| Affiliation |
(AGL) Chairman, Department of Ophthalmology, The Methodist Hospital, Houston, Texas; Professor of Ophthalmology, Weill Cornell Medicine, New York City, New York |
| Transcript |
"Today what I'd like to talk to you about is secondary IIH, and really this "I" is problematic because it means idiopathic so you might want to consider calling it: secondary pseudotumor cerebri. The ones that we're going to be talking about today are the medicines. The medicines that we typically ask about are steroids and normally that's steroid withdrawal, and not steroid administration. Vitamin A analogues including isotretinoinic acid which is Accutane, the tetracyclines and their derivatives including doxycycline and most commonly minocycline, and lithium- these are the top ones. Even though the book says oral contraceptive pills (OCPs) are included, normally, we consider this as a secondary cause only in the setting of cortical venous sinus thrombosis. We know that the estrogen preparations and the birth control pills increase your risk of thrombosis, and that is the presumed mechanism. Just having OCP alone is probably not a good linkage. These are the strongest ones, but you need to know that there's some other ones. So, when in doubt you should always just look it up. The one that is less common, but you need to know about is growth hormone- growth hormone patients also probably are reasonable. The way that we usually establish a cause-and-effect relationship between these drugs and the findings is to use the causality criteria, and there are a number of these causality criteria out there. I'll just share with you the ones that I like. Number one, there has to be a biologically plausible mechanism, so you can't just have an association. There has to be some reason that's biologically plausible as a mechanism for why that drug would cause the finding you're seeing. For example, for vitamin analogs that's thought to decrease outflow through the arachnoid granulations, and growth hormone similarly makes sense. The second is there has to be a temporal relationship between the cause and the effect, and that relationship has to be coherent with the pharmacokinetics of the agent. Normally, things have dose response curves like hydroxychloroquine and ethambutol. It takes a while for the drug to build up, and you don't get the toxicity, in hydroxychloroquine for example, until you're 5 because it's a cumulative dose effect. That's completely compatible with the temporal relationship between the bullseye maculopathy and hydroxychloroquine. The third is analogy, which is our fancy way of saying you can't be the only person that's had this, and that's why it's always wise to Google it and see if other people have had the same association. It could be human or animal study analogy. You have to have a specific effect, and pseudotumor cerebri is not that specific, so that's why there is this confusion with the birth control pill and pseudotumor cerebri. The same people that are taking the pill are the same people getting the pseudotumor, so in case control studies this doesn't actually fall out as an independent variable. That's why we'd like to have specificity of effect. Obviously, pseudotumor is not that specific. However you should really be thinking about it when you're a kid, if you're a thin person, a man, an elderly patient- those patients are much more likely to have an exogenous agent causing their pseudotumor than the typical obese female who just happens to be on tetracycline. You've got dose effect, so the dose gradient in the IIH cases are not good dose gradients. For the tetracyclines and the vitamin A analogs, there's not really a dose response curve where the more you take the more you're likely you are to get it. For ethambutol, if you're on 25 mg per kg, your chance of getting toxicity is a lot higher than 15 mg per kg. The same with hydroxychloroquine, which is 6.5 mg per kg, but you get toxicity at 5 mg per kg. The one that does have a dose gradient is the growth hormone. In the patients who have the growth hormone version, we can kind of reduce their dose or do half dose, and it might make them better. You need to be able to exclude chance alone, so you really need to in pseudotumor diagnosis of exclusion can't really be having any other feature here. That's why we're going to be more comfortable with thin elderly or male patients with exogenous causes than the alternative, which is an obese female who just happens to be on tetracycline. Then the strongest piece of evidence that we have is called a re-challenge or a de-challenge, and so if you take it away, their symptoms get better and if you give it to them again, they get the exact same thing again. A de-challenge/re-challenge is one of our strongest pillars for establishing a cause-and-effect relationship. For these four drugs [steroid withdrawal, vitamin A, tetracycline, and lithium] and probably growth hormone as well, we have pretty good concordance with the causality criteria. These are the ones that typically we're going to ask about. You could probably add nalidixic acid onto this list as well, but we don't use nalidixic acid anymore. This is kind of my list of secondary medicines that cause pseudotumor cerebri." |
