The herbal medicine Echinacea purpurea (E. purpurea) has been shown to induce cytochrome P450 3A4 (CYP3A4) both in vitro and in humans. This study explored whether E. purpurea affects the pharmacokinetics of the CYP3A4 substrate docetaxel in cancer patients.
Ten evaluable cancer patients received docetaxel (135 mg, 60 min IV infusion) before intake of a commercially available E. purpurea extract (20 oral drops three times daily) and 3 weeks later after a 14 day supplementation period with E. purpurea. In both cycles, pharmacokinetic parameters of docetaxel were determined.
Before and after supplementation with E. purpurea, the mean area under the plasma concentration–time curve of docetaxel was 3278 ± 1086 and 3480 ± 1285 ng ml−1 h, respectively. This result was statistically not significant. Nonsignificant alterations were also observed for the elimination half-life (from 30.8 ± 19.7 to 25.6 ± 5.9 h, P = 0.56) and maximum plasma concentration of docetaxel (from 2224 ± 609 to 2097 ± 925 ng ml−1, P = 0.30).
The multiple treatment of E. purpurea did not significantly alter the pharmacokinetics of docetaxel in this study. The applied E. purpurea product at the recommended dose may be combined safely with docetaxel in cancer patients.
The herbal immunostimulant Echinacea purpurea (E. purpurea) is widely used among cancer patients.
E. purpurea has been shown to induce cytochrome P450 3A4 (CYP3A4) in vitro and in clinical studies.
CYP3A4 is extensively involved in the metabolism of many anticancer drugs, such as docetaxel.
What this Study Adds
This is the first clinical study to investigate the pharmacokinetic interaction between E. purpurea and an anticancer drug metabolized by CYP3A4.
The commercially available E. purpurea extract did not significantly alter the pharmacokinetics of docetaxel.
The applied E. purpurea formulation at the recommended dose may be combined safely with docetaxel and presumably also with other anticancer drugs primarily metabolized by CYP3A4.
The use of complementary and alternative medicines among cancer patients and the associated risk of herb–drug interactions have increased over recent years [1, 2]. Especially for anticancer drugs, which usually have narrow therapeutic windows, these interactions could have serious consequences, such as an increased risk of toxicities or undertreatment.
Among cancer patients, Echinacea is a widely used herbal supplement. In a survey including 318 cancer patients, Echinacea was the most popular herbal medicine, used by 21% of all users of complementary and alternative medicines . Echinacea was also reported to be the second-most popular pharmacological complementary and alternative medicines agent among cancer patients enrolled into phase I clinical trials .
Echinacea is generally used to stimulate the immune system and to prevent the common cold and upper respiratory infections [5, 6]. The most common species of Echinacea are Echinacea angustifolia, Echinacea pallida and Echinacea purpurea (E. purpurea). The components of Echinacea responsible for the pharmacological effects are caffeic acid derivatives, alkylamides, polysaccharides and glycoproteins . Of these components, caffeic acid derivatives and the more bioavailable alkylamides are found in ethanolic liquid extracts for medicinal use .
The use of Echinacea by cancer patients may interfere with their conventional chemotherapy via interactions with the cytochrome P450 (CYP) 3A4 isoenzyme system. This enzyme system is involved in the metabolism of many anticancer drugs. Both in supersomes and in hepatocytes, it has been shown that Echinacea extracts have the potential to inhibit CYP3A4 in vitro [9, 10]. There are indications that Echinacea is also capable of inducing CYP3A4. Induction of CYP3A4 by E. purpurea has been shown in healthy volunteers, in whom the systemic exposure to the CYP3A4 probe midazolam was significantly decreased after supplementation with E. purpurea for 28 days . In another clinical study with midazolam in healthy volunteers, E. purpurea also affected CYP3A4 function . In this study on volunteers, the systemic clearance of intravenous (IV) midazolam was significantly increased, which reflects induction of hepatic CYP3A4 activity, while intestinal CYP3A4 was not significantly affected, as shown by the lack of significant alterations in oral clearance of orally administered midazolam . In a third clinical study, no significant effect of E. purpurea on midazolam pharmacokinetics was reported in healthy volunteers . Thus, both in vitro and clinical results showed that E. purpurea has the potential to affect CYP3A4, but results concerning inhibition and induction are inconsistent.
An anticancer drug for which systemic exposure may be affected via CYP3A4 modulation by E. purpurea is docetaxel. Docetaxel has, among other indications, been approved for the treatment of locally advanced or metastatic breast cancer, nonsmall cell lung cancer and hormone-refractory metastatic prostate cancer at doses ranging from 75 to 100 mg m−2, administrated as a 1 h IV infusion every 3 weeks. Pharmacokinetic interactions between docetaxel and E. purpurea could be expected, because docetaxel is extensively metabolized by CYP3A4. As docetaxel is administered intravenously, hepatic CYP3A4 is mainly involved in its metabolism. Expected induction of hepatic CYP3A4 by E. purpurea may lead to decreased plasma levels of docetaxel. For docetaxel, systemic exposure has shown to be a good predictor for its efficacy and toxicity . Thus, CYP3A4 induction by E. purpurea could lead to undertreatment in patients receiving docetaxel chemotherapy.
Currently, no clinical studies concerning pharmacokinetic interactions between E. purpurea and anticancer drugs have been reported. Results of the present study may provide valuable information about the safety of concomitant use of E. purpurea with other anticancer agents metabolized by CYP3A4. In the present study, the primary objective was to determine the effect of E. purpurea on the pharmacokinetics of docetaxel. The secondary objective was to assess the effect of E. purpurea supplementation on safety parameters, such as grade 3 and 4 toxicities induced by docetaxel [according to National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE, version 3.0)].
This clinical study was performed at the Netherlands Cancer Institute (NKI, Amsterdam, The Netherlands). Patients with histological or cytological proof of cancer for whom treatment with docetaxel was considered to be of therapeutic benefit (e.g. advanced breast, gastric, oesophagus, bladder, prostate, ovarian, nonsmall cell lung, head and neck cancer) were included. Other inclusion criteria were as follows: age ≥ 18 years, performance status ≤ 2 according to the World Health Organization scale, life expectancy > 3 months, absolute neutrophil count ≥ 1.5 × 109 l−1, platelet count ≥ 100 × 109 l−1, haemoglobin level ≥ 6.0 mmol l−1, hepatic function as defined by serum bilirubin ≤ 1.5 times the upper limit of normal and alanine aminotransferase and aspartate aminotransferase ≤ 2.5 times the upper limit of normal, renal function as defined by serum creatinine ≤ 1.5 times the upper limit of normal or creatinine clearance ≥ 50 ml min−1, able and willing to swallow and retain oral medication, to comply with the protocol procedures and to follow dietary restrictions.
Patients were excluded in the event of any treatment with investigational drugs within 30 days before the start of the study or the use of herbal supplements within 6 weeks prior to study treatment. Other exclusion criteria were as follows: alcoholism, drug addiction, psychotic disorders leading to non-adequate follow-up, concomitant use of multidrug resistance and CYP3A-modulating drugs, uncontrolled infectious disease, HIV-1 or HIV-2 type patients, unresolved (>grade 1) toxicities of previous chemotherapy, bowel obstruction or motility disorders that may influence the absorption of drugs, pregnancy, chronic use of H2-receptor antagonists or proton pump inhibitors, neurological disease that may render a patient at increased risk for peripheral or central neurotoxicity and presence of symptomatic cerebral or leptomeningeal metastases.
The study (EudraCT number: 2008-000886-41) has been approved by the Medical Ethical Committee of the NKI, and all patients provided written informed consent prior to study entry. All patients were treated between April 2009 and March 2010.
Docetaxel (Taxotere®; Aventis Pharma SA, Antony Cedex, France) was administered intravenously and was supplied in a 15 ml clear glass vial containing 2 ml of a 40 mg ml−1 docetaxel solution in polysorbate 80. Standard docetaxel pretreatment consisted of oral dexamethasone 8 mg two times daily for three consecutive days: 1 day before, on the day of docetaxel administration and 1 day after. Furthermore, commercially available E. purpurea drops were used (A. Vogel Echinaforce®, batch 08 K0302; Biohorma BV, Elburg, The Netherlands). These drops were labelled to contain 95% aerial parts and 5% roots of E. purpurea.
Study design and procedures
On day 1, all patients received docetaxel at an absolute dose of 135 mg given as a 60 min IV administration (cycle 1). The dose of 135 mg was based on a safe dose of 75 mg m−2 and a mean body surface area of 1.8 m2. From day 7 until the morning of day 22, the patients ingested 20 drops of the E. purpurea extract three times daily. On day 22, the second cycle of docetaxel was administered according to the same dosing schedule as on day 1 (Figure 1). The follow-up of each patient ended with an end-of-treatment visit 3 weeks after day 22.
Blood samples for assessment of docetaxel pharmacokinetics were drawn at predose, 0.25, 0.5, 0.75, 1, 1.5, 2, 4, 7, 10, 24 and 48 h after the start of the docetaxel infusion. Blood was collected in heparinized tubes and centrifuged at 1500g for 10 min at 4°C. Subsequently, plasma was separated and stored at −20°C until analysis.
Docetaxel plasma levels were quantified using a validated liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) assay with a lower limit of quantification of 0.25 ng ml−1 .
Analysis of E. purpurea constituents: alkylamides
In order to check the compliance to E. purpurea intake, patients had to keep diaries. In addition, they were called at regular intervals by the research team, and single blood samples were collected in heparinized tubes on days 7, 14 and 22. In a subset of four patients, the pharmacokinetics of alkylamides were studied by collection of blood samples at t = 0, 0.5, 1 and 2 h after administration of E. purpurea. Plasma was separated after centrifugation at 1500g for 10 min at 4°C and stored at −20°C until quantitative analysis of dodeca-2E,4E,8Z,10E/Z-tetraenoic acid isobutylamides (DTAI) using a validated LC-MS/MS assay .
Pharmacokinetic parameters were calculated using noncompartmental analysis with R software (version 2.10.1; R Development Core Team, Vienna, Austria) by employing validated scripts.
The following pharmacokinetic parameters of docetaxel were calculated: area under the plasma concentration–time curve from time zero to infinity (AUC0–∞), elimination half-life (t1/2) and maximum plasma concentration (Cmax).
For each patient, the values of AUC0–∞,t1/2 and Cmax of docetaxel in cycle 1 (before E. purpurea) were compared with values obtained in cycle 2 (after E. purpurea). After logarithmic transformation of these parameters, Student's paired t-test (α = 0.05) was performed by use of R.
Docetaxel-related adverse events
Docetaxel-related adverse events during cycles 1 and 2 were registered according to NCI CTCAE version 3.0. Adverse events were considered as docetaxel related when rated as ‘possibly’, ‘probably’ or ‘definitely’ related by the investigator.
Eleven patients were included (Table 1), of whom one patient needed to be replaced because the second docetaxel course was not administered due to her deteriorated physical condition. Hence, in total 10 patients were eligible for evaluation.
Table 1. Patient characteristics (n = 11)
World Health Organization performance status
Nonsmall cell lung carcinoma
Unknown primary tumour
Effect of E. purpurea on the pharmacokinetics of docetaxel
In Figure 2, the mean plasma concentration–time curves of docetaxel in the absence and presence of E. purpurea are presented. The individual differences in docetaxel AUC0–∞ are depicted in Figure 3. In five patients, an increase of the AUC0–∞ of docetaxel was observed after supplementation with E. purpurea, while in the other five patients the AUC0–∞ decreased. The mean values of pharmacokinetic parameters of docetaxel are shown in Table 2. Intake of E. purpurea did not result in statistically significant changes in AUC0–∞, Cmax and t1/2 of docetaxel.
Table 2. Summary of docetaxel pharmacokinetic parameters
Day 1: docetaxel alone
Day 22: with pretreatment of E. purpurea
Pharmacokinetic data are given as means ± SD and were obtained on day 1 (docetaxel alone) and day 22 (14 days after start of Echinacea purpurea). The P values were obtained from Student's paired t test. Abbreviations are as follows: AUC0–∞, area under the docetaxel plasma concentration–time curve extrapolated to infinity; Cmax, peak plasma concentration; and t1/2, half-life of the terminal disposition phase.
AUC0–∞ (ng ml−1 h)
3278 ± 1086
3480 ± 1285
Cmax (ng ml−1)
2224 ± 609
2097 ± 925
30.8 ± 19.7
25.6 ± 5.9
Docetaxel-related adverse events
The incidence of docetaxel-related adverse events differed between the two docetaxel courses. During the course without E. purpurea, 24 adverse events (grade 1–2, 22 events; and grade 3–4, two events) were reported in nine patients, while 16 adverse events (grade 1–2, 15 events; and grade 3–4, one event) occurred in six patients in the course after E. purpurea supplementation. However, in the majority of the patients the incidence of docetaxel-related adverse events did not correlate with changes in the AUC0–∞ of docetaxel.
The most common adverse events were fatigue (cycle 1, five events; and cycle 2, four events), alopecia (five events in both cycles), rash (cycle 1, three events; and cycle 2, one event) and allergic reaction (cycle 1, two events; and cycle 2, one event), and were mostly of grade 1–2.
Adherence to E. purpurea intake
Adherence was confirmed by inspection of patients' diaries, telephone calls and inspection of returned bottles. Bioanalysis of alkylamides in plasma samples was less applicable to demonstrate adherence. Several bioanalytical assays have been developed for quantification of alkylamides such as undeca-2-ene-8,10-diynoic acid isobutylamide  and DTAI . In the present study, DTAI, the most abundant alkylamides in E. purpurea, were determined. Unfortunately, in the majority of the patients' samples collected at random time points on days 14 and 22, DTAI plasma levels were below the lower limit of quantification of 0.01 ng ml−1. Pharmacokinetic blood sampling from 0 until 2 h after ingestion of E. purpurea revealed that DTAI had already reached the lower limit of quantification after 2 h .
Based on significant induction of CYP3A4 by E. purpurea in previous clinical studies with midazolam [11, 12], the pharmacokinetic interaction between E. purpurea and docetaxel was investigated in the present clinical study. In contrast to the significant interaction in our clinical study with the herbal antidepressant St John's wort and docetaxel , E. purpurea did not significantly affect systemic exposure to docetaxel in the present study. Intraindividual changes in AUC (Figure 3) were in line with an estimated intraindividual variability in docetaxel clearance of 25% . There are no clear explanations for the more remarkable 93% increase in AUC0–∞ in one patient. The incidence of docetaxel-related adverse events, however, did not increase in this patient (data not shown).
Compared with the clinical studies in which significant induction of CYP3A4 by E. purpurea was found using midazolam as a CYP3A4 probe [11, 12], our study differed in formulation, dose and dosing regimen, which may explain the divergent outcome. First, in the study of Penzak et al. , E. purpurea was administered for a longer time period (28 days). In the study of Gorski et al. , however, a shorter supplementation period (8 days) was applied, but the dosing frequency was higher (four times daily). Second, comparison between the contents of the E. purpurea formulations used in the midazolam studies [11, 12] and in the present study is complicated. The formulations of Penzak et al.  and Gorski et al.  contained 500 and 400 mg E. purpurea extract, respectively. Our commercial product was only labelled to contain 95% aerial parts and 5% roots of E. purpurea, and no information was provided about the total amount of extract used. Previously, this formulation has been shown to contain 18–34 μg ml−1 DTAI [5, 9]. In contrast, contents of DTAI or other alkylamides in the extracts used in the midazolam studies [11, 12] were not specified. Due to these differences in specifications and the absence of clinical interaction studies with Echinaforce® and midazolam, the effects of Echinaforce® on midazolam pharmacokinetics remain unknown.
Besides differences in the amounts and phytochemical content of E. purpurea extracts, differences in their origin may also have contributed to the conflicting clinical outcomes. For example, alkylamide content is known to vary considerably across different parts of E. purpurea plants , and DTAI are more abundant in roots than in leaves. Consequently, the root extract used by Gorski et al.  was likely to contain more DTAI than our product and could exert a more potent effect on CYP3A4. The ability of alkylamides and E. purpurea extracts to induce CYP3A4, however, is inconclusive according to in vitro data. Recently, Modarai et al. have shown that E. purpurea extract and isolated alkylamides did not significantly induce CYP3A4 in HepG2 cells . However, the lack of an effect on CYP3A4 could be explained by the use of HepG2 cells. It has been shown previously that LS180 cells are to be preferred over HepG2 cells to study CYP3A4 induction, because LS180 cells show higher CYP3A4 expression . In LS180 cells, our group has shown significant induction of CYP3A4 by isolated alkylamides and E. purpurea extracts using a gene reporter assay, which is a reliable method to assess the CYP3A4 induction potential of compounds . Induction of CYP3A4 became significant (P < 0.05) at relatively high concentrations of 10 and 100 μg ml−1 alkylamides and E. purpurea extract (data not shown).
In addition to the moderate CYP3A4-inducing properties of E. purpurea, its systemic exposure could be insufficient to induce hepatic CYP3A4 significantly in the present study. For example, pharmacokinetic analysis of DTAI indicated that plasma levels of these major alkylamides were undetectable or in the lower range of the calibration curve (<0.08 ng ml−1) halfway through the supplementation period. In addition, DTAI were also rapidly eliminated within 2 h after intake. This finding indicates that the absence of DTAI in plasma samples collected during the study was caused by low systemic absorption and rapid elimination of DTAI. As plasma levels of DTAI were not quantifiable or hardly quantifiable throughout the study period on days 14 and 22, compliance with E. purpurea supplementation could not be checked by pharmacokinetic analysis of DTAI. However, inspection of patient diaries and returned bottles of E. purpurea indicated that patients ingested their drops according to the schedule.
Besides the dosing regimen and content of the applied E. purpurea product, docetaxel pretreatment with dexamethasone may also have contributed to the lack of a significant effect of E. purpurea on the pharmacokinetics of docetaxel. Dexamethasone is a known inducer of CYP3A4 . Assuming induction of CYP3A4 by dexamethasone, systemic exposure to docetaxel could have been decreased already in both courses, thus making the inductive effect of E. purpurea during the second course less noticeable. However, results regarding clinical effects of dexamethasone on CYP3A4 are conflicting. A significant pharmacodynamic interaction has been shown between dexamethasone and the CYP3A4 substrate lapatinib , while dexamethasone did not significantly alter docetaxel pharmacokinetics in Asian patients . Presumably, these differences in outcomes resulted from differences in exposure to dexamethasone. In the lapatinib study, the median duration of treatment with dexamethasone was 11 days , which was substantially longer than the 3 day treatment period with dexamethasone in the study with docetaxel in Asians . These data suggest that treatment with dexamethasone for 3 days in the present study would have had only a modest inductive effect on CYP3A4.
While this study focused on CYP3A4, the drug efflux transporter P-glycoprotein (P-gp, ABCB1) is also involved in the pharmacokinetics of docetaxel. In accordance with CYP3A4, P-gp is also regulated by the nuclear pregnane X receptor. Consequently, upregulation of P-gp by E. purpurea could have resulted in decreased plasma levels of docetaxel. In clinical practice, however, the role of P-gp in docetaxel pharmacokinetics does not seem to be relevant. For example, the potent P-gp inhibitors R101933 [27, 28] and zosuquidar  did not significantly alter plasma levels of docetaxel in cancer patients. Furthermore, there were no significant associations between several P-gp polymorphisms and docetaxel clearance . Moreover, E. purpurea is unlikely to affect P-gp function, because no significant interactions were found in clinical studies with Echinacea extracts and the sensitive P-gp substrates fexofenadine  and digoxin .
Significant clinical interactions between E. purpurea and the CYP3A4 and CYP3A5 substrate midazolam indicate that E. purpurea also has the potential to interact with CYP3A5. Corresponding to CYP3A4, the polymorphic CYP3A5 enzyme is also regulated by pregnane X receptor and is involved in the metabolism of docetaxel . However, potential CYP3A5 induction is not likely to affect docetaxel pharmacokinetics significantly, because the affinity of docetaxel for CYP3A4 is approximately 10 times higher than for CYP3A5 . In agreement with this finding, no significant correlation was observed between the inactive CYP3A5*3 genotype, which is present in the majority of Caucasians, and docetaxel clearance in cancer patients [26, 30].
This study was not planned in a randomized crossover design. Considering the risk of tumour progression, randomization was not in the interest of patients with advanced cancer. Patients randomized to the group starting with E. purpurea intake would then have to wait for 14 days prior to receiving their first cycle of docetaxel. The absence of randomization may be seen as a limitation of this study. However, the fixed treatment sequence in this study is not likely to introduce substantial bias to the pharmacokinetic results.
No significant period effect of docetaxel exposure is expected based on data published in the literature [34, 35]. A modest intrapatient variability of the AUC0–24 h of docetaxel (mean ratio of cycle 2 to cycle 1 was 1.11 ± 0.14) was reported after repeated administration of docetaxel administered over 1 h at a dose of 55 mg m−2 every 3 weeks . Accordingly, a similar intrapatient variability was reported after repeated 3 weekly administration of docetaxel dosed at 100 mg m−2 over 1 h . In addition, the pharmacokinetic end-points in the present study are objective outcomes; therefore, biased results by learning effects are very unlikely. Furthermore, before the start of the second docetaxel treatment, patients underwent physical examination, and laboratory values were checked to ensure that inclusion and exclusion criteria were still met. It can thus be assumed that patients' basic medical conditions were comparable between the two cycles.
Carryover effects were also not likely to affect the pharmacokinetic results, because docetaxel levels were not quantifiable in the predose plasma samples of cycle 2. Thus, the washout period of 3 weeks was adequate.
Furthermore, a validated LC-MS/MS assay for docetaxel analysis was used, and for every patient the plasma samples of both cycles were analysed within the same analytical run. The sequence of treatment was therefore not likely to affect the bioanalysis of docetaxel.
It should be noted that the outcome of this study applies only to the specific E. purpurea formulation and dose used in the present study. As stated above, alkylamide distribution varies in different parts of E. purpurea plants  and also in several liquid E. purpurea preparations . Thus, the risk of CYP3A4-mediated interactions may be product dependent.
In conclusion, our findings showed that at the recommended dose and schedule of a commercially available E. purpurea extract no statistically significant interference with docetaxel pharmacokinetics could be demonstrated. This result indicates that the applied E. purpurea formulation may be combined safely with docetaxel.
All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: support from the Dutch Cancer Society for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work.
This work was supported by the Dutch Cancer Society [UU 2007–3795]. We would like to thank Roel Maas-Bakker (Department of Pharmaceutical Sciences, Division of Pharmacoepidemiology & Clinical Pharmacology, Utrecht University) for his technical assistance in the in vitro experiments performed at Utrecht University.