HealingWell
Search Close Search
☰

Can treating prostate cancer make it more aggressive? (More trend-analysis fallout...)

Support Forums
>
Prostate Cancer
✚ New Topic ✚ Reply
12
❬ ❬ Previous Thread |Next Thread ❭ ❭
Posted 3/22/2017 9:04 PM (GMT 0)
Worried Guy gave me an idea (yes, that's dangerous), pursuing a back-projection of my PSA history before diagnosis. I rolled that forward too, assuming it all sits on a linear log slope. The implications became a bit, well, unsettling, when compared to the present trend.

Are y'all up for a little more mathemagics?

The last PSA readings before I was diagnosed give a log slope of 18 months doubling time. On a side note, doing this also shows my PSA was and remained exactly "normal" from 2005-2010 at about 1.85 (a 30g organ at 0.066 production per gram), missing one minor PSA rise to 2.55 in 2011 that could have warned of a possible trend beginning.

So, starting at diagnosis and projecting that initial log slope, with no treatment my PSA would have hit 35 this month, March 2017. That's roughly the PSA value at which metastases may become visible on imaging, according to one study I read.

Well, we did do therapy after diagnosis. After radiation, and squashing my PSA to undetectable level with ADT2 for 3 years, the PSA slope has a new starting point at the end of that therapy. With rising PSA values starting in May 2016, the readings give a log slope with a 3 month doubling time (!). That's FAR more aggressive, alarmingly so if it holds. On that new slope, my PSA would hit 35 around June 2018.

So, putting those two scenarios together, it looks like all the radiation treatments, and 3 long years of hormone therapy, may have delayed the development of metastases by all of 15 months! Also, it somehow transformed the disease population into a far more aggressive type, with a doubling time that's 6 times faster.

Huh. Has anyone seen any study about such a comparison? This is based on the foundation assumption that a prostate-cancer-driven PSA follows a linear log slope. Could treatment have caused selection within the cell population for faster reproduction? Would the trend have steepened eventually anyway, as the cells mutated further or something? Did they become something even worse than the type 5 cells, of which I had aplenty at diagnosis?

Not that it really means anything now, but I'm sure curious about this. I won't bore you with the details unless someone asks, but the math is really pretty simple.

(Oh, and my wife thinks Worried Guy and I need to be quarantined from each other...).
Posted 3/22/2017 10:37 PM (GMT 0)
I am too new to this to have any studies to draw from, but I'll geek out with you for a moment (my math degree is old and dusty, but my head still likes to look for patterns).

it seems like a pretty big assumption that the PSA is necessarily on a linear log slope...I mean...I realize it looks that way in your signature from August to March...BUT your body was also still rebounding from being at an unnaturally low testosterone level due to the ADT. Do you know how your testosterone levels have climbed during the same time? Was the slope similar? It makes me wonder if the trend will level off (even if it still increases, maybe it'll be at a slower pace as your testosterone levels out).

Not trying to argue, just looking for other factors that might affect the extrapolation. smile
Posted 3/23/2017 12:12 AM (GMT 0)
It is indicated that l cancers respond to treatment by becoming more virulent if the treatment did not kill the cancer. A good example is is small cell prostate cancer. It is rare, but when found is usually after ADT. However it usually does not elevate PSA the traditional marker for PCa growth.

I too have played the PSA progression game of what if I hadn't treated where would I be today? For me based on a one year doubling rate I would be at about PSA 4,000. Of course it doesn't work that way as the increase is never linear. As the cancer becomes more mutated it gives of less PSA so we really don't know how to effectively forecast. And just how high will the PSAhetbefore we die and stop the progression? For some that may be 3,000 and for others it may be 10.

My niece had a very nasty strain of apendiceal cancer. All of it is bad. Hers responded specifically to each phase of her treatment completely changing the physical form. She made it about 18 months.

So, yes, I think many cancers respond to treatment by evolving and mutating to stay alive. Unfortunately it remains that too many cancers die only when the host expires.
Posted 3/23/2017 2:21 AM (GMT 0)

WifeofPie said...
I am too new to this to have any studies to draw from, but I'll geek out with you for a moment (my math degree is old and dusty, but my head still likes to look for patterns).

it seems like a pretty big assumption that the PSA is necessarily on a linear log slope...I mean...I realize it looks that way in your signature from August to March...BUT your body was also still rebounding from being at an unnaturally low testosterone level due to the ADT. Do you know how your testosterone levels have climbed during the same time? Was the slope similar? It makes me wonder if the trend will level off (even if it still increases, maybe it'll be at a slower pace as your testosterone levels out).

Not trying to argue, just looking for other factors that might affect the extrapolation. smile


Great question! Yes, I have plotted PSA vs. testosterone, and there's no direct relationship. Well, they've both gone up, but in no proportional fashion.

It's not really much of an assumption about the natural log progression. That's standard cell doubling math, and it does follow a straight line.
[Edit: Here's a paper exploring exactly that linearity. Section 4.2 specifically].

By the "saturation theory", as long as there's enough testosterone floating about to fully occupy the androgen receptors on the cells, then it doesn't matter how much more there is. Rather like once there's enough water to float a boat, it doesn't matter how much deeper it gets. That threshold level is about 50, the "castrate" level.

Even in May 2016, fresh off of hormone therapy, my testosterone was already 132 yet PSA was <0.1. After that, it rose rapidly, to 290 in August (PSA 0.2), 350 in December (PSA 0.5), and 544 in March (PSA 0.7).

I first plotted just the last 3 PSA log values, and then decided to plot May too calling it 0.1 (as high as it might have been, to be conservative). That point landed smack on the same straight line.

I'm thinking there may be competing populations, some wandering type 4, some wandering type 5, some heavily irradiated "normal" cells in what's left of my prostate, all producing varying amounts of PSA. So the total is the aggregate of all of them, and no doubt they all have different log slopes. That's where it may get interesting. Maybe treatment knocked out the long PSA doubling time (psadt) cells, leaving the short psadt cells to reveal themselves now. I don't know.

Of course, subsequent treatment if necessary will throw all of that out the window by radically changing the cell's hormonal environment.
Posted 3/23/2017 2:54 AM (GMT 0)

JNF said...
... For me based on a one year doubling rate I would be at about PSA 4,000. Of course it doesn't work that way as the increase is never linear. As the cancer becomes more mutated it gives of less PSA so we really don't know how to effectively forecast. ....


Jack, that's sort of what I was looking for. Do you have any source for that effect? Others have indicated it is rather linear. Of course we don't know what your untreated number "is" because you were treated, and very well obviously. Is there any real reason to suspect if untreated that it couldn't have been 4000? They do go that high. Just curious, since that's really the important part of this. Just how linear is this, anyway? My guess is we don't see that, since we don't let it grow without doing something to slow it down. You'd never just watch it keep progressing like that. Treatments of varying effectiveness would of course change the slope by affecting the cells involved.

Do the cells in fact become increasingly mutated, or do one's type 5 bizarro cells just keep reproducing as more and more of the same? Different Gleason scores make different PSA per unit volume (ng/ml/cc): 4 for G6, 3 for G7, 2 for G8, 1 for G9 and .5 for G10. "Normal" cells make 0.066/cc. I just can't quickly find the source for those numbers, but we've discussed them on the forum before [Edit: Corrected per JNF, quoting Strum. Thanks!]

In real cellular populations, the doubling process model has to assume unrestricted growth to remain on a linear log slope. Sufficient food, space, disposal of waste material, and so on. Any of those things can develop pressure limiting the population's growth.

Is there something analogous for PCa?:
- Does it eventually "use up" all of the available testosterone, for example? I assume there's some upper limit for our ability to make testosterone. Does that throttle the growth? Do folks with widespread metastatic cases, with PSA rising rapidly, find the testosterone level dropping as a consequence?

- Do little tumor populations in bone limit themselves somehow physically, by simple overcrowding or some such?

- Do they become unable to create enough blood supply channels to grow?

- Do they poison each other by generating too much waste material to get it out of the tumor zone?

In any case, the linearity range I'm looking at is way way down in reasonable values, 50 or less. I'd be surprised if there are many physiological processes that would constrain the curve at that relatively low level.

(I'm sorry, again I can hear the eyes rolling. "Oh, there he goes again. Give it a rest. Sheesh." smile )
Posted 3/23/2017 4:09 AM (GMT 0)
Very interesting questions.
Posted 3/23/2017 1:10 PM (GMT 0)
Jerry, great questions. Here are the answers, just not in order. Yes, no, maybe, I don't know, no one knows, somebody knows but we don't know whom, why would you ask that, and we need to find out. Pretty simple, right??

Regarding the PSA expression, I saw it in a book by Strum. He gave ranges for each Gleason grade and averages of 4 for G6, 3 for G7, 2 for G8, 1 for G9 and .5 for G10. And you have the .066 for benign non-stressed prostate tissue. Using some variation of this information coupled with the physical findings of my biopsy, my urologist estimated that I may have had a tumor load of about 19cc. My prostate was estimated at 38cc. So if the PCa was only within the prostate it made up about half the volume. But the ct scan showed the seminal vesicles were enlarged and a clear nodule was felt, all indicating a strong probability of local advancement.

When we hear of the really high PSA findings it seems to be in two areas. First the situation where it is the first PSA test like our good friend Todd. He was over 3,000 on his first test. Second when a person is failing and the PSA keeps going up despite treatment. We have seen men get back into the thousands then as well. There doesn't appear to be an absolute limit and I remember reading about a man that had something like 23,000 PSA at dx.

I understand that the progression could be linear if nothing else was happening, but that usually is not the case. Two things we would expect. First that many cancers will further mutate and become more virulent with successive generations. With PCa this could mean more cancer but less PSA being expressed. A friend fought for 9 years and finally small cell in the lungs did him in. He had extensive mets in the lymphatic system as well, but his PSA was never over 10. The second thing that probably occurs is that as mets become established they become the next generation of mother ships and the PSA starts going up exponentially as there are simply greater cancer volumes being created, but not necessarily higher grades. I would expect to see that in a G7 guy. Again look at Todd, he had wide spread bone and visceral mets, was G7 and 3,000 PSA.

As to progression and T levels. I have understood that some studies have linked advanced and high risk disease with lower T. Is the lower T a cause of advanced disease or a result of the advanced disease consuming more of the T to drop its level of detection?

And as you point out, there can come a time when the cancer has simply run out of room to continue to expand. I assume that is one of the aspects of organ failures that lead to the death of the host and thus the cancer.

In the immortal words of Zufus........The Twilight Zone.
Posted 3/23/2017 2:00 PM (GMT 0)
It is a twilight zone! Thanks for your thoughts. I like your random access answers.

Thanks, yes those are the numbers I was recalling (wrongly). I'll edit my post to match that.

You make an interesting point in that later, increasingly-mutated cells may produce even less PSA. That would actually lengthen the doubling time; a worse situation could have a longer doubling time. That is confounding for the psadt analysis, since either better or worse could cause a longer psadt!

The propagation of mets from mets has been knocked around here in the last few months, as part of the whack-a-mole discussion. Is it worth knocking out an asymptomatic met you can find just to keep it from seeding others? Or at that point is it already too late to do anything more, in that the metastatic-capable cells have already blown through your system and just need enough generations to double to detectable size? I seem to recall Tall Allen's position being the latter.

If PSA production is proportional to the number of cells, then that does support the cell-doubling-rate log slope calculation. Wherever they come from, doubling in place, or wandering somewhere and then starting their own doubling tumor, it's still the same number of cells that are doubling. Hence the PSADT from the log slope. By the time you have a detectable met, it is obviously proof the cancer cells have gained the ability to be metastatic. At that point, there's no hope of finding all the systemic initiation sites anymore. The key would be how long each generation takes to double. There must be some typical number of doublings to reach a health-threatening situation, so the longer each takes the longer the survival.

You can't clear your yard easily of dandelions by finding the seeds. You have to kill the first one before it seeds, or if too late for that, apply something that keeps the seeds from germinating. The cancer analogy seems to me that prostate surgery, or radiation, is an attempt to kill the plant before it goes to seed. If it has already seeded your system, we don't seem to have many things that can stop them from sprouting. We do seem to have things to slow them down though, like hormone therapies.
Posted 3/23/2017 2:59 PM (GMT 0)
All,
Great post and replies, very interesting questions and answers, need more replies and views on this, bring it on Brothers.
Hey Redwings, any chance you could post a clickable ink to your curves?
I like the site YANANOW where you can do a little plot and graph by entered your stats, and see some others too. But still, it's very limited and is a tiny representation of the whole PC community.
The other Wings
Posted 3/23/2017 3:09 PM (GMT 0)
This is a great thread. I can't participate in the mathematical aspects of it, but it highlights the variability of prostate cancer not only in its original presentation but in its progressive complexity. Since Redwing raised the question in the context of being diagnosed with outside the capsule cancer, this discussion is only or most relevant to that scenario. If one has a small amount of capsule contained G6 , or something close to that, there should be no discussion of whether or not to treat.
But if you are diagnosed with G 8 and above, capsule contained or not, the nanograms will tell you that you have a chronic condition that you will have to deal with.
I recently read that genomic testing of the original tumor is very limited for post recurrence treatment since the genomic structure of the tumor will have morphed between original Biopsy and recurrence.
Redwing's question therefore is not now answerable because the enemy is constantly donning new uniforms and taking on new weaponry and we are left shooting at old targets with inappropriate mechanisms.
It appears that notwithstanding the fact that these treatments may well be shooting at the wrong targets, they are blindly hitting some of the enemy and therefore slowing progression to some measurable extent.
What is less measurable is the side effect price one will pay to gain that extension.
As it now stands you have two choices: you can chase your cancer and treat it at every turn to extend time to metastasis.
Or you can ignore your cancer and deal with it upon metastasis.
I believe which choice you make will be highly influenced by your age.
In my mind, the paradigm shifts at age 65 plus.
Tall Allen has given us the statistics of how long it will take for your prostate cancer to kill you.
At the age of 65 , a reasonable argument can be made you should do nothing, and the argument becomes more reasonable the older you are. So many men " have to do something" whenever they are told there is evidence of cancer.
So prostate cancer becomes more of a mind game than a body game.
It has been said that the first physician you should see after diagnosis is a psychiatrist.
The point being if you don't realize your diagnosis will equally impact your mental state, you will be Pavlov's dog irrationally chasing your cancer and bringing on side effects that in the scheme of things were unnecessary.
Posted 3/23/2017 4:14 PM (GMT 0)
Great points. And yes, the psychology can wear you down. My risk here is what Paul Kalanathi cautioned about in his relatively brief autobiography as a neurologist with terminal cancer, "When Breath Becomes Air"; he said "Getting too deeply into statistics is like trying to quench thirst with salty water.". So true. Every analysis just opens vistas of new questions! I just added another layer by looking at my pretreatment numbers, and factoring out the constant baseline of healthy tissue. Thus the PSA increase was only due to tumor growth, leaving the base constant. That had an interesting effect on the log slope, since now it's a composite of a constant base + a log-linear growth term.

[Edit: Here's a graph, for those interested. I'd post the actual Excel file, but I'm not sure how to do that. It'll take a little time to make something that's more easily interpreted. Modifying the pretreatment PSA to isolate the tumor growth portion shortened the doubling time to about 11 months. So, 3 years of treatment may have delayed the same 35 PSA point by about 2 years. Also, the new doubling rate is "only" 4 times faster than the pretreatment. Whoopee.]

I suppose this could be interpreted as whether or not to treat, though I hadn't thought of it that way. It just kind of blew me away when I put the log curves together and saw the implication. At 55 when diagnosed, there was no question whatsoever about treating it or not, just what was the most reasonably aggressive path. I'm still there. If the next step now is hormone therapy again, as much as I didn't like it, I'll sure jump on it again.
Posted 3/23/2017 5:18 PM (GMT 0)
Redwing..at your age, or below age 65 ( an arbitrary to be sure cut off) I would recommend one do exactly what you are doing .
Thus, the gravamen of my comments are addressed to the age 65 plus group.
To that group, answer this question which I have posed to myself.
Take Person A. Gleason 8 or above or any grade outside the gland.
He has a prostatectomy that fails, radiation that fails, ADT that fails, second line ADT that fails, chemo that fails, experimental immuno or silmilar treatments that fail. He goes metastatic in year 8. He then treats his metastatic disease with whatever is recomended and lives to year 10.
Take Person B.He has the same cancer as A. He does nothing and goes Metastatic in year 5 .He does whatever he can do but since his tumor load should be higher since he was never scanned until he was symptomatic , he lives to year 6 .
You can insert whatever numbers and end points into the equation, you wish.
The point of the exercise is to determine how many years you are willing to give up to live without the mental trauma of the PSA cycle, the treatment side effects, the cost if you are financially constrained, the collateral effect on your family as you cycle through treatment failure , after failure.
There is no size fits all answer, but there is an answer for each person.
Posted 3/23/2017 6:17 PM (GMT 0)
Redwing! I love this question!!! I love it so much that even though I'm missing both nerve bundles and have total venous leakage in the parts still connected, I felt a twinge of tumescence 'twixt my thighs when I got into your description. I like the way you think!
I get it!
You and I both know the limitations of a single data point or a sample of one. We can obtain a result but that result will not be statistically significant. We need data from hundreds, or thousands of guys. Could we do that? Well, yes, we can! The data already exists! All we need to do is extract the nuggets. Stick with me for another minute or two.

What if we did a meta study and compared the results for thousands, or tens of thousand of guys?
How do we do that? By making a series of hypothetical patients based upon your journey and putting those hypothetical patients' stats into online calculators that already exist. Memorial Sloan Kettering has data from 10,000+ guys of all levels of PSA, Gleason scores, age, etc., with and without treatments including RP and SRT. You can compare results and odds survivability rates at 5, 10, 15 years for all the options.
Run some numbers before treatment and some after. This will help you confirm or debunk your 13 month number. (Honestly it is a bit of a morose exercise but it would be instructive.)

I did my own meta study immediately post RP. I was high risk, PSA 23, doubling time 5 months, etc and my 6 week PSA was 0.04. My doc wanted me to have salvage radiation right away since data has shown it to be more effective if applied within 12 weeks of surgery. Unfortunately, at that time I was totally incontinent and impotent and I had heard that once you are radiated you can expect to remain that way for the rest of your life. I ran the numbers and saw that without radiation I had a 30% chance of biochemical recurrence in 5 years. With radiation that number dropped to 25%. I decided that small difference was not worth the certainly of living with a wet noodle for the rest of my life. I believed there was a chance I would recover one or both functions in 3 or 6 months. I decided to forego radiation. I lucked out and at 7 years post surgery I can state with certainly I am in the 70% group.
I used the MSK Nomograms found here: https://www.mskcc.org/nomograms/prostate

Also in another thread, (the newly diagnosed 87 year old man) Tall Allen told us about a life expectancy simulation that compares life expectancy at 10 and 15 years with and without treatment. (I don't have that link in front of me now. MSK Webcore? ) That might be a useful one as well.

With this data you might be able to prove whether or not the old adage "What doesn't kill us makes us stronger." applies to PCa tumors as well as humans.
That is an interesting concept.
Thanks for making my day.

(Whew! I'm so hot and bothered I need to take a shower now!)

Jeff
Posted 3/23/2017 6:29 PM (GMT 0)
But Staying Tuned, you forgot to factor in the quality of life issues of untreated metastatic disease. This stuff can get very nasty for the patient and family. The use of the ADT, etc. may extend life and may also extend the period of time before the mets cause physical problems that could result in a lower quality of life than the treatments.

I think most people will choose to manage the disease as long as possible. The treatment side effects are usually not as bad as the advancing disease effects. That is why most all non-curable cancers (PCa or other cancer) are most always treated in order to lessen the effect the cancer presents. Palliative treatment is routine and beneficial for most patients even when it doesn't prolong life.

Of course, each situation has to be handled individually. I reckon that most people want to die of something other than the cancer. A heart attack, race car accident, jealous boyfriend, the bulls of Pamplona, skydiving...... take your pick. This becomes more likely the longer that life can be extended by the palliative care treatments.

Also you assume that death occurs a year after the disease becomes metastatic. What I have read is that for most PCa it is more like 5 to 7 years on average after being diagnosed as metastatic. That can be a long time of poor quality of life that can introduce similar stress financially, psychologically, and familial as treating the disease. Untreated metastatic disease in the bones is not only painful but causes fractures and disabilities that have to be addressed with medical care and living assistance. Probably more expensive and tougher on the patient and family than PSA anxiety, ADT cost and hot flashes.
Posted 3/23/2017 6:33 PM (GMT 0)
Afterthought while still enjoying the afterglow...

Someone skilled in the art could write a Greasemonkey script and automate this process so you can run a large number of simulations without having to type all the data in every time.
You could run 50 or 100 simulations that track both your journey and a hypothetical person who did not receive treatment Plot the results and you might end up with a statistically significant piece of information.
It won't affect your treatment in any way but it just might help answer the question.
Jeff
Posted 3/23/2017 6:58 PM (GMT 0)
Sounds like the Monte Carlo analyses we run to forecast probable investment results. Didn't want you engineers to think you are the only ones who know how to work with numbers.

It's just that we usually don't get that same "twinge of tumescence 'twixt our legs" when we run numbers. We save that for other non-numeric activities.....qualitative not quantitative...just sayin'.
Posted 3/23/2017 7:19 PM (GMT 0)

Worried Guy said...
...Memorial Sloan Kettering has data from 10,000+ guys of all levels of PSA, Gleason scores, age, etc., with and without treatments including RP and SRT. You can compare results and odds survivability rates at 5, 10, 15 years for all the options. ...

We find pleasure in the strangest ways, don't we? Glad you find it interesting; it seemed like something you might like!

Is that MSK database publicly accessible? I know nothing about Greasemonkey, but running this concept for a few people would sure be interesting. I can putter with my spreadsheet to see if I can get it more automated. Right now it's fairly manual to plot the log, get the curve fit, put it back in the formulas, and so on.

I asked a guy in one of my support groups if he'd send me his recent few PSA tests with dates. He's due for a PSA Friday, and I said I'd like to try predicting his number!

Trying some sort of simulation seems problematic. The only way I see to generate artificial cases would be predicated on the same rule I'm trying to explore. That's too circular to be of use. It would be useful if we had even a dozen or so pretreatment (i.e. at diagnosis) PSA trends, and then those same folks much later in a recurrence setting.

Of course, even if one could show something like this, it's not very useful. The only way I'd even question my original treatment choice would be in hindsight, knowing it's recurring. One can't know that at the outset. Still, it's kind of interesting to show what the actual benefit of some treatment might have been.
Posted 3/23/2017 7:36 PM (GMT 0)
Would you like to hear an explanation on the differences in you PSA kinetics?

I can't say that this is what happened, but this is what could explain it...

PSA kinetics following primary treatment is considered quite reliable. But PSA kinetics before primary treatment is NOT reliable because so many other things other than PSA from prostate cancer could be—and very often are—contributing to the final readings. This is exactly why AS programs specifically tell men NOT to use PSADT as a tool to gauge PC progression...it simply is not reliable/precise.

The non-cancerous contributors to PSA tend to be at generally low-ish levels, so IF you had some of them in your system pre-treatment, then the most significant effect they would have had would have been to elevate your lower (earlier) readings, which would have flattened out your exponential growth curve in the way your diagram indicates. The effect on your later (pre-treatment) values would be smaller and less significant because the PC contribution is a larger percentage.

Unfortunately, your post-treatment PSADT curve is probably the most accurate of the two.

Post Edited (JackH) : 3/23/2017 2:35:03 PM (GMT-6)

Posted 3/23/2017 8:27 PM (GMT 0)
I don't know if the data is available. I only know about it as a user. I plugged in my numbers (got depressed) and made my decision.
Experts say we should know why we own every stock in our portfolio. Similarly I feel we should know why we get the treatments we get. I made sure to go into it with my eyes wide open. I fretted over my decision for about a day but the data was enough to put my mind at ease. I could clearly state why I refused radiation.

As for predicting the guy's PSA. Bet him something serious, like a coffee. You'll be right on.

You know..... Maybe the curve would be closer to reality if, as JackH mentioned, you considered and subtracted out the non-cancerous component like an offset and added it in later. The curve fit would be something like f(a + b * (e^c)).
Or not.
My pee leakage curve was a 3rd order polynomial. After a year it predicted i would never be dry so I wnt ahead with sling surgery.
Now let's do something useful and figure out what the stock market is going to do. ;-)
Posted 3/23/2017 8:36 PM (GMT 0)
Jeff that is easy. We investment experts know exactly what the markets will do..........up and down and up and down and sometimes more up than down and sometimes more down than up but over a long period more up than down. Got it?
Posted 3/23/2017 9:47 PM (GMT 0)
While I haven't been able to follow all the math here... OK, I didn't try that hard to tell the truth... But, anyway, one thing that occurs to me is that there are two contributors to your PSA. One would be PSA produced by progression of prostate cancer, either local tissues that weren't quite done when the IMRT machine went *ding* or systemic disease somewhere else. The other would be PSA produced by your prostate -- irradiated but still present -- as your testosterone recovered. The contribution of prostate cancer is a function of time since treatment. Call the prostate cancer contribution as a function of time PCa(t). The contribution of irradiated prostate tissue response is a function of your testosterone level which, in turn, is a function of time. Call all that TResp(Test(t)). So:

PSA(t) = PCa(t) + TResp(Test(t))

What we are interested in is the shape of the PCa(t) function. We assume it is some sort of exponential growth. What is confounding us is the second term. Again we have some idea of the shape of Test(t). We have the data for one thing, and we expect some sort of sigmoid that starts like exponentiation but then has a shoulder and levels out. That may already have happened. The response curve for irradiated prostate tissue producing PSA in response to testosterone is also probably some sort of sigmoid that levels off when a saturation level is hit but we are pretty vague about that function.

What occurs to me is that once your testosterone levels out -- when Test(t) becomes constant -- then the TResp(Test(t)) term also becomes a constant and you should get better data about PCa(t) moving forward.

Or, to put it in terms so simple that even I can understand it, now that your testosterone has recovered your PSA going forward should tell you much more clearly about any possible recurrence and its aggressiveness.
Posted 3/23/2017 10:55 PM (GMT 0)
Thanks, guys.

Jeff, what you describe is what I did in the graph I posted. For the pretreatment numbers, I have 3 years of pre-diagnosis PSA virtually constant at 1.9 or so, which is pretty exactly 0.066*30g, as expected. I made an assumption that carried on and the eventual tumor was all additional to it. So, I subtracted off the constant, log plotted the remaining increase, extrapolated that line both directions at the same slope, and added back on the constant. I may not be saying that well, but I'm pretty sure it's what you describe.

Peter, that's really the $64 question. How much PSA is the remaining non-cancerous prostate tissue making, now that it has been cooked? By saturation theory, the function dependent on testosterone should have been saturated already at 50 ng/ml, achieved way back at the start in March 2016. Later increases of T shouldn't matter at all, not proportional to the presence of testosterone. So, increasing or stable, the "benign" prostate tissue's PSA production should be a constant. Perhaps though, it took time to restart PSA production after being put to sleep for 3 years? That's a real unknown.

[Edit: Here's a study of incidentally irradiated prostates. This study says it is quite unusual for irradiated prostate tissue to resume secreting any PSA. From the article:
CONCLUSION:
Recovery of prostatic secretory function is an uncommon event after external beam radiation. The concern that this might significantly confound new definitions of biochemical failure after radical radiation for prostate cancer that are based on progressively rising PSA values thus appears to be unfounded.]

Is there any idea how much PSA a heavily irradiated prostate should make? I've seen studies saying after modern RT, the PSA should stay at 0.2-0.5 or less. That would imply that irradiated prostate tissue should make very little PSA indeed. At normal-tissue 0.066 ng/ml/cc*30cc, or 1.9, the same value could exist post RT. So everyone who has primary RT would have post-treatment PSA of 1.9 and higher, unless RT knocks the prostate PSA output down permanently. Any ideas?

If the benign tissue's awakening is driving the number harder than any recurrent tumor tissue, then the next readings should greatly flatten out the log slope. Therein lies my hope, I believe; maybe this very quick PSA rise is not based on actual growth at all! Maybe there's little to no tumor growth. It's all dependent on what's being put out by the cooked prostate, and how fast that changes upon testosterone recovery.
Posted 3/23/2017 11:12 PM (GMT 0)
Q: How much PSA should irradiated prostate make?

A: The nadir amount, which varies a small amount person-to-person. No more, although readings may catch a post treatment "bounce." Post-treatment PSAreadings, with the minor exception of a bounce, accurately track PC progression; pre-treatment PSA does not accurately track PC.
Posted 3/23/2017 11:23 PM (GMT 0)
My understanding of the saturation theory is that the saturation level for testosterone is somewhere between castrate levels, at the low end, and the lower end of the normal range at the top. That puts the shoulder of the curve somewhere between about 20 and 175 ng/dL. Perhaps you have better data on that. Did your 50 ng/dL figure come from somewhere? Obviously, if the shoulder is higher then the contribution of your recovering testosterone would continue to accumulate longer into the process.

Also, I am not sure the saturation theory applies the same way to the PSA response of prostate tissue. It says that once there are enough activation sites active for cancer cells to successfully divide then more testosterone won't make them divide any faster. I'm not that sure the same concept necessarily applies to non-cancerous cells producing PSA in response to testosterone. It might, but it doesn't follow automatically. I think it's too complicated, too many unknowns, to make any real conclusions from three firkin' data points. The next couple will tell you more since your testosterone will probably stay fairly flat.
Posted 3/23/2017 11:25 PM (GMT 0)
JackH,

How does one figure the nadir for men on ADT?
✚ New Topic ✚ Reply
12