PsychEducation.org (home)

Details on the Lamotrigine Rash
(revised 5/2005; reviewed 3/2008, minor revisions)

Rash Rates
First, do we know how this medication compares with others which are also recognized to cause "Stevens-Johnson Syndrome", and other severe skin reactions?  A direct comparison was attempted using a German database.^Mockenhaupt  The authors are current or former employees of GSK, the company that makes lamotrigine, so these numbers are still somewhat suspect until some other source repeats this kind of analysis. However, it looks to me on reviewing the design of the study that a very earnest attempt was made to capture "real" rates, not those favorable to GSK.  For example, they made a series of assumptions about use rates then checked their numbers when different assumptions were made.  As you can see in the following table, the results were pretty similar either way (the right-hand column shows the rate adjusted with a different assumption about the rates of new prescription): 

"Bottom line":  the rate of serious rash in this analysis is between 2 and 4 cases per 10,000 new users of lamotrigine (the manufacturer's usual quoted rate is higher: 1 per 1,000).  The rate in several other commonly used antiseizure medications is somewhat lower, except for Dilantin.  

If rash, then what?
There are ways to identify the rashes that carry great risk (e.g. Stevens-Johnson syndrome, SJS; and toxic epidermal necrolysis, TEN. Sounds scary just naming it.  But that's all right, because this is indeed a very scary skin condition).  The problem is that simple rashes, lacking the known danger signs, can also be risky, so it's not easy to say "oh, this rash is safe".  And since there is a very low rate of severe skin reactions that can (rarely) even be fatal, any rash that appears while a person is taking lamotrigine should raise concern.

What are the signs of great risk?  Here are the versions I've heard so far: 

• "Anything above the neck"
• "Around or in the mouth"
• "Soft tissues (like mucous membranes of mouth, nose, eyes -- including the membrane over the eye (conjunctiva), so a red, sore eye counts) (or the anus, also a mucous membrane, also counts)
• "Anything on the face"

  Obviously these all describe nearly the same area, but the "soft tissues" version is the most specific. 

Three strategies for rash

1. Stop for any rash anywhere.
2. Have a dermatologist see the patient within 24-48 hours; hold the doses until seen.  
3. Stop for any rash above the neck; for anything else, reduce the dose to the previous level, and hold it there until you can tell whether the rash is going away (if so, continue upward again but more slowly and/or by smaller steps; use Benadryl or topical Caladryl to control itching while you're waiting).

Why the hurry to see the dermatologist?  Well, the obvious reason is because if there is a risk, you want to stop right away.  Secondly, if the medication is stopped for more than 3 days (for any reason; some sources say four days, a few say 5), the patient must start again from the very beginning of the dose steps.  Therefore if the patient can be evaluated very quickly, she can stop the medication, get a "green light" from the dermatologist, and resume the medication at the same dose. 

If you don't have a dermatologist handy, then strategy #2 may not be practical.  That leaves #1 and #3.   Number 3 has been recommended in several meetings I've attended, e.g. by Dr. Lauren Marangell at the Menninger Utah meeting, Winter 2002.  She spoke confidently of that approach there.  There is at least one article which notes having used this approach successfully.^e.g.Huang   But at a more recent meeting, the manufacturer's representatives were more cautious:  by their account, if you can't do #2, you should do #1. 

Update 4/2005:  Several neurologists in Denmark recently reported on their experience "re-challenging" patients who had a rash when given lamotrigine, which appears in the first line of the table below, along with two other reports they cited.   

My view

Here's one more bit of information which strikes me as overwhelmingly important on this question.  If starting at 25 mg lowers the risk of rash, compared to starting at 50 mg -- and that's clearly been shown -- , then is the risk lower yet if you start at 12.5mg?  The graph below shows another way of looking at this same issue (blue bars show dose per day, by week; yellow bars are guesses, extrapolated from existing data, shown in green):  

Think about it:  this medication has extremely few side effects, and does not cause weight gain (unlike all 3 main alternatives: lithium; Depakote/valproate; Zyprexa/Seroquel/aripiprazole).  It also has antidepressant effects like lithium.  So, this can be a great medication for someone. Why risk "blowing it" by going up a little faster on the dose and thereby raising the risk of getting the rash, and having to consider stopping entirely at that point?  For most patients considering lamotrigine, they've had symptoms for years.  Waiting another few weeks because of using a slow dose increase -- and thus buying a little more insurance that they might be able to benefit and stay on this medication -- just makes more sense to me. 

Thus my main strategy on "handling the rash" is to try to avoid it in the first place.  [Update 5/2005: Stanford's Bipolar Clinic, led by Dr. Terry Ketter, published their results from a strategy in which they warned patients not to add any new allergens: no new soaps, detergents, cosmetics, shaving creams, deodorants, etc; no new foods; and avoiding sunburn or poison oak.  They also waited to start if the patient had recently had a rash, symptoms of a viral infection, or a rash.  Using this strategy, they lowered the benign rash rate -- there were no serious rashes in this group of 100 patients -- to 3%, versus the common 10% rate for rash in other lamotrigine research studies.^Ketter][Update 2008: when they repeated that strategy with the pharmacist giving the skin instructions, the difference in rash rates was not significant.^{Ketter (b)}  Ketter says he still gives the same instructions, despite those data.]  

At the most recent lamotrigine experts' meeting, it was clear that slowing the rate of increase lowers the rash rate.  What is not so clear, at least to me so far, is the dosing rate at which the rash won't get any less frequent, even if you go slower (huh?  read that sentence again...).   Seems to me that is the rate we ought to be using.  Is this "lowest-reasonable-rate" 25 mg to start, as the manufacturer now recommends? (for adults not taking Depakote)  What about starting at 12.5 mg -- would that lower the rash risk a little further yet? 

When I started using 12.5 mg as a starting place, I think the rate of rash went lower.  But if there was much of a difference, it was pretty small.  So I now use the manufacturer's "starter kit", which begins with 25 mg per day for two weeks.  I point out to patients that the starter kit they give neurologists begins with 50 mg per day for 2 weeks.  Apparently they think psychiatrists are a bit more wary and need to have their wariness accommodated with a lower starting dose. (Someone suggested that this is not a global wimpiness on psychiatrists' part, as the neurologists are similarly wimpy when they use lithium!). In any case, it's nice to know that those with more years of experience with this medication use it more aggressively, not less.

When I want to be as cautious as possible, I sometimes still start at 12.5 mg and go up weekly by that amount, as this does not really introduce any additional delay compared to the starter kit, just some hassle for the patient, who has to chop a lot of pills.  I'll often jump not from 50 mg to 100 but rather take a week at 75 mg (all those chopped halves from the starter kit leave enough for this week), then go to 100 mg.  And if I'm really concerned, I'll use the 5 mg pediatric dose even in adults, going up by a pill a week until reaching 25 mg, then by 12.5 mg steps from there to 50, then 25 mg steps from there to 100, and then I start to relax a bit.  Remember, you can never relax completely.