Coralie Alabert, Marine Baixas, Clara Barnouin, Chloé Pontreau, Alice Vadurel


Diabetes is a disease caused by the dysfunction of the pancreas, which becomes incapable to regulate the rate of sugar in the blood in a natural way.

For a healthy human body, the pancreas secretes a substance, insulin, which is in charge of maintaining the level of sugar in the blood after every meal, keeping it between 80mg and 130mg for every decilitre of blood[1].

For a diabetic person, the pancreas does not work. Consequently, the sick person has to monitor his/her glycemia[2] to know the rate of sugar in the blood and adapt the insulin injection accordingly. The dose is injected artificially in the body via the stings of insulin or the insulin pump.

The classical treatment for diabetes, such as the couple glycemic-injections or the insulin pump, is a burden in the daily life and restricts the diabetics with permanent constraints: monitoring the sugar level, 4 to 6 injections of insulins per day, hypoglycemias, hyperglycemias etc. … As the start-up Diabeloop demonstrates, the technology seems capable to answer the challenges put forward by the disease. The company, with its artificial pancreas, collaborates medicine and technology. It answers the daily problems of diabetics by related products.



According to Pierre-Yves Benhamou[3], doctor in charge of the diabetology department of the CHU of Grenoble, Diabeloop has been innovating for 30 years and has been in the line with the innovations for treatment of diabetes. Some patients like the feeling of control and to decide on everything (dose, meal) while for others, this chronicity is unbearable. They adopt a practice of comfort which does not allows good control.

Dr Benhamou underlines that Diabeloop is a real breakthrough and represents a considerable contribution to handle the diabetes. Indeed, the treatment proposed by Diabeloop eliminates the stress which constraints the diabetic person to switch to the machine that controls the rate of sugar in an autonomous way. Thus it relieves the life of the patient which trusts it and even claims to do better than what the patient would have been able to do.

How does this new treatment works?  It comes in 3 elements: glycemia sensor, insulin pump (directly stuck on the skin of patient) and the terminal similar to a smartphone which checks the connection between this sensor and the insulin pump.




First of all, the process consists of a Dexcom sensor, an applicator applied on the skin and of a 7mm thread which goes under the skin. This thread of the size of a hair measures the interstitial glycemia, which is the blood glycemia with ten to fifteen minutes of delay. This thread sends back the data to a terminal which looks like a smartphone with a SIM card. This totally locked terminal needs a huge power processor and therefore cannot be the phone of the user, for safety reasons and the weight of the data. The data stored in the SIM card will only be accessible by the patient and his doctor (access to the summary table, the improvement suggestions by the device…). The terminal, according to the information received by the sensor of glycemia, orders the insulin pump to adapt the dose of insulin to be sent. The reserve of insulin contained in the terminal lasts three days.[5]

Dr Pierre-Yves Benhamou underlines that, in spite of a strengthened autonomy in daily life and a management of the rate of sugar completely left to the machine, constraints remains in the replacement of the different pieces of the equipment : every three days for the reserve of insulin in the pump, every week for the thread of the sensor of glycemia even if the ambition is to be able to change it every two weeks and finally, there also has to be a supervision for the terminal battery which for instance has a capacity of two days.


In terms of performance, Diabeloop demonstrates the efficiency of its product. By performance, one understands efficiency as the capacity to regulate the diabetes.

As Dr Guillaume Charpentier, president of Diabeloop and the CERITD, details: “the algorithm plans what is going to take place in a more reliable way than what the patient would do on his own. For an even more precise treatment, the patient can improve this forecast “by warning” the machine, if possible 2 hours in advance, when he is going to eat or to practise a particular physical exercise”[6]. Thanks to this continuous control and this automatic way of insulin delivery, which only gives the necessary dose, the hazards caused by the disease are reduced.

The performance is illustrated by the clinical trials of the prototype in 2014 [7] :


The artificial pancreas is able to reduce hypoglycemias by half and completely stop the night-hypoglycemias. According to Pierre-Yves Benhamou, thanks to this artificial pancreas, the patient is going to keep a good range 75 % of the time while without this automated system, he would spend only 50 to 60 %. There is thus, a good progress. We would move from 30 therapeutic decisions (such as: I think of my diabetes, have I eaten enough sugar, etc.) to only 5 information given to the system per day. It is thus a better solution for the treatment of diabetes but not a cure. The algorithm cannot actually manage all the events of the everyday life on itself. That is why it will be necessary to continue to give information to it.

However, in spite of a real performance, the way towards the market is complex and the reliability of the product requires to be proved by numerous tests. For Diabeloop, the pre-marketing tests are very encouraging.


Diabeloop began the clinical tests 4 years ago. The tests were not conducted on animals because the insulin used in their product is the same as the one already on the market. This stage was substituted by an electronic stage, which is authorized for the pre-launch tests on the market. The purpose of this first test was to create a kind of database to allow the improvement in the algorithm. The pancreas cannot be tested in fact, but the evolution of the glycemia at the time of the meals can be registered thanks to the samples of blood. Thus it allowed to paint a portrait of glucose on the computer and mathematical models were developed from the results, sorts of simulator to test the new algorithm.

Until today, 3 tests were conducted. The first stage of the clinical tests was patients’ recruitment. Several criteria were established: the patient has to have a type 1 diabetes, be handled by insulin for several years, be autonomous towards the management of the insulin pump, have had no complication and be obviously motivated.

For the first test, each of the 18 patients tested the product during a meal. Only one person reacted badly to the test, which was already very encouraging and allowed to improve the first version of the algorithm.

During the second test, 15 patients tested the new algorithm for more than 24 hours (meal + at night). The treatment was considered more effective than a classic treatment without creating more dangers.

Now the third test is conducted on 45 patients. Every patient tests the product during three days and it is aimed to validate the algorithm.

This series of tests is however every time made in private hospitals, over a short period, under professionals’ supervision. Next year there will be an ultimate study on 50 patients who will test the product during three months at home. Diabeloop will study if new technological limits appear and must be overtaken. The company will also test the resistance of the product, in particular the battery, and all what could happen in the practical life too (shock, interference with other electronic objects). Indeed, this technology is based on a machine or an algorithm, which is not infallible. In the field of medicine, it is thus necessary to make sure that in case of dysfunction, the life of the patients is not endangered. That is why Diabeloop sets up a system in case of failure.

If the terminal does not work, the pump returns to its reference flow which is prescribed by the diabetologist. There will not be self-regulation anymore according to the direct data but the regular regulation. In regard to the pump, no dysfunction is possible because it is directly implanted in the skin.

These studies show that Diabeloop runs a lot of tests to obtain the highest degree of possible reliability to improve its product and solve some existing problems. Besides, the injected product is already on the market, so for that, the reliability is not to be proved.

The reliability is an indispensable condition to penetrate into the world market of the diabetes which represented more than 40 billion euros in 2013[8]. Given that, every year, approximately 7 million people are diagnosed for diabetes [9], this market is increasing. There is a huge opportunity for the pharmaceutical companies to be taken, that has already included some companies including Diabeloop. Nevertheless, it is still necessary to master the complex technology.

According to the matrix of Teece [10], the technology possessed by Diabeloop is situated in the area « blocking-alliance ». That is to say: that it is protected from the competitors (possibility of obtaining patents[11] and mathematical algorithms which are hard to imitate). It is the reason why Diabeloop has only one single big competitor for the moment, Medtronic. Secondly, it is essential to create an alliance with the medical profession (hospitals, drugstore …) or to create partnerships with the main actors on the diabetes market. Which is why Diabeloop is in discussions with Air Liquide, a service provider for 80 % of the French market[12]. The objective is, on one hand, to take over the market and to evict the new potential incomers, and on the other hand, for a company as Diabeloop, to take advantage of infrastructures such as hospital centre, which they could not possess.


Thanks to this matrix we can thus suppose that from its marketing strategy, Diabeloop will know how to give way on a market already established by many actors. This success will be due to the innovation of the product but also to its reliability, guaranteed by the alliances with CHU, which brings in a notoriety to the project. Besides, the patented and complex technology of Diabeloop make the imitation complicated and guarantee to the company a distinction compared to its competitors. Diabeloop does not benefit from a big name in the medical area but will undoubtedly know how to give way on the market.


Its integration on the market is eagerly awaited by the diabetics. It is well described by this testimony collected on the French federation of the diabetics: “my controls are made from 6 to 7 times a day. That’s why I look forward to this new system of glycemic controls without pricking myself on fingers. But as usual, the French system is 10 years late. “[13]. Numerous testimonies confirm these words.

To whom should we address this treatment then?

In the first place, the innovators [14] of the technology Diabeloop, once launched on the market, could be testers. Indeed, they are well informed and like being the first to influence the other potential patients.

The early adopter is a category of consumer who is very likely to act favourably in front of a new product and to be among the first ones to spread it on the market. He would be a young man or women, from 18 to 30 years with type 1 diabetes and worried of protecting his health. He wishes to live a normal life, to practice sport, be free of his movements, and especially not to be forced by the usual treatment of diabetes. Furthermore, he reacts favourably to those new technologies.

The late majority describes the person who will trust the product when everybody will use it. He waits to be sure of its reliability and efficiency before using it.

Finally, laggards are the people reluctant to possess an insulin pump. It is often a psychological barrier as far as it affects the private life and asks for a capacity to assume this not very aesthetic equipment. Today, according to Erik Huneker[15], only 20 % of the diabetics of type 1 carry an insulin pump. We suppose that laggards would be more numerous among the old population. The latter are used to their treatment and do not like to change, especially concerning new technologies. However, according to Dr Pierre-Yves Benhamou[16], it is difficult
to fully categorize laggards.



The adoption of Diabeloop is essentially correlated with the refund by the Social Security. This refund is a binding but essential stage for the total adoption of this treatment.


The attainment of the CE[17] marking is also an administrative constraint and a compulsory stage for the product to be launched on the market. At the beginning of 2017, thanks to the ultimate study in ambulatory, it should be obtained. Diabeloop chose to acquire ISO certification 13485[18] which is optional but underlines quality security. Plus, to guarantee an exclusivity, Diabeloop is working to earn about ten patents.


Logistic constraints also take shape. As the product tends to develop, the start-up has to be capable to handle the demands. If 300 patients call, issues with training as well as scalability arises.

The financial limits must be also underlined. Diabeloop needs around 4 to 5 million euros to finish the development and the launch of the product in the market.

They receive state subsidy such as the one given by the alliance Aviesan, which gathers the main public research bodies. However, it remains insufficient.

Thus they are currently searching for financing options which should be found before the end of year 2016. This financing can be in shape by partnerships, allowing to develop a distribution network for the product. Thus the solution to exceed the financial limit is to reach agreements with the key actors of the diabetes.

Diabelopp stands out as a challenge in terms of evolution: the miniaturization of the process, in particular the miniaturization of both glycaemic sensor and insulin pump. This evolution would allow to remove the psychological barrier against wearing devices daily that some laggards consider too visible. Hence, the miniaturization would be the ultimate stage of comfort for a diabetic which would almost be unnoticeable. Dr Benhamou compares this possible evolution to the evolution of the cellphones. It is true that the first cellphone was thick and heavy, today they are thin and light thanks to the miniaturization of components. The similar evolution is aimed by Diabeloop.[19]

Technology can be seen as a simple stage towards the total autonomy too, because in the end the objective of the medical profession is to allow the patient to live completely without it. Diabeloop would be then a stage towards the complete autonomy. This approach of the technology as a simple stage is consolidated by Dr Pierre-Yves Benhamou.

Actually, he works enormously on researches on cellular therapy which aims at transplanting the generative cells of insulin within the body of the patient. Nevertheless, the transplant remains risky because it can have heavy side effects. Restoring the capacity of insulin creation would change the diabetic into a “normal” and autonomous person, but it still demands a lot of research despite encouraging tests.

To conclude, one of the major stakes in the evolution of the treatment for diabetics is an egalitarian access to the treatment. It is even the first objective according to Dr Benhamou. Unfortunately, the access to the novelties in terms of treatments, artificial pancreas or transplant, are first reserved for the patients whose diabetes is unbalanced. It was the case in particular for the adoption of the insulin pumps which are now even more widespread.



[1] French federation of the diabetics

[2] Rate of sugar in the blood obtained after test of glycemia. The diabetic pricks himself the fingertip to take a drop of blood which will be deposited on the device and obtains his rate of sugar.

[3] Interview of Pierre-Yves Benhamou, 13th April 2016, person in charge of the department of diabetology of the CHU of Grenoble

[4] Website of Diabeloop :

[5] Interview of Erik Huneker, 17th April 2016, cofounder of Diabeloop

[6] Article of CEA, « Diabeloop, le nouveau pancréas artificiel français : une révolution très attendue par les diabétiques de type 1 », 8th Mai 2016 :

[7] Same

[8] Xerfi, introduction of the study market « Le marché du diabète », 2013,

[9] Same

[10] David J.Teece (1986), « Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy »

[11] Article 6 of the law n°68-1 of January 2nd, 1968 on patents: ” are patentable the new inventions implying(involving) an activity creative and susceptible to industrial application “.

[12] Interview of Erik Huneker, 17th April 2016, cofounder of Diabeloop

[13] Testimony collected on the site of the French federation of the diabetics, « Je suis diabétique insulinodépendante depuis l’âge de 14 ans. », 2016 :

[14] According to the broadcasting theory of innovation of Everett Rogers (1962) which represents the categories of individuals according to their expectations and to their reactions to the innovation.

[15] Interview of Erik Huneker, 17th April 2016, cofounder of Diabeloop

[16] Interview of Pierre-Yves Benhamou, 13th April 2016, person in charge of the department of diabetology of the CHU of Grenoble

[17] Compulsory label on the European territory for a medical device. He obtains when the product answers a particular specification.

[18] The standard ISO 13485 specifies the requirements of the management systems of the quality for the industry of medical devices. She leans on the requirements of the more general standard ISO 9001 2000, in the context of this industry. ISO certification 13485 is not compulsory to market medical products and does not guarantee absolute of their quality or their safety. However, she gives the indication that the company follows a procedure recognized by an independent body which audits him regularly. The certification also simplifies the marking IT of the products marketed by a company, in particular by allowing internal self-certifications governed by the organized procedures.

[19] Interview of Erik Huneker, 17th April 2016, cofounder of Diabeloop

Leave a Reply

Your email address will not be published. Required fields are marked *