TCR: First of all, Dr. Koran, I can’t resist plugging your marvelous book on OCD (Obsessive-Compulsive and Related Disorders in Adults: A Comprehensive Clinical Guide, 1999, Cambridge University Press). It’s readable and chockfull of great information. How did you get into the field of OCD?
Dr. Koran: I had a patient who had OCD, anorexia, and depression who I treated for years and then read about Anafranil (clomipramine) which wasn’t available in this country. But Prozac (fluoxetine) was in development, so I called Lilly and asked if I could get it in a humane use protocol for her. They agreed, and she became remarkably improved. A few months later Lilly called me and said, “We are going to try it in OCD, would you like to be an investigator?” It sounded like fun and I did it. Eventually, I specialized in OCD.
TCR: Let’s start with the basics then, with the SSRIs. Are there any in particular that you like to start with in your practice? What is you rationale for approaching this decision?
Dr. Koran: We choose SSRIs using three criteria. First, if a first-degree relative has responded to a particular SSRI, we start with that, hoping there is some biologic similarity. If there isn’t such a relative, then we choose based on side effects and drug interactions, and that leads us to choose Lexapro (escitalopram) first. While there aren’t controlled trials of Lexapro in OCD, I believe it is just as effective as the other SSRIs. Furthermore, Lexapro (along with Celexa/citalopram) is least likely to cause drug interactions, and at least in depression trials, Lexapro is the SSRI least likely to be discontinued for intolerable side effects. The discontinuation rate for Lexapro is 5% and for the other SSRIs it is 9-18%. So we usually start with Lexapro.
TCR: With Lexapro, what kinds of doses are you shooting for?
Dr. Koran: We shoot for doses as high as 30-40 mg QD. We get there over a number of weeks, starting perhaps with 5 mg because with any SRRI we try to start at half doses to avoid queasiness. Once we get past the queasiness stage with Lexapro, we might try 10 mg for a week or two, then 20 mg for two or three weeks, and then shoot for 30 mg if the patient is willing, not having troubling side effects, and has shown no response. For all the SSRIs, we push people to the highest comfortably tolerated dose and then wait 6-10 weeks because we don’t know what the right dose is for any given patient. Meanwhile, we may try cognitivebehavioral therapy and refer the patient to www.ocfoundation.org for psychoeducation.
TCR: There has been a buzz in the psychiatric community that Anafranil may be more effective than SSRIs. What are your thoughts?
Dr. Koran: That was based on a meta-analysis by John Griest. Based on several later meta-analyses the evidence does not support an advantage of Anafranil over SSRIs. The advantage reported in that first meta-analysis might have reflected the fact that the Anafranil subjects in early studies were more treatment responsive because they had never had any other treatment, whereas some people in later trials had failed Anafranil already. So you weren’t looking at the same patient population as the trials evolved.
TCR: So based on that, you start patients on SSRIs rather than Anafranil?
Dr. Koran: Yes. I treat a patient with at least two and possibly three different SSRIs before considering Anafranil. One of the augmentation strategies, however, is adding low to moderate doses of Anafranil to an SSRI, based on some open label studies showing efficacy.
TCR: How would you start Anafranil in a patient already taking an SSRI?
Dr. Koran: It is specific to the SSRI. With Lexapro or Celexa, one doesn’t worry about drug interactions. Luvox, on the other hand, inhibits the metabolism of Anafranil and one expects a 4-fold higher clomipramine level with this combination. If I augment Luvox, I first decrease its dose to about 150 mg QD, then add 25 mg of Anafranil. I wait two weeks to get to steady state, and then measure a blood level. Then, depending on the level, I would raise the Anafranil dose to 50 mg and wait a couple of weeks, and so on. I’m shooting for a combination of clomipramine plus desmethylclomipramine levels between 200 and 500 ng/ml 12 hours after the dose, but no higher than that because of risks of seizures and cardiac effects.
TCR: Do you always get EKGs on patients on Anafranil?
Dr. Koran: No, only for those who are over 40.
TCR: What are the chances of getting a response simply by switching from one SSRI to another?
Dr. Koran: There is only a tiny literature on that. In one study of people who failed one SSRI and were given Celexa, 14 out of 18 responded. An unpublished poster argues the opposite in a Zoloft (sertraline) trial–that if you had been exposed to one SSRI and were given Zoloft, your chance of responding dropped from about 50% to about 30%. And there is a Prozac trial analysis which said that if drug-naive, you had a roughly 40% chance of responding to Prozac, but if you had been exposed to another SSRI, your chance dropped into the low 20s. My clinical experience is that if you fail one SSRI, you still have about a 50% chance of responding to the second, but that is simply a clinical impression.
TCR: So if you were to start the patient on your first choice, Lexapro, and the patient failed that, which SSRI would be your second choice?
Dr. Koran: Statistically, all the drugs are equally likely to help any given patient. So if there is no family history to guide me and drug interactions are not a concern, we might think about using either Luvox or Zoloft, and it would be a bit up to the patient. I would say, “If we give you Zoloft, you are a little more likely to have loose stools, and if we give you Luvox, you are a little more likely to have drowsiness. Which one of those things would you least like to have?” I lay out likely side effects of each SSRI so that the patient can participate in the choice.
TCR: So, if you have tried a second SSRI and the patient still is not doing well, what is your personal clinical algorithm for deciding on how to augment?
Dr. Koran: The best-supported augmentation strategy is adding an atypical antipsychotic in low dose, and the two that are supported by double-blind placebo-controlled trials are Risperdal (risperidone) and Zyprexa (olanzapine). There are multiple open label trials for Seroquel (quetiapine), and we are about to finish a double-blind trial for Geodon (ziprasidone). There are no data that I know of for Abilify (aripiprazole), but just as all SSRIs work for about the same proportion of OCD patients, all atypicals probably work equally well for augmentation, but you don’t know which patient will respond to which drug. In our Seroquel open label trial, we had patients who had failed one or two other atypical augmentations and still responded to Seroquel. So it is not yet predictable.
TCR: What sort of doses are we talking about with atypicals for OCD?
Dr. Koran: With Risperdal, you start at 0.25 mg or 0.5 mg QHS and then go up by 0.25 or 0.5 mg depending on the patient’s anxiety level about taking medications and the urgency of getting better. With Zyprexa, you might start at 2.5 mg QHS and again increase every four to five days by 2.5 mg. The response is usually rapid at a given dose: within two weeks people know whether they are better or not. With Seroquel, we usually start at 25 mg or 50 mg and we shoot for somewhere between 200 mg and 400 mg. In our Seroquel study we observed sedation at all dose levels but mostly it was transient.
TCR: So your favorite initial augmentation trials would be one of the atypicals?
Dr. Koran: Probably Risperdal starting at 0.25 mg and going from there.
TCR: Moving on from the atypicals, what would be number two for you?
Dr. Koran: Number two is adding Anafranil (discussed above). And after that there is one double-blind placebo-controlled trial suggesting that pindolol 7.5 mg a day (2.5 mg TID) is a reasonable augmenter, but the literature is mixed. And all the other augmentation strategies are supported only by case series or by case series that are subsequently contradicted by double-blind trials. For example, BuSpar (buspirone) looked quite good in initial case series but did not outperform placebo in two double-blind trials. Nonetheless, my clinical experience is that BuSpar augmentation works well in about one out of five patients. I think that the biological heterogeneity of the disease defeats attempts to find augmentation strategies that are effective for large proportions.