Advertisement - Members don't see this ad
Or do it backwards (axials first) if you can't resist a challenge.
FLAME TRIAL prostate boost - ?lymph nodes?
Started by BobbyHeenan
Does nodule contouring matter? Even though there was significant variation in contouring in the FLAME trial the results were still excellent, especially given that the authors state 90% of pts were high risk.
Also, I was just perusing the protocol recently while looking at their 10 year results and realized FLAME overall was a negative trial as it did not meet its primary endpoint. Am I reading it wrong? Did the authors disclose this in the manuscript? I could not find anywhere that this issue was addressed.
From the protocol:
The FLAME trial was a superiority trial designed to detect an increase of 10% in the 5 year biochemical disease-free survival (bDFS) in the focal boost arm.
But the difference in 5Y bDFS was only 7%. Negative, right?
I might overlook that issue since the 10 year results to some degree may outweigh the lack of meeting primary endpoint at 5 yrs but there was a lot of censoring in both arms. With quite a bit more censoring in the focal boost arm without any explanation by the authors as to why.
My thoughts remain that the only dose escalation trial showing an OS benefit (FLAME still did not) was GETUG 18 where patients were treated with conventional fractionation. Getug 18 OS at 10 yrs: 77% vs 66%. @Palex80 even convinced me that dose escalation decreases cardiac morbidity! WIN WIN!!!
Also, I was just perusing the protocol recently while looking at their 10 year results and realized FLAME overall was a negative trial as it did not meet its primary endpoint. Am I reading it wrong? Did the authors disclose this in the manuscript? I could not find anywhere that this issue was addressed.
From the protocol:
The FLAME trial was a superiority trial designed to detect an increase of 10% in the 5 year biochemical disease-free survival (bDFS) in the focal boost arm.
But the difference in 5Y bDFS was only 7%. Negative, right?
I might overlook that issue since the 10 year results to some degree may outweigh the lack of meeting primary endpoint at 5 yrs but there was a lot of censoring in both arms. With quite a bit more censoring in the focal boost arm without any explanation by the authors as to why.
My thoughts remain that the only dose escalation trial showing an OS benefit (FLAME still did not) was GETUG 18 where patients were treated with conventional fractionation. Getug 18 OS at 10 yrs: 77% vs 66%. @Palex80 even convinced me that dose escalation decreases cardiac morbidity! WIN WIN!!!
On X:
Advertisement - Members don't see this ad
Well, it depends on your theology, really.
For normal people: 60 months is long enough to offer it to patients with reasonable confidence.
For people who worship conventional fractionation like it’s still 1978: this approach was ready for clinical use the moment the protocol was written.
For SBRT/brachy purists: no, we need 25-year median follow-up, multigenerational toxicity analysis, and perhaps a notarized statement from the patient’s rectum before we feel comfortable
That guy has literally never had a thread about a study he likes.
That is not to say he’s wrong.
I think he is operating way outside the Overton window of the state of the modern study. 99% suck, so if you’re evaluating all of them on the 1% that are good, everything is going to suck. The mid point of the 99% is the metric basically.
Otherwise, it’s just “this sucks” and next.
That is not to say he’s wrong.
I think he is operating way outside the Overton window of the state of the modern study. 99% suck, so if you’re evaluating all of them on the 1% that are good, everything is going to suck. The mid point of the 99% is the metric basically.
Otherwise, it’s just “this sucks” and next.
Rumor is that someone licked all the red off his candy cane as a kid. He never got over it.
I like that he brings a different viewpoint to an issue.
The red can be licked off?!
What I see in the protocol is: "With this number of patients, the expected increase in five-year freedom from biochemical failure rate of 10% can be detected with a power of 80%."
The goal is to detect a statistically significant increase in bDFS (e.g., p < 0.05). If the *true* increase in bDFS were 1% and we enrolled a billion patients, the superiority trial would be considered positive. We're not going to enroll a billion patients, we're going to enroll N patients, and we're going to select N so that *if* the true increase were 10%, then we'd yield a positive result 80% of the time. However, because of randomness, the measured increase wouldn't be 10%, we would get some interval around 10% ... and 20% of the time, the trial would still turn out negative because of that chance.
Disclaimer: not a biostatistician
Similar threads
