Artemisinin as a therapeutic vs. its more complex Artemisia source material

Abstract

Many small molecule drugs are first discovered in nature, commonly the result of long ethnopharmacological use by people, and then characterized and purified from their biological sources. Traditional medicines are often more sustainable, but issues related to source consistency and efficacy present challenges. Modern medicine has focused solely on purified molecules, but evidence is mounting to support some of the more traditional uses of medicinal biologics. When is a more traditional delivery of a therapeutic appropriate and warranted? What studies are required to establish validity of a traditional medicine approach? Artemisia annua and A. afra are two related but unique medicinal plant species with long histories of ethnopharmacological use. A. annua produces the sesquiterpene lactone antimalarial drug, artemisinin, while A. afra produces at most, trace amounts of the compound. Both species also have an increasing repertoire of modern scientific and pharmacological data that make them ideal candidates for a case study. Here accumulated recent data on A. annua and A. afra are reviewed as a basis for establishing a decision tree for querying their therapeutic use, as well as that of other medicinal plant species.

1. Introduction: artemisinin, its derivatives, and Artemisia sp. vs. many diseases

Artemisinin (ARTⁱ; Fig.1), produced by the medicinal plant Artemisia annua L., was identified and isolated in the 1960’s by the Chinese Project 523 and in 2015 Tu Youyou was awarded the Nobel Prize in Medicine for the discovery of artemisinin as the primary active molecule in A. annua. ¹ Use of decoctions of the plant were documented, however, in the Chinese Materia Medica and used for >2,000 yrs ^{1, 2} to treat fever, thought to be malaria. Although many people in low- and middle-income countries (LIMC) still use decoctions of the plant to treat disease, artemisinin has been isolated and derivatized into more bioavailable forms that are now combined with another antimalarial and commercialized as artemisinin combination therapies (ACTs) mainly to thwart emergence of artemisinin drug resistance.

Figure 1.

Artemisia annua and A. afra and key phytochemicals including artemisinin and thujones (found in some A. afra cultivars) and artemisinin semisynthetic derivatives

Artemisinin and A. annua have broad antimicrobial potential and also are showing promise also in alleviating chronic conditions including diabetes, inflammation, and fibrosis (Table 1). With an increasing number of studies, it is becoming apparent that A. annua is at least equal in therapeutic efficacy to artemisinin at the same concentration and is often more potent. Together those results beg the question: why not instead use the plant? Here we examine the evidence and considerations related to a more traditional use of A. annua and A. afra and whether such use translates readily for other medicinal species to develop qualifying plants as Botanical Drugs as defined by the US Food and Drug Administration (FDA).

TABLE 1.

Examples of diseases or conditions in which artemisinin and A. annua showed efficacy.

Disease / Condition	Model	Reference
Malaria	in vitro / rodent	3–6
Schistosomiasis	in vitro / rodent	7–9
Lyme (Borrelia)	in vitro	10–12
Tuberculosis	in vitro	13–15
Leishmaniasis	in vitro / rodent	16–18
Trypanosomiasis	in vitro / rodent	19, 20
Acanthamoebiasis	in vitro / rodent	21
Periodontitis	in vitro	22
HIV	in vitro	23, 24
SARS-CoV-2	in vitro	25–30
Hepatitis B	in vitro	31
Human papillomavirus	in vitro	32, 33
Fibrosis	in vitro	34
Diabetes	In vitro / rodent	35–37
Inflammation	in vitro / rodent	38–40

This review aims to summarize recent research comparing efficacy and safety of Artemisia annua, A. afra and artemisinin in treatment of five exemplar infectious diseases. These terms were combined with Boolean operators to select 69 articles from a total of 504 articles: Artemisia annua, Artemisia afra, covid-19, tuberculosis, schistosomiasis, malaria, Borrelia, synergism, toxicity, and artemisinin. The search was undertaken December 7, 2022 on PubMed and SCOPUS, limited to 2017–2022 publications excluding most reviews and book chapters excepting some context setting older articles.

1.1. A. afra an Artemisia alternative

Compared to the annual species A. annua, perennial A. afra Jacq. ex Willd has not been as well studied. For two well-documented comprehensive reviews on this species, see those by the van der Kooy lab ^{41, 42}. However, like A. annua, this medicinal plant has been used for centuries with earliest reported documentation in 1932 mainly by indigenous southern Africans ⁴². Although A. afra has many phytochemicals in common with A. annua, there are two major differences. It produces no or scarcely detectable artemisinin, but unlike A. annua, it may contain thujone (Fig. 1) ⁴². Thujone has had a mixed reputation regarding toxicity as later noted. Because A. afra is also increasing in global traditional use, it has been included in some of the topics discussed herein.

2. Safety: artemisinin vs. Artemisia

Similar to pure drugs, a Botanical Drug used for therapeutic or prophylactic purposes must comply with rigorous regulatory oversight to meet two key requirements: efficacy and safety ⁴³. For approval, a Botanical Drug also must be available as a consistent product that does not vary in quality or dosing efficacy. Compared to pure molecules that are either extracted from plants, e.g. artemisinin, or synthesized into its final form, e.g., artesunate (AS), plants respond to environmental changes that can result in considerable variation in phytochemical content. Genetic variations can also arise from cross-pollination, favored by A. annua, so seed planting can yield wide variations in artemisinin and other phytochemicals. Long-term consistency of phytochemicals including artemisinin can be achieved by clonal propagation through rooted cuttings ^44–46 or controlled self-pollination to provide homozygous seed ^{47 48}. Having target molecules to track, e.g., artemisinin in A. annua, enables validation of stable production of plant material. A botanical drug should have at least one identified molecular constituent that can be tracked through its growth, harvest, and processing. Once stable production is achieved in a specific geographical location, other botanical product goals are achievable.

A. annua and A. afra have been used for millennia by indigenous peoples to treat various ailments and are accepted as safe traditional medicines. Such use, while important, belies scientific evidence of safety, so more recently both species have been studied using standard in vitro and in vivo methods to establish overall safety. A. annua is a generally regarded as safe (GRAS) herb ^{49, 50}, yet some countries, e.g., in the EU and New Zealand, have banned or limited its sale or use. Such occurrences are the exception and the result of not being listed in their pharmacopeias; concerns about misuse as a malaria prophylactic by travelers (EU) ⁵¹; an isolated case study ⁵²; or unusual formulations, e.g., as a supercritical CO₂ extract into grapeseed oil ⁵³. Until recent rodent studies ^54–56, far less was known about the safety of A. afra. Concerns had focused on the presence of thujones, generally present in A. afra but not in A. annua. Thujones are nervous system antagonists affecting 5-hydroxytryptamine (serotonin) ⁵⁷ and γ-aminobutyric acid-A (GABAA) ⁵⁸. However, given that the LD₅₀ of thujone is 192, 230, 396 mg/kg body weight in rats, mice, and guinea pigs, respectively ⁵⁹, the small amounts (0–0.86 mg/g dry plant mass) in aqueous A. afra extracts ⁶⁰, and thujone being only partially extractable into water ⁶¹, toxicity of orally consumed A. afra from a traditional infusion is not a concern.

Rodent studies measured acute toxicity in rats of p.o. delivered A. afra reconstituted tea infusions resuspended in 0.9% saline from the freeze-dried extract of 100 g dried leaves, boiled 30 min, cooled, and filtered ⁵⁴. The single p.o. doses were 0–24 g/kg body weight (BW) and monitored for 14 d . There were no symptoms of toxicity to 2 g/kg BW, after which symptoms began to appear. Chronic p.o. administration of A. afra at 100 mg/kg BW delivered daily for three months showed there were no adverse events including 10 hematological and biochemical blood parameters. Using the same A. afra extraction protocol, Sunmonu and Afolayan ⁵⁵ measured >20 hematological and biochemical blood parameters in streptozotocin-induced diabetic rats treated with 50–200 mg/kg BW of extract daily for 15 d. Compared to diabetic-induced rats, the A. afra extract restored all liver function and hematological indices to normal except for platelets and neutrophils that although greatly improved, were not equal to untreated controls. More recently Kane et al. ⁵⁶ compared extracts of A. afra dried leaves in dichloromethane, hexane, and ethanol at acute p.o. doses of 1,000–2,500 mg/kg BW in mice. Similar to prior studies ^{54, 55}, there was no apparent toxicity. Both AST (aspartate aminotransferase) and ALT (alanine aminotransferase), important indicators of liver function, improved or remained equivalent to untreated controls in all studies. Together these rodent studies suggested that p.o.-delivered A. afra is safe.

Although most studies do not adequately provide enough detail describing extraction of the plant material, there are rodent studies showing that A. annua also is not toxic. For example, in an acute toxicity test in mice Siddiqui et al. ⁶² provided 60% ethanolic extracts of A. annua p.o at 50–5,000 mg/kg BW, then observed animals for 14 days and observed no toxic effects at any level. In another acute study in mice, Park et al. ⁶³ compared healthy mice to those treated with LPS/D-galactosamine (LPS/D-gaIN) to induce liver failure. The LPS/D-gaIN mice were treated p.o. with either an ethanolic or water extract of A. annua at 100 mg/kg BW daily for 14 d. The water extract was significantly better than the ethanolic extract at improving the AST and ALT levels in animals. In a 13-week chronic toxicity study Park et al. ⁶³ fed a water extract of A. annua to rats and also observed no adverse effects in 14 hematological and 18 biochemical blood parameters at the 1,000 mg/kg BW dose. As noted for A. afra, AST and ALT levels in all A. annua treated animals either improved or remained equivalent to untreated healthy controls regardless of length of treatment, indicating a positive effect on the liver. More recently, an ethanolic (~70%) extract of A. annua stems and leaves were tested in male and female rats for acute oral toxicity according to the OECD Test Guideline 425 (2008) limit test and also showed no toxicity at the delivered dose.

Recently evidence of safety in humans was documented in a small randomized, double-blind controlled clinical trial by Han et al. ⁶⁴ who measured the safety of an undefined A. annua hot water extract powder (SPB-201). They compared the extract powder to a crystalline cellulose placebo p.o in patients with non-alcoholic mild liver dysfunction. For 8 weeks about 48 patients/group were provided with 2 tablets TID of SPB-201 for a daily dose of 480 mg/d of A. annua dried extract or placebo tablets. There were no significant differences in 23/24 hematological parameters between placebo and A. annua treatments; only hematocrit declined a bit more in SPB-201 treated patients than those given placebo (p=0.0462). Multidimensional fatigue scale (MFS) assessments were also significantly better in the SPB-201 group. Both AST and ALT were significantly reduced at 4 and 8 weeks in the SPB-201 patients compared to placebo indicating improved liver function, a result similar to a human trial using tea infusions of A. annua and A. afra in several hundred malaria patients ⁶⁵. When compared to the standard malaria treatment recommended by WHO for that region of Africa, neither ALT nor AST were significantly altered post-treatment with A. afra or A. annua ⁶⁵. Together these human trials demonstrated the safety of both A. afra and A. annua.

Compared to A. annua and A. afra, artemisinin and its derivatives seem to have more adverse effects, but direct comparison is challenging and even LD₅₀s where available cannot be directly compared (Table 2). Of 11 possible and monitored adverse effects, Munyangi et al. ⁶⁵ observed that only 5% of malaria patients treated with tea infusions of A. annua or A. afra reported nausea or vomiting, while 42.8% of patients treated with ASAQ (the recommended ACT for the study region) reported adverse effects, especially asthenia, abdominal pain, hypoglycemia, pruritis, nausea, and vomiting. Based on that small study it seemed the Artemisia infusions were better tolerated by patients.

Table 2.

LD₅₀s for orally delivered Artemisia annua, A. afra, artemisinin, and artemisinin derivatives.

Species or drug	LD50 mg/kg BW	solvent	animal	Ref.
A. annua	2,750 7,425 4,162	hexane ether dilute alcohol*	mouse	66 67 67
A. afra	>5,000 >5,000	water water	mouse rats	68 69, 70
ART	5,105 4,228	NA	mouse	67 71
AS	1,000 −1,300	NA	mouse	71
AM	>800 1,000	NA	rat dog	71
DHA (artenimol)	700–1,500	NA	mouse	71

^*Specific alcohol not stated, but likely ethanol; NA, not available.

3. Bioavailability: artemisinin vs. Artemisia-delivered artemisinin.

There are concerns that A. annua does not provide enough artemisinin into the bloodstream to combat the malaria parasite, and hence, fosters resistance ⁷². In vivo rodent studies measuring serum levels of artemisinin showed that artemisinin is >40-fold more bioavailable when delivered per os via A. annua ⁷³, and later studies supported those results ^{74, 75}. This is consistent with earlier pharmacokinetic studies ⁷⁶ showing A. annua tea infusions containing 94.5 mg artemisinin/delivered dose/subject delivered high levels of artemisinin into the serum of healthy human subjects, and with a C_max of 240 μg/L plasma, that was well above the 9 μg/L minimum antimalarial threshold required to kill the parasite ⁷⁷.

Artemisinin is four times more soluble in digestive fluid when delivered as plant material ⁷⁸, and in a Caco-2 cell model of intestinal permeability, plant-delivered artemisinin subsequently crossed the intestinal epithelial barrier more efficiently than pure artemisinin ⁷⁹. Furthermore, phytochemicals in the Artemisia extracts inhibit cytochrome P450 2B6 and 3A4, crucial in Phase I metabolism in the intestine and liver, thereby making artemisinin highly bioavailable when delivered via the plant ³⁸. In rats there was significantly greater serum artemisinin delivered from orally ingested A. annua than from similarly administered pure artemisinin, results consistent with the P450 data. Likewise, there was significantly more artemisinin in rat organs and tissues, and a greater decline in inflammation when delivered via A. annua than from pure artemisinin. Tea infusions of both A. annua and A. afra inhibited enzymatic activity of both CYP2B6 and CYP3A4 hepatic P450s and provided mechanistic data showing the effect was not only at the level of enzymatic activity, but also at the transcription and translation steps of these two P450s ⁸⁰.

4. Therapeutic examples

4.1. Malaria

Malaria is the most studied artemisinin-susceptible disease and artemisinin combination therapy (ACT) is the currently approved treatment for malaria patients. Despite widespread use, ACTs are still not available to all because of rural logistic challenges and cost. Although WHO states that a full ACT treatment costs <US $2, that is still too expensive for a family of four that each could easily be infected twice a year. Assuming $16 for a family of four infected twice a year, that is about 4–5% of an annual subsistence income of $350–400 where, for example, in Ghana ACT costs are likely much higher and people live on $3.2–5.5/day. ACTs actually cost ~$10 for a treatment course of artemether–lumefantrine ⁸¹, so a family of four would need ~$40 for one bout of malaria (as of July 30, 2021, https://www.statista.com/statistics/1221864/middle-income-poverty-rate-in-ghana-by-level/.) The goal is to produce a malaria treatment for <US $0.50 ^{82, 83}, but highly subsidized ACTs are well beyond that price point. Consequently, many with financial limitations, especially in LMI countries, are moving towards using a more traditional use of A. annua or A. afra as a tea infusion, a growing practice that unfortunately is discouraged by the WHO ⁷². A series of recent studies are providing a base of scientific knowledge to better guide the traditional use of Artemisia for treating malaria.

Malaria is caused by numerous Plasmodium species with P. falciparum being the most lethal, and it has numerous stages requiring therapeutic treatment: liver stages (asexual), blood stages (asexual and sexual), and dormant stages (hypnozoites in the liver). The sexual gametocyte stage transfers the parasite back to the mosquito vector with the next bite helping to complete the parasite’s life cycle. Asexual parasites include trophozoite, schizont, and ring stages. The sexual stage includes male and female gametocytes that can reside in the blood for some time acting as a parasite reservoir and are responsible for asymptomatic malaria. A. annua and A. afra tea infusions kill both blood asexual ^{5, 6, 84} and gametocyte stages ^{5, 6} in vitro, results that corroborate clinical observations ^{65, 85, 86}. Recently, Ashraf et al. ⁵ showed that tea infusions of the artemisinin-deficient species, A. afra, were more effective than artemisinin-containing A. annua against both rings and schizonts, and even more effective than dihydroartemisinin (DHA). Using bioassay guided fractionation of A. afra, Moyo et al. identified two guaianolide sesquiterpene lactones, yomogiartemin and 1α, 4α-dihydroxybishopsolicepolide, that were effective against both gametocytes and asexual blood stages of P. falciparum, with IC₅₀s vs. late-stage gametocytes of 15.8 and 6.3 μM, respectively ⁸⁷. Neither was cytotoxic in vitro and the latter was especially potent against late vs. early stage gametocytes.

A contrasting study used A. annua mutants that were impaired either in artemisinin or total flavonoid production ⁸⁸. The mutant CHI1–1, deficient in chalcone isomerase, shut down major flavonoid production. Two other lines produced little to no artemisinin: AMS-silenced had insufficient amorphadiene synthase, and CYP71AV1–1 was deficient in a key P450; both enzymes are required for artemisinin biosynthesis. Compared to WT the EC₅₀s (ng/mL) of 9:1 chloroform:ethanol extracts against the Dd2 strain of P. falciparum were: 15.6, 25.9, 350, and 4,220, for WT, CHI1–1, AMS-silenced, and CYP71AV1–1, respectively. Authors concluded that despite numerous reports of individual antimalarial activity of many flavonoids as summarized in Gruessner and Weathers ⁴⁶, flavonoids played no role in the antimalarial activity of A. annua. How can one explain this discrepancy with those other antimalarial flavonoid studies and more recent ones involving the traditional infusion ^{5, 6}? It may be in the methodology. For example, Snider and Weathers ⁶ initiated experiments using dichloromethane (DCM) extracts and did not see much activity and had control vehicle (DMSO) issues (data not included in the published report). Testing hot water infusions, however, provided greater antiparasitic activity in vitro and water-based infusions are also more biosimilar. In personal communications with Dominique Mazur (Sorbonne University/Inserm/CNRS), similar solvent problems were observed and more informative results were obtained with biosimilar hot water infusions ⁵. In testing A. annua extracts against SARS-CoV-2 variants, initial DCM extracts were at best challenging to interpret. Switching to the water-based infusion also yielded more reproducible and consistent results with no DMSO vehicle issues and far less cytotoxicity than with DCM (see section 4.2 for summary data).

Using P. yoelii in mice, Li et al. ⁸⁹ compared high (100 mg/kg) and low (25 mg/kg) oral doses of artemisinin to a combination treatment of artemisinin+AB+AA+scopoletin (each at 25 mg/kg) and observed the 4-drug combination was equal in efficacy to the 100 mg/kg dose of artemisinin in eliminating the parasite. AUC, C_max and T_max for the 4-drug combination were about double that of the high artemisinin dose. Efficacy of low dose artemisinin (25 mg/kg) was about a third that of the high dose artemisinin (100 mg/kg). Results indicated that nonartemisinin compounds present in A. annua synergistically improved the activity of artemisinin.

A recent report ⁹⁰ reexamined the original studies by the Project 523 team and discovered notes indicating that a NaOH-treated ether extract also had antimalarial activity in malaria patients. Based on that rediscovered information, ether extracts ± a subsequent alkali extraction were tested in P. yoelii-infected mice to search for alternative and synergistic antimalarial compounds. Of the 113 identified compounds that included sesquiterpenoids, flavonoids, and other phytochemicals, those with antimalarial activity fell within two activity groups: 1, artemisinin and non-artemisinin antimalarial compounds, and 2, synergistic compounds. The non-artemisinin antimalarial active phytochemicals included 5α-hydroperoxy-eudesma-4(15),11-diene while the synergistic phytochemicals included arteannuin B, arteannuin B analogues, and polymethoxy flavonoids.

In some malaria species, e.g., P. vivax, hypnozoites, a dormant parasite stage, form in the liver and can later relapse into a clinical infection. Because hypnozoite metabolism is so slow, they are not particularly susceptible to antimalarial drugs that require active growth to be therapeutically effective. Recently, Ashraf et al. ⁵ showed that tea infusions of both A. afra and A. annua disrupted hypnozoites. Considering that long term consumption of A. annua is safe with no significant adverse effects, hypnozoites could be eliminated using the plant. Together these studies^{5, 89} show that both A. annua and A. afra tea infusions are effective, and since A. afra lacks artemisinin, efficacy is not solely dependent upon artemisinin which was further substantiated by Shi et al. ⁹⁰.

There is an effort to discourage the use of A. annua and A. afra to treat malaria. This is unfortunate, but not surprising considering the low price point of the plant as it would outcompete ACTs thereby obviating the profits of modern pharmaceuticals. While widespread use of a tea infusion is challenging, this traditional preparation offers a cost effective and accessible first line treatment of malaria in regions where ACTs are not readily accessible.

4.2. SARS-CoV-2 (Covid-19)

COVID-19 (SARS-CoV-2) emerged in 2019 and now is a rapidly evolving pandemic virus with estimates of >650 million cases and >6.5 million deaths all of which are deemed underestimates due to inadequate global tracking (https://coronavirus.jhu.edu/map.html). Several groups have reported that extracts of A. annua and A. afra have potent anti-SARS-CoV-2 activity ^{25, 26, 28–30}. All testing thus far has been in vitro using hot water and/or solvent extracts of A. annua.

In 2021 ²⁵, seven different dried leaf cultivars from four continents were prepared as hot water extracts (DLAe) and tested against an original strain, US WA1. All were effective with IC₅₀s (based on artemisinin content) of 0.1–8.7 μM artemisinin representing 23.4–57.4 μg leaf dry mass. One dry leaf sample was from leaves stored for >12 yrs that was equally as potent as one-year old material, indicating long term stability of anti-SARS-CoV-2 activity in dried leaves. A. annua DLAe also minimally affected Vero E6 or Calu-3 human lung cells infected with a VSV-spike pseudovirus containing the SARS-CoV-2 spike protein. Those results suggested that A. annua inhibits SARS-CoV-2 primarily by targeting the virus post-entry step. Although artemisinin and several other antimalarials including amodiaquine (AQ), artemether (AM), deoxyartemisinin (dART), and DHA were tested, most were not as potent as DLAe, yet a Spearman correlation analysis of all the artemisinin-based IC₅₀s indicated that potency increased with decreasing artemisinin content, suggesting compounds other than artemisinin in the plant were responsible for the observed antiviral efficacy and that in complex combinations, artemisinin may be antagonistic. Nie et al. ²⁹ separately confirmed non-artemisinin activity of Artemisia when they showed that A. afra lacking artemisinin also had an IC₅₀ of 0.65 μg/mL, comparable to that from four different cultivars of A. annua (ranging from 0.88 to 3.42 μg/mL). Although the IC₅₀s for all the Artemisia cultivars tested were similar to that for artemisinin, the selectivity index (SI) varied substantially from 3.52–26.89 for the Artemisia cultivars and was 4.3 for artemisinin. The higher the SI the less toxic the extract; two A. annua extracts had an SI<7. It is possible that those two cultivars had artemisinin contents equivalent to the pure artemisinin that was tested. Unfortunately, no artemisinin contents were reported.

Nair et al. has continued to test the original (frozen) DLAe samples for efficacy against emerging SARS-CoV-2 variants including delta ²⁶, omicron, and BA.4 ³⁰ and to date the A. annua hot water extracts have remained potent. Although thus far there is no identification of the non-artemisinin active phytochemicals, there are several likely candidates including isorhamnetin, luteolin, and rosmarinic acid, which are present in A. annua with IC₅₀ values of 8.42, 11.81, and 9.43 μM, respectively, against 3CLpro, the main chymotrypsin-like protease involved in SARS-CoV-2 viral replication ⁹¹. Arteannuin B, present in A. annua, also seems to have post-viral entry activity with an IC₅₀ of 12.03 μM ⁹². Other A. annua phytochemicals, e.g., quercetin and myricetin, showed activity against SARS-CoV-2 NTPase/helicase ^{93, 94}. Many of these small molecules, especially flavonoids, are showing antiviral potential and likely work synergistically as antiviral agents. Future studies are needed to identify and validate the activity of these anti-SARS-CoV-2 therapeutic compounds, and to also validate efficacy of DLAe and putatively active phytochemicals in rodent models.

Other variations of artemisinin and A. annua extracts have been tested mainly in vitro against SARS-CoV-2. Although we and others showed artemisinin is not the sole anti-SARS-CoV-2 compound in Artemisia ^{25, 29} it does have some anti-SARS-CoV-2 activity ^{25, 27–29, 92, 95, 96}. There have been some small nonrandomized clinical studies in patients treated with artemisinin-piperaquine (ART-PQ) or placebo ⁹⁷. Compared to placebo at 19.3 d, the mean time for recovery for ART-PQ treated patients (no PCR-detectable virus) was 10.6 d. All ART-PQ treated patients were virus-free after 21 d compared to 36 d for placebo. In another small trial patients who used ArtemiC, an oral spray containing artemisinin, curcumin, frankincense, and vitamin C ⁹⁸ had faster recovery vs. placebo. To my knowledge, however, there are no reports of well controlled clinical trials solely using A. annua or its extracts.

4.3. Tuberculosis

Tuberculosis (TB), a communicable disease caused by respiratory transfer of Mycobacterium tuberculosis (Mtb), accounted for ~1.2 million deaths of an estimated 10 million cases in 2020 (https://www.who.int/teams/global-tuberculosis-programme/tb-reports/global-tuberculosis-report-2021). TB is currently treated with a series of drugs, mainly rifampicin and isoniazid, taken for about 6 mon and although deaths recently rose for the first time in years, likely attributed to the SARS-CoV-2 pandemic, there were fewer cases of drug resistant TB (https://www.who.int/news/item/14-10-2021-tuberculosis-deaths-rise-for-the-first-time-in-more-than-a-decade-due-to-the-covid-19-pandemic). Nevertheless, because of emerging drug resistance there is an ongoing search for new therapeutics.

A. afra and A. annua showed some promise as TB therapeutics. Using M. aurum and Mtb H37Rv, Ntutela et al. ¹⁴ compared efficacy of A. afra aqueous (AQ), DCM, and MeOH extracts in vitro and only the DCM extract yielded significant growth inhibition, albeit less than the isoniazid control. Upon fractionation of the DCM extract, the C8 fraction had a minimum inhibitory concentration (MIC) equivalent to the isoniazid control. However, when mice infected with the Mtb were treated with a daily 1 g dose of the C8 fraction, and DCM and AQ extracts and then compared to isoniazid, there was no significant decrease in either spleen or pulmonary bacilli even after 26 wks treatment.

These results were in contrast to results form a small human trial in North Kivu (Democratic Republic of Congo) where a small number of TB patients were treated with A. annua or A. afra tea infusions (10 g dry mass/L), 330 mL every 8 hr each day for 10 days ⁹⁹. Using the Ziehl-Neelsen sputum assay, samples were negative for bacilli 10 and 15 d after treatment with A. afra and A. annua, respectively, suggesting that both species were effective in vivo; however, more study is needed. Taken together these studies suggest there may be merit to using A. afra and A. annua in treating TB; however, considerably more work is needed with well-controlled randomized clinical trials (RCTs) and to discern the mechanism of action of artemisinin and the different plant extracts.

Artemisinin was reported to block DosRST, a key stress response system in Mtb ¹⁵, and appeared to reduce viability in hypoxia, the granulosis stage of TB infection where the bacilli are less susceptible to drugs. Even in a DosR deletion mutant, artemisinin-induced gene expression changes suggesting other possible targets. Other work suggests that artemisinin may also affect lipid peroxidation in Mtb ^{100, 101}. Subsequent MIC analysis of artemisinin, A. annua, and A. afra DCM extracts against Mtb (H37Rv ΔRD1 ΔpanCD) in vitro showed that the A. annua extract was strongly bactericidal ¹³. In contrast, artemisinin at the same concentration, although also bactericidal, had slower killing kinetics. Results suggested that A. annua extracts killed Mtb through a combination of artemisinin and other phytochemicals found in the plant. Indeed, the A. afra extract with only negligible amounts of artemisinin also killed Mtb, but much slower than the A. annua extract. Similar effects were seen when the more virulent Mtb Erdman strain was tested ¹³. While the non-artemisinin active components in A. annua have not yet been identified, Bhowmick et al. suggested that artemisinic acid (AA) and deoxyartemisinin (dART) may have antimycobacterial roles ¹⁰². They reported a four-fold decrease in MIC when the AA+dART combination was tested against M. smegmatis, a mycobacterium species often used as a model to screen for anti-Mtb compounds. Others reported synergistic effects of some terpenes with rifampicin, isoniazid, and ethambutol on Mtb ¹⁰³. Unfortunately, efficacy in M. smegmatis was not particularly predictive, as Martini et al. ¹³ reported that despite significantly affecting growth, neither artemisinin nor the Artemisia extracts fully inhibited growth or killed M. smegmatis.

4.4. Lyme disease

Lyme disease affects millions of people and is mainly caused by the spirochetes, Borrelia burgdorferi, B. afzelii, and B. garinii (https://emedicine.medscape.com/article/330178-overview#a5), but may also be attributed to or co-infected with other tick-borne parasites including human Babesia (https://www.bayarealyme.org/blog/lyme-with-a-side-or-two-of-babesia-the-most-common-co-infection-that-is-frequently-missed/). Although artemisinin ^{12, 104, 105} and A. annua ¹⁰⁶ have shown in vitro efficacy against B. burgdorferi Lyme, in a mouse model they were ineffective against Babesia microti ¹⁰⁷.

B. burgdorferi afflicts humans first by an acute infection that if caught early is readily treatable with a short course of doxycycline. However, infection is not always obvious and can be missed, leading to no treatment for several weeks often resulting in “post-treatment Lyme disease syndrome” (PTLDS) with clinical symptoms that can last 6 months or longer. Although not well characterized, one posited explanation for PTLDS is the formation of a slow-growing form of the spirochete termed a persister that is especially drug tolerant. Persisters can develop in vitro and display a round body morphology in response to exposure to Lyme antibiotics, e.g. amoxicillin (Amx), doxycycline (Dox), and cefuroxime (Cef) ¹⁰⁵. In a screen of an FDA approved drug library against stationary phase cultures of B. burgdorferi, one of the top hits was artemisinin showing about 50% greater potency than Dox and Amx, and equivalent to Cef, but less potent than daptomycin (DAP) ¹². In a follow-up study using Amx-induced persisters as well as stationary-phase cultures, artemisinin was compared with a number of drugs including Dox, DAP, and cephoperazone (CefP) and in combinations at 10 μg/mL ART+Dox was not substantially different from artemisinin in its effect on persisters or stationary phase cells. However, two 3-drug combinations, ART+Dox+CefP and ART+Dox+DAP, substantially reduced both persisters and stationary phase live cell numbers. Others observed similar effects from artemisinin ¹⁰⁴ and measured MIC for artemisinin, Dox, and DAP at 5.0, 0.08, and >10 μg/mL, respectively, against stationary phase B. burgdorferi. Using the C_max drug concentrations determined from the MIC analyses, cell viability was 53.6% and 65.8% for artemisinin and Dox, respectively. Artemisinin and Dox combinations were not reported, and other drug combinations showed greater potency than artemisinin or Dox. Feng et al. ¹⁰ recently compared a variety of natural product ethanolic extracts including A. annua in 60% ethanol and measured stationary phase cell viability at 44% at MICs of 0.5 and 1%. At 1.0 and 0.5%, however, residual cells in treated cultures regrew after subculture. Dox-treated cells had 74% viability after initial treatment and also regrew upon subculture. Overall, both artemisinin and A. annua showed activity against B. burgdorferi, but considerably more studies are needed including, e.g., A. annua water extracts, multiple dosing, mechanistic studies, and testing in an animal model, to determine therapeutic potential.

4.5. Schistosomiasis

Schistosomiasis (bilharzia) is the result of trematode (blood fluke) infections caused by the genus Schistosoma transmitted by freshwater snails. WHO reports that in 2019 >105 million people were treated out of >236 million who needed a preventative treatment (July 15, 2022 accessed, https://www.who.int/news-room/fact-sheets/detail/schistosomiasis). Currently the preferred therapy is the cost-effective praziquantel (PZQ), but there are concerns regarding a need for prophylaxis and possible emerging drug resistance ¹⁰⁸.

Earlier studies suggested that when PZQ was combined with an artemisinin, e.g., AM or AS, clinical outcomes improved ¹⁰⁹, but AS alone was not as effective as either PZQ or PZQ + AS. AM and AS also were prophylactic against schistosomiasis compared to placebo ¹⁰⁹. Due to the above concerns and anecdotal evidence suggesting Artemisia had anti-schistosomal efficacy, a large clinical trial compared PZQ to A. annua and A. afra tea infusions at 1 L infusion/d for 7 d and showed faster elimination of the parasite than PZQ with fewer side effects ¹¹⁰. Although the study was later retracted, it provided clinical safety data on Artemisia-treated patients and stimulated subsequent in vitro and rodent studies ^{8, 9, 111, 112} that confirmed the efficacy of Artemisia vs. schistosomiasis.

In a recent in vitro study, S. mansoni worms were treated with A. annua (ANAMED A3) and A. afra tea infusions or extracts at 10 and 100 μg/mL ⁹. After 24 h, only the 10 μg/mL dichloromethane (DCM) extract of A. afra killed 100% of worms, while at 72 h the hexane extract was also 100% effective. A. afra tea infusions were 100% effective at 24 and 72 h, but only at 100 μg/mL. Hexane and DCM extracts of A. annua were 100% effective, but only at 100 μg/mL and at 72 h incubation. Although both Artemisia species were effective, artemisinin-free A. afra was more potent and authors posited this was due to as yet unidentified novel molecules. Yang et al. ⁸ compared artemisinin and A. annua in per os treated mice to untreated infected and uninfected controls 7 d post-infection with the free-swimming larval stage (cercariae) of S. japonicum. Treatments were at 300 mg/kg of artemisinin dosing given once a week for 4 weeks. There was no significant difference between artemisinin and A. annua treatments in reduction of both worm (95.7 vs. 98.3%, respectively) and egg levels (98.4 and 100%, respectively), indicating the effect of the commercial plant extract (Henan Yuzhou Heima Pharmaceutical Co., Ltd. Batch No. 20210715) was equivalent to artemisinin, albeit less costly to produce. In another study, Ado et al. ¹¹¹ used methanolic and aqueous extracts of A. annua and showed efficacy even against shedding of S. mansoni cercariae by its snail vector, Biomphlaria pfeifferi. More recently, Fadladdin et al. ¹¹² confirmed the in vitro efficacy of a cold-water extract of A. annua against adult male and female S. mansoni worms with subsequent validation in Golden hamsters. Histochemical analysis of the hamster tissues showed improved liver parenchyma morphology and the absence of eggs compared to untreated infected controls. Together these studies indicate that A. afra and A. annua have anti-schistosomal activity, and while artemisinin may play some part in that response, nonartemisinin phytochemicals mainly appear to be responsible.

5. Artemisinin drug resistance

Despite concerns ¹¹³ about A. annua inducing artemisinin resistance, there is no evidence that use of the plant induces artemisinin drug resistance. Not only has the plant been used for millennia to treat various fever related diseases ² with no evidence of artemisinin resistance emerging from its use, there is also in vivo evidence ⁴ that in a rodent malaria model the plant delivered at the same artemisinin level as pure artemisinin controls was at least three times more resilient against the evolution of artemisinin drug resistance than an equal amount of pure artemisinin. Furthermore, A. annua proved effective at eliminating artemisinin-resistant parasites in another rodent model, in contrast to pure artemisinin ⁴. Indeed, there are many phytochemicals in the plant that have antimalarial activity, albeit weaker than artemisinin as summarized in Gruessner and Weathers ⁸⁴ and that may explain the observed in vitro antimalarial effects of artemisinin-free Artemisia sp. ^{5, 6}. Although the WHO argues that use of A. annua causes artemisinin resistance, they have not offered any direct evidence to support that stance. The WHO claims that A. annua does not provide enough artemisinin into the bloodstream to combat the parasite, and hence, fosters resistance. As noted previously, in vivo rodent studies measuring serum levels of artemisinin showed that artemisinin is >40-fold more bioavailable when delivered per os via A. annua ⁷³, and later studies supported those results ^{74, 75}. This is consistent with earlier pharmacokinetic studies ⁷⁶ showing A. annua tea infusions containing 94.5 mg artemisinin/delivered dose/subject delivered high levels of artemisinin into the plasma of healthy human subjects, well above the 9 μg/L minimum antimalarial threshold required to kill the parasite ⁷⁷.

6. Is artemisinin the sole Artemisia therapeutic

Although artemisinin is a potent antimicrobial molecule, it may not always be the solely active therapeutic in A. annua. Monotherapies for treating infectious diseases are to be avoided to prevent emergence of drug resistance and the World Health Organization (WHO) has continued to argue that use of A. annua was an artemisinin monotherapy ⁷². Indeed, in their 2019 whitepaper, WHO ignored at least one significant prior report that countered the claim that use of A. annua would induce artemisinin drug resistance. The 2015 study ⁴ showed that when A. annua dried leaves were given per os to mice infected with Plasmodium chabaudi, artemisinin drug resistance developed three-fold slower in the mouse model than for pure artemisinin delivery at the same artemisinin dose. Furthermore, in artemisinin-resistant strains of P. yoelii, the plant proved more effective than artemisinin at eliminating the parasite. Together those results indicated that compared to pure artemisinin, whole plant delivery of artemisinin was significantly more resilient to development of artemisinin drug resistance, a problem now emerging in Africa ¹¹⁴.

The WHO also focused only on artemisinin as the active principle in the plant. While artemisinin is well-established as a potent bioactive molecule with significant antimalarial activity, a growing number of studies have shown that there is considerable antimalarial activity from the plant that extends beyond artemisinin. For example, the studies already described in Section 2.3.1 on malaria ^{5, 6, 65, 84–86} clearly demonstrate that artemisinin is not the sole antimalarial molecule in A. annua. In fact, Ashraf et al., 2022 showed that artemisinin-free A. afra has potent antimalarial activity ⁵. Although the active phytochemical(s) are not yet identified, many with in vitro antimalarial activity found in A. annua and A. afra are summarized in a recent review ⁴⁶ and in other recent studies ^{89, 90} previously discussed in Section 4.1. These non-artemisinin antimalarial responses are often synergistic with artemisinin, and there was some evidence to that effect in early in vitro studies where the flavonoids chrysoplenol-D, cirsilineol, and eupatorin reduced the IC₅₀ against P. falciparum by >50% ¹¹⁵. To clearly elucidate such a complicated therapeutic response requires further extensive investigation using biochemometric methods and such studies are in progress for malaria, TB, and Covid-19.

Another example of artemisinin not functioning as a sole therapeutic in A. annua therapy is against SARS-CoV-2. In this case artemisinin-free A. afra had a potent anti-SARS-CoV-2 response ²⁹ and the results of Nair et al. ²⁵ suggested that artemisinin might be antagonistic to the antiviral effect of A. annua.

7. Conclusions

This case presentation aimed to serve as an example of the types of questions that require empirical answers while also demonstrating the challenges that are present in trying to establish a medicinal plant that could be widely and cost-effectively implemented as a Botanical Drug. For approval as a drug, a therapeutic must be efficacious and safe. A. annua and A. afra are also edible plants, and while not necessarily palatable, they are safe for per os consumption so that makes this case somewhat unusual. Nevertheless, any edible medicinal plant should first be considered as a holistic therapeutic or prophylactic entity with different potential delivery modes and their relevant caveats (Table 3).

Table 3.

Comparisons between traditional tea infusion and encapsulated Artemisia delivery modes.

Delivery mode	Pros	Cons
Traditional tea infusion at 5 g dried leaves infused in 1 L boiled water for 10 min, cool, strain, drink 1 L/d in 3 equal aliquots.	Meshes well with traditional cultures prevalent in low- and middle-income countries	Not all compounds fully extracted.
	Artemisinin fully extracted if properly prepared.	No storage beyond 12 h;
	Artemisinin very bioavailable
	Uses boiled water so provides modicum of safety during ingestion
	Most cost-effective delivery mode
Encapsulated dried leaves: ~0.33 g powdered dried leaves/ capsule (size 00)	Acceptable delivery mode for people in more developed countries	Not a traditional delivery mode, so may be less acceptable to people in low- and middle-income countries
	No losses due to extraction	Requires potable water to consume
	Artemisinin is very bioavailable	A bit more costly than tea infusion
	Can be used as a rectal suppository in neonates and children too young to swallow a capsule or drink tea.	Must ingest large number of capsules to equate with tea infusion delivery/day
	Offers long term storage

If there are active phytochemicals in the plant, those molecules would provide a molecular tracking system. In A. annua a key tracker molecule is artemisinin with others emerging as extracts are fractionated and analyzed for therapeutic activity. For A. afra there is no specific identified molecule to track. Nevertheless, one can establish phytochemical fingerprints or chemical profiles that link to therapeutic activity. Phytochemical fingerprints also help establish consistency of plant material. Examples of tracking methods include ultraviolet, infrared, Fourier transformed infrared, mass spectroscopy, high performance liquid chromatography (HPLC), gas chromatography (GC), thin layer chromatography (TLC), capillary zone electrophoresis, liquid chromatography–mass spectrometry (LC-MS), and GC-mass spectroscopy. Although not fully addressed here, consistency of plant crops is essential for establishing a medicinal plant as an approved Botanical Drug. Cultivation, processing, and use of a crude extract or a whole plant material, e.g., dried plant material used for infusion, powdered leaves encapsulated or compressed into a tablet, also are more sustainable options. They require fewer resources than pure molecules to process as a drug, they can be more locally sourced with less likelihood of counterfeiting, and they are considerably more cost effective. Nevertheless, key requirements must be met regarding product consistency, safety, and of course efficacy. In the US meeting those requirements enabled a product to be classified as a Botanical Drug ¹¹⁶.

For approval as a Botanical Drug that has specific therapeutic claims, the same rules generally apply as for a pure molecule drug and these have been recently summarized for Botanical Drugs ^{43, 117–120}. Although as for any drug off-label use is possible, a botanical drug’s approval is specific to its use as verified through RCTs. For approval of either A. annua or A. afra as Botanical Drugs, two key tasks are lacking: demonstration of long-term supply of plant material with a consistent phytochemical content, and successful RCTs specific to at least one disease. While Botanical Drug approval is possible in the US for specific diseases like Covid-19, tuberculosis, Lyme disease etc., it is likely not a practical goal for low- and middle-income countries mainly because of the stringent quality control requirements. Once quality control criteria are met, Botanical Drug approval would be possible and could lead to off-label use for other artemisinin/Artemisia-susceptible diseases. Obtaining such approval, e.g. by the US FDA, would also provide validated support for more global use of these important medicinal plants as cost effective therapeutics for many of the Artemisia-susceptible diseases that plague not only people in developed countries but also those in the developing world.