Data sharing is making a difference
Data sharing is becoming a fundamental part of clinical research throughout the world. We take a look at how WWARN, the first malaria data sharing network, has used pooled analysis of shared data to provide evidence to help improve dosing regimens of malaria treatments.
Why do we need to share data?
Over the past two decades, the sharing and pooling of research data has become critical in the advancement of many different scientific fields, and a requirement of some funders, regulatory agencies and scientific journals.
The need for data sharing within the tropical medicine research community is no different. In fact, data sharing might have even more value in tropical medicine because of the lower volume of data available, the substantial public health impact of ensuring evidence-based policies and practices, and the limited interest from commercial organisations in these poverty related diseases.
WWARN and shared data
Determining how effectively an antimalarial medicine performs is complicated. It requires treating hundreds of malaria patients with a drug and then following the response of each patient for a month or more in clinical trial settings. Clinical trials become even more powerful if the results can be combined and analysed across target populations, regions, countries and even continents, and include data on clinical response as well as drug concentrations, laboratory markers of resistance and drug quality.
WWARN was launched in 2009 to be a scientifically independent data sharing platform, supporting changes to policy and providing information to help slow or stop the spread or emergence of antimalarial resistance. The Network’s aim was to facilitate data sharing so that specific scientific questions could be investigated.
To accomplish this, international researchers agreed to contribute their clinical trial results to the WWARN Data Centre. WWARN then developed standards for data sharing, secure methods for online data submission, and approaches to harmonise diverse data formats. The results from all of these individual patient trials could then be combined into large data sets for pooled analysis.
This systematic approach took several years to develop, given the complexity of the patient data, the diversity of data sources and the challenging scientific questions being asked of the combined data sets.
The very first collaborative analysis - exploring the efficacy of dihydroartemisinin-piperaquine (DP) - examined responses of 7,000 patients from 26 individual studies. Though only 136 of these patients were not cured, closer analysis revealed that small children were three times more likely to fail treatment because they were not receiving a high enough drug dose. These findings were supported by PKPD models (Tarning et al. submitted).
The impact so far
Since the establishment of WWARN in 2009, research partners from over 260 institutions globally have worked with WWARN, and over 120,000 individual patient records have been contributed to the WWARN Data Centre. That equates to around 80% of all the available artemisinin combination therapy trial data.
Based on the results from the Dihydroartemisinin-Piperaquine (DP) Dose Impact Study Group and pharmacometric modelling of piperaquine, the World Health Organization has revised the recommended dose of DP, a commonly used antimalarial for young children. These revised dose regimens are predicted to provide similar piperaquine concentrations across all age groups.
Similarly, a meta-analysis undertaken by the Artesunate Amodiaquine (ASAQ) Dose Impact Study Group, based on 9,106 patients, found that although the overall efficacy of ASAQ is adequate in most settings, efficacy varies with the formulation and is affected by a range of risk factors including age. The Artemether lumefantrine (AL) Dose Impact Study Group found that cure rates were lowest in young children from Asia, especially those with high parasitemia, and young underweight children from Africa.
WWARN’s innovative approach has enabled the malaria community to ask practical, life-saving research questions using the pooled results and cross-disciplinary analyses to identify and understand common trends, particularly in populations who are likely to be most vulnerable to poor treatment outcomes. This is only feasible through an approach using individual participant data meta-analysis, as any single study lacks the statistical power to examine the many different risk factors, and meta-analyses from aggregated information do not provide sufficient detailed information to make accurate comparisons.
The results of meta-analyses are helping researchers to recommend adjustments to treatment approaches that can delay the emergence or spread of any resistant parasites and prolong the lifespan of available antimalarial medicines.
Recently a report was published detailing a summary of a workshop held in South Africa in March 2015 - Sharing Research Data to Improve Public Health in Africa. The workshop brought together public health researchers and epidemiologists primarily from the African continent, along with selected international experts, to talk about the benefits and challenges of sharing data to improve public health, and to discuss potential actions to guide future work related to public health research data sharing.
At the event, Karen Barnes, Head of WWARN’s Pharmacology Group and professor of clinical pharmacology at the University of Cape Town, presented on WWARN’s efforts to bring the antimalarial research community together to make malaria treatment more effective.
“The meeting highlighted the critical need to establish fair conditions of use, and mechanisms to ensure that the contribution of those generating the data is properly recognised. This is particularly important for researchers in low and middle income countries and is something that WWARN has worked hard to accomplish,” says Barnes.
WWARN strongly encourages those who share their data to also participate in any project where that data is used. Those involved in a project receive authorship and those who have contributed data are cited as contributors in any publication that results from the use of their data.
What lies ahead?
The WWARN partners continue to gather new study results to update the Data Centre and provide evidence to inform new drugs in development and to advocate for better quality and optimal dosing of medicines. Several more large-scale analyses of antimalarial medicines are currently in progress. WWARN is also adapting the WWARN data management system for other tropical diseases such as Ebola and visceral leishmaniasis.