MODELING THE INTERNAL DATA

1
Quantitative Human Pharmacology Modeling to
Accelerate SGLT2i Drug Development G Nucci, A
Ghosh, N Haddish-Berhane, P DaSilva-Jardine, J
Trimmer, M Reed and CJ Musante Pfizer Global
Research Development, Pfizer Inc, Groton CT
06340 Entelos inc. Foster City, CA 94404
INTRODUCTION The sodium-glucose cotransporter 2
(SGLT2) is responsible for the majority of the
glucose reabsorption by the kidney1 and has been
recognized as a novel therapeutic approach for
improving glucose control in type 2 diabetes
mellitus (T2DM). Several selective SGLT2
inhibitors (SGLT2i) are currently under
development2,3. SGLT2 inhibition results in
increased urinary glucose excretion (UGE) in both
healthy subjects (HS)4,5 and subjects with
T2DM6,7,8 leading to reduced plasma glucose and
glycosylated hemoglobin (A1c) levels with the
additional benefit of a negative caloric balance,
thereby combining glucose control with weight
loss9,10. UGE provides a mechanistic biomarker
for the clinical assessment of SGLT2 inhibition.

• MODEL BUILDING VALIDATION
• A physiologically based representation of
  competitive SGLT2 inhibition was added to the
  PhysioLab to account for the effect on glucose
  reabsorption in the proximal tubule (see poster
  M-5-4 from A. Ghosh et al).
• Assuming the SGLT2 transporter is uniformly
  distributed along the length L of the effective
  tubule, that we have an average urinary flow of
  velocity ?, a tubule glucose concentration C and
  applying competitive SGLT2 inhibition through
  Michaelis Menten kinetics we have

• MODELING THE INTERNAL DATA
• The FIH design (dosing protocols, meals, HS
  characteristics) was incorporated in the
  Physiolab. As soon as the internal candidate PK
  and UGE data became available its popPK
  parameters were included in the platform
  representation and the drug potency Ki was tuned
  in real time to match the observed exposure
  -response characteristics both as daily UGE as
  well as the time course of UGE excretion.

• CONCLUSIONS
• Quantitative human pharmacology modeling of SGLT2
  inhibition enabled the optimization of doses and
  dosing regimen for clinical trial design of the
  Pfizer SGLT2i program taking full advantage of
  studies conducted internally as well as
  quantitative understanding of published data.
• By increasing confidence in pharmacology and in
  understanding of the mechanism of action the
  model allowed the ability to take calculated risk
  to gain speed in clinical development (i.e.
  combining phase IIa and IIb)
• It also enabled a comparison of various
  inhibition profiles in T2D virtual patient
  populations to develop dosing strategies that
  optimize the efficacy and tolerability of the
  clinical candidate.
• In general the availability of quantitative human
  pharmacology models allows to
• Identify therapeutic targets
• Identify knowledge gaps (call for experiments to
  increase confidence or disprove a new target)
• Answer trial design questions and simulate
  outcome leveraging on quantitative understanding
  of the mechanism of action and human
  pathophysiology
• Improve decision making in drug development

OBJECTIVES Our aim was to integrate the available
data on an internal SGLT2i candidate, the
physiologic understanding of the mechanism of
action, and all published data on other SGLT2
targeting compounds in a mechanistic model to
inform the clinical development plan. In order to
improve the quality and speed of decision-making
(clinical trial design, doses, and dosing
regimens), we aimed to represent SGLT2 inhibition
in the Entelos Metabolism PhysioLab Platform
(PhysioLab), establishing a link between the
biomarker UGE in HS and improvements in glycemic
control (A1c) and body weight (BW) in longer term
studies in T2DM.
THE PHYSIOLAB PLATFORM The PhysioLab is a large
scale mathematical model of human T2DM
pathophysiology. It is based on an extensive
survey of published literature and includes a
mechanistic representation of the major
physiological systems involved in the regulation
of nutrient intake, utilization, storage, and
disposal in health and disease. T2DM virtual
patients are created by parametrically changing
the underlying disease pathophysiology (e.g.,
insulin and glucose effects on metabolic
pathways) to reflect the diversity observed in
T2DM trials. The model has been validated by
comparing virtual patient responses to published
and proprietary data from a variety of clinical
studies and trials, including oral glucose
tolerance tests, clamp procedures, mixed nutrient
meal tests, nutrient and hormone infusions, and
therapeutic interventions.

• REFERENCES
• Kanai, Y. et al. Journal of Clinical
  Investigation 1994, 93397-404.
• Isaji M. Curr Opin Investig Drugs. 2007 8
  285-292.
• Idris I. Donnelly R.. Diabetes Obesity and
  Metabolism. 2009, 1179-88.
• Komoroski B, et al. Clin Pharmacol Ther. 2009,
  85520-526.
• Hussey EK, et al. J Clin Pharmacol. 2010.
• Hussey E.K et al. 69th ADA Scientific Sessions.
  582-P.
• Polidori D. et al. 69th ADA Scientific Sessions.
  511-P
• Komoroski B., et al. Clin Pharmacol. Ther. 2009,
  85513519
• List J.F. et al. Diabetes Care 2009, 32,
  650-657.
• Wilding J.Pet al. Diabetes Care. 2009,
  321656-1662