Please visit the new BioModels platform to access the latest content. This website is no longer updated and will be retired on 31 May 2019.
BioModels Database logo

BioModels Database


Palmér et al., (2014). Effects of IL-1β-Blocking Therapies in Type 2 Diabetes Mellitus: A Quantitative Systems Pharmacology Modeling Approach to Explore Underlying Mechanisms.

April 2017, model of the month by Vincent Knight-Schrijver
Original model: BIOMD0000000620, BIOMD0000000621


Inflammatory cytokines are essential in a normal immune response. However, like the elements of many other highly regulated biological networks, the dysfunctional or disproportionate activity of these cytokines causes disruption. This is often to the detriment of the organism and manifests itself as a pathological condition. In this article the focus is on the effects of Interleukin 1 beta (IL-1β) by inducing pancreatic beta-cell destruction. This drives the progression of type-2 diabetes mellitus (T2DM). The targeting of IL-1β signalling is proposed as a viable therapeutic option seen in multiple studies including clinical trials of a compound, anakinra or the monoclonal antibody canakinumab. The mechanism behind both of these putative therapies is IL-1β blockade. The quantitative systems pharmacology (QSP) model showcased here explores the effects of IL-1β receptor (IL-1βRa) inhibition by anakinra [1].

The Model

There are two parametrisations of this model in the BioModels Database. A healthy condition model [BIOMD0000000621] and a disease condition [BIOMD0000000620] model. The encoded SBML model equations controlling the glucose homeostasis were modified from their original algebraic form.
Palmér et al (2014) focus upon the effects of anakinra in T2DM patients and assume the disease state. The model consists of the insulin, glucose and proinsulin homeostasis adapted from previous models such as that by de Gaetano et al [2] - available in the non-curated branch of the BioModels (MODEL1112110003). In addition to this, the model by Palmér et al includes β-cell function and turnover as well as a model for glycated haemoglobin (HbA1c). Finally, the model has the added value of incorporating IL-1β-mediated effects. This enables the model to simulate important effects of IL-1βRa blockade on β-cell function and predictively report suitable clinical observations. The schematic representation of the model is illustration in Figure 1

Figure 1

Figure 1The model schema illustrates the interaction of the various modelled components of IL-1β modulation in T2DM. IL-1βRa modulation regulates the rates of replication and apoptosis in β-cells. It also regulates the secretion of insulin. Figure taken from [1].

IL-1βRa modulation is described by a competitive inhibition model between anakinra and IL-1β for the receptor. Active IL-1βRa is then reported to inhibit the insulin secretion of pancreatic β-cells. IL-1β also modulates the opposing replication and apoptosis reactions of β-cells. We can see the model's reproduction of these phenomena in figure 2. The interactions imply that an increase in IL-1β would increase the IL-1βRa reponse and thus reduce the insulin secretion capacity. A similar effect can be seen in the rate of change in β-cell number. However, the β-cell rate of change response is bell-shaped and implies that an intermediate but not low or high activation of IL-1βRa actually increases the rate of β-cell replication and overall mass (figure 2). Perhaps an optimal level of IL-1βRa blockade exists which both lies within this region and improves insulin secretion capacity. Through which process does the current therapeutic regimen exert its therapeutic effect?

Figure 2

Figure 2IL-1βRa modulation.[3]. Left: The effect of IL-1βRa activity upon β-cell mass. Right: the effect of IL-1βRa modulation upon the insulin secretion capacity of the pancreas. Figure taken from [1].


Model simulations utilise a daily administration of 200 mg anakinra for thirteen weeks followed thirty nine weeks of observation. The results suggest that, during treatment, the insulin secretory capacity improves dramatically and the mass decreases (figure 3). However in the weeks following therapy the mass begins to improve beyond the initial disease state. Resulting from this is an agreement with clinical observations as proinsulin:insulin ratio (PI/I) and HbA1c data [3, 4] for responders were reproduced by the model during simulated anakinra administration (figure 3, lower panel). The effect upon the PI/I ratio and the HbA1c is caused by a combination of changes in β-cell function and mass.

Figure 3

Figure 3Simulations of 13 weeks of anakinra therapy (Blue line) or placebo (Red line). Upper Left: Improvements in Insulin secretory capacity. Upper Right: Improvements in β-cell mass. Lower Left: PI/I data and simulations. Lower Right: HbA1c data and simulations. Upper panels were simulated in Copasi, lower panels were taken from [1].

However, the key factor in linking the therapy to the clinical response is not fully understood. How then does the reduction in IL-1βRa modulation affect a combination of both facets seen in figure 2? To explore the contributions of each element of IL-1βRa response, the authors simulated separate cases for an improvement in either β-cell function or mass (figure 4). The model predicts that whilst increasing the β-cell mass alone has little effect upon short-term HbA1c or fasting plasma glucose, improvements in β-cell function produce a substantial decrease in both clinical markers.

Figure 4

Figure 4HbA1c response to separate improvements in β-cell function (yellow) or β-cell mass (green) over two years of the anakinra treatment regimen. Figure taken from [1].


The concluding thoughts on this are that the relatively fast benefit in glycaemic control is seen by inhibiting IL-1βRa's modulatory effect upon β-cell function through improvements in the insulin secretory capacity and the slowing of further IL-1β release. A further but less pronounced long-term benefit is likely seen through IL-1βRa's effect upon β-cell mass which could be significant only after multiple treatment episodes such as that simulated in figure 4. The model helps to understand the differences between elements of IL-1βRa blockade but also emphasises the synergistic action responsible for a full clinical response in patients. Perhaps then, in non-responsive patients, the model could be purposed to dissect if and how an individual fails to respond to a specific piece of IL-1βRa modulation by assessing the nature of their response. Finally QSP models like this continue to provide a quantitative interface between regimen optimisation and clinical endpoints for clinical use and trial design.


  1. Palmér, R. et al. Effects of IL-1β–Blocking Therapies in Type 2 Diabetes Mellitus: A Quantitative Systems Pharmacology Modeling Approach to Explore Underlying Mechanisms. CPT: pharmacometrics & systems pharmacology 3.6 (2014): 1-8.
  2. De Gaetano, A. et al. Mathematical models of diabetes progression. Am. J. Physiol. Endocrinol. Metab. 2008 Dec; 295(6): E1462-79.
  3. Larsen, C.M. et al. Interleukin-1-receptor antagonist in type 2 diabetes mellitus. N. Engl. J.Med. 356, 1517-1526 (2007).
  4. Larsen, C.M. et al. Sustained effects of interleukin-1 receptor antagonist treatment in type 2 diabetes. Diabetes Care 32, 1663-1668 (2009).