Prostate Health 2013

MD Anderson Cancer Center
Date: October 2013

John Davis, M.D.: I just wanted to wrap up with a short talk on a little bit pounding down on these--some of these themes further.  A lot of what gets talked about in prostate cancer is sort--again, risk assessment, time lines and so, this is a bit of an experimental talk.  I'm going to give you a few text slides, and maybe I'll give you the analogy of, you know, an empty house looks neat, right, 'cause it's empty and a messy house doesn't look so good but it has a lot of stuff in it, so maybe the idea of a house that's organized and has good stuff in it.  And I think one of the challenges for patients trying to investigate, you know, how this dilemma should be solved for them is how to organize all this information.  Because obviously, you've heard and read things up to the day, you get another two hours today.  And you'll continue to read in here more and the better you can organize new information the better it just doesn't seem so chaotic. 

I hope that analogy makes sense.  So, and then after that, I'm just going to run through a bunch of pictures.  Just images to drive home all these themes, few more questions and we'll get on with our Saturday.  So, this is the idea.  Again, assessment of prostate cancer with words like staging, morphology, problems with prediction, you've seen and heard examples already.  Assessment of one man with prostate cancer, and then sort of this noble way of thinking from our center of sort of biological classification of prostate cancer which is sort of an attempt to tie together multiple observations that can be useful.  So, again, staging thoughts.  I mean, everyone talks, when they have--if they hear of someone with cancer, they'll say, "What stage is it?"  And you're basically asking, where is it and how much is there?  So, for prostate, as we talked about, you do a 12-core biopsy.  Each one is looked at for grade, number, percent, and involvement.  To some degree, that gives the clinician initial idea of is it contained, is it likely to be aggressive locally or even metastatic. 

You can look at lymph nodes as we've talked about volume, location, and different imaging techniques can pick up different amounts of tumor or could be in the bones.  That's another metastatic site and then you have to look at number location and in some cases, an old injury or some arthritis can look abnormal.  So, a lot of imaging of the bones is trying to figure out what's real and what's, you know, just getting old, you know.  And then the odd situation, believe it or not, it didn't really get talked about.  But prostate is one of the few solid tumors where the staging system does not look at imaging of the local tumor.  Like when we think of the staging of prostate cancer, currently it's still the physician do an accrued exam with the finger, the digital rectal exam.  It's not the MRI.  Almost every other solid tumor, some staging picture is part of staging.  So anyway,  just a little background on that. 

What about morphology?  Actually, I think morphology as I've previewed earlier really drives a lot of thinking.  What does it look like under the microscope?  Gleason scoring, it's pathologist named Gleason from the '60s, remarkably useful as pure morphology, what does it look like on a microscope looking at architecture?  And in case we've skipped over and often we do assume some knowledge, but we get two numbers.  The first one is the primary pattern.  The second one is the secondary pattern, and/or its major pattern, minor pattern, if you will.  So, if you look here in order of increasing order of significance, Gleason 3,3, 3,4, 4,3, even though these two numbers equal 7.  4, 3 means it's predominantly pattern 4, potentially more significant and aggressive.  And then you get the latter combos 4,4 and combinations of 5.  They may be ordered--what happened to 1 and 2?  Well, they were sort of described on benign surgery from a long time ago, essentially for prostate screening and biopsy, 1 and 2 patterns are kind of thrown out so to speak.  So 3,3 is the lowest.  5,5 would be the highest.  Most people have combinations of 3's and 4's, almost 85, 90 percent.  There's only a handful--get that 5 number in there.  But there's another--other interesting observations that can be made.  Dr. Kim showed you example of how you can do a molecular profile to run more testing.  But sometimes, pathologists can throw in a few extra observations that have been looked at, lymphovascular invasion, perineural invasion.  I'll show you pictures of how people have analyzed that.  So--and the key point now, morphology is often a little more prognostic than staging alone.  But you look at both and there's whole sites that have put together the mathematics of that into nomograms that can spit out different predictions of different events down the stream.  So, another organization tool, so you've talked about risk, morphology, staging, how you put that together and counsel someone.  I think--to me, this is my favorite slide for patients.  This--if I only gave you one slide, this would be that type thing.  When you take all that information, want to meet with one man, question one is a sample of, are you going to do active surveillance or active therapy?  Just keep it that simple.  Look at all those facts and try to categorize them.  And if it's active surveillance, then you go down that road and we explain it and do that.  If it's active therapy 'cause either you--they don't want to do it or I don't think it's the good choice, then you're into box 2. 

Now, if you show the Gleason 9 in your 50, the amount of time spent on number 1 is going to be really short, right?  Now, if you're 80 and you've got 1-core Gleason 6, you know, all the time is on 1 and I'm not going to hardly let you get to 2, unless you're really, you know, but you get the point there.  But for other reasons, if any reason, if you get to active therapy, it's mainly a task of comparing the standards.  Radiation and surgery and some discussion of alternative therapies, so then you bring back Dr. Choy to talk about radiation, surgeons to talk about surgery.  We do it as a team if people want it that way, and then you go forward.  And then really like the third priority is nuances of technique.  And this is why I'm trying to reorganize you because if you listen to marketing efforts, they start with number three.  We have this technique.  Come check it out and, you know, one and two get kind of, you know, it's hard to make an active surveillance marketing theme unless, like, the doctors are fishing or something while you're being observed, you know, as opposed to in front of a new fancy technology.  But nuances is important but this is when it gets very detail-oriented, you know, or you'd have surgery, how long does it take?  You know, radiation, how long does it take?  What are the side, you know, nuances of technique.  In fact, anything in surgery has gotten easy.  It's top three because we basically do everybody robotically. 

Dr. Choi has a little more effort on part three because he's got, you know, three different sources commonly applied.  And so people can ask me all day long but, go ask Dr. Choi, you know.  Again, risk stratification is another way of looking at it.  D'Amico, you already--I had this in there but also Dr. Kim had even better version of that.  But simplified thinking, if you're low-risk, it's sort of an overlapping thought of you either need no active therapy or one at best, OK?  Intermediate risk, if in simplified terms would be yes, you need one therapy.  For the most part, it's highly curative, small risk of relapse that can be looked at.  But high-risk is almost a different entity altogether.  You're looking at one or two therapies or some combination of therapies with combination systemic and local therapy.  Often, clinical trials are specifically looking at this space.  Dr. Logothetis, if you want to go back to the last year's meeting, gave a whole session on this topic of clinical trials and the rationale for it. 

Because we actually expect up to a third of patients to relapse over time, if you just do a standard surgery or radiation, potentially, somewhere in that neighborhood.  So, we're trying to, you know, what is the relapse rate and intermediate risk is, you know, 10 percent or less.  And really, the relapse rate for low-risk, it's not quite zero but it's usually attached to what we showed you earlier, significant errors in diagnosis, if you will.  Obviously, if you take someone to surgery thinking you're doing Gleason 6, and you find Gleason 9, you know, everything is different.  So, active surveillance organizations. 

Again, you'll hear more about this but let me just give you--just all the details are there if you want them but look at the three bullet points and that's your--that's how you organize your drawer on that one.  What are the rules for inclusion?  Rules haven’t been finalized but that's the theme.  Rules for monitoring and rules for changing back to active therapy.  And a lot of interesting--even the statistics Dr. Kim just gave you--or literally, just tabulated recently.  The majority of people who are reclassified as high-risk with just incremental changes, so they stayed on active surveillance, but then some were treated on delayed fashion.  You're down to like five percent.  At five years, you know, that population will get followed further.  Again, radical prostatectomy, what are the issues for that in low-risk?  You know, it does play a role if someone has 10 cores of Gleason 6.  That's different than 1 core and if they're very young, that may drive them towards treatment.  There are some people who refuse surveillance 'cause, let's face it, there have to be repeat biopsies, tests, and actually this whole topic is a great, you know, you can have a whole separate debate on this because one expert could give all the facts that Dr. Papadopoulos provided and show you that over-screening and overtreatment or overdetection leads to overtreatment.  Even if you have that release valve of active surveillance that being said--I mean the--it's kind of a sad quote, but a prominent urologist came as a visitor here recently and said, "One way to prevent prostate cancer is to stop looking for it."

But most of us know that if you do that on a large scale, the bad ones slip under the radar and become difficult.  So, you're caught in this vice grip of you have to screen fairly well to find the bad ones, and as a sort of a secondary byproduct, you'll find the small ones.  And eventually, that's a burden for them to get retested.  So, that's really why this is a shared dilemma where men have to decide between those pathways.  So, for--again, with intermediate risk, if we're going to consider surgery, generally, it's--we think it's indicated, it's successful.  Mostly, we do nerve sparing, you know, the surgery is a little more of a friendlier version of the surgery. 

High-risk also appropriate in well-selected patients off and on trial.  Likely, it's a bigger surgery in terms of side effects 'cause you have to take wider margins and more lymph node tissue.  So--and Colleen always asks me throw in abbreviations 'cause we always just, you know, share with you all with all these terms.  But I mean, are you--do you feel expert by now?  Can you pass all these PSA, test marker, PCA, prostate cancer, Gleason I already did for you, RP for radical prostatectomy, XRT would be external radiation, RCT, randomized clinical trial, some people showed you pictures of randomized trials as opposed to, you know, opinions of the expert.  AUA, we threw around left and right, American Urological Association.  So, once you dive into these, all of a sudden, you have to have all these abbreviations down the heart.  And trust me, when I go to basic sciences lectures, they talk about the XP 3, 4, and 5, 6 gene.  I get lost in about two slides.  So, it happens to us too.  You know, obviously experts forget about the language that goes into their field.  Let's just play with images and figures.  The theme again is thought process and time and what changes it.  Dr. Chapin showed you versions of this.  I mean, just the reminder of the images of the prostate was placed way down the pelvis where God did not want us to go.  If you have the rectum nerves, bladder, bones, it is easily colon surgery, that's easy.  You can just pull it up, there it is, you know, but don't tell them I said that.  A front-end view of the prostate, again, it's showing why radiation or surgery is tricky to do.  This is like taking the bones out.  So, here's the rectum.  Here's the prostate, urethra, bladder.  There's not a lot of spare tissue here.  There's not a lot of free space. 

Technology is certainly enhanced to our ability to get down there with surgical tools or radiation beams.  But, you know, if you can just let it alone, if it's not going to kill you, then that's Dr. Kim's challenge, figure out how to let--just let it be, so to speak.  I won't go over biopsy a whole lot more but it shows how, I mean, tissue drives a lot of thinking, lots of ways to get it with schemes and patterns and what have you that does--as some of you--how many people in the room have had a biopsy?  I should've asked that earlier.  So, you'll know.  This is a very mechanical, practical thing.  You know, you lay on a table.  You've got a needle gun.  You got a bunch of specimen bottles.  You got a PA's--thinking--does anyone know Bob for the picture?  Anybody had a biopsy by Bob?  He retired after of a 20-plus year experience, although I heard he's coming back part-time or something like that.  A little bit of anesthesia.  We have a much newer machine about now.  What happens when you do a biopsy and it leads to treatment?  Here's the first slide of time. 
I always called these the sales.  The first guy that did this had this idea of displaying prostate cancer outcomes over time based on predictive factors.  So, the first idea was what happens if, you know, you never had curative treatment and you just kind of watched it for 15 years based on age and Gleason score.  So, anything in black means the patient died over time.  So obviously, at time zero, there's no black, right?  Gray means they died of something else along the way.  Now, in this--this is a different version of the same technique.  This is people who've had a prostatectomy based on age and grade.  So, the box up here that has almost no visible black is age under 60 with Gleason 6.  So, at year 5, nothing has really happened, right?  Year 10, maybe a few died of something else, car wreck or something.  15, maybe more have gotten heart disease.  And here's 20, and that you can--I mean, I can see a little black, and maybe you can, you know.  Let's go to the bottom right.  That one is easier.  This is age, you know, 70 to 90 with Gleason 8.  Yes, over time, you know, people can die of prostate cancer.  So, this is just a way to look, at over time, the effect of our intervention, how we, again, reassess risk as you go, as well it would be a way to show the effect of an intervention. 

So, here's a little example of time.  Here's known.  This is the standard Kaplan-Meier curve.  This was the original study of just sort of clinically-detected prostate cancer in a European location.  Watchful waiting versus radical prostatectomy, and then you look at it, probability of death from any cause, that's prostate or none, you know, they start, everybody was good at baseline then you go to 3, 6, 9, 12, 15.  And statistically, these curves started separating a little bit.  Watchful waiting add a few more events than radical prostatectomy.  The joke in the field is if you can get a laser pointer between two lines then it's a real effect, right, so.  You have to have a bigger effect to get them out, you know.  Now, but then they changed the rule on you here.  This is another, the same paper.  They said, "Well, what if you only checked people from 65 and up who got compared then what happened over time?"  These are together.  Then they said, "What if you're less than 65 and let’s see parts?"  So, if you look at this trial and if you look at the pivot trial Dr. Kim went over, those are big randomized trials.  You can kind of like put people in some practical categories.  If I look at a 50-year-old with intermediate to high-risk cancer, I would say there are two big trials that have looked at benefit of treatment and you fit under both qualify--you know, you've got moderate to high-grade cancer and you're under age 65.  Two trials are telling you, treatment is going to be better.  If you're 75 with Gleason 6, I would say there are two indicators of benefit and you don't have either one of them.  Well over 65 in low-grade cancer.  Of course, it gets tricky if they have one feature but not the other.  But it can give you some basic organization of the effect of doing all these.  Now, the other end of the scale is, what happens when you look at very well-selected low-risk patients like Dr. Kim showed you where this yellow line over time are people who die of other things after a prostate cancer diagnosis on surveillance.  But they were aged over 70 when they started, so yeah, 2, 4, 6, 8, 10, you know, people do die as they get older of other things. 

The blue line is people who are under age 70 when you started the clock.  So, it's a smaller line.  The two lines that you can barely read, right?  Can you see that there are lines down here?  These are people who died of prostate cancer.  Well-selected, lower into the scale, and there are two lines down there, above or below 70, and they're on top of each other so you probably can't even see them but you see the notation up here.  So, again, they're unusual to die of prostate cancer untreated in a study that's, you know, easily 10 years of follow-up, maybe a little bit more.  So, the benefit sort of starts there and moves beyond.  I'll come back to that theme of--all right, well, what happens if you're 10 and beyond?  Well, the thought is if you've got real cancer, it'll start spiraling out of control.  But if it's not destined to do that or if that's outside of your sort of window of longevity, then that's a consideration.  You already did screenings, so I'll move fully on.  Again, superimposed on all of these predictions are the fact that we don't always have the right information that's why we spent time on molecular risk and imaging risk 'cause here's just a classic case that Dr. Kim and I published a few years back. 
Let's just look at all the prostatectomies that had 10, 11 biopsies and met low-volume criteria.  So, this was the one where the biopsy showed to be Gleason 6, less than a millimeter, 1-core positive.  The morphology looked good, Gleason 6.  The staging stuff looked good.  But these patients, on their own said, "I want a radical prostatectomy." 

o, they did it, and had significant volumes of unlike this one, you can imagine, the biopsies just didn't reach that far.  You know, this one, you're seeing one cross-section, maybe when you veer off a centimeter or two, it's gone already.  You know, tumors are not perfect spheres.  Meaning they can have, you know, different dimensions in all three, really.  So maybe, we were just hitting this one perhaps, but dominant tumor, it was still Gleason 6 but very different from what we'd biopsy.  This is the anterior image that we published, same idea where at best maybe, we were just hitting a little piece of it up there.  And so, at the--we're always having to look at diagnostic problems with risk assessments with a cancer that shows as lethal variant and non-lethal variant in a slow-growing process.  Piece of cake, right?  So, anterior tumors, actually, here's the ultimate, and I always like shown.  This is for Dr. Kundra.  This is like the pre-MRI case.  First year of my practice here, a guy comes in from Dallas, four negative biopsies.  As his PSA quietly but surely went from like 10 to 20 to 50 to 100.  On ultrasound, he had 110 grams of tissue, very large prostate, normal is 20.  So, he's got five prostates rolled into one.

I'm sorry, his TRUS volume is 175.  His PSA was up to 110.  So, he's got even more prostates tucked away in there.  On the fifth biopsy, we hit one tiny core of Gleason 8, then as it--literally in '06, it was kind of like, "Hey, why don't we try an MRI just to be wild and crazy?"  You know, so--so as a noble thing, we did an MRI.  And basically, this is all tumor up here.  And we did a prostatectomy on and we had seminal vesicle invasion, Gleason 9, positive margins, positive nodes, detectable PSA.  And he's still alive eight years later but he's on hormones off and on, intermittent hormone therapy, we call it.  You know, maybe the survivorship would be OK.  But I don't--I haven't seen the sense, and I think it's 'cause when we start getting shown big glands, big PSAs, we're not just satisfied with three negative biopsies.  We're pushing it harder with imaging and other techniques.  So, again, we're heavily influenced by having a pathologist who once we give her a radical prostatectomy specimen, she covers it in five different colors, I think.  I only throw this up there.  This is maybe a self-promotion.  I mean, a lot of people are good at prostate surgery and radiation. 

I do love the fact that we have a lot of good team players.  We have a pathologist who's read every specimen since the '80s basically.  Five types of ink and she basically writes an essay.  Now, the patients complained though 'cause it takes her about four, five weeks to punch out the report, and you're having to wait on pins and needles, but once you get it, all this detail is laid out, tumor volume location. So, it's a rich resource for us to go back and pose new questions and understand, you know, what we did and/or what we didn't do.  So--and then once we give that to her, again, the effect of time is she shows this, you know, Gleason, anything, but stage-contained, negative margins, this is what the PSAs look at 24 months, here's five years and beyond.  You know, this is about a five plus year follow-up paper in cancer.  Once you see it out of the prostate, you see it deep down, that makes sense over time.  Basically, the Kaplan-Meier curve, I mean, we like to give you the big boy stuff here.  And not all fifth grade reading level.  Kaplan-Meier curve, simply, it's just the effect of measurement over time.  So, in this case, it's PSA relapse.  So, over here, the 1.0 means everyone's, you know, we assume were cured right after surgery.  Then at any time one patient relapses then the mathematics of the curve, it's just like it pokes the curve down, down, and down. 

That's why you see all these thick marks.  That's just the number of patients at that location as it goes down.  In here is like with a positive margin, extraprostatic with a positive margin.  This is T3b seminal vesicle invasion, so as you see more stage aggression, the effect of time changes.  This does allows to measure noble ideas.  So for example, you might say one population with seminal vesicle invasion that was left alone; you've got a curve drawn for them.  So then, the next question is, well, what happens if once they fail, you intercept them and you expose them to either new drug or postoperative radiation, can you correct that curve and get them back higher where the good people are, so to speak.  So, that's just a mathematical tool 'cause you'll see these as you move across.  So here--now, not everything you read and hear about is perfect new knowledge.  And I'll try to explain this, the best I can.  Maybe, Brian, you can correct me if I mess it up.  But pathologists will tell you things that may or may not matter.  Maybe they matter a little bit.  So, for example, they can say LVI, lymphovascular invasion.  They can show you a picture of it.  Now, in contrast to the molecular markers Dr. Kim talked about, this is free. 

As molecular markers are around 3,400 to 3,800 dollars to run a panel and people can basically get 90 percent of it covered.  Actually, it's fully covered under Medicare 'cause they haven't made a payment ruling on it.  But this is just an observation.  This is free just like all the other information.  Now, if you study that and you define it, in some populations, here's lymphovascular present--I'm sorry, absent up here, present down here.  It can certainly make two populations success of therapy look a lot different.  This is for PSA recurrence.  This is for cancer-specific survival.  And so that got published.  Perineural invasion, another observation, it seems to be more prevalent on biopsies.  Meaning, they can see cancer in around little nerve tracts.  The idea is if you see perineural invasion more likely to be outside the prostate along the nerve tracts, people can, you know, define populations with or without it, draw Kaplan-Meier curves.  In this particular example, they stratified by those same risk criteria we've talked about, low, intermediate, high. 

The finding didn't make a difference here but it did separate populations here.  But here was the thing that kind of affects my thinking of it.  Most of the people who had a perineural invasion also had a positive surgical margin.  So, one of the key tests of any new information is whether or not it gives independent value, and there's--statisticians do that for you.  They basically can tell whether or not two findings mimic each other.  In which case, they may by themselves be predictive but once you combine them, they fall off as being nonsignificant.  So, there are some little other features on a path report.  For the most part, Gleason score trumps everything, stage is helpful, and then these other things, perineural invasion and others help a little bit.  Let me move on to tertiary 5 again.  You know, it's nice having that 'cause I actually think that does matter.  So, this is Gleason 1 and 2 pattern in case we haven't shown it to you before.  Most people have 3 and 4.  Occasionally, they see just a little bit of 5 patterns, but it's so small, it can't be of primary or secondary pattern.  So, they actually, in some cases, give you a third pattern.  You're like a 4 plus 3/5 and yes, you know, people look at that together, and having tertiary 5 separated those Kaplan-Meier curves pretty wide.  So, that's another factor to consider with someone who might need extra therapy or more intensive monitoring, if you will.  Oddly enough, some post up nomograms don't--haven't even figured out how to integrate this 'cause not all pathologists read it the same.  So, if you've got the Sloan-Kettering nomograms, they don't even let you enter this data across there.  We look at it.  So, we could learn more on that.  And, again, let me just-- this is the personal value in that. 

Let me tell this story and then we'll probably wrap up here shortly.  Occasionally, I screw things up when I counsel patients 'cause you're trying to balance all these risks.  So, I have a patient who's got Gleason 4 plus 3 who doesn't want to have therapy which is fine.  If they would say, "I know there's a risk of progression and extraprostatic disease and I'm willing to take that 'cause I just don't want side effects."  Understandable and we would be fine except that then they throw in, "Oh, and by the way, I don't want do anything that would put me, my life at risk."  It's kind of like--OK.  So, you don't want to make any oncologic errors, but you don't want to have any side effects.  Well, who doesn't, right?  So--and life is full of choices.  And the other thing that you almost had that--got a little contentious with the two breaking it down is when you say you don't want to make an oncologic error, what window of risk are you talking about?  Are you talking about one percent risk or 30 percent risk of error?  'Cause here's an example.  This is some stuff Brian and I worked on.  If you just take intermediate-risk prostate cancer that has a surgery at MD Anderson, and look at the range of pathology results, here they are.  There's hundreds of cases like that.  10 percent of them have lymph node metastasis.  If the surgeon does a detailed template, that's another lecture, let's just say if you make a significant staging lymphadenectomy, 10--9 to 10 percent are positive.  So, that's clearly not contained, right?  Here you go.  25 percent had stage T3.  Five percent were upgraded to Gleason 8 to 10.  10 percent had that tertiary 5 pattern I just showed you.  That's--which is not a minor issue.  10 percent or so have the lymphovascular comment.  So then, that only left you with 52 percent of the whole cohort that just had routine Gleason 7 with no surprises, whatsoever. 

The other 50 percent had one of more these features.  If you reduce that to staging, yeah, I mean, there you go.  So, sure, this is pathologic.  If they had contained disease, some will have a positive margin or extraprostatic disease.  But this is the newer thing for a grant we put together.  What percent would have findings that would clearly trigger risk for more therapy?  So, 64 percent, none, 23 percent had either positive margin, extraprostatic disease, or seminal vesicle invasion.  That adds up to 23 percent, and by new guidelines, they should be considered for post-op radiation.  So clearly, that's not insignificant disease if they already need more therapy as soon as you're done.  Now, other people, again, six--this particular cohort around, it was six percent had positive nodes.  And then 7 percent have those other features, the tertiary 5, the Gleason 8 to 10.  These don't trigger automatic radiation but we know they're at higher risk of relapse.  So back to that patient's dilemma, how can I tell you you've got Gleason 7, 4 plus 3 and you don't want side effects but you don't want to be exposed to any being on the bad end of anyone of these predictions?  So, that's kind of the dilemma, you know, I took an hour to explain all that.  And I think they're still mad at me.  So, anyway, that's a career challenge.  So, let's just finish.  This is a--I'm going to try to reduce Chris Logothetis' brilliance to like four slides.  If it cannot be done, Jeri, don't tell him I said that.

Well, anyway, he put together this nice publication to then take all these things of risk and staging and morphology and actually try to almost redefine our thinking in terms of biologic classification 'cause here, I have these two slides of just interesting observations, mostly in advanced prostate cancer 'cause that's what he does.  But starting with the prostate, actually on the low-end, prostate cancer prevention trial, giving Finasteride, just got republished in the New England Journal.  It does decrease incidence of prostate cancer in a prevention mode but really only low-risk.  It almost ignores moderate to high-risk.  So, that's one observation.  Another one, for more advances, you can give hormone therapy and get a benefit.  But if you give chemotherapy early, you don't get a benefit.  That's different than other cancers.  A lot of, you know, my wife had breast, you know, breast cancer three times.  It's very well-worked out that you give chemotherapy early and you get--they can show you graphs just like I've showed you and they can show you the pointer in between the two lines and say, "You need chemo."  You know, in her case, it worked. 

Chemotherapy does work for prostate but only at late stages, not at early stages, kind of unexpected.  Metastases to bone versus lymph nodes have different observed patterns of outcomes.  Even these new interesting drugs that Dr. Logothetis went over last year like abiraterone given to advanced prostate cancer has about a four-month benefit but you're averaging a lot of different numbers.  You're averaging zeros with people who got greater than 12 months of therapies.  There's a different way to express that called a waterfall plot where they just show you how--which number of patients had each duration of therapy, and it's all over the place.  Yes, it adds up to four months, but for some people, it's substantial.  So, they clearly have--even though they can be staged the same way, they respond to drugs differently.  Enzalutamide has shown us the same thing.  And some of these that one of our other colleague, Ana Aparicio, works at or we call anaplastic.  They don't even look like prostate cancer anymore.  It's almost like a terminal differentiation.  So, in visual form, the model he's put together is called a spiral model of prostate cancer.  The idea of a spiral down here is that the spiral has turns and that as the cancer makes multiple turns, the biology changes, the drug response changes, and when you come off the spiral end, then you're in that anaplastic, you know, basically lethal form of it.  The other thing--the reason why I like the spiral model is that we know that for some people, we can alter it.  Meaning, if you're in turn 1 and sensitive to a particular marker except 17 then we actually start stretching the spiral.  Meaning, it takes longer before it turns again.  And other people maybe they fly right through that and you have to think of something else.  So, there we go.  Parting comments.  Back to the big picture and we're done.  Analyzing populations as I've shown you and creating guidelines is an interesting part--important part of this process but counseling one individual patient is its own work of art.

I think we accept bias and conflict of interest but I believe that we do behave with predominant interest of the individual patients.  So to do this as we've shown you, balancing risk and benefit and try to hit the right balance that the patient wants us to take.  We want to educate patients so they can individualize their treatment.  And you tell me if I'm wrong here, but as a default, I think many clinicians like being aggressive towards treatment, you know, screening, treatment, and cure, and accepting side effects.  Now, if one patient wants to go differently, that's fine.  You just have to tell us that and we have to work through how that works, right?  It's not--but, you know, for the most part, I would avoid extremes of strategy.  Meaning, absolutely no screening versus, you know, kill every cancer cell you can find. 

So, now, what's relevant to the media is that when you look at analyzed populations, then all these concepts of overtreatment cause conflicts.  I mean, it's just they all seem a lot more significant than when you're dealing with one individual patient.  And the media basically feed into that because, you know, it fits their profile of stories where there's a bad guy to be exposed, you know, the overzealous treatment people and whatnot.  And then the medical organizations as you've seen have to try to referee that and give us guidelines, advocacy, and most of our groups actually then try to reeducate the media on what should be done.  So, I think that gives you a good construct of almost anything you could read in prostate cancer.  It would fit into some criteria of something that helps with risk, with treatment, with guidelines.  Let's do this.  Don't do that.  I can't answer every question today but that's the construct.  So, well, let's call it there and generally, we'll try to spread out and take little more individual questions for, you know, another 10 minutes or so.  Thanks very much.

[ Applause ]