Prostate Cancer – Urological Research Foundation
One Man to Another: It’s In the Numbers
Why a patient’s column?
This is not a doctor’s column. I am, in fact, a patient of Dr. Catalona and a member of the URF Board. This column attempts to provide a patient’s perspective on prostate cancer to the readers of QUEST. I often discuss my own medical case in order to provide clear illustrative examples. I hope that you see some of yourself in these examples as well.
You may wish to review my prior columns on Dr. Catalona’s Website, www.drcatalona.com. Insert my name in the “QUEST Articles” search engine.
In search of a unified view of prostate cancer outcomes
My journey with prostate cancer began by learning about the individual numbers that are used for specifying the disease. In my case they were: Stage T1c (elevated PSA with no palpable tumor), PSA=5.4ng/ml (nanograms of protein per milliliter of blood), Gleason 3+4=7 (grade pattern 3 plus grade pattern 4 yields a score of 7). These are helpful numbers. They do not, however, answer the basic overall question that the patient would like to ask: “Do we have any way to look at all of the numbers together, in a unified way, to know where I stand?” There are many ways to answer such a question, but I will limit the discussion to three important approaches to such a unified view.
Problems in using prior data
The experience of prior patients is the basis for knowing what’s likely to happen to us, now and in the future. The use of such data always has at least the following two problems.
1. Using long-term data makes our estimates more reliable but it requires using old data that is very pessimistic, both because the men of that era were detected with prostate cancer at a later stage than men who are currently part of active screening programs, and because it usually implies the use of old analysis methods.
2. We almost always use results of what happened to the average man, as if such a man exists, but an individual result may be many times better or worse than the average. A famous article about using average estimates, by Stephen Jay Gould, Harvard professor, describes how his doctors advised him that he had an average life expectancy of 8 months. Gould explains why that doesn’t necessarily mean that he was only going to live for 8 months. He died only recently, 20 years after the original diagnosis!
A unified view of prostate cancer outcomes: three approaches
(Each approach is known by the name of the doctor who worked on it. Similar work has been done by others, but I believe that these are the best known. All of these approaches are available on the Internet and in printed form.)
1. Partin Tables. Partin Tables convert the common data obtained by your doctor – clinical stage, PSA value, and Gleason Score – into charts that show what is happening with the cancerous tumor in the prostate.
For each set of doctor’s (clinical) data, we can look up the probability that the tumor in the prostate is either organ (prostate) confined, or extracapsular extended (outside of the firm boundary of the prostate), or extended to the seminal vesicles, or extended into the lymph nodes. A very high probability of being organ confined is the best case.
If the Partin Tables are right, then the doctor knows about the extent of the tumor before any therapy is done. Almost every major published set of medical data follows the structure of the Partin Tables. The Partin Tables were issued in 1993, 1997, and then in 2001.
One concern of mine in using the Partin Tables is that when I inserted my own numbers into the 1997 Tables, and then into the 2001 Tables, my probability of organ confined disease had improved by 14%! That’s not possible since it’s for the same person, me. In addition, the truth was discovered after my surgery, namely, that I had extracapsular extension. The latest Tables would have given me poor guidance to expect what was found in surgery.
It has been suggested that one or more additional terms should be used in the Partin Tables to get more accurate results. In the last issue of Quest, Dr. Catalona suggested that “tumor volume” was another significant term in addition to PSA and Gleason score. The same Partin/Walsh medical team who create these Tables added tumor volume in a recent study. The results offer an improved insight, as we will see next.
2. D’Amico Risk Groups. In this approach, the terms that specify prostate cancer are grouped so that a patient can know the severity of his disease: that is whether he is in a low risk group, an intermediate risk group, or a high-risk group, for long-term non-recurrence after treatment with either surgery or radiation therapy.
D’Amico states that, after biopsy, if proper values are chosen to define each risk group (such as selecting the allowed values for Gleason Score, PSA, etc), then, the intermediate risk group disappears and everyone is either low risk or high risk.
We should, he says, be able to provide long-term estimates for non-recurrence for this expanded low risk group using surgery and radiation, and we should almost immediately look for additional therapy for the high-risk group.
D’Amico eliminates the intermediate risk group by determining the percent of biopsy cores that contain cancerous tissue. This percentage is proportional to the tumor volume and to the probability of organ confined disease. Generally, his expanded low-risk group includes patients with PSA’s less than 10, Gleason Scores of 7 or less, and less than 50% positive biopsy cores.
One problem with this method, however, is that it’s not always true that the very small biopsy cores hit the cancer. One can have a false low reading of the percentage of positive cores. In that case, the expanded low-risk group can lead to wrong conclusions because it puts Gleason Scores of 5 and 7, and PSA values of 2.5ng/ml and 9ng/ml in the same group. A wide range of expected outcomes could be grouped together and falsely thought to be excellent. In general, the wide extent of patient conditions included in each risk group tends to limit this method.
3. Kattan Nomograms. This approach provides us with a computer model that combines all of the terms that apply to the therapies of surgery, external beam radiation, and seed implants, into a tool that is easy to use on the Internet.
It provides several kinds of outcome data for the patient and compares therapy choices. However, it has several problems. Firstly, Kattan’s data are only for 5 years and that’s too short a data interval. Secondly, it does not agree very well with the 2001 Partin Tables. Thirdly, Kattan compares results for surgical and radiation therapies, but Dr. Walsh has pointed out that his 15-year non-recurrence rate for surgery would be improved by 22% if he used the non-recurrence criteria that are favored by the radiation community.
Comparing surgery and radiation outcomes may not be a fair comparison. And lastly, it doesn’t yet include any of the newer terms such as tumor volume, number of positive biopsy cores, percentage of biopsy cores that are positive, etc., that are changing the landscape of these kinds of predictions.
A recent result using these approaches
Dr. Walsh analyzed the recurrence rates of 1149 of his surgical patients. He included a selected measure of patient tumor volume (both percentage of positive biopsy cores and maximum percent of cancer in a single core of less or more than 50%).
He found the new result that Stage T1c patients (elevated PSA, non-palpable) were really two subgroups: a low risk group (T1cI, generally Gleason Score 6 or less) who had 96% rates of non-recurrence in 10 years; and a higher risk group (T1cII, generally Gleason Score 7 or more), who, when smaller tumor volumes were also present, had a 75% rate.
Dr. Walsh was recently on TV reporting that Senator Kerry’s probability of 10-year non-recurrence was 98%. Such high estimates of “cure” are startling. They reflect these newer analyses.
Conclusion about a unified view of prostate cancer outcomes
Estimates can be made of the probability of a tumor being contained or not contained to some extent, or that someone is in a particular risk group, or that he has some probability of non-recurrence if various therapy choices are made. All of this work is evolving.
The result of organizing all of the information into a unified view is that doctors are beginning to discover better terms for describing this disease, and therefore better guidance for proper courses of treatment. It also gives us, as patients, a better way to discuss our plans with our doctors. Finally, “If you treat it small, you get it all.” may summarize the newer story of applying tumor volume information.
Please feel free to offer comments or ask questions about this column by contacting Dr. Catalona at his website: send me an e-mail at firstname.lastname@example.org