The Opportunity in High Risk Prostate Cancer: De-escalation of ADT.
Our data is old creating opportunity.
Today, we’ll use 16 trials to make the argument that:
Our data is really very old and out of date
We should be looking to de-escalate the duration of ADT for high risk men.
(we’ll cover a lot of big trials: PACE-B, POP-RT, RTOG 8531, 9202, 0534, GETUG 18, and GU009 among others. So buckle up as we continue our “search for better”.)
As I like to semi-jokingly say:
PACE-B is best considered simply as a radiation trial.
I mean sure, it was an SBRT trial but to me, the single greatest achievement of that trial was demonstrating that we have moved to 95% or above 5 year cure rate for many intermediate risk men compared to surgery that reports closer to 80% (REF). Today we simply are better at defining the target and hitting the target than we have ever been. And cure rates have skyrocketed.
The anticipated cure rate at the time of the design of the trial was 85%. That was based on decades of data. Behind the scenes, I think many experts guessed that modern techniques might push that number towards 88% due to general improvements in our processes, but what we saw - in a large multi-institutional prospective trial - was a failure rate that was one third the expected rate.
And we have seen this same scenario play out in other modern radiotherapy trials for higher risk disease. Specially POP-RT where, if we include ADT and pelvic radiotherapy we can push control rates to 95% in high and very high risk disease. And this trial was not early high risk disease, but a cohort with an average expected nodal involvement rate of ~35%. Modern radiotherapy paired with our ability to accurately define the target has drastically improved long-term outcomes.
Which leads us to hazard rate discussions. As the population at risk of failure shrinks, any possible benefit to intensification becomes smaller and smaller. And yet, in high risk disease, we have continued to maintain a very conservative approach to not decrease the duration of ADT.
ADT duration is based on treatment outcomes when often 1 in 2 patients failed our primary course of treatment even with the addition of ADT. The data is just that old. I don’t think anyone knows exactly what high risk outcomes would be without consistent ADT usage today, but certainly, based on a wide collection of modern trials, one can feel confident that it is a fraction of the failures that we saw in the trials that define ADT duration.
Data Supporting our Current Approach is Very Old:
What does NCCN reference?
(In these massive topics, I think this is a reasonable starting point - to review all the references that make up our “standard of care” guidelines. So that is where we begin.)
NCCN has one paragraph reviewing the supporting data:
(NCCN blue blocked text - then my comments):
EBRT has demonstrated efficacy in patients with high-risk and very-high risk prostate cancer. One study randomized 415 patients to EBRT alone or EBRT plus 3-year ADT.
This EROTC study demonstrated less than 50% control in the combined arm. It ran from 1987-1995. The primary endpoint was CLINICAL disease free survival. It found an OS benefit but so very different from today.
NCCN continues…
In another study (RTOG 8531), 977 patients with T3 disease treated with EBRT were randomized to adjuvant ADT or ADT at relapse.
Again, this trial is from the mid 80s. It is so very old that a PSA was NOT required for the trial. Further, it required a machine with an SSD of at least 80 cm from source to isocenter. Again, different times. Below are the OS curves where it did produce benefit but clearly these are different times. For comparison, AJCC state relative survival at 99% today for “regional” ie LN+ disease (REF).
Two other randomized phase 3 trials evaluated long term ADT with or without radiation in a population of patients who mostly had T3 disease.
Again, these trials are so old that they examine the question of - Is radiation required in addition to permanent ADT?
UK PR07: This trial was “more modern” running from 1995-2005 but just to give you flavor, here was the radiation dose: “65-69 Gy to the prostate”. The question was, does radiation add to permanent ADT. And it did.
SPCG-7/SFUO-3: This was a similar design - ADT plus or minus radiation - again with ADT for life. And again, radiation did add benefit.
NCCN Concludes with this look at post-treatment kinetics:
In all four studies, the combination group showed improved disease-specific survival and OS compared to single-modality treatment. Patients with a PSA nadir >0.5 ng/mL after radiation and 6 months of ADT have an adjusted HR for all-cause mortality of 1.72 (95% CI, 1.17–2.52; P = .01) compared with patients who received radiation only.
In this final study (REF), the goal was to hit a nadir of 0.5 or less with combination therapy. Compare that to the current ICECaP analysis that recommends a PSA of 0.1 or less as prognostic. This old study found 1 in 4 men not reaching 0.5. In a quick review of my 120 patients within a similar cohort, I have two men at 0.5 with approximately 85% reaching 0.1 and 92.5% reaching 0.2 within 6 months of the completion of treatment.
Per NCCN, that is our data supporting 18-36 months of ADT use with radiation.
And one more: RTOG 9202
One thing I like about a Substack format is that I can add in the latest data. So I’ll cover one more study that just landed on the topic - available online February 1, 2024.
It begins, “From 1992 to 1995, 1554 men with locally advanced prostate cancer (T2c–T4, prostate-specific antigen <150 ng/ml) received radiotherapy with 4 mo (short-term [STADT]) versus 28 mo (longer-term [LTADT]) of ADT.”
Just to drive home the technique differences from today - “4 field, >=4MV to 65-70 Gy, prescribed to isocenter” - so a point not a volume, so even cooler on a relative basis. And yes, in this framework, longer was better.
But more importantly, it makes my point for me:
Everything you read on this topic relates to treatments that are absolutely not relevant to today’s approaches and today’s results.
And while, yes, the OS was improved and prostate cancer mortality was improved in this old setting, the longer ADT increased MI deaths and the trial trended (with a non-significant HR of 1.22, p=0.15) towards a similar finding of detriment for overall cardiovascular mortality.
Hazard rates matter. Treatments that have benefits also have real risks and our job in medicine is NOT to reduce these complex discussions to an advertisement or billboard. While that might be nice, science is not that clean.
And today, the reality today is quite different. The reality of treatment today is that our cure rate has simply been pushed far higher and even in the case of failures, our salvage approaches are far more effective. I believe our results are so vastly superior today that we should re-assess the role for ADT and look for opportunities to de-escalate rather than lean into studies that are clearly inferior.
I point out this data, not to say, these weren’t really great trials at the time or to discount NCCN, but rather to highlight the opportunity for modern trials to readdress the issue of ADT in high risk today. Heck, I’m old enough that permanent ADT was the standard when I arrived at MDACC. I literally first practiced based on the data that built the framework for these approaches. And I’m telling you, the disease and treatment then versus the disease and treatment today are not comparable.
The Mechanism of Action:
I think most would argue today that there is some real interplay in the effectiveness of radiation when ADT is given concurrently. Digging through that data is beyond todays scope, but somehow there is a level of synergy when the two are given together. From our large volume of data, I believe that is the most logical assessment. So today, let’s assume some additive benefit in a cohort of high risk due to concurrent ADT and radiation.
But even given that, what is the argument for continuing the ADT well beyond the conclusion of the radiation? I think that is the more pertinent question in today’s treatment environment.
Six months beyond the completion of local treatment, I can argue that there is little chance for ADT to have a synergistic effect via a direct interaction with radiation. Once treatment is complete and we are months out, additional months of hormonal treatment are, very likely, only serving to delay recurrence.
But at what cost? And I think that is question we should ask.
Today I believe we sit at a crossroad. We have demonstrated time and time again that radiation is an amazingly effective treatment for prostate cancer. Yet, we are slow to alter basic recommendations for ADT.
Just two weeks ago, GETUG 18 (a trial that accrued from 2009-2013) was presented. High risk to very high risk cohort: (T3 or T4 or Gleason 8 or higher or PSA 20-100 with no obvious nodes or metastatic disease).
It showed a couple of things: 1) high dose (80 Gy) improved OVERALL survival compared to 70 Gy and 2) 5 year progression free survival was north of 90%.
And per the Conclusion slide:
”High dose RT with LT-ADT: A new standard of care in high risk PC.”
And while I don’t disagree with that, I think it assumes we still really need the 33 months median ADT usage in that trial. And that is where I think we make pretty big assumptions. Fundamentally, the trial speaks to radiation dose - it doesn’t address ADT duration, yet both are assumed to be required.
In fact, I believe quite strongly:
Today, there is far greater risk of inappropriate intensification of treatment approaches using “new technology” for a highly curable disease. On the contrary, we should be emphasizing the strength of primary radiation in prostate cancer and looking for ways to de-escalate in higher risk disease as cure rates climb.
Two clear examples:
A recent AI analysis says that even in unfavorable risk patients two out of three patients where we give ADT do not benefit. (REF)
Or the NEPI trial: A randomized phase I/II study of neoadjuvant treatment with 177-Lutetium-PSMA-617 with or without ipilimumab in pts with very high risk prostate cancer (ISUP-GG 4+5 and cT3 plus cN+ or PSA > 20ng/ml) who are candidates for radical prostatectomy. (REF)
The greater risk, by far, in today’s world seems clearly towards over-treatment in prostate cancer. Remember that many men in the POP-RT trial are truly high risk (an average nodal risk of 35%, median PSA of 28, nearly 50% GG4 or 5, and 45% T3b or T4). So likely not that different from the NEPI trial - and yet they attained 5 year 95% biochemical disease free survival keeping it simple: Relying on good radiation with ADT. And so I’d ask:
Is this really a group that needs further intensification?
ADT is bad enough
For many men, ADT is a larger quality of life issue than is the primary definitive portion of the treatment - ie, the radiation. Today, I see great opportunity for us to decrease ADT duration for men with high risk disease (and increase avoidance in UIR disease).
We can use kinetics post-treatment to help select those cases that might have the best prognosis if that adds comfort. In this high risk cohort, where ADT is current mandatory, diagnostics might have the opportunity to point out those where we likely have very little clinical risk to de-escalate. But regardless, broadly I believe there is likely real opportunity to shorten ADT duration for 80%-90% of high risk men to ~6 months total duration from our current NCCN recommendation of 18-36 months.
(This estimate is based on the ICECap PSA kinetics analysis and my own database / kinetics data that I have described in detail previously).
I see this large cohort as a “forgotten” subset where we are using old data that needs to be “modernized”. When you look back at the data, I think it is surprising at the lack of current information in such an important and fundamental aspect of our approach for high risk disease. Today, I believe there is real potential for improvement across a broad spectrum of high to even very high risk patients to better define ADT duration. I think it is highly likely we are over treating the vast majority of men with the improvements in modern radiation and staging.
And briefly, “second generation” anti-androgens aren’t the answer - recent data from just six months ago demonstrates more cognitive decline and functional toxic effects as we evaluate the toxicity of these drugs (REF). Simply stated, ADT approaches - both the old and the new carry real risks and we need a new path forward.
What about GU-009? The Predict-RT trial
As a refresher, this is the Decipher High Risk trial: de-escalate 2 of 3 men (move them to 1yr of ADT) and intensify 1 in 3 men (add a second generation AA and keep the duration at 24 months) based on a strict dichotomous evaluation of Decipher risk score.
Honestly, I’m a keep it simple guy so I think about it this way. This will test 100 men to intensify 33% of them. Yet, in prospective data only 1 in 20 fail at 5 yrs. (think POP-RT). Ok… at least we are de-escalating some, but why 1 yr? Why not further to 6 months if this test is so great?
Remember, even by their own data Decipher has a maximum hazard rate improvement of ~1.28. So lets assume this math / advertising number remains (which I highly doubt). The 5 failures turn into 4 and we have tested 100 men to intensify 33 to help 1 (causing extra toxicity in 32). I won’t even touch the recent improvement in salvage therapies further weakening any “benefit”.
What is the other approach?
I’d argue we can simply de-escalate nearly all high risk and use PSA kinetics post treatment like found in the ICECap data and then simply extend ADT / intensify those men that don’t meet 0.1 at 6 months post-treatment. For comparison, this approach would test 0, removes 75% of the total ADT for this cohort (Math: Predict Approach: 12x67+24x33, Alt: 6x85 + 24*15) intensifying a group PROVEN not to be responding as well with early additional intensification.
Those are the two approaches. I favor the later quite strongly. In fact I’d bet money it outperforms the more costly, more intensive approach on multiple levels - costs, QoL, and likely breaking even on MFS. (My more in depth analysis of the data behind Decipher is HERE.)
One quick laugh:
Sitting at ASTRO this year, I watched as we presented RTOG 0534 (SPPORT trial - post-op plus or minus ADT and pelvis - Pollack). This years data “showed” that on quality of life surveys, ADT had no effect at 5 years out - at least that was the headline.
Now ask your yourself: Is that really “truth” or “narrative”? As the slide went up, I laughed out loud at the slide - literally - as I mumbled to myself:
“Was it that they fully recovered, or have they simply forgotten how good they had it before the ADT.”
Regardless of what the surveys show, I think that conclusion is a partial truth at the very best. ADT affects quality of life. One group had 6 months of active treatment - one did not. By default, one has worse quality of life (the longer ADT group). Do they equally recover? Who knows?
I honestly would bet they do not. Time “fixes” a lot. Perspectives matter and so do accurate memories five years later in an elderly population. Then again, I guess you can argue that with long enough follow-up, we all end with a quite similar “QoL”. But because we lean too much on ADT and not enough on the strong results of radiation, we tend to support the “narrative” that: “they recover and do well”. At least that is my perspective.
(I won’t belabor the point, but I’ve looked and I do not believe any QoL surveys have ever been validated to demonstrate that equalization from two different baselines truly represents no difference.)
One Final Point:
I return to POP-RT. We treated men with high risk disease (35% average nodal risk per the Roach formula) with PSMA scans for staging, modern radiation to include pelvic nodes and 2 years of ADT.
In simple terms, 5% had a recurrence of prostate cancer (bDFS of 95%) and an additional 5% passed away due to any cause deaths (DFS was 89.5% - there was one exception in the trial with a single neuroendocrine transformation).
I’ll repeat that: your odds of dying from other causes in 5 years are as high as your odds of having a PSA LABORATORY “failure” at 5 years.
Yes, these are recurrences of the cancer, but this is an aging population as evident by the competing risk of DEATH from other causes. And how many of the recurrences can be salvaged with SBRT or kept to remission with simple ADT? I’m not certain, but it is a growing percentage each year.
So with that context, is this really a group that one even needs to consider “intensifying” - leading to more falls and cognitive decline? Far more important is to look at de-escalation for the many men who are lower risk than this cohort who still lie within NCCN “High Risk.” Our primary questions should be: When can we use less ADT and when is the pelvis really required for 95% bDFS? To me, those are our priorities.
Instead trials like GU-009 create a false “new standard” without ever demonstrating superiority to a far less complex, less costly, less intensive treatment approach. It’s bad medicine and it’s happening in subtle ways more and more.
A Great Opportunity:
On this site, I talk a lot about searching for a path to better - more effective treatments - less toxicity - better outcomes. The strongest approaches towards better do it via simple reproducible pathways.
Higher risk prostate cancer appears to represent a great opportunity to find a path towards better by leaning more on good radiation and limiting our reliance on ADT - searching for paths to de-escalate treatment in this cohort. Our data is old and our results today are so very different that we should reassess the duration, and perhaps even frequency of use, in this cohort as a real path towards better outcomes for our patients.
At least that is my perspective. As always, re-read the data and ask yourself if we are really asking the questions that help push towards better?
Special thanks to Vendang Murthy MD at Tata Memorial for the call which led to this piece. He is working on a trial to evaluate just this approach - 9 months of ADT along with really strong modern radiation in a high risk cohort. I think it is a brilliant concept and one that can help push us towards better and less intense treatment for a number of men affected by prostate cancer.
Look for his trial and his publications to follow. Once again, impactful research looking to push for better out of that part of our shrinking world. Well done!!
If you made it here, hit the like button or pass the piece around - it is free content and that is “my advertising”. Thanks for the support.
PSA Kinectics Discussions on this site:
(Additional Reading)
If you are interested in PSA kinetics, I wrote a ton about them last year. Below is a subset of the most pertinent ones to today’s discussion.
Recent discussion with a colleague about "escalating local therapy and de-escalating systemic therapy when aiming for cure of a localised cancer (vice versa for metastatic ones)" motivated by your writings. Thank you Dr Storey.
Interesting...I'll share with my husband.
My husband was treated last Fall with 28 proton radiation treatments of the prostate only at Mayo (Phoenix), is on Orgovyx for at least two years and Abiraterone/Prednisone for two years. Cancer localized per MRI and PSMA/PET scan
Gleason 9, 8 out 20 cores had cancer, PSA 46.6. I believe the Stampede trial was the inspiration for the current treatment plan.
It was not recommended to have radiation of the nodes/pelvis.
Is another look at treatment plan warranted?
He will be 74 in a few months...quality of life is a concern ...he has coronary artery disease, chronic fatigue syndrome....previous history of bladder cancer (2020)
Any feedback appreciated!