Read the Full Video Transcript

Rashid Sayyid: Hello everyone. This is Rashid Sayyid. I'm a urologic oncology fellow at the University of Toronto and along with Zach Klaassen, Program Director and Associate Professor at Augusta University, we'll be discussing a recent publication from the Annals of Oncology looking at Transcriptomic and Clinical Heterogeneity of Metastatic Disease Timing with Metastatic Castration-sensitive Prostate Cancer. This article was published by Dr. Sutera and all in the Annals of Oncology.

We know that patients with mCSPC or metastatic castration-sensitive prostate cancer represent a heterogeneous patient population. For practical purposes, obviously based on clinical outcomes, this cohort is split into four groups commonly based on the disease volume, commonly using the charted criteria, and the timing of metastases. Using data from the CHAARTED and the GETUG-AFU15 trial, among men who received ADT alone, we can see that there are significant differences in the median OS based on how patients present, meaning timing, synchronous or metachronous. Synchronous means they present with de novo mets. Metachronous means later on. Then based on your volume. Synchronous high volume patients, so the worst group for prognosis, have a median OS of three years with ADT alone. If we compare that with a metachronous low volume patients, we see that it's much longer at about eight years.

Why does mCSPC risk grouping matter? This has important implications for treatment selection. Synchronous high volume patients commonly advocated that these patients receive early treatment intensification with triplet regimens. Conversely, patients with metachronous low volume, some have advocated for ADT alone. Obviously, there are things to consider, including patient age, comorbidity, their performance status and not just the disease, but many have advocated for ADT alone plus or minus an androgen receptor signaling inhibitor such as abi or enza. Also, this has implications for treatment of the primary disease site, whereby patients with synchronous low-volume disease may benefit from radiation to the prostate based on data from HORRAD and STAMPEDE Arm H. Although, this has been challenged somewhat by recent results from the PEACE-1 trial. Currently, current treatment recommendations and patient selection are based on clinical outcomes data, but the underlying mechanisms that help guide the treatment decisions are currently poorly understood. The interplay between disease timing and volume also is incompletely understood. We need to better understand how these patients differ from a genomic standpoint.

Transcriptomics may be the answer to this. What is tumor transcriptomics? It's essentially analysis of the entire collection of RNA sequences in the cell, or the transcriptome. Essentially in a simplified way, this provides information regarding what genes are turned on or off in the cells. Several transcriptomic signatures based on gene expression profiles of the primary tumor have been developed, and these have been shown to be correlated with clinical outcomes. Transcriptomic profiling has demonstrated predictive utility in determining benefit to combine systemic therapy within mCSPC. We can use this tool to better understand differences in the patients based on the disease, the timing, and the volume.

The study objective was to evaluate for differences in transcriptomic profiles relative to timing and volume of metastatic disease and associate them with clinical outcomes and response to therapy. For the purpose of this study, the author's performed an international multi-institutional retrospective analysis of mCSPC men who underwent RNA expression profiling of the primary tumor. I highlight primary because they did not get biopsies of bone mats or soft tissue mets. It was the primary tumor here. They performed RNA sequencing using a Tempus xT tissue assay, or they performed a microarray analysis using the Veracyte Decipher platform with the data subsequently normalized via an algorithm. They then perform batch correlation between the microarray and RNA sequencing using ComBat to all allow for combined analysis.

What did they evaluate in this setting? They importantly evaluated the gene expression of the androgen receptor activity. This is an important implication as androgen based therapy represents the backbone treatment for all these patients. They also looked at PAM50, which is a 50 gene signature that classifies prostate cancer into luminol and basal subtypes. They also looked at PORTOS, which is the post-op radiation therapy outcome score, the homologous recombination deficiency, the small cell/neuroendocrine variations. They also looked at different hallmarks of cancer such as androgen response, DNA repair and epithelial-to-mesenchymal transition. They also looked at WNT signaling mutations. The performed quality control of the samples by checking the proportion of reads aligning to the reference genome.

What data sources did they use? CHAARTED, STOMP and ORIOLE. STOMP and ORIOLE were two phase two trials that looked at metastases directed therapy for patients with metachronous oligo mets. They also used off trial data from patients from Ghent University and Johns Hopkins Hospital. RNA expression profiling was, in patients with synchronous metastases, was performed from the primary prostatic tissue, and in metachronous disease, this was also performed in the primary prosthetic tissue in the initial localized stage. You may ask, well, it's quite a homogenous sample with primary prosthetic tissue from both groups, but keep in mind when we're talking about metachronous disease, these patients are going to present with the initial disease and then later on develop the metastases. They may accumulate mutations over time. Although both use primary prostatic tissue, it's not really a fair comparison because metachronous disease patients may accumulate further mutations. When you sample the primary prostatic tissue, it may under-represent the mutations that are present. Something just to keep in mind as we're interpreting the results of the study.

The authors use the modified CHAARTED criteria to stratify patients by disease volume, with high volume disease defined as the presence of visceral metastases, or at least four bone mets with at least one outside the spine or pelvis. Then they relied on either conventional imaging or choline, PET/CT, so that's where the modified part comes from. That's because in the STOMP trial that we talked about, the phase two MDT trial, patients were staged using choline C-11 PET/CT as opposed to the classic conventional imaging that we see with the CHAARTED criteria. Follow-up clinical data was collected via serial physical exams, conventional imaging, and PSA measurements as well.

For the study endpoints, the primary was differences in transcriptomic profiles between the timing of disease. The median transcriptomic scores were reported for both metachronous and synchronous cohorts and compared using the Mann-Whitney U test, which is a nonparametric test given the smaller sample size. This was within the RNA sequence, the microarray and the combined batch, essentially trying to lump them together and then also parse them by the analysis performed. Then the proportion of patients with low androgen signaling was compared using the chi-square test, with low androgen receptor activity defined as less than equal to 11. This is important again because the backbone treatment for these patients is ADT, so that relies heavily on androgen signaling.

Secondary analysis included determining clinical and transcriptomic variables associated with time to CRPC and overall survival, with CRPC defined using Prostate Cancer Working Group 3 criteria. Then a subset analysis was performed to evaluate response to therapy and clinical transcriptomic differences accounting for both the timing and volume disease. As this is a time to event analysis, survival analysis using Kaplan-Meier curves was performed with between group comparisons, also analyzed using the log-rank test. Then variables associated with time to CRPC and overall survival were evaluated using multi-variable Cox regression modeling. At this point, I'll turn it over to Zach to go over the results and discussion for this paper.

Zach Klaassen: Thanks so much, Rashid. This is the demographic characteristics for the 252 patients included in this study. Median age was 64 years. The median PSA was 20.5, and the most common Gleason grade group was Gleason five at almost 50%. With regards to timing of metastatic disease, synchronous patients, there was 60% of patients, and metachronous was 40% of patients. When we look at disease burden, it was very well split; low volume, high volume 50/50, with a roughly 50/50 detection of low volume disease on conventional detection versus enhanced detection in the low volume patients, whereas high volume patients were predominantly detected with conventional imaging. When we look at treatments, we see several different varieties of treatments for these patients. ADT monotherapy 30.2%, ADT plus docetaxel 33.3%, MDT monotherapy 15.5%, MDT plus systemic therapy 17.5%, and observation was chosen in 3.6%.

This is the Kaplan Meier curve for time to CRPC in the entire cohort. Metachronous is in red, synchronous disease is in blue. The median time to CRPC for synchronous metastases was 15.4 months, and for metachronous metastatic disease was 91.0 months. The five-year time to CRPC rate was 24% for synchronous disease and 78% for metachronous disease. Similarly, this is the Kaplan Meier curve for overall survival. Median OS for synchronous disease was 48.8 months and was not reached among men with metachronous disease, and the five-year OS rate was 39% for synchronous metastases and 79% for metachronous metastases.
These are the box plots comparing androgen receptor activity and hallmark AR transcriptional scores for both metachronous and synchronous disease. To summarize this figure, basically we have androgen receptor activity A through C and hallmark AR transcriptional scores from D to F. Generally, we see that the synchronous disease had lower AR... excuse me, androgen receptor activity, as well as lower hallmark AR transcriptional scores. This was detected both on RNA sequencing and microarray as well as when combined and corrected in the batch score. Take home here is that synchronous disease has lower AR activity.

This is the multi-variable Cox regression for time to CRPC and overall survival. The top half of this table is for time to CRPC. We see that the bold variables, high versus low volume, very statistically significant with a hazard ratio of 5.22, but also hallmark androgen response. Higher hallmark androgen response, lower hazard ratio 0.56, both statistically significant. I have a box around synchronous versus metachronous because when adjusted for all of these variables, synchronous versus metachronous disease was not statistically significant for CRPC. When we look at the bottom of this table, similarly, we see that both high volume versus low volume was statistically significant as was hallmark androgen response. Again, for overall survival, when adjusted for these other variables, synchronous versus metachronous disease was not statistically significant for overall survival.

This is the Kaplan Meier curve for time to CRPC and overall survival for synchronous metastatic disease. Monotherapy treatment is in red and AR plus non-AR therapy is in blue. We see that the patients with synchronous metastatic disease had significantly improved time to CRPC when treated with AR plus non-AR therapy, as well as improved overall survival in the bottom right with a hazard ratio of 0.47 and a 95% confidence interval of 0.30 to 0.72. Similarly, for metachronous metastatic disease, similar analysis with monotherapy versus AR plus non-AR therapy. We do not see a benefit to AR plus non-AR therapy for time to CRPC nor for overall survival, with the overall survival hazard ratio of 1.37 and 95% confidence interval of 0.50 to 3.92.

Switching gears a little bit, this is the Kaplan Meier curve for time to CRPC and overall survival for high volume metastatic disease stratified by metachronous or synchronous disease. We see that for time to CRPC, no difference with a three-year time to CRPC for synchronous disease of 17% and 42% for metachronous. Similarly, for overall survival, no different for high volume metastatic disease, whether it be metachronous or synchronous with a three-year OS rate of 54% for synchronous versus 46% for metachronous disease. Similarly, for low volume metastatic disease, we do see differences in time to CRPC and overall survival. Three-year time to CRPC for synchronous metastasis, 64% versus 92% for metachronous, as well as a significant difference for overall survival; synchronous metastasis three-year rate of 76% and for metachronous 94%.

By way of discussion, this is the first report to demonstrate a biological difference between timing of metastatic disease with several notable findings. First of all, synchronous metastatic castrate-sensitive prostate cancer is associated with a lower androgen response transcriptomic profile. Secondly, although synchronous metastatic disease experiences a more aggressive clinical course on uni-variable analysis when accounting for disease volume and androgen response biology, timing of metastatic disease alone no longer appears to be a strong prognostic indicator. Patients with synchronous but not metachronous disease may experience improved outcomes with AR plus non-AR combination therapy, and as a result of a lower androgen response profile. Finally and notably patients with metachronous low volume disease do not derive the same benefit from ADT plus docetaxel, but they do benefit from ADT plus enzalutamide or apalutamide.

In conclusion, this multi-institutional series demonstrated for the first time a biological difference in timing of metastatic presentation of castration-sensitive prostate cancer for patients with low volume disease. Specifically, this study demonstrated that patients with metachronous low volume metastatic disease have a more ADT responsive transcriptional profile than those with synchronous low volume disease with timing of disease having a greater effect in patients with low volume metastatic disease. Finally, this study showed that patients with high volume or synchronous low volume disease may derive the greatest benefit from multimodal therapy targeting AR and non-AR biologies potentially as a result of this biological difference. We thank you very much for your attention and we hope you enjoyed this UroToday Journal Club discussion published in the Annals of Oncology.