I was privileged to virtually attend Prof Chris Parker’s investigator briefing on the initial results of the radiotherapy arms of the RADICALS trial on Tuesday, prior to his delivery of the late breaking abstract at the ESMO congress in Barcelona yesterday. This is practice-changing science and I’ve been honoured to have been allowed to contribute to it.
RADICALS is a fairly complex trial which was set up at a time when immediate adjuvant prostate-bed radiotherapy was recommended for men who had high risk features (such as involved surgical margins or high grade disease) following radical prostatectomy. Difficulties in defining the prostate bed have always necessitated a “blunderbuss” approach to the radiotherapy; complicated by the migration of normal organs into the space left behind by prostate removal. The study was set up to clarify whether close monitoring of blood levels of prostate specific antigen (PSA) could identify whether some men could be spared the toxicities of this treatment. A second randomisation – not currently reported – looked at whether there was a benefit to adding testosterone suppression to the radiotherapy.
One of the biggest challenges in conducting a trial like this is in selecting clinically relevant endpoints which will be achieved in a clinically useful timeframe. The primary endpoint was set as distant metastasis free survival. In prostate cancer, this often serves as a useful surrogate for overall survival where it can take many years (or decades) to achieve sufficient numbers of overall survival events to occur. These data are still immature and were not presented.
The current analysis is of a selection of the secondary endpoints; essentially looking to see if there was evidence of relapse found on PSA testing or scans, or initiation of 2nd line therapy for relapsed disease. First, Chris showed us the initial power calculations and the event rates seen for these endpoints closely match the estimations. He went on to show the Kaplan-Meier curves for the secondary endpoints. I hope to show these when slides are available, but it is sufficient to say the curves for immediate vs delayed radiotherapy were closely overlaid. The numbers told the same story. For example, the PSA control rate on the immediate radiotherapy arm at 5 years was 85% versus 88% on the PSA monitoring/delayed radiotherapy arm (HR 1.10 95%CI 0.81-1.49 p=0.56). There was no hint that immediate radiotherapy was better than PSA monitoring/delayed radiotherapy in terms of cancer control.
At 8 years follow up, only one third of men on the monitoring/delayed arm have required radiotherapy. Unsurprisingly, then, the toxicity rates on this arm are lower than those on the immediate radiotherapy arm. This was seen for all grades of toxicities. The rate of urinary incontinence was twice as high at one year in the men who received immediate treatment and the rate of severe stricture of the urethra (the tube connecting the bladder to the outside world via the penis) was also increased.
Pending the mature primary-endpoint analysis, these data are sufficient to recommend PSA monitoring for men who have high risk features after radical prostatectomy. Control rates appear similar and two thirds of men can be spared the toxicity of radiotherapy. It must be stressed that the PSA monitoring regimen used in the trial was much more intense than that which is often recommended, and that the median PSA at time of delayed therapy was 0.2 (IQR 0.1-0.3). Often, 0.2 is taken as the trigger to begin to discuss radiotherapy. That situation is no longer tenable.
There will be controversy in territories where immediate radiotherapy has remained standard practice, and where compensation drives these interventions. In the UK, at least, there has been a drift towards delaying radiotherapy over the past decade; so, beyond the change in intensity of PSA monitoring, this should be easy to implement. Congratulations to Chris and the team on this elegant and robust piece of research, and many thanks to the hundreds of men who volunteered for the trial.