Read the Full Video Transcript

Rashid Sayyid: Hello everyone. This is Rashid Sayyid. I'm a first year Urologic Oncology fellow at the University of Toronto. And along with Zachary Klaasen, assistant professor and program director of the Medical College of Georgia, we'll be presenting the Journal Club for the most recent publication of the ProtecT trial that looked at 15 year outcomes after active monitoring, surgery or radiotherapy for prostate cancer. This report was recently published in the New England Journal of Medicine on March the 11th and was led by Dr. Freddie Hamdy. We know that despite advances in prostate cancer with regards to diagnosis, risk stratification, or treatment, significant controversies with regards to over and undertreatment remain. The US Preventive Services Task Force recommendations against PSA screening in 2012 have been associated with an increased incidence of regional and advanced disease. However, in the short to medium term, at least, prostate cancer mortality rates have remained unchanged.

The ProtecT trial or the prostate testing for cancer and treatment, which was funded by the NIH and the University of Oxford, invited over 82,000 men between the ages of 50 and 69 years who had a life expectancy of 10 plus years. No other malignancies except skin cancer and were eligible for treatment. These men received invitations from 337 primary care centers between 1999 and 2009, and these men were invited to undergo a PSA test. And if the PSA test was between 3 and 19.9, so essentially less than 20 patients were invited to undergo a subsequent rectal exam and a 10-core TRUS-guided biopsy at one of nine secondary cancer centers. Now, if the biopsy was benign in these men they were invited to have a repeat biopsy. If their free to total ratio was less than 11%, they had evidence of ASAP or high grade PIN on their biopsy, and if the repeat biopsy was still benign, then they were excluded from the trial and managed by their primary care physician.

Eventually, prostate cancer was diagnosed in 2,664 men, and of these men just over 1,600 were randomized in either active monitoring, prostatectomy or radiotherapy with three to six months of neoadjuvant ADT. Without going to too much detail on this flow chart, which has previously been presented in prior reports, we see that there was essentially an equal split between the three arms, 545 in the active monitoring arm, 553 in prostatectomy and 545 to radiation neoadjuvant ADT. And what I want to draw your attention to here is that these men were analyzed to the groups which they were assigned. So irrespective of what they got in the interim, as part of the Intention-To-Treat Analysis, every patient who was randomized to a group was analyzed as part of that group at the end of the trial.

In terms of the characteristics of the patients, the median age was 62 years ranging between 50 and 69. The median PSA was 4.6, ranging between three and 18.9, just over three quarters had Gleason 6 disease and had evidence of clinical stage T1c disease as well. The 10-year outcomes of ProtecT were previously published by Hamdy et al in the New England Journal of Medicine 2016. And if I draw your attention to the right side, although we see that the active monitoring arm had an increased hazard of disease progression, this did not translate into any clinically or statistically meaningful differences in prostate cancer specific survival between the three arms. So the objectives and outcomes of this specific report were to provide updated 15-year outcomes by treatment arm for prostate cancer specific mortality, which was the primary outcome. And this was adjudicated by an independent data cause of death committee.

They also looked at all cause mortality, looked at metastases, which was defined as either evidence of medicine imaging or a PSA level greater than hundred. They also looked at the composite outcome of clinical disease progression, which was defined as metastases, having clinical stage T3, T4, initiation of long-term ADT development of ureteral obstruction, rectal fistula or requirement for urinary catheterization, secondary tumor growth. Another outcome of interest was initiation of long-term ADT. As for the treatments, all patients follow standardized trial specific clinical management pathways. PSA levels were checked every three months for the first year, and then from the second year onwards they had it checked every six to 12 months. Patients underwent clinical evaluation annually. Now specifically starting with the active monitoring group, if patients had a PSA increase of at least 50% or the patient and or the clinician had any concern for disease progression, this triggered a review.

With the result being one of two, either patients were continued on monitoring or they proceeded to undergo further testing with subsequent radical or palliative treatments. As for the prostatectomy group, specifically with regards to the topic of adjuvant salvage radiotherapy, this was discussed that patients had positive surgical margins, had evidence for extracapsular disease or post-op PSA level greater than 0.2. Of note, the study period was early on. Patients were recruited between 1999 to 2009, and this preceded the latest data looking at the role of adjuvant versus early salvage radiotherapy. Now, as for the radiotherapy group, all patients received neoadjuvant ADT for three to six months. And as for the technique, it was 3D conformal radiotherapy with 74 gray in 37 fractions. Management review was triggered in this group if the PSA increased by two or there was any concern about disease progression. In all three groups the active monitoring, radiation or prostatectomy, all patients underwent a bone scintigraphy, a bone scan if the PSA was greater than or equal to 10, and ADT was discussed for all patients if the PSA was at least 20 nanograms per ml.

As mentioned previously, the intent to treat principle was followed and patients were followed until they were censored or lost to follow up or death. Looking at the outcome of prostate cancer, specific mortality and other secondary outcomes between groups comparisons were performed using multivariable Cox models, adjusting for the trial center, the age, the Gleason score, and the log transformed baseline PSA. Pairwise comparisons, meaning, for example, the prostatectomy versus radiation or radiation versus active monitoring were planned only if the p-value for comparisons overall was less than 5%, and if no correction for multiplicity was not performed, and this has the consequence of an increased risk of a type one or false positive error, meaning we're more likely to find a significant difference even if it's not present. And as such, the authors accounted for that and only reported point estimates with 95% confidence intervals provided and no p-values included.

And then for subgroup analyses, we can see that prostate cancer mortality was assessed looking at the following strata, looking at age less than 65 or older. We looked at, they looked at grade group disease 1 versus 2 versus 3, the PSA level less than 10 or 10 to 20, stage T1 or T2. They also looked at the aggregate tumor length in the biopsies less than or greater equal to four. They also looked at the maximum tumor length in a single biopsy less than or greater or equal to two. And then given previous criticisms suggesting that this was essentially a low to potentially intermediate risk cohort, they performed risk stratification by the D'Amico score, the CAPRA, and then the Cambridge Prognostic Group criteria as well. At this point, I'll turn it over to Zach to discuss the results and the discussion.

Zachary Klaasen: Rashid, thanks so much for that great introduction. So hopping right into the primary and secondary outcomes of this long-term analysis of ProtecT. At the top of the table here you can see that the primary outcome of death from prostate cancer with active monitoring as the reference group, there was no statistically significant difference for those men that underwent prostatectomy or radiotherapy compared to active monitoring. And again, when we look at death from any cause, which was one of the key secondary outcomes, with active monitoring acting as the reference group, again, no difference between these two treatment modalities and active monitoring for death from other causes. However, as I've highlighted in this box at the bottom of the table, we do see that for metastatic disease, ADT therapy and clinical progression, any treatment, whether it be prostatectomy or radiotherapy, did lead to a clinically significant reduction in these outcomes compared to active monitoring. Looking at prostate cancer survival with prostatectomy in red, radiotherapy in blue, and active monitoring in gray, as you can see in this cumulative incidence figure, essentially completely overlapping curves suggesting no difference.

And when we look at the 10 and 15 year prostate cancer survival for active monitoring, prostatectomy and radiotherapy, essentially excellent outcomes for all of these groups, not just at 10 years, but also at 15 years at more than 96% in all groups at 15 years. What's important to note is if we look at underlying causes of death, only 45 patients that have prostate cancer, but we see the chronicity of prostate cancer in these patients, given that 164 men died of other cancers and 101 men died of cardiovascular disease. So certainly in the long term follow up, some of these men will die of prostate cancer. But looking at the highlighted box here with other cancers in cardiovascular, a significant number of men will die of other causes. Looking at some of the other secondary endpoints in terms of the cumulative incidence curves, this is metastasis free survival.

Again, the same color scheme as the previous figure I showed. We do see a difference between active monitoring in gray compared to radiotherapy and prostatectomy. And when we look at the ADT free survival and the progression free survival, active monitoring now in green, prostatectomy in red and radiotherapy in blue for ADT free survival on the left, again, we see a difference between active monitoring and the overlapping prostatectomy and radiotherapy curves. Essentially, again, in progression free survival we see worse outcomes for active monitoring compared to prostatectomy and radiotherapy. But as previously mentioned, both in this analysis and in the previous analysis that Rashid highlighted, this did not result in differences in prostate cancer survival. This curve looks at the probability of undergoing radical treatment during the follow-up period.

You can see here radiotherapy in blue, prostatectomy in red and active monitoring in gray. And as expected, the men that were randomized to radiotherapy and prostatectomy, 92% had radiotherapy, 90% had a radical prostatectomy. And over time, 61.1% of the men in the active monitoring group either had a radical prostatectomy or radiotherapy. What's important, and one of the take home messages from these results is that by the end of follow up, 24.4% of men in the active monitoring group were alive and had not received radical treatment nor ADT. This is the subgroup analyses looking for prostate cancer specific survival. As Rashid laid out, there were several groups in terms of age, Gleason grade group, tumor length and biopsies, PSA level clinical stage, and CAPRA and D'Amico risk score. And what's interesting is when we compare prostatectomy to active monitoring and when we compare radiotherapy to active monitoring, the only subgroup that had statistically significant benefit from treatment was in clinical stage T1c prostate cancer or prostatectomy versus active monitoring had a significant survival benefit.

So essentially the majority, if not all, except for this one group, had no difference in subgroup analyses when comparing prostatectomy to active monitoring and radiotherapy to active monitoring. So there's several important discussion points from this extended analysis of the ProtecT trial. This provides evidence of a high percentage of long-term survival, 97% from prostate cancer specific death, and 78% from death of any cause regardless of treatment group. Certainly as we discussed, radical prostatectomy and radiotherapy reduced the incidence of metastases, local progression and long-term ADT used by half compared to the active monitoring group. But these did not result in reductions in mortality at 15 years of follow-up. So even though the incidence of metastases increased in these men, the number of prostate cancer deaths remain low on intervals between metastases and death, continue to extend from 10 to almost 20 years in some cases, which does call into question whether metastasis can be used as a surrogate for lethality of prostate cancer in men that present with localized disease.

So in conclusion at a median follow-up of 15 years ProtecT found that mortality from PSA detected prostate cancer remained very low regardless of whether a man receive active monitoring, radical prostatectomy or radiotherapy. Thus, the choice of therapy involves weighing the trade-offs between benefits and harms associated with treatments for localized prostate cancer. We thank you very much for your attention. We hope you enjoyed this UroToday Journal Club discussion of the recently published ProtecT 15 year follow-up in the New England Journal of Medicine.