Christopher Wallis: Hello, and thank you for joining us for this UroToday journal club discussion. Today we're discussing the MARCAP study, looking at androgen deprivation therapy use and duration with definitive radiotherapy for localized prostate cancer: An individual patient data meta-analysis. I'm Chris Wallis, an assistant professor in the Division of Urology at the University of Toronto. And I'm joined today by Zach Klaassen, assistant professor in the Division of Urology at the Medical College of Georgia. This is the citation for this recent publication in Lancet Oncology led by Dr. Kishan, Dr. Spratt, and the MARCAP consortium group.

By way of background, meta-analyses of randomized control trials may be the most practice-changing evidence we have in medicine. However, despite a large number of randomized trials, this approach has been relatively underutilized in assessing the role of androgen deprivation therapy and prostate cancer. Thus, despite the numerous trials of ADT in conjunction with radiotherapy, there remains substantial heterogeneity in terms of the questions relating to radiotherapy dose, ADT modality, ADT duration, and ADT timing. And as a result, the aggregate benefit of these interventions is poorly quantified. Perhaps as a result of the relatively limited evidence base, real world utilization of ADT overall and of ADT prolongation is lower than we would expect or hope. To address this, the MARCAP consortium was established MARCAP refers to the meta-analysis of randomized trials in cancer of the prostate, which seeks to address these gaps by acting as a data repository for international trials groups to consolidate the results of randomized trial data.

So in this present analysis, using the MARCAP consortium, the authors sought to assess three goals. Quantifying first, the oncologic effects of the addition of ADT to radiotherapy, second, the effects of neoadjuvant ADT prior to radiotherapy, and third, the effects of adjuvant ADT following radiotherapy. To do this, they performed a systematic review and meta-analysis with a pre-specified search strategy that was registered in the PROSPERO online repository. They sought to evaluate randomized control trials that assess the question of ADT use and prolongation in men who are receiving definitive radiotherapy for prostate cancer. The search included a number of databases, including MEDLINE, Embase, the Cochrane Central Register of Clinical Trials, ClinicalTrials.gov, Web of Science, Scopus, as well as major conference abstracts. They assessed these between 1962 and 2020 and excluded studies that did not provide evidence or outcome data on metastasis or survival, that used non-steroidal antiandrogens, that were single center in nature, or that used lifelong androgen deprivation.

Following identification of relevant studies, data were requested from the investigators of these trials, including baseline characteristics, such as age, T stage, tumor grade, PSA, performance status, and allocated treatment. In terms of outcome measures, the authors assessed for both a binary outcome event and as well as a time to event for biochemical failure, local failure, regional failure, distant failure, and death, as well as the cause of death. The individual patient data from each of the included studies was then shared with the co-primary investigators, as well as the study statisticians.

Three different treatment regimes and questions were identified including the use of ADT with radiation compared to radiation alone, the role of neoadjuvant ADT prolongation comparing short versus long durations, and the role of adjuvant ADT prolongation, again comparing short versus long durations. The primary outcome for each of these comparisons was metastasis free survival with secondary outcomes, including biochemical recurrence, distant metastasis, and overall survival.

They constructed a pre-specified statistical plan and utilized the intention-to-treat analytic principle. Inter-trial heterogeneity was assessed in two stages by first comparing effect sizes and standardized errors for each study and then performing aggregated analysis using random effects models. In terms of the individual patient data meta analysis, a single stage approach was used with univariable Cox regression models stratified by trial identifier. For time to outcomes such as metastasis free survival and overall survival, Cox proportional hazard models were used. Whereas for biochemical recurrence, [inaudible] metastasis and prostate cancer, specific mortality competing risks models were used.

The authors performed a number of sensitivity analyses to assess the robustness of their analysis. They used a cause specific additive frailty models with a random trials effect in comparison to the previous analytic approach we discussed. And for each treatment group a 10 year restricted mean survival time and 10 year restricted meantime lost were calculated again to compare between treatment approaches in each of the three questions. ADT use meta-analysis was repeated while excluding the EORTC 22863 trial, given is differing methodology. Finally, the authors performed formal interaction testing on the basis of subgroups defined by age, NCCN risk group, and radiotherapy dose.

At this point in time, I'm going to hand discussion over does Zach to walk us through the results.

Zachary Klaassen: Thanks so much, Chris. So this is the trial profile. Looking at the studies and the systematic review you can see that there was 1,148 screen trials moving down to 59 full text articles that were assessed for eligibility, and ultimately ending up with 12 randomized trials that met eligibility criteria and importantly, being able to receive independent patient level data for all 12 trials for this study.

So the next couple slides will look at the trials included in this meta-analysis. And this first slide looks at the seven trials that were assessed for just ADT use. Yes or no. I've added asterisks to the trials that had a significant benefit in the intervention group versus the control group for metastasis free survival. So for instance, we can see here that the RTOG 8610 trial had a statistically significant benefit for two months of ADT followed by two months of radiotherapy plus ADT versus radiotherapy with an MFS hazard ratio of 0.80 and a 95% confidence interval of 0.65 to 0.98.

Similarly, for the EORTC 22863 trial, we can see there was a metastasis free benefit of 0.50, hazard ratio to 95% confidence interval, 0.38 to 0.65 for radiotherapy plus 36 months of ADT versus radiotherapy alone.

And finally dropping down to the bottom here. We see the TROG 96.01 trial. There was a statistical significant benefit for the hazard ratio of MFS of 0.59 and a 95% confidence interval of 0.46 to 0.75 for five months of ADT followed by one month of ADT plus radiotherapy versus radiotherapy alone.

This slide looks at the included trials for neoadjuvant ADT progression. The only statistically significant benefit was in the TROG 96.01 trial with the MFS hazard ratio of 0.71 and a 95% confidence interval of 0.55 to 0.92 for five months of ADT followed by one month of radiotherapy plus ADT versus the control group of two months of ADT followed by one month of radiotherapy plus ADT.

Finally, this is the table looking at adjuvant ADT prolongation. Several significant trials here. The RTOG 9202 trial with a MFS benefit of hazard ratio, 0.87 95% confidence interval of 0.77 to 0.97 favoring the intervention of two months of ADT followed by radiotherapy and an additional 26 months of ADT versus the control group of two months of ADT followed by two months of radiotherapy plus ADT. Also significant was the EORTC 22961 trial looking at radiotherapy plus 36 months of ADT versus radiotherapy plus six months of ADT with a hazard ratio favoring the intervention group of 0.60 and 95% confidence interval of 0.47 to 0.76. And finally, the DRTGICOR trial had a significant benefit for the intervention group of two months of ADT followed by radiotherapy and an additional 26 months of ADT compared to the control group of two months of ADT followed by two months of radiotherapy plus ADT with the MFS hazard ratio of 0.45 and a 95% confidence interval of 0.24 to 0.83.

This is the baseline characteristics across these trials. And you can see here that... We'll focus just on the total because there's really comparable characteristics across the ADT use, neoadjuvant ADT prolongation, and adjuvant ADT prolongation groups. Overall, there was 10,853 patients with a median follow up of 11.4 years. Looking at the median age was essentially set 70 for all of these groups looking at Gleason score, overall, the most common Gleason score was 7 41.9%. And we see that this is pretty standard across all three of these analyses. In terms of T category, a majority of these patients were T one to two at 68.4% with a median PSA of 12.2 ranging from about 10 to 16.6 in the ADT prolongation group, looking at NCCN risk group; low was 10.1%, intermediate was 43.3%, and high risk by NCCN was 46.6%.

This is the metastasis free survival and overall survival curves for ADT use. You can see MFS on the left and OS on the right, which will be the same lineup for the next couple of slides. And so looking at them, this is the ADT use yes or no. And we can see a significant metastasis free survival benefit of a hazard ratio of 0.83 and a 95% confidence interval of 0.77 to 0.89 favoring ADT versus no ADT. And again, we see this in the overall survival favoring ADT with a hazard ratio of 0.86 and a 95% confidence interval of 0.80 to 0.92.

These are the MFS curves for neoadjuvant ADT extension. And this is a MFS again on the left and OS on the right. And here we do not see a significant benefit for extended ADT versus short ADT. You can see on the MFS, the hazard ratio is 0.95, not statistically significant as well as on the right for overall survival hazard ratio of 0.95, not statistically significant. So no benefit for MFS or OS would Neo adjuvant ADT extension.

Finally looking at adjuvant ADT prolongation, we see a MFS and an OS benefit for long term ADT versus short term ADT. On the left, the MFS hazard ratio of 0.84 and 95% confidence interval of 0.78 to 0.91. And for overall survival on the right, hazard ratio favoring long term ADT of 0.85 and a 95% confidence interval of 0.78 to 0.94. This is the Forest plot showing the effects of ADT use on biochemical recurrence, distant metastasis, metastasis free survival, and overall survival. And this is the ADT yes or no. And we can see here that on the left, favoring ADT, across all of these metrics and endpoints, this is statistically significant for ADT utilization. A very similar forest plot for neoadjuvant ADT extension across these same endpoints. However, here we do not see any benefit for extended neoadjuvant ADT with the hazard ratio, all crossing the midline.

And again, looking at ADT adjuvant prolongation. Again, we see a benefit across these four endpoint favoring long term adjuvant ADT with hazard ratios ranging from 0.56 To 0.85 for overall survival. This is the subgroup effects of interaction test results for metastasis free survival. And you can see this is for broken down again by ADT use, neoadjuvant ADT extension, and adjuvant ADT prolongation. For ADT use looking at a high level of view here we see a benefit across all of these groups, especially for high intermediate NCCN risk for radiotherapy. We do not see a benefit for high dose radiotherapy, and we see no difference based on age with ADT use favoring both patients less or greater than 70 years.

Looking at the neoadjuvant ADT extension, similar to what we discussed as well in the previous slides, no benefit for ADT neoadjuvant extension for NCCN based on NCCN risk groups, as well as age. And again, looking at adjuvant ADT prolongation, we see a benefit for high and intermediate NCCN risk, and we see a benefit for low, but not for high radiotherapy dose. And again, we see a benefit regardless of age.

So several discussion points from this trial. There's a lot of data in this MARCAP consortium trial. This study from the MARCAP consortium represents the first IPD meta analysis of eligible RCTs evaluating the use and prolongation of ADT with radiotherapy. There were several novel clinically relevant insights that were unclear from looking at the individual trials alone. I've listed them here.

First, there was a significant MFS and OS benefit from the addition of ADT to radiotherapy. Secondly, adjuvant ADT prolongation to at least 18 months in conjunction with radiotherapy improved metastasis free and overall survival compared to short term ADT. Third, extension of new adjuvant ADT was not associated with improved biochemical recurrence, distant metastasis, MFS, or OS benefit, and thus should not be routinely recommended for treatment. And finally, the treatment effects of each intensification strategy were not significantly affected by radiotherapy dose, NCCN risk group, or patient age.

So in conclusion, this meta-analysis from the MARCAP consortium is the first systematic IPD meta-analysis of randomized trials evaluating the use and prolongation of ADT and localized prostate cancer. This study provides the strongest level of evidence for the routine recommendation for the addition of ADT to radiotherapy in men with intermediate risk disease and the use and prolonged adjuvant ADT for men with high risk disease, irrespective of radiotherapy dose. As we've mentioned, the strategy of prolongation of neoadjuvant ADT should not be routinely recommended in practice. And finally, given the similar relative treatment effect benefits seen by age or NCCN risk group, more accurate tools for risk stratification are needed to determine which cohorts of men derive clinically meaningful benefits from ADT treatment intensification.

Thank you very much for your attention. We hope you enjoyed this UroToday journal club, discussing the recent publication from the MARCAP consortium.