DDMODEL00000122: Samtani_2010_Hb1Ac_prediction
Short description:
Linear regression model to simulate the HbA1c value at the steady state from fasting plasma glucose (FPG)
PharmML 0.8.x (0.8.1) 



Paolo Magni

Context of model development:  Clinical endpoint; Diagnostic model; 
Model compliance with original publication:  Yes; 
Model implementation requiring submitter’s additional knowledge:  No; 
Modelling context description:  The objectives were to develop a translational model that will help select doses for Phase3 trials based on abbreviated Phase2 trials and understand the competitive landscape for oral antidiabetics based on efficacy, tolerability and ability to slow disease progression. Data for eight antidiabetics with temporal profiles for fasting plasma glucose (FPG) and hemoglobin A1c (HbA1c) from 12 publications were digitized. The monotherapy data consisted of FPG and HbA1c profiles that were modeled using a transit compartment model. Evaluation of the competitive landscape utilized HbA1c literature data for 11 drugs. For the safety metric, tolerability issues with antidiabetic drug classes were tabulated. For disease progression, the coefficient of failure method was used for assessing data from two longterm trials. The transit rate constants were remarkably consistent across 12 trials and represent systemspecific/drugindependent parameters. The competitive landscape analysis showed that the primary efficacy metric fell into two groups of ?HbA1c: >0.8% or ?0.8%. On the safety endpoints, older agents showed some tolerability issues while the new agents were relatively safe. Among the different drug classes, only the thiazolidinediones appeared to slow disease progression but may also increase heart failure risk. In conclusion, the ability of systemspecific parameters to predict HbA1c provides a tool to predict the expected efficacy profile from abbreviated dosefinding trials. To be commercially viable, new drugs should improve ?HbA1c by about 0.8% or more and possess safety profiles similar to newer antidiabetic agents. Thus, this study proposes a suite of simple yet powerful tools to guide type2diabetes drug development. ; 
Modelling task in scope:  simulation; 
Nature of research:  Clinical research & Therapeutic use; 
Therapeutic/disease area:  Endocrinology; 
Annotations are correct. 

This model is not certified. 
 Model owner: Paolo Magni
 Submitted: Dec 12, 2015 3:23:51 PM
 Last Modified: Oct 10, 2016 8:36:07 PM
Revisions

Version: 7
 Submitted on: Oct 10, 2016 8:36:07 PM
 Submitted by: Paolo Magni
 With comment: Edited model metadata online.

Version: 5
 Submitted on: Jun 2, 2016 8:14:41 PM
 Submitted by: Paolo Magni
 With comment: Updated model annotations.

Version: 2
 Submitted on: Dec 12, 2015 3:23:51 PM
 Submitted by: Paolo Magni
 With comment: Edited model metadata online.
Name
Generated from MDL. MOG ID: Method_1_Samtani_mog
Independent Variables

Function Definitions
$\mathrm{additiveError}:\mathrm{real}\left(\mathrm{additive}:\mathrm{real}\right)=\mathrm{additive}$

Covariate Model: $\mathrm{cm}$
Continuous Covariates
$\mathrm{FPG}$
Parameter Model: $\mathrm{pm}$
Random Variables
${\mathrm{EPS\_1}}_{\mathrm{vm\_err.DV}}~\mathrm{Normal2}\left(\mathrm{mean}=0,\mathrm{var}=1\right)$
Population Parameters
$\mathrm{BETA0}$
$\mathrm{BETA1}$
$\mathrm{RES}$
Structural Model: $\mathrm{sm}$
Variables
$\mathrm{HBA1C}=\mathrm{pm.BETA0}+\mathrm{pm.BETA1}\cdot \mathrm{cm.FPG}$
Observation Model: $\mathrm{om1}$
Continuous Observation
$Y=\mathrm{sm.HBA1C}+\mathrm{additiveError}\left(\mathrm{additive}=\mathrm{pm.RES}\right)+\mathrm{pm.EPS\_1}$
External Dataset
OID

$\mathrm{nm\_ds}$

Tool Format

NONMEM

File Specification
Format

$\mathrm{csv}$

Delimiter

comma

File Location

Simulated_Samtani_2010_ss_data.csv

Column Definitions
Column ID  Position  Column Type  Value Type 

$\mathrm{ID}$ 
$1$

$\mathrm{id}$

$\mathrm{int}$

$\mathrm{TIME}$ 
$2$

$\mathrm{idv}$

$\mathrm{real}$

$\mathrm{DV}$ 
$3$

$\mathrm{dv}$

$\mathrm{real}$

$\mathrm{FPG}$ 
$4$

$\mathrm{covariate}$

$\mathrm{real}$

$\mathrm{EV}$ 
$5$

$\mathrm{undefined}$

$\mathrm{real}$

Column Mappings
Column Ref  Modelling Mapping 

$\mathrm{TIME}$ 
$T$ 
$\mathrm{DV}$ 
$\mathrm{om1.Y}$ 
$\mathrm{FPG}$ 
$\mathrm{cm.FPG}$ 
Estimation Step
OID

$\mathrm{estimStep\_1}$

Dataset Reference

$\mathrm{nm\_ds}$

Parameters To Estimate
Parameter  Initial Value  Fixed?  Limits 

pm.BETA0 
$2.84$

false

$\left(,\right)$

pm.BETA1 
$0.5$

false

$\left(,\right)$

pm.RES 
$0$

true

$\left(,\right)$

Operations
Operation: $1$
Op Type

generic

Operation Properties
Name  Value 

algo

$\text{foce}$

Step Dependencies
Step OID  Preceding Steps 

$\mathrm{estimStep\_1}$
