Read the Full Video Transcript

Laurence Klotz: Missed windows of opportunity was my topic. So when does active surveillance break down in the real world? Goals are to avoid unnecessary treatment, and identify those who would benefit from treatment while they're still curable. So there's two sides to this. Unnecessary treatment due to overgrading, patient anxiety, and so on. Driving treatment, issues of work volume, fear of litigation, overtreatment of borderline Grade Group 2, favorable intermediate risk, or driven by parameters that are not considered true surrogates for progression like PSA kinetics, imaging, progression volume, progression Grade Group 1. Actually, probably that's the most egregious one, in my opinion. Failure to use existing nomograms.

And then there's the other side, inadequate or delayed treatment, undertreatment due to either pathology. Undergrading is a big problem in some parts of rural Canada, missed high-grade cancer due to either errors in co-registration of its target and so on. Progression of undiagnosed high-grade cancer due to lack of follow-up and missed progression due to reliance on imaging and PSA. I think the key is that despite all the limitations that we've had with surveillance prior to the MR and biomarker era, the metastasis rate is remarkably low. It was fairly high, relatively speaking, in our cohort, about 3% at 15 years, 0.1% at Hopkins. So we're not going to improve a lot on mortality; overtreatment's a more common problem than undertreatment.

Just a couple of brief comments. This is one phenomenon that's not as well known as it should be from Peter Carroll's group that tangential cutting on a Gleason grade, three acinus results in artifactual pattern four, and you see that particularly when there's 5% or less pattern four at ASCO. Yesterday it was a whole case of 5% pattern four, no one raised this point. This is actually a Grade Group 1 cancer with artifactual upgrading. And this is just to show that in these 5% or less pattern four, they have exactly the same RP distribution as 3+3.

Next slide. We've got four papers showing that. This is from the Canary group, risk calculators. The issue here is these are not nearly as widely used as they should be and these are a powerful way, using readily available parameters, to predict the risk of upgrading on subsequent biopsy. Really not widely used.

Next, this has been alluded to, the variation in the use of surveillance. This is from Quoc Trinh and colleagues, just to show it varies from zero to 80% adoption of AS, I mean this is as much small area variation as you get. This is from Matt Cooperberg. The average genomic risk, and this is very important, I think. The second line is the Gleason. So light blue is Grade Group 1, dark blue is higher grade and it's not quantized. And that's the key point. There's heterogeneity. And if you look at the highest average genomic risk quartile on the right, there's 2% Grade Group 1. As soon as you have elements of pattern four, it increases sevenfold to 14%.

So we have these cases of Grade Group 1 that have the worst Genomics, something to be aware of. Next, and this is the upgrading issue. Ruth Etzioni has the best paper on this, which Theo alluded to. This is our data, 1% per year, it's all in the same ballpark of one to 2%. So this happens. This is one of the issues of surveillance, that patients need to be followed, 15 to 20% at 10 years.

There are a few thoughts on the role of fear. So there's a sense that the role of fear is driving a lot of intervention. It's a U.S. problem much more than a Canadian problem. For what's that worth, it was 10% in our cohort. I think it's less than that now. Late acceptance of the concept of indolent cancer is much more widespread. 15 years ago you said to a patient, "You have cancer but you don't need any treatment." And they looked at me like I was kind of a lunatic. That doesn't happen anymore. So I think it's becoming less of a problem. So there's room for improvement in all these areas without changing the term. And obviously, changing the term is something that we're here to discuss. Thank you.