How accurate are the MSK nomograms?

I won't say, but I will hope that 10 years is only the start for you.

Fairwind:

Now you refer to YANA and previously you referred to HW as evidence that the MSK nomograph is a bunch of hooey.

I think folks have pointed out that you are looking at a skewed sample. You are ignoring the tons of folks that have done well and moved on. But they are included in the MSK nomographs, along with every one else.

I, too, am VERY disappointed that the probabilities did not work for me.

We both got the short end, along with many of the folks that post here

Mel

Very good discussions and observations. All of you have very valid points, Fairwind's included. I have a few observations of my own:

1. There were 200,000+ new prostate cancer diagnosed in the US a year and over a 10 years period, there were 2 million+ new prostate cancer patients. A 1% fatality of this population means there will be 20,000+ prostate cancer death per year, consistent with the number of cancer death per year. So, perhaps, the Sloan-Kettering prediction is not that far off.

2. As mentioned prior, the HealingWell population is perhaps not normal but skewed towards patients that have problems rather than those that are successfully treated and have no complications. In my cancer support group, those that were successfully treated tended to drop out quickly as they went on to live a normal life. Also other than me, I know of no one that participated in internet forums.

3. Prior to my treatment, I looked at the Patin table and ran the MSK nomogram for my PSA number, Gleason score, # of core positive and clinical T-stage (T2b) and was totally stressed out by my odds of:

Organ confine: 51%
EPE: 32%
SV Invasion: 11%
Lymph node: 3.3%

So, for me I can't say that the MSK prediction was optimistic. On the other hand, perhaps I may not be as "cancer free" as I hope??

So, fellow members, as I came to realize and accept, in life, we played the hand that were dealt to us. We win some, lose some and accept "luck of the draw" as it is handed to us. I now take everyday as a bonus and try to enjoy and be thankful for each and everyday I am alive. Perhaps, that is the best outcome of my being a member of this club, cancer free or not.

Regards to all.

I did not see the need to reply but the last few messages need to be addressed because people who read this in the future might become confused.

I believe in using the right tool for the right job. The Partin Tables and the Memorial Sloan Kettering nomogram for post surgical treatment also know as the Kattan nomogram are two DIFFERENT tools.

The Partin table was developed to be a predictive tool for those who are selecting a treatment plan. While directed toward the surgical option, the Partin table uses your PSA, Gleason and Clinical Stage and gives you an informed decision as to what your pathology should be once the prostate is removed. Again, this is a predictor of your pathology based on the thousands of cases studied by Johns Hopkins as they compared the pre surgical and post surgical results. The more people studied by Hopkins the better the model has become in providing an accurate assessment of the outcome someone will have.

The Kattan nomogram that is labeled pre-treatment is Sloan Kettering's version of the Partin nomogram. The SK nomogram adds the probability of post surgical freedom from recurrence to the result based on the data collected from patients who were tracked by SK urology department. The Partin table does not add the post surgical predictive model.

The post surgical SK nomograms are the potential for disease free progression over time based on the data collected from patients AFTER surgery when pathology trumps the predictive values of biopsy results for giving long term outcomes. I think we all agree that pathology results are better at predicting future outcomes as opposed to results from the biopsy.

The results are based on patient data and how they did over time for those treated at MSK in New York.

For those concerned with PSA biochemical recurrence after surgery, the Han tables from Hopkins address those concerns. Hopkins also has a set of unnamed tables that indicate survival times after biochemical recurrence based on PSADT (rate of rise of PSA over time)

The study I cited earlier is a compilation of data from several large academic settings so that over 15,000 patients over many years were studied to both see how actual patients did and to use some predictive statistical model to extend the vision out beyond a shorter time period to get a sense of how patients will do over time. This study breaks some new ground because of the massive size of the PC patients studies and the amount of time they kept in contact with them to chart their progress. This was a major effort that took a lot of work to accomplish.

A great feature of the SK nomogram is that it has some boxes to check if surgical outcome includes EPE, positive margins, etc. Some models don't account for that and leave men in limbo because they can't extract data from the patient outcomes that matches their own individual profile. The SK nomogram does this better than a few other models do because SK tracked the men with these variations and recorded how they did over time.

Anyway, my point was to not confuse the Partin Tables with the SK post surgical nomogram. They are tools used for two different purposes and employed at different times in the treatment experience.

Mike S.

On the MSKCC website, there is a Nomogram entitled: Risk of Dying of Prostate Cancer in Men With Rising PSA After Radical Prostatectomy. I plugged in Randy's numbers and it gave him a 72% chance of dying in five years and a 99% of dying within 10 years. It's been nearly 3 years, and so far, so good. While his PSA promptly rose (and quickly) as soon as he discontinued ADT, he still looks and acts like a man in his prime. There has been no detectable bone mets....or any other mets, for that matter and we hope to continue on. There is no way to determine what really comes next. All this nomogram did for us was provide us with a few sleepless nights. Unless something changes drastically, I expect that he will be one of the 28% that will survive the 5 year mark. I also pray that he's one of the 1% who also survive that 10 year mark.

RandyJoe,

I have always struggled to understand the MSK "Risk of Dying of Prostate Cancer in Men With a Rising PSA After Radical Prostatectomy" nomogram. I looked at it more closely today.

I believe there is a flaw in this web application's implementation (more probable) or maybe the underlying data, calculations, interpolations, etc. (yikes!).

Take the case of John Doe, who had RALP at age 50, pre-RALP 4.0 psa, favorable path, went five years before recurrence of .2 at age 55, and has favorable PSADT of 30 months. I think everyone would agree that, of a group of 100 men like John Doe, most would not have died of PCa at 5 or 10 years, and probably most not even at 15 years. Not only that, but probably not that many of them would be dead of *any* cause at those same years--they would all be age 70 at 15 years out from recurrence, and the average male lifespan is 77 (at least in U.S.).

Here is the data as entered on the nomogram:

Disqualifying Treatments:

Hormone Therapy Before Surgery No
Radiation Therapy Before Surgery No
Hormone Therapy After Surgery Before PSA Rise No
Radiation Therapy After Surgery Before PSA Rise No

Before Surgery:

Pre-Treatment PSA 4.0 ng/mL

After Surgery:

Extracapsular Extension No
Lymph Node Involvement No
Seminal Vesicle Involvement No
Primary Gleason at Surgery Pattern 3
Secondary Gleason at Surgery Pattern 3
Surgical Margins Positive No

At Recurrence:

Age at Biochemical Recurrence 55 years
Time to Biochemical Recurrence 60 months
PSA at Biochemical Recurrence .2 ng/mL
PSA Doubling Time 30 months

After hitting the "calculate" button, the 5 / 10 /15 years "death from prostate cancer" chart and figures show as: 64%, 98% and 99%.

The text under the displayed chart for the "10 year after recurrence" is this:

This number shows, as a percentage, the probability that you will die of prostate cancer 10 years from the time you experience biochemical recurrence after radical prostatectomy. This probability means that for every 100 patients like you, we expect that 98 will have died of prostate cancer 10 years from the time of biochemical recurrence after surgery, and 2 will NOT have died from prostate cancer within 10 years.

This prediction addresses survival related specifically to prostate cancer; it does not exclude the possibility of death from other causes, such as heart disease or accident, within this time period.

The information used to build this prediction tool was based on patients treated from 1987 to 2011...

I entered my own (less favorable) numbers and my results are not *anything* close to what I think are my real risks of dying of PCa or anything else in 5/10/15 years out. Yes, one never knows, but still...

Even with their caveat that the underlying data is somewhat "long ago", it is hard to believe that this group of men were dying (or will be dying) at the rates the nomogram shows.

Robert