If you haven't read Part 1, here is the scenario again, and at the bottom, I'll put things that should've jumped out at you that needed closer scrutiny - and I'm sure there are more I didn't think of - any more you can think of?
And again, the point of this critical-thinking exercise is to take these critical-thinking skills and apply them just as diligently to your political "beliefs" when it comes to politicians and their press and pundits trying to "sell" you on something :)
*Again, this is a FICTIONAL drug I just made up - it does not exist.*
And no, by the way, I haven't jumped up into diabetes range myself, I'm still a thin, low-range prediabetic (knock-on wood) with an A1c of 6.1 at last check in August.
____________________________________
Okay, so let's pretend it's 2025 and you are the VP of Pharmacy for a national commercial health insurance plan. In two days, you have to give your recommendation to the Pharmacy and Therapeutics committee for a vote, whether or not to add this product to your formulary.
Now - remember the the IRA (Inflation Reduction Act) will be in full effect by 2025, which will lower copay costs to us, the consumers, to no more than $50 - especially in diabetes - which Big Pharma will likely respond to by simply raising their prices, putting the burden on insurance companies, who will, in turn, raise their premiums ;)
So Fictional Product X is a new GLP-1 that has just entered the market for type 2 diabetes.
You already have 3 GLP-1's on the market:
Product Y and Z are both self-injected, either daily or once weekly, both proving to have similar efficacy at significantly reducing both hemoglobin A1c and weight and are fairly safe.
Product Y has also just published a post-launch 2-year study proving long-term prevention of cardiovascular events, with Product Z's results expected to be published soon with similar results.
Product Q is an oral pill, slightly less effective than Y and Z, but also proven in a post-launch study to offer some cardiovascular protection.
Through contracting, you received a bigger rebate for Product Z than Product Y - thus, Product Z is on the preferred tier (lower copay/coinsurance).
Product Q, the oral formulation is also co-preferred on the preferred tier (to give providers and patients a choice, especially those who have trouble self-injecting).
Thus, because of it's higher cost, Product Y is on the non-preferred tier (higher copay/coinsurance, likely a step through either Product Z and/or Product Q first before you can get to it).
Also, another recent anti-diabetes class is available, the class of SGLT2 inhibitors, which have been found to be just as effective.
Your company's PBM - Pharmacy Benefit Manager (the company that contracts and manages all your pharmaceutical products) tells you Product Y - the currently non-preferred drug - is now offering an even steeper rebate of an additional 20%, if they are made 1 of 2 preferred options on the formulary, after this new drug launches.
You don't care how you get to a better bottom-line net cost - whether high wholesale cost/higher rebate or low wholesale cost/low rebates, just so long as at the end of the day - if all products are equal or near equal in efficacy and safety - you get the best deal so you don't have to raise your premiums for members.
(The PBM, however, does care - because they make most of their large profits off high wholesale cost/high rebates ;)
Your key-opinion expert endocrinologist says the efficacy of this new longer-acting GLP-1 is actually somewhat less than the others, and she has some some safety concerns.
However, 10% of your plan's in-network endocrinologists have called you, telling you they believe this drug is actually better and they'd like to try it, especially for their patients that have failed the other GLP-1's, they don't care what it costs, especially now that their patients only have to pay no more than $50 with the Inflation Reduction Act.
*Now remember, this is a pharmaceutical company, trying to sell you a product - so think carefully and critically - think to yourself, "What's wrong with this picture?"*or "What is missing from this trial?"
FICTIONAL CLINICAL TRIAL OF FICTIONAL PRODUCT X:
INDICATION: An FDA-approved treatment of diabetes type 2 in patients 12 years and older.
TRIAL DESIGN: A double-blind (both patients and the doctors administering the drug are blind to who is receiving which), randomized (random draw as to whether the patient gets treatment or not), placebo-controlled (a placebo is the comparator, either a pill with no active ingredients or has active ingredients) controlled clinical trial.
TRIAL DURATION: 12 weeks.
TRIAL SIZE: 150 patients, all formally diagnosed with type 2 diabetes.
DOSAGE: A once-monthly injection - 8 mg, 12 mg, or 16 mg based on weight.
PLACEBO: Active comparator; metformin (biguanide drug class).
INCLUSIONS: Formal diagnosis of type 2 diabetes, 12 years of age or older, both treatment-naive and treatment-experienced patients.
EXCLUSIONS: Patients 0-11 years of age, patients with history of a vascular or cardiovascular event in the last two years. History of endocrine or thyroid tumors or cancer.
PRIMARY ENDPOINT: The reduction of hemoglobin A1c from baseline by 1% versus the active-comparator placebo (metformin) by Week 12.
SECOND ENDPOINT: Weight loss of 1.5 pounds by 0 pounds by active-comparator placebo (metformin) by Week 12.
CONCLUSION/RESULTS:
- Product X showed a 1% percent greater drop in hemoglobin A1c versus placebo.
- Product X showed an average loss of 1 pound more than placebo.
- Side effects were nausea, vomiting, and constipation in 33% of the patients.
- Severe adverse events were retinopathy reported in 10% of the patients by Week 12, and 2% thyroid and pancreatic tumor formation by Week 12.
- Discontinuation rate was 25%.
______________________
Now, the pharmaceutical company notes that there was no head-to-head comparison against other GLP-1's, nor SGLT2's (though both were on the market when the study began and have become the standard of care by many endocrinologists.)
However, in their presentation, they have a slide of side-by-side comparison with the market-leading GLP-1.
All previous trials for GLP-1's have gone on 26 weeks rather than 12 weeks, so let's pretend their results at 12 weeks were up to at least 10% better than Product X.
The WAC or wholesale acquisition price, or list-price range before rebates for GLP-1's already on the market is now $1,000 to $1,500 per month/$12,000 to $18,000 annually.
However, the manufacturer of Product X tells you that the WAC price before rebates for Product X will be at a 30% per-dose premium to the others - making it $1,300 to $1,950 a month ($15,600 to $23,000 annually).
They are justifying this 30% premium cost because they are once monthly instead of once weekly, which they say not only saves cost, but will promote better patient adherence, plus they are approved for younger patients, plus they provide a new self-injector with new technology that sends information about whether the medication was administered correctly, at the appropriate times, using Bluetooth technology, to both the pharmacy and the provider, also ensuring better patient adherence.
1) There are an estimated 37 million people in America with type 2 diabetes and 450 million in the world - 150 people is too small of a sample size!
2) The trial duration is only 12 weeks - which is not only of shorter duration than the prior studies of GLP-1's, but too short to determine long-term benefit!
Also, A1c's are your blood's median average over of 2 to 3 months, so they're never even checked sooner than 3 months. One blood draw after 3 months is not enough to determine long-term benefit.
3) The active-comparator of metformin is already known not to lower A1c as effectively as the newer agents, nor lower risk of cardiovascular events, nor significantly affect weight.
Thus, we've already known for years that the new mechanisms of actions outperform metformin, we don't need another study - give us something new, a new comparator!
4) There is no breakdown between how many participants in the trial were treatment-experienced or treatment-naive, only a mention that both were accepted in the trial.
Thus, we don't know how many had been on anything else right before the trial or not before to cause those results, or whether there was a "wash-out" period before participating in the trial (which should've been mentioned in the inclusion/exclusion criteria).
5) There is no mention of what the baseline A1c's of the participants were before treatment - not even a median average.
6) Note the 1% reduction in A1c is relative to metformin - not an overall drop in A1c!
So it's a 1% decrease in A1c versus whatever metformin's drop was.
So if metformin only dropped your A1c from 9% to 8.9%, then Product X dropped it from 8.9% to 8.1% - which may be statistically significant, but it isn't clinically significant - your A1c is still way above target range of below 7%.
But an overall absolute-value drop in A1c from 9% to 8.1%, would be less than 1% - so instead, they used the delta drop between Product X and metformin because the relative -value percentage between them is the greater number, see how that works? ;)
Tricky, tricky tricky :)
And either way, you're still not at target A1c range.
Also, it doesn't say whether this score was "placebo-adjusted" or not - meaning whether or not they accounted for/subtracted out placebo effect first.
7) Metformin is not known to significantly affect weight - thus a 1.5-pound loss would be expected with a GLP-1's mechanism of action, but losing just 1.5 pounds in 12 weeks could simply be the result of diet and exercise, not necessarily related to the drug.
8) The most recently-launched GLP-1's have included proven avoidance of cardiovascular events - where is this data, and why isn't it even mentioned that there would be a post-launch study regarding this?
(Likely because 12 weeks isn't nearly enough time to study this - even 26 weeks isn't enough. The other drugs did publish this data in a phase 3/4 study two years later, though.)
Also, what I neglected to mention was that the other GLP-1's only restricted people with prior cardiac events for 60 days prior to treatment with Product X - why two years for this one, with no mention of studying cardiovascular events at all?
9) The safety, well, sucks, versus the shorter-term GLP-1's.
If there are only 150 people in the trial and 10% of them got retinopathy after 12 weeks, that's 15 people and that's a problem - one that's not easily remedied.
And 2% developed endocrine and thyroid tumors (not mention if they were cancerous or not) may sound low and may not warrant a black-box warning, but 2% of 150 is 3 people, it's a pretty serious side effect - and now we're talking about 12 to 18-year-olds as well.
(Believe it or not, the 33% with GI problems is actually not unusual with most drugs and interestingly, is often the same in placebos without active ingredients.)
10) 25% of people discontinued the product - why? Where is that breakdown?
11) They assume patient adherence will be better with once-monthly dosing from a technologically-advanced device that gives administration readouts - but the results aren't showing better efficacy as a result of the device.
Thus, you would need a study to prove that the device itself provides better adherence (and thus efficacy) before trying to charge a premium for it ;)
*Remember, any percentage results in the small study population, positive or negative, will only be amplified in the real world because more people will be taking the drug than in the controlled clinical trial*
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.