EMA/HMPC 发布草药药品药物相互作用研究反思文件草案并征求意见（EMA/HMPC/362638/2025）
Draft reflection paper on the investigation of drug interactions of herbal medicinal products
EMA 草药药品委员会（HMPC）于 2026 年 9 月 24 日通过草药药品药物相互作用研究反思文件草案并公开征求意见，征求意见自 2026 年 10 月 15 日开始、2027 年 1 月 15 日截止。
草案针对 ICH M12 未同等涵盖的草药药品相互作用评价，按 WEU、TU 与第 8(3) 条申请分别给出数据来源与研究建议。
PDF 文字版；图形和原始排版请参阅官方 PDF。
第 1 页
1 24 September 2026
2 EMA/HMPC/362638/2025
3 Committee on Herbal Medicinal Products (HMPC)
4 Reflection paper on the investigation of drug interactions
5 of herbal medicinal products
6 Draft
Adopted by Committee on Herbal Medicinal Products (HMPC) for 24 September 2026
release for consultation
Start of public consultation 15 October 2026
End of consultation (deadline for comments). Comments should be 15 January 2027
provided using this template to [email protected]
7
8
Keywords Committee on Herbal Medicinal Products; HMPC; EU herbal monographs;
herbal medicinal products; HMPs; traditional herbal medicinal products;
THMPs; herbal substances; herbal preparations; drug-drug interactions;
DDIs; ICH M12 implementation
Official address Domenico Scarlattilaan 6 ● 1083 HS Amsterdam ● The Netherlands
Address for visits and deliveries Refer to www.ema.europa.eu/how-to-find-us
Send us a question Go to www.ema.europa.eu/contact Telephone +31 (0)88 781 6000 An agency of the European Union
© European Medicines Agency, 2026. Reproduction is authorised provided the source is acknowledged.
第 2 页
9 Reflection paper on the investigation of drug interactions
10 of herbal medicinal products
11 Table of contents
12 1. Introduction ............................................................................................ 3
13 2. Discussion ............................................................................................... 3
14 3. Conclusion ............................................................................................... 8
15 4. References .............................................................................................. 8
16
Reflection paper on the investigation of drug interactions of herbal medicinal products
EMA/HMPC/362638/2025 Page 2/9
第 3 页
17 1. Introduction
18 The scope of the ICH M12 guideline on drug interaction studies, in force from November 2024, is
19 limited to pharmacokinetic interactions with aspects generally applying to the development of small
20 chemical molecules. It formally supersedes the EMA Guideline on the investigation of drug interactions,
21 but does not address a number of topics to the same extent, e.g. recommendations for investigating
22 the drug interaction potential of herbal medicinal products (HMPs) including traditional herbal medicinal
23 products (THMPs). For further information on topics not covered by ICH M12, see document
24 implementation strategy-ICH M12. In general, the ICH M12 describes the pharmacokinetic
25 investigation of drug interactions for medicinal products. The general pharmacokinetic aspects in
26 ICH M12 are considered applicable also for (T)HMPs. The purpose of this reflection paper is to provide
27 specific recommendations on how drug interaction issues can be investigated and reported for (T)HMP
28 applications. The reflection paper should be read in conjunction with Directive 2001/83/EC, as well as
29 relevant European Union (EU) guidelines, including Notice to Applicants.
30 Directive 2001/83/EC provides definitions for herbal substances, herbal preparations, HMPs and
31 THMPs. According to these definitions a HMP is any medicinal product, containing as active ingredients
32 exclusively one or more herbal substances or one or more herbal preparations, or one or more such
33 herbal substances in combination with one or more such herbal preparations. Thus, (T)HMPs consist of
34 complex mixtures of phytochemical constituents. Moreover, although the products are manufactured in
35 controlled and standardised conditions, their composition may vary to some extent, even between
36 batches of the same product. The situation is further complicated when two or more herbal substances
37 and/or herbal preparations are combined in one (T)HMP. For most herbal substances and herbal
38 preparations, the constituents responsible for potential drug interaction effects are not known,
39 although, in some cases, certain constituents or groups of constituents may be recognised as
40 implicated.
41 In view of this specific and complex nature of (T)HMPs, guidance is needed on how to evaluate the
42 potential for clinically relevant drug interactions. The recommended approaches will depend on the
43 regulatory basis of the application of the herbal medicine, e.g. whether it is based on Article 10a well-
44 established use (WEU), Article 16a traditional use (TU) or application according to Article 8(3) of
45 Directive 2001/83/EC, as there is a difference in terms of existing data and clinical experience.
46 Therefore, the scope of this reflection paper is broader than the information covered by the ICH M12
47 on pharmacokinetic interactions, where ICH M12 generally applies to the development of small
48 chemical molecules. In addition, the reflection paper covers aspects related to EU herbal monographs
49 established by HMPC.
50 2. Discussion
51 MECHANISMS OF DRUG-DRUG INTERACTIONS (DDIS)
52 DDIs occur through several main mechanisms, typically classified into three categories:
53 physicochemical, pharmacodynamic, and pharmacokinetic interactions, all of which may be relevant for
54 (T)HMPs. The detailed mechanisms of DDIs are not within the scope of this reflection paper and hence,
55 described only briefly below.
56 Physicochemical interactions involve direct chemical or physical incompatibility between drugs, for
57 example leading to precipitation or inactivation when mixed together. However, physiochemical
58 interactions are out of scope of this reflection paper.
59 Pharmacodynamic (PD) interactions involve the combined effects of drugs at their sites of action.
60 Drugs may act on the same or related receptors or physiological systems, leading to additive,
Reflection paper on the investigation of drug interactions of herbal medicinal products
EMA/HMPC/362638/2025 Page 3/9
第 4 页
61 synergistic, or antagonistic effects. These interactions can be competitive (acting at the same receptor)
62 or non-competitive (affecting the same pathway through different mechanisms).
63 Pharmacokinetic (PK) interactions involve changes in the absorption, distribution, metabolism, or
64 excretion (ADME) of a drug:
65 • Absorption: One drug may alter the absorption of another drug by changing the pH of the
66 gastrointestinal tract, by modifying the residence time in a section of the tract (e.g. by
67 increasing bowel motility), by chelation or by affecting the activity of transport proteins.
68 • Distribution: A drug can inhibit or induce transport proteins involved in distribution of drugs
69 between different organs, leading to altered plasma and tissue levels of the involved drug.
70 One drug may also displace another drug from plasma or tissular proteins, leading to
71 increased levels of the free (active) form of the displaced drug.
72 • Metabolism: One drug may induce or inhibit the enzymes responsible for metabolising
73 another drug, leading to changes in its plasma levels.
74 − Inhibition of metabolic enzymes prevents the metabolism of other drugs which can lead
75 to increased plasma concentrations, increasing the risk of adverse effects or, if the
76 drug may not be converted to its active metabolite, reducing its therapeutic effect.
77 − Induction of metabolic enzymes occurs when a drug increases the expression of the
78 enzyme, leading to reduced plasma concentrations and consequently decreased
79 therapeutic effect (or increased effect in case of prodrugs).
80 • Excretion: One drug may alter the renal or biliary excretion of another drug by changing
81 urine pH or by affecting the activity of transport proteins in the kidneys and/or liver. A
82 drug may also alter the excretion by increasing bowel motility.
83 In ICH M12 it is emphasised that the potential for an investigational drug to be involved in
84 pharmacokinetic interactions, both as an object (effect of other drugs on the investigational drug,
85 victim) and as a precipitant (effect of the investigational drug on concomitant drugs, perpetrator)
86 should be evaluated. For (T)HMPs, the precipitant perspective is usually the most studied, while the
87 object perspective is less well understood, partly because the compounds responsible for effect and
88 adverse reactions are often not known, which makes it difficult to design meaningful pharmacokinetic
89 interaction studies.
90 GENERAL PRINCIPLES
91 Data sourcing
92 Literature data
93 Although the interaction potential of one specific (T)HMP is difficult to extrapolate from other products
94 produced from the same herbal substance, literature data to be considered in the evaluation of
95 potential DDI risk should be searched broadly. Data concerning herbal preparations that differ from the
96 intended product as well as food preparations with the same herbal substance ought to be included in
97 the search for interaction potential. In addition, data on isolated constituents should be included.
98 Post-marketing experiences
99 Information from post-marketing experience for the (T)HMP or other products with the same herbal
100 substance or preparation is of importance as it provides an opportunity to discover clinically relevant
101 DDIs. Post-marketing data can demonstrate interaction problems in the form of (increased) adverse
102 reactions or as an increased/attenuated effect of another drug. If adverse reactions or increased
103 effects of a medicinal product (i.e. object drug) occur when given together with an herbal preparation
Reflection paper on the investigation of drug interactions of herbal medicinal products
EMA/HMPC/362638/2025 Page 4/9
第 5 页
104 or its constituents, it may indicate that the elimination of the medicinal product is reduced. If, instead,
105 an attenuated drug effect is observed, the cause may be a reduced uptake or an increased elimination
106 of the drug. A clear example is the reports that revealed significant DDIs associated with the
107 consumption of St. John's wort. In the end of the 1990s, cases of prothrombin complex elevation
108 during warfarin treatment with concomitant use of St. John's wort were described. It is now known
109 that St. John's wort extracts induce cytochrome P450 enzymes and the efflux transporter protein Pgp,
110 which leads to a reduced effect of some drugs, e.g. warfarin, ciclosporin and hormonal oral
111 contraceptives.
112 New product-specific studies
113 When performing new product-specific pharmacokinetic drug-drug interaction studies, the general
114 outlines in ICH M12, are applicable (see also recommendations on Article 8(3) of Directive 2001/83/EC
115 applications below).
116 EU herbal monographs
117 When an EU herbal monograph is already established by HMPC, the Q&As R7 and R8 in the ‘Regulatory
118 questions and answers on herbal medicinal products’ (EMA/HMPC/345132/2010) describe how to relate
119 new products to that EU herbal monograph. In principle, the EU herbal monographs give applicants
120 and national competent authorities a clear reference point when preparing or assessing an application
121 for marketing authorisation or registration of herbal medicinal products in EU Member States. This also
122 includes known data on DDIs. It should also be noted, as described in the Guideline on the use of the
123 common technical document (CTD) format in the preparation of a registration application for traditional
124 herbal medicinal products (EMA/HMPC/71049/2007), that the applicant should perform own literature
125 research to fill the temporal gap between the application and the compilation of the assessment report
126 for the monograph. This should also be taken into account for WEU applications.
127 Data evaluation and interpretation
128 In the ICH M12 guideline, general information on in vitro and clinical evaluation of pharmacokinetic
129 DDIs mediated through enzyme or transporter inhibition or induction are described. There is also a
130 chapter on interpreting DDI study results. The applicant should also take into consideration herbal
131 specific guidance documents established by HMPC. Below, (T)HMP specific considerations are outlined
132 and summarised in Figure 1.
133 Data from different herbal preparations
134 In case interaction data are only available for other preparations that differ from the herbal substance
135 in question, the applicant should discuss to what extent the information is applicable to their product.
136 The general information in Q&As R7 and R8 in the ‘Regulatory questions and answers on herbal
137 medicinal products’ (EMA/HMPC/345132/2010) could support the discussion.
138 Pharmacodynamic interactions
139 Pharmacodynamic interactions may be caused by a large variety of mechanisms. It is therefore not
140 possible to give detailed guidance regarding pharmacodynamic interaction studies. The information
141 needed should be determined on a case-by-case basis. The potential for pharmacodynamic interactions
142 should be investigated for drugs which compete at the pharmacological target and/or have similar or
143 opposing pharmacodynamic (therapeutic or adverse) effects. If such drugs are likely to be used
144 concomitantly, pharmacodynamic interaction studies should be considered. Extensive pharmacological
145 and toxicological knowledge about the drug is important for the planning of pharmacodynamic
146 interaction studies. It is recommended that both in vitro studies and in vivo studies are used to
147 characterise the pharmacodynamic interaction profile.
Reflection paper on the investigation of drug interactions of herbal medicinal products
EMA/HMPC/362638/2025 Page 5/9
第 6 页
148 Pharmacokinetic in vitro data
149 Evaluating the DDI potential of an investigational drug as a precipitant involves characterising the
150 effect of the drug on major enzymes and transporters.
151 According to ICH M12, for small chemical molecules, this evaluation often starts with in vitro
152 experiments (enzyme inhibition and induction assays) to elucidate an eventual interaction potential,
153 followed by clinical DDI studies based on this mechanistic knowledge.
154 (T)HMP are often multi-component mixtures containing several chemical components that are
155 generally not defined in terms of quantity or quality. In vitro drug interaction studies may provide
156 information on the potential for interactions between (T)HMPs and other substances and on possible
157 mechanisms involved. However, for these mixtures it is challenging to draw conclusions about the
158 clinical relevance of an observed in vitro interaction, given that the identity(ies) and plasma level(s) of
159 the substance(s) responsible for the in vitro interaction are usually not known. In addition, tannins that
160 may be present in extracts, can cause enzymes (proteins) to precipitate, which could be misinterpreted
161 as enzyme inhibition. However, the in vitro studies when performed with a high concentration may
162 support the decision on the need for dedicated clinical interaction studies for evaluating the risk of the
163 specific (T)HMP causing interactions.
164 For extracts with one or a few known constituents, the situation may be more similar to small chemical
165 molecule medicinal products, taking into account the ‘Reflection paper on the level of purification of
166 extracts to be considered as herbal preparations’ (EMA/HMPC/186645/2008). In case information
167 regarding PK for the constituent(s) is available, in vitro findings can also be used to address the risk of
168 a clinically relevant DDI as described in ICH M12.
169 Pharmacokinetic animal data
170 Although animal studies have been a crucial part of drug development for many years, when it comes
171 to investigating DDIs for small molecules, these studies are not recommended due to limited
172 predictivity, due to species differences, as DDI mechanisms may not be accurately reflected in animal
173 models. However, if animal studies are available in the scientific literature, the applicant should discuss
174 the predictability of animal data for the clinical situation for the applied (T)HMP.
175 It is advised that no product-specific pharmacokinetic drug-drug interaction animal studies should be
176 performed for new HMPs.
177 Pharmacokinetic clinical data
178 Clinical DDI studies provide the best basis for evaluating the risk of specific (T)HMP causing
179 interactions, whether found as public data or as product specific studies. A clinical DDI study with a
180 sensitive and selective index substrate may elucidate a mechanism and the results can usually be
181 extrapolated to effects on other drugs as objects of interaction, i.e. drugs with similar metabolic and
182 distribution pathways. Considerations of importance when designing prospective clinical DDI studies
183 are described in ICH M12, section 3.2. Recommendations regarding the overall design, population and
184 number of subjects, dosing, route of administration etc. are valid also for studies with (T)HMPs.
185 Recommendations on different kinds of applications
186 Applications according to Article 10a (WEU) and Article 16a (TU) of Directive 2001/83/EC
187 Information about interactions (PK and PD) between medicinal products and HMPs based on WEU or TU
188 should, in general, be based on broadly sourced scientific literature, other contributing public data, and
189 post-marketing experiences, as described above.
Reflection paper on the investigation of drug interactions of herbal medicinal products
EMA/HMPC/362638/2025 Page 6/9
第 7 页
190 The available data should be evaluated as described above and the risk for DDI(s) should be discussed
191 in the application. In certain cases, if the mechanism of an observed interaction is known (i.e. if a
192 clinical DDI study with a sensitive index substrate has been performed), the risk can be extrapolated to
193 other drugs as objects of interaction, e.g. drugs with similar metabolic or distribution pathways.
194 Applications according to Article 8(3) of Directive 2001/83/EC
195 The potential for interactions (PK and PD) should be investigated by the applicant, based on publicly
196 available data, as described above, and as product-specific studies if sufficient published data are not
197 found.
198 In vitro studies on the enzyme inhibition and induction potential may be performed to indicate
199 potential interaction mechanism(s) (see ICH M12 guideline, section 2).
200 Determination of endogenous biomarkers of enzymes or transporters in blood or urine in clinical
201 studies may be useful to get indications of the HMP’s potential as a precipitant (see ICH M12 guideline,
202 section 3.2.7).
203 If clinical DDI studies are needed to obtain product specific information about an HMP, they should
204 follow the ICH M12 guideline, section 3, as far as possible. For extracts with one or a few known
205 constituents and a known mechanism of action, DDI studies should be considered also from the object
206 perspective.
207 The strategy used to determine which clinical studies to perform must be selected on a case-by-case
208 basis. The applicant should justify the choice of study design and substrates.
209 Topical products (Articles 10a (WEU), 16a (TU) and 8(3) of Directive 2001/83/EC)
210 For topical products, the degree of absorption needs to be evaluated. If systemic absorption is
211 negligible, the focus should be on possible interactions with other topical products, with attention to
212 medicinal products that are likely to be administered concomitantly to the target population. If
213 systemic absorption cannot be excluded, the risk for systemic DDIs needs to be addressed in line with
214 the sections above.
215 Product information
216 The Guideline on summary of product characteristics (SmPC) (EudraLex Vol. 2C) gives advice on how
217 to present information about interactions. Information about drug interactions should be presented in
218 the SmPC sections 4.5 and 5.2 and cross-referred to the sections 4.2, 4.3 or 4.4 when relevant.
219 Section 4.5 should contain all detailed information on drug interactions and only the recommendation
220 should be given in the cross-referred sections.
221 Regardless of the prescription status, the conclusions of the evaluation regarding potential risk of
222 interactions should be translated into actionable recommendations for use of the product, to be
223 included in its product information.
224 In case in vivo interaction studies from other preparations of the herbal substance in question are
225 considered relevant, the data should be presented in the product information. In such cases, the
226 magnitude of the interaction could be given in the SmPC of the (T)HMP together with information on
227 the difficulty to extrapolate the magnitude of interactions from other (T)HMPs or food products
228 manufactured with the same herbal substance.
229 For (T)HMPs, in line with EU herbal monographs, the SmPC section 4.5 ‘Interactions with other
230 medicinal products and other forms of interaction’, and corresponding text in patient leaflet (PL) and
231 labelling, commonly state that “None reported” or “No interaction studies have been performed”. Both
Reflection paper on the investigation of drug interactions of herbal medicinal products
EMA/HMPC/362638/2025 Page 7/9
第 8 页
232 statements mean that no interaction studies have been performed, and no clinically relevant
233 interactions have been found in the scientific literature for the time being.
234 Importantly, if there is a lack of data on DDI, patients should be made aware that there is no
235 information available on possible interactions and that interactions may occur. Since over-the-counter
236 (OTC) medicines are in most cases avoidable, it is important that the patients are made aware that
237 herbal medicinal products could cause interactions, just like any other medicinal product.
238 3. Conclusion
239 The complexity of (T)HMPs requires guidance on the evaluation of their potential for clinically relevant
240 drug interactions. The recommended approaches will depend on whether the product is applied for as
241 an Article 8(3) of Directive 2001/83/EC application, or as a WEU or TU, due to differences in terms of
242 data generation versus existing data and clinical experience.
243 In general, information to be considered in the evaluation of potential DDI risk should be gathered
244 broadly and include scientific literature, other contributing public data and post-marketing experiences.
245 For Article 8(3) of Directive 2001/83/EC application, published data could be a part of the evaluation,
246 but product-specific clinical DDI studies may be needed. Although the testing strategy may be decided
247 on a case-by-case basis, depending on the product characteristics, product-specific interaction studies
248 should follow the ICH M12 guideline, section 3, to the extent possible.
249 For WEU and TU applications the evaluation is based on available published data.
250 The risk for DDI(s) of the (T)HMP should be concluded based on all available data and translated into
251 general recommendations to be included in its product information.
252 4. References
253 Guideline on summary of product characteristics (SmPC), EudraLex Vol. 2C
254 Guideline on the investigation of drug interactions, CPMP/EWP/560/95
255 Guideline on the use of the CTD format in the preparation of a registration application for traditional
256 herbal medicinal products, EMA/HMPC/71049/2007
257 ICH M12 Guideline on drug interaction studies, EMA/CHMP/ICH/652460/2022
258 Implementation strategy of ICH Guideline M12 on drug interaction studies, EMA/460496/2024
259 Notice to Applicants Vol. 2A ‘Procedures for marketing authorisation, Chapter 1 Marketing
260 Authorisation
261 Reflection paper on the level of purification of extracts to be considered as herbal preparations
262 EMA/HMPC/186645/2008
263 Regulatory questions and answers on herbal medicinal products, EMA/HMPC/345132/2010
264
Reflection paper on the investigation of drug interactions of herbal medicinal products
EMA/HMPC/362638/2025 Page 8/9
第 9 页
265 Figure 1- Paths for consideration for the general collection of relevant DDI data.
266 The articles (Art.) refer to the corresponding Articles of Directive 2001/83/EC.
267
Reflection paper on the investigation of drug interactions of herbal medicinal products
EMA/HMPC/362638/2025 Page 9/9
来源：EMA Scientific Guidelines · ema.europa.eu