1. HCC Journal Club September 2009 Statistical Topic: Phase I studies Selected article: Fong, Boss, Yap, Tutt, Wu, et al. Inhibition of Poly(ADP-Ribose) Polymerase in Tumors from BRCA Mutation Carriers The New England Journal of Medicine July 9, 2009. Vol. 361, No. 2, pp. 123-134

2. Phase I studies  What are the goals? Dose-finding Safety PK and PD  How are they designed?  What is the rationale for dose- selection?

3. Fong et al.  Stated goals: Determine the following— Safety adverse-event profile dose-limiting toxicity maximum tolerated dose (MTD) Dose at which PARP is maximally inhibited PK profile PD profile

4. Study Design  “modified accelerated titration”  Not at all!  TRUE Accelerated titration design: Treat 1 person per dose until either  one DLT is observed  OR, two grade 2 toxicities Then, treat 3 patients per dose level Dose steps can be doubling or not.  Fong study: uses standard 3+3. probably called modified AT because allows doubling of dose in the absence of grade 2 or higher. Is NOT accelerated titration in the spirit of the original paper

5. Back to basics: Acceptable toxicity  What is acceptable rate of toxicity? 20%? 30%? 50%?  What is toxicity???? Standard in cancer: Grade 4 hematologic or grade 3/4 non-hematologic toxicity Always? Does it depend on reversibility of toxicity? Does it depend on intensity of treatment?  Tamoxifen?  Chemotherapy?

6. ‘3+3’ Design  “Standard” Phase I trials (in oncology) use what is often called the ‘3+3’ design (aka ‘modified Fibonacci’):  Maximum tolerated dose (MTD) is considered highest dose at which 1 or 0 out of six patients experiences DLT.  Doses need to be pre-specified  Confidence in MTD is usually poor. Treat 3 patients at dose K 1.If 0 patients experience dose-limiting toxicity (DLT), escalate to dose K+1 2.If 2 or more patients experience DLT, de-escalate to level K-1 3.If 1 patient experiences DLT, treat 3 more patients at dose level K A.If 1 of 6 experiences DLT, escalate to dose level K+1 B.If 2 or more of 6 experiences DLT, de-escalate to level K-1

7. Observed Data in Fong Study

8. Observed Data: with 95% confidence intervals

9. This is actually better than most  Most studies treat only 3 or 6 at each dose level  With 0 of 6 DLTs: Estimated DLT rate = 0% 95% CI for DLT rate = [0%, 45%]  With 1 of 6 DLTs: Estimated DLT rate = 17% 95% CI for DLT rate = [0%, 64%]

10. Why do we use it all the time?  It is terribly imprecise and inaccurate in its estimate of the MTD  Why? MTD is not based on all of the data Algorithm-based method Ignores rate of toxicity!!!  Likely outcomes: Choose a dose that is too high  Find in phase II that agent is too toxic.  Abandon further investigation or go back to phase I Choose a dose that is too low  Find in phase II that agent is ineffective  Abandon agent

11. We could use smarter designs!  Phase I is the most critical phase of drug development!  What makes a good design? MTD situation: Accurate selection of MTD  dose close to true MTD  dose has DLT rate close to the one specified Relatively few patients in trial are exposed to toxic doses  What makes a good design? Non-toxic agent situation: Accurate selection of dose (range) which hits target Relatively few patients are treated Relatively few patients are exposed to ineffective doses

12. This trial  Is MTD relevant?  What is the goal?  Should we be looking for hitting the target?  Toxicity ~ Efficacy?  PK and PD data presented  Although argument made for MTD, PARP inhibition is relatively constant for the higher doses

13. Novel Designs  Why not impose a statistical model?  What do we “know” that would help? Monotonicity (often) Desired level of DLT  Statistical models improve: Prediction Efficiency  Accelerated Titration: incorporates model (next slide)  Example: CRM (continual reassessment method) Originally devised by O’Quigley, Pepe and Fisher (1990) dose for next patient was determined based on toxicity responses of patients previously treated in the trial  Others out there (and variations of CRM)

14. Another Accelerated Titration Feature: Model fit

15. CRM example

17. CRM software example

18. How would CRM have worked in this study?  Would have accelerated quickly  Would have iterated at a few doses  May not have treated so many patients at MTD  Would likely have been a smaller study  Could have used PD data to help dose- finding.

19. Discussion  Phase 0 trials PK and PD single dose 6-10 patients  Goals: Define dose range for Phase I Improve chance of success in phase I and II Better planning of phase I  New and exciting!  First in man, pre-Phase I  Messy though: Phase 0 vs. Phase 1? How will this change Phase 1 goals?  My humble opinion: the development of Phase 0 strongly suggests that Phase I paradigm needs to be reconsidered