In today’s news reports we comment on articles dealing with:
- The declining rate of prostate cancer mortality in the USA
- Managing prostate cancer in the elderly
- Re-validation of the Partin tables
- RALP as salvage surgery for radiation-resistant patients
Several news organizations have covered a new report on the declining rate of cancer mortality, inclusive of an estimated annual decrease in prostate cancer mortality of 4.1 percent. It is gratifying to see this continuing decline in prostate cancer mortality, but we are still only beginning to address the over-treatment problem that has been associated with the mortality decrease. Way too many men are still being treated in ways that severely impact their quality of life while not necessarily offering a major mortality benefit.
Heinzer and Steuber have reviewed issues affecting the appropriate management of prostate cancer in the elderly, and the need nor better guidlines that can help clinicians to make the best possibole decisions with and for their patients. They note that prostate cancer is likely to assume greater importance with progressive aging of the population, and diagnosis of localized disease is now commonplace, even in elderly men, although cancer-specific mortality is generally low in such patients. They note that elderly men with indolent disease will not benefit from curative treatment,but that some patients with aggressive cancers will progress if not adequately treated. They also point out that comorbidities are the main predictors of life expectancy in the elderly. They conclude by stating that “there is a need for more risk stratified approaches for the management of prostate cancer in the elderly to avoid unnecessary intervention in men who unlikely benefit from such intervention, and allow treatment in those who might benefit from it.”
Yu et al. have re-validated the Partin tables as a predictive tool to estimate the probabilities of organ-confined disease, extracapsular extension, positive seminal vesicles, and positive lymph nodes at the time of surgery. Using the SEER database, they identified 11,185 men who underwent radical prostatectomy for prostate cancer in 2004 and 2005. Their analysis showed that the Partin tables discriminated well between patient groups at risk for positive lymph nodes and seminal vesicle invasion, but that discrimination of extraprostatic extension and organ-confined disease was more limited. They also note that the Partin tables performed best in younger men (aged 61 years or less).
Eandi et al. have reported on the use of robot-assisted laparoscopic prostatectomy (RALP) in salvage surgery of men progressing after first-line radiation therapy. They claim that their results demonstrate a similar level of effectiveness of this procedure compared to traditional open salvage surgery. The “New” Prostate Cancer InfoLink would remind readers that salvage surgery as second line treatment after radiation is associated with significant complications. In this series of 18 patients, Eandi et al. note that only 6 patients were continent post-surgery at a median follow-up of 18 months, and there was a range of complications during the surgery.
Filed under: Diagnosis, Living with Prostate Cancer, Management, Risk, Treatment, Uncategorized | Tagged: elderly, mortality rates, Partin tables, RALP, salvage
Search for new and
ongoing trials on the
CTAG PCa web site
O wise Webmaster, can you please explain to me what no one else can:
1. How screening reduces incidence of PCa — and if it doesn’t, then what is the reason behind the reduced incidence — the reduction is now 30% from the peak — and
2. Why a reduction in incidence has nothing to do with a reduced mortality rate? Surely if less men are being diagnosed with PCa each year (which is what the reduced incidence means, doesn’t it?) then surely there must be less PCa associated deaths.
I’ve tried to get some answers to these but keep getting told that I don’t understand or that I am a trouble maker. The first is true but if the latter is, it is unintentional. I really want to understand this puzzle.
Dear Terry:
OK … Let’s see if we can take this one step at a time.
1. How does “screening” reduce the incidence of prostate cancer? Well … we really don’t know whether it does or not. All we know is that in the USA, since the introduction of the PSA test, the incidence of prostate cancer first grew and then fell again. But we need to remember that the development of prostate cancer can take 15 to 20 years or more. So what we have been doing since 1990 is: (a) finding a large “pool” of patients with prostate cancer that had accumulated over the prior 30 years but who wouldn’t have been diagnosed without the availability of the PSA test. and (b) gradually reducing the size of that pool down toward a stable annual number of new diagnoses (which we have probably yet to reach). In other words, we probably haven’t been reducing the “real” incidence of prostate cancer at all. More likely, we have been reducing the prevalence of undiagnosed prostate cancer! Does that make sense?
2. Why is the incidence now 30% below the peak? Because we are now getting closer to whatever the “real” annual incidence of newly diagnosable prostate cancer is, because we must be getting closer to eliminating the accumulated pool of undiagnosed disease (at least in the USA). Which also explains why the vast majority of men in the USA are now diagnosed with T1c or at worst T2 disease as compared with the situation nearly 20 years ago.
3. Why does the reduction in incidence have “nothing” to do with a reduced mortality rate? This is pure speculation on my part, but I believe that a very high proportion of the cancers being discovered today — and especially those in men over 65 years of age — have never placed those patients at risk for prostate cancer-specific mortality. This does not mean that they had indolent cancer. It just means that they had indolent OR slowly growing forms of the disease that were always going to be managable with hormone therapy and other forms of care (if needed at all) until the man died of something else. By contrast, there is a subset of prostate cancer patients who get diagnosed with more aggressive forms of the disease who are rarely well-managed with first-line therapy, who respond onlyt temporarily to hormone therapy, and who are at high risk for prostate cancer death. Can I tell you how many men that is? No … but my guess is that it may be of the order of 15,000 to 20,000 new cases a year in the USA who may have a potential survival of only 10 years or less from the date of diagnosis. This would correlate with the gradual decline in prostate cancer-specific mortality, which is now down to about an estimated 27,500 for 2009. The question is going to be whether we see this decline “bottom out” at about the same time as we see the annual diagnosis rate stabilize.
The bottom line is that, at least in the USA, where PSA testing is widespread (albeit neither consistent nor competently applied) we are seeing a decline in both incidence and in disease-specific mortality — but I am as near to convinced as I can be that this is an ongoing effect based on an accumulated pool of patients whose risk has developed over some 30-40 years, and we have yet to reach a stable plateau in which we can say with any degree of safety that there is a real and predictable “annual incidence” of prostate cancer as opposed to an annual incidence of prostate cancer diagnoses!
Contrast this to a disease like pancreatic cancer, in which the vast majority of patients are still diagnosed with relatively late stage disease and die within 12-18 months of diagnosis. Now imagine that someone discovered “pancreas-specific antigen.” We would suddenly see a vast rise in the “incidence” of pancreatic cancer. But all those currently undiagnosed, early stage “localized” pancreatic cancers would probably have been there before. As yet, because we can’t diagnose them, we really don’t know how long it takes for pancreatic cancer to go from “diagnosable” at autopsy to clinically evident. It’s probably not 15 to 20 years — but it might be as much as 5 years.
Has this helped?
Thank you Mike. Very thorough and helpful, as ever.
Since I posted my original question I have been pursuing a line that made sense to me (and no one else apparently) as to whether the decline in prostate cancer incidence could be due to a change in the definition of prostate cancer. This being triggered by the puzzling shift in Gleason scores -– the so-called migration to higher levels -– and the lack of Gleason 5 diagnoses in the post-PSA era.
I say puzzling because it seemed very odd to me that, given that tumours were being discovered so much earlier, more of them were more differentiated. Given the current theory that tumours change their patterns and gradings as the disease progresses, it seemed logical to me that more early stage tumours would equal more lower grade tumours.
I was told in no uncertain terms in one recent exchange that there had been no change in the definition of PCa in the US and I have to say that the protest was so strong that I decided to hunt around more than I usually do and found that I might not be wrong.
One of the things I found was this -– extracted from Guidelines for processing and reporting of prostatic needle biopsies for The European Randomised Study of Screening for Prostate Cancer (ERSPC):
“As advocated by Epstein, Gleason scores of 2 to 4 should not be attributed to prostatic adenocarcinoma on needle biopsies. We would recommend that the lowest Gleason growth pattern that can be assessed in needle biopsies is growth pattern 3, implying that a Gleason score of 6 is the lowest possible on peripheral zone needle biopsies. … Thus, in sextant needle biopsies the Gleason score can range from 6 to 10.”
There were two references, which I am trying to track down, but it seemed clear from one of the in reference to the decision of members of the Cancer Committee, College of American Pathologists that they had created an updated protocol for the examination of specimens from patients with carcinomas of the prostatic gland –- and it seemed likely that this would be what the ERSPC committee was referring to.
So … Did this mean no more Gleason score 5 tumours in the USA? I’m still trying to track down pertinent information. In an article in Insight on Gleason migration it is said that there have been no reported PCa cases with a GS below 6 since 2002 at Cedars-Sinai. And another study I tracked down said that between 1988 and 2005 there had been a reduction in diagnoses of prostate cancer with Gleason score of 2 to 5. These had accounted for 32.8% of all diagnoses and now accounted for 0.2% of diagnoses, so it seems I may be on the right track.
If 32.6% of diseases that were formerly labeled as prostate cancer are no longer labeled in this way, would the incidence reduce? I think it would -– not by the full number of cases that were equivalent to the total of 32.6%, because I would expect many cases that would have been scored as Gleason 5 are now scored at a higher number –- Gleason 6, 7 or even 8. But I would expect the reduction to be significant.
And if less men were diagnosed with prostate cancer, would less men die? Again, I think so, after all prostate cancer deaths are due to failure of major organs affected by prostate cancerr. So, as I understand it, if for example a man with prostate cancer dies of pneumonia, that would be a prostate cancer death. A man dying of pneumonia who is not diagnosed with prostate cancer would, of course not be counted as a prostate cancer death. I mention pneumonia specifically because both Catalona and Walsh have mentioned that the reduction in mortality rates in the UK was due mainly to a change of the definition of prostate cancer death which excluded deaths from pneumonia in men diagnosed with prostate cancer.
Now I have found reference to another change in the Gleason protocol -– in 2005 one of the key aspects being a recommendation was for pathologists to report all higher tertiary grade components of the tumor as part of the Gleason score. Needless to say this is contributing to further migration. A study published a year later has this to say:
“After application of the modified GS on NB, a substantial shift in GS distribution occurred: The proportion of GS 6 and 7 were 48 and 26%, respectively, with conventional Gleason grading as compared to 22 and 68%, respectively, with modified grading.”
There was no discussion on Gleason 5 disease or lower diagnoses, simply because they no longer exist. I think these are significant changes in the very definition of prostate cancer, so maybe I wasn’t such a fool after all in asking the question.
Stand by for better mortality rates in both GS 6 and 7a (3 + 4) diagnoses.
Dear Terry:
I think you may be “over-thinking” some of this.
In the past decade, the “best” pathology laboratories in the USA abandoned Gleason grades 1 and 2 as “meaningless.” This led to any form of pathologically evident prostate cancer (on biopsy or on radical prostatectomy) having a Gleason score of at least 3 + 3 = 6. This pathological approach has now been accepted pretty much globally, and the next issue (still not formally accepted by the global pathology community and not adopted within guidelines) is the differentiation between Gleason 7a and 7b and the use of tertiary Gleason grades.
However, I honestly don’t think that any of this will be shown as meaningful in terms of prostate cancer incidence or survival over time. It will simply mean that comparisons of old to new data will have to group all patients with Gleason scores of 6 or less from the older data to compare to Gleason scores of 6 in the newer data.
Obviously I am slightly “over-simplifying,” but I would respectfully suggest that you may be “over-complicating” the issue.
We may well see better biochemical disease-free survival in men with Gleason 6 and Gleason 7a prostate cancers in years to come, but my bet is that the primary reason for that is that those patients will have also been diagnosed earlier, with lower clinical stages, less cancer in the prostate overall, and lower PSA levels too. Remember that the Gleason score is only one of the five major factors (the fifth being patient age) used by the Kattan nomograms to predict outcome after radical prostatectomy.
Thank you for being so kind to me and my wild ideas Mike. Can I beg your indulgence for a little further clarification?
It seems that I must have misinterpreted the ERPSC recommendation and when they said, “We would recommend that the lowest Gleason growth pattern that can be assessed in needle biopsies is growth pattern 3 …” they meant in fact that ANY atypical cells recognized as adenocarcinoma that were not clearly pattern 4 or 5 had to be graded as pattern 3, no matter how closely they conformed with normal cells. Have I got that right?
Then, moving on to the 2005 recommendations, as I understand those, one recommendation was for pathologists to report all higher tertiary grade components of the tumor as part of the Gleason score and to report any higher grade cancer, no matter how small quantitatively. Previously, any secondary grade that occupied less than 5% of the specimen would not be reported. What was proposed was that, as I understand it, even a small percentage of Gleason 4 or 5 would be incorporated into the scoring system.
So, from a practical point of view, prior to the first change in practice, a biopsy specimen which had 90% grade 2, 7% grade 3 and 3% grade 5 would be reported as a GS 5, probably indolent disease.
Following the initial change, this same specimen would now be given a GS 6 label — somewhat more aggressive.
And with the latest proposal the score would be GS8 — very aggressive.
As you say, this may not be meaningful in terms of prostate cancer overall survival over time. But I have to disagree when you say that the only effect will be to simply mean that comparisons of old to new data will have to group all patients with Gleason scores of 6 or less from the older data to compare to Gleason scores of 6 in the newer data.
One study has already shown a signficant jump in GS 7 (probably 7a) diagnoses so that would complicate comparison considerably, but my main concern is that the medical mainstream, which already demonstrates a lamentable lack of knowledge about the disease, will not undersatnd these fine nuances and will be recommending ever more aggressive treatment for conditions that, until now, have been regarded as potentially indolent.
The Kattan nomograms are not the only ones in use. I would expect that the Partin Tables may be the best known ones used during diagnosis to try to establish the posibility of spread — and they are certainly GS driven.
Again my thanks.
Dear Terry:
1. My interpretation of the ERSPC recommendation is that they intended for all cells clearly recognizable as adenocarcinoma and that were not clearly pattern 4 or 5 to be graded as pattern 3 (but this does NOT include all “atypical” cells).
2. Since the clinical application of tertiary Gleason grades is not yet approved practice, it is hard to make any decisions about interpretation of such data on a global basis. However, most of the specialty laboratories (such as Bostwick Laboratories) do now include tertiary Gleason grades in their reports.
3. With respect to your example of the biopsy specimen originally classified as 90% grade 2, 7% grade 3 and 3% grade 5 and reported as a Gleason score of 2 + 3 = 5, probably indolent disease: (a) you are correct that if the tertiary grade is ignored, this would now be classified as Gleason 3 + 3 = 6 (and it would still probably be categorized as low risk disease); (b) however, even if the tertiary grade was taken into account, according to Epstein and his colleagues, this should now be classified as Gleason 6.5 and therefore still relatively low-risk disease, not very aggressive (as you suggest).
The problem, of course, is that there is still no consensus about how to interpret and apply information from the tertiary Gleason score.
4. With respect to the Partin tables, actually they are also driven by the patient’s PSA level and clinical stage, as well as the Gleason score, and not solely by the Gleason score. As yet, even in their most recent evolution, these tables still apply the principle that a Gleason score of 5 is possible, since otherwise the Hopkins data on which they are based would not have enough accumulated data to provide statistical accuracy. However, they do now discriminate between Gleason 3 + 4 = 7 and Gleason 4 + 3 = 7. Neither the Partin tables nor the Kattan nomograms yet take account of tertiary Gleason grades.
5. Finally, with regard to the interpretation of all these nuances … unfortunately I would tend to agree with you that most urologists just want to operate, most radiation oncologists just want to radiate, most brachytherapists just want to “brachyate,” etc. As usual, this is why the intelligent patient needs to do his homework with care. What else can I tell you? At the other end of the scale, the recent paper from Memorial Sloan-Kettering Cancer Center dealing with the skill levels of urologists has opened the door to additional light on this issue.
Thank you for your hard work, patience, and wisdom — and all the best for Christmas and the New Year, Mike.