Christopher Wallis: Hello, and thank you for joining us for this UroToday Journal Club discussion. Today we're talking about a recent publication entitled, Long-term Outcomes Among Men Undergoing Active Surveillance for Prostate Cancer in Sweden.

I'm Chris Wallis, an Assistant Professor in the Division of Urology at the University of Toronto. With me today is Zach Klaassen, an Assistant Professor in the Division of Urology at the Medical College of Georgia.

You can see here, the citation for this recent publication in JAMA Network Open. Active surveillance has become the standard and most recommended treatment approach for men with low risk prostate cancer. And the rationale of this is that, we survey men with conversion from an active surveillance approach to an active treatment approach when they have disease progression. The goal of this is to minimize adverse effects of treatment, without increasing the risk of prostate cancer related death.

While uptake has been somewhat slower in the United States, in Sweden, active surveillance uptake occurred quite rapidly for men with low risk disease. As of 2019, more than 85% of men with low risk disease are preferentially treated with active surveillance.

An alternative to active surveillance that's less invasive is watchful waiting. And when compared to active treatment, the use of watchful waiting has shown an increase in mortality for patients treated in this manner. However, this often takes quite some time to manifest, and it may take several decades following diagnosis.

Because active surveillance did not become widely utilized before 2005, there's a general lack of data, in terms of active surveillance outcomes, beyond 15 years. However, due to increasingly long life expectancy, many men who are chosen to be treated with active surveillance may live for many years, and even decades, following the initiation of this treatment approach. So we need data to inform these men about potential long-term outcomes of surveillance.

The authors in this paper sought to understand long-term outcomes following active surveillance, using extensive modeling approach, modeling the disease trajectory, as we may consider multiple subsequent treatments for disease progression before men move to prostate cancer related death.

So to model this, the authors used a state transition model with the goal of assessing 30 year prostate cancer trajectories, men with low or intermediate-risk prostate cancer who undergo initial active surveillance.

You can see here, that to inform their modeling, they used data on men in Sweden, treated in the context of the prostate cancer database. The men were aged 40 to 75 years of diagnosis, and had very low, low, or intermediate-risk prostate cancer, diagnosed between 1992 and 2014.

They limited their cohort to patients within the dataset that had full disease trajectories to inform the models. They utilized standard data linkage, and used the National Patient Registry to collect information on comorbidities, as well as repeated prostate biopsies. And they used these repeated biopsies to define patients who are treated with an ongoing surveillance approach, as compared to watchful waiting. And so, when men ceased receiving regular biopsies, they were considered to transition to a watchful waiting approach.

They further used data from the Swedish Cancer Registry, in terms of pathologic characteristics, as well as survival endpoints.

This is the state transition model here, and it's premised on a few assumptions. The first of which is that, outcomes following transition to treatment after surveillance are equivalent to, if these treatments were utilized as primary therapy. So for example, if a man receives radical prostatectomy after a period of surveillance, when adjusting for the baseline characteristics at the time of treatment, we assume that the outcomes will be similar, whether they had this initial period of surveillance, or whether they moved straight to prostatectomy following diagnosis. And this is supported by observational research.

The authors performed a simulation of 100,000 men, utilizing all combinations of patient age, prostate cancer risk group, PSA level, and Charlson Comorbidity, and modeled this out to age 85 years, or 30 years of followup, whichever occurred first.

Each man's vital status was assessed at four week intervals. And at each time step where a man was still alive, the authors assessed whether they had remained on active surveillance, whether there've been a treatment change. And so you can see in the state transition model, that men could move from active surveillance straight to death from other causes, or they could transition to watchful waiting, to radical prostatectomy, or to radiotherapy. And then from each of these, they may move on to further treatment approaches. And treatment could occur within the same risk group, i.e., without a change in disease characteristics, or following a transition to a new risk group, i.e., grade or stage progression.

The authors calculated time specific prevalence estimates according to age, comorbidity level, and prostate cancer risk group; to examine the association between prostate cancer death, and the proportion of life-years after diagnosis without active treatment. The authors made a few assumptions, one, assuming that a lower risk of death, and two, that more life-years following diagnosis before the initiation of treatment were both indicative of greater benefit from an active surveillance approach.

They further undertook sensitivity analyses, acknowledging that in 2005 there was a change in histologic grading in prostate cancer, which resulted in a number of men who would previously have been categorized as Gleason six disease to be now graded as Gleason seven. And so, to account for this, they took 5, 10, 15, or 20% of men in the very low risk groups and upgraded those to a less favorable risk group. Thus, reflecting this shift.

Now, I'm going to hand it over to Zach, to walk us through the results of this interesting modeling study.

Zachary Klaassen: Thanks so much, Chris. This is the characteristics of men with prostate cancer in the database who under what active surveillance or watchful waiting, and we'll focus on the right hand part of the slide here, which is the active surveillance patients.

You can see that the median age at diagnosis was 67 years. The majority of these patients were over the age of 66, 30.6%, 66 to 70, and 19.9%, 71 to 80 years of age. As you can see, in Sweden, there was a tremendous uptick in active surveillance over the years of diagnosis. So going all the way from 1992 to 2014, you see a dramatic increase in the utilization of active surveillance, up to 45.7% from 2012 to 2014.

In terms of tumor stage, majority, three quarters, were T1 C. Not surprisingly, the majority of these patients did not have any clinical staging, so they were NX at 90.7%. Also, as would be expected with an active surveillance cohort, Gleason six patients, 90.8%. And three plus four patients, 6.9%. In terms of median PSA, 5.6. Detection, motor detection was screening in half the patients. Lower urinary tract symptoms in one third of the patients. And other symptoms in 14% of the patients.

In terms of Charlson Comorbidity Index, majority of these men, more than 80%, were healthy, with a Charlson Comorbidity Index of zero.

In terms of risk category, very low-risk, it was 34%, low-risk was 58%, and intermediate-risk was 7.1%.

So the next several slides will look very similar to this one. This is the prevalence of each state by prostate cancer risk category at each time point, and this is for very low-risk patients. At first, you can see the color delineation here at the bottom, active surveillance is in light blue, death from prostate cancer's in dark gray. You can see the rest of these, including watchful waiting, death from other causes, ADT, et cetera.

On the left here, is patients that went for conventional treatment, and patients that flipped to watchful waiting. So this is from active surveillance to conventional treatment, or from active surveillance to watchful waiting. And you can see at the top, these are broken down by age, 55, 60, 65 and 70. So to sort of summarize this slide, you can see that, among men with very low-risk prostate cancer that were 55 years of age, conventional treatment occurred in 55% of patients from age 60, 57% of patients for age 65, 44% of patients, and for age 70, 24% of patients.

Among men with very low-risk prostate cancer that flipped from active surveillance and watchful waiting, age 55 was 30% of patients, age 60 was 37% of patients, 65 was 52%, and age 70 was 76%.

For low-risk prostate cancer for those that flipped from active surveillance to conventional treatment, those that were age 55, 71%, age 60, 66%, age 65, 56%, at age 70, 33%. For those that went on to watchful waiting, 55 years of age, 27%, 60 years of age, 29%, 65 years of age, 37%, and in age 70, 60%.

Moving on to intermediate-risk prostate cancer, not surprising, we see some more patients flipping from active surveillance to conventional treatment. Among those age 55, 76%, eventually had conventional treatment. At age 60, 73%, 65 years of age, 64% and at age 70, 40%. Similar trends for watchful waiting. We see at age 55, watchful wedding, 23%, age 60, 25%, age 65, 31%, and age 70, 55%.

The authors had some explanations for the results. These are some of the highlights of the last few slides. So with regards to transition to radical treatment, this was much more common in younger men with intermediate prostate cancer, at 76% of men at age 55 years of age. And for older men, with less aggressive prostate cancer, which included 24% in men age 70 years of age with very low-risk prostate cancer.

With regards to transitioning to ADT after active surveillance or watchful waiting, this was more common in men aged 70 years with intermediate-risk prostate cancer at 27%, versus only 5% of men aged 55 years, with very-low risk prostate cancer.

This figure looks at estimates of remaining life-years without active prostate cancer treatment. The blue shading is proportion of estimated life-years without any treatment, and this is either active surveillance or watchful waiting. The light yellow was proportion of estimated life-years after ADT and post radiotherapy with active surveillance. And green is a proportion of estimated life-years after curative treatment, either radical prostatectomy or radiotherapy.

And I think, the most interesting highlights from this figure is probably the blue shading, which is a proportion of estimated life-years without any treatment. And you can see with very-low risk prostate cancer, as men get older, the likelihood of having no treatment is much higher. You can see at age 70, a majority of patients had no treatment. We see similar trends for low-risk patients. And as we get into the intermediate-risk patients, as you would expect, younger men are going to have less proportion of estimated life-years without any treatment. But even in men at age 70 with intermediate-risk prostate cancer, roughly half of these will have no treatment for their disease.

This figure looks at the association between proportion of life-years without active treatment and risk of prostate cancer death. You can see the green line is intermediate-risk prostate cancer, as highlighted here. The orange line is low-risk prostate cancer. And the blue line is very-low risk prostate cancer. And as you would expect, higher risk disease is going to have less life-years without any prostate cancer treatment, compared to very-low risk prostate cancer.

Several highlights from this figure. So the proportion of men who died of prostate cancer before 85 years of age, among men diagnosed at 55 versus 70 years, for low-risk prostate cancer was 9% versus 3%. For low-risk prostate cancer, it was 13% versus 6%. And for intermediate-risk prostate cancer, it was 15% versus 7%.

With regards to the mean proportion of remaining life-years without active treatment, for men with very-low risk prostate cancer, this was 48% for men diagnosed at age 55 versus 77% for men at age 70 years. For low-risk prostate cancer, 36% for men diagnosed at age 55, versus 66% for men at age 70 years. And for intermediate-risk prostate cancer, 29% for men age 55, versus 60% for men at age 70 years.

So several discussion points from this study. In this population based study of active surveillance from Sweden, there were several notable findings. First, that older men with low-risk prostate cancer had a long time without active treatment and low risk of prostate cancer death. Secondly, younger men with intermediate-risk prostate cancer had little benefit from active surveillance.

So as we know, there's several large institutional studies including Sunnybrook, Gothenburg, and Johns Hopkins University, that have active surveillance outcomes up to 15 years. However, there's several important differences between these institutional studies and the current study from Sweden.

First, this simulation was based on population level data in Sweden, with a capture of 99% of all men diagnosed with prostate cancer in the Swedish Cancer Registry, versus the existing series with long-term data, which began as experimental programs at tertiary referral centers.

Secondly, the follow-up schedule was at the discretion of the treating physician, which is difficult to standardize follow up schedules, and likely leading to a lack of adherence among these patients.

And this is two reasons why these outcomes are likely worse than the aforementioned tertiary cohorts, where the outcomes, especially for cancer specific survival are excellent.

So in conclusion, the findings of this Swedish cohort study suggest that men older than 65 years with low risk prostate cancer had a high proportion of treatment-free years, ranging from 53% to 70%, and a low risk of prostate cancer death, ranging from 6% to 8%. Hence, active surveillance was indicated among men in this subgroup.

In contrast, in men younger than 65 years of age, active surveillance appears to be indicated only in those with very low-risk prostate cancer.

Finally, this state transition model based on data for men who were diagnosed between 1992 and 2014 likely provides worst-case scenario that may improve in the future owing to enhance diagnostic technology, for instance, a multi-parametric MRI.

Thank you very much for your attention. We hope you enjoyed this UroToday Journal Club discussion.