dPlasma hacker team

Last month I wrote about being a mentor for one of the teams at the Stop COVID-19 hackathon by ConsenSys Health. My team was called dPlasma and amazingly won the first prize in the competition! 🏆

Let’s take a look at the project and its context.

Live hackathon winner announcement

I have been following the COVID-19 pharma news for a few months now in my monthly reviews and noticed that some potential treatment options have attracted more attention than others. The three most prominent are the cheap malaria drug hydroxy-chloroquine, Gilead’s remdesivir and convalescent plasma (CP) from survivors of the infection. On my radar, CP made its entry in the April 2020 review. This coincided with the start of the hackathon in mid-April. The team had already identified their goal to build a blockchain solution to improve CP supply. Perfect fit!

What is convalescent plasma therapy?

There is a lot of information online. So, I’ll keep it brief and leave some references for further reading at the end.

Plasma is extracted from healthy donors via a standard procedure called plasmapheresis. During this process the donor gives whole blood, the apparatus separates the liquid part, i.e. the plasma, from cells and other blood components, which are then channelled back to the donor. The plasma is stored for donation to patients.

Convalescent plasma therapy uses blood plasma from survivors of the COVID-19 (or other) infection as treatment because it contains antibodies against the virus. This is a passive immunisation strategy because it does not require the recipient to mount their own immune response. We essentially transfer the result of someone else’s immune reaction because the recipient’s defence isn’t well-prepared to fight the infection. This is comparable to the protection conferred upon a baby by breast milk.

Compatibility

We have antibodies in our blood plasma that are highly likely to cause complications between incompatible individuals. Plasma donor-recipient compatibility is sort of the opposite from ‘regular’ blood donations because in the latter you donate blood cells; as such, the other part of whole blood. Thus, whilst the universal blood donor has blood type O-, the perfect plasma donor has AB+.

Plasma donorDonor antibodiesMatching recipient
AB+noneall
AB-anti-Dall –
A+anti-BA+/-, O+/-
A-anti-B/DA-, O-
B+anti-AB+/-, O+/-
B-anti-A/DB-, O-
O+anti-A/BO+/-
O-anti-A/B/DO-
Blood types of matching plasma donors & recipients and antibodies present in donor plasma. D represents the Rhesus factor.

Now that we have some background, let’s get back to the hackathon project.

From ball and chain to blockchain

As far as I understand, the current supply chain of human biological material (such as plasma or organs) is convoluted and clunky. In fact, there is no ONE supply chain but several running in parallel because of disparate systems between jurisdictions.

dPlasma’s solution tackled this problem by aiming to build a reliable, central, quick & privacy-preserving global matching mechanism on the Ethereum blockchain. The project’s focus was on CP donations, but could be expanded to organ donations or other biological materials to grow a supply ecosystem.

The idea was to create a secure app for plasma donors to register key information directly relevant to the donation (e.g. blood type, age, COVID-19 test results) whilst preserving the privacy of information that’s not required (e.g. name or any other information associated with their medical record). The app prototype uses the FHIR standard for authenticated access to electronic health records. This means, for example, that app users do not need to rely on donors entering COVID-19 test results themselves in healthcare systems where such record-keeping is used, which improves data veracity.

At the other end, haematologists can register their need based on the patients they treat. Their requests are verified and validated through smart contracts. In fact, the entire design uses verifiable credentials to make sure that any submitted data and electronic signatures are trustworthy. Due to time constraints the hackathon project could not implement zero-knowledge proofs as part of the verification architecture. However, this would be an option to enhance privacy protection further if the project was taken forward.

As Dr Cheryl Pegus, one of the hackathon judges, said, the need for a hub like this has existed for years because haematologists don’t have a central place to share such medical information securely and access related services to fulfil their requirements.

In principle, this solution could be deployed across jurisdictions and enable global donations. One main limiting factor, then, would be the stability of the biological material during transport and storage.

I have blogged about the mismatch of demand & supply and the lack of incentives for organ donations before as well as past proposals for incentives and alternatives to organ harvesting. If it was possible to legally reward donors of biological material, this blockchain-based solution could be adapted to flawlessly integrate rewards paid in Ether.

What else is there?

Below are some other developments I have noticed that address the treatment of COVID-19 by passive immunisation.

Blood Drive

Out of another competition, the MIT COVID19 Challenge – Beat the Pandemic, COVID Blood Drive was born. Their aim is to increase the supply of blood and plasma donations as well as introduce best practices for data & sample collection and sharing. From their website it appears that they are setting out to be an info and connection hub for various stakeholders.

Antibody cocktails

Instead of treating patients with donor plasma, purified antibodies can be used. For examples of companies following that path, check out my May & June news reviews. Note that not all specific antibodies are effective against the virus even though they bind to it. Ideally, antibody preparation should contain neutralising antibodies. These are capable of inactivating the virus. There is some concern that non-neutralising antibodies could be harmful.

Antibody preparations can be derived from donor plasma or manufactured in the lab through other techniques. Generally, it is considered more effective to use a mix of antibodies, a so-called polyclonal product, instead of a monoclonal product that recognises only one specific part of the virus. The former is less likely to lose its protective effect when the virus mutates.

Cows

Following on from above, it appears that plasma or antibody preparations may no longer need to come solely from human donors. SAb Biotherapeutics has developed genetically modified cows that produce fully human antibodies. If proven safe and effective in clinical testing, this could reduce the need for human convalescent plasma. The bovine-produced human antibodies have been shown to protect rhesus macaques from Ebola in a pre-clinical study. This is a tremendous innovation! Back in my researcher days, there were humanised antibodies – part human, part mouse/rat. This is most certainly a leap forward.

Thanks

Many thanks to hackathon team member Dhruv Malik for advising on the technical dPlasma section of this article.

Further Reading