Letters

Use of hypericum as antidepressant

BMJ 2000; 320 doi: https://doi.org/10.1136/bmj.320.7242.1141/a (Published 22 April 2000) Cite this as: BMJ 2000;320:1141

Valid measure of antidepressant efficacy in primary care is needed

  1. Andrew Thornett, clinical research fellow (eanador{at}soton.ac.uk)
  1. Department of Psychiatry, University of Southampton, Southampton SO9 5NH
  2. Harvey Group Practice, St Albans AL3 5HB
  3. Integrative Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
  4. Arlington, Texas 76012-4704, USA
  5. Dunedin Hospital, Dunedin, New Zealand
  6. West Cumberland Hospital, Whitehaven, Cumbria CA28 8JG
  7. Bezirkskrankenhaus Landshut, D-84034 Landshut, Germany
  8. Imerem Institute for Medical Research Management and Biometrics, D-90478 Nuremberg, Germany
  9. Steiner Arzneimittel, D-12207 Berlin, Germany

    EDITOR—The study by Philipp et al comparing hypericum extract and imipramine or placebo seems to show that hypericum is as effective at treating moderate depression as imipramine.1 This impression is strengthened by the design of the study, which is double blind, randomised, and placebo controlled, using a widely used medication as a comparator and using globally accepted depression scales including the Hamilton depression score. Linde and Berner, however, in the accompanying commentary question the efficacy of hypericum because of its use in comparatively large doses and its comparison with low doses of standard antidepressants.1 They believe that these, together with the effect of unblinding on outcome, should be taken into account in the analysis of the results. The basis of this criticism is the lack of universal consensus on how the effects of antidepressant drugs should be measured in primary care. Difficulty arises because lower doses are often used to treat patients who may be less depressed than those seen in secondary care, and the treatments themselves may be more important as an adjunct to the interaction between doctor and patient than as a therapeutic intervention alone. Patients, too, may prefer to use treatment options that they see as more natural, including hypericum, and doctors are beginning to accept the importance of supporting patients' choices.2 Linde and Berne do not consider that most general practitioners use 20 mg of fluoxetine when treating depressed patients, and few are prepared to increase the dose beyond this level. This reluctance is less likely to be present with a treatment that is seen as part of alternative medicine and less likely to produce side effects.

    Further, in practice both patients and doctors know which medication has been prescribed, and the pragmatic nature of the trial conducted by Philipp et al reflects this, making the results more applicable to the situation in which the medication will be used. These difficulties in interpreting results of trials in general practice populations will be solved only when a primary care based system of measuring efficacy is developed that is relevant to this population and the treatment it receives.

    Footnotes

    • Competing interests None declared.

    References

    1. 1.
    2. 2.

    Naturalistic studies are needed

    1. Mike Walton, general practitioner (mike{at}walfri.u-net.com)
    1. Department of Psychiatry, University of Southampton, Southampton SO9 5NH
    2. Harvey Group Practice, St Albans AL3 5HB
    3. Integrative Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
    4. Arlington, Texas 76012-4704, USA
    5. Dunedin Hospital, Dunedin, New Zealand
    6. West Cumberland Hospital, Whitehaven, Cumbria CA28 8JG
    7. Bezirkskrankenhaus Landshut, D-84034 Landshut, Germany
    8. Imerem Institute for Medical Research Management and Biometrics, D-90478 Nuremberg, Germany
    9. Steiner Arzneimittel, D-12207 Berlin, Germany

      EDITOR—Philipp et al address an important question in attempting to test the effectiveness of hypericum extract in treating mild to moderate depression in a primary care setting.1 In the United Kingdom, general practitioners see and treat most depressive illness. Narrative experience of patients often illustrates that they have already tried, or intend to try, St John's wort or other products for their depression. For a general practitioner trying to practice evidence based mental health care, it is challenging to be able to answer the patient who asks whether St John's wort will work better than fluoxetine (or citalopram or whatever). Like many of my colleagues I welcome such research.

      I have concerns, however, over the ultimate applicability of research findings to the natural world of primary care. The authors say that, since hypericum products may vary considerably in composition, the results cannot be generalised to other extracts and also that the tested daily dosage of 1050 mg extract, which is equivalent to 6 g of the crude herb, is higher than that recommended. Given that patients who choose to treat their depression with hypericum products frequently buy the product directly over the counter, there is currently little control over or knowledge about the doses that are really taken by those who try this remedy. We need to establish the actual effective dose to be able to counsel our patients accordingly. Furthermore, the manufacturers and marketers who promote these various products should act in a responsible manner. Far too often the money spent on promoting a health product that is freely available over the counter is inversely proportional to the evidence of its efficacy. Let us hope for more quality research on the effectiveness of hypericum to enable us to give a clear message to our patients.

      Footnotes

      • Competing interests None declared.

      References

      1. 1.

      Use of placebo in depression trial was unethical

      1. Andrew Vickers, assistant attending research methodologist (vickersa{at}mskcc.org)
      1. Department of Psychiatry, University of Southampton, Southampton SO9 5NH
      2. Harvey Group Practice, St Albans AL3 5HB
      3. Integrative Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
      4. Arlington, Texas 76012-4704, USA
      5. Dunedin Hospital, Dunedin, New Zealand
      6. West Cumberland Hospital, Whitehaven, Cumbria CA28 8JG
      7. Bezirkskrankenhaus Landshut, D-84034 Landshut, Germany
      8. Imerem Institute for Medical Research Management and Biometrics, D-90478 Nuremberg, Germany
      9. Steiner Arzneimittel, D-12207 Berlin, Germany

        EDITOR—Although the trial on the use of hypericum extract for depression reported by Philipp et al is a valuable addition to the literature on St John's wort,1 I believe it to be unethical. The Declaration of Helsinki is quite explicit that patients in a clinical trial should be given the best proved medical care. In the reported trial, some patients with depression were denied effective treatment with antidepressant medication as a result of being allocated to placebo. According to the authors, one of these patients attempted to commit suicide. Ethics committees should not approve such trials, and patients should not consent to take part in them.

        Footnotes

        • Competing interests None declared.

        References

        1. 1.

        Active substances must be identified

        1. Tim Gorski, associate editor, The Scientific Review of Alternative Medicine (doctorg{at}writeme.com)
        1. Department of Psychiatry, University of Southampton, Southampton SO9 5NH
        2. Harvey Group Practice, St Albans AL3 5HB
        3. Integrative Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
        4. Arlington, Texas 76012-4704, USA
        5. Dunedin Hospital, Dunedin, New Zealand
        6. West Cumberland Hospital, Whitehaven, Cumbria CA28 8JG
        7. Bezirkskrankenhaus Landshut, D-84034 Landshut, Germany
        8. Imerem Institute for Medical Research Management and Biometrics, D-90478 Nuremberg, Germany
        9. Steiner Arzneimittel, D-12207 Berlin, Germany

          EDITOR—The report by Philipp et al provides additional helpful confirmation that St John's wort offers potential benefits in the treatment of depression.1 What is now needed, however, is not more clinical studies on raw extracts but isolation and characterisation of the active substance(s) and a determination of their likely pharmacological site and mechanism of action, their pharmacokinetics, and their metabolic fate. This has been done in the case of numerous other useful medications of plant derivation, from opium to salicylates. It is the only way to evaluate the potency, stability, duration of action, toxicity, and potential for drug interactions of the active substance(s), whatever it or they are. If the past is any guide, this is also the only way to further and deepen our understanding of human health and disease and to show the way to better therapeutic agents.

          Footnotes

          • Competing interests None declared.

          References

          1. 1.

          The herbalist will see you now

          1. David B Menkes, liaison psychiatrist (david.menkes{at}stonebow.otago.ac.nz)
          1. Department of Psychiatry, University of Southampton, Southampton SO9 5NH
          2. Harvey Group Practice, St Albans AL3 5HB
          3. Integrative Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
          4. Arlington, Texas 76012-4704, USA
          5. Dunedin Hospital, Dunedin, New Zealand
          6. West Cumberland Hospital, Whitehaven, Cumbria CA28 8JG
          7. Bezirkskrankenhaus Landshut, D-84034 Landshut, Germany
          8. Imerem Institute for Medical Research Management and Biometrics, D-90478 Nuremberg, Germany
          9. Steiner Arzneimittel, D-12207 Berlin, Germany

            EDITOR—Ambivalence among psychiatrists and general practitioners about herbal remedies will persist, despite the well designed trial of Philipp et al, which shows that hypericum (St John's wort) extract is comparable to imipramine in its effectiveness in treating depression.1 Complementary and conventional treatments in psychiatry do clash, particularly in the English speaking world, not least because of attitudes and prejudices held by practitioners.2 Nevertheless, we need to be familiar with such treatments because increasing numbers of our patients are attracted to them and may seek our advice. Illnesses such as serious depression are best treated by a competent doctor, irrespective of whether the treatment is conventional.

            Hypericum extracts are effective in mild to moderate depression3 and may also be effective in severe depression if titration above the usual dose is allowed. Their main advantage is their tolerability: side effects are even milder than those of the selective serotonin reuptake inhibitors, with notably less sexual dysfunction. Hypericum extracts seem to act like selective serotonin reuptake inhibitors in terms of possible interaction with monoamine oxidase inhibitors and (like most antidepressants) may provoke mania in patients with a bipolar tendency. It is important that appropriate cautions are given, in this case with regard to ingestion during pregnancy or in combination with monoamine oxidase inhibitors.

            Some of the scepticism about hypericum arises from the many different preparations available, and Philipp et al emphasise that their findings cannot be generalised to other extracts. Standardisation is attempted by the more reputable suppliers, generally with respect to content of hypericin and pseudohypericin (“total hypericins”). Hypericins are weak inhibitors of monoamine oxidase A but now seem to be less important to antidepressant action than other constituents of the herb, notably hyperforin.4 Consistent with the clinical profile of hypericum extracts, hyperforin enhances the synaptic availability of serotonin, as well as dopamine and noradrenaline.4 Hypericins and hyperforin also differ in their relative concentrations across the growth cycle of the herb, the latter being far more abundant towards the end of flowering.5 The conclusion is that current standardisation (with respect to hypericins) may correlate poorly with clinical potency (more likely due to hyperforin). This may explain the “modest” or “weak” results seen with some extracts and indicates the importance of further research and the urgent need for better standardisation. Steiner Arzneimittel (suppliers of the extract used by Philipp et al) specify the content of hyperforin as well as hypericin.

            The efficacy, tolerability, and popular appeal of hypericum pose a challenge to conventional medicine. Doctors require an up to date, working knowledge of effective herbal treatments—otherwise we will be eschewed by patients we could benefit.

            Footnotes

            • Competing interests None declared.

            References

            1. 1.
            2. 2.
            3. 3.
            4. 4.
            5. 5.

            Safety in overdose needs to be established

            1. Peter Balfour, self-constructing general practice trainee
            1. Department of Psychiatry, University of Southampton, Southampton SO9 5NH
            2. Harvey Group Practice, St Albans AL3 5HB
            3. Integrative Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
            4. Arlington, Texas 76012-4704, USA
            5. Dunedin Hospital, Dunedin, New Zealand
            6. West Cumberland Hospital, Whitehaven, Cumbria CA28 8JG
            7. Bezirkskrankenhaus Landshut, D-84034 Landshut, Germany
            8. Imerem Institute for Medical Research Management and Biometrics, D-90478 Nuremberg, Germany
            9. Steiner Arzneimittel, D-12207 Berlin, Germany

              EDITOR—The article by Philipp et al on the use of hypericum extract for depression and the accompanying commentary by Linde and Berner make it clear that the implications for the treatment of depression are serious,1 and I think that the main deciding factor in its use will be its safety in overdose. Has this been looked at yet? This has been one of the most important factors to recommend the selective serotonin reuptake inhibitors and that without this advantage, sales of drugs such as fluoxetine may be affected adversely. I just question the comment that 20 mg of fluoxetine is a “comparatively low dose”; rather, I would think of this as a standard dose for depression and certainly not enough to undermine the study mentioned.2 Finally, what is the current situation about prescribing extracts such as hypericum?

              References

              1. 1.
              2. 2.

              Authors' reply

              1. Michael Philipp, professor of psychiatry,
              2. Ralf Kohnen, professor of psychology,
              3. Karl-O Hiller, senior research fellow (kohiller{at}steiner-arzneimittel.de)
              1. Department of Psychiatry, University of Southampton, Southampton SO9 5NH
              2. Harvey Group Practice, St Albans AL3 5HB
              3. Integrative Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA
              4. Arlington, Texas 76012-4704, USA
              5. Dunedin Hospital, Dunedin, New Zealand
              6. West Cumberland Hospital, Whitehaven, Cumbria CA28 8JG
              7. Bezirkskrankenhaus Landshut, D-84034 Landshut, Germany
              8. Imerem Institute for Medical Research Management and Biometrics, D-90478 Nuremberg, Germany
              9. Steiner Arzneimittel, D-12207 Berlin, Germany

                EDITOR—We appreciate the positive feedback in most of the comments on our trial, which we would like to emphasise is not really comparable to other hypericum studies since the three arm method has been applied for the first time. The growing evidence on the effectiveness of some, but not all, hypericum preparations in mildly and moderately depressed patients is promising in many respects but contains risks when professional advice is not sought. The accumulating reports on possible interactions of hypericum preparations have to be taken into account seriously in further pharmacological, clinical, and epidemiological research.1

                The magnitude of the therapeutic effect of hypericum extract over placebo, which seemed to be questioned by Linde and Berner, corresponds to a Cohen's d (standardised mean difference) of 0.76(after six weeks) and 0.81 (after eight weeks of treatment).2 It is in line with effect sizes for placebo controlled hypericum studies (range of 0.5 to 1).3 Thus, the superiority of hypericum extract STEI 300 over placebo is impressive. Hypericum should be tested against newer antidepressants as well, but preferably in three arm trials.

                Thornett and also Walton correctly emphasise the need for more pragmatic or naturalistic trials like ours and point to the question of adequate evaluation of efficacy in trials with general practice populations. We believe that the undoubted merits of the HAMD scale should not hinder progress in developing a primary care based system of measuring antidepressive effects with other scales. By integrating self rating scales, our study approach already follows a promising alternative to psychiatric severity scales since we could show SDS and SF 36 scales to be sensitive for drug or placebo differences.

                The concern of Vickers about using a placebo group might have been triggered by our own misleading case report of a suicide attempt. The patient was suffering from an acute episode of a recurrent moderate depressive disorder and had been treated with imipramine before the study started. She was unfortunately randomised to the placebo group. After some improvement during the first weeks her status deteriorated. Around week seven she urged the investigator to open the emergency envelope since she was having suicidal thoughts and wanted to find out whether an effective treatment was being administered. The patient then was immediately treated with imipramine. No actual suicide attempt took place. We believe that control of suicidal risk in our study was efficient and the positive decision of the ethics committees was justified on these grounds. In most countries, regulatory bodies give approval to a drug as an antidepressant only when it has have been tested in placebo controlled studies.

                The proposal by Gorski to isolate and characterise the active substance(s) of St. John's wort is theoretically a straightforward strategy. We know from previous pharmacological work, however, that the picture is more complex.4 Extracts of St John's wort contain several pharmacologically relevant substances (besides hypericin and hyperforin), and it has been shown that substances like procyanidins may contribute indirectly to the antidepressant effects.5 Therefore we should concentrate our efforts towards the definition of well accepted ranges for all biologically relevant substances in St John's wort and study the potency, stability, duration of action, toxicity, and potential for drug interactions of such well characterised extracts. We hope that improved standardisation will lead to extracts with an even better relation between risks and benefits.

                References

                1. 1.
                2. 2.
                3. 3.
                4. 4.
                5. 5.
                View Abstract