The portion that will benefit is low and not likely to get high. Precision medicine is not a scam – it does benefit patients – but it is overhyped.
Vinay Prasad, a prominent skeptic, just published a study estimating that about 8% of cancer patients are eligible for genome-targeted therapy [1] . That’s not much. Worse, only a fraction – around half – of the patients identified will actually respond to therapy. Even fewer will see an increase in survival [2] .
Precision medicine is not doing much to fulfill the National Cancer Institute’s goal to “eliminate death and suffering due to cancer” – by 2015 [3] .
The onco-optimists rightly point out that the number of patients benefitting from precision medicine is bound to go up. We will have more data, more sophisticated analyses, and better tools to ID patients who will benefit. A doubling of PM-eligible cancer patients to 15% seems like a reasonable estimate.
But here’s the thing. We’ve probably already identified the best genomic signatures. The genes that are the strongest drivers of cancer get identified first. They are also the ones that provide the best drug targets as well – simple one gene: one protein associations that the pharma industry can throw its considerable talent and resources at.
Let’s look at Keytruda. It is an exemplar of a new class of cancer drugs (the immune checkpoint inhibitors) and the first with indications approved on the basis of genetic profile rather than tissue origin. Along with other immuno-oncology drugs, it is scoring massive sales:
From The Hype for Immune-Oncology Drugs Is Still Too Hot
But the actual clinical benefit of Keytruda? A median increase in overall survival for lung cancer patients of 4.2 months as compared to standard docetaxel therapy [4] . That’s good, but not good enough to eliminate death and suffering from cancer by 2015 or any other year.
Keytruda is no outlier. Post-approval analyses of increases in overall survival for new cancer therapies show that an additional 3 months of survival is the average:
Data mostly from Assessment of Overall Survival, Quality of Life, and Safety Benefits Associated With New Cancer Medicines. Red dot is Keytruda.
We expect technologies to get better and better over time. It happens with electronics, cars, aviation, communications and just about every aspect of modern material life. It’s natural to expect the same with drugs. Natural, but wrong.
If drug technology was getting progressively more powerful, there would be a rising trend in the graph above – bigger and bigger gains with each new generation of therapies, one breakthrough following another. Most of the dots on that graph, in fact, are for drugs that were officially designated as breakthrough therapies.
But the only breakthrough I see is in the prices that the market will bear for new drugs. Compare the two graphs – which has a flat trend line, and which is rising steadily? Precision medicine targets wallets better than it targets cancer.
Footnotes
[1] Estimation of Percentage of US Patients Who Benefit From Genome-Driven Oncology
[2] Availability of evidence of benefits on overall survival and quality of life of cancer drugs approved by European Medicines Agency: retrospective c… – PubMed – NCBI
[3] New science inspires FDA commissioner Andrew von Eschenbach
[4] Immune Checkpoint Inhibitors for Patients With Advanced Non-Small-Cell Lung Cancer: A Systematic Review.
I think this article is being overly pessimistic about the new therapies. Yes, every newly approved drug might only provide merely an average of 3 extra months in overall survival, but that is not 3 extra months compared to no treatment, but compared to the best available treatment to date. And the latter in itself already offers additional months of survival compared to no treatment or compared to the best treatment before that. Yes, we’re paying more for treatments, but we’re also earning more. And searching the internet for cancer survival statistics historical analyses brings nothing but data showing how cancer survival has been increasing from decade to decade. Progress at a steady pace is still progress. And if I were to be told by an evil cancer fairy that I will get cancer and my only choice is whether I get it today or in 3 decades, I will definitely choose to get it in 3 decades and I doubt any scientifically inclined person – including the author – would choose differently.
Personalised medicine might not be a magical solutions to all cancers, but it was never promised to be. It is not only about giving the drug to the patient who we are quite confident will respond, but as importantly, not giving it to patients who we know will not respond thus saving the patient from side effects and preventing the unnecessary cash flow to the drug manufacturer’s bank account.
Felix, thanks for such a thoughtful and well-reasoned reply. I don’t truly disagree with anything you say – ultimately these issues become tradeoffs between competing values.
What I find galling about the hype-enabled high prices of personalized cancer therapy is that it ends up diverting resources from far more effective interventions. Improved access (a uniquely American problem) and public health measures would save many more years of life than all the personalized cancer therapies put together. Their impact has been minimal – the rate of deaths per cancer diagnosis has stalled out after decades of steady progress. I’ll have more to say about this in a future post.