronic unremitting toxicity
Phase I Standard 3 + 3 Design
DLT
Dose
Recommended dose
3 pts
3 pts
3 pts
3 pts
3 pts
Eisenhauer EA et al. J Clin Oncol. 2000;18:684-692.
Starting dose
+
3 pts
DLT
3 pts
Classic Phase I Trial Design Limitations
• Wide confidence intervals
• Patients treated at ineffective doses in first cohorts
• High risk of severe toxicities in late cohorts
Intrapatient Dose Escalation
• Treat patients at dose level 1
• Dose level 2 is well tolerated and patients at dose level 1 have no toxicities
• Patients at level 1 are escalated to level 2
WHY NOT ALWAYS DO THIS?
• Makes evaluation of chronic toxicities difficult
• The proverbial 1 responder at dose level 1
Response to lonafarnib and EKB569
Pre-FTI
March 29, 2001
3 Months
September 19, 2001
Novel Designs – Wish List
• Maximise safety
–  patients exposed to DLT
– Safe RP2D
• Maximise chance benefit
–  patients exposed to likely subtherapeutic doses of drugs
• Efficiency ( N patients,  speed)
• Reduce time trial is on hold
Frequentist
Rule-based
Bayesian
Model-based
Frequentist Designs
Design
Dose
PK-guided
Double until target AUC (40% of murine at LD10)
then modified Fibonacci
Accelerated
titration
Double dose until target toxicity (<DLT) then
switch to more conservative (eg, Fibonacci until
MTD)
Accelerated Titrated Design
DLT
Dose
Recommended dose
GR 2 Toxicity
3 pts
3 pts
1 pt
1 pt
1 pt
Eisenhauer EA et al. J Clin Oncol. 2000;18:684-692.
Starting dose
+
3 pts
DLT
3 pts
Bayesian Designs
Escalation Scheme
Description
Modified continual
reassessment method
(mCRM)
Preset estimated MTD and dose levels. Update MTD
statistically on basis of each pt’s data
http://www.cancerbiostats.onc.jhmi.edu/software.cfm
TriCRM
Incorporates both toxicity and efficacy data into the
estimation of the biologically optimal dose – but OR
takes time to mature… (phase I/II better?)
EWOC (escalation with
overdose control)
Uses real-time toxicity data to make decisions
http://sisyphus.emory.edu/software_ewoc.php
Many more variations, some including PK
Phase I Trial Design: Targeted Agents
• MTD may not be the goal of phase I, as specificity of effect may be lost at
MTD
• Pharmacologic effect may not equal biologic effect
• Goal: Identify optimal biologic dose (OBD)
• Biomarkers can guide dose escalation and dose selection
Erlotinib PK
Patnaik A et al. Clin Cancer Res. 2006;12:7406-7413.
Phase II Study of the MEK Inhibitor, CI-1040
Steady State CI-1040 Plasma Concentration
Concentration (ng/mL)
1000
100
10
Course 1
Course 2
Course 3
Course 4 or later
Expected Concentration
1
0
4
8
12
Time (hr)
Rinehart et al. J Clin Oncol. 2004;22:4456-4462.
16
20
24
What’s the Target?
• Sorafenib (Raf kinase inhibitor) VEGFR1–3
• 5-Azacytidine (antimetabolite) methylation
• Imatinib (PDGFR) Bcr-Abl, kit
• Crizotinib (MET) EML4-ALK
• Iniparib (PARP) ??? alkylating agent forming adducts with cysteine-rich
proteins
• Tivantinib (MET) anti-tubulin
UNUSUAL TOXICITIES OF
TARGETED AGENTS
Signal Transduction Inhibitors (Dy and Adjei, CA-A, 2013)
Courtesy of Alex Adjei, MD, PhD.
Cutaneous Toxicities
Radiation dermatitis
Paronychia
Trichomegaly
Onycholysis
Courtesy of Alex Adjei, MD, PhD.
RASopathic Skin Toxicities
Photosensitivity
Radiation recall
Posterior Reversible Encephalopathy Syndrome (PRES)
Vatalanib
Bevacizumab
Sorafenib
MEK Inhibitor – Blurred Vision 2005
CSR: Subretinal fluid - 2012
SHOULD WE BE TREATING GENOMIC
SUBSETS IN PHASE I?
Somatic Mutations: Recurrent in Common Cancers
Dancey et al. Cell. 2012;148:409-420.
Phase I Study Design – Unselected Patients in Dose Escalation Followed by
Specific Expansion Cohorts
Dose Escalation
• PK, safety
• Define MTD
Cohort Expansion
Pharmacodynamics
Targeted Tumour Types
• Biopsies
• Functional imaging
• Molecular enrichment
• Histologic enrichment
PLX4032: Novel, Small-Molecule Inhibitor
Selectivity for BRAFV600E in vitro and in vivo
Selective for BRAFV600E kinase among 70 kinases screened
Selective in cellular assays
Phospho-ERK
IC50 (nM)
V600E
A375
COLO829
COLO205
20
10
30
WT
SW620
SKMEL2
>40,000
14,000
IC50 (nM)
10−100
100−1000
1000−10,000
Kinase domain binding
Selective regression of V600E
10
PLX4720
co-structure with
kinase domain of
BFAFV600E
Tumour
Size
100 mm3
5
0
Flaherty et al. ASCO 2009.
Tsai J et al. Proc Natl Acad Sci USA. 2008;105:3041-3046.
Vemurafenib (PLX4032) in Melanoma: Phase I Studies
Flaherty et al. ASCO 2009.
Tsai J et al. Proc Natl Acad Sci USA. 2008;105:3041-3046.
Phase I Study Design – Only Molecularly Enriched Patients
Dose Escalation
• PK, safety
• Define MTD
Cohort Expansion
Pharmacodynamics
Targeted Tumour Types
• Biopsies
• Functional imaging
• Histologic enrichment
Phase I Trial of BYL719: Study Design
MTD expansion phase
PIK3CA altered solid tumours
Dose-escalation phase
PIK3CA altered solid tumours
Oral, once-daily BYL719, 28-day cycle
(n=36)
Declaration
of MTD
Dose levels
Juric et al. AACR 2012.
400 mg QD BYL719
(Approximately 45 patients)
in PIK3CA-altered solid tumors
Number of Patients Needed to be Screened for
70 Marker + Patients
Prevalence of Marker
Expected Number to be Screened
0.05
1394
0.1
697
0.2
349
0.3
232
0.4
174
0.5
139
0.6
116
0.7
100
0.8
87
1
70
Sleijfer S et al. J Clin Oncol. 2013;31:1834-1841.
Molecular Selection of Patients for Phase I Trials
Nested clinical trial
Nested clinical trial
Nested clinical trial
• Multiplexed molecular prescreening at local site
• Reduces change of tissue exhaustion from multiple trial
demands
• Underscores the importance to have validated broad
molecular profiling capacity of actionable mutations and
other aberrations in local validated laboratory
• Multicentre participation – especially if seeking rare
molecular subsets
N-of-1 Trials: A Possible Design for Personalized Medicine?
Crossover study performed within a single individual
Drug A
Drug B
Drug B
Drug A
Drug B
Drug A
Drug A
Drug B
Drug A
Drug B
Average
treatment
effect of
Drug A
vs
Average
treatment
effect of
Drug B
70 year old man with Lung Cancer and a Braf Mutation :
Vemurafenib after Progression on Pemetrexed/Carboplatin
October 30, 2013
August 25, 2013
Conclusions
• Alternate designs are important but not the solution
– For most drugs, an MTD can be defined
– It may be chronic toxicity rather than classical cycle 1 DLTs
– Evaluate intermediate doses
– Biomarker end points are necessary for the few agents, such as antibodies, with no
dose-dependent toxicities
• Phase I trials in molecular subsets should be the exception at this time
– May miss other populations who will benefit
– Study can be slow and expensive
– May be more informative to treat a small subset of “wild-type” patients (eg, escalate
quickly and expand in molecular subset)