The specimen supply chain needs to go beyond reacting to orders and accurately anticipate demand. The industry is just starting to do this, and artificial intelligence is the key.
Precision medicine involves the development of personalized treatments for individual patients based on their unique biological composition and that of their specific disease. The driving principle is that a one-of-a-kind problem is best addressed by a one-of-a-kind solution.
Although a personalized approach makes the most sense for providing effective treatment, precision medicine is limited by its ability to scale. Each personalized treatment ends up requiring significant time and resources to develop.
The required research can also be constrained by a scarcity of materials – specifically, biofluids, tissue, and cells that very closely resemble the patient or patient class to be treated. Biobanks typically don’t store the necessary biospecimens; potential donors are small in number and dispersed throughout the world; and there are no standardized processes to collect and deliver the materials.
As with other 21st-century challenges, artificial intelligence can help mitigate the problem. It can help locate scarce biospecimens (patients or donors who can supply them) and support their collection for researchers in a timely and efficient manner. This is “just-in-time” biobanking or collecting prospectively.
The research challenge
Thanks to advancements in technology, doctors today are able to learn more about their patients than ever before.
This includes background knowledge about a patients’ genes, disease genotype, blood type, history, demographics, environment, lifestyles, comorbidities, etc. As such, precision medicine is made possible. The more data we have, the more we can theoretically tailor the potential treatment or even a cure.
Not surprisingly, precision medicine relies heavily on precision research. Precision in this case means that scientists and drug developers need to study people possessing specific, relevant traits, e.g., a particular genetic makeup, common to the patient(s) who would ultimately receive the novel treatment or cure.
Yet, rather than studying the whole patient, researchers often study biospecimens from patients having relevant and specific traits. A new, growing demand for precision biospecimens has created a big problem for biobanks and other health organizations that have traditionally supplied human biospecimens for research. Understandably, biobanks don’t always have the right specimens on hand.
For example, in the past, breast cancer researchers might have been able to complete their work using any 50 breast cancer tissue specimens. However, today, the researchers would be zeroing in on a subtype of disease.
Now the team might reasonably need 50 samples of tumor tissue from patients with metastatic, HER2-positive breast cancer, with a HER2 L755S mutation, refractory to Herceptin. Other sorting characteristics may be important as well, including patient age, gender, race, condition, severity, blood type, procedures, test results, outcomes, smoking status, family history, and more.
Finding a collection of 50 specimens with all the right traits is difficult, if not impossible. Biobanks that have been depending on revenue from specimen distribution are particularly struggling with these demands.
Meeting the precision research demand
Biobanks were designed for storage. When biobanks first emerged in the late 1990s, they focused on collecting and storing biospecimens acquired in the normal course of health care.
Specimens were occasionally shared with resident scientists. Distributing the specimens, especially outside the collecting institution, was not a priority. Many specimens languished untouched and were eventually disposed.
As the years passed and financial constrains arose, many biobanks needed revenue from their operations. As such, they began sharing biospecimens externally to defray costs. Over time, as researcher demand mounted, biobanks – as well as clinical labs, hospitals, labs, blood centers and other organizations with access to patients – went further.
To serve specific projects, biobanks participated in the prospective, or on-demand collection of specimens from large patient populations. Today, half of all specimens are collected prospectively, tapping more targeted patient populations than ever, using rich data about specimens, to do so.
With the advent of precision medicine, biobanks are again at a crossroads. Specimen collection for precision research is hard. When viable at all, it takes too long, involves too many suppliers, creates an administrative nightmare, and sometimes results in substandard specimen quality.
To solve these problems, the industry needs something more than big inventories of stored specimens and a willingness to perform on-demand collections. They need vast inventoriesof willing patients – i.e., populations from whom researchers can find the precise specimen matches for the project at hand. This transformation is an enormous opportunity for the industry.
The industry needs ‘just-in-time’ specimen supply
The specimen supply chain needs to go beyond reacting to orders. It must accurately anticipate demand and fulfill itself ahead of time with exactly what researchers will need. Call it just-in-time specimen supply. The industry is just starting to do this, and artificial intelligence is the key.
AI can aggregate and analyze extensive data about a) what specimens researchers around the world are seeking and b) which patients can be approached for specimens with what specific traits, including sex, age, race, blood type, medical conditions, lifestyle choices, treatments, tests, and outcomes, etc. The required de-identified patient data can be harvested at scale from electronic health records.
In this way, organizations have begun to combine demandand patientdata with inventorydata and artificial intelligence to better meet actual researchers’ precision-research needs, prior to the order coming in.
With increasing accuracy and efficiency, AI is anticipating demand and facilitating just-in-timeor even ahead-of-time collectionof research specimens, from carefully selected willing patients and donors. It’s similar to the way big retailers stock their regional warehouses.
As with any supply chain, the specimen supply chain will rely on a combination of tools for inventory management, demand forecasting, and supply of resources (patients and donors) from whom the product (biospecimens) can be acquired as needed.
Just-in-time specimen supply will help research flourish. More specifically, it will support precision research, which will unlock the miracle of precision medicine. And that means accelerating scientific discovery and more precision treatments and cures.
Tracy Curley is CEO of iSpecimen, an online global marketplace that connects scientists in need of biospecimens for medical research with a network of healthcare specimen providers.