Projection of 15-year prostate cancer-specific survival after radical prostatectomy

For several years we have been able to use the Kattan nomograms to project 5-, 7-, and 10-year recurrence-free survival before and after radical prostatectomy for localized prostate cancer, based on the patient’s age, stage, Gleason score, and other relevant data (associated with biopsy or surgical findings, as appropriate).

In a new paper, just published on line in the Journal of Urology, Eggener et al. have now published data that will help us to be able to predict 15-year disease-specific survival in men undergoing radical prostatectomy for localized prostate cancer.

In the same manner as prior risk projections developed by Kattan and his colleagues, this set of projections is based on a sophisticated risk regression analysis modeled using clinical, pathological, and long-term outcomes data from 11,521 patients treated by radical prostatectomy at four major academic centers between 1987 and 2005. The model was then validated using similar data on 12,389 other patients treated at a separate institution during the same time period. The average (median) follow-up of the real patients in the modeling cohort was 56 months (nearly 5 years) and 96 months (8 years) in the validation cohort.

Eggener et al. were able to show that:

  • Overall 15-year prostate cancer specific mortality (PCSM) in the modeling cohort was 7 percent.
  • Statistically significant predictors of PCSM were
    • Primary and secondary Gleason grade 4 or 5 (each p <0.001)
    • Seminal vesicle invasion (p <0.001)
    • Year of surgery (p = 0.002)
  • Only 3 of 9,557 patients (0.03 percent) with organ-confined prostate cancer and a pathological Gleason score of 6 or less actually died of prostate cancer post-surgery.

Based on these data, they have developed a nomogram that is capable of predicting 15-year PCSM which was accurate and highly discriminating when tested in the validation cohort of patients.

When patients are stratified by age at diagnosis, the nomogram predicts 15-year PCSM rates as follows:

  • For men with a pathological Gleason score 6 or less — 0.2 to 1.2 percent
  • For men with a pathological Gleason score of 3 + 4 = 7 — 4.2 to 6.5 percent
  • For men with a pathological Gleason score of 4 + 3 = 7 — 6.6 to 11.0 percent
  • For men with a pathological Gleason score of between 8 and 10 — 26.0 to 37 percent.
  • For men with organ-confined disease — 0.8 to 1.5 percent
  • For men with extraprostatic extension — 2.9 to 10.0 percent
  • For men with seminal vesicle invasion — 15.0 to 27.0 percent
  • For men with lymph node metastasis — 22.0 to 30.0 percent

The authors conclude that, “The prostate cancer specific mortality risk can be predicted with remarkable accuracy after the pathological features of prostate cancer are known” and that poorly differentiated cancer and seminal vesicle invasion are the prime determinants of PCSM after radical prostatectomy.

While it is fair to say that there are no real surprises in these data, based on cohorts of patients treated at some of the very best academic centers in the U.S.A., it will certainly add to the quality of the Kattan nomograms when these data can actually be built into the on-line prognostic tools available on the Memorial Sloan-Kettering web site.

38 Responses

  1. Patients can usually be assigned to multiple classes: for example, Gleason 4 + 3 and extraprostatic with SVI.

    Is there any information in the study how the different factors add?

  2. Interesting. And you’re right, little surprises ~ increases in mortality across all risk levels … Mike there are two factors shown, Gleason and stage. Do you know if combining these gives us more information. For example, my case of 4 + 3 = 7 and bilateral seminal vesicle invasion. Do you think the PCSM increases above the numbers shown or are the ranges shown on the high end taking that into account?

  3. Finally, a study that validates the study that was released from Johns Hopkins in September of 2009. Johns Hopkins claimed that organ-confined Gleason 6 was “virtually 100% curable.” I believe they said it was 99.6% according to them. I have not found similar success rates of treating organ-confined Gleason 6 prostate cancer from anywhere else, so I was beginning to doubt the study. I know this study was not specifically talking about Johns Hopkins but the results are just as good. Well almost, Johns Hopkins said they had no deaths at 15 years, but the patient cohort was only about 2,500 as apposed to the 9,500 here.

  4. It will be a while before the finalized nomogram is likely to be available on the MSKCC web site (I would assume). When it is, I am sure it will take account of the combination of risk factors mentioned by Tony and Reuven. In the meantime, it is safe to assume that the upper and lower levels of risk for men with a specific Gleason score are encompassing the presence or absence of all other risk factors. In other words, if you are a Gleason 4 + 3 with bilateral seminal vesicle invasion but negative lymph nodes, your potential risk for PCSM at 15 years is probably nearer 11% than 6.6%.

    With respect to the Johns Hopkins data, these data either were included in the model cohort or they were the data used to validate the findings of the model cohort. I haven’t seen the full paper yet.

  5. Yea, I noticed that Johns Hopkins was one of the participants. However, it is still good the other institutions did not see worse results.

  6. Also, Tony I am confused. Where do you see increases in mortality across all the risk levels?

  7. Chris:

    Mortality increased with higher GS (nothing new) …

    · For men with a pathological Gleason score 6 or less — 0.2 to 1.2 percent
    · For men with a pathological Gleason score of 3 + 4 = 7 — 4.2 to 6.5 percent
    · For men with a pathological Gleason score of 4 + 3 = 7 — 6.6 to 11.0 percent
    · For men with a pathological Gleason score of between 8 and 10 — 26.0 to 37 percent.

    One problem might be the difference between biopsy-derived Gleason score and post-surgical pathology-derived Gleason score when the determination of the biopsy sample is not done by an expert pathologist. A high percentage of cases are off by 30%, with mostly understaging. Obviously, these understaged patients will fail more often given the above stats.

  8. Also, Mike, how do they estimate 15 years out when mean followup is 5 and 8 years? I assume some patients went out to 15 years.

    Are these statistics based on the absence of second-line therapies for biochemical recurrence?


  9. Chris,

    This is just my own observance but prostate cancer-specific mortality is very rare in the first 10 years in all risk groups except for those with high Gleason’s wbo discovered their disease at a very late stage. However, I am seeing more men come to our Us TOO group, and at HealingWell, who are dealing with progressing disease more than 10 years after initial diagnosis and treatment. As I tell our group, prostate cancer is a 20+-year disease. And for most men diagnosed at a median age of 67 and above, that is good news. However to the younger guys, like myself, that is a whole different story. This is why we need longer studies to be carried out. Men under the age of 60 will more likely care for this information when deciding on which therapy to choose. I personally do not put too much stock in [a lot of] studies since they quite commonly lack the maturity necessary to give us the most important piece of information ~ prostate cancer-specific mortality. This study shows us that the mortality rates are higher than what any 10-year study I have seen shows us.

  10. Dear Chris:

    I do not pretend to be an authority on the statistical minutiae of risk regression analysis. What little I do know is that this is a very sophisticated method for projection of risk over time, and is the same method as Kattan and his colleagues used to develop the original Kattan nomograms, which have proven to be remarkably accurate.

    As I understand these statistical projections, they are based on the normal use of appropriate therapies in those men who were believed to need second-line and additional care over time. In other words, if you had a rising PSA 2 years after initial surgery for a Gleason 7 tumor, you probably got second-line radiation therapy.

  11. Also worth noting (Mike correct me if I am wrong) …

    Today the mortality rate in prostate cancer is considered to be somewhere around 3% of the entire diagnosed population of survivors in the US. This study shows the mortality rate to be 7%, or more than double that of the overall cohort. This does appear to be a significant increase in mortality rate …

  12. Tony:

    I am not sure that you can appropriately draw your last conclusion from the data above. This study is focused entirely on projecting risk for prostate cancer-specific mortality in an individual patient with certain specific characteristics. It tells us nothing about actual, overall prostate cancer-specific mortality, which is very likely lower than the accumulated, projected risk since a lot of those men at risk are still likely to die of something else before their prostate cancer does them in. In addition, one is still faced with the difficulty of interpreting the statistical “fact” that “only about 3 in every 100 men diagnosed with prostate cancer will die of their disease.” It is sometimes hard to know exactly how to interpret these statistical “facts” because of the assumptions being made in their development and derivation.

    The age-adjusted risk for prostate cancer-specific mortality is definitely still falling. If that is the case, I don’t think we can make the presumption that the overall risk of prostate cancer-specific mortality is rising based on what is “only” a statistical analysis of probabilities for an individual patient.

  13. I know these are all hypothetical results based on projections. But there is no question in my feeble little mind that the longer you live after being diagnosed with prostate cancer the more likely it is that the disease can prove to be fatal. These numbers don’t disagree with that thinking but they do challenge the contrary. But I do acknowledge it is improper to make any conclusions by combining results of different papers and studies, a mistake commonly made by many patients and advocates. Still I will always advocate for longer term studies preferably with Level 1 evidence. This paper totally misses that mark.

  14. I think there may be a problem with the analysis and results. I can’t find any indication that the type of treatment was taken into account. For example, say a man has 4 + 3 Gleason and has a prostatectomy. If after surgery they find SVI and/or extraprostatic extension, I would assume the mortality at 15 years is affected by the follow up salvage therapy (RT or RT+ ADT) or absence of it. Am I missing something?

  15. Reuven: No. You aren’t “missing” anything.

    Of course the results are “affected” by any follow-up treatment. This is not a study about the survival effects of surgery alone. This is a study about projecting prostate cancer-specific survival of men whose first-line treatment is surgery … followed by whatever other forms of treatment they might need later if they have progressive disease. If a patient had progressive disease after his first-line surgery, he would have been given second-line treatment, and then third-line treatment if he needed it.

  16. I am sorry, but in this case I wonder about the value of the study. I suppose people would be interested in finding out what the mortality is for first-line of treatment without secondary vs. first-line of treatment + secondary. That will help people make a decision as to whether to initiate atherapy vs. waiting for salvage (if necessary) and, based on statistics, which type of salvage, i.e. RT vs RT + ADT.

  17. Reuven:

    OK … Now you are missing the point. No one can tell any individual patient up front whether they are going to have progressive disease or not prior to treatment. What they can tell them, however, is the probability of progression-free survival (which is what the original Kattan nomograms do at 5 and 10 years prior to surgery and at 2, 5, 7 and 10 years after surgery alone). What the current study also shows is the probability of prostate cancer-specific survival at 15 years if surgery is your first-line treatment. This information is a supplement to the former information. On its own, you are correct, it has limited value, but the whole point is that it will need to be used in conjunction with the former information.

    To give you a specific example, consider a 60-year-old male, diagnosed with clinical stage T1c disease, a Gleason score of 3 + 4 = 7, a PSA of 4.8 ng/ml, and 2 of 12 biopsy cores positive for cancer. Before surgery, the pre-treatment Kattan nomograms can tell him that he has a 94% probability of progression-free survival at 10 years.

    So the man has surgery, and afterwards the picture isn’t quite as rosy because his pathological report shows he has pT3a disease (i.e., a positive seminal vesicle) and a Gleason score of 4 + 3 = 7. Three months after his surgery the post-treatment Kattan nomogram can tell him that he has an 86% probability of progression-free survival at 5 years but only a 78% probability at 10 years.

    However, the new study described above provides additional information. Despite the new 78% probability of progression-free survival at 10 years, we now can also know that he has a 73.0% to 85.0% percent probability of prostate cancer-specific survival at 15 years (based on the positive seminal vesicle).

  18. The survival rates are great as long as you treat cancer that does not need to be treated and you do not count deaths due to treatment. I was diagnosed 5 years ago with Gleason 3 + 4 and three PSA readings averaging 9.2 ng/ml.

    I chose to have no treatment other than auxiliary with vitamins and additives and now have PSA of 1.27 with no symptoms other than those of BPH.

  19. And John is completely correct … There is a large cohort of men who can reasonably manage their localized prostate cancer by monitoring it with care (“active surveillance”), making modifications to diet and exercise patterns, etc. The only problem is that — at present — we have no really good way to identify these patients with accuracy at the time of diagnosis.

  20. True that there are men with minimal prostate cancer that are being treated. How do these men impact the mortality rate? They don’t! The reduction in deaths is related to treating men that, if untreated, would have died of prostate cancer.

    The proof is the fact that 74% to 63% of men with high Gleason scores (8-10) treated in the study have a high probability of surviving 15 years.

  21. I wonder if the authors have other data available from this study such as biochemical recurrence rates by group. With such a large group of patients I would think the data would tend to show some valuable trends.

  22. Dear Chris: These are the data were all used to create the original Kattan nomograms we have all been using for some years now and that are available on the MSKCC web site.

  23. Maybe being a Gleason 6 is why I’m still alive. I’ve got mets scattered from hell to breakfast all over my body and my PSA is nearly 1,300 ng/ml more than 15 years post-diagnosis. I’ve never undergone androgen deprivation therapy, but I do take Proscar daily. I’ve had symptoms only twice in all this time: once when a prostate cacner met compressed a spinal nerve and I had to get radiation, which was successful and once when a pubic fracture called for more palliative radiation, also successful. I feel great now, but I wish some of the clinical trials didn’t require those of us in stage 4 with bone mets to first castrate ourselves either physically or chemically in order to participate.

    Sterling Greenwood/Aspen

  24. I should have mentioned in my prior post that I am taking monthly infusions of Zometa. And I did have the retropubic radical prostatectomy some months after I was diagnosed some 15 years ago by a Prostasure blood test. Obviously, there was a recurrence.

    Sterling Greenwood/Aspen.

  25. Sterling:

    I have heard of higher PSA levels in men with few or no evident symptoms … but not very often over ther past 20 years. You are clearly a walking exhibit for the variability of the aggressiveness of even metastatic prostate cancer.

  26. My husband (now 66) had a prostatectomy in 2009 and still had a PSA rating of 10.4 after the surgery. Recommended course of action was radiation (7 weeks) in conjuction with hormone therapy. October 2011 his PSA reading was 0.36. How is this possible with no prostate? What’s next? And where can I obtain some reliable information on what therapy options we can consider? Obviously, our lives have been impacted by this and trying to get real answers.

  27. Dear Mrs Jones:

    If your husband had a PSA of 10.4 after his surgery, then either some part of the prostate that was cancerous was left behind when the prostate was originally removed or your hasband’s cancer had spread outside the prostate. Your should ask the original surgeon for a copy of your post-surgical pathology report.

    The important thing now, after the radiation therapy, is whether your husband’s PSA is stable or rising. If you join our social network, we can discuss your husband’s case in more detail.

  28. Hello all.

    I need someone to clarify a specific situation my father is in, based on which I do not know how to fill in the nomograms.

    He had radical prostatectomy, and his pathology report was in some cases OK, in some not. He has been confirmed, post-surgery, to have a Gleason score of 3 + 4 = 7, pT2bN0, nothing in lymph nodes, seminal vehicles, urethra, … minimal infiltration to posterolateral margins (it literally says “minimal,” whatever that means) without trans-capsular spreading. So, while it looks like a “confined to the organ” situation, it also says that the bladder wall was invaded. His first PSA reading, 1 month after the surgery, is 0.03.

    So, how do I fill this? Also, how can tumor invade the bladder wall without transcapsular spreading?

    I suppose the picture is not so good, but I would like to see the chances he has. His tumor volume is 20% and he is 58 years old.

    Thanks in advance.

  29. Dear Tile:

    (1) The statement that there was “minimal infiltration to posterolateral margins without trans-capsular spreading” should probably be interpreted to mean that there were no significant positive surgical margins. However, you really need to check this with the surgeon.

    (2) If your father has cancer that has invaded the bladder wall, I cannot see how he can possibly be classified as pT2bN0. Such a finding would normally require him to be classified as pT4N0 (even though his original clinical stage may well have been T2b). Again, you are going to need to get some clarification from the surgeon. For example, was he able to surgically remove all the cancer from the bladder wall (which could then mean that there really were no positive surgical margins, even though the surgeon had to cut out tissue from beyond the actual prostate capsule (which is not like a real “capsule;” there is no real “wall” around the prostate).

    The situation may not be as worrisome as you think if the seminal vesicles and the lymph nodes were both negative and if the surgeon really was able to excise evident cancer that had grown into the bladder wall.

  30. Thank you so much.

    I managed to see him yesterday afternoon. There were 2 Ph reports. One took three large samples, one is malignant, 9 x 6 x 3 mm, and that one has infiltrated the wall. But there is absolutely no comment on how much. The other used 42 samples from left, right, apex, and two more sides, and this one concluded that there was no transcapsular spreading.

    The surgeon does not want to make a prediction; he only comments that the odds of cancer returning are against my father, but that he is not concerned that cancer is beyond point of “curability.” He also noted that he saw no physical presence of cancer outside prostate wall, but nevertheless he cut out all that may be affected,including the bladder neck.

    Let’s say that this is correct. (He is considered to be a top quality surgeon.) How would I calculate the odds?

    I am not searching for a cure, I am not that optimistic, but for a PSA-free probability in 5 and 10 years, if there is any.

    Thank you so, so much.

  31. Tile:

    If I use the Kattan post-surgical nomogram and the data you provided in your earlier post, and I make the assumptions specified by the surgeon in your last set of remarks (that your Dad had no positive surgical margins but he did have extracapsular extension), then the nomogram suggests (after just 1 month since his surgery) that at 5 years your Dad has a 94% probability of biochemical progression-free survival and at 10 years it is still 90%.

    However, the longer your Dad goes post-surgery with a stable PSA down around the current level (effectively undetectable), the better his probability of non-recurrence. So, just as an example, if we assume that his PSA is still down at about 0.03 ng/ml at 12 months after his surgery, then his probability of progression-free survival has increased to 95% at 5 years and to 91% at 10 years. Better still, if he is still disease-free after 5 years, then the projection is 96% at 10 years.

    The probability looks high that your Dad will still be around to tell you not to do something that seems perfectly reasonable to you (a father’s responsibility, after all) some 15 or 20 years from now — assuming he is otherwise in good health … so don’t let him spend too much time sitting around on the couch!

  32. Thank you very much. You just drove the tears to my eyes. Not because of a new “optimistic” probability, but when someone else takes care, its … I don’t know the word.

    I’m a little bit confused by your statement “that your Dad had no positive surgical margins but he did have extracapsular extension”

    Do you mean that minimal postlateral infiltration is considered as no positive surgical margins, and that bladder wall infiltration, which was, I hope, cut off, represents extracapsular extension? Or I misunderstood?

    I included both, and am still satisfied, as one can be, with the odds. I was just under impression that this bladder wall thing is really taking down all the odds. Also, how precise is 0.03? Surgeon says that he is not worried until 0.1 or even more, so am I to worry at 0.06 or anything less than 0.1?

    To confirm again, and I will not bother you further: Bladder wall was infiltrated, surgeon did not see it, but he cut off the area together with the bladder neck anyway. There was “minimal infiltration to the post lateral margins.” The rest says that apex is infiltrated by tumor, volume of 20%, urethra, lymph nodes, seminal vehicles are regular.

    Thanks again, sorry for “re-asking”, I just don’t want to get carried away. … And thanks a million times!

  33. Tile:

    There is always a certain amount of “judgment calling” involved in all of this type of prognostic estimation. What the pathology report and the surgeon both seem to be saying is that there was no significant evidence of a clear positive surgical margin anywhere around the prostate capsule but that there was a visible extension of the cancer at the base where it was attached to the bladder. The surgeon is saying he was able to remove all the cancer that was attached to the bladder and that again there was no positive surgical margin there either. So …

    Yes, I have made the judgment call “that your Dad had no positive surgical margins but he did have extracapsular extension.”

    Now … Please remember that I am not a physician. Please understand that I haven’t done or seen any of the things that the surgeon or the pathologist have done or seen. All I can do for you is use my judgment based on nearly 20 years of trying to help patients understand what they are being told. Your Dad may not be this lucky. Things go wrong. Statistics can only be used to assess probabilities … and I may be making the wrong judgments. BUT … I have told you what it sounds like to me … and I think you should just encourage your Dad to get his PSA checked routinely for the next 10 years. Even if he does have biochemical progression, there are plenty more things that can be done before he is at any significant risk of dying from prostate cancer … THAT is extremely unlikely.

  34. Thank you so much again.

    I am, again, not that optimistic, but, I never am about anything, so … I’m just searching for a straw to catch, as I am literally falling apart. I am just that kind of person, nothing much to do about it. But knowing that there are other people who care, is very, very encouraging.

    I really admire what you do. If I get any new info, although, I am afraid to ask for any, I will post them. Someone in similar situation might benefit from it. I hope…

    Thank you.

  35. Tile:

    If you join our social network you will find lots of men with prostate cancer and their family members talking about issues like this all the time. THAT is actually the best place for you to go to get questions answered … and don’t feel bad about asking questions. You are concerned for you Dad, but you don’t need to “fall apart” over this.

  36. Hi. Yes, I will join. Although I already joined some, but … still falling apart. :)

    I suppose this is who and how I am, but will try for sure. Will try anything. And again, thank you very much. What you do is impressive. Really impressive.

  37. Perhaps my experience will help address some of the questions presented here. My 15th “anniversary” since surgery will be 12/11/12 — in 2 months. My surgeon was the (in)famous Pat Walsh at Hopkins. (Walsh patient #2040. We all know our numbers, like the serial numbers on Steinway pianos.)

    Biopsy: Gleason 3 + 3. Two cores out of six malignant, 33% and 35%. (In 1997 biopsies were one “six pack”, rather than the standard two “six packs” today.) Ploidy: aneuploid.

    Procedure: radical prostatectomy with no prior treatment of any kind. (Walsh wouldn’t take patients with prior treatments nor those above 70 years of age.) Both neurovascular bundles retained (full potency).

    Age at time of surgery: 52.

    Pathology report: clear surgical margins.

    Post-surgery treatment: none.

    One reader asked: How did they have follow-up data for 15 years, when it is “typically” confined to 5 or 8 years? Answer: I receive a letter from Pat Walsh every year asking for this year’s PSA results. Without fail, I fax the form back to him every year. Also, despite an undetectable PSA every year; my internist insists on performing a DRE.

    It appears that I fall into the category that Hopkins claims is virtually 100% curable. Despite this history, I will continue to have PSAs beyond my 15th anniversary in December.

  38. Dear Ted:

    I think you can safely bet that you will continue to receieve the annual epistle from Pat Walsh (or from someone else at Hopkins on his behalf) for at least aother 10 years!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: