CATIE

April 16, 2006

CATIE: Sigh

1. You know, if you're going to be rigorous about BID dosing schedules because the FDA requires it, why so liberal with total dosing for Zyprexa? A mean dose of Zyprexa is 20.8 is way (150%) above FDA guidelines. For comparison, that would have meant dosing Geodon at 240mg, Seroquel at 1000mg, and Risperdal at 6mg. BTW: a mean of 20.8mg means that a lot of people were dosed with MORE than 20.8mg (max=30mg).

2. The miracle here isn't that Zyprexa won, but that Zyprexa 20mg barely won against Geodon 114mg.

3. Why Trilafon (perphenazine)? Originally you thought all conventionals were the same; so why not Haldol? Or Mellaril? You say it's because it had lower rates of EPS and TD, which is fine, but then why exclude TD patients from that arm?

4. So you excluded patients with tardive dyskinesia from the perphenazine group (fine) but then had the nerve to say people tolerated it as well as other meds? Do you think maybe people who have TD may have different tolerances to meds? Different EPS? Different max doses? That they're just different?

5. You can't generalize from an obviously slanted "typical" arm to all other typicals. If you chose Trilafon over Haldol because of better tolerability a priori, you can't now say that "typicals" have equal tolerability to atypicals. Why not pick two typicals of differing potencies (like Mellaril and Haldol) and infer from there?

6. Do you actually believe-- does anyone believe-- that any of these patients are compliant with BID regimens? Especially with sedating meds like Seroquel?


The secret to understanding CATIE 2 is to understand that there are two CATIE 2s.

CATIE2-Efficacy: People who dropped out of CATIE 1 because their med didn't work were randomized to Clozail, Zyprexa, Risperdal or Seroquel. On average, new Clozaril switches stayed on 10 months, everyone else only 3. 44% of Clozaril stayed on for the whole 18 month study; only 18% of the others completed the study.

CATIE2-Tolerability: People who dropped out of CATIE 1 because of side effects (not efficacy) were randomized to Zyprexza, Risperdal, Seroquel, and Geodon (not Clozaril.) Risperdal patients stayed on for 7 months, Zyprexa for 6, Seroquel for 4 and Geodon for 3.

CATIE2-Efficacy is fair. If you fail a drug, you're likely to do better on Clozaril than anything else.

CATIE2-Tolerabilty makes no sense at all. The reason Geodon was used is because it has "very different" side effects. Hmm. How? "In particular, ziprasidone [Geodon] was known not to cause weight gain." But this assumes that the intolerability of the first antipsychotic was its weight gain.

Most importantly is this: if a patient couldn't tolerate their first antipsychotic, how likely is it that it was effective? In other words, if it wasn't tolerable, it wasn't efficacious-- these patients could have been in CATIE2-Effectiveness study. So how did they choose?

Easy: they gave the patient the choice: Geodon or Clozaril? Out of 1052, half left altogether. 99 went into the Clozaril study (CATIE2-Effectiveness) and 444 went into Geodon (CATIE2-Tolerability.) Of the 444 in the Tolerability trial, 41% were actually labeled first drug non-responders. 38% were labeled as not tolerating their first drug, but of those, who knows how many were also nonresponders?

And 74% dropped out again.

If you take the 444 in the Tolerability study and divide them into two groups:

• those who left CATIE1 because of lack of efficacy: then switching to Zyprexa or Risperdal kept them on their meds longer. (Which makes no sense again: this is the same thing as the CATIE2-Effectiveness, where (except for Clozaril) there was no difference between Seroquel, Zyprexa and Risperdal.)
• those who left CATIE 1 because of lack of tolerability, then it made no difference what you switched to.

Sigh.

And what's with the blinding? In every other study with a clozapine arm, you equalize the weekly blood draws by making everyone have to submit to them. But in this case, they unblinded clozapine so as not to have to subject all these people to blood sticks. Except they were subjecting them already-- they were checking blood levels.

And where was perphenazine? "[CATIE1] did not anticipate this unexpected result [that perphenazine would be as efficacious] that challenged the widely accepted (but never proven) belief that the newer atypical antipsychotic medications are better than all older antipsychotic medications" and so was not considered for CATIE2. Apart from the fact that it is simply untrue that anyone thought the atypicals were more efficacious than the typicals, it is furthermore untrue that that the authors did not "anticipate this unexpected result." In 2003, after basically doing Medline meta-analysis, they found that "not all of them were substantially different from conventionals such as perphenazine."

What's funny about these guys is how they conveniently lump all typicals together but arguing for differential effects of individual atypicals; then argue typicals are different from each other to justify picking Trilafon; and then say atypicals are different from each other ("not all of them were different") but typicals are all pretty much the same ("conventionals such as perphenazine.")

Bottom line:

The stated purpose of CATIE2 was to help clinicians decide which drug to switch to if patients a) failed their first drug; b) couldn't tolerate their first drug.

The divorce rate in America is 40-50%. Say you get divorced, and a friend says, I have two women for you, Jane and Mary. If the problem with your first wife was that she didn't turn you on, you should marry Jane. If the problem with your ex was that she was annoying, you should marry Mary.

What's going to happen here is that your second marriage, to either girl, is doomed. Certainly more than the national average of 50%. How long is it going to take before your second wife doesn't turn you on either? How long before you find stuff intolerable about her? The answer is, more likely than your first marriage-- say, 75%-- because the problem isn't your wives, it's you. You've framed the question in an idiotic and arbitrary manner. You don't get married to get turned on OR to be with someone who isn't annoying. You want the marriage to have both simultaneously, and much more. These things are not separable. This is CATIE2. A meaningless dichotomy-- efficacy and tolerability are not separate, let alone opposites-- used to create a false paradigm of medication selection.

April 17, 2006

CATIE Reloaded

And enough with the notion that medication compliance is a good proxy for overall efficacy.

All of these horrible psychiatry studies-- CATIE, Lamictal and Depakote maintenance trials, etc-- keep telling us how long patients stay on medications, because they say this means the drugs are working. The authors think that if a drug is working, they patient will stay on it. But you would think this only if you didn't actually treat many patients. I can make a similar argument that staying on a medication is inversely related to efficacy-- because when a patient feels better, they simply stop taking their meds.

Think about antibiotics. People don't finish the full 14 day course, precisely because they feel well. If they felt sick, they would probably take them longer than 14 days. In fact, people overuse these antibiotics even when its a virus, despite the antibiotic having no efficacy at all. They will demand an antibiotic even though know that it shouldn't be doing anything.

Same with pain meds. Oh, that's an acute problem? How about the chronic problems of diabetes and hypertension. People will skip/miss/forget doses when they feel asymptomatic, and will be more compliant when they have symptoms associated with these illnesses (e.g. headache, dizziness, etc.)

Look, I'm not telling you that compliance and efficacy aren't related. I am saying that if you want to measure efficacy, don't use compliance as a proxy-- go measure actual efficacy. And don't tell me it's too hard. You got $67 million for this study. Find a way.