Presentation is loading. Please wait.

Presentation is loading. Please wait.

Pharmacometrics: A Business Case May 25, 2010 Pharmacometrics Task Force.

Similar presentations


Presentation on theme: "Pharmacometrics: A Business Case May 25, 2010 Pharmacometrics Task Force."— Presentation transcript:

1 Pharmacometrics: A Business Case May 25, 2010 Pharmacometrics Task Force

2 | 1 What is Pharmacometrics (PM)? Pharmacometrics (PM) analyses are: Quantitative analyses of data pertaining to: Pharmacokinetics Biomarkers Clinical outcomes Disease characteristics Trial characteristics Can include: – Mathematical modeling and simulation – Statistical analysis Often used to facilitate efficient drug development and approval process Needs multidisciplinary team consisting of quantitative clinical pharmacologists, statisticians, engineers, data management experts, and clinicians

3 | 2 Utilization of PM allows for a more efficient drug development process Propose best doses Estimate effect size Rescue discarding good drug Main advantages Target patient selection Maximize value of prior data Drug approval Labeling The cost of PM is marginal compared to the final cost of a trial. In most cases, the level of effort for PM is as low as 1 person for 2-6 months 1 1 Depending on the complexity of analysis, the level of effort may be slightly lower or higher

4 | 3 Common questions about PM No! Will it delay my NDA? Will it take a long time? Does it take many resources to deliver?

5 | 4 Example benefits (1/2) Scenario and impact examples Propose best doses Exposure – response relationships for organ rejection (effectiveness) and creatinine clearance (safety) were developed. Simulations explored alternative dosing regimens to optimize benefit – risk profile; level of effort was 1 person for less than 2 months. Pharmacometric analyses presented to Cardio- Renal Advisory Committee in 2005 with recommendations to conduct another clinical study using an optimized dosing regimen (pg 9) 1 Highly variable PK of Tacrolimus between ulcerative colitis patients and high trough concentrations in Phase II studies presented challenges to further development. Simulation of dose titration based on exposure-response was effective for identifying target trough concentration, demonstrating effectiveness and justifying Phase III studies (pg 13) 2 The exposure-viral load reduction model predicted the effect at different doses, resulting in a range of possible active oral doses used in the design of phase IIa trials (pg 15) 3 Estimate effect size Enhanced the trial success by increasing study duration; trial now suitable for registration (pg 21) 4 A new dosing regimen was selected based on pharmacometric analyses and evaluated in an additional clinical trial. Nesiritide was approved by FDA in 2001 (pg 30) Rescue discarding good drug 5

6 | 5 Example benefits (2/2) Scenario and impact examples Pharmacometric dose – response analysis identified the proportion of mildly diseased non-responders was the primary cause of lack of evidence of effectiveness. FDAs approvable letter suggested that sponsor conduct a future study including patients with moderate and severe disease (pg 35) Target patient selection 6 Maximize value of prior data Approval of oxcarbazepine monotherapy in pediatrics was based on demonstrating similar exposure – response relationship for seizure frequency in pediatrics and adults using prior data from adjunctive therapy trials. No additional monotherapy pediatric trials were required (pg 38) 7 Drug approval Confirmatory evidence provided by significant dose–response (chorea score) relationship. Internal consistency of results across one positive, one negative trial and their extensions. (pg 40) 8 Clinical trial simulations a paricalcitol dosage regimen based on iPTH/80 was predicted to significantly lower the rate of hypercalcemia, compared to the iPTH/60 based regimen tested in clinical trials, without significantly impacting efficacy. Oral paricalcitol was approved by the FDA for use in CKD Stage 5 patients at a dose of iPTH/80 TIW without the conduct of further clinical trials in patients (pg 41) 9 Labeling Cleviprex dosing regimen used in clinical trials resulted in overshooting and oscillations around the target blood pressure. Simulations of the exposure – response relationship were used to optimize the dosing regimen to quickly achieve and maintain target blood pressure (pg 44) 10

7 | 6 Three big challenges exist Perception that PM may slow the drug development timeline and raise costs Lack of awareness of full benefits and usage Lack of adequate training infrastructure in the US and abroad Challenges

8 | 7 Key success factors for enhancing adoption of PM Educate professionals around cost and speed implications of PM, recognizing that in some cases it may take longer time in Phase 2 Increase awareness of PM benefits with clinical development executives and scientists and widely share successes Align structurally PM in large organizations to reflect multi- disciplinary nature and clinical decision-making implications Enhance number of trained professionals Develop standardized analysis and reporting

9 | 8 Estimate effect size Rescue discarding good drug Target patient selection Maximize value of prior data Drug approval Labeling Propose best doses Detailed business cases

10 | 9 Determining appropriate dosing for future trials Cyclosporine trough ng/ml Probability of failure Percent Everolimus 9 ng/ml Everolimus 6 ng/ml Everolimus 3 ng/ml Azathioprine Everolimus 12 ng/ml Cyclosporine trough ng/ml Mean CrCL change from baseline ml/min 300 PM approach and impact Sponsor NDA for everolimus tablets (a prophylaxis of organ rejection following heart transplantation) needed to provide safe and effective dosing regimens that would minimize renal toxicity Key questions included: – Are the effectiveness and renal toxicity clinical outcomes related to drug exposure? – What is a rational dosing regimen that would maximize effectiveness (benefit) and minimize nephrotoxicity (risk)? Analysis conducted by 1 person for <2 months FDA performed simulations based on heart transplantation study to project range of outcomes of altered dosing schemes Determined the appropriate dosing to test in future trials PROPOSE BEST DOSES 1 SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clin Pharmacol Ther, Impact of pharmacometric reviews on new drug approval and labeling decisions--a survey of 31 new drug applications submitted between 2005 and Bhattaram VA, Bonapace C, Chilukuri DM, Duan JZ, Garnett C, Gobburu JV, Jang SH, Kenna L, Lesko LJ, Madabushi R, Men Y, Powell JR, Qiu W, Ramchandani RP, Tornoe CW, Wang Y and Zheng JJ. (reference citation) 81: , copyright 2007 (year of publication)

11 | 10 Enhancing trials through prodrug dosing 0 Response 70 Area under the curve (AUC) mg.hr/L 50 Toxicity 60 Effectiveness Response Weight kg Predicted toxicity – weight-based dosing Predicted effectiveness – weight-based dosing Predicted toxicity – fixed dosing Predicted effectiveness – fixed dosing SOURCE: Wang Y, Bhattaram AV, Jadhav PR, Lesko LJ, Madabushi R, Powell JR, Qiu W, Sun H, Yim DS, Zheng JJ and Gobburu JV (2008) Leveraging prior quantitative knowledge to guide drug development decisions and regulatory science recommendations: impact of FDA pharmacometrics during J Clin Pharmacol 48: PM approach and impact Sponsor developing a prodrug (test) for the treatment of a life threatening disease in which a noninferiority comparison to the parent (reference) drug failed to establish effectiveness Proposed a fixed dosing strategy for patent drug (as approved) with key question: – What is the appropriate dosing regimen (fixed vs. per kg dosing) for noninferiority comparison of the prodrug and parent drugs? Analysis conducted by 1 person for <2-4 weeks FDA analyzed data from failed test drug trials to develop the parent drug AUC, effectiveness, and toxicity relationships Proved dosing regimen for prodrug (mg vs. mg/Kg), which enhanced future trial to support an approval PROPOSE BEST DOSES

12 | 11 Recommending dosage warning label Probability of quitting at weeks 9-12 or nausea after treatment Percent AUC Ng/ml PM approach and impact Sponsor conducted five dose- finding and registration clinical trials to determine the appropriate dosing Key questions included: – What is the optimal dose? – Is there a need for dose adjustment in subjects with renal impairment? If so, by how much? Sponsor conducted population pharmokinetics analysis, model simulations and exposure-response analysis to determine basis for discussing the dose selection Led to recommendation of dosage warning to be included in current label PROPOSE BEST DOSES SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clin Pharmacol Ther, Impact of pharmacometric reviews on new drug approval and labeling decisions--a survey of 31 new drug applications submitted between 2005 and Bhattaram VA, Bonapace C, Chilukuri DM, Duan JZ, Garnett C, Gobburu JV, Jang SH, Kenna L, Lesko LJ, Madabushi R, Men Y, Powell JR, Qiu W, Ramchandani RP, Tornoe CW, Wang Y and Zheng JJ. (reference citation) 81: , copyright 2007 (year of publication)

13 | 12 Clarifying dosing and regulatory expectations NOTE: Only dose-finding studies shown Mar 02 CS02/N = 129, 6 mo CS07/N = 172, SD CS06/N = 82, SD Activity CS21 dose/ regimen not finalized EOP2A meeting CS14/N = 127, 12 mo CS12/N = 187, 12 mo Mar 05Mar 04Mar 06Mar 07Mar 03 CS21/N = 610 NDA submission – Feb 29, 2008; approval – Dec 24, 2008 Registration trial PM approach and impact Sponsor needed to determine the dosing for a drug 7 years in development for advanced prostate cancer patients Key questions were: – Is a loading dose needed to suppress testosterone, and, if so how much? – Is a maintenance dose and suppression regimen needed? Sponsor developed a mechanistic data model to explore dosing strategies via trial simulations Identified alternative dosing strategies and clarified regulatory expectations that led to approval SOURCE: FDA, Drug Approval Package, Degeralix Injection, Ferring Pharmaceuticals, December 24, 2008 PROPOSE BEST DOSES

14 | 13 Simulating dose to justify Phase 3 design PM approach and impact Tacrolimus, a potent approved immunosuppressant, had a promising expansion as an oral therapy for ulcerative colitis (UC) Highly variable PK between patients and high trough concentrations in Phase II studies presented challenges to further development Logistic analysis of PK/response in late Phase II demonstrated trough concentration is a good predictor for response Simulation of dose titration based on PK/response was effective for attaining target trough concentration, demonstrating efficacy and justifying Phase III studies SOURCE: Presented by Atsunori Kaibara (Astellas Pharma, Inc) at the PK/PD Internal Symposium on Modeling and Simulation in Drug Development and Clinical Applications", Yonsei University Medical Center, Seoul, Korea (2006) High trough concentrations in Phase II studies Simulated dose titration shows eliminated high trough PROPOSE BEST DOSES 2

15 | 14 Promoting innovative trial designs to determine effective and safe dosing SOURCE: Wang Y, Bhattaram AV, Jadhav PR, Lesko LJ, Madabushi R, Powell JR, Qiu W, Sun H, Yim DS, Zheng JJ and Gobburu JV (2008) Leveraging prior quantitative knowledge to guide drug development decisions and regulatory science recommendations: impact of FDA pharmacometrics during J Clin Pharmacol 48: Biomarker Genotype PM approach and impact Sponsor developing a new compound to treat type 2 diabetes Key questions include: – Is this a once-a-day drug? – Is dosing adjusted to match exposures reasonable to characterize the effectiveness and safety of the drug? Built a semimechanistic model, using parameters derived from FDAs prior experience with 26-wk trials, to describe the time course and drug concentrations Provided the FDA team with the ability to discuss innovative trial designs and the sponsor with a model to apply to other similar drugs Response rate Percent BIDQD PROPOSE BEST DOSES

16 | 15 Accelerating drug development through a model-based approach to understanding dosing PM approach and impact Sponsor needed to select a range of active oral doses for clinical phase IIa trials of a novel anti-HIV drug, maraviroc PK/PD model to link plasma concentration to inhibition of viral replication was adapted for short- term treatment based on disease model parameters from literature and preclinical data The model predicted the effect on viral load of different doses, resulting in a range of possible active oral doses used in the design of phase IIa trials SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clin Pharmacol Ther, Rosario MC, Jacqmin P, Dorr P, van der Ryst E, Hitchcock C. A pharmacokinetic-pharmacodynamic disease model to predict in vivo antiviral activity of maraviroc, (reference citation), 2005 Nov;78(5):508-19, copyright year of publication) Solid lines: Measured Dotted lines: Simulated Bold: 100 mg 2x / day Narrow: 25 mg 1x / day Time, days Diff log (BSL)-log (Day 6), copies/ml IC50 = 5.75 ng/ml IC50 = 0.64 ng/ml Simulated and observed viral load for 2 dosage regimens PROPOSE BEST DOSES 3

17 | 16 Promoting collaboration and more informed decisions to determine dosing SOURCE: Wang Y, Bhattaram AV, Jadhav PR, Lesko LJ, Madabushi R, Powell JR, Qiu W, Sun H, Yim DS, Zheng JJ and Gobburu JV (2008) Leveraging prior quantitative knowledge to guide drug development decisions and regulatory science recommendations: impact of FDA pharmacometrics during J Clin Pharmacol 48: PM approach and impact Chance of being the preferred dose Percent % Time Week 10mg BID 20 mg BID 40 mg QD Sponsor seeking appropriate dosing and trial design for a new class of antiviral compound in conjunction with another approved drug Key questions included: – Are the proposed dose and dosing regimen reasonable for the phase 2b trial? – Is dose selection based on the first 1-mo data reliable? Applied a mechanistic viral–dynamic model, including prior knowledge of the approved drug combination, to describe the time course of viral load reduction driven by concentrations Incorporating prior knowledge to build a model that leads to more informed decisions and FDA/sponsor collaboration PROPOSE BEST DOSES

18 | 17 Developing models that accurately identify dosing PM approach and impact Sponsor needed to determine which dose of a pain compound would be needed to achieve superiority to 400 mg ibuprofen in a post-oral surgery model PK/PD models were developed relating plasma concentrations to pain relief scores Clinical trial simulations conducted to recommend a dose of 360 mg Dose of 360 mg was corroborated with placebo- and positive-controlled study evaluating various doses and 400 mg ibuprofen SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clinical Pharmacology & Therapeutics, Kowlaski, KG, Olson S, Remmers AE, Hutmacher MM; Modeling and Simulation to Support Dose Selection and Clinical Development of SC-75416, a Selective COX-2 Inhibitor for the Treatment of Acute and Chronic Pain, (reference citation), June 2008 copyright (year of publication) Workflow for the development of the pain relief and dropout models Oral solution predictions and clinical trial simulations to support the design of the post–oral surgery pain study Model IA Pain relief Model IB Dropout Model IA Pain relief Model IB Dropout Oral solution PK CTS results Study results Model IIA/IIB TOTPAR6 predictions Model IA/IB PR prediction Model IIA/IIB Goodness of fit ModelIA/IB Goodness of fit SC capsule Valdecoxib SC Oral solution PROPOSE BEST DOSES

19 | 18 Determining best dosing to avoid failure PM approach and impact Sponsor seeking approval for an oral suspension product Key question was: What is the optimal dosing strategy to avoid clinical failure in the majority of patients? FDA conducted exposure- response analysis to compliment the sponsors pre- specified statistical analysis Enhanced the label and decided to conduct a Phase IV study to evaluate therapeutic advantages of monitoring and adjusting the dosing PROPOSE BEST DOSES 0 Steady-state concentration Ng/ml Patients with clinical failure Percent SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clin Pharmacol Ther, Impact of pharmacometric reviews on new drug approval and labeling decisions--a survey of 31 new drug applications submitted between 2005 and Bhattaram VA, Bonapace C, Chilukuri DM, Duan JZ, Garnett C, Gobburu JV, Jang SH, Kenna L, Lesko LJ, Madabushi R, Men Y, Powell JR, Qiu W, Ramchandani RP, Tornoe CW, Wang Y and Zheng JJ. (reference citation) 81: , copyright 2007 (year of publication)

20 | 19 Determining age-appropriate dosage for labeling SOURCE: Bhattaram VA, Booth BP, Ramchandani RP, Beasley BN, Wang Y, Tandon V, Duan JZ, Baweja RK, Marroum PJ, Uppoor RS, Rahman NA, Sahajwalla CG, Powell JR, Mehta MU and Gobburu JV (2005) Impact of pharmacometrics on drug approval and labeling decisions: a survey of 42 new drug applications. AAPS J 7:E503-E512. Age Months Age factor Age factor = for age >24 months PM approach and impact Sponsor seeking approval to use an adult tachycardia drug (already on the market) for pediatrics Key question was: is the pediatric dosing regimen proposed by the sponsor acceptable? FDA used data from two clinical trials and modified the sponsors exposure-response model to determine dosing regimen for neonates and infants, the subset of pediatrics in question Approved dosage based on outcome of age analysis and incorporated pediatrics-specific dosing into the label PROPOSE BEST DOSES

21 | 20 Rescue discarding good drug Target patient selection Maximize value of prior data Drug approval Labeling Detailed business cases Estimate effect size Propose best doses

22 | 21 Advancing decisions based on dual dose range Insomnia patients LPS % change from mean placebo response Healthy volunteers LPS, % change from mean placebo response Insomnia patients WASO % change from mean placebo response Healthy volunteers WASO % change from mean placebo response Y = 0.31x r 2 = 0.66 SOURCE: Wang Y, Bhattaram AV, Jadhav PR, Lesko LJ, Madabushi R, Powell JR, Qiu W, Sun H, Yim DS, Zheng JJ and Gobburu JV (2008) Leveraging prior quantitative knowledge to guide drug development decisions and regulatory science recommendations: impact of FDA pharmacometrics during J Clin Pharmacol 48: Sponsor developing a drug to treat insomnia held an end-of-phase 2a meeting (EOP2A) Key questions discussed were: – Is the dose range selected for the Phase 2b studies in insomnia patients reasonable? – What should be the duration of the Phase 2b studies? Analysis conducted by 1 person for <2-4 weeks FDA used data from 14 studies of internal FDA submissions and insomnia drug literature Proved: – Effective use of dual dose range (correlation between healthy subjects and insomnia patients) – Shorter trials could produce results PM approach and impact ESTIMATE EFFECT SIZE 4

23 | 22 Simulating clinical trials for successful design PM approach and impact Sponsor needed to determine if a new treatment was effective for Alzheimers disease (AD) A crossover design was evaluated to establish proof-of-concept using smaller and shorter duration trials Clinical trial simulation model was built from Phase I data and literature reports including PK/PD and disease progression of other AD treatments Eight alternative trial designs were simulated to determine trial design Trial resulted in more efficient trial design and a conclusive decision for further development SOURCE: Lockwood, Peter; Ewy, Wayne; Hermann, David; and Holford Nick; Application of Clinical Trial Simulation to Compare Proof-of-Concept Study Designs for Drugs with a Slow Onset of Effect; An Example in Alzheimers Disease, Pharmaceutical Research, 2006 ESTIMATE EFFECT SIZE

24 | 23 Determining precision in dose prediction PM approach and impact Sponsor evaluated pregabalin for pain treatment by determining how precisely pain reduction could be predicted and demonstrated PK/PD model relating pain relief to gabapentin plasma concentrations was derived from a phase 3 study and further modified to reflect pregabalin preclinical data Simulation data suggested that doses that identify predefined response may be imprecisely estimated Quantification of imprecision will drive phase 2 dose and trial design SOURCE: Lockwood, Peter A; Cook, Jack A; Ewy, Wayne E; and Mandema, Jaap W;The Use of Clinical Trial Simulation to Support Dose Selection: Application to Development of a New Treatment for Chronic Neuropathic Pain, Pharmaceutical Research, November 2003 Model 3 Drug effect (fraction of baseline value) Model 2 Model Drug effect models based only on data within this concen- tration range Gabapentin concentration Ug/ml Concentration–response profiles for efficacy ESTIMATE EFFECT SIZE

25 | 24 Mean pain score Time Days Mean pain score Time Days Placebo (predicted) 2,400 mg daily (predicted) 1,200 mg daily (predicted) 600 mg daily (predicted) 2,400 mg daily (observed) 1,200 mg daily (observed) 600 mg daily (observed) Placebo (observed) Conducting exposure-response analysis to prevent the need for additional trials 3,600 mg daily (observed) 3,600 mg daily (predicted) Mean pain score Time Days Mean pain score Time Days Time Days Mean pain score 1,800 mg daily (predicted) 1,800 mg daily (observed) PM approach and impact Clinical trial data of gabapetin for postherpetic neuralgia was comprised of single replicates of various doses Exposure-response analysis was used to avoid additional replicate trials for the representative doses A random-effects model was applied to the submitted studies, demonstrating predictability and establishing exposure-dependent decrease in pain scores and cross-confirmation of trials Regulatory approval was granted without the need for additional clinical trials SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clinical Pharmacology & Therapeutics, Lalonde RL, Kowalski KG, Hutmacher MM, Ew W, Nichols DJ, Milligan PA, Corrigan BW, Lockwood PA, Marshall SA, Benincosa LJ, Tensfeldt TG, Parivar K, Amantea M, Glue P, Koide H and Miller R; Model-based Drug Development, (reference citation), July 2007 copyright (year of publication) ESTIMATE EFFECT SIZE

26 | 25 Using quantitative decision criteria to understand when to terminate development of drug PM approach and impact Gemcabene, a cholesterol-lowering drug, was evaluated for market potential versus existing therapy using quantitative decision criteria based on estimated effect and distribution of trial results A lower reference value (LRV) of competitive efficacy and target value (TV) of estimated commercial viability were used for risk criteria PK / PD and disease model data were used to simulate trials using the uncertainty in treatment effect Both probability of success (a go decision) and probability of a correct decision were used to evaluate trial performance metrics, leading to a decision to terminate development Go Pause Stop Data-driven decision Total Stop (fail) PCT 20LRV or PCT 90TV Go (success) PCT 20 >LRV and PCT 90 >TV Truth (desired decision) Prob(>TV)Prob(Stop and >TV) Prob(Go and >TV) >TV (Go) Prob(TV)Prob(Stop and TV) Prob(Go and TV) TV (Stop) 1.0Prob(Stop)Prob(Go)Total Example of a decision rule based on dual criteria Example of a design performance summary ESTIMATE EFFECT SIZE SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clinical Pharmacology & Therapeutics, Lalonde RL, Kowalski KG, Hutmacher MM, Ew W, Nichols DJ, Milligan PA, Corrigan BW, Lockwood PA, Marshall SA, Benincosa LJ, Tensfeldt TG, Parivar K, Amantea M, Glue P, Koide H and Miller R; Model-based Drug Development, (reference citation), July 2007 copyright (year of publication)

27 | 26 Demonstrating drug dosages to allow early trial discontinuation PM approach and impact Proof-of-concept trial of acute stroke therapy used an early termination rule to end trial at the earliest point of establishing futility or efficacy An adaptive dose allocation and sample size scheme was continually updated to allocate patients to placebo or adaptively to best model the minimal dose providing maximal treatment effect Termination was based on standard deviation of response Dose-response relationship was flat, showing that only 3 of the 15 doses were needed to demonstrate efficacy over placebo, allowing early trial discontinuation E F Evaluable population Effect over placebo Dose of UK-279,276 (mg) The horizontal line at 0 indicates the line of no change, at 1 the futility threshold (F), and at 2 the efficacy threshold (E) Dose–effect curve of effect over placebo ESTIMATE EFFECT SIZE SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clinical Pharmacology & Therapeutics, Lalonde RL, Kowalski KG, Hutmacher MM, Ew W, Nichols DJ, Milligan PA, Corrigan BW, Lockwood PA, Marshall SA, Benincosa LJ, Tensfeldt TG, Parivar K, Amantea M, Glue P, Koide H and Miller R; Model-based Drug Development, (reference citation), July 2007 copyright (year of publication)

28 | 27 Using quantitative data to understand when to discontinue development PM approach and impact Sponsor wanted to compare two potential insomnia compounds Clinical Utility Index (CUI) was developed using various measures of residual sedation and efficacy Dose-response analyses were conducted on each end point, using sensitivity analysis to determine the degree of clinical meaning in CUI Peak CUI values were observed at doses not considered viable, leading to an expedited and quantitative decision to discontinue development SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clinical Pharmacology & Therapeutics, Ouellet D, Werth J, Parekh N, Feltner D, McCarthy B and Lalonde, RL; The Use of a Clinical Utility Index to Compare Insomnia Compounds: A Quantitative Basis for Benefit– Risk Assessment (reference citation), March 2009 copyright (year of publication) Line: median CUI Shaded: 80% conf. Symbols: observed Probability of observing a difference at peak clinical utility index (CUI) value Difference in CUI Clinical utility index for lead and backup compounds. CUI Lead compoundBack-up compound Dose, mg ESTIMATE EFFECT SIZE

29 | 28 Evaluating competitive potential of a drug to decide development PM approach and impact Sponsor needed to evaluate the efficacy of a gemcabene, novel cholesterol-lowering drug Dose response model was developed from publicly available data and proprietary patient data, using the Pharsight Drug Model Explorer visualization technology Unlike competitor ezitimibe, gemcabene was found to have little additional LDL-C lowering in combination with high statin doses Use of the model after the first phase II trial facilitated a quick decision to stop development SOURCE: Mandema, Jaap W; Hermann, David; Wang, Wenping; Sheiner, Tim; Milad, Mark; Bakker-Arkema, Rebecca; and Hartman, Daniel; Model-based development of Gemcabene, AAPS Journal, 2005 Dose, mg LDL % change from baseline With placebo (A 0) 10 mg atorvastatin (A10) 40 mg atorvastatin (A 40) 80 mg atrovastatin (A 80) Atorvastatin dose, mg LDL % change from baseline GemcabeneEzitimibe Atorvastatin alone (top) With agent (bottom) Dose-response relationship for gemcabene when combined with placebo Dose-response relationship of atorvastatin monotherapy and in combination with gemcabene and ezitimibe ESTIMATE EFFECT SIZE

30 | 29 Propose best doses Estimate effect size Rescue discarding good drug Target patient selection Maximize value of prior data Drug approval Labeling Rescue discarding good drug Detailed business cases

31 | 30 Revising dosing strategy for quicker approval Sponsor seeking approval for a drug to treat acute decompensated congestive heart failure (CHF) Key question: what is the optimal dosing regimen of nesiritide to achieve faster benefits and minimize risk (i.e., undesired hypotension)? Developed model based on exposure and rate data from original submission to explore alternative dosing strategies Sponsor revised dosing strategy based on PM results and received quicker approval PM approach and impact SOURCE: Bhattaram VA, Booth BP, Ramchandani RP, Beasley BN, Wang Y, Tandon V, Duan JZ, Baweja RK, Marroum PJ, Uppoor RS, Rahman NA, Sahajwalla CG, Powell JR, Mehta MU and Gobburu JV (2005) Impact of pharmacometrics on drug approval and labeling decisions: a survey of 42 new drug applications. AAPS J 7:E503-E512. RESCUE DISCARDING GOOD DRUG Nesiritide plasma concentrations ug/l Time Hours Placebo-corrected hemodynamics mmHg Systolic BP PCWP Time course of nesiritide plasma concentrations Indicates predicted Indicates observed ` Typical time course of nesiritide plasma concentrations and the effects on the pulmonary capillary wedge pressure after a 2 μg/kg bolus followed by a fixed-dose infusion of 0.01 μg/kg per minute (data for the initial 3 hours)

32 | 31 Incorporating dosing strategies into labeling Sponsor seeking approval for a drug to treat hypercalcemia of malignancy and osteolytic bone metastases Key questions included: – Is there a need to adjust dosing in patients with renal impairment? – If so, what doses should be recommended? Early phase PK studies were used to develop a population model leading to FDA recommendation for dose adjustment in mild and moderate renal impairment patients Incorporated FDAs dosing strategies into label PM approach and impact Creatinine clearance ml/min Risk of renal deterioration Percent Zoledronic acid Placebo Observed ` Predicted SOURCE: Bhattaram VA, Booth BP, Ramchandani RP, Beasley BN, Wang Y, Tandon V, Duan JZ, Baweja RK, Marroum PJ, Uppoor RS, Rahman NA, Sahajwalla CG, Powell JR, Mehta MU and Gobburu JV (2005) Impact of pharmacometrics on drug approval and labeling decisions: a survey of 42 new drug applications. AAPS J 7:E503-E512. RESCUE DISCARDING GOOD DRUG Risk of renal deterioration increases with decreasing renal function (assessed based on baseline creatinine clearance) following 4 mg infusion of zoledronic acid over 15 min and placebo in solid tumor and prostate cancer patients

33 | 32 Eliminating need for additional trials Week Symptom score PM approach and impact Sponsor had inconclusive results from two registration trials in patients with a debilitating neurological disorder without approved treatments Key question was: is there adequate evidence of effectiveness in the current clinical trial database? Analysis conducted by 1 person for <2-4 weeks FDA analyzed data across studies to investigate whether there was a consistent effectiveness signal Based on results and need to supply treatment for this disease, additional clinical trials were alleviated RESCUE DISCARDING GOOD DRUG SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clin Pharmacol Ther, Impact of pharmacometric reviews on new drug approval and labeling decisions--a survey of 31 new drug applications submitted between 2005 and Bhattaram VA, Bonapace C, Chilukuri DM, Duan JZ, Garnett C, Gobburu JV, Jang SH, Kenna L, Lesko LJ, Madabushi R, Men Y, Powell JR, Qiu W, Ramchandani RP, Tornoe CW, Wang Y and Zheng JJ. (reference citation) 81: , copyright 2007 (year of publication)

34 | 33 Specifying additional clinical trial needs to avoid trial failure SOURCE: Bhattaram VA, Booth BP, Ramchandani RP, Beasley BN, Wang Y, Tandon V, Duan JZ, Baweja RK, Marroum PJ, Uppoor RS, Rahman NA, Sahajwalla CG, Powell JR, Mehta MU and Gobburu JV (2005) Impact of pharmacometrics on drug approval and labeling decisions: a survey of 42 new drug applications. AAPS J 7:E503-E512. Sponsor seeking approval for a drug for patients with an unmet life threatening rheumatologic disorder was unable to demonstrate additional evidence of effectiveness after two trials Key questions included: – Is the laboratory concentration predictive of the clinical outcome? – What dose should be approved? Expanded analysis by simulating the estimated reduction required to achieve a clinical benefit resulting in a dose-response relationship Recommended to explore the maximally tolerated dose or a dose selected to achieve a greater reduction in this laboratory value PM approach and impact Series Change in biomarker Percent Relative risk of primary end point RESCUE DISCARDING GOOD DRUG Relationship between the relative risk of the clinical event and the percent change in the biomarker (laboratory concentrations). 95% confidence limits Best fit

35 | 34 Target patient selection Rescue discarding good drug Maximize value of prior data Drug approval Labeling Propose best doses Estimate effect size Detailed business cases

36 | 35 Selecting patient population Placebo-subtracted change in score A at Week 12 Dose mg Placebo-subtracted change in score A at Week 12 Dose mg PM approach and impact Sponsor proposed a drug to treat patients with mild, moderate or severe life- threatening disease with inconsistent results during three trials Key questions included: – What is the reason for the inconsistent results across the three registration trials? – How can the success rate of future trials be improved? Analysis conducted by 1 person for <2-4 weeks FDA developed a new exposure-response model based on analysis across the three studies Led to suggestion to conduct future study in patient segmentation (moderate and severe disease patients only) TARGET PATIENT SELECTION 6 SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clin Pharmacol Ther, Impact of pharmacometric reviews on new drug approval and labeling decisions--a survey of 31 new drug applications submitted between 2005 and Bhattaram VA, Bonapace C, Chilukuri DM, Duan JZ, Garnett C, Gobburu JV, Jang SH, Kenna L, Lesko LJ, Madabushi R, Men Y, Powell JR, Qiu W, Ramchandani RP, Tornoe CW, Wang Y and Zheng JJ. (reference citation) 81: , copyright 2007 (year of publication)

37 | 36 Saving time and allowing rapid development decisions by targeting patients PM approach and impact Early development decisions were needed for Alzheimers Disease compound to evaluate efficacy and dose-response versus competitor Original 12-week trial design had 6 parallel groups, 5 dose levels, and patients per group PK/PD and disease data were simulated in various dose-response relationships, using the slowest time patterns to evaluate crossover The most robust design demonstrated that while hundreds of patients were still needed, shorter trials were sufficient, resulting in time savings Negative results allowed rapid termination of development Effect size at 25mg = -3 Effect/dose slope = Effect/conc slope = ED50 ~ 8 mg EC50 ~ 21 ng/mL Hill coefficient = 4 Linear dose response model CI-1017 tid dose (mg) Hyperbolic (E max ) dose response model CI-1017 tid dose (mg) Sigmoidal (Smax) dose response model CI-1017 tid dose (mg) U-shaped dose response model CI-1017 tid dose (mg) CI-1017 average steady state plasma concentration (ng/mL) Effect size at 25 mg = 100% Emax = -3 ED50 ~8.5 mg EC50 ~21 ng/ml ½ maximal effect = -2 Effect size at 25 mg tid = -3 = 75% Emax Change in ADAS-cog score Overall effect size at 10mg tid = -3 ED50 Emax ~ 7 mgED50 Imax ~ 15 mg Combined agonist-antagonist dose/response Agonist (Emax) dose/response Antagonist (Imax) dose/response TARGET PATIENT SELECTION SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clinical Pharmacology & Therapeutics, Lalonde RL, Kowalski KG, Hutmacher MM, Ew W, Nichols DJ, Milligan PA, Corrigan BW, Lockwood PA, Marshall SA, Benincosa LJ, Tensfeldt TG, Parivar K, Amantea M, Glue P, Koide H and Miller R; Model-based Drug Development, (reference citation), July 2007 copyright (year of publication)

38 | 37 Target patient selection Maximize value of prior data Rescue discarding good drug Propose best doses Estimate effect size Drug approval Labeling Detailed business cases

39 | 38 Using data to enhance trial design when expanding drug usage PM approach and impact Sponsor seeking approval to use an adult seizure drug (already on the market) for pediatrics Key questions included: – Is there adequate evidence for approving drug in pediatrics without the need for additional controlled clinical trials? – What are the appropriate dosing instructions for this indication? FDA used originally submitted data to build an exposure-response model for quantitative analysis to: – Test whether placebo responses in adult and pediatric patients were similar – Test whether exposure-response relationship in the two populations were similar – Derive reasonable dosing recommendations in pediatrics Proved dosing recommendations in pediatrics matched adults providing an approval without additional trials SOURCE: Bhattaram VA, Booth BP, Ramchandani RP, Beasley BN, Wang Y, Tandon V, Duan JZ, Baweja RK, Marroum PJ, Uppoor RS, Rahman NA, Sahajwalla CG, Powell JR, Mehta MU and Gobburu JV (2005) Impact of pharmacometrics on drug approval and labeling decisions: a survey of 42 new drug applications. AAPS J 7:E503-E512. MAXIMIZE VALUE OF PRIOR DATA 7

40 | 39 Target patient selection Rescue discarding good drug Propose best doses Estimate effect size Maximize value of prior data Labeling Detailed business cases Drug approval

41 | 40 Identifying treatment dosage and potential harmful effects PM approach and impact Number of patients, alkaline phosphatase 3x ULN Proportion of patients with favorable endoscopic response Percent Dose mg Series Sponsor needed further analysis for an antifungal drug Key question is: what is the appropriate dose of micafungin for the treatment? FDA reviewer used data from two studies to model the relationship between dose and effectiveness Dose–response analysis recommended the treatment dose needed for approval as well as a package insert indicating harmful effects with greater dosage DRUG APPROVAL 8 SOURCE: Reprinted by permission from Macmillan Publishers Ltd: Clin Pharmacol Ther, Impact of pharmacometric reviews on new drug approval and labeling decisions--a survey of 31 new drug applications submitted between 2005 and Bhattaram VA, Bonapace C, Chilukuri DM, Duan JZ, Garnett C, Gobburu JV, Jang SH, Kenna L, Lesko LJ, Madabushi R, Men Y, Powell JR, Qiu W, Ramchandani RP, Tornoe CW, Wang Y and Zheng JJ. (reference citation) 81: , copyright 2007 (year of publication)

42 | 41 Determining dosing for drug approval PM approach and impact Sponsor conducted 3 trials in chronic kidney disease Stage 5 patients with hyperparathyroidism for an oral paricalcitol drug Used a validated exposure-response model to define new dosing regimen that reduces the rate of hypercalcemia while maintaining acceptable therapeutic response Results of trial confirmed predicted results from exposure-response modeling and simulations, maintaining a pre-defined efficacy rate of over 80% and limiting observed rate of hypercalcemia Drug approved by FDA without need for further clinical trials in patients SOURCE: Zemplar Capsules United States Package Information, June Accessed at 18March Parameter Model simulated Observed Efficacy (2 consecutive serum iPTH measurements decreased by 30% from baseline) Safety (2 consecutive serum calcium measurements over 11 mg/dL; hypercalcemia) % of subjects DRUG APPROVAL 9

43 | 42 Target patient selection Rescue discarding good drug Propose best doses Estimate effect size Maximize value of prior data Drug approval Detailed business cases Labeling

44 | 43 Incorporating dosing recommendations into labeling PM approach and impact Sponsor sought approval for Parkinsons disease drug for acute use in patients Key questions included: – Is the maximum recommended dose and the titration strategy proposed by the sponsor appropriate? – Is there a need for adjusting dose in the renally impaired? Used data from the dose-finding strategy for simulations using an exposure- response model Dosing recommendations suggested by the exposure-response analysis were incorporated in the labeling after discussions with the sponsor SOURCE: Bhattaram VA, Booth BP, Ramchandani RP, Beasley BN, Wang Y, Tandon V, Duan JZ, Baweja RK, Marroum PJ, Uppoor RS, Rahman NA, Sahajwalla CG, Powell JR, Mehta MU and Gobburu JV (2005) Impact of pharmacometrics on drug approval and labeling decisions: a survey of 42 new drug applications. AAPS J 7:E503-E512. LABELING

45 | 44 Improving dosing regimen to incorporate into label PM approach and impact Drug evaluated in the pre-, during and post- surgical settings for patients with mild to moderate hypertension Cleviprex dosing regimen used in clinical trials resulted in overshooting and oscillations around the target blood pressure PK/PD modelling used to simulate the exposure–response relationship to optimize the dosing regimen to quickly achieve and maintain target blood pressure Resulted in a safer and more effective dosing regimen than employed in clinical trials, which was favored to be incorporated into the label SOURCE: Drug Approval Package. Cleviprex® (clevidipine butyrate) Application No b. 10 LABELING


Download ppt "Pharmacometrics: A Business Case May 25, 2010 Pharmacometrics Task Force."

Similar presentations


Ads by Google