We spend a good deal of time discussing targeted therapies in cancer treatment, but don’t pay as much attention to what I am going to call “targeted diagnostics.”
Yet, this is probably as important a part of the current revolution in cancer treatment since diagnostic and prognostic tests will play an increasing role in helping us improve our ability to decide which cancer patients will need treatment in which circumstances, as well as help us to more accurately provide prognostic information for patients and their physicians.
Two articles and an editorial in today’s New England Journal of Medicine point out how important these types of tests are becoming in cancer treatment.
Over the years we have developed a number of ways to help us identify which patients are going to do better or worse than others.
For example, in most cancers, we are able to provide information regarding the stage of the disease. In some cases the staging information is based on clinical data, such as physical examinations, x-rays, and so on.
In other situations, we have much more information available to stage a patient, such as surgical specimens of the cancer as well as the nearby lymph nodes. When coupled with the clinical information, this can provide us with what we call the pathologic stage of the disease.
Some of the additional information that we can take into consideration in determining whether or not a person has a good or bad prognosis includes the information from the pathologist’s examination of the tissue under the microscope.
The pathologist will comment in many cases whether a tumor is well or poorly differentiated, or perhaps whether or not the tumor invades blood and lymph vessels contained within the tumor.
There are some other laboratory tests that can be done which can help us determine the aggressiveness of the cancer, or whether it is likely to respond to certain therapies.
Still, these are imperfect ways to offer a patient a prognosis since there is no fundamental analysis of the cancer cells themselves.
If we had this information, we might better predict who will do well, and who will not. We may be able to understand why some patients who have tumors that we now predict will do well in fact have early recurrences, and why some patients whose cancers look aggressive in fact do very well.
For example, let’s examine what happens to a woman with breast cancer has no clinical evidence of spread of her disease.
The surgeon operates, and takes out the cancer and either one or many lymph nodes from under the axilla (depending on whether or not the sentinel node is positive or negative).
The pathologist then examines the cancer under the microscope, and tells the doctor the type of the cancer, whether or not it is well or poorly differentiated, and whether or not the lymph nodes contain cancer.
The pathologist also orders additional tests, such as hormone receptors, which provide both prognostic and treatment information (positive is better than negative; positive breast cancers have a higher likelihood of responding to hormone treatments).
The pathologist also does additional studies, such as HER2/neu markers, which help identify (again) good vs. poorer prognoses, and indicate whether a woman should be treated with Herceptin as part of her adjuvant therapy protocol.
But as good as these tests may be they really don’t examine the detailed genetic makeup of the cancer. And, because of human and laboratory variability, not all pathologists and not all laboratories produce the same information from the examinations of the same material.
That’s where these new types of genetic tests come in.
By analyzing the genetic makeup of the cancer cells in a particular tumor, the hope is that we will have a better chance of determining how aggressive a particular cancer may be.
There are some very practical issues here that are worthy of consideration and discussion.
For example, we know that we treat many more women than we have to with adjuvant chemotherapy for breast cancer. It’s not that we want to expose them to the toxicities, cost and inconvenience of chemotherapy. The reality is that we do not have a foolproof way of sorting them into groups where we can say with reasonable certainty who will do well without chemotherapy, and who needs chemotherapy to improve their outlook.
The result is that we have to treat more women than who will benefit from the treatment, to be certain we don’t “miss” anyone who needs the treatment.
If we had a test that would help us identify which women had a better prognosis, and for whom adjuvant chemotherapy would add little or no benefit, then we would improve cancer care considerably.
In one of the articles in the Journal, the authors report on an evaluation of five different tests that are currently available either in the clinic or the lab which have been found to be useful in helping determine which breast cancer patients are at greatest risk of recurrence.
The simple explanation of this complex research study is that the authors found four of the five tests produced similar prognostic results. This was interesting, since each of the tests looks at difference cancer cell genes and gene patterns.
The authors concluded that these tests, although looking at different genetic markers in the cancer cell, were in fact measuring the same “bottom line” cancer processes in the cells.
The editorial that accompanied the article made the point, as have others, that “at present, therefore, it is not clear that the quantification of the level of expression of dozens or hundreds of genes provides more information about the potential of a cancer for metastasis, virulence and response to therapy for an individual patient that does an optimal analysis of the standard and readily available histopathological prognostic factors.”
In plain English, although these tests are promising, they are not yet sufficiently accurate for patients and doctors to put their complete faith in the results for any individual patient. They don’t significantly improve the information available to us in managing a breast cancer patient beyond what is offered by currently available technology.
As also pointed out by the editorialist, there are randomized clinical trials currently underway which are going to assess the practical impact of these tests
Until those trials are completed (and I suspect until more accurate tests are available), these genetic profiling tests for breast cancer remain an item of interest and may be informative for some patients.
However, my oncologist colleagues tell me they are not yet sufficiently accurate for them to have complete comfort in recommending to their patient that they not take adjuvant chemotherapy for breast cancer based on a “good prognosis” result from one of these tests.
The article on the use of gene profiles in lung cancer was also interesting especially since, as the authors’ noted, previous reports of research in this area left uncertainty regarding the value of the use of these profiles in treating patients with lung cancer. Consequently, in their opinion, theirs is the first study of its kind that has a practical application in helping doctors and patients make clinical treatment decisions.
In this paper, the authors did a very elegant analysis of multiple genetic profiles of lung cancer tissue from patients with two forms of non-small cell lung cancer (adenocarcinoma and squamous cell carcinoma).
Their test was reasonably accurate in predicting who would do well and who would not. Again, however, it wasn’t perfect and there remains room for improvement.
One interesting finding was that in patients with stage IA lung cancer, which is a very early stage of the disease, the researchers were able to sort out who would do well and who would not.
In early stage lung cancer, most patients are advised to receive adjuvant chemotherapy based on clinical trials that show it has benefit. But in stage IA patients, there is no proven benefit for adjuvant chemotherapy for the typical patient.
I suspect that we are going to be seeing more patients with Stage IA disease in the next several years, as more heavy current and former smokers decide to get screened for lung cancer with CT scans.
As a result, the issue of adjuvant chemotherapy may become more pressing, since 25% of these patients will relapse and die from their disease.
If that’s the case, then it would be valuable to know if we can predict which patients with stage IA disease are going to do well and who will not. Then, we could consider what treatment options are available to help improve the survival for those patients who appear headed for a poor outcome.
What is exciting about this current gene-profiling research is that the authors were in fact able to sort the Stage IA patients into a group that did well, and a group that did not.
The implications of this finding were such that the authors suggested in the paper the design of a clinical trial to answer the question. I certainly concur with their suggestion based on the results they obtained from their gene assay.
Where does all of this lead us?
First, I think it is important to understand that this type of research with its improved diagnostic and prognostic impact is part of a larger picture.
We are seeing incredible strides in the treatment of cancer. This is in no small part the result of a huge investment in research that has helped us to understand the behavior of cancer cells.
The results, as recounted not infrequently in this blog, are showing up in many places, not the least of which is the development of the new targeted therapeutics.
But we are also developing targeted diagnostics, and targeted prognostic tools that will be able to tell us not only the details of the cancer’s behavior, but also which drugs to use to give the patient the best opportunity to survive their disease.
This brave new world was envisioned but not realized even a couple of years ago.
Now we are living in the midst of a revolution that will continue to develop and be refined at what I suspect will be a fairly quick pace.
Our doctors will be able to take cancers and perform analyses in the laboratory that will yield the secrets of these cells, and tell us who needs treatment and who does not, and then be able to monitor patients for relapse before the recurrent cancer becomes visible.
The studies reported in the Journal are but a very small sample of the type of research is currently underway in the area of genetic analysis of various cancers, and its applications to the care of cancer patients.
It has taken us decades to get here, and I suspect the journey going forward will be considerably shorter.