Musings of a Cancer Doctor

Wide-ranging views and perspective from George W. Sledge, Jr., MD

Saturday, December 29, 2012

Mega-Trends in Oncology

It’s the end of 2012, and the Mayan Apocalypse failed to materialize, which is good news for my readers, who will continue to enjoy my pontifications for years to come.  But our narrow escape from doom left me in a thoughtful mood.  I’ve been thinking about our collective future.


Where are we headed in oncology? Not just the Fiscal Cliff and the Affordable Care Act, as pressing as they seem today, but the bigger trends, the long reads in cancer’s playbook. Let’s call them the Mega-Trends. I’m sure all of us could generate a list, but here are some of mine.


Our Demographic Destiny:  The first trend was locked in place decades ago. The Baby Boomers, the product of the post-World War II surge in births, are now reaching an advanced age, as my aching joints remind me on a daily basis. Cancer is an age-related disease. There will be a surge in cancer incidence over the next decade, and with that surge an increased need for oncology services.


A great deal follows from this simple fact. We will have more cancer patients than warm bodies to take care of them. The cost of cancer care to society will climb significantly. And this is true even before you factor in the many new (read: obscenely expensive, usually non-curative) treatments headed our way.


How we deal with insufficient personnel resources and rising costs is partly under the control of our masters in Washington. But deal with our demographic destiny we must, and I would prefer that the solution arise from those of us who actually treat cancer patients.


The Survivorship Dilemma:  File this one under “the problem with success.” Increasing number of cancer patients plus improved survival rates for those patients (already apparent in several diseases) equals increased numbers of cancer survivors. And the number of survivors is surging.


Cancer survivors are not the same as appendicitis survivors. They have a greater burden of treatment-induced complications. They are at greater risk than the general population for subsequent malignancies. They require greater use of ancillary services (e.g., genetic counseling). And they often bear the emotional and financial scars of their disease.


We are still learning how to deal with the survivorship dilemma. At a simple and selfish level, who takes care of these patients in the personnel-constrained environment we face as a result of the demographic changes mentioned above? It’s not like we have an excess of primary care physicians and nurse practitioners just sitting around.


The Information Era:  We might also call this the Genomic Era, since the DNA code is, ultimately, just a 3.2 billion-long string of numbers. If every patient, a decade from now, undergoes ubiquitous, cheap genomic testing (host and tumor), the data load will be immense, and the comprehension chasm deep. We don’t currently train oncologists in information technology. We need to think how to do this, and we need to start yesterday.


Maybe I’m over-estimating the depths of the problem: if I can describe a patient with a few data points (ER-positive, HER2-negative, 2cm, 2+ LN), do all those other data points really matter? Yes, I think so: the trend, a decades-long trend, is for logarithmically increasing data loads. I think that trend will continue, and probably accelerate.


Measuring, Smaller and Better:  Part of the information load comes from improvements in measuring tumor burden. I date to an era dominated by plain film chest X-rays and liver ultrasounds, and (I really date myself) physical examination.  The old technologies were crude and not particularly reproducible.


This continues to change. I cannot even begin to guess how we will be measuring tumor location, volume, and function a decade from now. But of this I am certain: we will, sometime in the next decade, become capable of measuring sub-millimeter tumor deposits.


This will be a mixed blessing, as all recognize.  What do you do when, in 2022, you find a clump of eighteen cancer cells in the liver of a colorectal cancer survivor?  Are those cells dormant micrometastases? Will they progress to overt metastatic disease? Are they sensitive to any particular available drugs? Do we need to change our current (non)surveillance strategies? I don’t have any answers, but I will be surprised if this problem does not emerge from evolving imaging technologies. And, one hopes, the solution as well.


Global Integration of Cancer Care:  I know, this one started long ago, but I doubt it is finished. Globalization has already affected drug development and distribution. Dictations are processed in Bangalore while we sleep, and CAT scans are increasingly likely to be read wherever they may be read most cheaply.


Thank God my patients still need to see me in my clinic, right? My cognitive services are still important, and it’s not like my patient will fly to Beijing every week for a paclitaxel infusion, right? Well, maybe, maybe not. The infusion part of the equation could easily migrate, like the dialysis business, to specialized centers divorced from the rest of the healthcare system.  IBM’s Watson, given enough time, might replace a lot of the cognitive stuff we do. And if Watson needs a human face, then some holographic version of Skype might be just the ticket a decade from now.


We just do not know what the practice model will look like in ten years. But the conjoined twins of digitalization and globalization put lots of new possibilities in play.


The General versus the Particular: And, last but not least, the changing science of systemic therapy for cancer. The last decade has been dominated by the particular in systemic therapy: HER2, bcr-abl, ALK, and all the rest. In 2012 the FDA approved 12 new cancer indications, and almost all were targeted (or, to use the emerging cliché, precision medicine).


The genomics is suggesting a large (though not infinite) number of driver mutations, and often with many occurring simultaneously in smart cancers. If each driver mutation demands its own targeted therapy, the number of new drugs will increase relentlessly. And will every combination of driver mutations require a combination of targeted therapies? Try and imagine the general medical oncologist, faced with a dozen new drugs per year, employed in some factorial of combinations, across a wide array of tumors. The mind boggles.


And because the mind boggles, it is time to move from the particular back to the general? Oncology used to be dominated by “general” drugs: the platinums, the anthracyclines, the taxanes, and alkylating agents.  Are we starting to see a return to the “generalist” drug scheme? There is nothing particularly precise about ipilimumab other than its molecular target, nor is there anything smelling of true targeted therapy about the CDK 4/6 inhibitor that stole the show at the recent San Antonio Breast Cancer Symposium. They both target global systems (T regs, cell division), as their respective toxicities suggest.


Might this be a way around having to test hundreds of targeted therapy/precision medicine/whatever you want to call them combinations? As a “generalist” researcher once told me, all snowflakes are different, but they all melt. Maybe. We’ll see. Whatever works, and we don’t know what works yet. Probably some combination of the two.


I suspect that the tension between agents attacking particular cancer targets, and those affecting more general targets, will play out again and again over the next decade.


Anyways, my delight at the world not ending in 2012 stems in great part from my curiosity about how this will all play out. The Mega-Trends are there, and will define a great part of the next decade. But other surprises will keep popping up, and I cannot wait to see what they will be.