Efficacy and tolerability of Hypericum extract for the ... - Livar.net
Efficacy and tolerability of Hypericum extract for the ... - Livar.net
Efficacy and tolerability of Hypericum extract for the ... - Livar.net
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
European Neuropsychopharmacology (2010) 20, 747–765<br />
REVIEW<br />
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong><br />
treatment <strong>of</strong> mild to moderate depression<br />
Siegfried Kasper a,1 , Filippo Caraci b,1 , Bruno Forti c ,<br />
Filippo Drago b, ⁎, Eugenio Aguglia d<br />
a<br />
Department <strong>of</strong> General Psychiatry, Medical University <strong>of</strong> Wien, MUV, Austria<br />
b<br />
Department <strong>of</strong> Experimental <strong>and</strong> Clinical Pharmacology, University <strong>of</strong> Catania Medical School, Catania, Italy<br />
c<br />
CSM, Belluno, Italy<br />
d<br />
Department <strong>of</strong> Clinical Psychiatry, University <strong>of</strong> Catania School <strong>of</strong> Medicine, Catania, Italy<br />
Received 13 May 2010; received in revised <strong>for</strong>m 13 July 2010; accepted 20 July 2010<br />
KEYWORDS<br />
Depression;<br />
<strong>Hypericum</strong>;<br />
St. John's wort;<br />
<strong>Efficacy</strong>;<br />
Review<br />
1. Introduction<br />
Abstract<br />
Depression is a common condition in <strong>the</strong> general community<br />
<strong>and</strong> <strong>the</strong> most prevalent <strong>for</strong>m <strong>of</strong> psychiatric morbidity,<br />
⁎ Corresponding author. Department <strong>of</strong> Experimental <strong>and</strong> Clinical<br />
Pharmacology, University <strong>of</strong> Catania Medical School, Viale A. Doria 6,<br />
95125 Catania, Italy. Tel.: +39 095 7384236; fax: +39 095 7384238.<br />
E-mail address: f.drago@unict.it (F. Drago).<br />
1 These authors equally contributed to this work.<br />
www.elsevier.com/locate/euroneuro<br />
Depression is a common condition in <strong>the</strong> community with a significant impact on affected<br />
individuals, <strong>the</strong>ir relatives <strong>and</strong> society. Many patients with depression do not seek treatment <strong>and</strong><br />
are <strong>of</strong>ten concerned about <strong>the</strong> possible adverse effects <strong>of</strong> antidepressant drugs. Extract <strong>of</strong><br />
<strong>Hypericum</strong> per<strong>for</strong>atum (St. John's wort) has long been recognized as a treatment <strong>for</strong> depression.<br />
Several published trials <strong>and</strong> meta-analyses have demonstrated <strong>the</strong> efficacy <strong>and</strong> <strong>tolerability</strong> <strong>of</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> mild to moderate depression. Recent comparative trials <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> <strong>and</strong> o<strong>the</strong>r antidepressants, including selective serotonin reuptake inhibitors (SSRIs),<br />
provide support <strong>for</strong> <strong>Hypericum</strong> <strong>extract</strong> efficacy. However, since <strong>the</strong> constituents <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> differ between <strong>the</strong> individual manufacturers, <strong>the</strong> efficacy cannot be extrapolated from<br />
one <strong>extract</strong> to ano<strong>the</strong>r. In this review, WS 5572, LI 160, WS 5570 <strong>and</strong> ZE 117 <strong>Hypericum</strong> <strong>extract</strong>s<br />
have been shown to be significantly more effective than placebo with at least similar efficacy <strong>and</strong><br />
better <strong>tolerability</strong> compared to st<strong>and</strong>ard antidepressant drugs.<br />
© 2010 Elsevier B.V. <strong>and</strong> ECNP. All rights reserved.<br />
0924-977X/$ - see front matter © 2010 Elsevier B.V. <strong>and</strong> ECNP. All rights reserved.<br />
doi:10.1016/j.euroneuro.2010.07.005<br />
although figures depend on <strong>the</strong> population studied <strong>and</strong><br />
diagnostic criteria used. Studies that employ Diagnostic <strong>and</strong><br />
Statistical Manual <strong>of</strong> Mental Disorders, Fourth Edition (DSM-<br />
IV) criteria <strong>of</strong> major depression provide lower prevalence<br />
estimates than those that allow more self-reporting <strong>of</strong><br />
depressive symptoms <strong>and</strong> less stringent diagnostic criteria<br />
(Battaglia et al., 2004; Mulsant <strong>and</strong> Ganguli, 1999; Olsen et<br />
al., 2004; Wilhelm et al., 2003). A large-scale survey <strong>of</strong> 5566<br />
individuals assessed using <strong>the</strong> modified-Mini-International<br />
Neuropsychiatric Interview (MINI) showed that <strong>the</strong> prevalence<br />
<strong>of</strong> major depression in Italy was 10.8% (Battaglia et al.,
748 S. Kasper et al.<br />
2004), a figure that corresponds with those from o<strong>the</strong>r<br />
epidemiological studies (Mulsant <strong>and</strong> Ganguli, 1999). However,<br />
<strong>the</strong> actual impact <strong>of</strong> such depressive disturbances is<br />
likely to be underestimated as patients may not be<br />
sufficiently severe to meet st<strong>and</strong>ard diagnostic criteria<br />
(Aguglia <strong>and</strong> Forti, 2001).<br />
One <strong>of</strong> <strong>the</strong> most important factors in treatment <strong>of</strong> mild to<br />
moderate depression is <strong>the</strong> <strong>tolerability</strong> <strong>of</strong> pharmacological<br />
<strong>the</strong>rapy, since <strong>the</strong> majority <strong>of</strong> patients have to continue<br />
normal work activities (Sommer <strong>and</strong> Harrer, 1994). Side<br />
effects <strong>of</strong> psychoactive drugs, such as dependence <strong>and</strong> mood<br />
alterations, are <strong>of</strong>ten feared by patients, which is thought to<br />
lead to an overall low rate <strong>of</strong> acceptance <strong>of</strong> st<strong>and</strong>ard<br />
pharmacological treatment (Röder et al., 2004). Moreover,<br />
<strong>the</strong> presence <strong>of</strong> somatic symptoms in mild <strong>for</strong>ms <strong>of</strong><br />
depression may cause fur<strong>the</strong>r problems with compliance<br />
because <strong>of</strong> <strong>the</strong> side effects <strong>of</strong> antidepressant <strong>the</strong>rapy<br />
(Vorbach et al., 1994). Thus, treatment <strong>of</strong> patients with<br />
mild to moderate depression represents a particularly<br />
clinically challenging problem.<br />
Extracts <strong>of</strong> <strong>Hypericum</strong> per<strong>for</strong>atum, commonly called St.<br />
John's wort, have been used <strong>for</strong> centuries in herbal medicine<br />
<strong>and</strong> have been clinically studied since <strong>the</strong> early 1990s (Röder<br />
et al., 2004). Drugs based on <strong>Hypericum</strong> <strong>extract</strong> are widely<br />
employed in Europe <strong>and</strong> are gaining popularity in <strong>the</strong> United<br />
States (Kasper <strong>and</strong> Dienel, 2002; Lecrubier et al., 2002).<br />
<strong>Hypericum</strong> <strong>extract</strong> contains at least 10 active constituents<br />
that may contribute to its pharmacological effects (Wagner<br />
<strong>and</strong> Bladt, 1994). Components known, or suspected, to play a<br />
role in antidepressant activity include phloroglucines (e.g.<br />
hyper<strong>for</strong>in), naphthodianthrones (e.g. hypericin) <strong>and</strong> <strong>the</strong><br />
flavonoids (e.g. quercitrin). Although <strong>the</strong> labeling <strong>of</strong> <strong>extract</strong>s<br />
indicates that several <strong>of</strong> <strong>the</strong> products studied should be<br />
pharmaceutically equivalent, dissolution under biorelevant<br />
conditions revealed that <strong>the</strong>y have quite different release<br />
pr<strong>of</strong>iles<strong>and</strong>cannotbeconsideredswitchable(Westerh<strong>of</strong>f et<br />
al., 2002). Clinical evaluation <strong>and</strong> comparison <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong>s is problematic due to <strong>the</strong> differences in study design,<br />
inclusion criteria <strong>and</strong> <strong>extract</strong> used <strong>and</strong>, in many earlier<br />
studies, <strong>the</strong> <strong>Hypericum</strong> preparations are not adequately<br />
characterized. In some studies, <strong>the</strong> preparations are st<strong>and</strong>ardized<br />
by ei<strong>the</strong>r <strong>the</strong>ir hypericin or hyper<strong>for</strong>in content or,<br />
occasionally, both. Although <strong>the</strong>se individual substances alone<br />
have been shown to have antidepressant activity, <strong>the</strong> total<br />
<strong>extract</strong> appears to be more effective (Reichling et al., 2003).<br />
Some more recent studies use several active components or<br />
<strong>the</strong> whole <strong>extract</strong> to characterize a preparation.<br />
The mechanisms <strong>of</strong> action <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> are not<br />
fully understood (Butterweck, 2003), although some authors<br />
have hypo<strong>the</strong>sized that hyper<strong>for</strong>in is <strong>the</strong> main active principle<br />
responsible <strong>for</strong> <strong>the</strong> antidepressant activity (Mennini <strong>and</strong><br />
Gobbi, 2004). The antidepressant activity <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> has been demonstrated in different animal models<br />
<strong>of</strong> depression (Butterweck, 2003). <strong>Hypericum</strong> <strong>extract</strong>s can<br />
inhibit <strong>the</strong> synaptosomal uptake <strong>of</strong> noradrenaline, serotonin<br />
<strong>and</strong> GABA in a concentration-dependent manner (Butterweck,<br />
2003; Schrader, 2000). In particular it has been hypo<strong>the</strong>sized<br />
that non-selective inhibition <strong>of</strong> monoamines occurs in part<br />
via modulation <strong>of</strong> Na + gradient membranes, with <strong>Hypericum</strong><br />
<strong>extract</strong> causing sodium influx into <strong>the</strong> neuron, which finally<br />
leads to <strong>the</strong> release <strong>of</strong> intracellular calcium (Singer et al.,<br />
1999; Marsh <strong>and</strong> Davies, 2002). In vivo studies have demon-<br />
strated that <strong>Hypericum</strong> <strong>extract</strong>s lead to a downregulation<br />
<strong>of</strong> β-adrenergic receptors (Butterweck, 2003), a common<br />
biochemical marker <strong>of</strong> antidepressant efficacy (Simbrey<br />
et al., 2004). Fur<strong>the</strong>rmore <strong>Hypericum</strong> <strong>extract</strong> is known<br />
to increase dopaminergic activity in <strong>the</strong> prefrontal cortex<br />
(Yoshitake et al., 2004).<br />
Stress <strong>and</strong> hypothalamic–pituitary–adrenal (HPA) axis<br />
abnormalities have a central role in <strong>the</strong> pathogenesis <strong>of</strong><br />
depression, with an impaired negative feed-back <strong>of</strong> glucocorticoids<br />
on <strong>the</strong> activity <strong>of</strong> <strong>the</strong> HPA axis, which results in<br />
elevated cortisol levels (Krishnan <strong>and</strong> Nestler, 2008). H.<br />
per<strong>for</strong>atum significantly attenuated restraint stress-induced<br />
increases in plasma ACTH <strong>and</strong> corticosterone levels (Kumar<br />
et al., 2010; Grundmann et al., 2010; Butterweck <strong>and</strong><br />
Schmidt, 2007). Studies in healthy male volunteers have<br />
found that <strong>Hypericum</strong> <strong>extract</strong>s can modulate salivary <strong>and</strong><br />
serum cortisol levels (Franklin et al., 2006; Schule et al.,<br />
2001). Fur<strong>the</strong>r studies are needed to better underst<strong>and</strong> <strong>the</strong><br />
molecular mechanisms underlying <strong>the</strong> antidepressant activity<br />
<strong>of</strong> H. per<strong>for</strong>atum <strong>and</strong> <strong>the</strong> specific effects <strong>of</strong> <strong>the</strong> different<br />
<strong>Hypericum</strong> preparations.<br />
Much <strong>of</strong> <strong>the</strong> heterogeneity seen in <strong>Hypericum</strong> clinical trial<br />
data can be explained by <strong>the</strong> differences in <strong>extract</strong><br />
preparation. The aim <strong>of</strong> this review is to examine clinical<br />
evidence related to <strong>the</strong> efficacy <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> available<br />
<strong>Hypericum</strong> <strong>extract</strong>s <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> depression <strong>of</strong> mild<br />
to moderate severity, <strong>and</strong> to compare <strong>the</strong> efficacy <strong>and</strong><br />
<strong>tolerability</strong> <strong>of</strong> specific preparations <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong><br />
with st<strong>and</strong>ard antidepressant drugs <strong>and</strong> allows individual<br />
<strong>extract</strong>s to be evaluated on a peer-<strong>extract</strong> basis. We will<br />
focus our attention on patients with mild to moderate<br />
depression which is <strong>the</strong> population most likely to benefit<br />
from <strong>Hypericum</strong> <strong>extract</strong> <strong>the</strong>rapy due to <strong>the</strong> favourable<br />
<strong>tolerability</strong> pr<strong>of</strong>ile. Relevant studies were selected from<br />
literature sourced using comprehensive Medline searches <strong>for</strong><br />
articles on <strong>Hypericum</strong> <strong>extract</strong> from 1993 to <strong>the</strong> present using<br />
<strong>the</strong> following search terms: <strong>Hypericum</strong>, St. John's wort,<br />
Johanniskraut, Johanniskrautextrakt, <strong>and</strong> depression. Additional<br />
manual searches were also per<strong>for</strong>med in <strong>the</strong> reference<br />
lists <strong>of</strong> published meta-analyses <strong>and</strong> r<strong>and</strong>omized controlled<br />
trials identified in <strong>the</strong> literature search. Most active control<br />
studies in this review were designed to determine noninferiority<br />
with st<strong>and</strong>ard antidepressant treatment in terms<br />
<strong>of</strong> efficacy <strong>and</strong> to examine <strong>the</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> in comparison to st<strong>and</strong>ard antidepressant drugs.<br />
2. Clinical evidence <strong>for</strong> efficacy <strong>and</strong> <strong>tolerability</strong><br />
<strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong><br />
The clinical studies were selected <strong>for</strong> this review on <strong>the</strong> basis<br />
<strong>of</strong> <strong>the</strong> patients' baseline clinical status. Only studies <strong>of</strong><br />
patients with mild to moderate depression were included.<br />
Mild to moderate depression is defined by a baseline<br />
Hamilton Depression Rating Scale (HAMD) score <strong>of</strong> 16–24 or<br />
mean HAMD score <strong>of</strong> 16–24 or by The Tenth Revision <strong>of</strong> <strong>the</strong><br />
International Classification <strong>of</strong> Diseases <strong>and</strong> Related Health<br />
Problems (ICD-10), F32.0, F32.1, F33.0 <strong>and</strong> F33.1.<br />
Of <strong>the</strong> various <strong>extract</strong>s available, LI 160, with a hypericin<br />
content <strong>of</strong> 0.72–0.96 mg in a 900 mg dose, has been <strong>the</strong> most<br />
frequently studied. O<strong>the</strong>r studies have used <strong>extract</strong>s WS<br />
5570, WS 5572, WS 5573, STEI300, ZE 117, psychotonin,
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
esbericum, LoHyp-57, calmigen, STW3, <strong>and</strong> an unnamed<br />
<strong>extract</strong> from <strong>the</strong> Swiss Herbal Remedies Company, which<br />
have differing hypericin <strong>and</strong> hyper<strong>for</strong>in contents (Table 1).<br />
Clinical studies have been reviewed by <strong>the</strong> specific preparation<br />
<strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> used, to determine whe<strong>the</strong>r<br />
some <strong>extract</strong>s might have a greater antidepressant effect<br />
than o<strong>the</strong>rs.<br />
2.1. <strong>Hypericum</strong> <strong>extract</strong> vs. placebo<br />
Numerous studies have compared <strong>the</strong> efficacy <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> with placebo in patients with major depression<br />
(Hänsgen <strong>and</strong> Vesper, 1996; Hübner et al., 1994; <strong>Hypericum</strong><br />
Depression Trial Study Group, 2002; Kalb et al., 2001;<br />
Table 1 Preparation details <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>s used in clinical trials.<br />
Extract Daily dose (mg) St<strong>and</strong>ardized<br />
hypericin content<br />
LI 160 900 (3×300) (Bjerkenstedt<br />
et al., 2005; Brenner et al.,<br />
2000; Hänsgen <strong>and</strong> Vesper,<br />
1996; Harrer et al., 1994;<br />
Hubner et al., 1994;<br />
<strong>Hypericum</strong> Depression Trial<br />
Study Group, 2002; Shelton<br />
et al., 2001; Sommer <strong>and</strong><br />
Harrer, 1994)<br />
WS 5570 900 (3×300) (Kasper et al.,<br />
2004; Lecrubier et al., 2002)<br />
WS 5572 900 (3×300 (Kalb et al., 2001;<br />
Laakmann et al., 1998) or<br />
3×600 (Szegedi et al., 2005))<br />
0.12–0.28% (<strong>Hypericum</strong><br />
Depression Trial Study<br />
Group, 2002; Lecrubier<br />
et al., 2002) (0.72–0.96 mg<br />
in 900 mg/day dose)<br />
(Wheatley, 1997)<br />
0.12–0.28% (Lecrubier<br />
et al., 2002; Szegedi<br />
et al., 2005)<br />
Laakmann et al., 1998; Lecrubier et al., 2002; Lehrl <strong>and</strong><br />
Woelk, 1993; Philipp et al., 1999; Schmidt <strong>and</strong> Sommer,<br />
1993; Schrader et al., 1998; Shelton et al., 2001; Sommer<br />
<strong>and</strong> Harrer, 1994; Witte et al., 1995; Kasper et al., 2006,<br />
2008a,b). A summary <strong>of</strong> <strong>the</strong>se studies including responder<br />
rates is shown in Table 2. In <strong>the</strong> majority <strong>of</strong> studies, <strong>the</strong><br />
responder rate is defined as <strong>the</strong> proportion <strong>of</strong> patients with a<br />
HAMD score <strong>of</strong> b10 at endpoint or a ≥50% reduction in HAMD<br />
score from baseline to endpoint. Studies comparing LI 160<br />
<strong>and</strong> placebo, all conducted in Germany, showed a significantly<br />
greater response with LI 160 vs. placebo. Two studies,<br />
both conducted in <strong>the</strong> US (<strong>Hypericum</strong> Depression Trial Study<br />
Group, 2002; Shelton et al., 2001), showed no significant<br />
difference between LI 160 <strong>and</strong> placebo. However, <strong>the</strong><br />
greater severity <strong>of</strong> depression <strong>of</strong> <strong>the</strong> patients included in<br />
St<strong>and</strong>ardized<br />
hyper<strong>for</strong>in content<br />
Preparation method<br />
Not stated Extracted by methanol<br />
80% w/w (drug to <strong>extract</strong><br />
ratio 4–7:1) (Bjerkenstedt<br />
et al., 2005)<br />
3–6% (Lecrubier<br />
et al., 2002;<br />
Szegedi et al.,<br />
2005)<br />
Not stated 5% (Laakmann<br />
et al., 1998)<br />
Extracted from Herba<br />
hyperici (drug to <strong>extract</strong><br />
ratio 3–7:1) (Lecrubier<br />
et al., 2002; Szegedi<br />
et al., 2005)<br />
Not stated<br />
WS 5573 900 (3×300) Not stated 0.5% Not stated<br />
Psychotonin f 200–240 (Witte et al., 1995) Not stated Not stated Not stated<br />
ZE 117 500 (2×250) (Schrader, 2000; 0.2% (Woelk, 2000) or 0.5% Not stated Extracted by ethanol 50%<br />
Schrader et al., 1998; Woelk, (Schrader et al., 1998)<br />
w/w (drug to <strong>extract</strong> ratio<br />
2000)<br />
4–7:1) (Schrader, 2000;<br />
Woelk, 2000)<br />
STEI300 1050 (3 ×350) (Philipp 0.2–0.3% (Philipp et al., 2–3% (Philipp Extracted by ethanol 60%<br />
et al., 1999)<br />
1999)<br />
et al., 1999) w/w (Philipp et al., 1999)<br />
Esbericum 75–300 (1–4×75) (Shaper <strong>and</strong> Not stated Not stated Extracted by ethanol 60%<br />
Brummer, 2005)<br />
w/w (drug to <strong>extract</strong> ratio<br />
2.0–5.5:1) (Shaper <strong>and</strong><br />
Brummer, 2005)<br />
LoHyp-57 800 (4×200) (Harrer et al., Not stated Not stated Extracted by ethanol 60%<br />
1999)<br />
w/w (drug to <strong>extract</strong> ratio<br />
5–7:1)<br />
Calmigen 300 (2×150) (Behnke<br />
et al., 2002)<br />
Not stated Not stated Not stated<br />
Unnamed <strong>extract</strong> 900 (3×300) (van Gurp 0.3% hypericin (note: <strong>the</strong> Not stated Not stated<br />
from Swiss et al., 2002)<br />
paper states “<strong>Hypericum</strong><br />
Herbal<br />
content <strong>of</strong> 0.3%”) (van Gurp<br />
Remedies<br />
et al., 2002)<br />
STW3/Laif 600 612 (Gastpar et al., 2005) Not stated Not stated Not stated<br />
749
Table 2 Overview <strong>of</strong> design <strong>and</strong> response data from comparative studies <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> depression.<br />
Study Design Disease severity at baseline <strong>Hypericum</strong> <strong>extract</strong> (mg/day) Comparator Population<br />
(mg/day)<br />
a N Responders n/N (%)<br />
<strong>Hypericum</strong> Comparator<br />
Placebo as comparator<br />
(Lehrl <strong>and</strong> Woelk, 1993) DB, R HAMD mean 22.7 LI 160 (900) Placebo ITT/EP 50 4/25 (16.0) b<br />
2/25 (8.0)<br />
(Schmidt <strong>and</strong> Sommer, DB, R Mild to moderate LI 160 (dose not<br />
Placebo PP 60 (66.6) (26.7)<br />
1993)<br />
stated in abstract)<br />
(Hubner et al., 1994) DB, R, Mild: ICD-9 300.4,<br />
LI 160 (900) Placebo ITT 39 14/20 (70) 9/19 (47)<br />
SC 309.0; mean HAMD ~12<br />
(Sommer <strong>and</strong> Harrer, DB, R, Mild: ICD-9 300.4 <strong>and</strong><br />
LI 160 (900) Placebo PP 89<br />
1994)<br />
MC 309.0; mean HAMD ~16<br />
c<br />
28/42 (67) 13/47 (28)<br />
(Hänsgen <strong>and</strong> Vesper, DB, R Moderate: HAMD<br />
LI 160 (900) Placebo EP 107 35/53 (66.0)<br />
1996)<br />
mean 20.7<br />
c<br />
12/54 (22.2)<br />
(Shelton et al., 2001) DB, R, Mild to moderate: HAMD LI 160 (900–1200) Placebo ITT 200 (all) 26/98 (26.5) 19/102 (18.6)<br />
MC 20–22<br />
109 (mild to<br />
moderate)<br />
20/59 (33.9) 11/50 (22)<br />
(<strong>Hypericum</strong> Depression DB, R, Mild, moderate <strong>and</strong> severe: LI 160 (900) Placebo EP 229 43/113 (38.1) 50/116 (43.1)<br />
Trial Study Group, 2002) MC HAMD ≥20, mean ~23 (77.6%<br />
mild to moderate, data are<br />
shown <strong>for</strong> <strong>the</strong> whole cohort<br />
since separate data <strong>for</strong> <strong>the</strong><br />
mild to moderate subgroup<br />
were not provided)<br />
PP 166 39/82 (47.6) 47/84 (56)<br />
(Lecrubier et al., 2002) DB, R, HAMD 18–25, mean<br />
WS 5570 (900) Placebo ITT/EP 375<br />
MC HAMD 21.9<br />
e<br />
98/186 (52.7) 80/189 (42.3)<br />
(Laakmann et al., 1998) DB, R, Mild to moderate: mean WS 5572 (900)<br />
Placebo ITT 98 24/49 (49.0)<br />
MC HAMD ~20–21<br />
(WS 5572 only; not<br />
including <strong>the</strong> WS<br />
5573 group <strong>of</strong> <strong>the</strong><br />
study, N=49)<br />
c<br />
16/49 (32.7)<br />
(Kalb et al., 2001) DB, R, Mild to moderate: mean WS 5572 (900) Placebo ITT 72 23/37 (62.1)<br />
MC HAMD ~20<br />
c<br />
15/35 (42.9)<br />
(Witte et al., 1995) DB, R Moderate: ICD-10 F32.1; Mean<br />
HAMD 23.6<br />
Psychotonin f. (200–240) Placebo EP 97 34/48 (70.8) c<br />
25/49 (51.0)<br />
(Schrader et al., 1998) DB, R, Mild to moderate: ICD-10 F32.0, ZE 117 (500) Placebo ITT 159<br />
MC F32.1; HAMD 16–24, mean ~20<br />
e<br />
45/80 (56.3) 12/79 (15.2)<br />
(Philipp et al., 1999) DB, R, Moderate: ICD-10 F32.1, F33.1 STEI 300 (1050) Placebo ITT 146 67/100 (67) 22/46 (47.8)<br />
MC<br />
EP 153 67/106 (63.2) 22/47 (46.8)<br />
(Kasper et al., 2006) DB, R Mild or moderate major WS 5570 (600–1200) Placebo PP 324 159/243 26/81 (31.1%)<br />
MC depressive episode:<br />
mean HAMD 22.8<br />
(65.4%)<br />
750 S. Kasper et al.
Placebo as comparator in long-term trial<br />
(Kasper et al., 2008b) DB, R<br />
MC<br />
Tricyclic as comparator<br />
(Vorbach et al., 1994) DB, R,<br />
MC<br />
(Harrer et al., 1994) DB, R,<br />
MC<br />
(Wheatley, 1997) DB, R,<br />
MC<br />
Recurrent episode <strong>of</strong> moderate<br />
major depression HAMD ≥20,<br />
mean ~23<br />
Depressive disorder according<br />
to DSM-III-R; mean HAMD ~20<br />
Moderate: ICD-10 F32.1; HAMD<br />
16–24, mean HAMD ~21<br />
(Bergmann et al., 1993) DB Mild to moderate: mean<br />
HAMD 15.6<br />
(Woelk, 2000) DB, R, Mild to moderate: mean<br />
MC HAMD 22.4 (mild N=189;<br />
moderate N=135)<br />
(Philipp et al., 1999) DB, R, Moderate: ICD-10 F32.1<br />
MC <strong>and</strong> F33.1<br />
SSRI as comparator<br />
(Brenner et al., 2000) DB, R Mild to moderate: HAMD ≥17,<br />
mean HAMD ~21<br />
(<strong>Hypericum</strong> Depression<br />
Trial Study Group, 2002)<br />
DB, R,<br />
MC<br />
(Harrer et al., 1999) DB, R,<br />
MC<br />
(Schrader, 2000) DB, R,<br />
MC<br />
WS 5570 (900) Placebo PP 426 Relapse rates<br />
51/282 (18.1%)<br />
LI 160 (900) Imipramine<br />
(75)<br />
LI 160 (900) Maprotiline<br />
(75)<br />
Mild to moderate: HAMD 17–24 LI 160 (900) Amitriptyline<br />
(75)<br />
Mild, moderate <strong>and</strong> severe:<br />
HAMD ≥20, mean ~23 (77.6%<br />
mild to moderate; data are<br />
shown <strong>for</strong> <strong>the</strong> whole cohort<br />
since separate data <strong>for</strong> <strong>the</strong> mild<br />
to moderate subgroup were not<br />
provided)<br />
Mild to moderate: ICD-10 F32.0<br />
<strong>and</strong> F32.1; mean HAMD ~17<br />
(mild N=72, moderate N=77)<br />
Mild to moderate: HAMD 16–24,<br />
mean HAMD ~20<br />
Esbericum (dose not stated) Amitriptyline<br />
(30)<br />
ZE 117 (500) Imipramine<br />
(150)<br />
STEI 300 (1050) Imipramine<br />
(100)<br />
LI 160 (900) Sertraline<br />
(75)<br />
LI 160 (900) Sertraline<br />
(50)<br />
LoHyp-57 (800) Fluoxetine<br />
(20)<br />
ZE 117 (500) Fluoxetine<br />
(20)<br />
Relapse rates<br />
37/144 (25.7%)<br />
ITT 135 42/67 (62.7) 37/68 (54.4)<br />
PP 130 54/66 (81.8) 40/64 (62.5)<br />
EP 102 27/51 (53) 28/51 (55)<br />
PP 86 27/44 (61) 28/42 (67)<br />
ITT 156 40/87 (46) 42/78 (53.9)<br />
PP 121 40/67 (59.7) 42/54 (77.8)<br />
EP 80 32/40 (80) c<br />
28/40 (70)<br />
EP 324 e<br />
68/157 (43.2) 67/167 (40.1)<br />
ITT 205 67/100 (67) 66/105 (62.9)<br />
EP 216 76/106 (71.7) 70/110 (63.6)<br />
ITT 28 7/13 (54) 6/15 (40)<br />
EP 30 7/15 (46.7) 6/15 (40)<br />
ITT 222 43/113 (38.1) 53/109 (48.6)<br />
EP 224 46/113 (40.7) 55/111 (49.5)<br />
PP 161 39/82 (47.6) 48/79 (61)<br />
ITT 149 (all) 50/70 (71.4) 57/79 (72.2)<br />
72 (mild) 27/33 (81.8) 30/39 (76.9)<br />
77 (moderate) 23/37 (62.2) 27/40 (67.5)<br />
EP 161 (all) 50/77 (64.9) 57/84 (67.8)<br />
ITT 240 f<br />
75/126 (60) 45/114 (40)<br />
(continued on next page)<br />
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
751
Table 2 (continued)<br />
Study Design Disease severity at baseline <strong>Hypericum</strong> <strong>extract</strong> (mg/day) Comparator<br />
(mg/day)<br />
SSRI as Comparator<br />
(Behnke et al., 2002) DB, R,<br />
MC<br />
(van Gurp et al., 2002) DB, R,<br />
MC<br />
(Gastpar et al., 2005) DB, R,<br />
MC<br />
(Szegedi et al., 2005) DB, R,<br />
MC<br />
Mild to moderate: mean HAMD Calmigen (300) Fluoxetine<br />
~20–21<br />
(40)<br />
HAMD ≥16, mean HAMD ~19 Swiss Herbal Remedies (900– Sertraline<br />
1800)<br />
(50–100)<br />
Moderate: ICD-10 F32.1, F33.1; STW3/Laif 600 (612) Sertraline<br />
HAMD 20–24, mean HAMD ~22<br />
(50)<br />
Moderate to severe: HAMD ≥22,<br />
mean HAMD 25.5 (44% had<br />
moderate depression, HAMD<br />
22–24 but data are shown <strong>for</strong><br />
<strong>the</strong> whole cohort since separate<br />
data <strong>for</strong> <strong>the</strong> moderate subgroup<br />
were not provided)<br />
(Bjerkenstedt et al., 2005) DB, R, Mild to moderate major<br />
depressive disorder DSM-IV<br />
Mean HAMD ~24–25<br />
(Fava et al., 2005) DB, R,<br />
MC<br />
Mild to moderate major<br />
depressive disorder DSM-IV<br />
HAMD score (17 items) Mean<br />
HAMD ~22<br />
(Moreno et al., 2006) DB, R, Mild to moderate major<br />
depressive disorder DSM-IV<br />
HAMD score 14–17<br />
(Gastpar et al., 2006) DB, R,<br />
MC<br />
Moderate depression: ICD-10<br />
F32.1, F33.1 mean HAMD ~22<br />
WS 5570 (900–1800) Paroxetine<br />
(20–40)<br />
LI 160 (900) Fluoxetine<br />
(20)<br />
Placebo<br />
LI 160 (900) Fluoxetine<br />
(20)<br />
Placebo<br />
<strong>Hypericum</strong> <strong>extract</strong> Iperisan<br />
3×1 (900 mg)<br />
Fluoxetine<br />
(20)<br />
Placebo<br />
STW3-VI (900) Citalopram<br />
(20)<br />
Placebo<br />
Population a N Responders n/N (%)<br />
<strong>Hypericum</strong> Comparator<br />
ITT 70 16/35 (45.7) c<br />
EP 90 a<br />
20/45 (44.5) c<br />
21/35 (60.0)<br />
22/45 (48.9)<br />
PP<br />
(12 weeks)<br />
70/102 (68.6) 69/98 (73.5)<br />
24 weeks 161 68/81 (84.0) 65/80 (81.3)<br />
ITT 244 d<br />
86/122 (70.5) 73/122 (59.8)<br />
200 d<br />
ITT 163 22/54 (40.7%) Fluoxetine<br />
20/54 (37%)<br />
Placebo<br />
21/55 (38.1%)<br />
ITT 135 17/45 (37.7%) Fluoxetine<br />
14/47 (29.8%)<br />
Placebo<br />
9/43 (20.9%)<br />
ITT 66 4/20 (20%) Fluoxetine<br />
11/20 (55%)<br />
Placebo<br />
11/26 (42.3%)<br />
PP<br />
(12 weeks)<br />
388 71/131 (54.2%) Citalopram<br />
71/127 (55.9%)<br />
Placebo<br />
51/130 (39.2%)<br />
a ITT/EP/PP data are presented where available; b Data are taken from <strong>the</strong> review <strong>of</strong> Linde et al. (2005a,b); c ITT data were not reported; d PP data not reported; e The PP analysis (data not<br />
reported) was undertaken <strong>and</strong> yielded similar results to <strong>the</strong> ITT analysis; f The ITT analysis (data not reported) was undertaken <strong>and</strong> yielded similar results to <strong>the</strong> PP analysis. DB = double-blind;<br />
EP = evaluable patients (safety analysis population); HAMD = Hamilton Depression Rating Scale; ITT = intention to treat population; MC = multicenter; NS = not significant, pN0.05; PP = per<br />
protocol (protocol compliant) population; R = r<strong>and</strong>omized.<br />
752 S. Kasper et al.
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
<strong>the</strong>se studies limits <strong>the</strong> generalization <strong>of</strong> <strong>the</strong>se findings;<br />
whereas many o<strong>the</strong>r well designed studies have demonstrated<br />
positive results <strong>for</strong> <strong>Hypericum</strong> <strong>extract</strong> in mild depression<br />
patients.<br />
Of <strong>the</strong> o<strong>the</strong>r placebo-controlled studies, two, using<br />
<strong>extract</strong> WS 5572, showed a significant response compared<br />
with placebo (Kalb et al., 2001; Laakmann et al., 1998) <strong>and</strong>,<br />
in a large study, WS 5570 <strong>extract</strong> was shown to be<br />
significantly more effective than placebo (Lecrubier et al.,<br />
2002). Findings from a study comparing two <strong>Hypericum</strong><br />
<strong>extract</strong>s with different concentrations <strong>of</strong> hyper<strong>for</strong>in (WS<br />
5572, 5% <strong>and</strong> WS 5573, 0.5%) suggest that <strong>the</strong> efficacy <strong>of</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> in mild to moderate depression might<br />
depend on <strong>the</strong> hyper<strong>for</strong>in content (Laakmann et al., 1998).<br />
A double-blind, r<strong>and</strong>omized, placebo-controlled, multicenter<br />
clinical trial (Kasper et al., 2006) has demonstrated<br />
<strong>the</strong> superior antidepressant efficacy <strong>of</strong> WS 5570 600 mg/day<br />
(in one dose) <strong>and</strong> <strong>of</strong> WS 5570 1200 mg/day (in two daily<br />
doses) compared to placebo in <strong>the</strong> treatment <strong>of</strong> patients<br />
with a mild or moderate major depressive episode after<br />
6 weeks <strong>of</strong> treatment. Interestingly <strong>the</strong> primary outcome<br />
measure, <strong>the</strong> HAMD total score change vs. baseline after<br />
6 weeks, indicates no significative differences in efficacy<br />
between WS 5570 600 mg/day <strong>and</strong> 1200 mg/day, although<br />
<strong>the</strong> number <strong>of</strong> patients who experienced remission was<br />
higher in <strong>the</strong> WS 5570 1200 mg/day group than <strong>the</strong> WS 5570<br />
600 mg/day group.<br />
The antidepressant efficacy <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>s has<br />
been studied in different controlled clinical trials per<strong>for</strong>med<br />
in patients with mild or moderate, or with moderate or<br />
severe depression. No studies have been specifically conducted<br />
in mild depression patients. In view <strong>of</strong> <strong>the</strong> high<br />
response rate normally seen in mild depression, <strong>the</strong> efficacy<br />
<strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>s in this subtype <strong>of</strong> depression has been<br />
questioned, also considering <strong>the</strong> large <strong>and</strong> unpredictable<br />
response to placebo.<br />
Interestingly a recent sub-analysis <strong>of</strong> data from three<br />
different controlled clinical trials has clearly shown that St.<br />
John's wort <strong>extract</strong> WS 5570 has a significant beneficial<br />
effect during acute treatment <strong>of</strong> patients suffering from mild<br />
depression (Kasper et al., 2008a). A subgroup <strong>of</strong> 217 patients<br />
with a pre-treatment total score ≤20 points on <strong>the</strong> 17-item<br />
HAMD was selected <strong>for</strong> <strong>the</strong> sub-analysis from a total <strong>of</strong> more<br />
than 1200 patients included into <strong>the</strong>se trials. The rates <strong>of</strong><br />
responders (i.e., patients with a HAMD total score decrease<br />
≥50%) were 73%, 64%, 71%, <strong>and</strong> 37% <strong>for</strong> WS 5570 600 mg/day,<br />
900 mg/day <strong>and</strong> 1200 mg/day, <strong>and</strong> placebo, respectively.<br />
These data suggest that <strong>Hypericum</strong> <strong>extract</strong>s, compared to<br />
placebo, were significantly able to reduce depression<br />
severity in mild depressed patients. Fur<strong>the</strong>rmore this study<br />
shows that treatment with WS 5570 leads to a substantial<br />
increase in <strong>the</strong> probability <strong>of</strong> remission within 6 weeks <strong>of</strong><br />
treatment (Kasper et al., 2008a).<br />
Continuing treatment with antidepressant drugs has been<br />
considered as <strong>the</strong> most accepted strategy to prevent relapse<br />
in major depression (Geddes et al., 2003). Different trials<br />
have investigated <strong>the</strong> efficacy <strong>of</strong> antidepressant drugs in <strong>the</strong><br />
short-term treatment <strong>of</strong> an acute episode, whereas a paucity<br />
<strong>of</strong> methodologically adequate studies have analyzed <strong>the</strong><br />
effect <strong>of</strong> such drugs during long-term prophylaxis (Fava et<br />
al., 2003). Interestingly recent studies have been per<strong>for</strong>med<br />
to assess <strong>the</strong> long-term efficacy <strong>and</strong> safety <strong>of</strong> <strong>Hypericum</strong><br />
753<br />
<strong>extract</strong>s in treatment <strong>of</strong> depression. (Kasper et al., 2004,<br />
2007, 2008b).<br />
A double-blind, r<strong>and</strong>omized, placebo-controlled trial<br />
has also been recently conducted to analyze <strong>the</strong> prophylactic<br />
efficacy <strong>and</strong> safety <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> WS 5570<br />
900 mg/day (3–6% hyper<strong>for</strong>in) compared to placebo in<br />
preventing relapse during 6 months' continuation treatment<br />
<strong>and</strong> 12 months' long-term maintenance treatment<br />
after recovery from an episode <strong>of</strong> recurrent depression<br />
(Kasper et al., 2004, 2008b). Patients who continued<br />
treatment with WS 5570 were at a 30% lower risk <strong>of</strong> relapse<br />
during <strong>the</strong> first six months than those in whom <strong>the</strong><br />
antidepressant medication was replaced by placebo after<br />
recovery from <strong>the</strong> acute episode. Interestingly, <strong>the</strong><br />
prophylactic effect <strong>of</strong> WS 5570 during long-term maintenance<br />
treatment was particularly pronounced in patients<br />
with an early onset <strong>and</strong> consequently with a more<br />
extended history <strong>of</strong> depression. Fur<strong>the</strong>rmore <strong>the</strong> <strong>tolerability</strong><br />
<strong>of</strong> WS 5570 in long-term maintenance treatment was<br />
on <strong>the</strong> placebo level (Kasper et al., 2008b).<br />
Placebo-controlled studies showed <strong>Hypericum</strong> <strong>extract</strong>s<br />
to be well tolerated with no noticeable differences<br />
between <strong>Hypericum</strong> preparations. Typically, adverse<br />
events were reported at incidences similar to placebo.<br />
Schmidt <strong>and</strong> Sommer (1993) specifically assessed cognitive<br />
effects <strong>and</strong> reported that no impairment <strong>of</strong> any cognitive<br />
functions occurred. In an LI 160 study, no notable side<br />
effects were reported in ei<strong>the</strong>r group (Sommer <strong>and</strong> Harrer,<br />
1994). In <strong>the</strong> ZE 117 study by Schrader et al. (1998)<br />
adverse events, similar in both groups, were described as<br />
transient, self-limiting <strong>and</strong>, in <strong>the</strong> case <strong>of</strong> ZE 117, mainly<br />
non-specific gastrointestinal complaints. A placebo-controlled<br />
study showed a greater incidence <strong>of</strong> specific<br />
adverse events with <strong>Hypericum</strong> <strong>extract</strong> vs. placebo<br />
(<strong>Hypericum</strong> Depression Trial Study Group, 2002). However,<br />
<strong>the</strong>se events were mild in severity <strong>and</strong> thought to be due<br />
to spurious associations created in <strong>the</strong> analysis process. In<br />
an LI 160 study by Shelton et al. (2001) LI 160 was well<br />
tolerated; only headache was reported in more patients<br />
taking LI 160 than placebo: in 39/95 (41%) <strong>and</strong> 25/100<br />
(25%), respectively (p =0.02).<br />
2.2. <strong>Hypericum</strong> <strong>extract</strong> vs. tricyclic antidepressants<br />
A number <strong>of</strong> studies have compared <strong>Hypericum</strong> <strong>extract</strong>s with<br />
<strong>the</strong> older tricyclic antidepressants <strong>for</strong> depression <strong>and</strong> shown<br />
<strong>the</strong>m to be at least as effective <strong>and</strong> better tolerated than<br />
tricyclics (Table 2) (Bergmann et al., 1993; Harrer et al.,<br />
1994; Philipp et al., 1999; Vorbach et al., 1994; Wheatley,<br />
1997; Woelk, 2000).<br />
In a double-blind study comparing <strong>Hypericum</strong> <strong>extract</strong> LI<br />
160 was as effective as maprotiline in reducing HAMD score at<br />
4 weeks <strong>and</strong> <strong>the</strong> responder rates were similar (61% <strong>and</strong> 67%<br />
<strong>for</strong> LI 160 <strong>and</strong> maprotiline, respectively, based on <strong>the</strong> perprotocol<br />
analysis). A higher proportion <strong>of</strong> patients experienced<br />
adverse events in <strong>the</strong> maprotiline group (13 [25%] <strong>and</strong><br />
18 [35%] with LI 160 <strong>and</strong> maprotiline, respectively) (Harrer<br />
et al., 1994). Increased tiredness, dryness <strong>of</strong> <strong>the</strong> mouth <strong>and</strong><br />
cardiac complaints were more commonly reported with<br />
maprotiline than with LI 160. In a 6-week, double-blind study<br />
comparing <strong>Hypericum</strong> <strong>extract</strong> LI 160 with imipramine
754 S. Kasper et al.<br />
(Vorbach et al., 1994), patients treated with LI 160 (N=67)<br />
showed a reduction in HAMD score from 20.2 to 8.8, similar to<br />
that seen in <strong>the</strong> imipramine-treated group (from 19.4 to<br />
10.7; N=68); both significant reductions from baseline<br />
(pb0.001). Interestingly, in a subgroup analysis <strong>of</strong> patients<br />
with a baseline mean HAMD score ≥21, LI 160 (N=26) was<br />
significantly more effective than imipramine (N=25) in terms<br />
<strong>of</strong> HAMD <strong>and</strong> Clinical Global Impression (CGI) severity<br />
reduction (pb.05). Patients receiving LI 160 had a lower<br />
frequency <strong>and</strong> severity <strong>of</strong> adverse effects compared with<br />
imipramine.<br />
One criticism made <strong>of</strong> <strong>the</strong> above study (Vorbach et al.,<br />
1994) was that <strong>the</strong> interpretation <strong>of</strong> similar efficacy was not<br />
based on an adequately a priori planned statistical hypo<strong>the</strong>sis.<br />
This shortcoming was addressed in ano<strong>the</strong>r LI 160 study<br />
using amitriptyline as <strong>the</strong> chosen comparator (Wheatley,<br />
1997). Patients were r<strong>and</strong>omized to receive <strong>Hypericum</strong><br />
<strong>extract</strong> LI 160 (N=87) or amitriptyline (N=78). At 6 weeks,<br />
<strong>the</strong> responder rate was statistically similar in both treatment<br />
groups (p=.064; per=protocol analysis), <strong>and</strong> adverse effects<br />
were reported significantly less frequently among LI 160treated<br />
patients (37% vs. 64% <strong>for</strong> amitriptyline; p≤0.05) with<br />
<strong>the</strong> most common side effect in <strong>the</strong> amitriptyline group being<br />
dry mouth <strong>and</strong> in <strong>the</strong> LI 160 group, headache. In total, 24% <strong>of</strong><br />
patients discontinued treatment in <strong>the</strong> LI 160 group,<br />
compared with 31% with amitriptyline; <strong>the</strong> most common<br />
reason <strong>for</strong> discontinuation in both groups was adverse<br />
events. In a comparative trial, <strong>Hypericum</strong> <strong>extract</strong> ZE 117<br />
was at least as effective as imipramine (Woelk, 2000); <strong>the</strong><br />
reduction in mean HAMD scores in patients receiving ZE 117<br />
was similar to imipramine at 6 weeks (12.00 <strong>and</strong> 12.75, <strong>for</strong> ZE<br />
117 <strong>and</strong> imipramine, respectively). Doses were chosen to<br />
avoid possible criticism from using overly low doses <strong>of</strong><br />
syn<strong>the</strong>tic antidepressants. Interestingly, at endpoint, <strong>the</strong><br />
mean score on <strong>the</strong> anxiety-somatization subscale <strong>of</strong> <strong>the</strong><br />
HAMD was significantly lower in <strong>the</strong> <strong>Hypericum</strong> <strong>extract</strong> group<br />
(3.79) than <strong>the</strong> imipramine group (4.26) (p=.03). As with LI<br />
160 in <strong>the</strong> previous study, <strong>Hypericum</strong> <strong>extract</strong> ZE 117 was<br />
significantly better tolerated than <strong>the</strong> tricyclic comparator<br />
(pb0.01): adverse effects were observed in 39% <strong>of</strong> patients<br />
given <strong>Hypericum</strong> <strong>extract</strong> <strong>and</strong> in 63% <strong>of</strong> patients administered<br />
imipramine, with 3% <strong>and</strong> 16% discontinuing treatment<br />
because <strong>of</strong> adverse events, respectively.<br />
The finding that <strong>Hypericum</strong> <strong>extract</strong> had an efficacy at<br />
least equal to that <strong>of</strong> imipramine confirmed data from an<br />
earlier study using <strong>Hypericum</strong> <strong>extract</strong> STEI 300; Philipp et al.<br />
(1999) carried out <strong>the</strong> first trial in which <strong>Hypericum</strong> <strong>extract</strong><br />
was compared with tricyclic antidepressants <strong>and</strong> placebo.<br />
Patients were r<strong>and</strong>omized to STEI 300, imipramine or<br />
placebo <strong>for</strong> 6 weeks followed by 2 weeks follow-up. STEI<br />
300 was significantly more effective than placebo <strong>and</strong> at<br />
least as effective as imipramine in terms <strong>of</strong> responder rate at<br />
6 weeks, <strong>and</strong> significant greater improvements in quality <strong>of</strong><br />
life (using <strong>the</strong> SF-36 scale) were also observed with STEI 300<br />
vs. placebo. No safety concerns with STEI 300 were noted,<br />
with <strong>the</strong> most frequent adverse effect in <strong>the</strong> imipramine<br />
group being dry mouth (38% [13% with placebo]) <strong>and</strong> in <strong>the</strong><br />
STEI 300 group, nausea (8% [2% with placebo]).<br />
In a meta-analysis by Röder et al. (2004), (N=1231) <strong>the</strong><br />
mean responder rates were 55.0% <strong>for</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>and</strong><br />
53.9% <strong>for</strong> tricyclic antidepressants (p=.13). The responder<br />
rate ratio [RR], defined as <strong>the</strong> ratio <strong>of</strong> <strong>the</strong> responder rates<br />
observed with comparator <strong>and</strong> <strong>Hypericum</strong> <strong>extract</strong> was 0.98<br />
(95% CI=0.87–1.10, p=.70) (Röder et al., 2004). The number<br />
<strong>of</strong> adverse events was significantly lower in patients<br />
administered <strong>Hypericum</strong> <strong>extract</strong> than those receiving tricyclics;<br />
<strong>the</strong> proportion <strong>of</strong> patients reporting adverse effects<br />
was 49.0% vs. 28.3% <strong>for</strong> tricyclics vs. <strong>Hypericum</strong> <strong>extract</strong>,<br />
respectively (pb.00001). Consequently, <strong>the</strong> frequency <strong>of</strong><br />
<strong>the</strong>rapeutic discontinuation was 3.6 times higher in patients<br />
treated with tricyclics than those treated with <strong>Hypericum</strong><br />
<strong>extract</strong>.<br />
Overall studies suggest that <strong>Hypericum</strong> <strong>extract</strong>s, when<br />
compared to tricyclic antidepressants, demonstrate a similar<br />
efficacy <strong>and</strong> a more favourable side-effects pr<strong>of</strong>ile.<br />
2.3. <strong>Hypericum</strong> <strong>extract</strong> vs. SSRIs<br />
There have been a number <strong>of</strong> comparative studies between<br />
<strong>Hypericum</strong> <strong>extract</strong> <strong>and</strong> SSRIs, with recent studies incorporating<br />
more rigorous methodology <strong>and</strong> greater patient<br />
numbers than earlier trials as with <strong>the</strong> o<strong>the</strong>r studies<br />
described (Table 2) (Behnke et al., 2002; Brenner et al.,<br />
2000; Gastpar et al., 2005; Harrer et al., 1999; <strong>Hypericum</strong><br />
Depression Trial Study Group, 2002; Schrader, 2000; Szegedi<br />
et al., 2005; van Gurp et al., 2002; Fava et al., 2005;<br />
Bjerkenstedt et al., 2005; Moreno et al., 2006; Gastpar et<br />
al., 2006; Papakostas et al., 2007).<br />
Several trials have compared <strong>Hypericum</strong> <strong>extract</strong> with<br />
sertraline. In a small preliminary study comparing <strong>Hypericum</strong><br />
<strong>extract</strong> LI 160 with sertraline in 30 outpatients with mild to<br />
moderate depression (Brenner et al., 2000), a clinical<br />
response was seen in 54% <strong>of</strong> patients receiving LI 160 <strong>and</strong><br />
40% <strong>of</strong> those receiving sertraline (p=NS, ITT analysis) <strong>and</strong><br />
both treatments were well tolerated. In a r<strong>and</strong>omized<br />
controlled trial in primary care, 90 patients with major<br />
depression (HAMD ≥16) were r<strong>and</strong>omized to treatment with<br />
ei<strong>the</strong>r sertraline (50–100 mg/day) or an unnamed <strong>Hypericum</strong><br />
<strong>extract</strong> containing 0.3% hypericin (900–1800 mg/day) (van<br />
Gurp et al., 2002). Assessments were carried out at entry <strong>and</strong><br />
at weeks 2, 4, 8 <strong>and</strong> 12. At week 12, <strong>Hypericum</strong> <strong>extract</strong><br />
achieved similar improvements in HAMD <strong>and</strong> Beck Depression<br />
Inventory (BDI) scores to sertraline (18.9–9.4 <strong>and</strong> 19.5–12.1<br />
<strong>for</strong> <strong>Hypericum</strong> <strong>extract</strong> vs. 19.7–11.5 <strong>and</strong> 18.4–12.0 <strong>for</strong><br />
sertraline, <strong>for</strong> HAMD <strong>and</strong> BDI respectively; p=NS). During <strong>the</strong><br />
early stages <strong>of</strong> treatment, significantly more adverse effects<br />
were reported among sertraline-treated patients than<br />
among those receiving <strong>Hypericum</strong> <strong>extract</strong>. The authors<br />
concluded that <strong>Hypericum</strong> <strong>extract</strong> has a role as a first<br />
treatment option <strong>for</strong> mild to moderate depression in primary<br />
care settings.<br />
In ano<strong>the</strong>r recent study, 123 patients with moderate<br />
depression (ICD-10 F32.1 or F33.1, HAMD 20 to 24) were<br />
r<strong>and</strong>omized to <strong>Hypericum</strong> <strong>extract</strong> STW3 (612 mg) <strong>and</strong> 118<br />
patients to sertraline (50 mg) with approximately 100<br />
patients in each group being treated <strong>for</strong> a period <strong>of</strong><br />
12 weeks followed by 12 weeks follow-up (Gastpar et al.,<br />
2005). At 12 weeks, reductions in HAMD scores were similar<br />
between groups <strong>and</strong> responder rates were 68.6% <strong>and</strong> 73.5%<br />
<strong>for</strong> STW3 <strong>and</strong> sertraline, respectively (p value not stated, PP<br />
analysis). At 24 weeks, in <strong>the</strong> 161 patients who continued on<br />
treatment <strong>for</strong> <strong>the</strong> 12-week follow-up (81 patients receiving<br />
STW3 <strong>and</strong> 80 receiving sertraline), responder rates were
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
84.0% in STW3 recipients <strong>and</strong> 81.3% in <strong>the</strong> sertraline group (p<br />
value not stated). Overall, <strong>tolerability</strong> was deemed ‘good’ or<br />
‘very good’ in most patients (99.2% <strong>and</strong> 94.9% <strong>of</strong> patients<br />
with STW3 <strong>and</strong> sertraline, respectively). Adverse events<br />
attributed to study medication <strong>and</strong> discontinuations due to<br />
adverse events were more common with sertraline than<br />
STW3; treatment-related adverse events occurred in 12<br />
[9.8%] <strong>and</strong> 16 [13.6%] patients treated with STW3 <strong>and</strong><br />
sertraline, respectively, with drug-related discontinuations<br />
seen in 5 (4.1%) <strong>and</strong> 10 (8.5%) patients, respectively. One <strong>of</strong><br />
<strong>the</strong> strengths <strong>of</strong> this study is that <strong>the</strong> sample size was<br />
calculated to ensure adequate power to demonstrate <strong>the</strong><br />
non-inferiority <strong>of</strong> STW3 compared with sertraline.<br />
In <strong>the</strong> <strong>Hypericum</strong> Depression Trial; nei<strong>the</strong>r <strong>Hypericum</strong><br />
<strong>extract</strong> (LI 160; 900–1500 mg/day) nor sertraline (50–<br />
100 mg/day) were shown to have significantly greater effect<br />
on HAMD score or significantly greater responder rate (based<br />
on HAMD <strong>and</strong> CGI scores) than placebo (<strong>Hypericum</strong> Depression<br />
Trial Study Group, 2002). However, Linde et al. (2002)<br />
suggest that as patients had relatively long-term chronic<br />
depression, this lack <strong>of</strong> efficacy vs. placebo might be due in<br />
part to a high proportion <strong>of</strong> <strong>the</strong> patients being treatment<br />
resistant (Cott <strong>and</strong> Wisner, 2002; Linde et al., 2002;<br />
Wheatley, 2002). Since patients significantly related to <strong>the</strong><br />
treatment arm to which <strong>the</strong>y had been r<strong>and</strong>omized, loss <strong>of</strong><br />
blinding may have been ano<strong>the</strong>r limitation <strong>of</strong> this study.<br />
O<strong>the</strong>r authors have suggested that unsuitable outcome<br />
measures (Spielmans, 2002; Volp, 2002), <strong>and</strong> overoptimistic<br />
effect sizes are <strong>the</strong> cause <strong>of</strong> <strong>the</strong>se unexpected findings<br />
(Linde et al., 2002). Adverse events reported at a significantly<br />
greater rate with sertraline than placebo <strong>and</strong> LI 160<br />
were diarrhoea, nausea, anorgasmia, <strong>and</strong> sweating. However,<br />
<strong>the</strong> authors note that <strong>the</strong>se events, inconsistent with<br />
<strong>Hypericum</strong>'s previously reported safety pr<strong>of</strong>ile, were all<br />
mild, <strong>and</strong> multiple comparisons may have produced spurious<br />
associations.<br />
Several studies have compared <strong>Hypericum</strong> <strong>extract</strong> with<br />
fluoxetine. A study comparing <strong>Hypericum</strong> <strong>extract</strong> LoHyp-57<br />
800 mg/day <strong>and</strong> fluoxetine 20 mg/day was carried out in 161<br />
elderly patients with mild to moderate depressive episodes<br />
(ICD-10 F32.0 <strong>and</strong> F32.1) (Harrer et al., 1999). <strong>Efficacy</strong> in<br />
terms <strong>of</strong> improvement in HAMD score <strong>and</strong> responder rate was<br />
similar with <strong>Hypericum</strong> <strong>extract</strong> <strong>and</strong> SSRI. LoHyp-57 was<br />
better tolerated than fluoxetine with a total <strong>of</strong> 12<br />
undesirable events considered to be possibly or probably<br />
related to study drug observed in <strong>the</strong> LoHyp-57-treated<br />
group compared with 17 in <strong>the</strong> group treated with<br />
fluoxetine.<br />
In a similar comparative, r<strong>and</strong>omized study in 240<br />
patients, <strong>Hypericum</strong> <strong>extract</strong> ZE 117 was at least as effective<br />
as fluoxetine <strong>for</strong> mild to moderate depression in terms <strong>of</strong><br />
HAMD score reduction, <strong>and</strong> achieved a significantly greater<br />
responder rate (p=.005) <strong>and</strong> significantly greater improvements<br />
in CGI item 1 (p=.03) than fluoxetine (Schrader,<br />
2000). ZE 117 had a superior <strong>tolerability</strong> pr<strong>of</strong>ile with a<br />
lower global incidence <strong>of</strong> adverse effects (14% vs. 25%,<br />
respectively; p b.07). No patients discontinued due to<br />
adverse events with ZE 117. In a study comparing <strong>the</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> calmigen with fluoxetine (Behnke et al.,<br />
2002), both treatments were significantly effective with no<br />
differences between groups in any <strong>of</strong> <strong>the</strong> efficacy or<br />
<strong>tolerability</strong> endpoints. O<strong>the</strong>r studies have compared<br />
755<br />
<strong>Hypericum</strong> <strong>extract</strong>s with fluoxetine (Fava et al., 2005;<br />
Bjerkenstedt et al., 2005; Moreno et al., 2006; Papakostas<br />
et al., 2007).<br />
Bjerkenstedt et al. (2005) conducted a 4-week, multicenter,<br />
double-blind, r<strong>and</strong>omized controlled study comparing<br />
LI 160 900 mg/day <strong>and</strong> fluoxetine 20 mg/day in 163<br />
patients affected with mild to moderate depression.<br />
Un<strong>for</strong>tunately both active treatments failed to prove<br />
superiority over placebo regarding <strong>the</strong> primary efficacy<br />
measure, <strong>the</strong> HAMD total score reduction from baseline to<br />
4 weeks <strong>of</strong> treatment. The number <strong>of</strong> responders was not<br />
significantly different in <strong>the</strong> three groups (Table 2). Never<strong>the</strong>less,<br />
considering <strong>the</strong> stricter criterion <strong>of</strong> remission (final<br />
HAMD b8) to assess treatment success, both <strong>Hypericum</strong> (24%)<br />
<strong>and</strong> fluoxetine (28%) were significantly superior to placebo,<br />
but <strong>Hypericum</strong> was significantly better tolerated than<br />
fluoxetine. One <strong>of</strong> <strong>the</strong> limits <strong>of</strong> this study is <strong>the</strong> duration <strong>of</strong><br />
treatment (4 weeks), which seems to be too short, considering<br />
that separation <strong>of</strong> active drug from placebo may<br />
require 6 <strong>and</strong> more weeks <strong>of</strong> treatment.<br />
Similar results have been observed in an 8-week doubleblind<br />
trial in patients with mild to moderate depression,<br />
assessing <strong>the</strong> efficacy <strong>and</strong> safety <strong>of</strong> H. per<strong>for</strong>atum in<br />
comparison with fluoxetine (Moreno et al., 2006). Outcome<br />
was monitored using <strong>the</strong> HAMD, Montgomery-Åsberg depression<br />
rating scale, CGI. The mean HAMD score was low<br />
compared to o<strong>the</strong>r studies (14–17) <strong>and</strong> only 66 patients<br />
completed <strong>the</strong> study. The authors found no differences<br />
between <strong>the</strong> groups regarding mean HAMD changes <strong>and</strong> <strong>the</strong><br />
lowest response to treatment in <strong>the</strong> <strong>Hypericum</strong> group, when<br />
compared with both placebo <strong>and</strong> fluoxetine groups. The<br />
number <strong>of</strong> patients included in this study does not allow<br />
definitive conclusions due to <strong>the</strong> statistical power <strong>of</strong> <strong>the</strong><br />
sample size.<br />
In a 12-week, r<strong>and</strong>omized, active- <strong>and</strong> placebo-controlled,<br />
parallel-group, double-blind study Fava et al.<br />
(2005) have analyzed <strong>the</strong> efficacy <strong>and</strong> safety <strong>of</strong> H.<br />
per<strong>for</strong>atum (LI-160 900 mg/die) relative to both placebo<br />
<strong>and</strong> fluoxetine (20 mg/die) in a population <strong>of</strong> 135 outpatients<br />
with mild to moderately severe MDD. The primary efficacy<br />
end point <strong>for</strong> this study was <strong>the</strong> final HAMD-17 total score<br />
after 12 weeks <strong>of</strong> r<strong>and</strong>omized treatment on end point. The<br />
mean HAMD score in <strong>the</strong> three groups was 19. Interestingly<br />
<strong>the</strong> authors found that St. John's wort treatment was<br />
associated with a significantly (Pb0.05) greater decrease in<br />
HAMD-17 scores compared with fluoxetine <strong>and</strong> showed a<br />
trend toward statistically significant superiority over placebo.<br />
An higher rate <strong>of</strong> remission (HAMD-17b8) was observed in<br />
<strong>the</strong> St. John's wort group (38%) compared with <strong>the</strong> fluoxetine<br />
group (30%) <strong>and</strong> <strong>the</strong> placebo group (21%). There were no<br />
adverse events related treatment discontinuations in <strong>the</strong><br />
<strong>Hypericum</strong> group <strong>and</strong> placebo-treated patients, whereas 4%<br />
(2/47) <strong>of</strong> <strong>the</strong> fluoxetine-treated patients dropped out<br />
because <strong>of</strong> side effects. Surprisingly in this study fluoxetine<br />
lacked <strong>of</strong> efficacy <strong>and</strong> St. John's wort treatment was not<br />
significantly superior to placebo. The authors suggest that<br />
<strong>the</strong>se data may be explained with <strong>the</strong> smaller than planned<br />
sample size. Recruitment was stopped be<strong>for</strong>e planned<br />
sample size was reached due to decision <strong>of</strong> <strong>the</strong> sponsor.<br />
Larger studies are <strong>the</strong>re<strong>for</strong>e needed to fully compare <strong>the</strong><br />
efficacy <strong>of</strong> H. per<strong>for</strong>atum vs. fluoxetine in <strong>the</strong> treatment <strong>of</strong><br />
mild to moderate depression.
756 S. Kasper et al.<br />
One <strong>of</strong> <strong>the</strong> most discussed issues in <strong>the</strong> field <strong>of</strong> antidepressant<br />
trial is <strong>the</strong> relationship between timing <strong>of</strong> clinical<br />
improvement <strong>and</strong> resolution <strong>of</strong> depressive symptoms during<br />
<strong>the</strong> treatment <strong>of</strong> major depressive disorder (Papakostas<br />
et al., 2007). Some authors have proposed that patients<br />
showing a significant symptom improvement during <strong>the</strong> first<br />
2 weeks <strong>of</strong> treatment, might not respond to antidepressant<br />
drugs per se, but <strong>the</strong>y might have a ‘placebo’ pattern <strong>of</strong><br />
response (Stewart et al., 1998; Nierenberg et al., 2004). On<br />
<strong>the</strong> contrary o<strong>the</strong>r authors have demonstrated that a clinical<br />
response during <strong>the</strong> first 2 weeks <strong>of</strong> treatment can highly<br />
predict a greater drug vs. placebo difference in depressive<br />
symptoms improvement at endpoint (Stassen et al., 1998;<br />
Szegedi et al., 2003).<br />
Along this line Papakostas et al. (2007) have examined<br />
data from <strong>the</strong> 39 responders <strong>of</strong> <strong>the</strong> above discussed 12-week<br />
double-blind study (Fava et al., 2005) comparing H.<br />
per<strong>for</strong>atum, fluoxetine or placebo (Papakostas et al.,<br />
2007). The authors defined an early clinical response as a<br />
25% decrease in 17-item HDRS after 2 weeks <strong>and</strong> found that,<br />
among <strong>the</strong> 39 responders, earlier clinical improvement<br />
predicted lower HDRS-17 scores at week 12 in <strong>the</strong> patients<br />
treated with ei<strong>the</strong>r fluoxetine or <strong>Hypericum</strong>, but not in<br />
placebo-treated patients. The small sample size <strong>of</strong> <strong>the</strong> study<br />
does not allow definitive conclusions <strong>and</strong> larger studies are<br />
needed to demonstrate that an early clinical improvement<br />
during treatment can predict a greater symptom resolution<br />
at endpoint in depressed patients responding to <strong>Hypericum</strong><br />
or fluoxetine.<br />
Szegedi et al. (2005) conducted a well designed, 6-week,<br />
multicenter, double-blind, r<strong>and</strong>omized controlled study<br />
comparing WS 5570 900 mg/day <strong>and</strong> paroxetine 20 mg/day<br />
in 251 patients affected with moderate to severe depression.<br />
This study has been thought worthy <strong>of</strong> inclusion in this review<br />
because it is <strong>the</strong> only study to compare <strong>Hypericum</strong> <strong>extract</strong><br />
with paroxetine <strong>and</strong>, although patients had a more severe<br />
<strong>for</strong>m <strong>of</strong> depression than <strong>the</strong> o<strong>the</strong>r studies reviewed, nearly<br />
half (44%) had a baseline HAMD <strong>of</strong> 22–24 (moderate<br />
depression). To avoid criticisms <strong>of</strong> inadequate dosing doses<br />
were increased to 1800 mg/day WS 5570 or 40 mg/day<br />
paroxetine if no response was observed within 2 weeks.<br />
Outcome was monitored using <strong>the</strong> HAMD, Montgomery-<br />
Åsberg depression rating scale, CGI, <strong>and</strong> BDI. After<br />
2 weeks, 57% <strong>and</strong> 48% <strong>of</strong> patients in <strong>the</strong> WS 5570 <strong>and</strong><br />
paroxetine groups, respectively, were switched to <strong>the</strong> higher<br />
doses. Between baseline <strong>and</strong> day 42, <strong>the</strong> average reduction<br />
in <strong>the</strong> HAMD score was 14.4 (SD 8.8) <strong>for</strong> WS 5570 <strong>and</strong> 11.4 (SD<br />
8.6) <strong>for</strong> paroxetine (pb.01; ITT analysis). Fig. 1 shows <strong>the</strong><br />
total HAMD scores over time in <strong>the</strong> intent-to-treat population.<br />
WS 5570 showed a numerically greater responder rate<br />
compared with paroxetine, although this did not reach<br />
significance (70.5% vs. 59.8%; p=.08). In addition, a greater<br />
proportion <strong>of</strong> patients treated with paroxetine reported<br />
adverse events. Data show that WS 5570 was at least as<br />
effective as SSRI even in patients with moderate depression<br />
with greater <strong>tolerability</strong>, <strong>and</strong> suggest that <strong>Hypericum</strong><br />
<strong>extract</strong> is a valid alternative to more frequently prescribed<br />
antidepressants such as SSRIs.<br />
Finally Gastpar et al. (2006) conducted a 6-week, large<br />
multicenter, double-blind, r<strong>and</strong>omized placebo-controlled<br />
study to compare <strong>the</strong> efficacy <strong>and</strong> safety <strong>of</strong> H. per<strong>for</strong>atum<br />
(STW3-VI 900 mg/day) to citalopram 20 mg/day in 388<br />
Figure 1 Total HAMD scores over time (intention to treat<br />
analysis, means <strong>and</strong> 95% confidence intervals) from a trial<br />
comparing <strong>the</strong> <strong>Hypericum</strong> <strong>extract</strong> WS 5570 <strong>and</strong> paroxetine in<br />
244 patients with moderate to severe depression. On day 42 WS<br />
5570 was statistically superior to paroxetine (pb0.01, two-sided<br />
t-test) (modified from Szegedi et al. (2005) with permission).<br />
patients affected with moderate depression. Outcome was<br />
monitored using <strong>the</strong> HAMD, <strong>the</strong> von Zerssen's Adjective Mood<br />
scale <strong>and</strong> <strong>the</strong> CGI. The mean HAMD score in <strong>the</strong> three<br />
treatment groups was 22. The authors found a statistical<br />
significant <strong>the</strong>rapeutic equivalence <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong><br />
STW3-VI to citalopram <strong>and</strong> <strong>the</strong> superiority <strong>of</strong> this <strong>Hypericum</strong><br />
<strong>extract</strong> over placebo. Fur<strong>the</strong>rmore <strong>the</strong> percentage <strong>of</strong><br />
responders at <strong>the</strong> end <strong>of</strong> treatment was 54.2% in <strong>the</strong><br />
<strong>Hypericum</strong> group, 55.9% in <strong>the</strong> citalopram group <strong>and</strong> 39.2%<br />
in <strong>the</strong> placebo group. More adverse events were observed in<br />
patients treated with citalopram than in o<strong>the</strong>r group<br />
treatment, thus suggesting both a non-inferiority in terms<br />
<strong>of</strong> clinical efficacy <strong>and</strong> a better safety <strong>and</strong> <strong>tolerability</strong> <strong>of</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> in comparison to citalopram (Gastpar<br />
et al., 2006).<br />
3. Evidence from meta-analyses<br />
Several systematic reviews <strong>and</strong> quantitative meta-analyses<br />
<strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> trials, compared with both placebo <strong>and</strong><br />
st<strong>and</strong>ard antidepressants, provide a comprehensive overview<br />
<strong>of</strong> clinical studies involving <strong>Hypericum</strong> <strong>extract</strong> (Table 3)<br />
(Kim et al., 1999; Linde et al., 1996, 2005a,b, 2008; Linde<br />
<strong>and</strong> Knuppel, 2005; Röder et al., 2004; Volz <strong>and</strong> Laux, 2000;<br />
Whiskey et al., 2001; Rahimi et al., 2009). In Table 3, in all<br />
meta-analyses except Röder et al., <strong>the</strong> RR was defined as <strong>the</strong><br />
ratio <strong>of</strong> <strong>the</strong> responder rates observed with <strong>Hypericum</strong><br />
<strong>extract</strong> <strong>and</strong> comparator, whereas in Röder et al. it is<br />
inverted (comparator/<strong>Hypericum</strong>).<br />
Over time, <strong>the</strong>re has been a trend towards an improved<br />
quality <strong>of</strong> meta-analysis relating to <strong>Hypericum</strong> <strong>extract</strong> in<br />
terms <strong>of</strong> studies included <strong>and</strong> <strong>the</strong> criteria used to select<br />
studies. Recent meta-analyses <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> have<br />
sought to encompass more methodologically sound studies<br />
than those used in earlier meta-analyses, which has been<br />
made possible by improvements in <strong>the</strong> quality <strong>of</strong> individual<br />
clinical studies. This has led to a greater number <strong>of</strong> large<br />
r<strong>and</strong>omized studies with rigorous criteria <strong>for</strong> major
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
Figure 2 Change from baseline (means <strong>and</strong> 95% confidence intervals) in <strong>the</strong> re-scaled cluster 1 (core symptoms <strong>of</strong> depression) <strong>and</strong><br />
cluster 2 (mainly anxiety <strong>and</strong> insomnia-related symptoms) scores in 544 patients with mild to moderate depression treated with<br />
<strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> 42 days (Kasper <strong>and</strong> Dienel, 2002). Re-scaled cluster scores were calculated as <strong>the</strong> means <strong>of</strong> <strong>the</strong> scores <strong>of</strong> all<br />
items included in <strong>the</strong> cluster re-scaled to a 0–100 scale (where 0=all items in <strong>the</strong> cluster score 0; 100=all items in <strong>the</strong> cluster scored<br />
<strong>the</strong> maximum score). Negative values indicate symptom improvement.<br />
depression <strong>and</strong> higher baseline depression scores. Fur<strong>the</strong>r<br />
evidence <strong>for</strong> <strong>the</strong> greater selectivity <strong>of</strong> included studies in<br />
more recent meta-analyses relates to <strong>the</strong> diagnostic criteria<br />
used; in Röder et al. (2004), <strong>the</strong> majority <strong>of</strong> studies included<br />
used ICD-9 <strong>and</strong> ICD-10 <strong>and</strong> DSM (DSM-III, III-R, <strong>and</strong> IV) criteria.<br />
A minimum HAMD score <strong>of</strong> 16 to 21 was stated as inclusion<br />
criteria in 15 studies <strong>and</strong>, in 7 studies, a maximum score <strong>of</strong><br />
20, 24 or 25 was specified. One <strong>of</strong> <strong>the</strong> major limitations,<br />
however, is <strong>the</strong> inclusion in all meta-analyses <strong>of</strong> studies using<br />
different preparations <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>.<br />
Linde et al. (2005a) point out that, compared with trials<br />
published be<strong>for</strong>e 1995, newer trials have larger sample sizes,<br />
are <strong>of</strong> longer duration, are more likely to use a placebo run-in<br />
phase, are more <strong>of</strong>ten restricted to patients who met criteria<br />
<strong>for</strong> major depression <strong>and</strong> tend to include patients with higher<br />
baseline scores on depression scales. In addition, documentation<br />
<strong>of</strong> methodology <strong>and</strong> daily doses are more thorough in<br />
<strong>the</strong> more recent trials. In <strong>the</strong>ir most recent meta-analysis<br />
Linde et al. (2008) have examined several new well-designed<br />
trials restricted to patients with major depression.<br />
Linde <strong>and</strong> Mulrow (2000) carried out a systematic review<br />
<strong>and</strong> meta-analysis <strong>of</strong> <strong>the</strong> efficacy <strong>and</strong> <strong>tolerability</strong> <strong>of</strong><br />
<strong>Hypericum</strong> <strong>extract</strong>, which was published as a Cochrane<br />
review. This was an update <strong>of</strong> a systematic review carried<br />
out in 1996, one <strong>of</strong> <strong>the</strong> first <strong>for</strong> <strong>Hypericum</strong> <strong>extract</strong>. Only<br />
r<strong>and</strong>omized studies were included in which <strong>Hypericum</strong><br />
<strong>extract</strong> was compared to placebo or conventional antidepressants<br />
<strong>for</strong> <strong>the</strong> treatment <strong>of</strong> depression. Outcomes were<br />
measured using st<strong>and</strong>ardized clinical evaluation scales.<br />
Moreover, <strong>for</strong> inclusion, <strong>the</strong> methodology <strong>of</strong> each individual<br />
study required assessment by at least two independent<br />
observers using <strong>the</strong> evaluation scale proposed by Jadad et al.<br />
(1996). Of 45 studies identified, 27 met <strong>the</strong> inclusion criteria<br />
<strong>for</strong> a total <strong>of</strong> 2291 patients. Seventeen <strong>of</strong> <strong>the</strong> included trials<br />
757<br />
were placebo-controlled. Ten studies compared <strong>Hypericum</strong><br />
<strong>extract</strong> (8 <strong>Hypericum</strong> <strong>extract</strong> alone, two in combination with<br />
Valeriana <strong>extract</strong>) with o<strong>the</strong>r antidepressants or sedatives.<br />
Among <strong>the</strong>se various trials, diagnostic criteria were quite<br />
broad <strong>and</strong> included dysthymia <strong>and</strong> also patients with mild<br />
symptoms who did not meet criteria <strong>for</strong> major depression.<br />
O<strong>the</strong>r potential drawbacks among <strong>the</strong>se studies include a<br />
limited follow-up period, which <strong>for</strong> <strong>the</strong> most part did not<br />
exceed 4 weeks, <strong>and</strong> <strong>the</strong> use <strong>of</strong> low doses <strong>of</strong> st<strong>and</strong>ard<br />
antidepressants in comparative trials. None<strong>the</strong>less, in spite<br />
<strong>of</strong> <strong>the</strong>se negative aspects, <strong>the</strong> methodology was considered<br />
generally acceptable. Overall, <strong>Hypericum</strong> <strong>extract</strong>s were<br />
significantly superior to placebo (RR 2.47, 95% CI=1.69–<br />
3.61) <strong>for</strong> short-term <strong>the</strong>rapy <strong>of</strong> mild or moderate depression<br />
<strong>and</strong> were very well tolerated. In this meta-analysis,<br />
<strong>Hypericum</strong> was also shown to be at least as effective as<br />
st<strong>and</strong>ard antidepressants (<strong>Hypericum</strong> <strong>extract</strong> alone, RR 1.01,<br />
95% CI=0.87–1.16; combined <strong>extract</strong>s, RR 1.52, 95%<br />
CI=0.78–2.94). However, data from this meta-analysis did<br />
not allow to establish whe<strong>the</strong>r <strong>Hypericum</strong> <strong>extract</strong> has<br />
equivalent clinical efficacy to o<strong>the</strong>r antidepressants <strong>for</strong> <strong>the</strong><br />
heterogeneity <strong>of</strong> <strong>the</strong> studies, <strong>the</strong> short duration <strong>of</strong> <strong>the</strong><br />
observation period <strong>and</strong> <strong>the</strong> low proportion <strong>of</strong> studies using an<br />
SSRI as comparator.<br />
In a subsequent meta-analysis by Linde et al., (2005a,b),<br />
published in 2005, which is an update to <strong>the</strong> 2000 systematic<br />
Cochrane review, only r<strong>and</strong>omized, controlled, double-blind<br />
studies were included; <strong>of</strong> 68 possible studies identified, 37<br />
trials met <strong>the</strong> inclusion criteria, including 26 comparisons<br />
with placebo <strong>and</strong> 14 with st<strong>and</strong>ard syn<strong>the</strong>tic antidepressants<br />
(13 <strong>of</strong> <strong>the</strong>se provided efficacy data: 6 tricyclics, 7 SSRIs).<br />
Responders were defined as those who experienced objective<br />
improvement using <strong>the</strong> HAMD <strong>and</strong> CGI scales. The studies<br />
with placebo involved 3320 patients, many <strong>of</strong> whom were
758 S. Kasper et al.<br />
Table 3 Overview <strong>of</strong> several meta-analyses <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>.<br />
Review Year Inclusion<br />
criteria<br />
(Linde et al., 1996) 1996 RCT a , depressed patients,<br />
single or combination<br />
preparation, comparison<br />
with placebo or o<strong>the</strong>r<br />
antidepressants, all<br />
clinical outcome measures<br />
(Linde <strong>and</strong> Mulrow, 2000) 2000 RCT, depressed patients,<br />
single or combination<br />
preparation, comparison<br />
with placebo or o<strong>the</strong>r<br />
antidepressants, all<br />
clinical outcome measures<br />
(Kim et al., 1999) 1999 Blinded/controlled, single<br />
preparation, depression<br />
(ICD10, DSM-IIIR, DSM-IV),<br />
outcomes by HAMD<br />
(Volz <strong>and</strong> Laux, 2000) 2000 Placebo-controlled trials<br />
<strong>of</strong> <strong>Hypericum</strong> or fluoxetine<br />
in mild depressive disorders<br />
(HAMD ≤24)<br />
(Whiskey et al., 2001) 2001 RCT, depression, single<br />
preparation, comparison<br />
with placebo or o<strong>the</strong>r<br />
antidepressants, outcomes<br />
by HAMD<br />
(Kasper <strong>and</strong> Dienel, 2002) 2002 Double-blind, RCT, mild<br />
to moderate depression,<br />
comparison with placebo,<br />
outcomes by HAMD<br />
(Röder et al., 2004) 2004 Double-blind, RCT,<br />
depression, single<br />
preparation, comparison<br />
with placebo or o<strong>the</strong>r<br />
antidepressants<br />
(Linde et al., 2005a) 2005 Double-blind, RCT, included<br />
patients with depression,<br />
single preparations,<br />
comparison with placebo or<br />
o<strong>the</strong>r antidepressant, all<br />
clinical outcomes<br />
(Rahimi et al., 2009) 2009 Double-blind, RCT, included<br />
patients with depression<br />
single or combination<br />
preparation, comparison<br />
with placebo or SSRI,<br />
outcomes by HAMD<br />
Number <strong>of</strong> studies<br />
(total number <strong>of</strong> patients)<br />
15 vs. placebo (1008)<br />
8 vs. antidepressants/<br />
sedatives (749)<br />
17 vs. placebo (1168)<br />
10 vs. antidepressants (1123)<br />
2 vs. placebo (169)<br />
4 vs. tricyclic<br />
antidepressants (482)<br />
17 <strong>Hypericum</strong> vs.<br />
placebo (1739)<br />
9 fluoxetine vs.<br />
placebo (1873)<br />
14 vs. placebo<br />
9 vs. antidepressants<br />
Responder rate ratio*<br />
(95% CI)<br />
2.67 (1.78–4.01)<br />
1.1 (0.93–1.31) <strong>for</strong><br />
single preparations<br />
1.52 (0.78–2.94)<br />
<strong>for</strong> combinations<br />
2.47 (1.69–3.61)<br />
<strong>for</strong> single preparations<br />
2.00 (1.14–3.52) <strong>for</strong><br />
combination<br />
1.01 (0.87–1.16) <strong>for</strong><br />
single preparations<br />
1.52 (0.78–2.94) <strong>for</strong><br />
combination<br />
1.48 (1.03–1.92)<br />
1.11 (0.92–1.29)<br />
Responder rate ratio<br />
not stated<br />
HAMD reduction<br />
10.2 (51.4%)<br />
12.5 (55.5%)<br />
1.98 (1.49–2.62)<br />
1.00 (0.90–1.11)<br />
3 vs. placebo (544) 1.35 <strong>for</strong> cluster<br />
1 HAMD items<br />
1.40 <strong>for</strong> cluster<br />
2 HAMD items<br />
19 vs. placebo (2129)<br />
15 vs. antidepressants<br />
(2231)<br />
26 vs. placebo (3320)<br />
14 vs. antidepressants<br />
(2283) e<br />
13 SSRI vs. placebo (1995)<br />
11 <strong>Hypericum</strong> vs. SSRI (1652)<br />
0.66 (0.57–0.78)<br />
0.96 (0.85–1.08)<br />
1.15 (1.02–1.29) b <strong>and</strong><br />
2.06 (1.65–2.59) c<br />
depression only<br />
1.71 (1.40–2.09) c <strong>and</strong><br />
6.13 (3.63–10.38) d<br />
including o<strong>the</strong>r<br />
diagnoses<br />
1.01 (0.93–1.10) overall<br />
0.97 (0.85–1.12) vs. SSRIs<br />
1.03 (0.93–1.14) vs. older<br />
antidepressants<br />
1.22 (1.03–1.45)<br />
0.99 (0.91–1.08)
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
Table 3 (continued)<br />
Review Year Inclusion<br />
criteria<br />
(Linde et al., 2008) 2008 Double-blind, RCT, included<br />
only patients with major<br />
depression, single<br />
preparations, comparison<br />
with placebo or o<strong>the</strong>r<br />
antidepressant, all<br />
clinical outcomes<br />
affected by major depression, <strong>and</strong> were treated <strong>for</strong> more<br />
than 4 weeks using a mean dose <strong>of</strong> 800 mg/day <strong>Hypericum</strong><br />
<strong>extract</strong>. The 14 trials using st<strong>and</strong>ard antidepressants as <strong>the</strong><br />
comparator involved 2283 patients (Linde et al., 2005a;<br />
Linde <strong>and</strong> Knuppel, 2005). The proportion <strong>of</strong> responders was<br />
comparable in <strong>Hypericum</strong> <strong>extract</strong>-treated patients compared<br />
with that seen with both tricyclic antidepressants<br />
(RR=1.03, 95% CI =0.93–1.14) <strong>and</strong> SSRIs (RR=0.98, 95%<br />
CI=0.85–1.12). There was a trend towards fewer dropouts<br />
<strong>for</strong> any reason or due to adverse effects in patients treated<br />
with <strong>Hypericum</strong> <strong>extract</strong> compared with placebo <strong>and</strong> also less<br />
dropouts compared with those administered st<strong>and</strong>ard<br />
antidepressants.<br />
To avoid compromising results by inclusion <strong>of</strong> studies with<br />
weak methodology, Kasper <strong>and</strong> Dienel (2002) carried out a<br />
meta-analysis using trials that also met relatively strict<br />
criteria based on patient selection (diagnostic criteria),<br />
outcome measures used, design (double-blind, placebocontrolled),<br />
period <strong>of</strong> observation as well as quality <strong>of</strong><br />
investigator rating. The meta-analysis included three r<strong>and</strong>omized,<br />
double-blind, placebo-controlled studies including<br />
a total <strong>of</strong> 544 patients affected by mild to moderate<br />
depression who received <strong>Hypericum</strong> <strong>extract</strong> (3×300 mg/<br />
day) or placebo. These studies all utilized st<strong>and</strong>ardized <strong>and</strong><br />
well-defined diagnostic criteria (DSM-IV) <strong>of</strong> mild to moderate<br />
major depression. Primary outcome was measured using <strong>the</strong><br />
total HAMD score <strong>and</strong> <strong>the</strong>rapy lasted <strong>for</strong> 6 weeks.<br />
The objective <strong>of</strong> <strong>the</strong> meta-analysis was to evaluate <strong>the</strong><br />
<strong>the</strong>rapeutic pr<strong>of</strong>ile <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> by cluster analysis<br />
<strong>and</strong> determine if <strong>Hypericum</strong> <strong>extract</strong> has a selective effect on<br />
particular signs <strong>and</strong> symptoms <strong>of</strong> depression. Cluster analysis is<br />
a statistical technique that attempts to overcome potential<br />
shortcomings <strong>of</strong> factor analysis, particularly difficulties in<br />
interpretation. For <strong>the</strong> HAMD, most analyses include two<br />
factors, one related to core symptoms <strong>of</strong> depression (e.g.<br />
pr<strong>of</strong>ound, non-reactive sadness, somatic symptoms, <strong>and</strong><br />
lethargy) <strong>and</strong> ano<strong>the</strong>r to elements such as agitation <strong>and</strong><br />
anxiety. For this meta-analysis, two clusters were also<br />
identified. The first (HAMD items 1, 2, 3, 7, 8, 12, 13, 14,<br />
<strong>and</strong> 16) represented core symptoms <strong>of</strong> depression including<br />
somatic symptoms. The second (HAMD items 4, 5, 6, 9, 10, 11,<br />
15, <strong>and</strong> 17) was composed primarily <strong>of</strong> items that evaluated<br />
anxiety <strong>and</strong> insomnia correlated with depression. <strong>Hypericum</strong><br />
Number <strong>of</strong> studies<br />
(total number <strong>of</strong> patients)<br />
18 vs. placebo (3064)<br />
17 vs. antidepressants (2810)<br />
Responder rate ratio*<br />
(95% CI)<br />
1.28 (1.10–1.49) in <strong>the</strong><br />
nine larger trials<br />
1.87 (1.22–2.87) in <strong>the</strong><br />
nine smaller trials<br />
1.02 (0.90–1.15) vs. tri- or<br />
tetracyclic antidepressants<br />
1.00 (0.90–1.11) vs. SSRIs<br />
*The responder rate ratio is <strong>the</strong> ratio <strong>of</strong> <strong>the</strong> responder rates observed with <strong>Hypericum</strong> <strong>and</strong> comparator (a value N1 denotes a greater effect<br />
with <strong>Hypericum</strong> vs. comparator) with <strong>the</strong> exception <strong>of</strong> Röder et al. (2004), where it is inverted <strong>and</strong> defined as <strong>the</strong> ratio <strong>of</strong> <strong>the</strong> responder<br />
rates <strong>of</strong> comparator <strong>and</strong> <strong>Hypericum</strong> (in this case a value b1 denotes a greater effect with <strong>Hypericum</strong> vs. comparator). a Or quasi-r<strong>and</strong>omized<br />
(e.g. alteration); b five larger trials; c six smaller trials; d five smaller trials; e 13 provided quantitative data. HAMD = Hamilton Depression<br />
Rating Scale; RCT = r<strong>and</strong>omized controlled trial; SSRI = selective serotonin reuptake inhibitor.<br />
759<br />
<strong>extract</strong> was found to be particularly efficient in reducing core<br />
symptoms <strong>of</strong> depression in a global manner with a <strong>the</strong>rapeutic<br />
pr<strong>of</strong>ile similar to SSRIs. Fig. 2 shows <strong>the</strong> change <strong>of</strong> <strong>the</strong> cluster<br />
scores vs. baseline, <strong>and</strong> demonstrates that score changes<br />
over time were similar <strong>for</strong> both clusters but <strong>the</strong> extent <strong>of</strong><br />
score changes was more pronounced <strong>for</strong> core symptoms <strong>of</strong><br />
depression (cluster 1).<br />
The meta-analysis by Röder et al. (2004) examined <strong>the</strong><br />
available clinical data by considering only <strong>the</strong> average<br />
responder rate to <strong>Hypericum</strong> <strong>extract</strong> <strong>the</strong>rapy, compared<br />
with placebo or o<strong>the</strong>r antidepressants. This meta-analysis<br />
found that, compared with newer generation antidepressants,<br />
<strong>the</strong> mean responder rate was highly similar in <strong>the</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> <strong>and</strong> antidepressant groups (51.0% vs. 48.0%, respectively)<br />
(Röder et al., 2004). This review also differentiated<br />
rates <strong>of</strong> treatment success from <strong>Hypericum</strong> <strong>extract</strong> or<br />
comparator <strong>the</strong>rapy according to subgroups <strong>of</strong> patients,<br />
based on severity. Inclusion criteria <strong>for</strong> <strong>the</strong> meta-analysis<br />
were a double-blind, r<strong>and</strong>omized design with placebo or o<strong>the</strong>r<br />
antidepressants as <strong>the</strong> comparator, diagnosis <strong>of</strong> depression<br />
<strong>and</strong> use <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> alone <strong>and</strong> not in combination<br />
with o<strong>the</strong>r plant <strong>extract</strong>s. The quality <strong>of</strong> <strong>the</strong> studies was<br />
quantified using a scoring system derived from criteria used by<br />
<strong>the</strong> authors. Thirty studies satisfied both <strong>the</strong> inclusion <strong>and</strong><br />
quality criteria <strong>for</strong> <strong>the</strong> meta-analysis, 12 <strong>of</strong> which had been<br />
previously included in <strong>the</strong> meta-analysis by Linde et al. (Linde<br />
et al., 2005a; Linde <strong>and</strong> Knuppel, 2005). However, <strong>the</strong> metaanalysis<br />
from Röder et al. (2004) contained more studies using<br />
SSRIs as <strong>the</strong> comparator. The results showed a significantly<br />
higher efficacy with <strong>Hypericum</strong> <strong>extract</strong> compared with<br />
placebo (N=2129, RR=0.66, 95% CI=0.57–0.78). The average<br />
number <strong>of</strong> patients responding to <strong>the</strong>rapy was also significantly<br />
higher <strong>for</strong> <strong>Hypericum</strong> <strong>extract</strong> compared with placebo<br />
(53.3% vs. 32.7%, respectively; pb.00001, number needed to<br />
treat [NNT]=4.2, 95% CI=3.0–6.6). <strong>Hypericum</strong> <strong>extract</strong> also<br />
showed an efficacy that was similar to st<strong>and</strong>ard antidepressants<br />
(N=2231, RR=0.96, 95% CI=0.85–1.08) <strong>and</strong> <strong>the</strong> average<br />
number <strong>of</strong> patients responding to <strong>the</strong>rapy was also similar <strong>for</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> <strong>and</strong> o<strong>the</strong>r antidepressants [52.3% vs.<br />
51.3%, respectively (p=.509)]. In <strong>the</strong> subgroup with mild to<br />
moderate depression, <strong>Hypericum</strong> <strong>extract</strong> showed significantly<br />
higher efficacy with respect to st<strong>and</strong>ard antidepressants<br />
(N=1166, RR=0.85, 95% CI=0.75–0.97, p=.01, NNT=14.3,
760 S. Kasper et al.<br />
95% CI=8.3–100.0). The mean responder rate was 59.5% in <strong>the</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> group compared with 52.9% <strong>for</strong> st<strong>and</strong>ard<br />
antidepressants (p=.14).<br />
Linde et al. (2008) carried out a systematic review <strong>and</strong><br />
meta-analysis on <strong>the</strong> efficacy <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> in major depression, which was published as a<br />
Cochrane review. Only r<strong>and</strong>omized studies were included in<br />
which <strong>Hypericum</strong> <strong>extract</strong> was compared with placebo or<br />
st<strong>and</strong>ard antidepressants <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> major<br />
depression. Outcomes were measured using st<strong>and</strong>ardized<br />
clinical evaluation scales. Moreover, <strong>for</strong> inclusion, <strong>the</strong><br />
methodology <strong>of</strong> each individual study required assessment<br />
by at least two independent observers, using scales<br />
developed by Jadad et al. (1996).<br />
Of 79 studies identified, 29 met <strong>the</strong> inclusion criteria <strong>for</strong> a<br />
total <strong>of</strong> 5489 patients. Eighteen <strong>of</strong> <strong>the</strong> included trials were<br />
placebo-controlled. Seventeen studies compared <strong>Hypericum</strong><br />
<strong>extract</strong> with o<strong>the</strong>r antidepressants. Eight new trials were<br />
included since <strong>the</strong> last update (Linde et al., 2005b) with a<br />
total <strong>of</strong> 1947 patients. The severity <strong>of</strong> depression was<br />
described as mild to moderate in 19 trials, <strong>and</strong> as moderate<br />
to severe in 9 trials. Responders were defined as those who<br />
experienced objective improvement using <strong>the</strong> HAMD <strong>and</strong> CGI<br />
scales. The studies with placebo involved 3064 patients,<br />
whereas <strong>the</strong> seventeen trials using st<strong>and</strong>ard antidepressants<br />
as <strong>the</strong> comparator involved 2810 patients. Patients receiving<br />
<strong>Hypericum</strong> <strong>extract</strong> were significantly more likely to be<br />
responders (RR=1.48; 95% CI=1.23–1.77), but placebocontrolled<br />
trials were highly heterogeneous. In <strong>the</strong> nine<br />
larger trials <strong>the</strong> combined response rate ratio <strong>for</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> compared with placebo was 1.28 (95% CI=1.10–<br />
1.49), whereas effects in favour <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> were<br />
more pronounced in <strong>the</strong> nine smaller trials (RR=1.87 95%<br />
CI=1.22–2.87).<br />
In <strong>the</strong> meta-analysis, <strong>Hypericum</strong> was also shown to be at<br />
least as effective as tricyclic antidepressants (RR=1.02 95%<br />
CI=0.90–1.15) or SSRI (RR=1.00 95% CI=0.90–1.11). Interestingly<br />
patients treated with <strong>Hypericum</strong> <strong>extract</strong> were less<br />
likely to drop out from studies due to adverse effects than<br />
patients allocated to older st<strong>and</strong>ard antidepressants<br />
(OR=0.24; 95% CI=0.13–0.46) or to SSRIs (OR=0.53; 95% CI<br />
0.34–0.83).<br />
This meta-analysis examines <strong>for</strong> <strong>the</strong> first time only clinical<br />
trials conducted in patients with major depression <strong>and</strong><br />
suggests that <strong>Hypericum</strong> <strong>extract</strong> have a modest effect over<br />
placebo in a similar range as st<strong>and</strong>ard antidepressant drugs,<br />
finally showing that <strong>Hypericum</strong> <strong>extract</strong>s have fewer side<br />
effects than st<strong>and</strong>ard antidepressant drugs. However it<br />
should be underlined that trials from German-speaking<br />
countries have significantly more positive results than trials<br />
from o<strong>the</strong>r countries. The reasons <strong>for</strong> <strong>the</strong> “country effect”<br />
are unclear <strong>and</strong> fur<strong>the</strong>r studies are needed to investigate this<br />
difference.<br />
In a recent meta-analysis Rahimi et al. (2009) have<br />
examined all studies comparing <strong>the</strong> efficacy <strong>and</strong> <strong>tolerability</strong><br />
<strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>and</strong> SSRIs. Only r<strong>and</strong>omized <strong>and</strong> double<br />
blinded conducted in patients with major depressive<br />
disorder were included. Outcomes were measured using<br />
HAMD. Duration <strong>of</strong> treatment ranged between 4 <strong>and</strong><br />
12 weeks in various studies. The methodological quality <strong>of</strong><br />
included trials was assessed using <strong>the</strong> Jadad score. 13 eligible<br />
controlled clinic trials were included in <strong>the</strong> meta-analysis.<br />
Comparison <strong>of</strong> SSRIs with placebo yielded a significant<br />
relative risk (RR) <strong>of</strong> 1.22 (95% CI: 1.03–1.45) <strong>for</strong> clinical<br />
response, whereas a non significant RR <strong>of</strong> 0.96 (95% CI=0.71–<br />
1.29) was observed <strong>for</strong> remission.<br />
In this meta-analysis <strong>Hypericum</strong> was also shown to be at<br />
least as effective as SSRI. The authors, by comparing<br />
<strong>Hypericum</strong> <strong>extract</strong> with SSRI, found that rate <strong>of</strong> clinical<br />
response (RR=0.99 95% CI=0.91–1.08), remission (RR=1.10<br />
95% CI=0.90–1.35), <strong>and</strong> reduction <strong>of</strong> HAMD score (RR=0.32<br />
95% CI=−1.20–0.64), was similar between <strong>Hypericum</strong> <strong>and</strong><br />
SSRIs. Lower withdrawal due to adverse events was observed<br />
in patients treated with <strong>Hypericum</strong> <strong>extract</strong> compared to SSRI<br />
(RR=0.53 95% CI=0.35–0.82).<br />
The results obtained from this meta-analysis demonstrate<br />
both a similar efficacy <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>and</strong> SSRI <strong>and</strong> a<br />
better pr<strong>of</strong>ile <strong>of</strong> <strong>tolerability</strong> <strong>for</strong> <strong>Hypericum</strong>, finally suggesting<br />
that <strong>Hypericum</strong> <strong>extract</strong> might be considered as an<br />
alternative to SSRIs in <strong>the</strong> management <strong>of</strong> mild to moderate<br />
depression.<br />
4. Tolerability with <strong>Hypericum</strong> <strong>extract</strong>: evidence<br />
from meta-analyses <strong>and</strong> systematic reviews<br />
The absence <strong>of</strong> adverse effects is an important factor in <strong>the</strong><br />
treatment <strong>of</strong> patients affected by mild to moderate depression,<br />
given that <strong>the</strong>se patients are less likely to accept drugs<br />
with undesirable effects than patients with more severe <strong>for</strong>ms<br />
<strong>of</strong> depression. Poor <strong>tolerability</strong> has been shown to have a<br />
negative impact on compliance, with <strong>the</strong> potential <strong>for</strong><br />
worsening <strong>of</strong> depressive symptoms (Keller et al., 2002). In<br />
general, <strong>Hypericum</strong> <strong>extract</strong>s have a favourable <strong>tolerability</strong><br />
pr<strong>of</strong>ile <strong>and</strong> are generally better tolerated than st<strong>and</strong>ard<br />
antidepressant drugs (Ernst et al., 1998; Rahimi et al., 2009).<br />
The meta-analysis by Röder et al. (2004) found a lower<br />
incidence <strong>of</strong> adverse events with <strong>Hypericum</strong> <strong>extract</strong> (24.5%)<br />
compared to that seen with syn<strong>the</strong>tic antidepressants<br />
(32.8%); although this did not reach significance (p=.09),<br />
<strong>and</strong> a similar number <strong>of</strong> withdrawals due to adverse effects<br />
(p=.2). Ano<strong>the</strong>r methodologically sound overview <strong>of</strong> <strong>the</strong><br />
adverse effects <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> comes from a<br />
systematic review by Knuppel <strong>and</strong> Linde (2004). Part <strong>of</strong> this<br />
review was per<strong>for</strong>med in <strong>the</strong> context <strong>of</strong> updating <strong>the</strong> metaanalysis<br />
<strong>of</strong> Linde, et al. discussed above (Linde et al., 2005a;<br />
Linde <strong>and</strong> Knuppel, 2005). Overall, <strong>the</strong> systematic review<br />
examined dropout rates <strong>and</strong> adverse effects in 24 doubleblind,<br />
r<strong>and</strong>omized trials, 17 large-scale observational studies<br />
as well as numerous case reports, case series <strong>and</strong> data sets<br />
from surveillance agencies (Knuppel <strong>and</strong> Linde, 2004). Based<br />
on <strong>the</strong> results <strong>of</strong> this review, <strong>the</strong> authors made several<br />
statements regarding <strong>the</strong> rates <strong>of</strong> adverse effects <strong>and</strong><br />
dropouts in patients treated with <strong>Hypericum</strong> <strong>extract</strong>. Firstly,<br />
in r<strong>and</strong>omized trials, rates were similar <strong>for</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> <strong>and</strong> placebo, lower <strong>for</strong> <strong>Hypericum</strong> <strong>extract</strong> compared<br />
with tricyclic antidepressants <strong>and</strong> slightly lower <strong>for</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> compared with SSRIs. In non-r<strong>and</strong>omized<br />
observational studies, rates <strong>of</strong> discontinuation due to<br />
adverse effects among <strong>Hypericum</strong> <strong>extract</strong>-treated patients<br />
are low even in long-term studies <strong>and</strong> effects are generally<br />
mild.<br />
Several longer-term studies on <strong>Hypericum</strong> <strong>extract</strong> have<br />
also been completed, which provide an assessment <strong>of</strong> <strong>the</strong>
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
<strong>tolerability</strong> <strong>of</strong> this agent over a longer period <strong>of</strong> follow-up.<br />
Data from four large post-marketing surveillance studies<br />
with a total <strong>of</strong> 14, 212 patients taking various preparations <strong>of</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> have been published (Lemmer et al.,<br />
1999; Rychlik et al., 2001; Schakau et al., 1996; Woelk et al.,<br />
1994). In <strong>the</strong>se studies, <strong>the</strong> overall incidence <strong>of</strong> side effects<br />
was 0.1–2.4% with 0.1–0.9% <strong>of</strong> patients discontinuing due to<br />
drug-related adverse events (Schulz, 2006).<br />
Linde <strong>and</strong> Knuppel (2005) published a systematic review<br />
<strong>of</strong> 16 large-scale, open-label observational studies on<br />
<strong>Hypericum</strong> <strong>extract</strong>, including an assessment <strong>of</strong> methodological<br />
quality, involving a total <strong>of</strong> 34,804 patients. Two studies<br />
were long-term (52-week) observational studies including<br />
<strong>the</strong> WS 5572 study (Lemmer et al., 1999) in mild to moderate<br />
depression reviewed by Schulz (2006), <strong>and</strong> a study <strong>of</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> Laif 600 in depressive disorders, published<br />
in abstract <strong>for</strong>m (Zeller, 2000). In <strong>the</strong>se two studies,<br />
3.4% <strong>and</strong> 5.7% <strong>of</strong> patients discontinued due to adverse<br />
effects. All adverse events reported with <strong>Hypericum</strong> <strong>extract</strong><br />
were mild <strong>and</strong> <strong>the</strong> most common were gastrointestinal<br />
symptoms, light sensitivity <strong>and</strong> o<strong>the</strong>r skin problems.<br />
In a 1-year, open-label study on 313 patients with mild to<br />
moderate depression according to <strong>the</strong> DSM-IV scale, <strong>the</strong><br />
<strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> was rated as good or very<br />
good after one year by both patients <strong>and</strong> physicians (Hubner<br />
<strong>and</strong> Arndt, 2000). At <strong>the</strong> end <strong>of</strong> <strong>the</strong> 12-month follow-up<br />
period, <strong>the</strong>re were no significant differences between <strong>the</strong><br />
initial <strong>and</strong> final parameters concerning EEG, blood pressure,<br />
heart rate or haematological parameters. Only 11.2% <strong>of</strong><br />
cases reported undesirable events during <strong>the</strong> entire year <strong>of</strong><br />
<strong>the</strong>rapy, <strong>the</strong> most frequently reported adverse events being<br />
infections <strong>of</strong> <strong>the</strong> upper airway, headache <strong>and</strong> gastrointestinal<br />
disturbances.<br />
Acute mania was reported in one patient taking 1800 mg<br />
<strong>Hypericum</strong> <strong>extract</strong> in a r<strong>and</strong>omized, double-blind comparison<br />
with sertraline (van Gurp et al., 2002). Several o<strong>the</strong>r<br />
cases <strong>of</strong> acute mania have been reported (Fahmi et al., 2002;<br />
Guzelcan et al., 2001; Moses <strong>and</strong> Mallinger, 2000; Nierenberg<br />
et al., 1999). No causal relationship between <strong>the</strong> use <strong>of</strong> <strong>the</strong><br />
<strong>extract</strong> <strong>and</strong> <strong>the</strong> mania has been established but <strong>the</strong><br />
propensity <strong>for</strong> affective switching, as seen with syn<strong>the</strong>tic<br />
antidepressant use, cannot be ruled out. <strong>Hypericum</strong> <strong>extract</strong>s<br />
should be <strong>the</strong>re<strong>for</strong>e avoided in patients with bipolar<br />
disorder.<br />
Phototoxicity has been raised as a potential adverse<br />
effect <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>. In <strong>the</strong> <strong>of</strong>ficial register <strong>of</strong><br />
spontaneous events (Schulz, 2001), only 1 case per 300,000<br />
treated cases has been reported. Volunteer studies have<br />
shown a non-clinically relevant increase in photosensitivity,<br />
only detectable using highly sensitive photometric measurement.<br />
This might become relevant, but only in fair-skinned<br />
individuals, in those with skin disorders or after extended<br />
solar irradiation (Schempp et al., 2000) <strong>and</strong> most individuals<br />
would be highly unlikely to experience problems. Investigations<br />
in volunteers have also shown that <strong>the</strong> threshold dose<br />
<strong>for</strong> an increased risk <strong>of</strong> photosensitization is about 2–4 g/day<br />
<strong>of</strong> a usual commercial <strong>extract</strong> (equivalent to approximately<br />
5–10 mg <strong>of</strong> hypericin), considerably more than doses used in<br />
clinical practice (Schulz, 2001). It has been estimated that a<br />
<strong>Hypericum</strong> <strong>extract</strong> dose 30–50 times greater than <strong>the</strong><br />
recommended daily dose taken at one time would be<br />
required to cause severe phototoxic reactions (Ernst et al.,<br />
1998). Indeed, only few cases have been reported with<br />
respect to a serious phototoxicity reaction necessitating<br />
treatment cessation (Bove, 1998; Golsch et al., 1997; Lane-<br />
Brown, 2000). In one, a woman developed recurrent<br />
ery<strong>the</strong>matous lesions in areas exposed to light following<br />
consumption <strong>of</strong> a commercial <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> 3 years.<br />
The condition resolved after discontinuation <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> (Golsch et al., 1997). In ano<strong>the</strong>r case, a woman<br />
experienced neuropathy after taking <strong>Hypericum</strong> <strong>extract</strong><br />
(Bove, 1998) <strong>and</strong> three cases <strong>of</strong> severe blistering <strong>and</strong> burns<br />
were reported in patients taking <strong>Hypericum</strong> <strong>extract</strong>s (ei<strong>the</strong>r<br />
internally or topically) be<strong>for</strong>e exposure to sunlight (Lane-<br />
Brown, 2000).<br />
According to Brockmoller et al. (1997), given that<br />
photosensitivity is unlikely at doses <strong>of</strong> up to 1800 mg/day,<br />
potential phototoxic effects may only become apparent<br />
when <strong>Hypericum</strong> <strong>extract</strong> is taken with o<strong>the</strong>r drugs with<br />
phototoxic effects (Schulz, 2001).<br />
5. Compliance<br />
761<br />
Compliance with antidepressant medication is essential to<br />
ensure treatment response <strong>and</strong> prevent relapse <strong>and</strong> symptom<br />
recurrence (Keller et al., 2002). However, compliance<br />
with antidepressant medication is poor, with one set <strong>of</strong><br />
published data indicating that between 30% <strong>and</strong> 60% <strong>of</strong><br />
patients do not take <strong>the</strong>ir medications as prescribed<br />
(Demyttenaere et al., 2004), a figure which agrees broadly<br />
with o<strong>the</strong>r commonly quoted estimates <strong>of</strong> antidepressant<br />
non-compliance (Olivier-Martin, 1986). The rate <strong>of</strong> onset <strong>and</strong><br />
poor <strong>tolerability</strong> <strong>of</strong> antidepressant drugs has a considerable<br />
influence on patient compliance. To illustrate <strong>the</strong> point, one<br />
US study that examined claims data among 2012 patients<br />
found that initiating treatment with a tricyclic antidepressant<br />
reduced <strong>the</strong> probability <strong>of</strong> antidepressant treatment<br />
compliance vs. SSRI plus psycho<strong>the</strong>rapy (Tai-Seale et al.,<br />
2000). In contrast, <strong>Hypericum</strong> <strong>extract</strong> is associated with a<br />
good <strong>tolerability</strong> pr<strong>of</strong>ile <strong>and</strong> good compliance. In a placebocontrolled<br />
study <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> (ZE 117) involving 162<br />
patients, an electronic counter was inserted in <strong>the</strong> medicinal<br />
container in order to evaluate compliance (Schrader, 2000).<br />
A high level <strong>of</strong> compliance was observed, with a <strong>the</strong>rapeutic<br />
coverage <strong>of</strong> 81.7% noted. Adherence to <strong>the</strong>rapy was<br />
attributed, at least in part, to a low number <strong>of</strong> undesirable<br />
effects (5 in <strong>the</strong> placebo group <strong>and</strong> 6 in <strong>the</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> group), generally limited to acute gastrointestinal<br />
complaints. O<strong>the</strong>r investigators have suggested that <strong>the</strong><br />
superior <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>s relative to o<strong>the</strong>r<br />
antidepressants, particularly tricyclics, could account <strong>for</strong><br />
advantages in terms <strong>of</strong> compliance (Wheatley, 1997).<br />
Finally a recent trial from Kasper et al. (2008b) has<br />
demonstrated that long-term treatment with <strong>Hypericum</strong><br />
<strong>extract</strong> is not associated with any unexpected drug-specific<br />
risks or problems <strong>of</strong> intolerance (Kasper et al., 2008b).<br />
Fur<strong>the</strong>rmore <strong>the</strong> rates <strong>of</strong> potentially attributable events<br />
reported during double-blind treatment with <strong>Hypericum</strong><br />
were lower than in <strong>the</strong> placebo group. Recent meta-analyses<br />
have also clearly demonstrated that treatment with <strong>Hypericum</strong><br />
<strong>extract</strong> is associated with a lower withdrawal due to<br />
adverse events when compared to st<strong>and</strong>ard antidepressant<br />
drugs, thus suggesting an advantage <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> in
762 S. Kasper et al.<br />
terms <strong>of</strong> <strong>tolerability</strong> <strong>and</strong> compliance (Linde et al., 2008;<br />
Rahimi et al., 2009).<br />
6. Drug interactions with <strong>Hypericum</strong> <strong>extract</strong><br />
Several systematic reviews have examined <strong>the</strong> published<br />
literature on <strong>Hypericum</strong> <strong>extract</strong>-mediated drug interactions<br />
(Henderson et al., 2002; Izzo, 2004; Mills et al., 2004; Whitten<br />
et al., 2006). Izzo (2004) published a review <strong>of</strong> <strong>the</strong> clinical<br />
evidence related to drug interactions with <strong>Hypericum</strong> <strong>extract</strong>s<br />
which pulls toge<strong>the</strong>r many small studies <strong>and</strong> case reports.<br />
Interactions <strong>of</strong> clinical relevance <strong>of</strong>ten relate to <strong>the</strong> potential<br />
<strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>s <strong>for</strong> inducing cytochrome P450 enzymes<br />
(particularly CYP 3A4) <strong>and</strong> also increasing expression <strong>of</strong> Pglycoprotein<br />
(Pgp). Current evidence suggests that <strong>the</strong>se<br />
effects can be due to hyper<strong>for</strong>in, one <strong>of</strong> <strong>the</strong> main constituents<br />
<strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong>, which increases <strong>the</strong> expression <strong>of</strong> <strong>the</strong><br />
pregnane X receptor, <strong>and</strong> <strong>the</strong>n induces Pgp expression as well<br />
as CYP 3A4 activity (Moore et al., 2000). Moreover pharmacodynamic<br />
interactions have been described between <strong>Hypericum</strong><br />
<strong>extract</strong> <strong>and</strong> o<strong>the</strong>r antidepressants (e.g. serotonin<br />
reuptake inhibitors such as paroxetine <strong>and</strong> sertraline) with<br />
<strong>the</strong> possible occurrence <strong>of</strong> a serotonin syndrome (Zhou <strong>and</strong><br />
Lai, 2008).<br />
Izzo (2004) have reviewed <strong>the</strong> data relating to interactions<br />
with oral contraceptives (OCs); several cases have been<br />
reported in which women have had breakthrough bleeding<br />
while taking <strong>Hypericum</strong> <strong>extract</strong> with OCs. In two open-label<br />
post-marketing surveillance studies (Hall et al., 2003;<br />
Pfrunder et al., 2003) <strong>of</strong> concomitant OC <strong>and</strong> <strong>Hypericum</strong><br />
<strong>extract</strong> use over 3 months, <strong>the</strong>re was no loss <strong>of</strong> contraceptive<br />
control; endogenous hormone levels showed an absence <strong>of</strong><br />
ovulation in all patients taking <strong>Hypericum</strong> <strong>extract</strong> throughout<br />
<strong>the</strong> study period.<br />
A systematic review was conducted by Whitten et al.<br />
(2006) to examine <strong>the</strong> quality <strong>and</strong> outcomes <strong>of</strong> clinical trials<br />
analyzing <strong>the</strong> effect <strong>of</strong> St. John's wort <strong>extract</strong>s on <strong>the</strong><br />
metabolism <strong>of</strong> drugs by CYP3A. All <strong>of</strong> <strong>the</strong> 19 studies that used<br />
high-dose hyper<strong>for</strong>in <strong>extract</strong>s (N10 mg =day) had outcomes<br />
consistent with CYP3A induction, while <strong>the</strong> three studies<br />
using low-dose hyper<strong>for</strong>in <strong>extract</strong>s (b4 mg=day) demonstrated<br />
no significant effect on CYP3A (Whitten et al.,<br />
2006).<br />
7. Conclusion<br />
Whe<strong>the</strong>r examined individually or combined in quantitative<br />
meta-analyses, <strong>the</strong> results <strong>of</strong> numerous clinical studies<br />
strongly support <strong>the</strong> effectiveness <strong>of</strong> some <strong>Hypericum</strong><br />
<strong>extract</strong>s <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression.<br />
Study methodology <strong>and</strong> quality <strong>of</strong> reporting have increased<br />
in recent years, which should make prescribers <strong>and</strong><br />
patients more confident in choosing <strong>Hypericum</strong> <strong>extract</strong><br />
over conventional antidepressants. Although a small<br />
number <strong>of</strong> comparative trials have shown <strong>Hypericum</strong><br />
<strong>extract</strong> to be effective in severe depression, <strong>the</strong> majority<br />
do not support this view, <strong>and</strong>, <strong>the</strong>re<strong>for</strong>e, weighing current<br />
evidence, <strong>Hypericum</strong> <strong>extract</strong> is not recommended <strong>for</strong><br />
patients with severe depression. There is some evidence<br />
that <strong>Hypericum</strong> <strong>extract</strong> is effective in controlling symp-<br />
toms in patients with disorders associated with or related<br />
to depression such as anxiety (Muller et al., 2003),<br />
somat<strong>of</strong>orm disorders (Muller et al., 2004; Volz et al.,<br />
2002; Woelk, 2000), <strong>and</strong> obsessive compulsive disorder<br />
(Taylor <strong>and</strong> Kobak, 2000).<br />
Overall, current evidence appears to support <strong>the</strong><br />
efficacy <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild<br />
to moderate depression; meta-analyses <strong>of</strong> clinical trials<br />
suggest that <strong>Hypericum</strong> <strong>extract</strong> is significantly more<br />
effective than placebo <strong>and</strong> at least as effective as<br />
st<strong>and</strong>ard antidepressants. Those <strong>extract</strong>s which demonstrate<br />
efficacy in clinical trials can <strong>and</strong> should be<br />
recommended <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate<br />
depression. Based on our review <strong>of</strong> individual clinical<br />
trials, WS 5572, LI 160, WS 5570 <strong>and</strong> ZE 117 <strong>Hypericum</strong><br />
<strong>extract</strong>s have been shown to be significantly more<br />
effective than placebo with similar efficacy to st<strong>and</strong>ard<br />
antidepressants. Initial studies <strong>of</strong> <strong>extract</strong>s LoHyp-57,<br />
psychotonin, calmigen, Swiss Herbal Remedies, STW3,<br />
STEI 300 <strong>and</strong> esbericum have produced promising results<br />
but fur<strong>the</strong>r clinical data are needed to confirm <strong>the</strong>ir<br />
effectiveness in mild to moderate depression.<br />
Considering <strong>the</strong> favourable safety pr<strong>of</strong>ile <strong>of</strong> all <strong>extract</strong>s<br />
studied (equivalent to placebo <strong>and</strong> better tolerated than<br />
st<strong>and</strong>ard antidepressants), which also correlated with a<br />
lower frequency <strong>of</strong> dropouts in clinical studies than seen<br />
with st<strong>and</strong>ard antidepressants, <strong>Hypericum</strong> <strong>extract</strong> is a<br />
valid <strong>the</strong>rapeutic approach in patients with mild to<br />
moderate depression. For <strong>the</strong>se patients, <strong>for</strong> whom<br />
adverse effects <strong>of</strong> treatment may be difficult to accept,<br />
<strong>Hypericum</strong> <strong>extract</strong>s provide a high rate <strong>of</strong> pharmacological<br />
compliance. They provide an effective <strong>and</strong> well tolerated<br />
alternative to st<strong>and</strong>ard tricyclic <strong>and</strong> SSRI antidepressants<br />
<strong>and</strong> are advocated <strong>for</strong> first line <strong>the</strong>rapy <strong>of</strong> mild to<br />
moderate depression.<br />
Role <strong>of</strong> <strong>the</strong> funding source<br />
This study was not supported by external funding.<br />
Contributors<br />
All authors have been actively involved in <strong>the</strong> construction <strong>of</strong> <strong>the</strong><br />
paper from <strong>the</strong> outset <strong>and</strong> have had full editorial control over its<br />
content.<br />
Conflict <strong>of</strong> interest<br />
Siegfreid Kasper has received grant/research support from Eli Lilly,<br />
Lundbeck, Bristol-Myers Squibb, GlaxoSmithKline, Organon, Sepracor<br />
<strong>and</strong> Servier; has served as a consultant or on advisory boards <strong>for</strong><br />
AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Eli Lilly,<br />
Lundbeck, Pfizer, Organon, Schwabe, Sepracor, Servier, Janssen,<br />
<strong>and</strong> Novartis; <strong>and</strong> has served on speakers' bureaus <strong>for</strong> AstraZeneca,<br />
Eli Lilly, Lundbeck, Schwabe, Sepracor, Servier <strong>and</strong> Janssen.<br />
Filippo Caraci has received grants or honoraria as a consultant<br />
from Eli Lilly, Janssen, <strong>and</strong> Innova Pharma.<br />
Bruno Forti <strong>and</strong> Eugenio Aguglia declare that <strong>the</strong>y have no<br />
conflict <strong>of</strong> interests.<br />
Filippo Drago has served as a consultant or on advisory boards <strong>for</strong><br />
Eli Lilly, GlaxoSmithKline, AstraZeneca, Novartis, Pfizer, SIFI,<br />
Amdipharm, Grunenthal.
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
References<br />
Aguglia, E., Forti, B., 2001. La depressione fra normalità e malattia.<br />
Aspetti clinici, sociali e terapeutici dei disturbi depressivi<br />
sottosoglia. Il Pensiero Scientifico Editore, Roma.<br />
Battaglia, A., Dubini, A., Mannheimer, R., Pancheri, P., 2004.<br />
Depression in <strong>the</strong> Italian community: epidemiology <strong>and</strong> socioeconomic<br />
implications. Int. Clin. Psychopharmacol. 19, 135–142.<br />
Behnke, K., Jensen, G.S., Graubaum, H.J., Gruenwald, J., 2002.<br />
<strong>Hypericum</strong> per<strong>for</strong>atum versus fluoxetine in <strong>the</strong> treatment <strong>of</strong> mild<br />
to moderate depression. Adv. Ther. 19, 43–52.<br />
Bergmann, R., Nussner, H., Demling, J., 1993. Beh<strong>and</strong>lung leichter<br />
bis mittelschwerer depressionen. Therapiewoche Neurol. Psychiatr.<br />
7, 235–240.<br />
Bjerkenstedt, L., Edman, G.V., Alken, R.G., Mannel, M., 2005.<br />
<strong>Hypericum</strong> <strong>extract</strong> LI 160 <strong>and</strong> fluoxetine in mild to moderate<br />
depression: a r<strong>and</strong>omized, placebo-controlled multi-center study<br />
in outpatients. Eur. Arch. Psychiatry Clin. Neurosci. 255, 40–47.<br />
Bove, G.M., 1998. Acute neuropathy after exposure to sun in a<br />
patient treated with St John's wort. Lancet 352, 1121–1122.<br />
Brenner, R., Azbel, V., Madhusoodanan, S., Pawlowska, M., 2000.<br />
Comparison <strong>of</strong> an <strong>extract</strong> <strong>of</strong> <strong>Hypericum</strong> (LI 160) <strong>and</strong> sertraline in<br />
<strong>the</strong> treatment <strong>of</strong> depression: a double-blind, r<strong>and</strong>omized pilot<br />
study. Clin. Ther. 22, 411–419.<br />
Brockmoller, J., Reum, T., Bauer, S., Kerb, R., Hübner, W.D., Roots,<br />
I., 1997. Hypericin <strong>and</strong> pseudohypericin: pharmacoki<strong>net</strong>ics <strong>and</strong><br />
effects on photosensitivity in humans. Pharmacopsychiatry 30<br />
(Suppl 2), 94–101.<br />
Butterweck, V., 2003. Mechanism <strong>of</strong> action <strong>of</strong> St John's wort in<br />
depression: what is known? CNS Drugs 17, 539–562.<br />
Butterweck, V., Schmidt, M., 2007. St. John's wort: role <strong>of</strong> active<br />
compounds <strong>for</strong> its mechanism <strong>of</strong> action <strong>and</strong> efficacy. Wien. Med.<br />
Wochenschr. 157, 356–361.<br />
Cott, J., Wisner, K.L., 2002. St John's wort <strong>and</strong> depression. JAMA<br />
288, 448 author reply 448–449.<br />
Demyttenaere, K., Bruffaerts, R., Albert, A., Mesters, P., Dewé, W.,<br />
Debruyckere, K., Sangeleer, M., 2004. Development <strong>of</strong> an<br />
antidepressant compliance questionnaire. Acta Psychiatr.<br />
Sc<strong>and</strong>. 110, 201–207.<br />
Ernst, E., R<strong>and</strong>, J.I., Barnes, J., Stevinson, C., 1998. Adverse effects<br />
pr<strong>of</strong>ile <strong>of</strong> <strong>the</strong> herbal antidepressant St. John's wort (<strong>Hypericum</strong><br />
per<strong>for</strong>atum L.). Eur. J. Clin. Pharmacol. 54, 589–594.<br />
Fahmi, M., Huang, C., Schweitzer, I., 2002. A case <strong>of</strong> mania induced<br />
by <strong>Hypericum</strong>. World J. Biol. Psychiatry 3, 58–59.<br />
Fava, G.A., Ruini, C., Sonino, N., 2003. Treatment <strong>of</strong> recurrent<br />
depression: a sequential psycho<strong>the</strong>rapeutic <strong>and</strong> psychopharmacological<br />
approach. CNS Drugs 17, 1109–1117.<br />
Fava, M., Alpert, J., Nierenberg, A.A., Mischoulon, D., Otto, M.W.,<br />
Zajecka, J., Murck, H., Rosenbaum, J.F., 2005. A double-blind,<br />
r<strong>and</strong>omized trial <strong>of</strong> St. John's wort, fluoxetine, <strong>and</strong> placebo in<br />
major depressive disorder. J. Clin. Psychopharmacol. 25, 441–447.<br />
Franklin, M., Hafizi, S., Reed, A., Hockney, R., Murck, H., 2006.<br />
Effect <strong>of</strong> sub-chronic treatment with Jarsin (<strong>extract</strong> <strong>of</strong> St John's<br />
wort, <strong>Hypericum</strong> per<strong>for</strong>atum) at two dose levels on evening<br />
salivary melatonin <strong>and</strong> cortisol concentrations in healthy male<br />
volunteers. Pharmacopsychiatry 39, 13–15.<br />
Gastpar, M., Singer, A., Zeller, K., 2005. <strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> STW3 in long-term treatment with a oncedaily<br />
dosage in comparison with sertraline. Pharmacopsychiatry<br />
38, 78–86.<br />
Gastpar, M., Singer, A., Zeller, K., 2006. Comparative efficacy <strong>and</strong><br />
safety <strong>of</strong> a once-daily dosage <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> STW3-VI <strong>and</strong><br />
citalopram in patients with moderate depression: a double-blind,<br />
r<strong>and</strong>omised, multicentre, placebo-controlled study. Pharmacopsychiatry<br />
39, 66–75.<br />
Geddes, J.R., Carney, S.M., Davies, C., Furukawa, T.A., Kupfer, D.J.,<br />
Frank, E., Goodwin, G.M., 2003. Relapse prevention with<br />
763<br />
antidepressant drug treatment in depressive disorders: a systematic<br />
review. Lancet 361, 653–661.<br />
Golsch, S., Vocks, E., Rakoski, J., Brockow, K., Ring, J., 1997.<br />
Reversible increase in photosensitivity to UV-B caused by St.<br />
John's wort <strong>extract</strong>. Hautarzt 48, 249–252.<br />
Grundmann, O., Lv, Y., Kelber, O., Butterweck, V., 2010.<br />
Mechanism <strong>of</strong> St. John's wort <strong>extract</strong> (STW3-VI) during chronic<br />
restraint stress is mediated by <strong>the</strong> interrelationship <strong>of</strong> <strong>the</strong><br />
immune, oxidative defense, <strong>and</strong> neuroendocrine system. Neuropharmacology<br />
58, 767–773.<br />
Guzelcan, Y., Scholte, W.F., Assies, J., Becker, H.E., 2001. Mania<br />
during <strong>the</strong> use <strong>of</strong> a combination preparation with St. John's wort<br />
(<strong>Hypericum</strong> per<strong>for</strong>atum). Ned. Tijdschr. Geneeskd 145,<br />
1943–1945.<br />
Hall, S.D., Wang, Z., Huang, S.M., Hamman, M.A., Vasavada, N.,<br />
Adigun, A.Q., Hilligoss, J.K., Miller, M., Gorski, J.C., 2003. The<br />
interaction between St John's wort <strong>and</strong> an oral contraceptive.<br />
Clin. Pharmacol. Ther. 74, 525–535.<br />
Hänsgen, K.D., Vesper, J., 1996. Antidepressive wirksamkeit eines<br />
hochdosierten <strong>Hypericum</strong>-extraktes. Münch. Med. Wochenschift<br />
138, 29–33.<br />
Harrer, G., Hubner, W.D., Podzuweit, H., 1994. Effectiveness <strong>and</strong><br />
tolerance <strong>of</strong> <strong>the</strong> <strong>Hypericum</strong> <strong>extract</strong> LI 160 compared to maprotiline:<br />
a multicenter double-blind study. J. Geriatr. Psychiatry<br />
Neurol. 7 (Suppl 1), S24–S28.<br />
Harrer, G., Schmidt, U., Kuhn, U., Biller, A., 1999. Comparison <strong>of</strong><br />
equivalence between <strong>the</strong> St. John's wort <strong>extract</strong> LoHyp-57 <strong>and</strong><br />
fluoxetine. Arzneimittel<strong>for</strong>schung 49, 289–296.<br />
Henderson, L., Yue, Q., Bergquist, C., Gerden, B., Arlett, P., 2002.<br />
St John's wort (<strong>Hypericum</strong> per<strong>for</strong>atum): drug interactions <strong>and</strong><br />
clinical outcomes. Br. J. Clin. Pharmacol. 54, 349–356.<br />
Hübner, W.D., L<strong>and</strong>e, S., Podzuweit, H., 1994. <strong>Hypericum</strong> treatment<br />
<strong>of</strong> mild depressions with somatic symptoms. J Geriatr.<br />
Psychiatry Neurol. S1, S12–S14.<br />
Hubner, W.D., Arndt, K.H., 2000. Johanniskraut-Einjahresstudie:<br />
Offene multizentrische Therapiestudie mit Johanniskraut-<br />
Extrakt LI 160 bei Patienten mit depressiven Erkrankungen.<br />
Zeitschr. f. Phyto<strong>the</strong>rapie. 21, 306–310.<br />
<strong>Hypericum</strong> Depression Trial Study Group, 2002. Effect <strong>of</strong> <strong>Hypericum</strong><br />
per<strong>for</strong>atum (St John's wort) in major depressive disorder: a<br />
r<strong>and</strong>omized controlled trial. JAMA 287, 1807–1814.<br />
Izzo, A.A., 2004. Drug interactions with St. John's wort (<strong>Hypericum</strong><br />
per<strong>for</strong>atum): a review <strong>of</strong> <strong>the</strong> clinical evidence. Int. J. Clin.<br />
Pharmacol. Ther. 42, 139–148.<br />
Jadad, A.R., Moore, R.A., Carroll, D., Jenkinson, C., Reynolds, D.J.,<br />
Gavaghan, D.J., McQuay, H.J., 1996. Assessing <strong>the</strong> quality <strong>of</strong><br />
reports <strong>of</strong> r<strong>and</strong>omized clinical trials: is blinding necessary?<br />
Control. Clin. Trials 17, 1–12.<br />
Kalb, R., Trautmann-Sponsel, R.D., Kieser, M., 2001. <strong>Efficacy</strong> <strong>and</strong><br />
<strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> WS 5572 versus placebo in<br />
mildly to moderately depressed patients. A r<strong>and</strong>omized doubleblind<br />
multicenter clinical trial. Pharmacopsychiatry 34, 96–103.<br />
Kasper, S., Dienel, A., 2002. Cluster analysis <strong>of</strong> symptoms during<br />
antidepressant treatment with <strong>Hypericum</strong> <strong>extract</strong> in mildly to<br />
moderately depressed out-patients. A meta-analysis <strong>of</strong> data from<br />
three r<strong>and</strong>omized, placebo-controlled trials. Psychopharmacology<br />
(Berl) 164, 301–308.<br />
Kasper, S., Dienel, A., Kieser, M., 2004. Continuation <strong>and</strong> long-term<br />
maintenance treatment with <strong>Hypericum</strong> <strong>extract</strong> WS 5570 after<br />
successful acute treatment <strong>of</strong> mild to moderate depression—<br />
rationale<strong>and</strong>studydesign.Int.J.Meth.Psychiatr.Res.13,176–183.<br />
Kasper, S., Anghelescu, I.G., Szegedi, A., Dienel, A., Kieser, M.,<br />
2006. Superior efficacy <strong>of</strong> St John's wort <strong>extract</strong> WS 5570<br />
compared to placebo in patients with major depression: a<br />
r<strong>and</strong>omized, double-blind, placebo-controlled, multi-center<br />
trial [ISRCTN77277298]. BMC Med. 4, 14.<br />
Kasper, S., Anghelescu, I.G., Szegedi, A., Dienel, A., Kieser, M., 2007.<br />
Placebo controlled continuation treatment with <strong>Hypericum</strong> <strong>extract</strong>
764 S. Kasper et al.<br />
WS 5570 after recovery from a mild or moderate depressive<br />
episode. Wien. Med. Wochenschr. 157, 362–366.<br />
Kasper, S., Gastpar, M., Müller, W.E., Volz, H.P., Dienel, A., Kieser,<br />
M., Möller, H.J., 2008a. <strong>Efficacy</strong> <strong>of</strong> St. John's wort <strong>extract</strong> WS<br />
5570 in acute treatment <strong>of</strong> mild depression: a reanalysis <strong>of</strong> data<br />
from controlled clinical trials. Eur. Arch. Psychiatr. Clin.<br />
Neurosci. 258, 59–63.<br />
Kasper, S., Volz, H.P., Möller, H.J., Dienel, A., Kieser, M., 2008b.<br />
Continuation <strong>and</strong> long-term maintenance treatment with <strong>Hypericum</strong><br />
<strong>extract</strong> WS 5570 after recovery from an acute episode <strong>of</strong><br />
moderate depression—a double-blind, r<strong>and</strong>omized, placebo<br />
controlled long-term trial. Eur. Neuropsychopharmacol. 18,<br />
803–813.<br />
Keller, M.B., Hirschfeld, R.M., Demyttenaere, K., Baldwin, D.S.,<br />
2002. Optimizing outcomes in depression: focus on antidepressant<br />
compliance. Int. Clin. Psychopharmacol. 17, 265–271.<br />
Kim, H.L., Streltzer, J., Goebert, D., 1999. St. John's wort <strong>for</strong><br />
depression: a meta-analysis <strong>of</strong> well-defined clinical trials. J.<br />
Nerv. Ment. Dis. 187, 532–538.<br />
Knuppel, L., Linde, K., 2004. Adverse effects <strong>of</strong> St. John's wort: a<br />
systematic review. J. Clin. Psychiatry 65, 1470–1479.<br />
Krishnan, V., Nestler, E.J., 2008. The molecular neurobiology <strong>of</strong><br />
depression. Nature 455, 894–902.<br />
Kumar, A., Garg, R., Prakash, A.K., 2010. Effect <strong>of</strong> St. John's wort<br />
(<strong>Hypericum</strong> per<strong>for</strong>atum) treatment on restraint stress-induced<br />
behavioral <strong>and</strong> biochemical alteration in mice. BMC Complement.<br />
Altern. Med. 10, 18.<br />
Laakmann, G., Schule, C., Baghai, T., Kieser, M., 1998. St. John's<br />
wort in mild to moderate depression: <strong>the</strong> relevance <strong>of</strong> hyper<strong>for</strong>in<br />
<strong>for</strong> <strong>the</strong> clinical efficacy. Pharmacopsychiatry 31 (Suppl 1), 54–59.<br />
Lane-Brown, M.M., 2000. Photosensitivity associated with herbal<br />
preparations <strong>of</strong> St John's wort (<strong>Hypericum</strong> per<strong>for</strong>atum). Med. J.<br />
Aust. 172, 302.<br />
Lecrubier, Y., Clerc, G., Didi, R., Kieser, M., 2002. <strong>Efficacy</strong> <strong>of</strong> St.<br />
John's wort <strong>extract</strong> WS 5570 in major depression: a double-blind,<br />
placebo-controlled trial. Am. J. Psychiatry 159, 1361–1366.<br />
Lehrl, S., Woelk, H., 1993. Ergebnisse von messungen der kognitiven<br />
leistungsfähigkeit bei patienten unter der <strong>the</strong>rapie mit johanniskraut.<br />
Nervenheilkunde 12, 281–284.<br />
Lemmer, W., von den Driesch, V., Klieser, E., 1999. Johanniskrautspezialextrakt<br />
WS 5572 bei leichter bis mittelschwerer depression.<br />
Wirksamkeit und vertraglichkeit von neuroplant 300<br />
filmbletten. Fortschr. Med. 117, 143–154.<br />
Linde, K., Knuppel, L., 2005. Large-scale observational studies <strong>of</strong><br />
<strong>Hypericum</strong> <strong>extract</strong>s in patients with depressive disorders—a<br />
systematic review. Phytomedicine 12, 148–157.<br />
Linde, K., Mulrow, C.D., 2000. St John's wort <strong>for</strong> depression.<br />
Cochrane Database Syst. Rev. CD000448.<br />
Linde, K., Ramirez, G., Mulrow, C.D., Pauls, A., Weidenhammer, W.,<br />
Melchart, D., 1996. St John's wort <strong>for</strong> depression—an overview<br />
<strong>and</strong> meta-analysis <strong>of</strong> r<strong>and</strong>omised clinical trials. BMJ 313,<br />
253–258.<br />
Linde, K., Melchart, D., Mulrow, C.D., Berner, M., 2002. St John's<br />
wort <strong>and</strong> depression. JAMA 288, 447–448 author reply 448–449.<br />
Linde, K., Berner, M., Egger, M., Mulrow, C., 2005a. St John's wort<br />
<strong>for</strong> depression: meta-analysis <strong>of</strong> r<strong>and</strong>omised controlled trials. Br.<br />
J. Psychiatry 186, 99–107.<br />
Linde, K., Mulrow, C.D., Berner, M., Egger, M., 2005b. St John's wort<br />
<strong>for</strong> depression (review). Cochrane Database Syst. Rev. 3,<br />
CD000448.<br />
Linde, K., Berner, M.M., Kriston, L., 2008. St John's wort <strong>for</strong> major<br />
depression. Cochrane Database Syst. Rev. CD000448.<br />
Marsh, W.L., Davies, J.A., 2002. The involvement <strong>of</strong> sodium <strong>and</strong><br />
calcium ions in <strong>the</strong> release <strong>of</strong> amino acid neurotransmitters from<br />
mouse cortical slices elicited by hyper<strong>for</strong>in. Life Sci. 71,<br />
2645–2655.<br />
Mennini, T., Gobbi, M., 2004. The antidepressant mechanism <strong>of</strong><br />
<strong>Hypericum</strong> per<strong>for</strong>atum. Life Sci. 75, 1021–1027.<br />
Mills, E., Montori, V., Wu, P., Gallicano, K., Clarke, M., Guyatt, G.,<br />
2004. Interaction <strong>of</strong> St John's wort with conventional drugs:<br />
systematic review <strong>of</strong> clinical trials. BMJ 329, 27–30.<br />
Moore, L.B., Goodwin, B., Jones, S.A., Wisely, G.B., Serabjit-Singh,<br />
C.J., Willson, T.M., Collins, J.L., Kliewer, S.A., 2000. John's wort<br />
induces hepatic drug metabolism through activation <strong>of</strong> <strong>the</strong><br />
pregnane X receptor. Proc. Natl Acad. Sci. U. S. A. 97, 7500–7502.<br />
Moreno, R.A., Teng, C.T., Almeida, K.M., Tavares Junior, H., 2006.<br />
<strong>Hypericum</strong> per<strong>for</strong>atum versus fluoxetine in <strong>the</strong> treatment <strong>of</strong> mild<br />
to moderate depression: a r<strong>and</strong>omized double-blind trial in a<br />
Brazilian sample. Rev. Bras. Psiquiatr. 28, 29–32.<br />
Moses, E.L., Mallinger, A.G., 2000. St. John's wort: three cases <strong>of</strong><br />
possible mania induction. J. Clin. Psychopharmacol. 20,<br />
115–117.<br />
Muller, D., Pfeil, T., von den Driesch, V., 2003. Treating depression<br />
comorbid with anxiety—results <strong>of</strong> an open, practice-oriented<br />
study with St John's wort WS 5572 <strong>and</strong> valerian <strong>extract</strong> in high<br />
doses. Phytomedicine 10 (Suppl 4), 25–30.<br />
Muller, T., Mannel, M., Murck, H., Rahlfs, V.W., 2004. Treatment <strong>of</strong><br />
somat<strong>of</strong>orm disorders with St. John's wort: a r<strong>and</strong>omized,<br />
double-blind <strong>and</strong> placebo-controlled trial. Psychosom. Med. 66,<br />
538–547.<br />
Mulsant, B.H., Ganguli, M., 1999. Epidemiology <strong>and</strong> diagnosis <strong>of</strong><br />
depression in late life. J. Clin. Psychiatry 60 (Suppl 20), 9–15.<br />
Nierenberg, A.A., Burt, T., Mat<strong>the</strong>ws, J., Weiss, A.P., 1999. Mania<br />
associated with St. John's wort. Biol. Psychiatry 46, 1707–1708.<br />
Nierenberg, A.A., Quitkin, F.M., Kremer, C., Keller, M.B., Thase, M.E.,<br />
2004. Placebo-controlled continuation treatment with mirtazapine:<br />
acute pattern <strong>of</strong> response predicts relapse. Neuropsychopharmacology<br />
229, 1012–1018.<br />
Olivier-Martin, R., 1986. Psychological factors, compliance <strong>and</strong><br />
resistance to antidepressant treatment. Encephale 12, 197–203.<br />
Olsen, L.R., Mortensen, E.L., Bech, P., 2004. Prevalence <strong>of</strong> major<br />
depression <strong>and</strong> stress indicators in <strong>the</strong> Danish general population.<br />
Acta Psychiatr. Sc<strong>and</strong>. 109, 96–103.<br />
Papakostas, G.I., Craw<strong>for</strong>d, C.M., Scalia, M.J., Fava, M., 2007.<br />
Timing <strong>of</strong> clinical improvement <strong>and</strong> symptom resolution in <strong>the</strong><br />
treatment <strong>of</strong> major depressive disorder. A replication <strong>of</strong> findings<br />
with <strong>the</strong> use <strong>of</strong> a double-blind, placebo-controlled trial <strong>of</strong><br />
<strong>Hypericum</strong> per<strong>for</strong>atum versus fluoxetine. Neuropsychobiology<br />
56, 132–137.<br />
Pfrunder, A., Schiesser, M., Gerber, S., Haschke, M., Bitzer, J.,<br />
Drewe, J., 2003. Interaction <strong>of</strong> St John's wort with low-dose oral<br />
contraceptive <strong>the</strong>rapy: a r<strong>and</strong>omized controlled trial. Br. J. Clin.<br />
Pharmacol. 56, 683–690.<br />
Philipp, M., Kohnen, R., Hiller, K.O., 1999. <strong>Hypericum</strong> <strong>extract</strong> versus<br />
imipramine or placebo in patients with moderate depression:<br />
r<strong>and</strong>omised multicenter study <strong>of</strong> treatment <strong>for</strong> eight weeks. BMJ<br />
319, 1534–1538.<br />
Rahimi, R., Nikfar, S., Abdollahi, M., 2009. <strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong><br />
<strong>of</strong> <strong>Hypericum</strong> per<strong>for</strong>atum in major depressive disorder in<br />
comparison with selective serotonin reuptake inhibitors: a<br />
meta-analysis. Prog. Neuropsychopharmacol. Biol. Psychiatry<br />
33, 118–127.<br />
Reichling, J., Hostanska, K., Saller, R., 2003. St. John' wort<br />
(<strong>Hypericum</strong> per<strong>for</strong>atum L.)—multicompound preparations versus<br />
single substances. Forsch. Komplementärmed. Klass. Naturheilkd<br />
10 (Suppl 1), 28–32.<br />
Röder, C., Schaefer, M., Leucht, S., 2004. Meta-analysis <strong>of</strong> effectiveness<br />
<strong>and</strong> <strong>tolerability</strong> <strong>of</strong> treatment <strong>of</strong> mild to moderate depression<br />
with St. John's wort. Fortschr. Neurol. Psychiatr. 72, 330–343.<br />
Rychlik,R.,Siedentop,H.,vondenDriesch,V.,Kasper,S.,2001.<br />
St. John's wort <strong>extract</strong> WS 5572 in minor to moderately<br />
severe depression. Effectiveness <strong>and</strong> tolerance <strong>of</strong> 600 <strong>and</strong><br />
1200 mg active ingredient daily. Fortschr. Med. Orig. 119,<br />
119–128.<br />
Schakau, D., Hiller, K.O., Schultz-Zehden, W., 1996. Nutzen/risiko-pr<strong>of</strong>il<br />
von johanniskrautextrakt. Psychopharmako<strong>the</strong>rapie 3, 116–122.
<strong>Efficacy</strong> <strong>and</strong> <strong>tolerability</strong> <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> <strong>for</strong> <strong>the</strong> treatment <strong>of</strong> mild to moderate depression<br />
Schempp, C.M., Ludtke, R., Wingh<strong>of</strong>er, B., Simon, J., 2000. Effect <strong>of</strong><br />
topical application <strong>of</strong> <strong>Hypericum</strong> per<strong>for</strong>atum <strong>extract</strong> (St. John's<br />
wort) on skin sensitivity to solar simulated radiation. Photodermatol.<br />
Photoimmunol. Photomed. 16, 125–128.<br />
Schmidt, U., Sommer, H., 1993. St. John's wort <strong>extract</strong> in <strong>the</strong><br />
ambulatory <strong>the</strong>rapy <strong>of</strong> depression. Attention <strong>and</strong> reaction ability<br />
are preserved. Fortschr. Med. 111, 339–342.<br />
Schrader, E., 2000. Equivalence <strong>of</strong> St John's wort <strong>extract</strong> (Ze 117)<br />
<strong>and</strong> fluoxetine: a r<strong>and</strong>omized, controlled study in mild<br />
−moderate depression. Int. Clin. Psychopharmacol. 15, 61–68.<br />
Schrader, E., Meier, B., Brattstrom, A., 1998. <strong>Hypericum</strong> treatment<br />
<strong>of</strong> mild−moderate depression in a placebo-controlled study. A<br />
prospective, double-blind, r<strong>and</strong>omized, placebo-controlled,<br />
multicenter study. Hum. Psychopharmacol. 13, 163–169.<br />
Schule, C., Baghai, T., Ferrera, A., Laakmann, G., 2001. Neuroendocrine<br />
effects <strong>of</strong> <strong>Hypericum</strong> <strong>extract</strong> WS 5570 in 12 healthy male<br />
volunteers. Pharmacopsychiatry 34, S127–S133.<br />
Schulz, V., 2001. Incidence <strong>and</strong> clinical relevance <strong>of</strong> <strong>the</strong> interactions<br />
<strong>and</strong> side effects <strong>of</strong> <strong>Hypericum</strong> preparations. Phytomedicine 8,<br />
152–160.<br />
Schulz, V., 2006. Safety <strong>of</strong> St. John's wort <strong>extract</strong> compared to<br />
syn<strong>the</strong>tic antidepressants. Phytomedicine 13, 199–204.<br />
Shaper & Brummer. Esbericum capsules. Release date: 2005.<br />
Accessed on: 28 February, 2005. Available from: http://www.<br />
schaper-bruemmer.de/SUB_ICM_EN.NSF/(html)/Frameset_<br />
load?OpenDocument&003Products_004Esbericum_002EsbericumKapseln_000EsbericumCapsules?OpenDocument&Allg-<br />
Nav_dsp?OpenForm&Kat0_root.<br />
Shelton, R.C., Keller, M.B., Gelenberg, A., Dunner, D.L., Hirschfeld,<br />
R., Thase, M.E., Russell, J., Lydiard, R.B., Crits-Cristoph, P.,<br />
Gallop, R., Todd, L., Hellerstein, D., Goodnick, P., Keitner, G.,<br />
Stahl, S.M., Halbreich, U., 2001. Effectiveness <strong>of</strong> St John's wort in<br />
major depression: a r<strong>and</strong>omized controlled trial. JAMA 285,<br />
1978–1986.<br />
Simbrey, K., Winterh<strong>of</strong>f, H., Butterweck, V., 2004. Extracts <strong>of</strong> St.<br />
John's wort <strong>and</strong> various constituents affect beta-adrenergic<br />
binding in rat frontal cortex. Life Sci. 74, 1027–1038.<br />
Singer, A., Wonnemann, M., Muller, W.E., 1999. Hyper<strong>for</strong>in, a major<br />
antidepressant constituent <strong>of</strong> St. John's wort, inhibits serotonin<br />
uptake by elevating free intracellular Na + . J. Pharmacol. Exp.<br />
Ther. 290, 1363–1368.<br />
Sommer, H., Harrer, G., 1994. Placebo-controlled double-blind<br />
study examining <strong>the</strong> effectiveness <strong>of</strong> an <strong>Hypericum</strong> preparation<br />
in 105 mildly depressed patients. J. Geriatr. Psychiatry Neurol. 7<br />
(Suppl 1), S9–S11.<br />
Spielmans, G.I., 2002. St John's wort <strong>and</strong> depression. JAMA 288,<br />
446–447 author reply 448−449.<br />
Stassen, H.H., Kuny, S., Hell, D., 1998. The speech analysis approach<br />
to determining onset <strong>of</strong> improvement under antidepressants.<br />
Eur. Neuropsychopharmacol. 8, 303–310.<br />
Stewart, J.W., Quitkin, F.M., McGrath, P.J., Amsterdam, J., Fava,<br />
M., Fawcett, J., Reimherr, F., Rosenbaum, J., Beasley, C.,<br />
Roback, P., 1998. Use <strong>of</strong> pattern analysis to predict differential<br />
relapse <strong>of</strong> remitted patients with major depression during 1 year<br />
<strong>of</strong> treatment with fluoxetine or placebo. Arch. Gen. Psychiatry<br />
55, 334–343.<br />
Szegedi, A., Muller, M.J., Anghelescu, I., Klawe, C., Kohnen, R.,<br />
Benkert, O., 2003. Early improvement under mirtazapine <strong>and</strong><br />
paroxetine predicts later stable response <strong>and</strong> remission with high<br />
sensitivity in patients with major depression. J. Clin. Psychiatry<br />
64, 413–420.<br />
Szegedi, A., Kohnen, R., Dienel, A., Kieser, M., 2005. Acute<br />
treatment <strong>of</strong> moderate to severe depression with <strong>Hypericum</strong><br />
<strong>extract</strong> WS 5570 (St John's wort): r<strong>and</strong>omised controlled double<br />
blind non-inferiority trial versus paroxetine. BMJ 330, 503.<br />
765<br />
Tai-Seale, M., Croghan, T.W., Obenchain, R., 2000. Determinants <strong>of</strong><br />
antidepressant treatment compliance: implications <strong>for</strong> policy.<br />
Med. Care Res. Rev. 57, 491–512.<br />
Taylor, L.H., Kobak, K.A., 2000. An open-label trial <strong>of</strong> St. John's<br />
wort (<strong>Hypericum</strong> per<strong>for</strong>atum) in obsessive−compulsive disorder.<br />
J. Clin. Psychiatry 61, 575–578.<br />
van Gurp, G., Meterissian, G.B., Haiek, L.N., McCusker, J.,<br />
Bellavance, F., 2002. St John's wort or sertraline? R<strong>and</strong>omized<br />
controlled trial in primary care. Can. Fam. Physician 48,<br />
905–912.<br />
Volp, A., 2002. St John's wort <strong>and</strong> depression. JAMA 288, 447 author<br />
reply 448−449.<br />
Volz, H.P., Laux, P., 2000. Potential treatment <strong>for</strong> subthreshold <strong>and</strong><br />
mild depression: a comparison <strong>of</strong> St. John's wort <strong>extract</strong>s <strong>and</strong><br />
fluoxetine. Compr. Psychiatry 41, 133–137.<br />
Volz, H.P., Murck, H., Kasper, S., Moller, H.J., 2002. St John's wort<br />
<strong>extract</strong> (LI 160) in somat<strong>of</strong>orm disorders: results <strong>of</strong> a placebocontrolled<br />
trial. Psychopharmacology (Berl) 164, 294–300.<br />
Vorbach, E.U., Hubner, W.D., Arnoldt, K.H., 1994. Effectiveness <strong>and</strong><br />
tolerance <strong>of</strong> <strong>the</strong> <strong>Hypericum</strong> <strong>extract</strong> LI 160 in comparison with<br />
imipramine: r<strong>and</strong>omized double-blind study with 135 outpatients.<br />
J. Geriatr. Psychiatry Neurol. 7 (Suppl 1), S19–S23.<br />
Wagner, H., Bladt, S., 1994. Pharmaceutical quality <strong>of</strong> <strong>Hypericum</strong><br />
<strong>extract</strong>s. J. Geriatr. Psychiatry Neurol. 7 (Suppl 1), S65–S68.<br />
Westerh<strong>of</strong>f, K., Kaunzinger, A., Wurglics, M., Dressman, J.,<br />
Schubert-Zsilavecz, M., 2002. Biorelevant dissolution testing <strong>of</strong><br />
St John's wort products. J. Pharm. Pharmacol. 54, 1615–1621.<br />
Wheatley, D., 1997. LI 160, an <strong>extract</strong> <strong>of</strong> St. John's wort, versus<br />
amitriptyline in mildly to moderately depressed outpatients—a<br />
controlled 6-week clinical trial. Pharmacopsychiatry 30 (Suppl<br />
2), 77–80.<br />
Wheatley, D., 2002. St John's wort <strong>and</strong> depression. JAMA 288, 446<br />
author reply 448–449.<br />
Whiskey, E., Werneke, U., Taylor, D., 2001. A systematic review <strong>and</strong><br />
meta-analysis <strong>of</strong> <strong>Hypericum</strong> per<strong>for</strong>atum in depression: a comprehensive<br />
clinical review. Int. Clin. Psychopharmacol. 16,<br />
239–252.<br />
Whitten, D.L., Myers, S.P., Hawrelak, J.A., Wohlmuth, H., 2006. The<br />
effect <strong>of</strong> St John's wort <strong>extract</strong>s on CYP3A: a systematic review <strong>of</strong><br />
prospective clinical trials. Br. J. Clin. Pharmacol. 62, 512–526.<br />
Wilhelm, K., Mitchell, P., Slade, T., Brownhill, S., Andrews, G.,<br />
2003. Prevalence <strong>and</strong> correlates <strong>of</strong> DSM-IV major depression in an<br />
Australian national survey. J. Affect. Disord. 75, 155–162.<br />
Witte, B., Harrer, G., Kaptan, T., Podzuweit, H., Schmidt, U., 1995.<br />
Treatment <strong>of</strong> depressive symptoms with a high concentration<br />
<strong>Hypericum</strong> preparation. A multicenter placebo-controlled double-blind<br />
study. Fortschr. Med. 113, 404–408.<br />
Woelk, H., 2000. Comparison <strong>of</strong> St John's wort <strong>and</strong> imipramine <strong>for</strong><br />
treating depression: r<strong>and</strong>omised controlled trial. BMJ 321,<br />
536–539.<br />
Woelk, H., Burkard, G., Grunwald, J., 1994. Benefits <strong>and</strong> risks <strong>of</strong> <strong>the</strong><br />
<strong>Hypericum</strong> <strong>extract</strong> LI 160: drug monitoring study with 3250<br />
patients. J. Geriatr. Psychiatry Neurol. 7 (Suppl 1), S34–S38.<br />
Yoshitake, T., Iizuka, R., Yoshitake, S., Weikop, P., Müller, W.E.,<br />
Ogren, S.O., Kehr, J., 2004. <strong>Hypericum</strong> per<strong>for</strong>atum L (St John's<br />
wort) preferentially increases extracellular dopamine levels in<br />
<strong>the</strong> rat prefrontal cortex. Br. J. Pharmacol. 142, 414–418.<br />
Zeller, K., 2000. A convincing safety pr<strong>of</strong>ile <strong>of</strong> St. John's wort <strong>extract</strong><br />
(Laif 600) shown in a large post marketing surveillance.<br />
Phytomedicine 7, 107.<br />
Zhou, S.F., Lai, X., 2008. An update on clinical drug interactions with<br />
<strong>the</strong> herbal antidepressant St. John's wort. Curr. Drug Metab. 5,<br />
394–409.