One of the things I enjoy about coming to meetings like the current annual session of the American Society of Clinical Oncology (ASCO) is that it gives me a chance to give thought to some larger questions that face cancer care. A presentation I attended Friday afternoon on the impact of genomics on cancer diagnosis and treatment in the future has offered just such an opportunity.
Most of you I suspect give little thought to the actual processes that we use to diagnose cancer. One has a tumor somewhere in the body, the doctors take a specimen, send it to the pathologist and the pathologist makes the diagnosis. Simple and straightforward. Get it done and get on with treatment.
But in fact it isn't so simple and straightforward. And in the world we live in, it is getting more and more complex.
Looking at cancer tissue under the microscope is something that has been done for over a century. More recently, we have seen the advent of special additional tests that tell us for example whether or not a cancer such as breast cancer is hormone sensitive or whether it has other markers such as HER2. We can send specimens of the cancer to a lab to find out whether or not it is more or less aggressive and we can even do tests to find out whether or not--for example--a woman with a breast cancer really needs to take traditional cancer chemotherapy. There are even special stains that can be applied to tumor tissue through a variety of techniques that can further refine the characteristics of a particular tumor and help us determine what kind of cancer it may be, or what subgroup of a family of cancers, such as lymphoma, a particular cancer fits in to.
All of that is well and good, but unfortunately that simple explanation does neither justice to how doctors diagnose cancer, nor does it say much about the problems that can occur in making cancer diagnoses, especially with all of the new tests that are available. I suspect that many physicians will agree that simply looking at the tissue under the microscope just doesn't tell us anymore all the things we should know about a particular individual's cancer.
Some cancers have become exceedingly complex to diagnose in this fashion. For example, we have learned over the past several years that there are a large number of subtypes of malignant lymphoma. The situation has become so complicated that many experts believe the pathologist who actually makes the diagnosis of lymphoma should be someone who specializes specifically in that type of cancer. Understanding that the majority of patients with lymphoma are not diagnosed at cancer centers who possess such expertise, one begins to wonder whether patients with suspected lymphomas should be referred to specialty centers even before the first biopsy is obtained.
And then there are the issues surrounding quality control and quality improvement in cancer diagnosis.
I was fascinated a number of years ago by a study that tried to look at quality control in the diagnosis of cervical lesions. The conclusion by the authors who reported that study was that a quality control program looking at the diagnoses made by the pathologists led to several questions about the final diagnoses. Trying to institute such a program was fraught with complexities. And that was at one small group of hospitals that were committed to addressing those questions. Now, there are research studies which show that differences in how pathologists make their diagnoses have real impact on how patients are treated with a variety of cancers, including common ones such as breast and colon cancers. These studies show that expert pathologists not infrequently disagree with their community brethren on the actual interpretation of cancer tissue specimens when a second opinion is obtained.
It doesn't stop there. We have known for years that some labs don't do a great job with tests to show, for example, whether a breast cancer is hormone sensitive. And those studies have been done for decades. More recently, researchers have reported wide variations in results from other tests such as HER2, and now there is research to show that genomic tests such as one which predicts whether or not a breast cancer patient needs adjuvant chemotherapy are frequently ordered inappropriately or the oncologists don't use the results to guide their patients' treatment, which raises the question of why they ordered the test and spent thousands of their patients' dollars in the first place.
So into this less than ideal world enters the developing science of genomics, a world where will be able to go way beyond the current state of art with optical microscopes and special stains. Now we are going to be able to look at the entire genetic makeup of a cancer on a routine basis. It is indeed going to be a brave new world, and it is going to be a routine part of our cancer diagnosis armamentarium sooner than many people realize.
In the session I attended here at ASCO, experts talked about the current state of the art in genomic diagnosis. To say the least, hearing from these experts was fascinating. But there is a bit of the future that is already here, and it is going to be very interesting to see how this all plays out, especially in the near term.
Understanding that we now have the potential to examine either the entire genetic sequence of a person's cancer or even pre-selected parts of that sequence is truly amazing. Already, researchers have discovered certain genetic changes which give us valuable information about a person's cancer. There are even circumstances where analysis of that data can lead to selecting treatments that are substantially different than would have been considered routinely based on the information from examining the tissue under the light microscope. There is even a center at the University of Michigan (and elsewhere) devoted to harnessing this new technology which offers patients and their physicians a detailed analysis of their cancer, along with suggestions of possible treatments that may work based on a genetic analysis.
In short, this is truly amazing stuff. However, for me, this raises some parallel questions that--once again--most of us don't think about routinely.
One such question is a simple one: how are we going to enter this "gee-whiz" world where we would like to have enough cancer tissue to analyze when our biopsies are more and more often done with small needles that give us small samples of that tissue? And what about the recent research that has shown us that even within one person's primary cancer there can be many different genetic "signatures" which may point in different directions? And how will we be certain the information we get from all these possible analyses will really make a difference? (At Michigan, for example, they actually empanel a board of experts--including a patient advocate--to review all the data on a particular patient before issuing a report and recommendation. They can do that now because this is such a unique service. But as the demand increases it will be interesting to see if they have to adjust their approach.)
And then there is another fundamental question that interests me: who is going to pay for all this?????
In the cost-conscious era we are entering, there are going to be higher expectations that what we offer patients makes a difference. Right now, doing that means research and research is expensive. Not only that, but research dollars are shrinking.
There is also another consequence of the current changes in health care that few of you are thinking about: much of this research is done at very specialized cancer centers. There is a shrinking pool of payors who want to pay for routine cancer care at these outstanding centers of excellence.
These centers do a lot of research and education. They provide complicated clinical services to their patients, their communities and frankly their country and the world. That is not necessarily efficient but I do believe it is necessary.
As we move our treatment of illness and by implication our treatment of cancer to more community based resources, those new programs (think "Accountable Care Organizations" or ACOs) aren't going to willingly send their patients--or their pathology--to the fancy centers. (I can tell you for a fact that I have advised several patients with lymphoma to have their diagnostic procedures done at a cancer center or at least have the pathology reviewed by a cancer center, and their HMO has said no. And that's despite a clear recommendation from a nationally recognized group of guideline experts who make explicit recommendations to that effect. Some of those patients on further review were found to have different diagnoses than suggested by the hospital that was "in plan" but not experienced in the treatment of large numbers of patients with lymphoma.)
So, as I reflect on the current state of the art in the diagnosis of cancer, I see us breaking the shackles of over a century of using the regular microscope and moving to a new world as outlined above. But I also see barriers that will make the necessary research and transition to genomic based diagnosis even more difficult.
How we resolve these issues over the next several years will give us a hint as to what kind of health care system we are going to have. More importantly, I suspect it will have a significant impact on whether or not we move forward in improving the care we offer our patients with cancer.
In closing, we cannot lose sight of the fact that the journey to the successful treatment of cancer starts right at the very beginning of that journey. Getting the diagnosis "right" at that beginning is in fact a key part of success. We are moving ever-more-closely to the day when we will be able to not only say it's a "colon cancer" for example, but we will be able to pinpoint the actual specific changes that led to that cancer developing in the first place.
As I wrote above, this is truly amazing stuff. But I ask you think about one more question: We might have the technology, but will we also have the will and the way to embrace that technology and make it work for all of us?
