Rationale
Platform trial
Study design
Statistical aspects
Simulations
Biostat meeting - May 16, 2019
Outline
Rationale
Clinical challenges
In Sweden, more than 10000 new diagnosed prostate cancer cases every year
\(\sim\) 2500 patients develop metastatic castrate resistant prostate cancer (mCRPC)
Multiplicity of available treatments for mCRPC and new therapies are expected to soon enrich the landscape of available therapies
Very heterogeneous response rates and increasing costs
How to optimize the treatment selection and identify the optimal sequencing of available therapies?
Biomarkers
A molecule, gene, or characteristics which identify a particular pathological process or disease
- Selected binary biomarkers (mutated/not mutated) based on retrospective studies:
- AR alterations (ARA)
- DNA repair deficiencies (DRD)
- PTEN pathway activation (TP53)
- TMPRSS2-ERG gene fusion (TEfus)
Are biomarkers
predictive (not only prognostic) of treatment response?
Statistical challenges
Multiplicity of treatments
Heterogeneous population
Different prevelances (and combinations) of biomarkers
No clear definition of control group
Multiple research questions
Continuous improvements
Platform trial
Why not a traditional clinical trial?
- Traditional clinical trials are ineffective for addressing multiple questions
- Conventional 2-group clinical trials are simple but inefficient (separate control population, different protocols, too large or not sufficiently powered)
- High costs, slow progress, and a high failure rate
- Platform trial may be an efficient strategy for evaluating multiple treatments
Platform trial
- Focus is on the disease rather than any particular experimental therapy.
- One master protocol and consent process, common IT infrastructure, logistics, DSMB.
| Characteristic | Traditional Trial | Platform trial |
|---|---|---|
| Aim | Efficacy of a single agent | Efficacy of multiple agents in heterogeneous population |
| Duration | Finite (to answer one primary question) | Potentially long-term |
| No. treatments | Generally limited | Multiple treatments; new agents may be introduced, other may leave |
| Stopping rules | Interim analysis | Some treatments may be removed (efficacy/futility) but the trial continues |
| Randomization | Fixed | Response-adaptive |
Study design
ProBio objectives
| Primary objective | Investigate whether treatment decision based on biomarkers improves progression free survival (PFS) compared with standard of care. |
| Secondary objectives | Investigate whether treatment decision based on biomarkers improves response rate at 2 months, time to PSA progression, time to radiographic progression, overall survival, quality of life, and health economy. In addition, we will compare adverse events. |
| Design | Randomized platform trial. |
| Study centers | Nation wide study (14 centers), interests from other Nordic countries, Belgium, and the UK. |
ProBio overview
Trial design
Biomarker subgroup combinations
Characteristics specific to the tumor or patient
Defined as the combination of the 4 biomarkers (\(2^4 = 16\) possibilities)
| ARA | DRD | TP53 | TEfus | prev |
|---|---|---|---|---|
| \(-\) | \(-\) | \(-\) | \(-\) | 32.4 |
| \(-\) | \(-\) | \(-\) | \(+\) | 6.7 |
| \(-\) | \(-\) | \(+\) | \(-\) | 17.1 |
| \(-\) | \(-\) | \(+\) | \(+\) | 11.4 |
| \(-\) | \(+\) | \(-\) | \(-\) | 6.7 |
| \(-\) | \(+\) | \(-\) | \(+\) | 4.8 |
| \(-\) | \(+\) | \(+\) | \(-\) | 1.9 |
| \(-\) | \(+\) | \(+\) | \(+\) | 1.0 |
| \(+\) | \(-\) | \(-\) | \(-\) | 4.8 |
| \(+\) | \(-\) | \(-\) | \(+\) | 3.8 |
| \(+\) | \(-\) | \(+\) | \(-\) | 1.0 |
| \(+\) | \(-\) | \(+\) | \(+\) | 3.8 |
| \(+\) | \(+\) | \(-\) | \(-\) | 2.9 |
| \(+\) | \(+\) | \(-\) | \(+\) | 1.0 |
| \(+\) | \(+\) | \(+\) | \(-\) | 1.0 |
Biomarker signatures
Groupings of biomarker subgroup combination (or biomarker) that are potential indications for treatment decision and/or prognosis
- Signatures of interest:
- All patients (null hypothesis of the trial)
- TP53- & AR- (hormonal therapies)
- TP53+ (poor prognosis)
- DRD+ (Carboplatin)
- TEfus+ (chemiotherapy)
| signatures | —- | —+ | –+- | –++ | -+– | -+-+ | -++- | -+++ | +— | +–+ | +-+- | +-++ | ++– | ++-+ | +++- | prev |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| all | X | X | X | X | X | X | X | X | X | X | X | X | X | X | X | 100.0 |
| TP53- & AR- | X | X | X | X | 50.5 | |||||||||||
| TP53+ | X | X | X | X | X | X | X | 37.1 | ||||||||
| DRD+ | X | X | X | X | X | X | X | 19.0 | ||||||||
| TEfus+ | X | X | X | X | X | X | X | 32.4 |
Subgroup combinations vs. signatures
Each patient belongs to one and only one biomarker subgroup combination, while he may belong to multiple biomarker signatures
Prevalences of some subgroup combinations may be low, while prevalences of biomarker signatures are generally moderate
Key points
Patients are stratified based on their biomarker subgroup combination, and then randomized to either the control (standard of care) or one of the active arms
Outcome-adaptive randomization is implemented to assign more patients to more promising (effective) treatments within the biomarker subgroup combinations
Treatments are constantly (monthly) evaluated within the biomarker signatures
Highly effective treatments will graduate from the platform trial and enter into a validation trial (fixed randomization 1:1)
Patients who progress, will be re-genotyped and re-randomized (max 2 randomizations)
Statistical aspects
(Outcome-Adaptive) Randomization
Fixed before enrolling 50 patients in the active arms
After, randomization probabilities will be updated monthly based on the accumulated data (PFS)
Proportional to \(\pi_{ij}\), the (Bayesian) probability of superiority for a treatment \(i\) in the biomarker subgroup combination \(j\):
\[P_{ij}(t) \propto P_{ij}^2(t-1) + \pi_{ij}^2\]
Probability of superiority (within biomarker subgroup combination)
Evaluation of therapies
Treatments are compared within the biomarker signatures of interest using the control group as comparator
the main outcome is a survival time. We will use Bayesian parametric model to contrast the distributions of PFS
Monthly, we will decide if continuing enrollment of new patients to a treatment signature combination, or to early stop (graduation, futility, max patients)
Decision rules for treatment \(i\) in the biomarker signature \(s\)
- graduation:
- \(n_{is} \ge 20\)
- \(\pi_{is} \ge .85\)
- \(\pi_{ij} \ge .65\), with \(j \in s\)
- \(n_{is} \ge 20\)
- stop futility:
- \(n_{is} \ge 20\)
- \(\pi_{is} \le .15\)
- \(\pi_{ij} \le .3\), with \(j \in s\)
- \(n_{is} \ge 20\)
- stop max patients:
- \(n_{is} \ge 150\)
Fictitious data
| Control | Treatment |
|---|---|
| 0.93, 1.15, 1.42, 2.55, 2.63, 2.87, 3.08, 3.97, 5.49, 5.81, 6.34, 6.43, 6.68, 6.95, 7.43, 7.43, 7.99, 8.69, 10.29, 10.88, 11.91, 16.88, 19.93, 20.00+, 20.00+ | 0.31, 0.48, 1.19, 2.66, 3.18, 3.89, 4.81, 5.23, 5.26, 5.62, 6.09, 6.78, 8.12, 8.46, 8.49, 10.51, 15.35, 17.06, 19.61, 20.00+, 20.00+, 20.00+, 20.00+, 20.00+, 20.00+ |
Probability of superiority over time
Additional details
- \(T_{is} \sim \text{Weibull}(\lambda_{is}, k)\)
- \(f(t_{is};\lambda_{is},k) = \lambda_{is} kt_{is}^{k-1}\exp(-\lambda_{is} t_{is}^k)\)
\(k = 1.05\) based on the ProBio pilot data
- \(\lambda_{is} \sim \text{Gamma}(a, b)\), with \(a = 10\), and \(b = 80\)
- \(\lambda_{is} | \underline{t_{is}} \sim \text{Gamma}(a^*, b^*)\), with \(a^* = a + \sum_{j = 1}^{n_{is}} \delta_{jis}\) and \(b^* = b + \sum_{j = 1}^{n_{is}} t_{jis}^k\)
\(P(\mu_{is} > \mu_{os})\) is computed through MCMC
Simulations
Operating characteristics
- Complicated design – complicated assessment of operating characteristics
- Statistical simulations
- Different types of errors
- Graduation when there is no effect
- Fail to graduate when there is an effect
- Graduate drug-signature combination where the biomarker subgroup is too large
- Fail to graduate drug-signature combination where the biomarker subgroup is too small
- Graduation when there is no effect
- Discovery trial, relatively liberal graduation criteria
- Graduated combinations will be validated in a side trial nested within the ProBio platform
- Simulation Scenarios:
- no treatment works better in any signature
- one treatment (Carboplatin) works better (HR \(\approx\) 2)
- one treatment (Carboplatin) works better (HR \(\approx\) 2.4) only in the biomarker subgroup combination belonging to the signature DRD+
- hormonal therapies work better in the signature TP53- & AR-, Carboplatin in the signature DRD+, while chemotherapies work better in the signature TEfus+ with HR ranging from 2 to 3.
- no treatment works better in any signature
- Results:
- Assumed PFS times (median)
- Error rate and power
- Average number of enrolled participants
- Probabilities of superiority
- Graduation times
Scenario 1
True PFS times
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel ---- 6.32 6.32 6.32 6.32 6.32 ---+ 6.32 6.32 6.32 6.32 6.32 --+- 6.32 6.32 6.32 6.32 6.32 --++ 6.32 6.32 6.32 6.32 6.32 -+-- 6.32 6.32 6.32 6.32 6.32 -+-+ 6.32 6.32 6.32 6.32 6.32 -++- 6.32 6.32 6.32 6.32 6.32 -+++ 6.32 6.32 6.32 6.32 6.32 +--- 6.32 6.32 6.32 6.32 6.32 +--+ 6.32 6.32 6.32 6.32 6.32 +-+- 6.32 6.32 6.32 6.32 6.32 +-++ 6.32 6.32 6.32 6.32 6.32 ++-- 6.32 6.32 6.32 6.32 6.32 ++-+ 6.32 6.32 6.32 6.32 6.32 +++- 6.32 6.32 6.32 6.32 6.32
Error rate
[1] 0.978
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel
0.116 0.086 0.104 0.084 0.114
Average number of participants
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 153.1 156.2 146.1 144.1 149.3 TP53- & AR- 80.9 82.9 70.7 65.5 68.6 TP53+ 54.6 55.5 55.4 58.5 60.5 DRD+ 23.5 24.8 44.1 24.9 23.6 TEfus+ 46.8 47.5 47.4 52.0 53.4
Probabilities of superiority
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 0.39 0.40 0.40 0.38 0.40 TP53- & AR- 0.41 0.42 0.42 0.41 0.41 TP53+ 0.44 0.44 0.44 0.43 0.45 DRD+ 0.44 0.45 0.47 0.44 0.44 TEfus+ 0.43 0.43 0.44 0.42 0.43
Time to graduation
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 35.88 35.93 35.71 35.91 35.88 TP53- & AR- 34.22 34.64 34.85 35.05 34.73 TP53+ 35.73 35.70 35.57 35.81 35.54 DRD+ 35.93 35.94 35.67 35.95 35.96 TEfus+ 35.81 35.77 35.80 35.68 35.86
Scenario 2
True PFS times
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel ---- 6.32 6.32 12.81 6.32 6.32 ---+ 6.32 6.32 12.81 6.32 6.32 --+- 6.32 6.32 12.81 6.32 6.32 --++ 6.32 6.32 12.81 6.32 6.32 -+-- 6.32 6.32 12.81 6.32 6.32 -+-+ 6.32 6.32 12.81 6.32 6.32 -++- 6.32 6.32 12.81 6.32 6.32 -+++ 6.32 6.32 12.81 6.32 6.32 +--- 6.32 6.32 12.81 6.32 6.32 +--+ 6.32 6.32 12.81 6.32 6.32 +-+- 6.32 6.32 12.81 6.32 6.32 +-++ 6.32 6.32 12.81 6.32 6.32 ++-- 6.32 6.32 12.81 6.32 6.32 ++-+ 6.32 6.32 12.81 6.32 6.32 +++- 6.32 6.32 12.81 6.32 6.32
Error rate
[1] 0.424
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel
0.116 0.114 0.172 0.094 0.102
Power
[1] 0.806
Average number of participants
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 160.6 156.9 96.2 149.2 149.4 TP53- & AR- 85.6 82.4 45.2 67.7 66.1 TP53+ 56.1 55.3 38.2 60.2 61.6 DRD+ 25.6 26.2 30.4 25.3 25.1 TEfus+ 49.8 49.9 26.9 55.2 56.3
Probabilities of superiority
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 0.40 0.38 0.98 0.38 0.38 TP53- & AR- 0.41 0.41 0.96 0.40 0.39 TP53+ 0.44 0.42 0.88 0.42 0.44 DRD+ 0.44 0.44 0.90 0.43 0.42 TEfus+ 0.44 0.43 0.86 0.42 0.43
Time to graduation
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 35.83 35.78 20.38 35.71 35.83 TP53- & AR- 34.55 34.40 21.95 34.96 35.02 TP53+ 35.76 35.81 27.94 35.67 35.85 DRD+ 35.92 35.79 25.22 35.95 35.95 TEfus+ 35.72 35.84 31.20 35.68 35.54
Scenario 3
True PFS times
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel ---- 6.32 6.32 5.11 6.32 6.32 ---+ 6.32 6.32 5.11 6.32 6.32 --+- 6.32 6.32 5.11 6.32 6.32 --++ 6.32 6.32 5.11 6.32 6.32 -+-- 6.32 6.32 15.13 6.32 6.32 -+-+ 6.32 6.32 15.13 6.32 6.32 -++- 6.32 6.32 15.13 6.32 6.32 -+++ 6.32 6.32 15.13 6.32 6.32 +--- 6.32 6.32 5.11 6.32 6.32 +--+ 6.32 6.32 5.11 6.32 6.32 +-+- 6.32 6.32 5.11 6.32 6.32 +-++ 6.32 6.32 5.11 6.32 6.32 ++-- 6.32 6.32 15.13 6.32 6.32 ++-+ 6.32 6.32 15.13 6.32 6.32 +++- 6.32 6.32 15.13 6.32 6.32
Error rate
[1] 0.334
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel
0.106 0.112 0.054 0.092 0.096
Power
[1] 0.802
Average number of participants
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 163.0 162.7 101.1 151.6 150.3 TP53- & AR- 87.5 84.5 47.3 68.3 67.5 TP53+ 56.5 59.6 38.0 62.6 62.4 DRD+ 23.3 23.3 37.7 23.7 24.1 TEfus+ 49.0 49.0 35.9 53.4 54.1
Probabilities of superiority
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 0.39 0.40 0.82 0.38 0.39 TP53- & AR- 0.40 0.43 0.89 0.40 0.40 TP53+ 0.43 0.43 0.54 0.43 0.44 DRD+ 0.43 0.43 0.97 0.43 0.43 TEfus+ 0.42 0.43 0.65 0.42 0.43
Time to graduation
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 35.91 35.88 33.61 35.78 35.62 TP53- & AR- 34.40 34.46 32.05 35.16 34.99 TP53+ 35.66 35.80 35.85 35.90 35.63 DRD+ 35.93 35.86 21.30 35.90 35.96 TEfus+ 35.88 35.81 35.09 35.63 35.62
Scenario 4
True PFS times
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel ---- 11.06 12.23 5.11 5.11 5.11 ---+ 11.06 12.23 5.11 15.70 15.70 --+- 5.11 5.11 5.11 5.11 5.11 --++ 5.11 5.11 5.11 15.70 15.70 -+-- 11.06 12.23 13.97 5.11 5.11 -+-+ 11.06 12.23 13.97 15.70 15.70 -++- 5.11 5.11 13.97 5.11 5.11 -+++ 5.11 5.11 13.97 15.70 15.70 +--- 5.11 5.11 5.11 5.11 5.11 +--+ 5.11 5.11 5.11 15.70 15.70 +-+- 5.11 5.11 5.11 5.11 5.11 +-++ 5.11 5.11 5.11 15.70 15.70 ++-- 5.11 5.11 13.97 5.11 5.11 ++-+ 5.11 5.11 13.97 15.70 15.70 +++- 5.11 5.11 13.97 5.11 5.11
Error rate
[1] 0.078
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel
0.012 0.012 0.048 0.074 0.074
Power
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel
0.268 0.384 0.532 0.356 0.360
Average number of participants
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 162.5 167.8 115.3 119.6 120.6 TP53- & AR- 103.6 105.6 49.0 34.2 35.5 TP53+ 43.6 46.3 47.1 64.7 64.7 DRD+ 19.8 19.3 50.5 19.2 19.0 TEfus+ 34.8 34.0 36.5 62.5 61.4
Probabilities of superiority
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 0.17 0.24 0.44 0.55 0.55 TP53- & AR- 0.63 0.74 0.57 0.37 0.37 TP53+ 0.15 0.15 0.37 0.69 0.68 DRD+ 0.30 0.32 0.90 0.33 0.33 TEfus+ 0.19 0.22 0.31 0.88 0.88
Time to graduation
Enzalutamide Abiraterone Carboplatin Cabazitaxel Docetaxel all 35.90 35.81 35.95 35.92 35.95 TP53- & AR- 31.27 29.25 35.16 35.93 35.89 TP53+ 35.97 35.96 35.91 34.92 34.93 DRD+ 35.94 35.99 26.52 35.96 35.98 TEfus+ 35.97 36.00 35.86 30.01 29.96