Advertisement

Experiment design for nonparametric models based on minimizing Bayes Risk: application to voriconazole\(^{1}\)

  • David S. Bayard
  • Michael NeelyEmail author
Original Paper

Abstract

An experimental design approach is presented for individualized therapy in the special case where the prior information is specified by a nonparametric (NP) population model. Here, a NP model refers to a discrete probability model characterized by a finite set of support points and their associated weights. An important question arises as to how to best design experiments for this type of model. Many experimental design methods are based on Fisher information or other approaches originally developed for parametric models. While such approaches have been used with some success across various applications, it is interesting to note that they largely fail to address the fundamentally discrete nature of the NP model. Specifically, the problem of identifying an individual from a NP prior is more naturally treated as a problem of classification, i.e., to find a support point that best matches the patient’s behavior. This paper studies the discrete nature of the NP experiment design problem from a classification point of view. Several new insights are provided including the use of Bayes Risk as an information measure, and new alternative methods for experiment design. One particular method, denoted as MMopt (multiple-model optimal), will be examined in detail and shown to require minimal computation while having distinct advantages compared to existing approaches. Several simulated examples, including a case study involving oral voriconazole in children, are given to demonstrate the usefulness of MMopt in pharmacokinetics applications.

Keywords

Bayes Risk Experiment design Nonparametric Population model Voriconazole 

Notes

Acknowledgements

Support from NIH Grants GM 068968 and HD 070886 is acknowledged.

Supplementary material

10928_2016_9498_MOESM1_ESM.pdf (64 kb)
(PDF 65 kb)

References

  1. 1.
    Andes D, Marchillo K, Stamstad T, Conklin R (2003) In vivo pharmacokinetics and pharmacodynamics of a new triazole, voriconazole, in a murine candidiasis model. Antimicrob Agents Chemother 47(10):3165–3169CrossRefPubMedPubMedCentralGoogle Scholar
  2. 2.
    Atkinson AC, Donev AN, Tobias RD (2007) Optimum experimental designs, with SAS. Oxford University Press, New YorkGoogle Scholar
  3. 3.
    Baek Y (2006) An interior point approach to constrained nonparametric mixture models. PhD Thesis, University of WashingtonGoogle Scholar
  4. 4.
    Bayard DS (1991) A forward method for optimal stochastic nonlinear and adaptive control. IEEE Trans Autom Control 36(9):1046–1053CrossRefGoogle Scholar
  5. 5.
    Bayard DS, Schumitzky A (2010) Implicit dual control based on particle filtering and forward dynamic programming. Int J Adapt Control Signal Process 24(3):155–177PubMedPubMedCentralGoogle Scholar
  6. 6.
    Bayard DS, Jelliffe R, Schumitzky A, Milman M, Van Guilder M (1995) Precision drug dosage regimens using multiple model adaptive control: theory and application to simulated vancomycin therapy. In: Selected topics in mathematical physics, Prof. R. Vasudevan Memorial volume. World Scientific Publishing Co., MadrasGoogle Scholar
  7. 7.
    Bayard DS, Jelliffe R, Neely M (2013) Bayes risk as an alternative to Fisher information in determining experimental designs for nonparametric models. In: PODE 2013 workshop: population optimum design of experiments, Lilly, UK, 15 June 2013Google Scholar
  8. 8.
    Bayard DS, Jelliffe R, Neely M (2014) Experiment design based on Bayes risk and weighted Bayes risk with application to pharmacokinetic systems. In: PODE 2014 workshop: population optimum design of experiments, Roche, Basel, 11 September 2014Google Scholar
  9. 9.
    Bender CM, Orszag SA (1978) Advanced mathematical methods for scientists and engineers. McGraw-Hill Book Company, New YorkGoogle Scholar
  10. 10.
    BestDose software package. http://www.lapk.org. Accessed 1 Jan 2016
  11. 11.
    Blackmore L, Rajamanoharan S, Williams BC (2008) Active estimation for jump Markov linear systems. IEEE Trans Autom Control 53(10):2223–2236CrossRefGoogle Scholar
  12. 12.
    Box REP, Lucas HL (1959) Design of experiments in non-linear situations. Biometrika 46:77–90CrossRefGoogle Scholar
  13. 13.
    Chaloner K, Verdinelli I (1995) Bayesian experimental design: a review. Stat Sci 10(3):273–304CrossRefGoogle Scholar
  14. 14.
    D’Argenio DZ (1981) Optimal sampling times for pharmacokinetic experiments. J Pharmacokinet Biopharm 9(6):739–756CrossRefPubMedGoogle Scholar
  15. 15.
    D’Argenio DZ (1990) Incorporating prior parameter uncertainty in the design of sampling schedules for pharmacokinetic parameter estimation experiments. Math Biosci 99:105–118CrossRefPubMedGoogle Scholar
  16. 16.
    D’Argenio DZ, Schumitzky A, Wang X (2009) ADAPT 5 user’s guide. Biomedical simulation resource. University of Southern CaliforniaGoogle Scholar
  17. 17.
    Draper NR, Hunter WG (1967) The use of prior distributions in the design of experiments for parameter estimation in non-linear situations. Biometrika 54(1–2):147–153CrossRefPubMedGoogle Scholar
  18. 18.
    Duda RO, Hart PE, Stork DG (2001) Pattern classification, 2nd edn. Wiley, New YorkGoogle Scholar
  19. 19.
    Fedorov VV (1972) Theory of optimal experiments. Academic, New YorkGoogle Scholar
  20. 20.
    Fedorov VV, Leonov SL (2013) Optimal design for nonlinear response models. CRC Press, New YorkGoogle Scholar
  21. 21.
    Gibaldi M, Perrier D (1982) Pharmacokinetics, 2nd edn. Marcel Dekker, Inc., New YorkGoogle Scholar
  22. 22.
    Goodwin GC, Payne RL (1977) Dynamic system identification: experiment design and data analysis. Academic, New YorkGoogle Scholar
  23. 23.
    Hennig S, Nyberg J, Fanta S, Backman JT, Hoppu K, Hooker AC, Karlsson MO (2012) Application of the optimal design approach to improve a pretransplant drug dose finding design for ciclosporin. J Clin Pharmacol 52(3):347–360. doi: 10.1177/0091270010397731 (Epub 4 May 2011)
  24. 24.
    Hu C, Lovejoy WS, Shafer SL (1994) Comparison of some control strategies for three-compartment PK/PD models. J Pharmacokinet Biopharm 22(6):525–550CrossRefPubMedGoogle Scholar
  25. 25.
    Huurneman LJ, Neely M, Veringa A, Prez FD, Ramos-Martin V, Tissing WJ, Alffenaar JWC, Hope W (2016) Pharmacodynamics of voriconazole in children: further steps along the path to true individualized therapy. Antimicrob Agents Chemother 60(4):2336–2342CrossRefPubMedPubMedCentralGoogle Scholar
  26. 26.
    Jelliffe R, Schumitzky A, Bayard D, Van Guilder M, Leary RH (2002) The USC*PACK programs for parametric and nonparametric population PK/PD modeling. Population Analysis Group in Europe, ParisGoogle Scholar
  27. 27.
    Karthaus M, Lehrnbecher T, Lipp HP, Kluge S, Buchheidt D (2015) Therapeutic drug monitoring in the treatment of invasive aspergillosis with voriconazole in cancer patients: an evidence-based approach. Ann Hematol 94(4):547–556CrossRefPubMedGoogle Scholar
  28. 28.
    Katz D, D’Argenio DZ (1984) Discrete approximation of multivariate densities with application to Bayesian estimation. Comput Stat Data Anal 2(1):27–36CrossRefGoogle Scholar
  29. 29.
    Kendall M, Stuart A (1979) The advanced theory of statistics, vol 2, 4th edn. Charles Griffin and Company, LondonGoogle Scholar
  30. 30.
    Kistos CP (2013) Optimal experimental design for non-linear models: theory and applications. Springer, New YorkGoogle Scholar
  31. 31.
    Landaw EM (1980) Optimal experiment design for biologic compartmental systems with applications to pharmacokinetics. PhD Dissertation, University of California, Los AngelesGoogle Scholar
  32. 32.
    Leary R, Jelliffe R, Schumitzky A, Van Guilder M (2001) An adaptive grid non- parametric approach to pharmacokinetic and dynamic (PK/PD) population models. In: Proceedings of the 14th IEEE symposium on computer-based medical systems, 2001, CBMS 2001. IEEE, p 389–394Google Scholar
  33. 33.
    Lindley DV (1972) Bayesian statistics, a review. Society for Industrial and Applied Mathematics (SIAM), PhiladelphiaCrossRefGoogle Scholar
  34. 34.
    Lindsay B (1983) The geometry of mixture likelihoods: a general theory. Ann Stat 11:86–94CrossRefGoogle Scholar
  35. 35.
    Luong ML, Al-Dabbagh M, Groll AH, Racil Z, Nannya Y, Mitsani D, Husain S (2016) Utility of voriconazole therapeutic drug monitoring: a meta-analysis. J Antimicrob Chemother 71(7):1786–1799CrossRefPubMedGoogle Scholar
  36. 36.
    Mallet A (1986) A maximum likelihood estimation method for random coefficient regression models. Biometrika 73:645–656CrossRefGoogle Scholar
  37. 37.
    Merle Y, Mentre F (1995) Bayesian design criteria: computation, comparison, and application to a pharmacokinetic and a pharmacodynamic model. J Pharmacokinet Biopharm 23(1):101–125CrossRefPubMedGoogle Scholar
  38. 38.
    Neely M, Rushing T, Kovacs A, Jelliffe R, Hoffman J (2010) Voriconazole pharmacokinetics and pharmacodynamics in children. Clin Infect Dis 50(1):27–36CrossRefPubMedPubMedCentralGoogle Scholar
  39. 39.
    Neely M, Margol A, Fu X et al (2015) Achieving target voriconazole concentrations more accurately in children and adolescents. Antimicrob Agents Chemother 59:3090–3097CrossRefPubMedPubMedCentralGoogle Scholar
  40. 40.
    Pronzato L, Pazman A (2013) Design of experiments in nonlinear models: asymptotic normality, optimality criteria and small-sample properties. Springer, New YorkCrossRefGoogle Scholar
  41. 41.
    Pronzato L, Walter E (1985) Robust experiment design via stochastic approximation. Math Biosci 75:103–120CrossRefGoogle Scholar
  42. 42.
    Raiffa H, Schlaifer R (2000) Applied statistical decision theory. Wiley Classics Library, Wiley, New YorkGoogle Scholar
  43. 43.
    Ryan EG, Drovandi CC, McGree JM, Pettitt AN (2015) A review of modern computational algorithms for Bayesian optimal design. Int Stat Rev 84(1):128–154CrossRefGoogle Scholar
  44. 44.
    Rodman JH, D’Argenio DZ, Peck CC (2004) Analysis of pharmacokinetic data for individualizing drug dosage regimens. In: Buron ME, Shaw LM, Schentag JJ, Evans WE (eds) Applied pharmacokinetics: principles of therapeutic drug monitoring, 4th edn. Applied Therapeutics, SpokaneGoogle Scholar
  45. 45.
    Schumitzky A (1991a) Application of stochastic control theory to optimal design of dosage regimens. In: D’Argenio DZ (ed) Advanced methods of pharmacokinetic and pharmacodynamic systems analysis. Plenum Press, New York, pp 137–152CrossRefGoogle Scholar
  46. 46.
    Schumitzky A (1991b) Nonparametric EM algorithms for estimating prior distributions. Appl Math Comput 45(2):143–157Google Scholar
  47. 47.
    Sheiner LB, Beal SL (1982) Bayesian individualization of pharmacokinetics: simple implementation and comparison with non-Bayesian methods. J Pharm Sci 71:1344–1348CrossRefPubMedGoogle Scholar
  48. 48.
    Sheiner LB, Rosenberg B, Melmon KL (1972) Modeling of individual pharmacokinetics for computer-aided drug dosage. Comput Biomed Res 5:441–459CrossRefGoogle Scholar
  49. 49.
    Sheiner LB, Halkin H, Peck C, Rosenberg B, Meimon KL (1975) Improved computer-assisted digoxin therapy: a method using feedback of measured serum digoxin concentrations. Ann Intern Med 82:619–627CrossRefPubMedGoogle Scholar
  50. 50.
    Silvey SD (1980) Optimal design: an introduction to the theory for parameter estimation. Chapman and Hall, LondonCrossRefGoogle Scholar
  51. 51.
    Tod M, Rocchisani J-M (1996) Implementation of OSPOP, an algorithm for the estimation of optimal sampling times in pharmacokinetics by the ED, EID and API criteria. Comput Methods Programs Biomed 50(1):13–22CrossRefPubMedGoogle Scholar
  52. 52.
    Tod M, Rocchisani JM (1997) Comparison of ED, EID, and API criteria for the robust optimization of sampling times in pharmacokinetics. J Pharmacokinet Biopharm 25(4):515–537CrossRefPubMedGoogle Scholar
  53. 53.
    Tse E, Bar-Shalom Y, Meier L III (1973) Wide-sense adaptive dual control for nonlinear stochastic systems. IEEE Trans Autom Control 18(2):98–108CrossRefGoogle Scholar
  54. 54.
    Vajjah P, Duffull SB (2012) A generalisation of T-optimality for discriminating between competing models with an application to pharmacokinetic studies. Pharm Stat 11(6):503–510CrossRefPubMedGoogle Scholar
  55. 55.
    Vozeh S, Steimer J-L (1985) Feedback control methods for drug dosage optimisation: concepts, classification and clinical application. Clin Pharmacokinet 10:457–476CrossRefPubMedGoogle Scholar
  56. 56.
    Vozeh S, Berger M, Wenk M, Ritz R, Follath F (1984) Rapid prediction of individual dosage requirements for lignocaine. Clin Pharmacokinet 9:353–363CrossRefGoogle Scholar
  57. 57.
    Walter E, Pronzato L (1987) Optimal experiment design for nonlinear models subject to large prior uncertainties. Am J Physiol Regul Integr Comp Physiol 253:R530–R534Google Scholar

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  1. 1.Laboratory of Applied Pharmacokinetics and Bioinformatics, The Saban Research InstituteChildren’s Hospital of Los AngelesLos AngelesUSA
  2. 2.Keck School of MedicineUniversity of Southern CaliforniaLos AngelesUSA

Personalised recommendations