Chris Wallis: Hello, and thank you for joining us for a UroToday Journal Club discussion. Today, we're talking about a recent paper entitled "Sequencing of Androgen-Deprivation Therapy for Short Duration with Radiotherapy for Nonmetastatic Prostate Cancer, the SANDSTORM Study: A Pooled Analysis of 12 Randomized Trials".

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 the Journal of Clinical Oncology, led by the MARCAP Consortium investigators.

The question of sequencing of systemic therapy with radiotherapy is important across a number of cancers. And interestingly, the sequencing question has been associated with differential survival for many cancer sites. In prostate cancer, the combination of androgen deprivation with radiation therapy is well established in patients with intermediate and high-risk localized disease, and is standard of care. However, the optimal sequencing of ADT is controversial, given that ADT is administered for a longer time than it is necessary to administer radiotherapy. There are alternate approaches, concurrent followed by adjuvant ADT has been compared to neoadjuvant followed by concurrent ADT, and based on two recent studies, one in individual patient data meta-analysis of the RTOG 9413 and Ottawa 0101 trials, and another, an IPD meta-analysis of three neoadjuvant trials versus four adjuvant trials. The current data suggests that a concurrent fall by adjuvant approach may be beneficial. However, there's substantial limitations to these data.

The hypothesis the author sought to test in this study was that concurrent or adjuvant ADT improves metastasis-free survival, compared with neoadjuvant and concurrent ADT, in patients who are receiving short-term ADT associated with radiotherapy.

To do this, they relied on the meta-analysis of randomized trials in cancer of the prostate, or MARCAP consortium, which has compiled individual patient data from randomized control trials performed by multiple organizations.

In the context of this particular study, they included randomized control trials including patients receiving neoadjuvant or concurrent, or concurrent and adjuvant ADT, defined as four to six months of therapy, with radiation on at least one of the trial arms. Neoadjuvant concurrent ADT was defined as that therapy were started at least two months before the start of radiation, which concurrent or adjuvant ADT started at the commencement of radiation, and lasted for at least two months after its completion. All patients included in this analysis were dichotomized into one of these two groups without any overlap.

There was a number of exclusions required because of differences between the included studies. They excluded patients who received purely adjuvant ADT, such as those performed in the RTOG 9413 trial. They excluded patients where greater than 80% of the ADT duration was neoadjuvant with insufficient concurrent therapy. They excluded trials where they were unable to calculate metastasis-free survival due to a lack of data. They excluded patients treated with non-steroidal antiandrogen monotherapy. And there was a number of trial specific considerations such as in CHHiP, patients who received non-steroidal antiandrogen monotherapy were excluded in an EORTC 22991. They had to impute the use of whole pelvic radiotherapy, as this was one of the key stratification factors.

So the primary endpoint, as alluded to in the hypothesis, with metastasis-free survival defined as the times from random assignment until metastasis or death from any cause. There are a number of secondary endpoints including overall survival, the cumulative incidents of biochemical recurrence, and just a metastasis of prostate cancer specific mortality, and of other cause mortality.

The authors used IPTW weighting to account for differences between groups, and provide a less biased estimate of the effective ADT sequencing on outcomes. These weights included age, the initial PSA diagnosis, Gleason score, T stage, radiotherapy dose, mid-trial enrollment year.

Cox proportional hazard models were used to calculate hazard ratios for metastasis-free survival and overall survival, where Fine and Gray computing risk models were used to calculate sub-distribution hazard ratios for biochemical recurrence, distant metastasis, prostate cancer specific mortality, and all other cause mortality. Prostate cancer specific mortality, non-prostate cancer mortality was considered the competing event. Conversely, for other cause mortality, prostate cancer mortality was the competing event. For both distant metastasis and biochemical recurrent, death of any cause was considered a competing event. All adjusted models further include the exposure variable of ADT sequencing, as well as the use of whole pelvic radiotherapy and an interaction to account for potential effect modification.

Where an interaction between field size, i.e., prostate only versus whole pelvic radiation, and the clinical impact of ADT sequencing was identified, the authors planned a priori to repeat all analyses within subgroups of cohorts defined as those were receiving prostate only and receiving whole pelvic radiotherapy. Within each of these cohorts, propensity scores were recalculated, and then IPTW adjusted Kaplan-Meier analyses were performed separately within each cohort.

In addition to using accumulative incidents, the authors further used restricted mean time lost as an alternate endpoint within these subgroup analyses to measure treatment effect. They performed sensitivity analyses excluding EORT 22991, given that it required imputation for the identification of whole pelvic radiotherapy status.

The authors further performed pre-specified analyses for the interaction between NCCN risk group and the impact of ADT sequencing on metastasis pre-survival.

I'm now going to hand it over to Zach, to walk us through the results of this trial, this pool of analysis.

Zachary Klaassen: Thanks so much, Chris. So this is the baseline characteristics for this analysis. We can see in the middle here, this is a comparison of neoadjuvant and adjuvant ADT combined with concurrent for both of these groups.

You can see at the top here, there was no difference between in age between these groups, median age of 70. In terms of radiotherapy dose, there was more high dose radiotherapy in the adjuvant group at 47%, versus 38% in the neoadjuvant group. In terms of pelvic nodal radiotherapy, more common in the adjuvant group at 57%, versus the neoadjuvant group at 38%. The adjuvant group was also higher with regards to T stage, with higher proportion of T3/T4 at 34%, versus 20% in the neoadjuvant group. Again, this was comparable. In terms of Gleason score between these two groups, with higher PSA in the adjuvant group at 13, versus 11 in the neoadjuvant group. In terms of NCCN risk group, again, we see that there was higher risk patients at 49% NCCN high-risk, in the adjuvant group, versus the neoadjuvant group at 36%.

The next several slides will look very similar to this. This is going to be Kaplan-Meier and cumulative incidence curves evaluating different outcomes. This is looking at adjuvant versus neoadjuvant ADT. So looking on the left metastasis-free survival was improved in the adjuvant group, which is in red, and the neoadjuvant group, which is in blue, hazard ratio favoring adjuvant ADT was 0.83, with a 95% confidence interval of 0.74 to 0.93. Again, we see an improvement of biochemical recurrence for the adjuvant group versus the new adjuvant group, hazard ratio of 0.77, and 95% confidence interval of 0.68 to 0.87.

This is a similar comparison, but looking at distant metastases, interestingly, there was no difference in distant metastases between these two groups. However, we do see an improved prostate cancer specific mortality with adjuvant versus neoadjuvant ADT, with a hazard ratio of 0.76, and a 95% confidence interval of 0.59 to 0.99.

The final curves for this specific analysis look at other cause mortality and overall survival. There was an improvement in other cause mortality for adjuvant ADT, hazard ratio of 0.77, 95% confidence interval 0.67 to 0.88. And we also see an improvement in overall survival, with adjuvant versus neoadjuvant ADT, hazard ratio 0.79, and a 95% confidence interval of 0.70 to 0.89.

The next analyses looks at these outcomes specific to just prostate only radiotherapy. Again, the similar breakdown in terms of color scheme for these slides as the previous ones. In terms of metastasis-free survival, we see a significant benefit with adjuvant versus neoadjuvant ADT, with a hazard ratio of 0.65, and a 95% confidence interval of 0.54 to 0.79. Again, a statistically significant benefit for biochemical recurrence for adjuvant ADT, hazard ratio of 0.46, and a 95% confidence interval of 0.35 to 0.59.

Similarly, we see a benefit for adjuvant ADT versus neoadjuvant for distant metastases, hazard ratio of 0.52, 95% confidence interval of 0.33 to 0.82. Again, a benefit for adjuvant versus neoadjuvant ADT for prostate only radiotherapy for prostate cancer specific mortality, with a hazard ratio of 0.30, and 95% confidence interval of 0.16 to 0.54.

Finally, in terms of prostate only radiotherapy, other cause mortality, no difference between these two, but we do see, again, a benefit of adjuvant versus neoadjuvant ADT for overall survival, with a hazard ratio of 0.69, and a 95% confidence interval of 0.57 to 0.83.

Similar set of slides, looking now specifically at whole pelvis radiotherapy. There was no difference between adjuvant and neoadjuvant ADT with regards to metastasis-free survival, as well as biochemical recurrence-free survival. When we look at the distant metastasis, interesting here, we see a flip. So there was a benefit for neoadjuvant ADT versus adjuvant ADT, with a significant hazard ratio of 1.57, 95% confidence interval of 1.20 to 2.05. No benefit for prostate cancer mortality when comparing adjuvant versus neoadjuvant ADT for patients receiving whole pelvis radiotherapy.

Finally, looking at other cause mortality, we do see a benefit for adjuvant versus neoadjuvant ADT, with a hazard ratio of 0.59, 95% confidence interval of 0.47 to 0.73. And we do see a statistically significant benefit for adjuvant versus neoadjuvant ADT, in terms of overall survival for those men receiving whole pelvis radiotherapy, with a hazard ratio of 0.82, 95% confidence interval of 0.68 to 0.99.

This is a nice Forest plot, basically summarizing the last nine slides, again, to highlight favoring concurrent or adjuvant ADT. We see significant benefit amongst all outcomes here for prostate only radiotherapy, with a benefit and overall survival for whole pelvis radiotherapy for adjuvant. In terms of neoadjuvant ADT, among men receiving whole pelvis radiotherapy, we do see a benefit with regards to the outcome of distant metastases.

So in this study, concurrent adjuvant ADT was associated with improvements in multiple oncologic endpoints, including metastasis-free survival and overall survival, in men receiving prostate only radiotherapy. In those men receiving whole pelvis radiotherapy, neoadjuvant concurrent ADT was associated with improved distant metastases, but not metastasis-free survival.

This is the first time in this study that a significant association was seen with concurrent or adjuvant ADT and overall survival.

And finally, the underlying mechanism behind superior outcomes with concurrent adjuvant compared with neoadjuvant concurrent ADT, coupled with prostate only radiotherapy, but not men with whole public radiotherapy, remains to be fully elucidated.

So in conclusion, ADT sequencing exhibits a significant interaction with radiotherapy field size, such that concurrent adjuvant short term ADT sequencing is associated with optimal oncological outcomes with men receiving prostate only radiotherapy. The effects are not as clear for patients receiving whole pelvis radiotherapy, although neoadjuvant concurrent short-term ADT sequencing may be preferred given its distant metastasis benefit.

These data strongly suggest that when prostate only radiotherapy is being delivered with short-term ADT, concurrent adjuvant ADT sequencing should be standard of care. These findings also support the recommendation that if whole pelvis radiotherapy is used, neoadjuvant concurrent sequencing should be the standard of care.

Finally, future trials, such as RTOG 0924 and the PIVOTALboost trial, will provide level one evidence evaluating the benefit of whole pelvis radiotherapy with neoadjuvant concurrent ADT sequencing in selected patients with intermediate-risk and high-risk prostate cancer.

We thank you very much for your attention, and we hope we enjoyed this UroToday Journal Club discussion of the recently published SANDSTORM analysis.