Version classiqueVersion mobile
OpenEdition Books

Des sources du savoir aux médicaments du futur

Jacques Fleurentin
Jean-Marie Pelt
Guy Mazars

3. Médicaments du XXIe siècle

Efficacy and safety of Harpagophytum and Salix extract preparations

Sigrun Chrubasik


In Europe, extracts from Harpagophytum procumbens and Salix species are widely used for the treatment of rheumatic pain. Declaration of the quantity of active principles helps to optimize treatment success. For both herbal medicinal products, pharmacological and clinical studies have proven antirheumatic effectiveness which is superior to placebo and equally effective compared to synthetic medication. Treatment of osteoarthritic and low back pain with Harpagophytum procumbens and Salix extracts is associated with a low incidence of adverse events lower than that known from NSAIDs. In terms of quality, safety and efficacy, Harpagophytum procumbens and Salix extract preparations may fulfill the requirements that characterize synthetic medication.

Texte intégral

Biopharmaceutical quality of Harpagophytum preparations

1The monographs of the European Scientific Cooperative on Phytotherapy (ESCOP, 1996) refer to pharmacopeia definitions of herbal material. Requirements for drug preparations include the cut, dried secondary root tubers that contain at least 2.2 % of iridoid glycosides or 1.2 % harpagoside for Harpagophytum procumbens and the dried bark of young branches which should contain not less than 1 % of total salicin for Salix species.

Harpagophytum preparations

2For the treatment of painful arthrosis or tendinitis, the Harpagophytum ESCOP Monograph (1996) recommends 1.5 - 3.0 g drug in decoction three times per day or equivalent aqueous or hydroalcoholic extracts. With 9 g crude 1 % drug patients consume 100 mg harpagoside (Chrubasik et al., 1996a). An analysis of the Harpagophytum preparations available in Germany 1996 revealed that the mean harpagoside content of the tablets varied between 1.5 and 9.3 mg with a daily consumption of harpagoside between 0.2 and 50 mg, respectively (Chrubasik et al., 1996b). Tea (aqueous extract) of 9 g crude drug contains a mean of 1 80 mg harpagoside, depending on the harpagoside content of the drug, which varied in the tea drugs 1996 between 1.1 and 3.6 % (Chrubasik et al., 1996a). Producer extracts with up to 2 % harpagoside (Chrubasik et al., 1996c) provide tablets with 50 mg harpagoside or less in the daily dosage. A special Harpagophytum extract, WS 1531* (drug extract ratio 6-9:1, harpagoside enriched, minimum 5 %), has recently been used for a dose-finding study (daily dosage of harpagoside 50 and 100 mg) (Chrubasik et al., 1 999). A high drug-extract ratio or a high quantity of extract are not necessarily associated with a high content of active principles (Sporer and Chrubasik, 1999).

Salix preparations

3For the treatment of feverish conditions, symptomatic treatment of mild rheumatic complaints, reliefs of pain, including mild headache, the Salix ESCOP Monograph (1997) recommends 3 - 6 g of powdered drug or extracts equivalent to up to 240 mg salicin, a prodrug of salicylate metabolites. This particular salicin dose is bioequivalent to 50 mg acetylsalicylate (Schmid, 1998), a cardioprotective rather than a pain-relieving dose.

Effect mechanisms

4Herbal medicinal products for the treatment of pain have a broader effect mechanism than synthetic antirheumatic drugs that are more or less selective cyclooxygenase inhibitors. Harpagophytum extract inhibits not only both arachidonic acid metabolism pathways, the cyclo oxygenase and the lipoxygenase (Tippler et al., 1997) but also cytokine release (TNFa) (Kammerer and Fiebig, 2000) and may exert still unknown effects. Likewise, Salix extract interacts with both both arachidonic acid metabolism pathways, cytokine release and provides antioxidative activity (Chrubasik et al., 2000).

Clinical studies in patients suffering from various osteoarthritic pain

Harpagophytum procumbens

5Nine exploratory (ESCOP, 1996; Chrubasik, 2000) and 3 confirmatory studies give evidence for the antirheumatic efficacy of Harpagophytum preparations. A double-blind study (Chrubasik et al., 1996d) included 118 patients suffering for longer than 6 months from chronic low back pain not attributable to identifiable causes. Upon their written informed consent, they received either 2400 mg Harpagophytum extract (drug-extract ratio 2.5:1) with 50 mg harpagoside per day or a placebo. Prior to treatment, the groups were comparable with regard to their history and biochemical data, circulatory and laboratory parameters, as well as the Arhus low back pain index (Manniche et al., 1994) and its scores for pain, invalidity and physical impairment. After 4 weeks, the median Arhus low back pain index had improved by 20 % in the Harpagophytum group and by 8 % in the placebo group (p < 0, 059). This significant trend of effectiveness was based on a significant decrease in the pain index (p = 0,016). In patients with pseudoradiating pain into the leg(s), the median effect of Harpagophytum extract was like that of placebo. However, in patients suffering from pain not pseudoradiating into the leg(s), the pain index decrease was 25 % in the Harpagophytum group and zero in the placebo group (Fig. 4). Nine of 54 patients receiving Harpagophytum extract (a total of 20 %) were completely pain-free in the fourth week of treatment, as compared one patient (2 %) in the placebo group (p = 0,008). Only minor and nonspecific adverse effects occurred during the Harpagophytum treatment.

6Another double-blind study (Chrubasik et al., 1999) included a similar patient population: 197 patients suffering from chronic low back pain (local as well as pseudoradiating) that was not attributable to identifiable causes. After written informed consent, they received tablets with either 200 or 400 mg Harpagophytum special extract (batch 9601) or placebo three times daily. The extract tablets contained wither 50 or 100 mg harpagoside, respectively. Based on the information of the above study, the main criterium of this 4 week randomised, double-blind, placebo-controlled study was chosen as being the number of painfree patients in the 4th week of treatment. The treatment and placebo groups were well matched in physical characteristics; in the severity of pain, duration, nature and accompaniments of their pain; the Arhus low back pain index and in laboratory indices of organ system function. 182 patients completed the study. The number of painfree patients was 3, 6 and 10 in groups P, H600 and H1200, respectively (p=0.040) and higher in patients suffering less than 42 days of pain. Sensitivity analysis with alternative principle outcome variables tended to confirm this result as well as subgroup sensitivity analyses, however only for severe and radiating pain. The Arhuse low back pain index and its components improved in all groups by about 20 %. Only the component current pain differed in the treatment groups from that of placebo, however, showing a larger relative change in the low dose Harpagophytum group. Subgroup analysis confirmed this trend for patients with shorter exacerbations, less pain and no radiation. However, the changes in the pain components of the Arhus low back pain index do not support the effect on the primary outcome variable. Multiple regression analysis revealed that the pain components of the Arhus index were significantly affected by the tramadol consumption whereas prior duration of exacerbation, initial intensity of pain or radiation of pain had no effect. As previously shown, the consumption of the analgesic rescue medication tramadol did not correlate well with the amount of pain. The number of adverse effects did not differ between groups, however, gastrointestinal adverse effects seem to be related to Harpagophytum extract.

7A recent randomized, double-blind, multicenter study (Chantre et al., 2000) included 122 patients suffering from osteoarthritis of the knee and/or the hip (grade 1, 2 or 3 in Kellgren's scale), with subjective pain of at least 50 mm on a 100 mm visual analog scale (VAS) and a score of at least 4 in the Lequesne algofunctional index. Patients received either freeze-ground secondary Harpagophytum procumbens roots containing not less than 3 % iri- doid glucosides (6 capsules a 435 mg with 60 mg harpagoside per day) or diacerhein (2 capsules a 50 mg per day). The hypothesis on which the study was based was to demonstrate a difference of around 10 mm on a 100 mm VAS between diacerhein and Harpagophytum procumbens.

8There was no significant difference at baseline in the efficacy assessment between the two treatment groups. The primary efficacy parameter "spontaneous pain" showed a significant improvement in both treatment groups but there was no statistical difference between the two groups. Likewise, both treatments resulted in improvement of functional parameters of the knee or of the hip. At completion of the four-month treatment phase, patients taking Harpagophytum procumbens were using significantly less NSAIDs and antalgic drugs. Significant less patients in the Harpagophytum procumbens group (n= 10) reported at least one adverse reaction imputable to the treatment compared to the patients in the diacerhein group (n=21). The most common adverse events reported were gastro-intestinal disturbances.

Salix species

9Five exploratory (ESCOP, 1997; Chrubasik, 2000) and 2 confirmatory studies give evidence for the antirheumatic efficacy of Salix preparations. A double-blind study investigated 78 inpatients of a clinic for rheumatic diseases (Schmid, 1998). They suffered from osteoarthritis of the knee and/or hip joint, as verified by the ARA criteria. After a wash-out period of four days, patients received daily 1360 mg Salix bark extract (in form of coated tablets) or placebo for two weeks. Verum medication corresponded to 240 mg salicin/day. An analgesic effect of the verum medication was observed by monitoring the change of the WOMAC pain index from day 0 to day 14 as the main study criterion. Using 10 cm visual analogue scales, a statistically significant difference (p<0.05) between verum and placebo group was observed both in the intention- to-treat population (n = 78 patients) and in the per-protocol population (n = 68 patients). Daily 10 cm VAS pain scales, and final overall assessments by patient and doctor, were used as secondary criteria and confirmed the positIve effect of the verum medication. Only minor adverse events occurred.

10In another placebo-controlled, randomised, double-blind study (Chrubasik et al., 2000) oral Salix bark extract with 120 and 240 mg salicin per day was compared with placebo over 4 weeks. 210 patients suffering from exacerbations of chronic low back pain and reporting current pain of more than 5 out of 1 0 on a visual analogue scale were included. Patients were allowed tramadol as the sole rescue medication and the principle outcome measure was the number of patients who were pain-free without tramadol for 5 days out of the last week. The treatment and placebo groups were well matched at baseline in 114 of 120 clinical features. Patients in both Salix groups tended to have higher Arhuse scores than those receiving placebo. Those receiving the larger dose of extract tended to have suffered pain for longer than those in the other groups, with osteoporosis contributing more to the overall pain; more of them suffered from depression and their Beck depression score tended to be higher and more had previously tried opioids and a third of them had changed their occupation because of unbearable pain. A total of 191 patients completed the study. The respective numbers of pain-free patients were 4, 15 and 27 in the placebo group (P), the Salix bark extract groups with 1 20 mg salicin and 240 mg salicin, respectively (p<0.001, one-tailed Cochrane-Armitage test, intention-to-treat analysis on 210 patients with "drop-outs" considered as "non-responders" to treatment). Neither radiation into the leg(s) nor the duration of the propensity to low back pain nor the duration of the exacerbation affected the response. A significant response in the Salix group with 240 mg salicin was evident after only one week of treatment. For each week of the study, significantly more patients in the placebo group required tramadol. There was no evidence for Salix bark-related side-effects, though one patient suffered a severe allergic reaction, possibly to the extract.


11The results of the clinical studies indicate that preparations from Harpagophytum procumbens or Salix species are useful for the treatment of osteoarthritis not only in order to improve clinical symptoms but also in order to save on the consumption of antiinflammatory and analgesic drugs. The good tolerance of the herbal medicinal products is of advantage in the treatment of a degenerative chronic pathology that requires prolonged periods of treatment. Evidence-based data (ESCOP monographs) and the recent clinical studies have not revealed any drug-related specific adverse events. Thus, in terms of efficacy and safety, both herbal medicinal products are recommendable for the treatment of pain and should be administered as first-line drugs.



CHANTRE P., CAPPELAERE A., LEBLAN D., GUEDON D., VANDERMANDER J., FOURNIE B. (2000) Efficacy and tolerability of Harpagophytum procumbens versus diacerhein in treatment of osteoarthritis, Phytomedicine, 7, 177-183.

CHRUBASIK S., SPORER F., WINK M. (1996a) Zum Wirkstoffgehalt in Teezubereitungen aus Harpagophytum procumbens, Forsch Komplementärmed, 3, 116-119.

CHRUBASIK S., SPORER F., WINK M. (1996b) Zum Harpagosidgehalt in Arzneimitteln aus Harpagophytum procumbens, Forsch Komplementärmed, 3, 57-63.

CHRUBASIK S., SPORER F., WINK M. (1996c) Zum Harpagosidgehalt verschiedener Trockenextraktpulver aus Harpagophytum procumbens, Forsch Komplementärmed, 3, 6-11.

CHRUBASIK S., ZIMPFER CH., SCHÜTT U., ZIEGLER R. (1996d) Effectiveness of Harpagophytum procumbens in treatement of acute low back pain, Phytomedicine, 3, 1-10.

CHRUBASIK S., JUNCK H., BREITSCHWERDT H., CONRADT C., ZAPPE H. (1999) Effectiveness of Harpagophytum extract WS 1531 in the treatment of exacerbation of low back pain: a randomized placebo-controlled doubleblind study, Eur J Anaesthesiol, 16, 118-129.

CHRUBASIK S., SPORER F., DILLMANN-MARSCHNER R., FRIEDMANN A., WINK M. (2000) Physikalisch-chemische Eigenschaften von Harpagosid und in vitro, Freisetzung von Harpagosid aus Harpagophytum-Tabletten, Phytomedicine, 6, 469-473.

CHRUBASIK S., EISENBERG E. (2000) Efficacy and Safety of Salix Extract Preparations, in Chrubasik S., Roufogalis B. (eds.) Herbal medicinal products for the treatment of pain, Lismore University Press, Lismore, Australia.

CHRUBASIK S. (2000) Efficacy and Safety of Harpagophytum Extract Preparations, in Herbal medicinal products for the treatment of pain, Eds. Chrubasik S., Roufogalis B., Lismore University Press, Lismore, Australia.

CHRUBASIK S., EISENBERG E., BAIAN E., WEINBERGER T., LUZZATI R., CONRADT C. (2000) Treatment of low back pain exacerbations with willow bark extract: a randomized double-blind study, Am J Med, in press.

(1996) ESCOP Monographs, Harpagophyti Radix, Fascicule 2, ISBN 1 - 901964-03-5.

(1997) ESCOP Monograph. Salicis Cortex (Willow Bark), Fascicule 4, ISBN 1-901964-03-5.

KÄMMERER N., Fiebig B. (2000) Antiinflammatory effects of Harpagophytum lysates in human monocytes, in Chrubasik S., Roufogalis B. (eds) Herbal medicinal products for the treatment of pain, Lismore University Press, Lismore, Australia.

SCHMID B.M. (1998) Behandlung von Cox- und Gonarthrosen mit einem Trockenextrakt aus Salix purpurea and daphnoides, PhD-thesis, University of Tübingen.

SPORER F., CHRUBASIK S. (1999) Präparate aus der Teufelskralle (Harpagophytum procumbens), Zschr. Phytotherapie, 20, 235-6.

TIPPLER B., SYROVETS T., LOEW D., SIMMET Th. (1996) Harpagophytum procumbens: Wirkung von Extrakten auf die Eicosanoidbiosynthese in Ionophor A23187-stimuliertem menschlichem Vollblut, in: Phytopharmaka II. Forschung und klinische Anwendung, Hrsg. D. Loew, N. Rietbrock, Steinkopf-Verlag, Darmstadt, 95-100.

© IRD Éditions, 2002

Conditions d’utilisation :

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.