Exposure-Response in Oncology Yaning Wang, Ph.D. Associate Director for Science Division of Pharmacometrics Office of Clinical Pharmacology OTS/CDER/FDA Disclaimer: My remarks today do not necessarily reflect the official views of the FDA
Outline • Motivation • Case study • Summary 2
Declining Success Rate in Oncology Phase 3 Trials 21% 80 67 70 60 Success Rate (%) 46 50 40 30 20 10 0 1997-2002 2002-2007 Drivers of Attrition-McKinsey & Co. Report 2008 3
Drivers for Oncology Phase 3 Failure (1990-2007) 47 50 40 32 Percent 30 20 15 6 10 0 0 Efficacy vs placebo Safety-unexpected Efficacy vs active Safety vs active Safety-expected 4 Drivers of Attrition-McKinsey & Co. Report 2008
Poor Estimation of Efficacy from Earlier Clinical Trials • Novel or ‘unprecedented’ mechanisms • Empirical selection of dose and dosing schedules • Poor correlation between biomarkers and clinical response • Short, single-arm phase 2 trials with small sample size • Large uncertainty on survival benefit 5 after phase 2 trials
Case Study: Herceptin • Regimen approved for HER2-overexpressing breast cancer was selected as the Phase 3 regimen for metastatic HER2- overexpressing gastric cancer • Initial assumption: median OS of 7 months for fluoropyrimidine (capecitabine or 5-fluorouracil) and cisplatin (FC) and 10 months for Herceptin+FC • Three interim safety analyses and one interim efficacy (50% death, 124) analysis • Third safety interim analysis showed combined median OS of 12 months • New assumption: median OS of 10 months for FC and 13 months for Herceptin+FC • Two interim efficacy analyses (50% death, 230; 75% death 345) • Second efficacy interim analysis (348 deaths) was significant at pre-defined significance level and was used as final analysis • Close of study was one year after 2 nd interim analysis, the time projected to have 460 deaths (full data, OS benefit: 1.4 month) 6
Phase 3 Clinical Trial Chemotherapy FC : Cisplatin+ R (FC) A capecitabine/5-FU N D O M I Z A T Trastuzumab + FC (T+FC) I O N Trastuzumab: 8mg/kg followed by 6mg/kg q3w N=594, 1:1 7
Trastuzumab Demonstrated Overall Survival Benefit of 1.4 Months FC (n=296) Trastuzumab+FC (n=298) Patients with Events 227 (76.7%) 221 (74.2%) Patients without Events 69 (23.3%) 77 (25.8%) Median OS in Months (95%CI) 11.7 (10.3, 13.0) 13.1 (11.9, 15.1) Hazard ratio (95%CI) 0.80 (0.67, 0.97) 8
Steep Exposure-Survival Relationship Patients with Cmin <12 µg/mL Had 7-10 Months Shorter Median OS 1.0 Median Survival (95% CI) Q1:Cmin <11.8 Q1:Cmin <11.8 7.7 (6.3-10.6) 0.9 Q2:Cmin 11.8-16.2 Q2:Cmin 11.8-16.2 14.1 (9.5-19.3) Overall Survival Q3:Cmin 16.2-21.4 Q3:Cmin 16.2-21.4 15.5 (13.1-24.3) 0.8 Q4:Cmin >21.4 Q4:Cmin >21.4 17.9 (14.8-21.2) 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0 10 20 30 40 50 9 Study Month
Confounding Risk Factors for Survival Lowest exposure pts (Q1) vs. the rest pts (Q2-Q4) Q2-Q4 (N=199) Q1 (N=67) 100 70% 87% 90 80 Percentage 40% 62% 70 42% 54% 60 50 40 5% 24% 30 20 10 0 ECOG Prior Asia (No) Number of Poor ECOG No Prior # of metastatic 10 Performance Gastrectomy Non-Asian Metastatic Sites Gastrectomy performance sites >2 Status (2) (No) (>2)
Is Herceptin Shortening Life? 1.0 Median Surival (95%CI) FC (all) 11.7 ( 10.5 , 13.1 ) 0.8 Survival Distribution H+FC (Q1) 7.7 ( 6.5 , 10.9 ) 0.6 0.4 0.2 0.0 0 10 20 30 40 50 11 Time (month)
Percentage of High Risk Patients Before and After Matching within Low Exposure Patients Before Matching After Matching FC (N=296) H+FC (N=67) FC (N=67) H+FC (N=67) 100 100 90 90 80 80 70 70 60 Percentage 60 50 50 40 40 30 30 20 20 10 10 0 0 ECOG Prior Asia (No) Number of ECOG Prior Asia (No) Number of Performance Gastrectomy Metastatic Sites Performance Gastrectomy Metastatic Sites Status (2) (No) (>2) Status (2) (No) (>2) 12
No Survival Benefit in Low Exposure Patients Compared to Matched Control Remaining (75%) Low Exposure (25%) 1.0 1.0 1.0 Median Surival (95%CI) Median Surival (95%CI) Median Surival (95%CI) FC FC FC 7.5 ( 6.4 , 10.8 ) 12.8 ( 11.2 , 14.2 ) 12.8 ( 11.2 , 14.2 ) 0.8 0.8 0.8 Survival Distribution FC+Trastuzumab FC+Trastuzumab FC+Trastuzumab 7.7 ( 6.5 , 10.9 ) 15.7 ( 14.1 , 19 ) 15.7 ( 14.1 , 19 ) 0.6 0.6 0.6 8 - 5.3 = 2.7 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 0 10 20 30 40 50 0 10 20 30 40 50 13 Time (month) Time (month)
Reason for Lack of Benefit in Selected Subgroup (Matched ) • Low trastuzumab exposure (Cmin <12 ug/mL) – Implication: higher dose may work • Low exposure (high risk) patients may be non-responders? – Implication: higher dose may not work 14
Support for Higher Dose Per 10 ug/mL increase in Cmin 0.69 Low Exposure (25%) 0.85 Remaining (75%) 0.2 0.4 0.6 0.8 1.2 Hazard Ratio • All patients are sensitive to higher trastuzumab exposure • Low exposure (high risk) patients may be more sensitive to higher trastuzumab exposure 15
Lower PK Exposure in Advanced Gastric Cancer Patients Breast cancer pts Ctrough Advanced gastric cancer pts Time (weeks) • Steady state Ctrough 24-63% lower than that in breast cancer patients using the same dosing regimen
Publication 17
Summary of Case Study • Lack of exposure-response data to justify regimen, leading to no survival benefit for 25% population • Lack of justification for assumed effect size led to under-powered study • Exposure-response (ER) and case-control analyses provided the rationale for post marketing requirement (PMR) study for a higher dose • Trial design (dose selection, patient population and effect size assumption) for PMR study was based on ER and case- control analyses 18
Conclusion • Better efficacy estimation is needed for oncology phase 3 trials • Pharmacometric analyses, such as biomarker-survival relationship, ER modeling, should be routinely conducted to support dosing regimen selection, effect size estimation, patient population selection. • Clinical trial simulation should become the standard procedure for phase 3 trial design 19
Acknowledgment • Jun Yang • Hong Zhao • Christine Garnett • Joga Gobburu • Atik Rahman • William Pierce • Genny Schechter • Patricia Keegan • Jeff Summers 20
Recommend
More recommend
