Inflammation and Immune

This is the part of the Artemisia annua story where the claims are loudest and the human evidence is thinnest, so it is worth stating the conclusion at the top rather than the bottom.

The anti-inflammatory, autoimmune, anticancer and antiviral literature on artemisinin is overwhelmingly preclinical — cell culture and rodent models, thousands of papers deep, mechanistically coherent, and largely untested in people. The human evidence consists of one small placebo-controlled trial in osteoarthritis, a 23-patient pilot in colorectal cancer that did not meet its primary endpoint, a phase I dose-finding study in metastatic breast cancer, and a body of Chinese-language work on artesunate in autoimmune disease that has not been replicated internationally.

That is not nothing. It is also not a reason to take artemisinin for lupus, rheumatoid arthritis, Crohn’s disease, cancer or a viral infection. Where the choice is between an established treatment and this, take the established treatment. Nothing below changes that.

Table of Contents

1. Where This Literature Actually Stands
2. The Mechanism: Radicals, NF-κB and T Cells
3. Osteoarthritis: The One Human Trial
4. Autoimmune Disease: Lupus, RA, Sjögren’s
5. Inflammatory Bowel Disease
6. Cancer: A Large Lab Literature and Two Human Studies
7. Antiviral Claims, Including COVID-19
8. What Would Change the Picture
9. Cautions
10. Key Research Papers
11. Connections

Where This Literature Actually Stands

A useful way to read any claim about artemisinin outside malaria is to ask three questions: Which compound? Which model? Which endpoint?

The Mechanism: Radicals, NF-κB and T Cells

Artemisinin’s immunological pharmacology is genuinely interesting, and understanding it explains both why researchers keep pursuing it and why the results have been so hard to translate.

The starting point is the same peroxide bridge that kills malaria parasites. Iron cleaves the –O–O– link and produces reactive carbon-centred radicals. In a parasite stuffed with heme, that is lethal. In a human cell, the same chemistry at lower intensity acts as a redox signal rather than a poison — and redox signalling is upstream of several of the master switches of inflammation.

Downstream, the reported effects cluster into four themes:

1. NF-κB suppression. NF-κB is the transcription factor that turns on TNF-α, IL-1β, IL-6 and the rest of the acute inflammatory programme. Artemisinin and its derivatives suppress NF-κB activation in macrophages across many models — for example reducing atherosclerotic lesions in mice via an AMPK/NF-κB/NLRP3 inflammasome axis.
2. Macrophage repolarisation. Artesunate has been shown to shift macrophages from the pro-inflammatory M1 phenotype toward the resolving M2 phenotype, with intestinal-barrier protection in mouse colitis via STAT6 signalling.
3. Th17/Treg rebalancing. This is the theme with the most direct autoimmune relevance. Artemisinin derivatives suppress Th17 differentiation and expand regulatory T cells in multiple models — and Th17-versus-Treg imbalance is a central abnormality in rheumatoid arthritis, lupus and psoriasis.
4. B-cell and antibody suppression. Artesunate reduces plasma-cell activity and autoantibody production in lupus-prone mouse strains, which is why lupus keeps reappearing in this literature.

So the mechanism is not hand-waving. The trouble is the gap between it and a patient. Artemisinin is cleared within hours, induces its own metabolism, and reaches modest plasma concentrations from oral dosing. Many of the concentrations that produce these elegant effects in a dish are higher than what an oral dose produces in blood, and far higher than a herbal tea produces. A mechanism that requires a concentration you cannot reach is a mechanism that will not become a medicine without reformulation.

Osteoarthritis: The One Human Trial

The single most relevant human study of an Artemisia annua preparation for inflammation was run in New Zealand on a commercial extract sold as a dietary supplement.

Design. Forty-two people with osteoarthritis of the hip or knee were randomised into three arms of 14: A. annua extract 150 mg twice daily (low dose), 300 mg twice daily (high dose), or placebo twice daily, over 12 weeks. Outcomes were the WOMAC index and a visual analogue pain scale.

Result. The low-dose group improved significantly from baseline: WOMAC total fell by a mean of 12.2 points and VAS pain by 21.4 mm. Placebo improved too, but not significantly (WOMAC −7.8, VAS −11.5). The high-dose group showed no significant change in anything. The extract was well tolerated. A subsequent six-month open-label extension study reported continued tolerability and benefit.

How to read this honestly. Three points, and they all cut the same way.

• These are within-group comparisons. “Improved significantly from baseline while placebo did not” is a much weaker statement than “improved significantly more than placebo.” With 14 people per arm, a between-group difference of this size would be very hard to detect, and the paper does not report one.
• The dose-response is backwards. If the drug worked, doubling it should not abolish the effect. The most parsimonious explanation for “low dose works, high dose does nothing” in a 42-person trial is chance.
• The open-label extension cannot fix either problem. Open-label pain studies reliably show improvement; that is what open-label pain studies do.

The authors themselves concluded that further studies are warranted, which is the right conclusion. Ten years on, the confirmatory trial has not appeared. In the meantime, artemisinin supplements have been linked in case reports to cholestatic liver injury — a real risk against a benefit that a 42-person pilot could not establish.

Autoimmune Disease: Lupus, RA, Sjögren’s

The autoimmune literature on artemisinin is large, growing fast, and almost entirely preclinical outside China. Reviews published in Frontiers in Immunology (2021), Immunotherapy (2023) and Heliyon (2024) collect the mechanistic work: artesunate and dihydroartemisinin suppressing Th17 expansion, promoting Tregs, dampening B-cell autoantibody production, and reducing disease scores in mouse models of lupus, collagen-induced arthritis, experimental autoimmune encephalomyelitis and psoriasis.

There is genuine intellectual pedigree here. Hydroxychloroquine — another antimalarial — is a cornerstone treatment for lupus and a real disease-modifying agent in rheumatoid arthritis. The idea that a second antimalarial class might have immunomodulatory value is not fringe reasoning; it is a well-motivated hypothesis based on a working precedent.

What is missing is the trial. In English-language literature there is no adequately powered, randomised, placebo-controlled trial of artesunate or artemisinin in lupus, rheumatoid arthritis or Sjögren’s syndrome. Chinese clinical reports exist and are cited in the reviews, but they have the same limitations that dog much of that literature — small, often unblinded, single-country, and typically published where international readers cannot appraise the methods.

If you have an autoimmune disease, the practical point is narrow and important: artemisinin does not substitute for a DMARD or a biologic, and stopping or reducing established immunosuppression in favour of a supplement can produce irreversible joint damage or organ involvement. If you want to try it alongside proper treatment, tell your rheumatologist — not least because of the CYP interactions described in the forms and dosing article, and because artemisinin-associated liver injury would be hard to distinguish from methotrexate or leflunomide toxicity.

Inflammatory Bowel Disease

Artesunate and dihydroartemisinin reduce colitis severity in mouse models — dextran-sulfate and TNBS colitis — with effects on macrophage polarisation, tight-junction proteins and epithelial barrier integrity. Delivery-focused work (for example chitosan-coated artesunate targeting the colon) reflects the same recurring theme: getting enough drug to the tissue is the limiting step.

There is no randomised human trial of artemisinin in Crohn’s disease or ulcerative colitis to report. Given that untreated inflammatory bowel disease causes strictures, fistulae, surgery and colorectal cancer risk, this is one of the worst possible conditions in which to experiment with an unproven agent instead of proven therapy.

Cancer: A Large Lab Literature and Two Human Studies

The anticancer hypothesis is easy to explain, which is part of why it spread so widely. Cancer cells often take up more iron than normal cells, expressing high levels of transferrin receptor to fuel rapid proliferation. Artemisinin needs iron to activate. Therefore, the argument goes, artemisinin should be selectively activated inside cancer cells.

The laboratory literature is enormous and broadly consistent: artemisinin derivatives induce apoptosis and ferroptosis, arrest the cell cycle, inhibit angiogenesis and reduce xenograft growth across many tumour types. Thomas Efferth’s 2017 review in Seminars in Cancer Biology is the standard entry point and is scrupulous about the field’s preclinical status.

The human evidence amounts to two studies worth knowing.

Colorectal cancer pilot (2015). Twenty-three patients awaiting curative resection were randomised to 14 daily doses of oral artesunate 200 mg or placebo before surgery; 20 completed per protocol. The primary endpoint — the proportion of patients whose tumours showed apoptosis in more than 7 percent of cells — was not met: 67 percent on artesunate versus 55 percent on placebo, a difference this trial had no power to test. Bayesian analysis of secondary markers suggested probabilities of 0.89 and 0.79 for effects on Ki67 and CD31 respectively. During a median 42 months of follow-up, 1 of 9 artesunate patients versus 6 of 11 placebo patients developed recurrent disease — a striking number, from a handful of events in a pilot that was never designed to measure recurrence. It is a signal worth a real trial, not a result.

Metastatic breast cancer, phase I (2017–2019). The ARTIC M33/2 study was a prospective, open, uncontrolled phase I study defining a well-tolerated dose of oral artesunate as an add-on in metastatic breast cancer, followed by a compassionate-use long-term add-on report. A phase I study answers one question — what dose can people take — and this one is notable for what it also found: a dedicated audiological sub-study raised questions about ototoxicity with prolonged artesunate dosing. That is a meaningful safety observation for anyone contemplating months of high-dose artemisinin.

The bottom line has not moved: artemisinin is not a cancer treatment. People do delay or decline conventional therapy on the strength of the laboratory literature, and the window in which curative treatment works does not stay open.

Antiviral Claims, Including COVID-19

Artemisinin has in-vitro activity against several viruses — cytomegalovirus, hepatitis B and C, HIV in some assays — and a much-cited 2012 paper on the “remarkable anti-HIV activity” of A. annua extracts noted that the activity did not track artemisinin content, implicating other constituents.

Then came 2020. In-vitro studies reported that artemisinin, artesunate and artemisinin-based combinations inhibited SARS-CoV-2 replication in cell culture, with the caveat that active concentrations were at or above what oral dosing achieves in blood. Meanwhile Artemisia preparations were promoted politically in several African countries as a COVID-19 remedy, most prominently as a branded herbal tonic, well ahead of any trial. Researchers in the region published a measured assessment in the American Journal of Tropical Medicine and Hygiene in 2020 laying out the anecdotal use, the political hype, the genuine treatment potential and the road map to actual randomised trials.

The randomised trials that would have settled it never produced a convincing positive result, and vaccines and antivirals with real trial evidence arrived. The episode is a useful case study: in-vitro antiviral activity is common, cheap to demonstrate, and predicts almost nothing about clinical benefit.

One related finding deserves a mention because it is legitimately interesting: A. annua and A. afra extracts show strong bactericidal activity against Mycobacterium tuberculosis in vitro, in a way not fully explained by artemisinin. That is a real laboratory observation about a disease that badly needs new drugs. It is also, once again, a dish.

What Would Change the Picture

It is fair to ask what evidence would actually justify using this plant for inflammation. Concretely:

• A between-group result, not a within-group one. A trial of at least a few hundred people showing the extract beats placebo on a validated scale, with a sensible dose-response.
• A standardised product. Trials of “Artemisia annua extract” are uninterpretable unless the artemisinin content per capsule is stated and consistent. See the forms and dosing article.
• Liver monitoring built in. Given the case reports, any serious trial of months-long dosing needs scheduled liver function tests, and those results need publishing.
• An answer on resistance. Long-term, low-dose artemisinin exposure in a population — for arthritis, say — is a novel selection pressure on any Plasmodium those people later encounter. Nobody has quantified that risk.

Cautions

• Do not replace disease-modifying treatment. For autoimmune disease, IBD or cancer, artemisinin has no established role. Substitution causes irreversible harm.
• Liver injury. Multiple published case reports link artemisinin supplements to cholestatic liver injury, including one with intrahepatic ductopenia and one assessed by updated RUCAM as immune-mediated herb-induced liver injury. A case report also documents hepatotoxicity from artesunate combined with temozolomide and Chinese herbs. Watch for jaundice, dark urine, pale stools, itching, right-upper-quadrant pain — and get liver enzymes checked if you use it for more than a few weeks.
• Ototoxicity with prolonged high-dose artesunate. Raised by the phase I breast-cancer audiological sub-study. Short antimalarial courses have generally not shown hearing loss in formal testing, but months of dosing is a different exposure.
• Pregnancy. Avoid. Artemisinins are embryotoxic in early-pregnancy animal studies.
• Asteraceae allergy. Cross-reactivity with mugwort, ragweed and sage is well documented.
• Drug interactions. Artemisinin induces CYP2B6 and CYP3A4 and autoinduces its own clearance; whole dried-leaf material can inhibit the same enzymes. This matters most for people on immunosuppressants, anticonvulsants, anticoagulants, antiretrovirals or chemotherapy — exactly the people this literature attracts.
• Never as malaria prophylaxis or treatment. See the malaria article.

Key Research Papers

Every PMID was verified live against NCBI E-utilities before publication. Each entry states whether the finding is human, animal or cell-culture.

Human studies — all of them

1. Stebbings S, Beattie E, McNamara D, Hunt S. A pilot randomized, placebo-controlled clinical trial to investigate the efficacy and safety of an extract of Artemisia annua administered over 12 weeks, for managing pain, stiffness, and functional limitation associated with osteoarthritis of the hip and knee. Clinical Rheumatology. 2016;35(7):1829–1836. Human RCT, n=42. Low dose improved from baseline; high dose did not change anything.
2. Hunt S, Stebbings S, McNamara D. An open-label six-month extension study to investigate the safety and efficacy of an extract of Artemisia annua for managing pain, stiffness and functional limitation associated with osteoarthritis of the hip and knee. New Zealand Medical Journal. 2016;129(1444):97–102. Human, open-label — no control group.
3. Krishna S, Ganapathi S, Ster IC, et al. A randomised, double blind, placebo-controlled pilot study of oral artesunate therapy for colorectal cancer. EBioMedicine. 2015;2(1):82–90. Human pilot RCT, n=23. Primary apoptosis endpoint not met.
4. von Hagens C, Walter-Sack I, Goeckenjan M, et al. Prospective open uncontrolled phase I study to define a well-tolerated dose of oral artesunate as add-on therapy in patients with metastatic breast cancer (ARTIC M33/2). Breast Cancer Research and Treatment. 2017;164(2):359–369. Human phase I, uncontrolled.
5. König M, von Hagens C, Hoth S, et al. Investigation of ototoxicity of artesunate as add-on therapy in patients with metastatic or locally advanced breast cancer: new audiological results from a prospective, open, uncontrolled, monocentric phase I study. Cancer Chemotherapy and Pharmacology. 2016;77(2):413–427. Human safety sub-study.
6. von Hagens C, Walter-Sack I, Goeckenjan M, et al. Long-term add-on therapy (compassionate use) with oral artesunate in patients with metastatic breast cancer after participating in a phase I study (ARTIC M33/2). Phytomedicine. 2019;54:140–148. Human, compassionate use.

Immunology and mechanism (cell and animal)

1. Qiu F, Liu H, Duan Y, et al. Immunoregulation by artemisinin and its derivatives: a new role for old antimalarial drugs. Frontiers in Immunology. 2021;12:751772.
2. Xie K, Li Y, Chen L, et al. Artemisinin and its derivatives as promising therapies for autoimmune diseases. Heliyon. 2024;10(7):e27972.
3. Gu J, Zhang Q, Xu Y, et al. Role of artesunate in autoimmune diseases and signaling pathways. Immunotherapy. 2023;15(14):1183–1193.
4. Jiang Y, Du H, Liu X, Fu X, Li X, Cao Q. Artemisinin alleviates atherosclerotic lesion by reducing macrophage inflammation via regulation of AMPK/NF-κB/NLRP3 inflammasomes pathway. Journal of Drug Targeting. 2020;28(1):70–79. Mouse.
5. Tao Y, Zhang Q, Meng M, Huang J. Chitosan-coated artesunate protects against ulcerative colitis via STAT6-mediated macrophage M2 polarization and intestinal barrier protection. International Journal of Biological Macromolecules. 2024;254(Pt 1):127680. Mouse.
6. Feng X, Cao S, Qiu F, Zhang B. Traditional application and modern pharmacological research of Artemisia annua L. Pharmacology & Therapeutics. 2020;216:107650.

Cancer (preclinical)

1. Efferth T. From ancient herb to modern drug: Artemisia annua and artemisinin for cancer therapy. Seminars in Cancer Biology. 2017;46:65–83. The standard review; explicit about the field’s preclinical status.
2. Ma Z, Woon CY, Liu CG, et al. Repurposing artemisinin and its derivatives as anticancer drugs: a chance or challenge? Frontiers in Pharmacology. 2021;12:828856.
3. Efferth T, Schöttler U, Krishna S, Schmiedek P, Wenz F, Giordano FA. Hepatotoxicity by combination treatment of temozolomide, artesunate and Chinese herbs in a glioblastoma multiforme patient: case report and review of the literature. Archives of Toxicology. 2017;91(4):1833–1846. Human case report.

Antiviral and antimicrobial (in vitro)

1. Cao R, Hu H, Li Y, et al. Anti-SARS-CoV-2 potential of artemisinins in vitro. ACS Infectious Diseases. 2020;6(9):2524–2531. Cell culture.
2. Zhou Y, Gilmore K, Ramirez S, et al. In vitro efficacy of artemisinin-based treatments against SARS-CoV-2. Scientific Reports. 2021;11(1):14571. Cell culture.
3. Kapepula PM, Kabengele JK, Kingombe M, et al. Artemisia spp. derivatives for COVID-19 treatment: anecdotal use, political hype, treatment potential, challenges, and road map to randomized clinical trials. American Journal of Tropical Medicine and Hygiene. 2020;103(3):960–964. The clearest contemporaneous account of the 2020 episode.
4. Lubbe A, Seibert I, Klimkait T, van der Kooy F. Ethnopharmacology in overdrive: the remarkable anti-HIV activity of Artemisia annua. Journal of Ethnopharmacology. 2012;141(3):854–859. Cell culture; activity did not track artemisinin content.
5. Martini MC, Zhang T, Williams JT, Abramovitch RB, Weathers PJ, Shell SS. Artemisia annua and Artemisia afra extracts exhibit strong bactericidal activity against Mycobacterium tuberculosis. Journal of Ethnopharmacology. 2020;262:113191. In vitro.
6. Bilia AR, Santomauro F, Sacco C, Bergonzi MC, Donato R. Essential oil of Artemisia annua L.: an extraordinary component with numerous antimicrobial properties. Evidence-Based Complementary and Alternative Medicine. 2014;2014:159819. In vitro.

Safety

1. Thio J, Rahman A, Cheah D, et al. Artemisinin-induced cholestatic liver injury and intrahepatic ductopenia. Oxford Medical Case Reports. 2024;2024(7):omae070.
2. Mathavan A, Mathavan A, Reddy R, Jones K, Cavanagh Y. Immune-mediated herb-induced liver injury: a potential association with herbal artemisinin use as supported by the updated RUCAM. BMJ Case Reports. 2023;16(5).
3. Ruperti-Repilado FJ, Haefliger S, Rehm S, et al. Danger of herbal tea: a case of acute cholestatic hepatitis due to Artemisia annua tea. Frontiers in Medicine. 2019;6:221.
4. Clark RL. Embryotoxicity of the artemisinin antimalarials and potential consequences for use in women in the first trimester. Reproductive Toxicology. 2009;28(3):285–296.

Live PubMed Searches

1. Artemisinin and NF-κB
2. Artesunate, Th17 and Treg
3. Artemisia annua and osteoarthritis
4. Artesunate and lupus
5. Artesunate and colitis
6. Artemisinin and ferroptosis in cancer
7. Artesunate cancer clinical trials
8. Artemisinin and SARS-CoV-2
9. Artemisinin and liver injury

Connections

• All Herbs
• Artemisia annua Benefits — the hub
• Artemisia annua (Sweet Wormwood)
• Artemisinin and Malaria — where the evidence is real
• Antiparasitic Uses
• Forms, Dosing, and the Herb vs Drug Question — interactions and label reading
• Malaria
• Andrographis — another bitter herb with strong preclinical immune data and thin human trials
• Wormwood (Artemisia absinthium)
• Mugwort (Artemisia vulgaris)

Back to Table of Contents