New tools to identify optimal drug combinations

2018
Dr Nicole Andenmatten, Project Manager, Translational Science, MMV

Today, in antimalarial drug development, we are at a juncture that offers many challenges and opportunities. Owing to the threat of resistance, we are aligned with the World Health Organization’s (WHO) recommendation to develop combination therapies to treat malaria, so that one compound in the combination can kill any remaining parasites if resistance to the other compound is generated.

Today, we have 14 compounds in preclinical and clinical development. Given the number of possible combinations this provides, a significant amount of data must be sifted through in order to select the best. Ideal combinations should have matched pharmacokinetics (PK; duration of activity in the body), different molecular targets and different paths of resistance. They must not interact in a negative way, for example, by increasing each other’s metabolism or showing additivity in safety signals, but ideally show additivity in their pharmacological activity. In addition, each time we take a combination forward, the owners of each molecule must consent.

Dr Nicole Andenmatten, Project Manager, Translational Science, MMV, explains how new tools are being employed to determine which drug combinations should progress further.

1. How were drug combinations selected traditionally?

In the past, the first time a combination would be evaluated would be in a clinical setting once the efficacy of the individual compounds had been confirmed in patients. This is a very expensive and time-consuming process. It’s okay if you only have one combination to consider, but once you have many, we need a way to prioritize them. We need to give our clinical colleagues two or three options that will work; not too many, and not too few.

2. What tools are being used to develop drug combinations at MMV?

Today, we have an in silico combination tool (go to the graphic), which began as a manual tool to develop combinations. We fondly call it the Combo tool. It was originally developed by Dr Wesley Van Voorhis from the University of Washington when he came to MMV on sabbatical 3 years ago.

We gather extensive data on each compound to determine which to take into development. This tool helps us organize those data. But the tool is only as good as the data that we enter into it. Those data come from a range of different studies and analyses. Using SCID mouse studies and volunteer infection studies (VIS)1 we look at the efficacy of individual compounds against the parasite that infects humans. As part of the standard drug development process we also evaluate the PK, safety, resistance potential and chemical characteristics of each compound.

Fed with all these data on individual compounds, the Combo tool can then generate a matrix of the compounds that form the best combinations. It’s a little bit like an online dating platform that collects data on each compound and then sorts them by compatibility, and suggests the best possible match.

We then select a subgroup of the most promising, and perform SCID combination studies to look at the combined action of these compounds. Using these data and human pharmacodynamics (PD)/PK information, we can predict how the combination might work in humans. From this, a decision to move forward into combination VIS and clinical trials can be made.

3. What are the advantages of these tools?

By evaluating the combination more thoroughly before phase IIb in patients, we “de-risk” the research programme, as we have more information on how the compounds will behave together, and so stand a much higher chance of choosing an optimal and efficacious combination to progress. It also means that we may be able to conduct a clinical development plan with fewer trials in fewer patients.

Ultimately, by evaluating combination therapies earlier, we aim to bring down development timelines and reduce the costs of bringing a product to market.

4. What have these tools helped us achieve?

Using the Combo tool, we’ve been able to analyse the compounds post candidate selection in MMV’s portfolio and rank them in terms of their compatibility. In collaboration with numerous partners, the SCID platform is helping us assess the efficacy of 16 of the most compatible combinations. We are in the midst of analysing the data.

5. Who have been the key players in this effort and what has it been like to work with them?

Externally, we have worked with Merck, Novartis and Sanofi on their respective compounds. They have been very happy and supportive of the work we have done with the Combo tool. For the SCID combo studies, we worked closely with two groups in Spain: The Art of Discovery, Bilbao and GSK’s Diseases of the Developing World ‘open lab’ in Madrid. The great thing is that both groups follow the same protocols, so the data can be compared between different project teams.

Internally at MMV, it’s also been a real cross-departmental effort. The discovery team, together with partners, developed the tools for assessing compounds as monotherapy. The translational medicine team then took these compounds to see how they work in combination – applying the expertise from the discovery team. The medical team provides support for the safety assessment of these different compounds. Important modelling work is also being conducted by the pharmacometrics team, who use the data to predict efficacy in humans. Preclinical studies support the clinical development path and so there is a really close interaction with the clinical science team as well, to ensure the studies answer the right questions for clinical development.

Subsequently, the IT team provided substantial support to automate the Combo tool. The work also involved an extensive effort from MMV’s legal and business development teams, who ensured all appropriate contractual agreements were quickly put in place: some of these molecules are owned by MMV, but many of them have pharmaceutical companies as their guardians. It has been a stimulating collaboration between all partners, internal and external, to develop this preclinical combination approach and I look forward to continuing the exciting work.


1. Volunteer infection studies (VIS) involve the inoculation of human volunteers with a low level of malaria parasites in a tightly controlled environment. Parasitaemia is closely monitored and the volunteers are administered a study drug 7 days later to assess its activity.