Identifying Synergy

A specific theme in our pre-clinical work is the assessment of combination strategies. Part of this research is done by evaluating pre-clinical growth inhibition data in various cancer cell lines. Nevertheless, to obtain an understanding of the generated combination data is not straightforward. How do we know that measured effects are amplified? After all, we are adding a drug to another, so most often we will obviously expect to see an increased cytostatic or cytotoxic effect. This question is actually a century old one and received its first answer from Dr Siegfried Walter Loewe in 1923, with the widespread mathematical model known as the isobole equation and the additivity concept.

Except in the most straightforward cases, appropriate interpretation of generated dose response data requires a mathematical model and software implementation. The classical models that provide insight in terms of drug combination effects are the aforementioned Loewe’s model but also the Bliss model and simpler Highest Single Agent (HSA) procedure. Nevertheless, each of these models is based on concepts which, in our opinion, are not widely applicable. Thus, we have developed the new Synergy, Antagonism or Neutrality Estimation (SANE) model to replace them (more details to follow soon).

Synergy studies using Combenefit software

In practice, most synergy studies will investigate effects obtained in specific systems (organisms, cell-line, proteins etc.), with a range of concentrations for drug A (generally four to ten concentration points) and a range of concentrations for drug B. For instance, in our lab we investigate mostly cancer-cell line growth inhibition and use a 8×8 format. The experimental results are then analysed based on the specific reference mathematical model (e.g. the Loewe model or our new preferred SANE model). We have developed the Combenefit software which allows automatic model-based analysis.

Combenefit only requires the user to save replicate data in the predefined .xls template files. It is based on an interactive platform which collects all experimental replicates and generates surface analyses for selected models. A wealth of outputs can be obtained and the software populates the project folder with the results. Our new SANE model as well as the Loewe, Bliss and HSA models are all available in Combenefit. Combenefit can be deployed on win64 machines and does not require any other pre-installed software (e.g. Matlab).

Not only drugs

The problem of identifying synergy is not confined to cancer pre-clinical studies. It is common to many other disciplines (e.g. Anaesthesiology, Endocrinology, Microbiology, Toxicology etc.) and does not necessarily involve two drugs. Studies can involve other agents such as hormones, neuromodulators, radiation etc. and different type of agents can be combined. We hope that Combenefit will help a large audience and we encourage people to get in touch with us via e-mail if help about experimental design or software analysis is needed. Combenefit is licensed under the MIT Licence and is therefore freely available for everyone, but please acknowledge us appropriately (a reference will hopefully be available soon).

Combenefit can be downloaded from the source forge website.

Please cite our publication: Di Veroli GY, Fornari C, Wang D, Mollard S, Bramhall JL, Richards FM, Jodrell DI. Combenefit: an interactive platform for the analysis and visualization of drug combinations. Bioinformatics. 2016 Sep 15;32(18):2866-8. Epub 2016 Apr 25. DOI; PubMed: 27153664; PubMed Central: PMC5018366

A Combenenefit users’ guide and guidelines about the design of synergy studies are in preparation and will be available to download soon.