Imaging-Based Prostate Cancer Screening Among BRCA Mutation Carriers - A UroToday Journal Club
January 20, 2021
Christopher J.D. Wallis, MD, Ph.D., Instructor in Urology, Vanderbilt University Medical Center, Nashville, Tennessee
Zachary Klaassen, MD, MSc, Urologic Oncologist, Assistant Professor Surgery/Urology at the Medical College of Georgia at Augusta University, Georgia Cancer Center
Full Text Article: Imaging-based prostate cancer screening among BRCA mutation carriers—results from the first round of screening
Christopher Wallis: Hello. Thank you for joining us for this UroToday Journal Club. Today, we're discussing a recently published paper, entitled "Imaging-based prostate cancer screening among BRCA mutation carriers, results from the first round of screening." I'm Chris Wallis, Fellow in Urologic Oncology at Vanderbilt University Medical Center. With me today is Zachary Klaassen, assistant professor at the Medical College of Georgia. Here is the citation for the paper we are discussing with the senior author, David Margel.
By way of background, most people are aware that BRCA mutations, particularly BRCA2, increase both the incidence and severity of prostate cancer. Recently published data from the PROREPAIR-A cohort provides some perspective data on this. This is data that was presented at ESMO this year. So highlighting the study design here we have a matched prospective cohort study with cases being germline BRCA2 carriers, and controls being noncarriers which were matched in a one to two fashion on the basis of a histologic grade and stage. When we look at this, we can see that looking at cancer-specific mortality, this is much higher among BRCA2 carriers than noncarriers with a difference of more than eight years and median cancer-specific survival. So this speaks to the increase in the severity of prostate cancer among these men.
So the question is if we know that these men are at a higher risk of prostate cancer and a worse prostate cancer phenotype, can we intervene? And, so this is the impact study that looked at targeted prostate cancer screening among men with BRCA mutations. This study schema is highlighted here, how they examine men with PSA's and thresholds for biopsy as well as ongoing surveillance approaches. And, in summary, the authors concluded that BRCA2 is associated with increased prostate cancer incidence, younger age of diagnosis, and a higher rate of clinically significant disease. learning what we discussed already.
So current guidelines recommend PSA based screening for men with BRCA2 mutations. However, when we look at women with BRCA mutations, imaging-based approaches, particularly MRI are routinely used for screening in this context for breast cancer screening. So the authors of the paper we are discussing today sought to examine the role of prostate cancer screening using both MRI and PSA among men with known BRCA2 mutations.
So they enrolled men aged 40 to 70 years of age who had known BRCA1 or BRCA2 mutations in Israel. The patients to be eligible for inclusion had to have no prior history of prostate cancer or prostate biopsy, and a good performance status. And this is how they proceeded through this study. These eligible men were invited to participate, and those who accepted following the informed consent underwent PSA testing and an MRI with a prostate biopsy performed as indicated. And then for those who had prostate cancer diagnosed, they received a treatment as suggested by their treating physician. And for those who did not have cancer diagnosed on this initial screening round, they had ongoing observation with annual visits for five years.
So these are the testing procedures undertaken in the trial. So PSA testing was utilized with age-based thresholds. As you can see here, MRI was performed with a 3T MRI with an external coil, and multi-parametric images were obtained, including T2 weighted images, diffusion-weighted images, and dynamic contrast-enhanced images. All studies were read by two uro-radiologists and interpreted using the PI-RADS, a version two criteria and MRIs with PI-RADS 3 or greater lesion were considered abnormal.
So prostate biopsy was offered to all men with abnormal screening results, either on the basis of their PSA levels or their multi-parametric MRI. So 12 core systematic biopsies were performed for men with abnormal PSA's, whereas a 12 core systematic biopsy and targeted biopsy were performed among those with abnormal MRIs.
The authors sought to assess the primary outcome of the prevalence of prostate cancer among male BRCA mutation carriers in Israel. And so they calculated that the baseline prevalence of prostate cancer in the Israeli male population was 0.65%. And the impact study suggested a prevalence of 2.8% among BRCA carriers. Thus, they sought to assess whether the prostate cancer rate among Israeli BRCA carriers was at least 2.8% using standard alpha 0.05 and beta 0.2 requiring a sample size of 190 men. The secondary outcomes were to evaluate various screening strategies, utilizing MRI and PSA based thresholds. So now I will pass it over to Dr. Klaassen to discuss the results and interpretation of the study.
Zachary Klaassen: Thanks, Chris. So you can see here, this is the Concort diagram for the study. They assessed 233 people for eligibility and ultimately recruited 188 patients for the study. 108 of these patients had BRCA1 mutations and 80 had BRCA2. You can see here that in the second to last area here, 67 patients had a normal MRI, normal PSA, 35 had an abnormal MRI with a normal PSA, 43 had a normal MRI with an elevated PSA and 32 patients had an abnormal MRI with elevated PSA. So ultimately 92 patients underwent prostate biopsy, 16 of these patients of that overall patients refused a biopsy. So the sample size for prostate biopsy was 92 patients.
This is a table looking at the baseline characteristics. You can see here, the variables on the left on the second column is the total overall and then broken down by BRCA1 and BRCA2, P-value comparing BRCA1 and BRCA2 is on the far right. You can see here, the only difference between the BRCA1 and BRCA2 patients was that the BRCA1 patients were a mean roughly three years older than the BRCA2 patients. If we look down at the bottom of the table, some interesting findings in terms of patients with family history, basically almost 100% of these patients had a family member that had previous cancer. Most commonly, not surprisingly was breast cancer at 80% overall and then, second was ovarian cancer. In terms of a personal family history of cancer, you can see that almost 13% of these patients had a personal family history of cancer. Most commonly was melanoma and non-melanoma skin cancer. So as per what we'd expect with a BRCA mutation, personal and both a family history of cancer for these patients.
This is a study looking at the initial PSA results stratified by age. You can see here at age 40-50, 50-60 and 60-70. Not surprisingly, the PSA's were a little higher as men got older. PSA densities here 0.03 for 40-50-year-olds, 0.03 for 50-60-year-olds, and 0.04 for 60-70-year-olds. Also, not surprisingly, the median prostate volume increased with age as well, from 23 MLs at 40-50 years of age to 45 MLs.
Also you can see here, the percent of patients with age-specific elevated PSA, 38% for 40-50-year-olds, 31% for 50-60-year-olds, and 47% for 60-70-year-olds.
In this table here, we have a summary of the multi-parametric MRI findings stratified by age. Similar to the last table, in this first row here, looking at multi-parametric MRI, normal MRI 69% in 40-50-year-olds, 44% in 50-60-year-olds, and 55% in 60-70-year-olds.
Then we get into the PI-RADS scoring. Not surprisingly, the most common PI-RADS was three. 14% in the younger patients, 24% in 50-60-year-old men, and 20% in 60-70-year-old men. And then we see several PI-RADS 4, 22% in the 50-60-year-olds and 14% in the 60-70-year-olds. And several PI-RADS 5, none of which were in the younger patients, 7% in the 50-60-year-olds, and 5% in the 60-70-year-olds.
In terms of the biopsy findings, also stratified by age, you can see here, the Gleason Groups in this column here, 1 through 5. Looking quickly at the totals, we can see here, I'll draw your attention to the bottom right, where 8.5% of these patients had prostate cancer. And the majority of these were Gleason grade group 1 and Gleason grade group 2. You can see here that there was one Gleason grade group 5 in the 50-60-year-olds. And then most of these, like I said were Gleason grade groups 1 and 2 overall.
This is the busiest table but it's basically a breakdown of the 16 patients that they found prostate cancer on. And you can see here that even amongst a screening cohort, this is just men with BRCA mutations, several intraductal, three patients with an intraductal or cribriform pattern. You can see here the PI-RADS correlation to these, PI-RADS 5 for this patient, PI-RADS 5 for this other patient with intraductal, PI-RADS 4 for patient number 16. The majority of these were actually diagnosed overall by a fusion MRI biopsy and not by PSA. 69% of these patients were diagnosed with a fusion of the multi-parametric MRI alone.
So I'll take a little time explaining this table. These are the several strategies that the authors came up with for detecting prostate cancer in BRCA carriers. So the first strategy, the first two are PSA only screening. Strategy 1, PSA greater than three. Strategy 2, elevated age-stratified PSA. The third strategy is basically just an MRI only screening strategy with anybody over or equal to a PI-RADS 3. Strategies 4 and 5 are combinations where PSA first, over three followed by as well, a PI-RADS greater than or equal to three. Strategy 5, elevated age-stratified PSA, and then an MRI PI-RADS greater than three or equal to three lesions. So if you go back up to this table here, strategy 1, which is PSA greater than three, you can see that 73% of biopsies would have been avoided. They would have only detected 31% of cancers, 69% would have been missed with an okay negative predictive value of 83.6%, but a positive predictive value of only 20%.
Strategy 2, which is elevated age-stratified PSA, you're avoiding fewer biopsies. You are detecting the same amount of cancers as Strategy 1 and you are missing the same amount. So your entire predictive value is essentially the same as Strategy 1 at 84%, but you get even a worse positive predictive value. So both of these are not great if you combine them together.
And when we look at Strategy 3, which is PI-RADS only, this is an MRI only screening strategy, you're only avoiding 32% of biopsies. You are detecting a lot of cancers at 94% and you're only missing six. So you've got a very good negative predictive value, the positive predictive value is okay at 23.8. So you are going to biopsy most folks, and you're going to detect a lot of cancer. And so the one comment the authors made here is that you're doing the most biopsies using strategy three out of the five,
When we start to combine the PSA and PI-RADS scoring, Strategy 4, a PSA greater than three, first, followed by an MRI, you're avoiding 80% of biopsies, you're detecting 56% of cancers, you're missing 44%. So you are improving on your positive predictive value at 50%, as well as maintaining your negative predictive value at 90.6%.
And finally, Strategy 5, which is the elevated age PSA followed by PI-RADS 3 lesion. You're avoiding 66% of biopsies and detecting 63% of cancers. You are missing 27% of cancers. And so the authors do note that five out of these six patients were Gleason grade group 1 only. So not missing high-grade cancers with a negative predictive value of 90% and a positive predictive value of 31.3%.
So basically, they are trying to figure out what the best strategy is in terms of combining PSA and MRI. These are the results of the decision curve analysis. You can see the net benefit is on the Y-axis and the threshold probability of prostate cancer is on the X-axis. This is for all carriers. So with a threshold probability of zero, you can see here the red line is PI-RADS 3, the MRI is certainly optimal here. And so when you get to the threshold probability of prostate cancer of 0.2, the net benefit is quite similar between the PI-RADS 3 and adding the PSA greater than three before the PI-RADS. So you are sort of getting this conglomeration here, of including MRI but also considering having a PSA beforehand in triage. Now, this flips a little bit when we look at younger men. So this is men under the age of 50. Once again, MRI, PI-RADS greater than or equal to three is in red. That's the top line here.
The dotted green line here is PSA greater than three, and PI-RADS greater than or equal to three. That is this line here. So, if we are looking at a threshold probability of 20%, by far MRI is key to diagnosing prostate cancer in these younger men. Conversely, when we look at men over the age of 55, the same lines here, the red line is not as beneficial in these older men and we see that PSA can first stratify these older men much, much better than just an MRI alone. So the take-home message from this, is that in these younger men you want to get an MRI first, whereas in the older men, more than 55 years of age, you want to get a PSA followed then by an MRI.
So a couple of discussion points from this study; we certainly have evidence that there is a benefit to using an MRI prior to prostate biopsy, data from PRECISION, MRI-FIRST, and, the 4M study. It's interesting to note from the EAU guidelines, that currently they specifically object to using MRI as a screening tool. Now, this is not for high-risk men like the BRCA study we are talking about, but this is just a general recommendation that MRI not be used as a screening tool. There has been one study published by my colleague on the line, Dr. Wallis was part of this study with Dr. Nam of Toronto, where they looked at MRI as a screening tool. It was a pilot study, a feasibility study of 47 men, and interestingly enough they diagnosed prostate cancer in 38.3%.
So I think looking at this data in this high-risk population, certainly, Dr. Margel and colleagues make a very strong argument that particularly in young men, MRI as a screening tool is certainly a viable and reasonable option. So in conclusion, the rate of prostate cancer in the first round of screening of BRCA-carriers was high at 8.5%. Among carriers younger than the age of 55, multi-parametric MRI had the best clinical utility. And among carriers older than 55 years of age, PSA should be used first to triage patients followed by multi-parametric MRI and appropriate patients. Thank you very much for attending this UroToday Journal Club.