Big data, cancer genomics and other “-omics”, and the interplay of cost and value will define the path of cancer research and care for the next 15 years, said Allen S. Lichter, MD, FASCO, CEO of the American Society for Clinical Oncology (ASCO), to a group of about 300 researchers and clinicians gathered for the University of Colorado Cancer Center Fall Retreat 2014.
First the good news: according to Lichter, the field he abbreviates as “-omics” that includes genomics, epigenomics, proteomics, and many more, is allowing and will continue to allow the discovery of new treatments targeting the very specific differences between healthy and cancerous cells.
“We’re moving toward a day in which every cancer can be matched with not just one, but a combination of targeted treatments,” Lichter says. By 2030, Lichter expects that a significant share of all cancers will be molecularly well-understood and highly treatable (see also the ASCO report Shaping the Future of Oncology: Envisioning Cancer Care in 2030).
But new treatments come with new challenges, a major one of which is our ability to pay for these treatments. Conservatively, a regimen of three new drugs could cost $350,000 per patient per year.
“Unless we dramatically adapt the cost, value and who we treat with these drugs, our system of oncology care will collapse,” Lichter says.
The challenge is even more acute in the current environment of decreased funding and rising costs. In response, Lichter quoted Ernest Rutherford, saying, “Gentlemen, we have run out of money. Now we have to start thinking.”
One necessary direction of thought is the difficult decision of how to best prescribe cancer treatments that will do the most good while keeping costs reasonable. Lichter showed slides of inflammatory media articles decrying efforts to “ration” drugs, but, he says, ASCO will continue to produce and publish guidelines for the best use of therapies to ensure efficient use.
Lichter also points to the need to decrease the cost of clinical trials. For example, due to budget constraints, in 2015 the National Institutes of Health will cap cancer clinical trial enrollment for studies run through cooperative groups to 17,000 patients, of whom 5,000 will be pediatric patients. This is a tiny percent of the approximately 1.4 million patients who will receive cancer treatment in the United States this year.
Lichter then described the difference between an innovative Norwegian cardiology study published in the New England Journal of Medicine with an enrollment cost of $50/patient, compared with the average cancer clinical trial cost of about $50,000/patient.
The difference, Lichter says, is that the Norwegian study was able to draw its conclusions from an existing data resource; the information of cardiology patients treated in Norwegian hospitals had simply been added to a database describing their treatments, demographics and outcomes. From this existing database, researchers could see that it didn’t matter whether a blood clot was removed in conjunction with the insertion of a stent or whether the clot was left in place – a powerful conclusion from an inexpensive study.
Similarly, Lichter spoke about the necessity to look beyond clinical trials for data and into the world of clinical patient data. Instead of paying $50,000 per patient enrolled in a clinical trial, oncology researchers could learn from Norwegian cardiologists and pull data straight from patient charts. The massive amount of data generated from 1.4 million patients as opposed to 17,000 per year would accelerate the pace at which we discover what works in what settings, while decreasing the cost of this new knowledge.
ASCO’s CancerLinQ, currently in the planning stages, will be this data resource. Lichter described a feasibility study for the project meant to collect what seemed an ambitious number of 30,000 breast cancer patient records into a database that would allow mining this data for (among other things) treatment recommendations. The study was forced to cap its data gathering at 140,000 records when it ran out of funds needed anonymize the patient data.
“There are challenges,” Lichter says, among which is the need to standardize the ways oncologists evaluate, code and report data (e.g. “disease progression” doesn’t mean the same thing in all cancers and in all settings), but Lichter sees a path forward for CancerLinQ toward the goal of collecting data for all patient-oncologist interactions in the United States.
Speaking of oncologists, Lichter hopes that as cancer care becomes more “rules based” – i.e. driven by treatment recommendations based on our understanding of patients, data and outcomes – routine cancer care may be provided by primary care physicians and not necessarily by oncologists, further decreasing the cost of care and also alleviating the projected shortage of oncologists. As precedent, Lichter spoke about the similar transition from specialist to primary care of treatment for hypertension and preventative cardiology medicine.
Outside Lichter’s three main points of targeted treatments based on “-omics”, big data, and cost/value, he spoke briefly about the promise of immunotherapy, which, he said, “has been promised every year since I finished my training, but is now finally taking shape before our eyes.” And he spoke about the need for new and more training of data scientists capable of keeping pace with the amount of data generated by “-omic” and patient sources.
Lichter closed by emphasizing the need for cancer researchers and clinicians to control the direction of oncology research rather than giving in to the pressures that “buffet” the field (presumably but not stated as such including legislative, media and pharma).
If Lichter’s vision of oncology in 2030 proves accurate, we will know much more about cancer and be able to do much more for patients fighting the disease. The challenge is dramatically adapting the ways we test treatments and pay for their use.
“By anticipating the future, we can shape it,” Lichter says.