What MDACC and their like will never say, but is undoubtedly true (take a look at the demographics that Palex posted), is that protons are for the rich elite. Either people who are comfortably wealthy with good insurance who can afford to spend a couple of months in Houston or the obscenely rich for whom money is never an object.
Just try and offer the protons. Really hype them. Put the patient on a guilt trip to make a proton trip. And then if all that fails, you did the right thing and the burden of feeling like you have mistreated the patient is removed. Let the insurance thing fall on the academic center. (If it’s Medicare, no problemo.) I’m sure if they can’t get protons approved they will send the patient back to you for IMRT.
Actually seeing a patient who has a GBM demanding they receive protons. I'm sure we all have had these experiences because they all fed false hopes of a cure with protons and nothing else is acceptable.
The centers around here will kidnap my patients even if they were to give IMRT. They use protons as a marketing tool to get them through the door. By the time I see them again, they were already told how much better their Trubeam is compared to mine.
I mean we all know what happens next.
Like the Shania Twain song “You’re Still a One”
Well keep in mind there are two downward ticking events in PFS curves: progression or death. And protons tended to show survival advantage so PFS separation at longer interval could make sense if protons are less murdery than photons. That said divining much from the way the KM curves dance or don’t is bootless mostly.
Perhaps the protons are indeed less murdery betwixt the 4 and 5-year mark and then similarly murdery once again. Seems unlikely, but possible.
Especially with all the resources they have
It might be because of all the resources they have...
This could happen if the proton patients receive adequate palliative (and sometimes even curative) treatment of the recurrence they experienced at 2 years out of primary treatment. Not necessarily because they had protons, but because perhaps as more wealthy patients had better follow-up, access to more salvage treatments and overall supportive services.
The initial recurrence would/should have registered as an event at the time it occurred for PFS for those patients though. They'd have been censored as an event prior to death.
Fully correct, I was adressing the dip of the OS, not the PFS curve. The PFS would not mirror that.
This is all missing the exceedingly important element of temporality. The raw can be very much different than the actuarial, and can even be flipped one versus the other. (And lost to follow up when? Day 0? Day 1000? Either way they’re censors; this is normal.)
??? That needs to be better explained
Gotcha. Agree with you as well. We were talking about 2 different things, lol.
I understand all of the above and agree. I don't have access to the temporal data. But we can "assume" some elements of temporality from historic observational trials that show 95% of oropharyngeal recurrences occur within 24 months, and almost 0% occur after 48 months.
About 52% never smokers, 42% had quit smoking, 8% smokers. 95% HPV positive. 76% ECOG 0.
Could explain feeding tube rates. I'm cool with my grizzled old population. They tank far less frequently than the young healthy guys.
Yep...
How are the hazard ratios so different when the curves look very similar? And I must be stupid because I always thought if the hazard ratios confidence intervals cross 1 that that means the result isn’t statistically significant?
He fixes the cable?
This is true to an extent. For NIH-sponsored studies, CMS will cover the treatment. This creates a bit of an ethical issue for some studies that plan to enroll patients across an age range (i.e. those under 65 would not have Medicare). Otherwise insurers are not obligated to cover experimental treatment on trial. Some private insurers allowed for coverage of proton therapy for patients on trial, or enrolled in registries, but this was likely negotiated ahead of time. If insurers were obligated to pay for treatment on study you can only imagine the abuse that would follow (and the amount of garbage studies that would be created and likely never get published).
Would be interesting to see feeding tube placement by location of treatment.
It’s an upside down world I think with non inferiority trials
Hazard ratios can vary significantly with similar looking curves when the outcomes are rare in both arms. Here, both progression and death are pretty rare, so this explains the HR peculiarity for this trial. It's a bad number to emphasize to the public...like emphasizing HR of chest wall recurrence for DCIS treated with XRT after mastectomy. There may be a 0.5 HR. It is probably not clinically significant.
Yeah, I have a problem with the study itself and the insane process of allowing discontinuation/crossover based on insurance approval/disapproval. By definition, this means that your randomization is bogus. Was there even really intent to treat per protocol, or rather was the intent to "treat per protocol if payment available", As discussed above, the latter is not remotely the former.
While I could more loquaciously quibble with some of this, in general I don't see any "HR peculiarity" here (or "pretty rare"-ness). It's all seemingly straightforward, no chicanery, etc., to my eyes.
The "9% margin" short changes just how unambitious they were.
That seemed to be the intent.
Yet another example of a large (clinically unacceptable?) delta in an NI trial. In general the delta for NI trials should be roughly 50% of the delta in contemporary superiority trials; assuming the investigators have equipoise.
