Q:  Depakote & Liver Problems, Enzyme Levels, Pancreatitis

Dr. Phelps,

Thank you so much for this web site. My son has bi-polar. He is 15-years-old and is the typical adolescent with bi-polar. We recently went to the Dr. for his regular review of meds. He has been on Depakote for about a year. We increased it about a month ago from 500mg P.M. to 1000 mg P.M.  His blood work just came back with a bad report about liver enzymes and possible underactive thyroid. The doc wanted to monitor it for another month if I was ok with it. My husband wanted to take my son off so we went back to 500mg for now, (don't want to simply stop).

Because of the info on pancreatitis, (potentially fatal linked to depakote), liver problems, and possible thyroid problems I'm quite concerned. He is also on Risperdol, 1mg am and 2 mg pm, and clonidine .1  to ,2 mg pm. Any advice you have would be greatly appreciated.

Thank you



Dear Ms. S' -- 
Your question raises at least the following three issues about risk: liver risk, pancreas risk, and what to do when you are worried about risk!  Here we go with each (the third part goes on a bit, as you'll see, well beyond your question to a couple of general points). 

1. About liver
In a review of 35 years of experience using valproate (Depakote is one version), it was noted that severe liver problems with Depakote occur "with an overall incidence of 1 in 20,000, but a frequency as high as 1 in 600 or 1 in 800 in high-risk groups such as infants below 2 years of age receiving anticonvulsant polytherapy."^Perucca  Although Depakote has its problems, a high risk of causing severe liver problems (unless taking an additional antiseizure medication, which risperidone and clonidine are not) is not among them.  A rate of 1/20,000 is pretty close to the rate for the general population not taking any medications. 

However, it can cause an increase in liver enzyme levels that indicate some liver dissatisfaction with Depakote but are not necessarily a marker for the severe version.  Nevertheless, when we see that increase in liver enzyme levels in a blood test, indicating that some liver cells are being injured, we often watch to see that this is really going to turn out to be a problem, if the increase is small -- say, up to twice normal levels of those enzymes, and some doctors say up to three times normal levels is okay. That might sound like a lot, but where the average upper limit of normal for these tests is around 35 in most labs (and thus "three times normal" would be about 100), alcohol can cause these numbers to go up to 300-400, and some forms of viral hepatitis can cause it to go much higher. I saw a patient with enzymes around 1,000 once. So up to "three times normal" is not much and not a cause of any lasting liver damage unless we leave it that way.  Just "watching", for few weeks or a month, then repeating the test, is a routine thing to do. Sounds like that's what your son's doc' was doing. 

2. . About pancreas
I've had difficulty nailing down a similar number for the rate of pancreas problems with Depakote (is it higher than the 1/20,000 figure above, for example?) An article in 2001 noted that there were 40 reported cases in the world (reported in English).^Taira I'm not sure how many patients in the world have taken valproate (using the generic name to capture everyone who would be at risk). But I'll bet that comes to a risk figure quite a bit lower than the liver risk rate. 

3. About decision-making
Here comes  the tricky part. What do you do with this kind of information?  What should a person do when she learns about risks the doctor didn't mention?  This happens all the time. The doctor will never tell you about all the risks of a medication. It would take too long.  We should have a sign over the door: "Caution, medications used here. Enter at your own risk." Nearly any medication, even something so seemingly innocent as aspirin or penicillin, carries with it some degree of risk. For example, if you do an internet search (e.g. Google) with the drug name followed by product information pdf , in the top 5 you'll usually find the "PI", the information sheet that the pharmacist usually gives you with a new medication -- you know, the one that looks like a stock market ticker-tape, unfolding down to the floor? This includes a section, called Adverse Events, which includes things that have happened to less than 1% of people taking the medication. Here you'll find all sorts of nasty sounding things, for nearly any medication. 

So, instead doctors select risks to tell you about. I pick the common side effects which are likely to occur and bother people. For Depakote that's weight gain, which will bother most people if it happens, and it happens commonly (unless one lowers the dose back to the point at which appetite goes back to normal, below which point most people don't gain weight, in my experience; don't do that on your own, anybody -- you'll see that's the point of this little essay I'm going on with here...).  And nausea at first; and hair loss, which at low doses is generally mild and not enough to change appearance but freaks people out.  

And then I tell them about risks, including both short-term risks like a rare liver or pancreas or amylase-related reaction; and long-term risks, which for Depakote are nearly nil as far as we know after several decades of use (except that the pancreas case reported by Taira above occurred after 19 years of use, which I didn't know; that is so rare it was worth Dr. Taira and colleagues writing it up after 40 other cases had been reported, which all occurred sooner).  I hope I'm not missing anything there: one can never say "never" about a medication, you see, that's the point.  There are rare cases of all sorts of weird things happening. And think about it: if some weird thing is really rare, it would take a long time to figure out it was associated with Depakote.  So in fact, in that way, medications we've used for a long time, like Depakote, may be "safer" in the sense that we've had a long time to learn about the possible rare problems -- such as pancreatitis, which was recognized as a problem only just a few years ago. This makes you wonder what other medications' long term risks might turn out to be, doesn't it?

So, there are two points I'm trying to make here.  First, everyone who takes a medication (or parents who watch as their children are offered a medication) should be aware: you're taking a risk. There is always an unknown level of risk, as well as some known ones we may be able to tell you about. You shouldn't consider taking the stuff if you have such mild symptoms that you are getting by pretty darn well, as long as those symptoms are not a marker of some possible bigger problem which will grow if you don't treat the symptoms you have now (and there is reason to think that bipolar symptoms, even when mild, may for some people -- we just don't know how to tell who -- mark a risk for worsening that can be reduced with treatment, although this concept is still a working theory and not very well established).  There should be some sort of "risk/benefit ratio" evaluation: do the benefits you may get from the medication justify the risk you face in taking it?  Doctors are supposed to be able to help you evaluate that risk/benefit ratio, but as you can see from this lengthy letter, doing so well can take a really long time, far longer than most doctors can really be expected to spend with you, you see? So they give you the big side effects, and the big risks, and you have to trust their judgment and experience to believe that taking into account everything they didn't tell you, the risk/benefit ratio is still strongly in favor of taking the medication. 

The second point:  what should you do when you discover later, perhaps through your own search for information about the medication or perhaps from a pharmacist, or from reading that ticker tape Product Information, that there are risks you haven't heard about from the doctor?  Pancreatitis from Depakote, for example?  Here's what I recommend: if you have concerns which rise up later, please talk about them with your doctor (here's a little essay on how to do that without getting brushed off as might happen sometimes with some doctors). I hope you can see from this long story here that evaluating the risk/benefit ratio is difficult for a doctor, because she can't really know all the possible risks. But she can know the common benefits, and she is in a much better position than you are to see those!  Depakote for example: she uses it all the time. She sees people get much better on it. She knows that part really well.  You need to give the doctor a chance to reassure you about the risk/benefit ratio when you discover some risk you hadn't heard about. 

Finally, though, you should also be prepared for some irritation in the doctor's voice when you do this.  If we all had plenty of time, this wouldn't be such an issue. If you make a specific separate appointment to raise your concerns, and the doctor is going to get specifically and separately (from the usual routine of care) paid for it, then you are not as likely to encounter this problem. But if the doctor will not get paid extra to deal with the concern you raise, then I hope you can understand: it's like going back to the grocer and asking for some more lettuce, when he just sold you a head of lettuce at what he thought was a fair price.  If you want it for free, you'd better be prepared for some irritation. If you say "hey, you didn't tell me there was some dirt on one of the leaves of this lettuce", when that is just part of the deal when you buy lettuce (analogous to the risk everyone takes when they take a medication, if you follow me here), you'd better be prepared for some irritation.  On the other hand, a good grocer may take the time to explain to you that all lettuce comes with this risk; or why that particular lettuce had more dirt on it than usual. He may want to keep your connection to his grocery. Your doctor may want to keep your connection because she knows that the connection itself helps the medication work well (there is good evidence for this), and it certainly is likely to help you keep taking a medication about which you have some concerns. But since for most doctors, there is some connection between time and money, asking for time without any payment (whether from you or their employer) is always going to run the risk of that irritation. That's why I wrote the essay on how to talk with doctors and increase the likelihood of getting what you want, linked above. 

Whew. That was more than you were after, perhaps. Sorry. Obviously I needed to write that (I'm posting parts of this reply, with no mention of you or your letter of course, on my own site, because I want to be able to refer my own patients to it). Thanks for your question. 

Dr. Phelps

Published October, 2005