Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Faith In Holistic modelling for First-In-Human PK

379 views

Published on

Presented by Neil Miller

Published in: Health & Medicine
  • Be the first to comment

  • Be the first to like this

Faith In Holistic modelling for First-In-Human PK

  1. 1. F.I.H 4 F.I.H Paris PBPK SYMPOSIUM 2019 Neil Miller 4th April 2019 Faith In Holistic modelling for First-In-Human PK
  2. 2. • Have faith in using PBPK modelling for predicting First-In-Human (FIH) pharmacokinetics • Use everything that we/you know by connecting the parts • See failures as opportunities to learn as “In a complex world, failure is inevitable.” Matthew Syed author of Black Box Thinking: The Surprising Truth About Success Summary Take home messages 2 • Scene setting • Components • Case studies • Summary Neil Miller April 2019 FIH 4 FIH.pptx
  3. 3. • Scene setting  Belief, parts and interconnection • The parts: optimising components of a PBPK prediction  Flow diagrams/decision trees • Case studies: connecting the parts  Applications • Summary Content Faith In Holistic modelling for First-In-Human PK (for PO dosing) 3Disclaimer: The views expressed in this presentation are those of the presenter and are not those of GlaxoSmithKline Neil Miller April 2019 FIH 4 FIH.pptx
  4. 4. • Holistic: characterized by the belief that the parts of something are intimately interconnected and explicable only by reference to the whole Scene setting Definition of Holistic 4 Neil Miller April 2019 FIH 4 FIH.pptx Proof that the approach works Components: 1. Oral Absorption 2. Distribution 3. Metabolism & Elimination Platform for integration/interconnecting of the parts = PBPK • Scene setting • Components • Case studies • Summary
  5. 5. Scene setting Proof that PBPK works for F.I.H PK predictions 5 Neil Miller April 2019 FIH 4 FIH.pptx “The simulation results using PBPK were shown to be superior to those obtained via traditional one compartment analyses. In many cases, this difference was statistically significant.” “Our prospective human PK prediction methods yielded good prediction results.” • Scene setting • Components • Case studies • Summary
  6. 6. Scene setting Proof that PBPK works for F.I.H PK predictions 6 Neil Miller April 2019 FIH 4 FIH.pptx “In the majority of cases, PBPK gave more accurate predictions of pharmacokinetic parameters and plasma concentration-time profiles than the Dedrick approach.” “This evaluation demonstrates that PBPK models can lead to reasonable predictions of human pharmacokinetics.” • Scene setting • Components • Case studies • Summary
  7. 7. Scene setting Parts of a PBPK F.I.H PK prediction 7 Neil Miller April 2019 FIH 4 FIH.pptx • Oral Absorption: Examining the multifactorial process driving the fraction of drug that reaches the systemic circulation is critical • Distribution: Understanding tissue distribution is essential for PBPK modelling, and has been facilitated by mechanistic equations • Metabolism and Elimination: PBPK modelling requires quantitative understanding of the main mechanism(s) of drug clearance • Scene setting • Components • Case studies • Summary
  8. 8. Scene setting PBPK = Platform for integration of the parts 8 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary
  9. 9. Components PBPK Modelling for FIH Predictions 9 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary
  10. 10. Components Oral Absorption 10 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary ASF absorption scale factors, BSSR bile salt solubilisation ratio, MPT mean precipitation time a Other processes-transporters: efflux transporters can be incorporated in GastroPlus models with a simple method (e.g. adjusting permeability based on preclinical observations or in-vitro data) to more complex methods (e.g. specifically incorporating effects of transporters) 1Sutton SC. Role of physiological intestinal water in oral absorption. AAPS J. 2009;11(2):277–85. 2Kesisoglou F. Use of preclinical dog studies and absorption modelling to facilitate late stage formulation bridging for a BCS II drug candidate. AAPS PharmSciTech. 2014;15(1):20–8.
  11. 11. Components Distribution 11 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary BPR blood/plasma ratio, Fup fraction unbound in plasma, Kp tissue-to-plasma partition coefficient, SpecPStc specific in-vivo diffusional clearance per millilitre of tissue cell volume 1Lukacova V, Parrott N, Lavé T, Fraczkiewicz G, Bolger M, Woltosz W. General approach to calculation of tissue:plasma partition coefficients for physiologically based pharmacokinetic (PBPK) modeling. AAPS National Annual Meeting and Exposition; 16–20 Nov 2008; Atlanta (GA). 2Xia B, Heimbach T, Lin TH, He H, Wang Y, Tan E. Novel physiologically based pharmacokinetic modeling of patupilone forhuman pharmacokinetic predictions. Cancer Chemother Pharmacol. 2012;69(6):1567–82. 3De Buck SS, Sinha VK, Fenu LA, Nijsen MJ, Mackie CE, Gilissen RA. Prediction of human pharmacokinetics using physiologically based modeling: a retrospective analysis of 26 clinically tested drugs. Drug Metab Dispos. 2007;35(10):1766–80. 4Samant TS, Lukacova V, Schmidt S. Development and qualification of physiologically based pharmacokinetic models for drugs with atypical distribution behavior: a desipramine case study. CPT Pharmacometrics Syst Pharmacol. 2017;6(5):315–21.
  12. 12. Components Metabolism and Elimination 12 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary CLR renal clearance, CLR,u unbound renal clearance, ECCS Extended Clearance Classification System, Fup fraction unbound in plasma, GFR glomerular filtration rate, IVIVE in-vitro in-vivo extrapolation 1Mathialagan S, Piotrowski MA, Tess DA, Feng B, Litchfield J, Varma MV. Quantitative prediction of human renal clearance and drug-drug interactions of organic anion transporter substrates using in vitro transport data: a relative activity factor approach. Drug Metab Dispos. 2017;45(4):409–17. 2Scotcher D, Jones C, Rostami-Hodjegan A, Galetin A. Novel minimal physiologically-based model for the prediction of passive tubular reabsorption and renal excretion clearance. Eur J Pharm Sci. 2016;94:59–71. 3Kimoto E, Bi YA, Kosa RE, Tremaine LM, Varma MVS. Hepatobiliary clearance prediction: species scaling from monkey, dog, and rat, and in vitro-in vivo extrapolation of sandwich-cultured human hepatocytes using 17 drugs. J Pharm Sci. 2017;106(9):2795–804.
  13. 13. Case studies: connecting the parts Applications of the flow diagrams 13 Neil Miller April 2019 FIH 4 FIH.pptx • Looking at real-world examples of how PBPK was applied for FIH PK • Compounds with challenging properties • Accuracy of predictions often under the spotlight, but support to internal decision making and opportunities to learn must not be over looked • Scene setting • Components • Case studies • Summary
  14. 14. Case studies: Compound 1 Applications of the flow diagrams 14 Neil Miller April 2019 FIH 4 FIH.pptx • Physchem = neutral and highly lipophilic (clogP >5) • Plasma free fraction & aqueous solubility too low for accurate quantification  These properties meant it was challenging to verify an IVIVE • Preclinical in vitro and in vivo pharmacology promising  Extrapolation of pharmacokinetics to estimate a clinical dose was conducted • Scene setting • Components • Case studies • Summary
  15. 15. Case studies: Compound 1 Applications of the flow diagrams 15 Neil Miller April 2019 FIH 4 FIH.pptx • Predicting Oral Absorption was challenging due to low solubility: • Scene setting • Components • Case studies • Summary Solubility was enhanced in biorelevant media Preclinical data showed decreasing bioavailability with increasing dose with an influence of formulation and feeding status The percentage water in small and large intestine compartments was reduced from 40 & 10% to 10 & 0.1% respectively Particle size: 1μm for nano-suspension 80μm for tablets Physiological model parameter changes are not recommended best practice, but there is considerable uncertainty and ongoing debate over the relevant parameterization of intestinal water volumes! Due to the predicted food effect, the first clinical study was conducted in the fed state
  16. 16. Case studies: Compound 1 Applications of the flow diagrams 16 Neil Miller April 2019 FIH 4 FIH.pptx • Distribution: • Scene setting • Components • Case studies • Summary Predicted volumes using the Lukacova method based on the physicochemical properties and assuming a Fup of 0.1% for all species, were in reasonable agreement with the observed data
  17. 17. Case studies: Compound 1 Applications of the flow diagrams 17 Neil Miller April 2019 FIH 4 FIH.pptx • Metabolism & Elimination: • Scene setting • Components • Case studies • Summary Metabolically cleared based on in vitro data Scaling of in vitro data to predict clearance should account for differences in binding in vitro and in vivo, but binding could not be measured. Assuming that in vitro free fraction was equivalent to Fup resulted in a large overprediction of clearance in the rat Human clearance was predicted using an empirical scaling factor derived from the rat which was verified by scaling of hepatocyte intrinsic clearances measured in dog and monkey
  18. 18. Case studies: Compound 1 Evaluating the predictions for 25mg tablet in the fed state 18 Neil Miller April 2019 FIH 4 FIH.pptx • Despite challenging properties, the pharmacokinetics prediction was good: • Scene setting • Components • Case studies • Summary Cmax well-predicted Greater than 2-fold under- prediction of AUC and t1/2 longer than expected, but a later intravenous microdose study confirmed good predictions of bioavailability and Vss Cmax Vss Bioavailability  Decision making: Compound with appropriate pharmacokinetics progressed  Learning opportunity: Systemic clearance at the lower end of the predicted range and consideration of uncertainty in clearance would have avoided a protocol amendment to adjust sampling times
  19. 19. Case studies: Compound 3 Applications of the flow diagrams 19 Neil Miller April 2019 FIH 4 FIH.pptx • Physchem = weak acid and highly lipophilic (clogP = 5) • Low solubility presented a challenge to in vitro assays • Prospective PBPK modelling was conducted to understand the compound’s PK properties and support the decision to move the compound into clinical development • Scene setting • Components • Case studies • Summary
  20. 20. Case studies: Compound 3 Applications of the flow diagrams 20 Neil Miller April 2019 FIH 4 FIH.pptx • Absorption modelling focused on monkey as formulation similar to clinical: • Scene setting • Components • Case studies • Summary FaSSIF solubility critical for absorption modelling in monkey Low solubility in simulated gastric fluid and FaSSIF, but formulation improved solubility Improved solubility and supersaturation of formulation meant permeability was more of a concern in terms of uncertainty Intestinal metabolism had to be considered because it is a CYP3A4 substrate in humans, but due to its relatively high in vivo permeability and metabolic stability, it was determined that gut extraction would be minimal in human In vivo permeability determined to be high based on PK data in rat and dog
  21. 21. Case studies: Compound 3 Applications of the flow diagrams 21 Neil Miller April 2019 FIH 4 FIH.pptx • Prediction of distribution challenging as Vss varied 18-fold in animals: • Scene setting • Components • Case studies • Summary The Lukacova method underestimated Vss across all preclinical species and this was mainly due to an in silico predicted acidic pKa value Sensitivity analysis showed that removing the acidic pKa from the calculation allowed the calculated Vss to be sensitive to other parameters In vitro Fup was <0.1% across species, but could not be precisely determined, and a range of 0.01-0.1% enabled reasonable Vss predictions in preclinical species
  22. 22. Case studies: Compound 3 Applications of the flow diagrams 22 Neil Miller April 2019 FIH 4 FIH.pptx • Metabolism & Elimination: • Scene setting • Components • Case studies • Summary Metabolically cleared based on in vitro data Predicting clearance for acidic compounds by scaling in vitro data is often challenging, and consideration of binding is important Hepatocytes provided one estimate of human clearance Given the limitations of the in vitro assays, single- species scaling based on monkey was also used for a range of predicted human clearance
  23. 23. • Reasonably accurate human PK predictions: Case studies: Compound 3 Evaluating the predictions for 25mg tablet in the fed state 23 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary Clearance and Vss predictions uncertain for this compound and it seemed they were both linked to Fup which could not be measured Therefore, two combinations of parameters were explored: • Low clearance & Low Vss due to a lower Fup • High clearance & High Vss due to a higher Fup  Decision making: Increased confidence that efficacious exposures would be achieved in the clinic  Learning opportunity: Assessing the in vitro inputs against in vivo PK profiles in preclinical species, and determining alternative parameters when in vitro data were inconsistent
  24. 24. • Have faith in using PBPK modelling for predicting First-In-Human (FIH) pharmacokinetics • Use everything that we/you know by connecting the parts • See failures as opportunities to learn as “In a complex world, failure is inevitable.” Matthew Syed author of Black Box Thinking: The Surprising Truth About Success Summary Take home messages 24 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary
  25. 25. • People stick to simple empirical approaches because they can be right • People are wary of complex mechanistic approaches because they can be wrong • Empirical approaches are fixed and if they are wrong there is no room to grow • Do the same again and hope for a better outcome • PBPK approaches take into account all of the parts and grow with knowledge Summary Final thoughts 25 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary Have Faith In Holistic modelling for First-In-Human PK
  26. 26. Acknowledgements Team effort 26 Neil Miller April 2019 FIH 4 FIH.pptx • Scene setting • Components • Case studies • Summary • Co-authors of the FIH PBPK manuscript: • Neil Parrott (Roche) • Micaela Reddy (Array BioPharma) • Viera Lukacova (Simulations Plus) • Aki Heikkinen (Admescope)

×