The rules should not be bent!
An open trial is where a physician gives patients a medication and measures how the patients do. Patients know that they are taking a medication and know what that medication is. They very likely think that the medication is expected to help them and that alone may increase the chance that they will get better. The physician almost assuredly wants to help the patient and wants to think that the treatment he is prescribing will do some good. This is actually the every day practice of medicine, but it's not science. The "art" of medicine should be solidly based on scientific research, but it is not research itself.
Numerous studies have shown that open trials are so full of biases that it is scientifically impossible to tell whether the medication is helping. In almost every condition, people will tend to improve. Perhaps it is human nature to jump to the conclusion that something you did caused that improvement. Unfortunately, it is just not possible to prove cause-effect with an open trial. And yet, numerous top psychiatry journals run articles of open trials containing wording suggesting the study gave some evidence of a cause-effect relationship. Over 40 open trial articles from psychiatrists at major universities claimed that gabapentin helped bipolar disorder and other psychiatric conditions and every one of them was wrong. See the Gabapentin Scandal.
Surprisingly, published open trials far outnumber scientifically sound, well controlled double-blind research studies. The current campaign by academic psychiatry to pretend to be scientific as shown by its use of the new catch phrase "evidence-based medicine" is little more than wishful thinking. The rule of evidence-based medicine is that medication treatments should be supported by double-blind random-assignment trials.
Open trials are frequently funded by pharmaceutical companies because they known that the publication of such trials will increase the sales of their drugs. These irresponsible articles do have a powerful effect on psychiatrist prescribing behavior, especially when quietly promoted by drug reps. In some cases, no harm is done, because the medication in question actually does help the condition being treated even if the open trial is of no value in proving such a connection. Later double-blind studies will provide the evidence that the medicine actually does work. However, in many cases, later double-blind studies have shown that the medicines touted in the open trials are actually worthless for the condition in question. For many patients of doctors influenced by the publication of the open trials, it is too late. The patients have already suffered side-effects of a useless medication and have been deprived of an opportunity to be treated with an alternative medication which has been proven to be of benefit for the condition the patient was suffering from.
Since many psychiatric diseases, especially unipolar and bipolar depressive disorders, have an elevated risk of suicide, wasting time with an ineffective treatment inevitably means more people will die and have died from suicide. Indeed, recent evidence that untreated depression damages the hippocampus, perhaps irreparably, suggests that useless treatments probably allow the damage to occur rather than providing the protective function that effective anti-depressants appear to have.
There is a very limited place for open trials. They are primarily to help researchers test whether any harm is occurring and whether there is any possibility that benefit is occurring before undertaking more elaborate double-blind studies. If patients get worse while on the medication, the medication is probably not helping. If patients develop severe side-effects or if patients can very easily tell that they are on the medication, this can prove useful knowledge to the research team before starting an expensive double-blind study. I, myself, did an open trial before starting a Phase II double-blind study of a medication I was studying for alcoholism. However, I knew that the apparent benefit in the open trial might be an illusion. I never considered for a single moment the possibility of publishing the results of my open trial despite the fact that I had many more patients in the open trial and for much longer periods of time than many of the open trials I have read.
Whether open trials should ever be published is another issue. I would argue that they should not since they intentionally or inadvertently mislead huge numbers of physicians to turn many of their patients into guinea pigs without the protections present in a study done under the supervision of an Independent Review Board (IRB) and without adequate informed consent. The rare exception to this rule is where an open trial is a run-up to a proposed double-blind study and found no evidence of possible benefit, leading to the cancellation of the double-blind trial. Publication of such a trial might protect other patients from harm.
Even if open trials of psychiatric illness are published, the wording in the open trial should never suggest that the results support a beneficial impact from the medication. Indeed, the wording should explicitly caution that the study provides no evidence of benefit since open trials can never provide evidence of cause and effect, especially in psychiatric illnesses which wax and wane in intensity. If an open trial of AIDS patients found 100% had an abrupt disappearance of all signs of illness and all signs of viral infection, obviously there would be some evidence of benefit, since spontaneous remissions in AIDS are extremely rare. Also, AIDS has objective blood tests that can't be changed by saying "I feel better" or altered by the wishful thinking of the physician, whereas psychiatric illnesses have no similar clear biological measures.
I have spoken to a few excellent academic researchers who are very concerned about the publishing of open trials. I know I am not alone, but I don't see any effort to convince journal editors to stop publishing such worthless and potentially harmful reports. I think I am obligated to do my part to point out the dangers in what is occurring. I know I am not sophisticated enough or well enough connected to get an article published, so my website will have to do.
Of course, many journals make a lot of money selling advertising space to drug companies. You don't bite the hand that feeds you. Below, you will see my very brief review of short-comings in a small percentage of the articles that appear to falsely promote the drug used in an open trial as having been shown to be effective in some way. You should be aware that I may be totally wrong in my opinions, but I don't think that is very likely.
(Please note: I started collecting the open trials for this page when I was reviewing reboxetine. This is absolutely no reflection on the value of reboxetine, which appears to be a useful anti-depressant medication that would be a good addition to those available in the U.S.).
More Irresponsible Claims From Harvard and CNS Spectrum for Manufacturer of Adderall-XR: In an uncontrolled, unscientific report from Harvard, supposedly to "assess the long-term safety and effectiveness of mixed amphetamine salts extended release" (Adderall-XR) in ADHD adults combined subtype, the psychiatrists report on a 24-month, open-label extension of a very short 4-week, DB PC study of 223 adults given up to 60 mg/day. At monthly visits, supposedly, "efficacy" was assessed with the self-report ADHD Rating Scale IV (ADHD-RS-IV). Self-reported symptoms decreased by 7.2 points (P<.001). The title: "Long-term safety and effectiveness of mixed amphetamine salts extended release in adults with ADHD." Biederman J, et al. Harvard. CNS Spectrum 2005 Dec;10(12 Suppl 20):16-25. Ed: This is deceptive and unethical publishing. An open trial of a highly subjective illness using self-reported symptoms while on an abuse-prone medication with a high street value cannot be called research. Many patients are known to sell their medications while most enjoy taking them personally and don't want the crash. If you were selling your medications, how would you make out those forms? If a person were addicted to marijuana and could get free marijuana every month, how do you think that person would fill out self-report forms? Harvard reports that "ADHD symptoms significantly improved for all subjects." How likely is that in a real world? Harvard, the home of Timothy Leary, has long been insensitive to the harm of controlled substances. Open trials are worthless, misleading reports which should be banned. They cannot prove cause-effect as claimed in the authors' title. Harvard and the CNS Spectrum should be ashamed.
Grossly Irresponsible Promotion of Topiramate (Topamax) for Bipolars by University of Toronto and Canadian Journal of Psychiatry: In a highly-biased, manufacturer-funded, open-label report of 16 weeks of topiramate added to the treatments of 109 bipolar patients (80% depressed, only 3% manic and 8% hypomanic), by the end of the 16 weeks, depression and manic symptoms had decreased. Three of the four authors are employees of the manufacturer. Only the head of psychiatry at the University of Toronto, Dr. McIntyre, is not an employee, although the university received funding for this study. The authors make the totally unsupported claims that, "Adjunctive topiramate treatment can reduce the severity of manic and depressive symptoms, as well as reducing tremor and weight in outpatients with BD I or II." Incredibly, the study was approved by Health Canada and various "ethics" boards. The authors do note that four [unpublished (i.e. suppressed), manufacturer-financed], placebo- and comparator-controlled trials conducted in patients with acute mania have failed to support the efficacy of topiramate as a monotherapy in bipolar disorder and are on file with the manufacturer, Johnson & Johnson. Open-label adjunctive topiramate in the treatment of unstable bipolar disorder. McIntyre RS, et al. University Health Network, Toronto, Ontario. . Can J Psychiatry. 2005 Jun;50(7):415-22. See Email Sent for any reply by the author.
Junk Psychiatry from the Journal of Psychopharmacology: In yet another misleading open trial, Italian psychiatrists used mirtazapine on 44 adults with generalized anxiety disorder. The claimed that they were studying the efficacy of mirtazapine, despite the fact that it is impossible to study efficacy with open trials in psychiatric disorders due to the extreme amount of bias that creeps into the large majority of studies. They claimed to have found a very high "response rate" in patients on mirtazapine, suggesting that these patients were responding to mirtazazepine and that in some way they had shown that mirtazazepine was effective. They claimed that their study "supports the notion that mirtazapine is an efficacious," which is absolutely untrue. Of course, virtually every anti-depressant ever researched for generalized anxiety disorder in double-blind research helps, but this is no excuse for publishing a worthless report. With the horrendous record of open trials misleading thousands of psychiatrists in the recent past in the gabapentin and topiramate bipolar scandals, you would think that academic journals would stop their slavish publishing of phony open trials. It is likely that this study was financed by the manufacturer and that the journal receives advertising income from the manufacturer. Mirtazapine treatment of Generalized Anxiety Disorder: a fixed dose, open label study. Gambi F, et al. University 'G. D'Annunzio' of Chieti, Italy. . J Psychopharmacol. 2005 Sep;19(5):483-487. The editor is David Nutt of the University of Bristol in the UK. . The for-profit publisher is SAGE publications. On Sept. 21, I emailed both the lead author and the editor. I will post any replies here.
Junk Science from the Journal of Affective Disorder: Psychiatrists from the University of Alcala in Spain make the wild claim that there retrospective chart review of just 34 charts of bipolar outpatients treated with lamotrigine as a sole or add-on treatment is able to "evaluate the effectiveness and safety of lamotrigine in bipolar spectrum patients." Clinic notes were used to score the Clinical Global Impression for Bipolar Disorders (CGI-BP-M) scale and to identify manic, hypomanic mixed or depressive relapses. Treatment duration ranged very widely from 6-96 weeks. While only 47% had their depression lift and only 32% remained not depressed in follow-up, they not only just to the conclusion that "lamotrigine was safe and effective" but also that the "effectiveness of lamotrigine was greater for those patients diagnosed with bipolar spectrum disorders other than bipolar I." Lamotrigine for the treatment of bipolar spectrum disorder: a chart review. Montes JM, et al. Universidad de Alcala, Madrid, Spain. . J Affect Disord. 2005 May;86(1):69-73. Ed: Shame on the Journal of Affective Disorder for publishing such dishonest and wild claims of effectiveness. Such open retrospective chart reviews are even worse that open studies which have already been repeatly shown to be virtually worthless and extremely misleading. I encourage you to email the editors of the journal, Hagop S. Akiskal of the University of California, San Diego, and protest this promotion of junk science. . and Cornelius Katona at the University of Kent in the UK. (The University of Kent emailed to note that the study had been accepted by Akiskal in the U.S.
More Junk Science from the Journal of Affective Disorder: Psychiatrists at the University of Tennessee treated 24 women with cyclothymic temperament and refractory depression with lamotrigine in a totally uncontrolled study. Eighteen (75%) met DSM-IV criteria for bipolar II disorder. In two thirds of the cases, lamotrigine was add-on therapy to an antidepressant. The only measure of response to therapy was the DSM-IV Global Assessment of Functioning (GAF). The authors claim that 16 (70%) had sustained "responses" instead of sustained improvement, suggesting that all 16 were helped by lamotrigine and that their study was proof of effectiveness. They state, "Robust, sustained responses to lamotrigine monotherapy were seen in 4 patients (17%)," and 12 had a remission of their depression for at least 12 months. Only "seven patients (30%) received no apparent benefit from lamotrigine." They conclude, "Lamotrigine induced prolonged illness remissions in a substantial number of female patients whose symptoms were both complex and refractory." Sustained remission with lamotrigine augmentation or monotherapy in female resistant depressives with mixed cyclothymic-dysthymic temperament. Manning JS, et al. University of Tennessee-Memphis. . J Affect Disord. 2005 Feb;84(2-3):259-66.
Irresponsible Claims From Harvard and Canadian Journal of Psychiatry: The authors claim that their tiny, open-label, uncontrolled, mixed diagnosis, 16-week study "assess(es) the antidepressant efficacy and tolerability of adjunctive ropinirole in 10 outpatients (7 MDD, 3 Bipolar) with treatment-resistant depression." The report on giving the Parkinson's drug ropinirole 0.25 to 1.5 mg daily added to tricyclic antidepressants or selective serotonin reuptake inhibitors found that MADRS depression scores decreased from 29.6 at baseline to 16.9 at endpoint (P < 0.02). The Harvard psychiatrists wrongfully claim that 4 of 10 (40%) patients "responded to the medication." They only know that four patients got better while on the medication. They also wrongfully claim that their poorly designed study "suggest(s) that, in selected cases of TRD, ropinirole augmentation of antidepressants is effective." Ropinirole in treatment-resistant depression: a 16-week pilot study. Cassano P, et al. Harvard-Massachusetts General. . Can J Psychiatry. 2005 May;50(6):357-60. Ed: The researchers did get informed consent and were monitored by an IRB. However, an uncontrolled, open-label study can never be used to determine possible efficacy in the treatment of depression. Maybe, ropinirole will someday be found useful. But, such open label, preliminary investigations should never be published and, if published, should emphasize that they cannot evaluate efficacy. Open-label studies have repeatedly misled thousands of American psychiatrists to give patients medications before they have ever been correctly tested for efficacy only later to find out that the medications were worthless for what the doctor was trying to treat.
Dr. Paolo Cassano Replies: Dear Dr. Radecki, I fully appreciate you concerns. Our study was indeed an open trial, and I will try to explain to you why we made the decision to publish it. In my opinion, you are raising a very important question: should open-label trials be published at all? On one hand , as you stated , there is a risk of clinicians jumping to conclusions and using medications inappropriately. On the other hand , there is a clear need to improve the treatment of patients who do not achieve adequate response with antidepressants, and yet no augmentation strategy used in practice has ever produced adequate evidence to obtain an FDA indication for resistant depression. Therefore, in the total absence of adequate evidence for any augmentation strategy currently used, clinicians struggle with what to do next for their patients. In addition, clinicians are better trained to understand the level of evidence and they usually know that the ultimate proof of efficacy comes from adequately powered double-blind trials. Moreover, there are still clinicians who use off label medications because their patients cannot tolerate or are resistant to other classes of drugs; open label studies can quickly provide some guidance in terms of safety and warn clinicians of any potential concern. Ultimately, researchers capitalize on open studies to plan double-blind trials. In fact, the estimate of the sample size for a double-blind study is usually based on the effect-size for the therapeutic effect of the medication observed in open trials . I think your question is open and deserves some research to rigorously assess the advantages and disadvantages of publishing the results of open label studies . For example, many of the SSRIs have indications for the treatment of anxiety disorders (PTSD, panic disorder, social phobia) based on double-blind studies that followed original open-label, pilot studies of off-label uses of the SSRIs. Patients with anxiety disorders have been treated successfully for their anxiety disorders with SSRIs even prior to establishing their efficacy with double-blind studies, and this was made possible by the publication of preliminary, open-label studies. This is basically why we decided to publish our open-label study. I hope you understand us. Sincerely, Paolo
Bogus "Research" from the University of Chicago on Children with Pervasive Developmental Disorder: The authors claim to have "assess(ed) the effect of escitalopram in the treatment of pervasive developmental disorders by doing a 10-week "study" using an open-label design, i.e., no scientific controls at all. Twenty-eight children with a PDD were given escitalopram (Lexapro) at doses up to 20 mg/d. There was significant improvement in ABC-CV Irritability Subscale Scores (20.5 to 10.9; p </= .001) and a slight Improvement on a Clinical Global Improvement Scale (5.2 to 4.6; p </= .001). The researchers made the totally unproven claim that 25% of children "responded at a dose less than 10 mg" and and an additional 36% "responded" at a dose greater than or equal to 10 mg. They claim that their bogus open-label study found "escitalopram to be useful" in treating PDDs. An open-label trial of escitalopram in pervasive developmental disorders. Owley T, Walton L, Salt J, Guter SJ Jr, Winnega M, Leventhal BL, Cook EH Jr. Department of Psychiatry, University of Chicago. J Am Acad Child Adolesc Psychiatry. 2005 Apr;44(4):343-8. Ed: It is absolutely impossible to attribute any improvement during the study to the medication since there was no control group. This study is intellectually dishonest and will mislead numerous psychiatrists to use a drug that has never been scientifically tested on similar children. Instead of irresponsibly publishing this "study," these psychiatrists should have done the right thing, i.e., done a real DB PC study.
Brazilian Academics Claim Open Trial Proved Reboxetine Effective for HIV Depression: In a very small open trial of 20 HIV patients who were suffering from Major Depressive Disorder, the authors report that 100% of the 15 who completed with 12 week trial of 8 mg/d of reboxetine had greater than 50% improvements in their MADRS depression scores. The authors write, "Within this open trial, reboxetine was found to be effective in reducing depressive symptoms in HIV illness." Surprisingly, they also call a 28% side-effect dropout rate as evidence reboxetine was "well-tolerated." Actually, this is almost twice as high a dropout rate as found in most large scale studies of standard anti-depressants. An open trial of reboxetine in HIV-seropositive outpatients with major depressive disorder. Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil. Carvalhal AS, de Abreu PB, Spode A, Correa J, Kapczinski F. J Clin Psychiatry. 2003 Apr;64(4):421-4
Irresponsible "Research" on Children From University of Illinois, Chicago: In small prospective 6-month open (uncontrolled) trial which supposedly examined the effectiveness and safety of divalproex sodium (Depakote) in pediatric mixed mania, 34 children with a mean age of 12 were treated with divalproex. The researchers claim an "effect size" based on change scores from baseline was 2.9 for the YMRS and 1.23 for the Child Depression Rating Scale-Revised and a response rate (>/=50% change from baseline YMRS score and </=40 score on CDRS-R at the end of study) of 73.5%. They also claim the medicine was safe and well tolerated. Divalproex sodium for pediatric mixed mania: a 6-month prospective trial. Pavuluri MN, Henry DB, Carbray JA, Naylor MW, Janicak PG. University of Illinois at Chicago (UIC). Divalproex sodium for pediatric mixed mania: a 6-month prospective trial. Bipolar Disord 2005: 7: 266-273. Ed: This is totally bogus. The researchers are claiming that in a uncontrolled study that the improvement was due to the medication. It is impossible to make such a claim from an uncontrolled study.
University of Toronto Psychiatrists Make Exaggerated Claims Favorable to Reboxetine Based on a Very Messy Retrospective Open Trial: In a report based on retrospective data on 45 patients treated by 7 different psychiatrists at 6 different test sites, the authors claim, "" The patients were a mixture of bipolar and unipolar depressed. Almost half were only moderately depressed. All had been on at least one anti-depressant for at least six weeks for the current episode of depression before being started on reboxetine through the Canadian Special Access Program. Many had received a large variety of treatments earlier in their lives, but there was no documentation in the report stating how many patients briefly improved while on these treatments. Half of the patients were simultaneously on a second psychotropic medication for their depressions during the study, although the authors never suggest that any of the improvement might be related to any of these medications. Some of these second medications lack double-blind evidence of any benefit for depression, i.e. gabapentin and dextroamphetamine. The only measure used in the study was the very rough measure of Clinical Global Improvement. The report is full of conclusory sentences in selected individual vignettes claiming that this or that patient responded or didn't respond to various individual medications, instead of simply stating that the patient improved or deteriorated while on the medication. Despite such a poorly controlled study, the authors make the exaggerated claim that exactly 56% were helped by the reboxetine. This was based on the fact that charts revealed that 56% were said to be much improved by their psychiatrists after being on reboxetine. Without saying it, the authors also suggest a 0% placebo response rate, while in double-blind studies of even severe refractory depression, some percentage of patients have always improved on the placebo. Of course, 11% were reported to be much worse, but the authors don't conclude that the reboxetine caused 11% of patients to become much more depressed! The authors list numerous financial links to the pharmaceutical industry. Reboxetine: a preliminary report on its use through the Special Access Program. Kennedy SH, Lam RW, Cohen NL, Rosenbluth M, Sokolov ST, McIntyre RS, Chue P, Craigen G; Canadian Network for Mood and Anxiety Treatments. J Psychiatry Neurosci 2002 Nov;27(6):418-22
Israeli Researchers Claim Open Trial Proves Reboxetine Helps Panic Disorder: An open trial of 29 panic disorder patients who had improved inadequately on SSRI medication were given a 6 week trial on reboxetine 8 mg/d. The Panic Self-Questionnaire (PSQ), the Hamilton Rating Scale for Anxiety (HAM-A), the 17-item Hamilton Rating Scale for Depression (HRSD) and the Global Assessment of Functioning (GAF) Scale scores all showed significant improvement. The authors conclude, "The 24 patients who completed the study responded well to reboxetine treatment." Of course, this suggests they all did better, which is highly unlikely, and suggests that the study proved that reboxetine caused the improvement in all of them, something that is just not true. They write, "Reboxetine appears to be effective in the treatment of SSRI-refractory panic disorder patients...." In fact, no one really knows. Double-blind trials are not that hard to set up. Why didn't they do one? The efficacy of reboxetine in the treatment-refractory patients with panic disorder: an open label study. Dannon PN, Iancu I, Grunhaus L. Hum Psychopharmacol. 2002 Oct;17(7):329-33. (Ed: Inevitably, drug detail salesmen will use this unscientific report to push reboxetine for Panic Disorder and some doctors will start using it as their first line of treatment. The only somewhat controlled study of reboxetine in Panic Disorder suggests that reboxetine is inferior to SSRIs. Dr. Dannon sent a very kind reply to this review, stating, "I am totally agree with you about your concerns. However, we can not control every published paper in this field." His study was not supported by a drug company and they are not suggesting reboxetine as an initial treatment. Dr. Dannon says that there are two double-blind studies published on the issue. I didn't find any on PubMed as of 11/22/03 or in the journals I track, but I will try to find them.)
German Researcher Encourages Reboxetine for Parkinson's Depression Based on Open Trial: A very small 16-patient and very short 4-week open trial of 16 Parkinson's disease patients suffering from depression and not helped or having side-effects with a previous anti-depressant were given reboxetine in an open trial. Hamilton Rating Scale for Depression, the Self-Rating Depression Scale, the Snaith-Hamilton Pleasure Scale, and the Social Adaptation Self-Evaluation Scale were measured. Patients improved while on reboxetine. The author concludes, "Reboxetine appears to be effective." He urges physicians "to consider reboxetine for treatment of depression in Parkinson's disease." Effect of reboxetine on depression in Parkinson's disease patients. Lemke MR. J Clin Psychiatry 2002 Apr;63(4):300-4. (Ed: Why not do a double-blind study? Why not compared two anti-depressants? Of course reboxetine has been found to help depression like any other anti-depressant and I don't recall seeing a study showing that one works better than another in Parkinson's disease. Obviously, Lemke is promoting reboxetine without any scientific evidence.)
Drug Industry Open Trial Claimed to Show Efficacy of Reboxetine Based on Open Trial: A study financed at least in part by the drug company of 128 MDD patients who had not adequately improved on fluoxetine in 6-12 weeks of treatment had the patients switched to reboxetine 8-10 mg/d in twice a day dosing. Patients improved over the 8 weeks of the study with significant improvement during the first week. The most common side-effects were insomnia, headache, dry mouth, diaphoresis, and constipation. Unfortunately, there was no control group, e.g. a group staying on fluoxetine or a group put on placebo or another anti-depressant medication. Therefore, it can only be said that some of the improvement during the study was probably due to the reboxetine, but there is absolutely no evidence that the patients did better than if they had simply been left on the fluoxetine. Switching to reboxetine: an efficacy and safety study in patients with major depressive disorder unresponsive to fluoxetine. Fava M, McGrath PJ, Sheu WP; Reboxetine Study Group. J Clin Psychopharmacol 2003 Aug;23(4):365-9. (Ed: This report is really junk science. There is simply no excuse for an industry funded study of a medication that has been around for years not to be designed in such a way as to advance science. To claim in the title that this open trial is to examine efficacy is intentionally misleading their non-research audience. For a little extra expenditure, a real study could have been done.)
Viennese University Psychiatrists Recommend Reboxetine for Bulimia Based on Only 7 Cases in Open Trial: An extremely small 7-patient open trial is reported where three of the seven dropped out with increased in symptoms or laxative abuse. Monthly binge frequency decreased 73% over the 12 week study and vomiting decreased 67%. Despite this extremely small trial and despite all of the biases of open trials, the researchers encourage clinicians to consider reboxetine in bulimia management saying, "Reboxetine seems to be an option for the treatment of bulimia nervosa." University Hospital, Vienna. Reboxetine in the treatment of bulimia nervosa: a report of seven cases. El-Giamal N, de Zwaan M, Bailer U, Lennkh C, Schussler P, Strnad A, Kasper S. Int Clin Psychopharmacol. 2000 Nov;15(6):351-6
Industry-Funded Italian University Open Trial of Reboxetine Declares Proof It is Effective; Results Were Incredible: A 160-patient, 52-week trial of reboxetine in the elderly (65-94 years old) report 89% much improved by week six and 95% much improved by week 52. HAM-D scores decreased from 24 at baseline to 7 at 52 weeks. Side-effects caused 15% to drop-out, a normal rate of side-effects for an anti-depressant. The University of Trieste researchers declare, "Results from this study show reboxetine to be effective." Reboxetine in the maintenance therapy of depressive disorder in the elderly: a long-term open study. Aguglia E. Int J Geriatr Psychiatry. 2000 Sep;15(9):784-93. (Ed: While it is wonderful to see nice long follow-ups, this report is yet another example of the pharmaceutical industry ignoring the importance of controlled trials, knowing they can do better with open trials. Money clearly wasn't a motivator in not doing a controlled trial. If the results of this study seem too good to be true, they probably are. I have never seen any anti-depressant in a double-blind study do anywhere near this well. In contrast to these phenomenal rates of improvement, double-blind research on reboxetine has found reboxetine to be no more effective than imipramine, desipramine, fluoxetine, i.e. any other anti-depressant to which it has ever been compared as best I can tell.)
University of British Columbia Researchers Say Open Trials Suggest Usefulness and Provide Initial Evidence of Benefit: In an excellent review article of the studies which have been published on anti-convulsant medication for bipolar disorders, UBC researchers unfortunately use misleading language causing readers to think much more highly of open trials than they should. They write, "Open reports suggest usefulness of gabapentin as an adjunct in bipolar disorder." However, open trials can never suggest usefulness, because they cannot provide any evidence of efficacy. Favorable open trials have a notoriously bad track record. At most, open trials might "suggest the possibility of usefulness." The authors write that several anti-convulsants with favorable open trials "require confirmation of their efficacy from double-blind studies." This wording suggests that the open trials provide some initial evidence of benefit which is simply untrue. Third generation anticonvulsants in bipolar disorder: a review of efficacy and summary of clinical recommendations. Yatham LN, Kusumakar V, Calabrese JR, Rao R, Scarrow G, Kroeker G. J Clin Psychiatry. 2002 Apr;63(4):275-83. Ed: I have asked the authors for their thoughts on my criticism of their wording. 11/26/03
Harvard Psychiatrist Evins Claims Open Trials Give Promising Evidence of Efficacy: A Massachusetts General Hospital researcher in an excellent review of anti-convulsants for bipolar disorders uses misleading wording that I find very troubling. Evins notes that psychiatrists have been prescribing several new anti-convulsants without any double-blind research to support the practice, but he doesn't find fault with this practice. He purports to be reviewing the efficacy of the newer anti-convulsants, but uses wording that leads the reader to think that open trials provide significant evidence. "Evidence for the efficacy of topiramate in acute and refractory mania is promising." However, 100% of this so-called evidence is based on open trials. But, open trials cannot provide evidence. Their only purpose is in a run-up to a real double-blind scientific study. Evins notes that gabapentin is still being widely used for bipolar disorder even though is has now been proven worthless for bipolar disorder. He fails to note that there were even more open trials claiming gabapentin effective. He fails to note the sad state of affairs in the field of psychiatry exposing thousands of patients with a potentially deadly disease to unproven and ineffective treatments. Instead of protest, he ends his article with the usual platitude "More controlled trials are needed." Efficacy of newer anticonvulsant medications in bipolar spectrum mood disorders. Evins AE. J Clin Psychiatry. 2003;64 Suppl 8:9-14
Tiagabine Promoted by Open Trial for Bipolar Disorder: A small open trial 22 adult BPD patients who were inadequately improved had tiagabine added to their medications for at least 6 months. The authors say that 36% "responded" to the tiagabine. All of the other 64% non-responders stopped tiagabine because of side-effects. They claim that based on totally uncontrolled trial, "low-dose tiagabine appears to have mood-stabilizing and antimanic properties." Sacramento. An open case series on the utility of tiagabine as an augmentation in refractory bipolar outpatients. Schaffer LC, Schaffer CB, Howe J. J Affect Disord. 2002 Sep;71(1-3):259-63. Ed: It is incredible that the authors can assume that all of the patients who improved, improved because of tiagabine. Indeed, bipolar disorder has a natural tendency to wax and wane. Every placebo group I have ever seen has had 20% or more of patients who did well. How can authors claim to measure a response rate in an open trial? I asked Dr. Schaffer 11/26/03 via email. Hopefully, he will give us the answer.
University of Texas Psychiatrists Claims Efficacy of Tiagabine Can be Determined with an Open Trial: While this open trial of tiagabine for bipolar disorder wasn't very flattering to tiagabine in that only 23% of patients improved while on the medicine, the wording of the report is very troubling. Like numerous other psychiatrists, the University of Texas team claim that open trials can be used to determine efficacy. They write, "We sought to evaluate the effectiveness of tiagabine..." and go on to explain their open trial of 17 treatment refractory patients. They did note that tiagabine appeared to cause serious side-effects including two seizures. They still claimed that their 23% improved patients "demonstrated limited efficacy." Tiagabine in treatment refractory bipolar disorder: a clinical case series. Suppes T, Chisholm KA, Dhavale D, Frye MA, Altshuler LL, McElroy SL, Keck PE, Nolen WA, Kupka R, Denicoff KD, Leverich GS, Rush AJ, Post RM. Bipolar Disord. 2002 Oct;4(5):283-9. Ed: How UT Southwest reached a cause-effect conclusion of some "limited efficacy" from an open trial is beyond me. Might not some or all of the 23% improved patients been due to the natural progression of the disease or placebo effect? How can they say there is any efficacy there? I sent an email 11/26/03 and hope to share the answer with you.
Clinicians Urged to Consider Tiagabine Based on Only Three Patients in Open Trial by Robert Wood Johnson Psychiatrist: A report from a Robert Wood Johnson Medical School psychiatrist says that two bipolar and one schizoaffective bipolar type patients who had tiagabine added to their treatment regimen improved. Based on this, the author writes, "Clinicians are advised to consider this new anticonvulsant as a potential adjunctive agent in the treatment of bipolar and schizoaffective disorders." Adjunctive tiagabine treatment of psychiatric disorders: three cases. Kaufman KR. Ann Clin Psychiatry. 1998 Dec;10(4):181-4. Ed: Encouraging psychiatrists to start using a medication based on only three cases in a totally uncontrolled open trial appears to be poor advice and the very opposite of encouraging "evidence-based medicine." Perhaps I am wrong. I emailed Dr. Kaufman 11/27/03.
Keio University Tokyo Psychiatrists Claim Proof for Zonisamide for Mania Based on Small Open Trial: A report of the anti-convulsant zonisamide (Zonegran) used on 24 manic patients with diagnoses of bipolar, schizoaffective, or schizophrenic disorders claims that 25% of all and 33% of bipolars had remarkable improvement and 80% of bipolars had moderate improvement. This 1994 report has never been followed up with a double-blind study. The first open study of zonisamide, a novel anticonvulsant, shows efficacy in mania. Kanba S, Yagi G, Kamijima K, Suzuki T, Tajima O, Otaki J, Arata E, Koshikawa H, Nibuya M, Kinoshita N, et al. Prog Neuropsychopharmacol Biol Psychiatry. 1994 Jul;18(4):707-15
Harvard Psychiatrist and Reviewer Shares His Thoughts: In response to an email, Dr. Ron Pies of Mass General (His excellent review of anti-convulsants and bipolar disorder can be found in Ann Clin Psychiatry. 2002, Dec;14(4):223-32) wrote, "While I share many of the feelings and concerns you express, I also think the issues you raise need to be placed in the context of "medical realities"--i.e., that very often, physicians in general and psychiatrists in particular find themselves "up against the wall" with many very difficult patients. (I'm sure you know this well!). I believe that, in such circumstances, the "rules" about double-blind controlled studies must sometimes be "bent" a bit, in the service of the patient. Of course, in such cases, the threshold for meeting accepted standards of informed consent should and must be higher. In contrast, the cavalier use of unproven agents simply for the sake of "trying out" a new treatment is certainly to be deplored--especially when such behavior is heavily driven by pharmaceutical company "lobbying"." He notes that more than one DB PC study have now found topiramate to be ineffective.
Dr. Pies suggests the following model doctor-patient discussion before starting a medication with only open trial reports: "Mr. Jones, as you know, you have not responded well to the standard agents used to treat your condition. There is a new drug, called X, which has shown some promise in several case reports, and in two uncontrolled studies. It does not have labeling approved by the FDA for your condition. Nevertheless, it might be worth trying, based on what we know about the drug's mechanism of action and its overall safety profile. Here are some of the potential benefits, and here are some of the potential risks...[full discussion]...Do you understand what I've said? Do you have any questions or concerns?"
Ed: I disagreed with Dr. Pies and sent a second email because there has never been an investigation as to how many of the thousands of psychiatrists using gabapentin and topiramate were "up against the wall." It appears very likely that very few of these patients were tried on adequate trials lithium, carbamazepine, anti-psychotics, and lamotrigine, i.e. PC DB proven medications, before being given the experimental gabapentin or topiramate. Psychiatrists have largely abandoned lithium and carbamazepine because they don't have drug companies funding studies and salesmen promoting them. In over 15 years of very busy clinical practice, I cannot think of a single case where I used a treatment not supported by a double-blind study. I think the gabapentin-topiramate fiasco supports my concerns. I also don't agree with Dr. Pies doctor-patient discussion. No where does it say that there is no real evidence that medication X works at all. Remember, there were over 40 open trials claiming gabapentin helped bipolar disorder and now we know it was worthless! The rules should not be bent!
Wild Claims for Oxcarbazepine (Trileptal) from Italian University: Only 18 patients were studied in an open trial using oxcarbazepine as an add-on medication with no blinding of raters or control group. In the 12 month follow-up, seven patients did poorly and seven patients remained stable with four having mild to moderate mood difficulties. Authors make the wild claim that "our study suggests the potential usefulness of OXC as adjunctive therapy to lithium both in acute and long-term treatment of bipolar disorder. Oxcarbazepine as add-on treatment in patients with bipolar manic, mixed or depressive episode. Benedetti A, Lattanzi L, Pini S, Musetti L, Dell'Osso L, Cassano GB., University of Pisa. J Affect Disord. 2004 Apr;79(1-3):273-7. Ed: Trileptal is a old seizure medication recently resurrected and pushed for bipolar disorder without a single scientific study ever having being done so far as I can tell as of 4/7/04. A Harvard psychiatrist and the above University of Pisa psychiatrist have been making claims that it is effective based only on small open trials. There is no way to make any determination on effectiveness using an open trial.
Extremely Irresponsible Claims from Harvard Psychiatrist: In a small, open study of only 42 patients and using only a very crude measure of improvement, Clinical Global Impressions-Improvement scale, a Harvard psychiatrist claims to have proven that oxcarbazepine is quite effective at treating mixed symptoms of bipolar disorders. Some patients received only oxcarbazepine and other received additional psychiatric medications. There was no comparison group, no randomization, no blinding of raters or patients. Ghaemi claims, " Oxcarbazepine was moderately to markedly effective in 24 subjects (57%)." Sedation (17/42, 40%) was the most common side effect. Twenty-two patients (52%) stopped treatment, mostly due to side effects (12/22). Oxcarbazepine treatment of bipolar disorder. Ghaemi SN, Berv DA, Klugman J, Rosenquist KJ, Hsu DJ. J Clin Psychiatry. 2003 Aug;64(8):943-5. Ed: This study is nothing short of bizarre. Ghaemi deceives the reader into thinking that every patient who was reported to have improved while on oxcarbazepine, improved because of the medication. Yet, every DB PC study ever done of bipolars has found a significant number of patients improve on placebos. Of course, Ghaemi has worked closely with many pharmaceutical companies. It is just possible that some conscious or unconscious bias might also be present.
Thomas E. Radecki, M.D., J.D.