Evaluating a Sequential Screening Approach for Prostate Cancer Using PSA, Kallikrein Panel, and MRI in the ProScreen Trial, Journal Club - Rashid Sayyid & Zachary Klaassen

May 7, 2024

Rashid Sayyid and Zach Klaassen explore the findings of the ProScreen trial, a rigorous examination into sequential prostate cancer screening using PSA, a kallikrein panel, and MRI. Highlighting the study's aim to refine screening methods by enhancing detection rates of high-grade cancers while reducing unnecessary diagnoses of low-grade conditions, they delve into the process and preliminary results of this innovative trial, which was featured in JAMA. The discussion illuminates how the ProScreen trial, through its methodical use of advanced screening technologies, seeks to balance early detection with minimizing the risks of over-diagnosis, potentially setting new standards for prostate cancer screening practices. The outcomes of this trial could influence future screening protocols and guidelines, particularly in light of the ongoing debate regarding the efficacy and utility of routine PSA screenings as influenced by historical data and past trials.


Rashid Sayyid, MD, MSc, Urologic Oncology Fellow, Division of Urology, University of Toronto, Toronto, ON

Zachary Klaassen, MD, MSc, Urologic Oncologist, Assistant Professor Surgery/Urology at the Medical College of Georgia at Augusta University, Well Star MCG, Georgia Cancer Center, Augusta, GA

Read the Full Video Transcript

Rashid Sayyid: Hello, everyone, and thank you for joining us in this UroToday Journal Club recording. I'm Rashid Sayyid, a urologic oncology fellow at the University of Toronto, and I'm joined by Zach Klaassen today, associate professor and program director at Wellstar MCG Health, where we'll be discussing the recently published ProScreen randomized trial that looked at prostate cancer screening with a combination sequentially of PSA, a kallikrein panel, and MRI. This study was recently published in JAMA with Dr. Anssi Auvinen as the first author.

So just by way of a background on PSA screening, as most of our listeners will be aware, there's significant controversy regarding routine PSA screening for prostate cancer. And between 2008 and 2016, actually, the USPSTF or the United States Preventive Services Task Force recommended against screening and gave this a Grade D. Whereas in 2016 we saw this altered down to a Grade C, whereby the USPSTF recommends discussing PSA screening in men aged 55-69 years only. But this is hardly a glowing endorsement. And really, the recommendations or the evidence to inform the recommendations are based on the results of two large RCTs of organized versus opportunistic PSA screening. And one of them was in the US, the PLCO trial, and the other was the European ERSPC.

And just to refresh your mind, the PLCO trial, and PLCO stands for prostate, lung, colorectal, and ovarian cancer, this trial randomized about 76,000 men between the ages of 55 and 74 from 10 US centers to receive an annual PSA test for six years plus a rectal exam for four years versus opportunistic screening. So this isn't a screen versus no screen trial, it's organized versus opportunistic screening trial. And long-term follow-up at the mean of 15 years showed that there was no significant difference in prostate cancer mortality with the organized PSA screening with a relative risk of 1.04. And so really this was essentially a negative trial against the use of organized PSA screening.

But then conversely, we saw the ERSPC or the European study, and there are really important differences between the two trials that are beyond the scope of this discussion. But essentially, this was a screening trial across eight European countries in the early 1990s of almost 200,000 men between 50 and 74 years, and conversely, here we see that organized PSA screening was associated with a 20% reduction in the prostate cancer mortality rate with a relative risk here of 0.8. And as we see in the Kaplan-Meier curves, we see that the curves diverge at about 3-4 years and then continue to diverge onwards, which really highlights some of the differences between these two trials.

And so beyond the issue of benefit versus no benefit, one of the major criticisms of PSA-based screening is that it often leads to over-diagnosis of low-grade lesions. And so this leads to unnecessary procedures, adverse effects, including anxiety along with other side effects as well. And so the question is, can we incorporate additional tools, novel tools, into such screening approaches in order to maintain the specificity for detecting high-grade lesions while minimizing the detection of these low-grade benign lesions for the most part? And so there are really two agents here that we can consider. So there's the 4K score or the kallikrein-4 panel, which includes total PSA, free PSA, intact PSA, and human kallikrein-2. And it's been shown in numerous studies to reduce the number of men referred for biopsy while retaining the sensitivity for high-grade cancer. Another tool is MRI, as our listeners will be aware of. And there are many trials in this setting whereby the Precision trial, among many others, has shown that this modality can improve the detection of clinically significant prostate cancer while lowering the incidence of clinically insignificant cancer.

So it becomes quite attractive to incorporate these two different modalities in PSA-based screening approaches. And so this really was the rationale for the ProScreen trial, which was designed specifically to reduce unnecessary prostate cancer diagnoses while theoretically or hopefully, maintaining the reduction in prostate cancer mortality that we saw in the ERSPC trial. And so this really adopted a three-phase or a sequential screening algorithm with patients undergoing a PSA and then the kallikrein panel score and then followed by that, the MRI with biopsies and, namely, we'll talk about this later, targeted biopsies reserved for men having a positive result from all three tests. Although patients with a negative MRI and a PSA density over .15 were also recommended for a systematic biopsy. Again, we'll discuss that later.

And so this isn't the final report of this trial. This is the preliminary exploratory report, where essentially, they're comparing the rates of low-grade, which we don't want, and high-grade prostate cancer, which we want to find, between patients in the intervention arm and the control group. So the intervention group is the group invited for screening. Now whether they accept the screening or not is another matter, but they were invited to undergo the PSA screening versus the control group, which was not. And then also they compared the rates between those who actually underwent the screening versus those in the control group. So there's two different layers here at play.

This was a randomized population-based pragmatic screening trial that was conducted or patients were invited actually, between 2018 and 2020. It included men between 50-63 years. And the reason 63 was chosen as the upper cut-off because they wanted to allow for at least two screening opportunities prior to age 71, and we'll see that this can be between every two to four to six years, and so they felt that 63 was a fair upper cut-off to apply here. Patients underwent 1:3 randomization meaning one to screening and three to the control to undergo the screening with the PSA, then 4K panel, then the MRI versus the control group whereby there was no screening invitation. And it's important to be aware that these patients can undergo rescreening. And the rescreening interval was risk stratified based on the results of the first PSA test results. If patients had a PSA greater or equal to three, they were rescreened up to two years between 1.5 and three after four years, and then if less than 1.5, then six years. It's important to note that men who declined the initial screening invitation were or are planned at least to be re-invited to subsequent screening rounds.

The primary outcome was prostate cancer mortality rates at 10 and 15 years of follow-up. And without going into too much detail, the sample size calculation mandated 111,000 men to provide almost 90% power to detect a 25% of relative risk reduction in 15 years. We saw this was 20% in the ERSPC trial. There are obviously differences and things have changed over the last 30 years, but again, quite an ambitious target here of 25% of relative risk reduction. In terms of secondary outcomes, these will include the incidence of high-grade defined as Grade Group 2-5 disease and low-grade prostate cancer as Grade Group 1. And these were ascertained through the Finnish Cancer Registry and the hospital pathology databases with quite good historical data in terms of completion of capturing these diagnoses. They also looked at the sensitivity of screening for high-grade prostate cancer, the incidence of advanced disease with the screening approaches, quality of life, and cost-effectiveness of this approach.

Going into further detail of the study participants, we said men aged 50-63 years, and they included patients from two cities in Finland, Helsinki and Tampere, who were residing in those two cities as of 2018 through the Population Data Services Agency, and the men with prior prostate cancer diagnoses were excluded. This is really a fun epidemiologic exercise in how you conduct a large-scale or population-based pragmatic trial. And so there are some nuances to this that differ from your classic clinical trials and other studies that we are more familiar with. And so, patients here were randomized prior to having informed consent obtained. And so, patients weren't even aware that this was happening. There's really no masking to where patients were randomized to.

Once patients were chosen to be in the screening invitation group, they were mailed an information package that described the aims and procedures of the trial, and a consent form was sent by the trial coordinator. And then, men who were okay with being included in the study provided written informed consent. And they also provided informed prior PSA measurements, prostate biopsies, and family history of prostate cancer. For the most part, men in the control group were not contacted as opposed to those in the screening invitation, although 13% were eventually contacted to evaluate healthcare costs and quality of life data for the purposes of the research study.

Now, how does the screening group intervention happen? So men who are randomized within the intervention arm accept to be included in this study had 30 CC's of blood drawn. First of all, PSA was measured in the local lab. If it was greater or equal to three, the plasma was shipped to Malmö, Sweden, for measurement of the 4K score. And the 4K score is calculated based on the patient's age and the 4-kallikrein marker levels. And if the result is greater or equal to 7.5%, this can range from 0-100, this was considered abnormally high, and patients were referred to the urology department, either the Helsinki or Tampere University Hospitals.

Once that happens, they were selected to undergo a 1.5 or 3 Tesla MRI. When we say multiparametric MRI, this includes all three phases because we know that there's a lot of work being done now to evaluate the biphasic MRIs in order to reduce costs. But in this setting, a PIRADS three, four, or five was considered suspicious for high-grade prostate cancer, and patients were referred for a transrectal fusion targeted biopsy with 2-4 cores obtained from each lesion. For patients who had a PIRADS 3 lesion above and a PSA density less than or equal to 0.15, no biopsy was performed. But if they did not have the PIRADS 3 lesion, but the PSA density was greater than 0.15, then they were recommended for a systematic biopsy. And many people have argued that really, is there a lot of value in doing this? And are we decreasing the specificity setting? Nonetheless, the authors felt that this was warranted, at least a biopsy in the setting, even if the MRI did not show a PIRADS 3 or higher lesion.

And at this point, I'll turn it over to Zach, who will go over the results and discussion for this ProScreen trial.

Zach Klaassen: Thanks so much for that great introduction, Rashid, and for really laying the groundwork for these initial results for the ProScreen trial. So, this is the flow diagram for the trial, and I'll spend a little bit of time going through this because I think there are some important points to be made here. So there were 61,193 men that were randomized. 448 were ultimately excluded. And subsequently, 15,299 patients were randomized to be invited to be screened, which was the intervention group. And 45,544 were randomized to no screening invitation and this was the control group. And so, if we work from the right side of the screen, of those 45,000 randomized to the control group, there were 65 low-grade cancers, 282 high-grade cancers. And then when we look at the 15,299, 7,744 were invited and screened as randomized. However, there were 7,457 that refused participation. So, in the intention of treated analyses, we'll see that this was all of these patients that were invited. Of course, patients can be invited and not want to participate. This is roughly 50%. Of these refused participation, we saw that there were seven low-grade, 44 high-grade cancers.

Now when we look at the patients that were randomized to invitation for screening and accepted the invitation, there were 7,519 that had negative prostate cancer results. This is a PSA less than three or a kallikrein risk score less than 7.5% or a PIRADS less than three. There were 53 patients out of these patients that had a PSA density greater than 0.15 that ultimately had a systematic biopsy. And there were 10 low-grade and 15 high-grade cancers found. Now among the patients that had an invitation to screening, accepted screening, and then subsequently had a positive prostate cancer result, this could be defined as greater than or equal to three PSA, greater than or equal to 7.5% kallikrein risk, and a greater than or equal to three PIRADS. So, they had to have all three, 225 of these patients. Ultimately, 209 were biopsied. We found 22 low-grade and 113 high-grade cancers.

So, we get this huge flow diagram of 15,000 patients invited to screening. We see that ultimately, 209 were biopsied in the positive prostate cancer group, 53 with a PSA density greater than or equal to 0.15. So, of these patients, the biopsy rate was 3.4%. Now looking at the characteristics of these patients, as Rashid laid out, this was 50-63 years of age. So, we look at the quintiles here based on age 50-54, 55-59, and 60-64. Among the controls, all the invited men, and the invited men who underwent screening, roughly 1/3 in each of these quintiles. So, well balanced based on age. Roughly 3/4 of the patients were from Helsinki, the rest were from Tampere. And when we look at the 6,513 patients that had a previous PSA, this was 44% of the patients. 13% had a PSA within 12 months of the trial, 13% within 1-2 years of the trial, 12% within 3-5 years of the trial, and 6% had a PSA greater than five years prior to entering the trial. 3% of patients had a prior biopsy, and 14% of patients had a prostate cancer first-degree relative.

So, this is a summary of the findings per 1,000 men. So, they broke this down per 1,000 men and broke it down by men invited for screening as well as men screened. So, if we look at the top part of this panel, the men invited for screening, of the 1,000 men, 509 participated, 49 had a PSA greater than or equal to three, 35 had a kallikrein panel score greater than or equal to 7.5%, 14 had a PIRADS greater than or equal to three. Three had a PSA density greater than 0.15. Ultimately, 17 out of 1,000 men were biopsied, finding eight high-grade prostate cancers and two low-grade prostate cancers.

When we look at the screened men, of 1,000 men that were screened, 97 had a PSA greater than or equal to three, 68 had a kallikrein panel score greater than or equal to 7.5%, 26 had a PIRADS greater than or equal to three, eight had a PSA density greater than 0.15, 34 were biopsied. So you can see about double the amount of biopsies for men that were screened compared to those that were invited for screening and finding more high-grade cancers at 17 per 1,000 men screened, and four low-grade prostate cancers.

This is a little bit more detail looking at the prostate cancers that were detected at screening. And I'm focusing on the grade group 2-5 because these are the more clinically significant prostate cancers. And we see some trends here, obviously a little bit more Grade Group 2-5 with increasing age. We see roughly no trend in terms of the men that had prior PSAs. In terms of prior biopsies, 5.1% versus 1.6% among the men with no prior biopsies. Family history had a slight signal here at 2.5% versus no family history of 1.6%. As you'd expect, for increasing PSA, more higher Grade Group prostate cancer up to 48.4% for PSA greater than 10. As expected, a higher kallikrein panel risk score, higher incidence of Grade Group 2-5 prostate cancer, a kallikrein panel risk score greater than or equal to 25, associated with 46.2% of the Grade 2-5 prostate cancers. And again, as expected, higher PIRADS score, higher Grade Group 2-5 incidence.

This looks at the prostate cancers diagnosed by trial group, screening participation, and grade group. The first thing we can look at here is that among the control group, 0.78% prostate cancer compared to all invited men, 1.44%, and for invited men who underwent screening, 2.08%. And we can see that the median follow-up of the control group here was 3.2 years. What's also interesting is the risk difference between all invited men versus the control group showed a slight increase in low-grade prostate cancer detection, 0.11%, and also in the high-grade Grade Group 2-5, 0.51%. And this was statistically significant. The last point on this slide is looking at the detailed classification by group. We do not see any difference between Grade Group 4 and 4 for all invited men versus control, but we do see an increase in Grade Group 1, Grade Group 2, and Grade Group 3, as you can see listed on the far right of this table.

Now the authors also provided recently at the European Association of Urology meeting more of a figure view of these similar results. So, we have the table view on the last slide and the figure view on this slide. Similar obviously data, but just a different way of presenting it. We can see the screened men here, 32 Grade Group 1 cancers and 128 Grade Group 2-5. When we combined the entire screening arm, those that were invited and those that participated, obviously slightly more. And we see the control group here. And then at the bottom, we see the detailed breakdown by Grade Group, 1, 2, 3, and then 4-5. What happens when these men are diagnosed with prostate cancer? They broke this down for treatment. The entire screening group is listed here and the control group on the right. Slightly more active surveillance utilized for the screening group versus the control group. Slightly more prostatectomies performed for the control group versus the screening group, 33% versus 28%. Roughly similar utilization rates of radiotherapy as well as androgen deprivation therapy.

So when we're looking at prostate cancer screening, the goal is increasing detection of high-grade and decreasing the detection of low-grade cancer. And so if we look at some of the key screening trials that have been published over the last several decades, we see that the high-grade cancer detection rate for ProScreen was 1.7%, which is comparable to the ERSPC trial that Rashid discussed at 1.8%, and slightly more in the UK Cluster Randomized Trial at 2.7%, and quite low detection rates of high-grade cancer in PLCO at 0.5%. When we look at low-grade cancer detection, ProScreen had 0.4% compared to a much higher low-grade cancer detection in ERSPC at 3.2%. And this may be secondary to ProScreen having a younger patient population compared to ERSPC.

Now there are two key recently published RCTs looking at MRI-based screening pathways. And so the first was out of Sweden by Huggoson and colleagues published in the New England Journal of Medicine in 2022. This is the Göteborg2 screening trial, 11,986 men aged 50-60 years, so similar to the ProScreen age group. They were randomized to an experimental arm, which was a PSA followed by MRI plus targeted biopsies if the PSA was greater than or equal to three, and a control group of PSA followed by MRI plus systematic plus targeted biopsies if the PSA was greater than three. We see very similar high-grade detection rates between these two arms at 0.9% and 1.1% compared to ProScreen, which is 1.7%. And we see low-grade cancer detection rates, less low-grade detected in the experimental arm at 0.6% versus 1.2% in the control, which is similar to ProScreen at 0.4%. Additionally, we see a similar proportion of men undergoing biopsies in the Göteborg2 trial at 2.8% versus 2.7% in ProScreen.

The second trial was also out of Sweden by Eklund and colleagues in 2021 published in the New England Journal of Medicine. This is the STHLM3-MRI study. Smaller study, 1,532 men, different age group, 50-74 years of age randomized to the experimental arm of an MRI plus systematic plus targeted biopsies if the PIRADS is greater than three in men that had a PSA greater than three and the control arm of systematic biopsy for PSA greater than three. We see clinically significant prostate cancer in 21% of the experimental arm versus 18% in the control arms, so quite similar, but we do see less clinically insignificant cancer at 4% in the experimental arm compared to 12% in the control arm. What's interesting here is these two trials compare different approaches in a screened cohort, so there's no control arm. And so it's somewhat difficult to compare this approach screen, but these two randomized trials do provide information about the effects of adding an MRI to a PSA-based screening regimen.

As with any big population-based study, there are some limitations and the authors do an excellent job of laying these out. The first is that the absolute difference between the two randomized groups in ProScreen was small, thus the clinical importance does remain somewhat unclear. Secondly, prior screening reported by a substantial number of participants may have reduced cancer detection compared with the population not undergoing screening. However, they did do a sensitivity analysis and they found that there were no differences between those with and without previous PSA tests. Third, the results are based on just a single invitation for screening, and there's a high likelihood that some of the high-grade cancers were missed. And finally, there's no data on cancers missed at screening, and the interval cancer incidence is needed to assess the sensitivity of the screening protocol and that will be assessed in additional follow-up.

So in conclusion, in the preliminary report of the ProScreen trial, compared with a control group not offered screening, a single prostate cancer screening intervention that included a PSA test, a Kallikrein panel for patients with a PSA greater than three, and an MRI, the screening intervention detected one additional high-grade cancer per 196 men, as well as one additional low-grade cancer per 909 men. And indeed, these preliminary findings should be interpreted provisionally pending the results of the primary mortality outcome, which will be published in subsequent years. Thank you very much for your attention. We hope you enjoyed this Journal Club discussion for UroToday discussing the ProScreen trial recently published in JAMA.