When trying to predict what will happen to pharma and biotech in 2019, let’s take comfort from Heisenberg’s Uncertainty Principle: if you know the position of something with great accuracy, you’re likely to be way out on its speed. In January 2019, we know precision medicine is shaking the US drug system to its core, robot radiologists are poised to replace the real thing and CRISPR-edited babies are here. Who knows how fast this Brave New World will become mainstream, but here are some guesses on how it might roll out during 2019.
1. Playing Catch-Up on Precision Medicine
During 2019, regulators, pharma, physicians and software engineers will continue to adjust to the new realities of precision medicine. Precision medicine involves treating a patient according to his or her disease-causing genetic mutations instead of a disease label such as “melanoma” or “pancreatic cancer.”
It’s been a massive paradigm shift. In 2017, the US Food and Drug Administration (FDA) approved the first drug (Keytruda) based on a genetic mutation rather than tumor type, closely followed by Vitrakvi. The FDA also announced that precision medicine submissions will now be handled by a newly created division, the Office of Drug Evaluation Science (ODES). In 2019, industry will be figuring out what this means for them.
Tumor-agnostic drugs are the holy grail of cancer because they allow treatments to be targeted only to patients who can benefit. One drug that works in many different diseases is also a win for the pharmaceutical industry. However, to reap the rewards pharma must first feel the pain. We will continue to see clinical trials remodelled mid-stream as biomarker-based therapy rapidly changes the playing field, as happened with Lung-MAP. Lung-MAP closed when the FDA approved Opdivo and re-opened two years later with new objectives.
In the meantime, physicians are trying to figure out what to do in the clinic. “Most oncologists are not well trained in cancer genomics,” says Dr. Richard Schilsky, Chief Medical Officer of the American Society of Clinical Oncology (ASCO). To fill the gap, ASCO broke with tradition and started its first-ever clinical trial, TAPUR, involving 113 sites and 1260 patients. The organic design of the trial – where patients are assigned off-label treatments based on their genomics – was crucial, says Schilsky: “This is done routinely in clinical practice, but the information is not gathered in a way that medicine can benefit.” The National Cancer Institute is running similar studies. Results will be presented at this year’s ASCO meeting.
With less fanfare, a quiet but unstoppable revolution is taking place in epilepsy, where there is increasing pressure to reclassify the disease, from simple “epilepsy” to hundreds of genetic-based diagnoses, such as “SCN8A encephalopathy” or “GRIN2 encephalopathy.” Individual mutations in rare epilepsies are being targeted with non-epileptic drugs, such as quinidine, based on the drug’s known mechanism of action.
2. Low-Cost Gene Analysis, Computational Bioinformatics and AI Applications
Precision medicine is now at an intersection point where the barriers are going down and the payoff is going up. Scores of bioinformatics companies have responded, and some are being snapped up by Big Pharma. For example, drugmaker Roche recently purchased genomics research company Foundation Medicine to accelerate biomarker-driven drug development.
Massive data, super-fast processing – sounds like artificial intelligence. Or is it? In 2019, arguments will continue over the definition of AI.
Dr. Muhammad Mamdani, Director of the Li Ka Shing Centre for Healthcare Analytics Research and Training in Toronto, Canada, says AI does what humans have always done – process data – only faster and with memories that don’t forget. Creepy creativity in AI, such as writing poems, is “all just complex math,” reassures Mamdani.
At one recent count, over 30 drug developers, including big players such as AbbVie, AstraZeneca, Astellas, Bayer, GSK, Janssen and Merck, are exploring AI partnerships, not only to find and validate new drug targets, but also for adherence monitoring, better clinical trial planning and operations, identification of proteins, pharmacovigilance (monitoring adverse events), drug repurposing, and prediction of disease. Many of these will bear fruit in 2019.
Clinics have been slow to allow AI software access to electronic health records (EHR) systems. However, clinical AI products that do emerge in 2019 will likely fall into three categories: image analysis, decision support and patient triage. Most significantly, AI image analysis, such as finding TB tumors, is so fast and accurate it’s driving fears that robot radiologists may soon replace the real thing.
Another field to watch in 2019 will be AI in disease prediction. For example, Janssen recently announced a partnership with Winterlight Labs to use Winterlight’s voice analysis technology to predict who will develop dementia. UK’s Francis Crick Institute recently showed that AI outperformed physicians at predicting heart disease death by deciphering the large amounts of data stored in EHRs.
“If we can predict what will happen to patients, we can manage them better,” says Mamdani. His team is working on an AI surveillance system that constantly patrols EHRs to alert physicians if one of their patients is heading for the emergency room. “AI will affect almost everyone within five to 10 years,” predicts Mamdani.
3. First Human Trials of CRISPR-Edited Therapies
There will be plenty of action to watch in the CRISPR gene-editing field in 2019. On the political front, the world watches the fate of He Jiankui, the Chinese researcher who claimed on November 25, 2018, to have created the first gene-edited babies and is now living under armed guard. Designer babies are still illegal in the US.
The first clinical trials of CRISPR-edited cells, in sickle-cell anemia, hematological cancers, sarcoma and pancreatic cancer, got underway in the Fall of 2018 and will be recruiting during 2019. A study of CAR T-cell therapy in leukemia and lymphoma plans to start recruiting in March, 2019. Results won’t appear this year but discussion of similar studies will be a hot ticket at this year’s ASCO meeting.
There will be continued expansion this year of companies offering cheaper and faster CRISPR technology. The World Health Organization estimates that over 10,000 diseases are caused by a single error in a single gene, providing almost unlimited potential for custom CRISPR kits. This year will also yield CRISPR-driven results in animal models, pathways for valuable chemicals and exponential advances in gene research.
4. Liquid Biopsies
Conventional biopsies, where doctors harvest a small piece of tissue to diagnose or monitor a cancer, are painful for patients, expensive and prone to error. Liquid biopsies – which look for cancer biomarkers in the blood, urine or saliva – may be the answer.
In prostate cancer, biomarker studies are already predicting how the patient’s disease might progress. Such tests could transform the treatment of prostate cancer, where men agonize over the decision to have surgery, with its risks of erectile dysfunction and incontinence, or watch and wait. With surgery for prostate cancer costing more than $100,000 in many centers, it’s surprising that commercially available tests such as ExoDxTM Prostate (IntelliScore) and 4KScore have remained unchallenged in the market so long. As precision medicine accelerates, we may see some strong liquid-biopsy contenders in 2019.
The theme in 2019 is that everything’s connected. This exponential growth of cross-disciplinary technologies provides unprecedented opportunities for researchers and developers and, if society can find a way to pay for them, a healthier future for all of us.