In France, a team of researchers, doctors and engineers are developing software to improve kidney transplant patient's follow-up

Six university hospitals. Three leading companies in the IT and medical sectors. Hundreds of volunteers. Thousands of data points...

These few figures show the scale of KTD-innov, a public-private consortium aiming to gain a better understanding of kidney transplants and then develop diagnostic and prediction software for use by doctors and patients alike. How do the researchers involved work, and what phase of the project are they in today? 
​
Magali Giral (professor of nephrology and director of clinical research at Nantes University Hospital) was kind enough to answer our questions.

First phase: developing algorithms to better predict the evolution of kidney transplant

KTD-innov, the project led by Dr. Sophie Brouard, CNRS research director at Nantes University Hospital,  and Prof. Alexandre Loupy, nephrologist at Necker Hospital and researcher at Inserm, is being carried out in two phases. The first part (KTD 1) is coming to an end at the moment, while the second part (KTD 2) is getting underway.

KTD 1 consisted of collecting biological samples from 750 patients, to validate a large number of biomarkers, that are already been used independently by hospitals and that have never before been treated together. “It’s a holistic approach”, says Professor Giral. “It is the combination of biomarkers of interest that will enable us to better diagnose renal graft rejection.”

The biomarkers studied fall into two categories. On the one hand, biomarkers are collected through non-invasive sampling such as blood or urine collection. Information that would indicate graft rejection is sought through the analysis of antibodies, proteins and genes. In addition, some biomarkers are derived from invasive tests that require a biopsy. For example, the expression of certain genes can be observed locally.

All these samples are currently being processed and analysed by the consortium in order to build algorithms that will make these data ‘speak’ by exploiting significant correlations between these biomarkers. The work is at the frontier of artificial intelligence and Big-Data. 
The first KTD 1 study results are expected by the end of spring.

It is not just a matter of producing algorithms; they must also be available and easily usable by the people concerned. On this subject, Prof. Giral tells us a little more: “Our aim is to produce  a software with a twofold function. On the one hand, it will be a decision-making tool for the clinician. On the other hand, it will be a tool for monitoring and supporting the patient". The software is being developed by Pierre-Antoine Gourraud’s team at the Nantes University Hospital and published by Sigma, a French company responsible for designing the future programme and its graphic interface.

Second phase: checking whether the algorithms can provide results within a reasonable time

Researchers are convinced that the software will provide relevant analyses, useful to both patients and practitioners. But will it be able to do so quickly enough? This is what KTD 2, a feasibility study, is seeking to verify, and Professor Giral summarises the issues at stake: “When there is a suspicion of rejection, clinicians need a result in a short time. If we provide them with this information two years later, it's pointless!” 

Indeed, the whole process takes time: there is a lot to do between the time of collecting the samples from the patient, sending and analysing the samples in the laboratories, compiling the data and then processing this information via the software... Can all these operations be carried out in a reasonable time frame for the patient and his nephrologist? “We want to prove that we are capable of doing it in 20 days,” says Prof. Giral.

In addition to this main objective, KTD 2 has secondary objectives:

• To validate the relevance of the algorithms developed during the first KTD 1 phase of the study, on a new cohort of 300 individuals.
• To present the software to practitioners to gather their opinions. Do they consider the data to be useful? Will it have an impact on medical decisions?
• Similarly, to present the software to patients. Does the data give them a better understanding of their situation, the risks of transplant rejection, and the treatment strategy that is used in their case?

In 2022, once this second phase of the KTD-innov study has been completed, new research will certainly be conducted to measure the real impact of the software on the transplants’ evolution.

A collective research made in France

For Prof. Giral, the collaborative aspect of the science involved in this project is an obvious added value: “It means that we don’t have to do what other research teams have already done. We are pooling our strengths and brains; we are thinking together with ten people rather than alone. Finally, we pool resources, which are increasingly difficult to mobilize in research".
This synergy is all the more powerful as the KTD-innov project is based on a network of well-established people and institutions: “We are used to working together, we know each other, we know who does what. With some, like the CHUs of Lyon or Paris, we have been sharing biocollections for 15 years and databases for almost 30 years!”

Innovative and ambitious, KTD-innov is also a French project, 100% financed by the State, with a total €5-million budget.
Moreover, for Professor Giral, the French context is particularly favourable to research in this field: “We are structured, we have common rules. And the patients’ profiles are both diverse and well distributed throughout the country. This is important because the results that will be produced by this project will be applicable throughout France.” 
And not necessarily applicable abroad? “This is yet to be demonstrated,” admits Professor Giral. Alexandre Loupy, coordinator of the KTD-innov project, is already working on this with EU-TRAIN, a project similar to KTD-innov based on European data and collaboration.

Before concluding, Prof. Giral wishes to acknowledge researchers’ and doctors’ participation in the project, in particular Emmanuelle Papuchon and Maud Racapé, who, together with Inserm Transfert, coordinate the activities of the clinical research teams of all the CHUs, as well as biology platforms and industrial partners.