Is overall survival a useful endpoint for evaluating therapies for intermediate-risk patients?

In a recent commentary, we looked at the utility of surrogate endpoints in evaluating therapies. In an abstract of a paper to be presented ASCO this year, Malouf et al. examined the large National Cancer Data Base to determine whether there is an association between the use of brachytherapy (BT), external beam radiation (EBRT), or a combination of both (CT) and overall survival in intermediate-risk patients.

They found records on 122,405 patients treated between 2004 and 2013 who were staged IIA. IIA is an AJCC risk category that is similar to the NCCN intermediate-risk category, except that it excludes those clinically staged with cancer in both lobes (stage T2c). The average age of the patients at diagnosis were:

  • EBRT: 69 years of age
  • BT and CT: 66 years of age

The study provides no information about the radiation doses used.

The average survival, and the percentages of those who survived 10 years were:

  • EBRT: 109 months, 61.5 percent
  • BT: 116 months, 72.9 percent
  • CT: 116 months, 73.1 percent

Survival differences were statistically significant between the EBRT cohort and those who received the two other therapies.

The authors conclude that:

The method of radiotherapy used contributes to the survival of patients with stage IIA prostate cancer, with brachytherapy with or without EBRT having improved survival. Careful selection of the proper treatment regimen should be used.

Now, when we look at US actuarial tables, we see the additional expected life expectancy for a 66-year-old man is 16.93 years (203 months), and 14.81 years for a 69-year-old man (177 months). So the men treated with EBRT should have lived 26 months less; yet they lived only 7 months less –- a relative survival gain for some unknown reason. It is also unknown why overall survival in both cohorts was so much less than actuarial expectations.

Using the Memorial Sloan Kettering nomogram for life expectancy where intermediate-risk cancer has been diagnosed but not yet treated, and assuming no significant co-morbidities or risk factors, and allowing only for the difference in age, the expected 10-year survival statistics for untreated prostate cancer are as follows:

  • Among the 66-year-old men (the BT and CT cohorts):
    • 71 percent would still be alive, which is close to the observed 73 percent among those who were treated
    • 20 percent would have died of other causes
    • 9 percent would have died of prostate cancer
  • Among the 69-year-old men (EBRT cohort):
    • 67 percent would still be alive, which is somewhat higher than the observed 63 percent among those who were treated
    • 25 percent would have died of other causes
    • 8 percent would have died of prostate cancer

What we learn from this is that for a man who has a life expectancy of 10 years or less, watchful waiting may be a better choice than radical treatment.

We also see that it is impossible to attribute the difference in the overall survival to prostate cancer, let alone to any of the treatments received. What we needed to know is the patients’  prostate cancer-specific mortality, and we have no idea from their analysis how, if at all, it was affected. Because of the very low rate of prostate cancer-specific mortality at 10 years, even in untreated patients, it takes a very long time to be able to detect differences in the efficacy of various treatments based on this endpoint; hence, the importance of surrogate endpoints. The authors’ conclusions seem to be completely unfounded based on the data they intend to present.

Editorial note: This commentary was written for The “New” Prostate Cancer InfoLink by Allen Edel.



20 Responses

  1. Dear David:

    There is a huge difference between the values of different endpoints in clinical trials of new drugs and the value of specific endpoints in assessing the value of screening.

    I agree with the BMJ that the only really 100% meaningful endpoint in screening trials is the effect on overall survival, which is why it will always be near to impossible to prove that screening for prostate cancer can save lives. (It would take a 40-year-long study to get an accurate result.) And screening is never going to lower the risk of a diagnosis with prostate cancer!

    However, when it comes to treatment other endpoints do have real value if they have been properly validated. For example, if we can lower the risk that men get metastatic prostate cancer (even if they then die of something else) isn’t that a valid endpoint?

  2. Obsolescence analysis calls authors’ conclusions into question

    About two-thirds of these patients had a 10-year survival, per study data. Therefore, while a final date of the study is not given, but, assuming a study cut-off for follow-up of 2015, by subtracting 10 years from 2015, about two-thirds were treated during or before 2005, but not before 2004, the earliest year for treatment. A large proportion of those treated later –- some as late as 2013 (just 3 years ago) likely had very short follow-up for the purpose of an overall survival analysis of men with intermediate-risk prostate cancer, rendering their data fairly meaningless, right? Thus, the question seems to be whether we can glean points of value for treatment decision-making in 2016 from data mostly on men treated during 2004 and 2005.

    As a patient with a challenging case who was carefully watching radiation technology develop since 2000, hoping for substantial improvement for challenging cases but not seeing enough progress for some years, finally treated with TomoTherapy in 2013, I remember clearly some major developments in this period, especially in dosing, imaging, and targeting. The short version is that EBRT dosing was too low at many centers around 2000 and for some years after that as research percolated through centers that there were better results for non-recurrence with higher dosing (around 78 to 81 Gy). Similarly, it was not until around the middle of the decade that advances in imaging began to migrate from leading practices to community practice, but that took a number of years and is still continuing. One of the benefits in imaging was in more accurate placement of seeds, thereby minimizing cold spots and lowering recurrence rates. Another benefit was better dose planning and daily targeting for EBRT. These advances resulted in substantial, not incremental improvement in non-recurrence outcomes. With this as context, it seems reasonable to conclude that a large proportion of men in this study were treated with radiation that is now obsolete and clearly inferior to technology available in 2016.

    Therefore, not even looking further than this analysis of the impact of obsolescence, I’ll join Allen in questioning the validity of the authors conclusion about the survival advantage of certain radiation therapy.

  3. Doubts about “watchful waiting” for these groups


    Thanks for noticing this abstract and reporting it. Perhaps just the following statement deserves another look.

    You wrote: “… What we learn from this is that for a man who has a life expectancy of 10 years or less, watchful waiting may be a better choice than radical treatment.” A small quibble is that “active surveillance” would be wiser than “watchful waiting” for most of these men, unless they had quite short life expectancies, such as with significant other life-threatening disease.

    But even substituting “active surveillance” in the statement, it seems that treatment, especially with 2016 technology, would be able to keep alive most of the 9% of men in the 66 (average)-year-old group who died of prostate cancer, which is a hefty percentage against the 20% who died from other causes. Consider that 9% is nearly a third of the 29% who are dying in this group –- pretty large. Similarly, among the 69 (average)-year-olds, 8% is also a substantial portion of the 33% who died of all causes –- about a quarter. If it were me, I would jump at treatment in a minute with those odds.

  4. Dear Jim:

    I will let Allen get into the details if he wants to, but regrettably your comment about many of these men needing treatment or at least active surveillance is flawed by the fact that you would have to actively treat about 9 such men with Stage IIA disease to have a chance of extending the survival of a single patient by even a month. This means that 90 percent of these men would only have a chance of side effects of treatment with no chance whatsoever of therapeutic benefit, and the men who might benefit might benefit by very little in terms of their survival.

    While it is possible that one could identify a small subset of these patients for whom treatment might be a good idea, Allen’s general assessment that the vast majority of these men would have been at least as well off on simple watchful waiting is utterly correct.

  5. Having trouble analyzing this for a response to Sitemaster and Allen

    I’m not sure this discussion is worth more time for any of us, but here are some of the obstacles that make this analysis like trying to untangle fishing line.

    – The MSK nomogram is not for men with a 10-year life expectancy, but rather for all men, some of whom have shorter expectancies and some longer, depending on the many conditions addressed in the nomogram.

    – I was not able to replicate the findings stated in Allen’s article using an age of 66 and no adverse findings. At 10 years the results were only 3 deaths from untreated prostate cancer, 20 from other causes, and 77 surviving, with 59 still alive 15 years later. (Used age 66, stage T2b, Gleason 6, PSA 15). These odds would be a lot more favorable for a course of active surveillance if not watchful waiting, as suggested by Allen, than the odds scenario in the article. Of course, today, wouldn’t it take a dunce not to have at least ADT if not chemo and other recently approved drugs to extend survival for patients who developed late stage cancer, or even non-metastatic CRPC, which is treatment rather than no treatment, making this ethereally hypothetical?)

    – The average tumors for the CT group were characterized as “poorly differentiated,” which suggests Gleason scores of 8 to 10, making these patients high-risk, rather than intermediate risk, right?

    I’m thinking it’s best to let this rest and get on to other news.

  6. I concur with the Sitemaster. I meant watchful waiting, not active surveillance. Among healthy 69-year-old intermediate-risk men, only 9% would have died of prostate cancer without treatment in 10 years, showing the very slow natural course of their disease. But this is just for healthy 69-year-olds. Now imagine the case for older men with significant comorbidities who have a life expectancy of just 10 years. With just hormone therapy (which is what watchful waiting entails), many of those would have their lives extended well past the point that they would have died of something else. In fact, using the same nomogram for a 69-year-old who has all the comorbidities they list, we see that none would have died of prostate cancer, and none would still be alive in 10 years. Radical treatment (immediate or delayed) would be a horrible choice for such a man. This is why NCCN only recommends observation (i.e., watchful waiting) for men with a life expectancy of less than 10 years.

    To your other point about when the men were treated in this retrospective study, they used Kaplan-Meier survival curves to determine extent of survival. Because this was a large database with 122,405 patients qualifying for inclusion, there were still a large number who survived 10 years or more, and the difference was statistically significant based on that sample. In a Kaplan-Meier curve analysis, the ones who are in it for a shorter time contribute to the overall robustness of the data — it is not “meaningless.”

  7. Jim,

    I assume those treated with combined therapy had a Gleason score of 4 + 3 and thus their tumors were predominantly “poorly differentiated.” I used the most common findings for the IIA risk group — Stage T2a, Gleason 3 + 4, and PSA of 12 (given in the abstract). The point you seem to be missing is not the precise number, but the fact that for this cohort overall survival is overwhelmed by age and comorbidities and cannot be used as a useful endpoint. I used the MSKCC nonogram to make that point rather than to show what 10-year survival was. That, I pulled off of US actuarial tables. I hope this clears up your confusion.

  8. Huh?: “hormone therapy (which is what watchful waiting entails)”

    Allen — to my knowledge “watchful waiting” has never entailed “hormone therapy.” At the least, adequately done hormonal therapy involves quite a bit of monitoring, with PSA at a minimum, but often at least testosterone and DHT also, which would make it much more like “active surveillance.” To many of us, “watchful waiting” is typically watching and waiting for symptoms to develop with no treatment, other than perhaps some lifestyle tactics such as diet and exercise. Are you following sources that regard hormonal therapy as a kind of watchful waiting?

    I hope to add some other comments on the other issues when time permits (if ever).

  9. Dear Jim:

    Hormone therapy can be an inevitable consequence of watchful waiting — if the patient lives long enough and if the cancer progresses far enough during the patient’s remaining lifetime. That’s what happened to the great advocate for watchful waiting in the 1970s and 1980s — Dr. Willett Whitmore. He just left his own prostate cancer alone, but ended up on ADT for the last couple of years before his death. Hormone therapy in such patients remains the most effective method for controlling bone pain in patients who start to get early symptoms of bone pain from metastasis after watchful waiting.

  10. Jim,

    Watchful waiting means that the patient has no intention of pursuing curative treatment or any other treatment whose purpose is to stop progression until it becomes necessary in order to prevent or stop pain or skeletal events. He monitors progression, often with PSA and bone scans,and at some point may decide to have a palliative therapy, which is almost always hormone therapy. Sometimes he will decide to have spot radiation too. This is a very different use of hormone therapy from that used by men who have longer life expectancy and are using it to delay progression. There are many different ways in which hormone therapy can be used. I’m glad I could clear up your confusion about this.

  11. Okay Allen. I guess, based on my own experience with hormone therapy and familiarity with that treatment, I consider it as effective in delaying progression for almost all users for varying lengths of time, with those employing combined and triple hormonal therapy generally enjoying more effectiveness. In other words, it sure seems like a “treatment” to me, though rarely curative.

    Sitemaster — It always makes me cringe when I think about those men who have waited so long before actively countering prostate cancer that they have bone mets. Of course that happens, but it is sad, especially for such a great physician as Dr. Whitmore. Unfortunately, he passed on before much was known about better forms of ADT and how to employ them. The doctors I follow closely are convinced that approach — delaying ADT — is nuts, and it kind of drives them (and me) crazy; they see many patients who have been on that approach and come to them with PSAs nearing or exceeding 100 (and much higher); sometimes those doctors are able to catch the disease before it has developed too much momentum. Their experience is that hormone therapy — a.k.a. ADT for androgen deprivation therapy — is best used early where it is far more effective, generally providing cancer control for many years. Other doctors, such as Dr. Patrick Walsh, who prominently declare that they do not generally use ADT early because they feel it has very limited effectiveness, and therefore have little experience with it in that role), are convinced, apparently on the basis of very limited if any actual experience in using ADT early, that ADT is best used when metastases have developed.

    We had a presentation by a well-regarded local radiation oncologist at our Us TOO group this week, and I was encouraged to hear that his practice has become more comfortable with using ADT to support radiation, though wisely (with the duration tailored to risk level and no ADT for men who want radiation despite being candidates for active surveillance).

    Thanks for your responses.

  12. Dear Jim:

    There is a vast difference between ignorance (i.e., just not doing anything and ending up with severe bone metastasis) and the appropriate use of active surveillance and watchful waiting in carefully selected groups of patients that can help them quite deliberately to avoid the overly early use of ADT. It appears that you don’t understand the distinctions that can be made here because you are seeing everyone else’s situations in the light of your own … which is, in fact, unusual in the extreme. Most men do not remain hormone sensitive for 15 or so years (regardless of what type of ADT they implement).

    At 68 years of age, with a reasonable life expectancy of another 20+ years based on family history, I can assure you that if I was diagnosed with prostate cancer tomorrow I would now go out of my way to avoid any treatment that wasn’t absolutely essential. If that meant that I ended up on some form of ADT for the last 3 to 5 years of my life or so … I would see that as a huge win! I would have avoided all the side effects and complications of treatment for 15 or so years, and at worst I would only need a few years of ADT (initiated early enough to avoid any bone pain, and late enough to minimize the problems and risks of castration resistance too).

    You may see things a very different way, but this is not a matter of “right” or “wrong” it is about how individuals think about managing their health, which, thank goodness, comes in many different flavors.

  13. Alert for Allen: you are at risk of becoming an advocate of primary ADT

    Allen, what you are describing as a form of watchful waiting has been described by many of us as primary ADT. There is actually quite a body of research on this, though a dearth of published trials. This is the therapy I was on for at least 13 years, though my intent was always to go for a cure when that looked promising, which I did in 2013, as you know.

    You might want to try this PubMed search: (“primary androgen deprivation therapy” OR “primary hormonal therapy” AND prostate cancer NOT CRPC NOT abiraterone NOT enzalutamide That string, with a filter for abstracts, yielded 113 hits. Here is a particularly interesting recent 2016 paper co-authored by Dr. Cooperberg, MD, well-known for his work at UCSF, and Japanese colleagues:

    Cooperberg MR1,2, Hinotsu S3, Namiki M4, Carroll PR1, Akaza H5. Trans-Pacific variation in outcomes for men treated with primary androgen-deprivation therapy (ADT) for prostate cancer. BJU Int. 2016 Jan;117(1):102-9. doi: 10.1111/bju.12937. Epub 2015 May 14.

    The conclusion of the paper drives me nuts as it suggests there is a mystery why the Japanese patients do far better (cancer-specific mortality) than their American counterparts, while the body of the abstract clearly notes that the Japanese have combined blockade 50% more often. Grrrrrrrrrr! Come on guys; connect the dots!!!

  14. Dear Jim:

    I am sorry, but I have now reached the conclusion that you really don’t understand the premise behind watchful waiting at all. You have never practiced it. It would never have been relevant for your particular diagnosis. And the paper you refer to by Cooperberg et al. has nothing even closely relevant to do with its use.

  15. Dear Sitemaster:

    Respectfully, we have a difference regarding watchful waiting.

    Regarding the timing and effectiveness of combined (here focusing on an LHRH-agonist plus an antiandrogen) or triple ADT (also adding a 5-alpha reductase inhibitor), that therapy seems to be obsolete for late-stage disease as newer and better drugs are now available and approved for that use.

    Where there is recurrence after an initial curative attempt, or as primary therapy – situations for which the new drugs have not been approved, there is a fair body of lesser-quality evidence, as in informal publications and very limited peer-reviewed non-trial research, that early ADT is considerably more effective than late ADT. My own case, which I’ll use because I know it best and can see where it illustrates general principles, suggests this, though of course does not prove it.

    In brief, starting with a PSA of 113.6 in the context of an apparent doubling time of 3 to 4 months, if deciding to defer ADT until the disease were late stage, I would likely be starting ADT 1 to 2 years after diagnosis, as at that time my PSA would be, using a PSADT of 4 months, 909 at 1 year and 7,270 at 2 years. Virtually all doctors who have weighed in on late use of ADT, including the ones I follow closely, would expect effective cancer control of likely under 2 years (being generous) in that situation. Allowing another two years for chemo before death (during the 2000s), I would have lived for an initial do-nothing 2 years, 2 years of ADT, plus 2 years of chemo, or about 6 years. Instead, at the time I was taking a shot at a cure in 2013, I had the cancer under excellent control at the 13 year point.

    Based on research by Memorial Sloan Kettering Cancer Center, and others, such duration of control for fairly early use of ADT seems typical to me, rather than exceptional. Food for thought.

  16. Dear Jim:

    You aren’t talking about watchful waiting at all. You are talking about the decision when to implement ADT of some type and in whom, and then which type of ADT to implement.

    Watchful waiting is a very deliberate process for management without any treatment at all in a carefully defined set of patients, but also with the recognition that if and when any treatment is implemented it will not be curative.

    We can debate (a) when and in whom that form of management is appropriate and (b) exactly what type of ADT would be appropriate in which patients who have been managed on watchful waiting and are found to need treatment. What is not up for debate is that watchful waiting as a process of monitoring with no treatment is a highly appropriate form of care for carefully selected men — and usually ones with limited life expectancies, e.g, a man of 80 diagnosed with a PSA of 15 ng/ml, a Gleason score of 3 + 4, and several comorbid conditions who has a life expectancy of < 10 years, and who is also extraordinarily unlikely to die of prostate cancer within that time frame. Active surveillance is not appropriate for this patient. He is never going to be a “good” candidate for any form of invasive treatment to supposedly cure a disease that is extraordinarily unlikely to kill him (perhaps 1 chance in 100,000). It is possible that, very near the end of his life, he might need a short course of ADT to relieve some bone pain, but even that is actually unlikely too.

    The whole point of watchful waiting is to avoid all treatment for as long as possible in men who are highly unlikely to gain any benefit from such treatment — thereby maximizing their quality of life.

    Your personal history is completely irrelevant to this discussion. No one with even quarter of a brain would ever have considered you to be an appropriate candidate for watchful waiting.

  17. Dear Sitemaster,

    I fully agree with the substantive points in your response of 8:02 am. This whole topic has been like a tricky curve ball for me, and I’ll be glad to drop it. My main concern has hinged on Allen’s statement that “With just hormone therapy (which is what watchful waiting entails)”. Your reply counters that statement, and I suspect Allen too really was referring to the hormonal therapy at the end of watchful waiting when pain relief was needed.

  18. Dear Jim:

    That is exactly all that Allen was referring to, which was extremely clear to me at the time. I think you completely misinterpreted his original comment.

  19. Jim,

    I agree with the Sitemaster, who has been extraordinarily patient in trying to explain what you are obviously (to us) misunderstanding. Rather than reply again, I suggest you sit on it and re-read at a later date. At any rate, none of this applies to your situation.

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 )

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: