What History Tells Us About Vaccine Timetables
The time needed for the technical development of a vaccine is almost invariably and grossly underestimated
When will there be a coronavirus vaccine?
More than 30 years ago, I was asked the same question about an AIDS vaccine. Being a realistic optimist, my answer was: I do not expect a vaccine earlier than five years from now. None of the many journalists who asked me that question ever quoted those five years. Far too pessimistic! Other scientists spoke about one or two years. They were cited in the newspapers (and I was no longer harassed by journalists).
Three decades later, we are still waiting for an AIDS vaccine effective enough to be licensed.
I learned a valuable lesson then. Today, when I get asked the same question about the novel coronavirus, I respond by repeating the question with a different emphasis: “When do WE (all) get the coronavirus vaccine?” And there are other important questions to be added: Will it be safe? Will it be reliably protective?”
An average vaccine takes about 10-12 years to be developed. This does not include those vaccine projects that did not make it to market but it does include a number of “me too” vaccines (vaccines undifferentiated from those already on market.) Those me-too vaccines are clearly the majority among past vaccine development projects and greatly influence the average time needed.
The figure below provides an overview on normal vaccine development, its phases and its average duration.
For comparison here is how coronavirus vaccine development is being explained to the general public:
“Make some milligrams of the desired antigen. Can be done in few weeks or months. Or, if you aim at a DNA or RNA vaccine, it may even be synthesized within days. Immunize mice with it and test the serum for antiviral antibodies. If the serum contains virus-neutralizing antibodies, you have your vaccine.”
Is this really a vaccine? No, it is only a potential vaccine candidate—one of about 100 candidates with a <5% chance to make it to the market.
The questions we ought to be asking are: Is there a process to manufacture this “vaccine” at large scale? Can it be made reproducibly at the same quality—and of what quality? How can we purify it to a degree that unwanted components are avoided? Which methods are required to analyze raw materials, intermediates and the final vaccine? How much investment is needed to set up the required facilities and to qualify these for GMP (Good Manufacturing Practice)? Or do we find an experienced contract development organization?
Complex questions from a different world—totally different from an academic research environment. No wonder that the time needed for the technical development of a vaccine is almost invariably and grossly underestimated. In practice, the technical development phase is not even a defined phase with an average duration; it marks the start of development and does not end until a vaccine is licensed. It continues throughout the entire clinical phase and often beyond.
In an urgent situation one might choose the fastest way forward: Take established technologies and methods, make about 1000 doses of the vaccine and then check if it is safe in laboratory animals and stable with regard to essential characteristics.
If so, apply for a Phase I clinical study, recruit a few volunteers and test different doses to see if this vaccine can be applied without causing undesired or unexpected side reactions. All this can be accomplished in a rather short period of time and the chances of success until this phase are quite reasonable. But we are still far away from a real vaccine!
On average, and including the high number of “me-too” vaccine projects, about 50% of the vaccine candidates pass Phase I successfully. Phase II clinical trials are much more demanding but still ~ 30% of vaccine projects survive this phase. However, the majority of vaccine development projects end here.
For new indication vaccine candidates, less than 5% will proceed to Phase III clinical trials. These are the ones with promising clinical data, competitive technologies and reasonable cost of goods—relevant factors to justify the high expenses expected for the clinical Phase III. A vaccine for Phase III studies must be identical to the desired end product. If not, the ultimate Phase III clinical proof-of-safety and efficacy under practical conditions will have been in vain because the tested vaccine is not the same as the final one.
For a coronavirus vaccine we need millions of doses. To do this you most likely need a dedicated facility: a new building, new equipment, facility qualification, process validation, trained operators which might take several years. Let us hope for a technology that fits well into an existing vaccine plant which is not needed for other life-saving vaccines.
So make your own estimates on how long it might take until there is a coronavirus vaccine for millions who want it. Be optimistic! Assume that every technical work is successful right away. Assume that the required manufacturing technology is already established and can be applied and scaled up as required. Assume that the antigen candidate does induce protective responses and no unexpected adverse effects. Assume that unlimited resources and money is provided. You may also correctly assume that licensing authorities will be very pragmatic and treat a coronavirus vaccine with absolute priority. However, please do not assume that the vaccine will be accepted, if one cannot prove that it is safe, effective, and of reproducible quality.
In other words, any substance, which when injected into research models, produces a scientific paper (or a news story in the press) is NOT yet a vaccine!