Atlas chapter 17 · Gastrointestinal

Agaricus subrufescens

E2-3 ±± — Materially mixedR1 — one controlled clinical programme was identified for the grade-owning AndoSan ulcerative-colitis symptom and quality-of-life proposition (Therkelsen et al., 2016). Multiple publications arising from the same recruitment programme or cohort are not counted as independent replications. The companion Crohn's disease study used the same investigators and preparation and did not demonstrate clear placebo-relative superiority on its principal between-group comparison. Other AndoSan studies and clinical studies of different Agaricus-containing preparations concern different populations, interventions or outcomes and are not counted as replications of this proposition. R1 describes replication depth, not trial quality, effect direction or the wider existence of human investigation.

Proposition / claim: Preparation-specific symptom and quality-of-life human signals, strongest in a short controlled ulcerative-colitis study of the mixed formula AndoSan.
Evidence owner: The mixed formula AndoSan owns the strongest controlled ulcerative-colitis symptom and quality-of-life signal. Other human findings belong to the particular Agaricus-containing preparation or formulation actually studied. Effects observed with mixed formulas cannot automatically be assigned to Agaricus subrufescens alone, and evidence from one ABM/ABMK-labelled preparation cannot automatically be transferred to another.
Dominant uncertainty: Benefit and harm attribution. Benefit: what, if anything, does A. subrufescens itself contribute to effects observed with mixed formulas or defined commercial preparations? Harm: which exact exposure owns the liver-injury signal? These are mirror-image attribution problems and must be resolved using the same evidence-ownership discipline.

Organism: Agaricus subrufescens
Tested intervention: Defined Agaricus subrufescens-containing preparations and mixed fungal formulations used in the human clinical programme, including AndoSan. AndoSan is a mixed preparation and is not Agaricus subrufescens alone: published descriptions identify it as approximately 82.4% Agaricus material, 14.7% Hericium erinaceus and 2.9% Grifola frondosa. Those proportions do not establish that the Agaricus component caused the observed clinical effect.
Evidence target: Standardized preparation
Intervention description: The human literature includes materially different interventions: whole fungal preparations, extracts, ABM/ABMK-labelled products and mixed formulations such as AndoSan. Product, manufacturer, fungal material, extraction or processing method, dose and batch information must be retained wherever available. Acronym continuity does not establish preparation continuity. Evidence remains attached to the intervention actually tested. If AndoSan produces a clinical effect, the evidence owner is AndoSan — not automatically Agaricus subrufescens. Historical medicinal material is frequently reported as 'Agaricus blazei Murrill' and sometimes 'Agaricus brasiliensis'; taxonomic and molecular work supports conspecificity of the Brazilian cultivated medicinal lineage with A. subrufescens, but the printed historical binomial alone does not establish modern molecular authentication. The name reported by the original study is preserved and crosswalked only where the medicinal lineage is reasonably identifiable from provenance.
Population: Selected human clinical populations studied with defined Agaricus-containing preparations, with the strongest controlled symptom and quality-of-life signal arising in 50 people with symptomatic ulcerative colitis receiving AndoSan for 21 days. Other clinical populations may be acknowledged in the intervention programme description but do not inherit the ulcerative-colitis result.
Comparator: Placebo in the strongest ulcerative-colitis randomized, single-blind controlled study; comparator structures vary across the wider preparation-specific human literature.
Outcome: Selected symptoms, fatigue and health-related quality-of-life outcomes. In the ulcerative-colitis study, several patient-reported outcomes improved relative to placebo, without a corresponding major fecal-calprotectin or general blood-measure change. Symptom improvement is not demonstrated inflammatory control, and patient-reported improvement is not proof of broad anti-inflammatory or immune-regulating activity.
Evidence maturity (E): E2-3 ±
Direction: ± — Materially mixed
Historical flag (H): Not specified at dossier level
Replication (R): R1 — one controlled clinical programme was identified for the grade-owning AndoSan ulcerative-colitis symptom and quality-of-life proposition (Therkelsen et al., 2016). Multiple publications arising from the same recruitment programme or cohort are not counted as independent replications. The companion Crohn's disease study used the same investigators and preparation and did not demonstrate clear placebo-relative superiority on its principal between-group comparison. Other AndoSan studies and clinical studies of different Agaricus-containing preparations concern different populations, interventions or outcomes and are not counted as replications of this proposition. R1 describes replication depth, not trial quality, effect direction or the wider existence of human investigation.
Claim Gap (CG): CG0-1: A short randomized trial of the mixed preparation AndoSan reported improvements in selected symptoms, fatigue and quality-of-life outcomes in ulcerative colitis. CG1: Some defined Agaricus-containing preparations have produced preliminary human symptom, quality-of-life or immune-related signals. CG2-3: "Agaricus subrufescens supports immunity." The intervention and outcome have become vague. CG4: "Agaricus is clinically proven to fight cancer." The direct human evidence does not justify the broad treatment claim. CG4: "Agaricus is a safe natural immune booster." Both efficacy and safety have been overgeneralized. Claim ceiling: CG1-2 for formulation-specific statements.
Evidence Transfer Vector: Transferred proposition: "Any A. subrufescens supplement safely boosts immunity." — T / I / O / R. T — Taxon: historical medicinal "A. blazei" → modern product identity. I — Intervention: AndoSan or another defined preparation → arbitrary supplement. O — Outcome: selected symptoms or biomarkers → broad immune function. R — Replication: small preparation-specific programme → established general effect. Safety runs beside this vector rather than inside it: preparation-specific adverse-event evidence is converted neither into universal safety nor into universal toxicity.
Taxonomic Transfer Confidence: MODERATE-HIGH contextually. Historical medicinal "A. blazei" → A. subrufescens: HIGH where the Brazilian/Japanese medicinal lineage is clear from provenance. Generic historical "A. blazei": MODERATE where contextual or biological documentation is insufficient. The taxonomy is recoverable in many cases but must not be assumed in all cases.
Intervention Transfer Confidence: LOW across formulas and generic species products. AndoSan → A. subrufescens alone: LOW. One named ABM/ABMK preparation → arbitrary Agaricus supplement: LOW. A standardized A. subrufescens preparation → a chemically and biologically equivalent replication product: MODERATE to HIGH only if equivalence is demonstrated. Formula evidence cannot be decomposed by assumption.
Safety: S1-2, with a serious liver-injury signal requiring preparation-specific attribution. A 2006 case series described three people with advanced cancer who developed severe hepatic injury while taking an Agaricus blazei extract; two died with fulminant hepatitis, and one case included improvement after withdrawal and worsening after re-exposure. This is a meaningful serious safety signal, although other causes could not be completely excluded, and later reports involve Agaricus-containing complementary products sometimes taken alongside other supplements or anticancer treatment. Balancing exposure evidence: a six-month phase I study of 78 cancer survivors taking a defined Agaricus blazei granulated powder at several doses found that most participants did not experience significant problems within measured clinical and laboratory parameters; adverse events were mainly gastrointestinal, one participant had a liver-dysfunction-related allergic finding, and events were not dose-dependent. The conclusion is preparation-specific human exposure with an unresolved but meaningful serious-harm signal. Neither "safe" nor "hepatotoxic" is an acceptable species-wide verdict.
Conservation: C0/C+ — A. subrufescens is cultivable and commercial production need not depend on high-pressure harvesting of rare wild populations. Controlled cultivation may additionally improve authentication, batch standardization and forensic investigation of safety signals. Cultivability does not imply therapeutic equivalence.
Resolution zone: RZ-5 → RZ-2/3
Next best experiment: Stage one — intervention and safety reconstruction. For every grade-driving human study and serious adverse-event report, reconstruct where recoverable: reported species identity; product and manufacturer; fungal tissue or culture material; other ingredients; extraction/manufacturing method; dose; duration; co-medications; clinical population; outcomes; batch information. Where retained suspected-toxicity product is available, consider fungal authentication, ingredient verification, contaminant screening, chemical fingerprinting, batch consistency, and comparison with uneventful batches where available. Contaminants, adulterants or interactions are not assumed to explain the signal without evidence. Stage two — attribution trial. If the mixed formula remains the strongest clinical lead, compare the full formula, a standardized Agaricus-only preparation matched as closely as scientifically possible, and placebo. A formula-without-Agaricus arm may be added only if scientifically and practically justified. Build prospective liver-function monitoring and structured adverse-event surveillance into the study. The objective is the smallest direct comparison capable of determining whether Agaricus contributes materially to the observed benefit while simultaneously improving preparation-specific safety knowledge.
Primary references: BENEFIT PROGRAMME — grade-owning study. Therkelsen SP, Hetland G, Lyberg T, Lygren I, Johnson E. Effect of a Medicinal Agaricus blazei Murill-Based Mushroom Extract, AndoSan™, on Symptoms, Fatigue and Quality of Life in Patients with Ulcerative Colitis in a Randomized Single-Blinded Placebo Controlled Study. PLoS One. 2016 Mar 2;11(3):e0150191. doi:10.1371/journal.pone.0150191. PMID 26933886. — Randomized single-blind placebo-controlled study, ~50 patients with symptomatic ulcerative colitis, 21 days. Symptom, fatigue and IBDQ quality-of-life outcomes improved relative to placebo; no corresponding major fecal-calprotectin change. Owns the grade for this dossier.
BENEFIT PROGRAMME — companion Crohn's disease study, same investigators, same product, same recruitment programme. Therkelsen SP, Hetland G, Lyberg T, Lygren I, Johnson E. Effect of the Medicinal Agaricus blazei Murill-Based Mushroom Extract, AndoSan™, on Symptoms, Fatigue and Quality of Life in Patients with Crohn's Disease in a Randomized Single-Blinded Placebo Controlled Study. PLoS One. 2016 Jul 14;11(7):e0159288. doi:10.1371/journal.pone.0159288. PMID 27415795. — The principal between-group symptom and quality-of-life comparisons did not demonstrate clear superiority to placebo; the positive findings were within-group improvement from baseline. NOT a positive replication.
BENEFIT PROGRAMME — secondary biomarker analysis of the SAME patients reported in the two 2016 PLoS One papers. Therkelsen SP, Hetland G, Lyberg T, Lygren I, Johnson E. Cytokine Levels After Consumption of a Medicinal Agaricus blazei Murill-Based Mushroom Extract, AndoSan™, in Patients with Crohn's Disease and Ulcerative Colitis in a Randomized Single-Blinded Placebo-Controlled Study. Scand J Immunol. 2016 Dec;84(6):323-331. doi:10.1111/sji.12476. PMID 27588816. — Publication overlap: not an independent cohort and not a replication.
BENEFIT PROGRAMME — earlier IBD pilot cohort (mixed ulcerative colitis and Crohn's disease), cytokine and calprotectin endpoints, same Oslo group. Førland DT, Johnson E, Saetre L, Lyberg T, Lygren I, Hetland G. Effect of an extract based on the medicinal mushroom Agaricus blazei Murill on expression of cytokines and calprotectin in patients with ulcerative colitis and Crohn's disease. Scand J Immunol. 2011 Jan;73(1):66-75. doi:10.1111/j.1365-3083.2010.02477.x. PMID 21129005. — Biomarker sub-study; does not test the symptom/quality-of-life proposition.
BENEFIT PROGRAMME — healthy-volunteer cohort (n=8), oral AndoSan; first of two biomarker papers from the same subjects. Johnson E, Førland DT, Sætre L, Bernardshaw SV, Lyberg T, Hetland G. Effect of an extract based on the medicinal mushroom Agaricus blazei murill on release of cytokines, chemokines and leukocyte growth factors in human blood ex vivo and in vivo. Scand J Immunol. 2009 Mar;69(3):242-250. doi:10.1111/j.1365-3083.2008.02218.x. PMID 19281536.
BENEFIT PROGRAMME — same healthy-volunteer cohort, different biomarker panel. Johnson E, Førland DT, Hetland G, Sætre L, Olstad OK, Lyberg T. Effect of AndoSan™ on expression of adhesion molecules and production of reactive oxygen species in human monocytes and granulocytes in vivo. Scand J Gastroenterol. 2012 Sep;47(8-9):984-992. doi:10.3109/00365521.2012.660544. PMID 22564240. — Publication overlap with the 2009 paper; one cohort, two analyses.
BENEFIT PROGRAMME — separate indication and population; cannot replicate the ulcerative-colitis proposition. Tangen JM, Tierens A, Caers J, Binsfeld M, Olstad OK, Trøseid AM, Wang J, Tjønnfjord GE, Hetland G. Immunomodulatory effects of the Agaricus blazei Murrill-based mushroom extract AndoSan in patients with multiple myeloma undergoing high dose chemotherapy and autologous stem cell transplantation: a randomized, double blinded clinical study. Biomed Res Int. 2015;2015:718539. doi:10.1155/2015/718539. PMID 25664323. — n=40, immunological endpoints only.
NON-ANDOSAN AGARICUS CONTEXT — different preparation (ABMK), different indication, independent group. Ahn WS, Kim DJ, Chae GT, Lee JM, Bae SM, Sin JI, Kim YW, Namkoong SE, Lee IP. Natural killer cell activity and quality of life were improved by consumption of a mushroom extract, Agaricus blazei Murill Kyowa, in gynecological cancer patients undergoing chemotherapy. Int J Gynecol Cancer. 2004 Jul-Aug;14(4):589-594. doi:10.1111/j.1048-891X.2004.14403.x. PMID 15304151. — Context only; not transferable to AndoSan or to ulcerative colitis.
NON-ANDOSAN AGARICUS CONTEXT — different preparation and indication. Hsu CH, Liao YL, Lin SC, Hwang KC, Chou P. The mushroom Agaricus blazei Murill in combination with metformin and gliclazide improves insulin resistance in type 2 diabetes: a randomized, double-blinded, and placebo-controlled clinical trial. J Altern Complement Med. 2007 Jan-Feb;13(1):97-102. doi:10.1089/acm.2006.6054. PMID 17309383.
NON-ANDOSAN AGARICUS CONTEXT — very weak design (open-label pilot, n=4), noted for completeness only. Hsu CH, Hwang KC, Chiang YH, Chou P. The mushroom Agaricus blazei Murill extract normalizes liver function in patients with chronic hepatitis B. J Altern Complement Med. 2008 Apr;14(3):299-301. doi:10.1089/acm.2006.6344. PMID 18370584. — No randomization, blinding or control.
SAFETY PROGRAMME — index serious-harm report. Mukai H, Watanabe T, Ando M, Katsumata N. An alternative medicine, Agaricus blazei, may have induced severe hepatic dysfunction in cancer patients. Jpn J Clin Oncol. 2006 Dec;36(12):808-810. doi:10.1093/jjco/hyl108. PMID 17105737. — Three patients with advanced cancer developed severe hepatic dysfunction while taking an Agaricus blazei (Himematsutake) extract; two died with fulminant hepatitis; one improved after withdrawal and worsened on re-exposure. The exposure is described generically, without a named manufacturer or batch, so the harm cannot be attributed to a specific preparation.
SAFETY PROGRAMME — later hepatic case report. Hisamochi A, Kage M, Arinaga T, Ide T, Miyajima I, Ogata K, Kuwahara R, Koga Y, Kumashiro R, Sata M. Drug-induced liver injury associated with Agaricus blazei Murill which is very similar to autoimmune hepatitis. Clin J Gastroenterol. 2013;6(2):139-144. doi:10.1007/s12328-013-0359-0. — Single case of severe drug-induced liver injury histologically resembling autoimmune hepatitis; product not identified by brand or manufacturer. PMID reported as 26181451 but not independently confirmed on PubMed during this pass.
SAFETY PROGRAMME — non-hepatic serious adverse event under mixed exposure. Iwai N, Okuda T, Sawada R, Ohara T, Hattori C, Taniguchi M, et al. Interstitial Lung Disease Associated with Agaricus blazei Murill in a Patient with Pancreatic Ductal Adenocarcinoma Receiving Gemcitabine-Based Therapy. Case Rep Gastroenterol. 2022;16(1):229-234. doi:10.1159/000522639. — Agaricus blazei extract taken concurrently with gemcitabine and nab-paclitaxel, both of which can cause pneumonitis; attribution is confounded by co-medication. PMID not confirmed during this pass.
SAFETY PROGRAMME — recent hepatic case, product identity unresolved. Strobbe G, Jeanjacquot M, Potey C, Carnot A, Marliot G. Grade 3 cytolysis in a patient with metastatic colorectal cancer consuming a mushroom powder-based alternative therapy. J Oncol Pharm Pract. 2024 (online 2 Sep 2024); assigned 2025;31(2):329-335. doi:10.1177/10781552241280617. — Hepatic cytolysis and cholestasis during treatment for metastatic colorectal cancer. The exposure is described as a mushroom-powder blend; Agaricus specificity is NOT confirmed from the abstract and requires full-text verification before this case is counted in the Agaricus safety ledger. PMID not confirmed during this pass.
SAFETY PROGRAMME — balancing exposure evidence. Ohno S, Sumiyoshi Y, Hashine K, Shirato A, Kyo S, Inoue M. Phase I Clinical Study of the Dietary Supplement, Agaricus blazei Murill, in Cancer Patients in Remission. Evid Based Complement Alternat Med. 2011;2011:192381. doi:10.1155/2011/192381. PMID 21584278. — Approximately six months of a defined Agaricus blazei Murill granulated powder at several dose levels in cancer patients in remission; most participants had no significant problems within measured clinical and laboratory parameters, adverse events were mainly gastrointestinal, and events were not dose-dependent. Exact adverse-event wording, including the liver-dysfunction-related allergic finding, requires full-text confirmation against the published adverse-event table.
SAFETY / TOXICOLOGY CONTEXT — narrative review. Firenzuoli F, Gori L, Lombardo G. The Medicinal Mushroom Agaricus blazei Murrill: Review of Literature and Pharmaco-Toxicological Problems. Evid Based Complement Alternat Med. 2008;5(1):3-15. doi:10.1093/ecam/nem007. PMID 18317543. — Narrative, not a systematic review; discusses the Mukai 2006 hepatotoxicity cases and preparation-quality concerns.
SAFETY / TOXICOLOGY CONTEXT — species-level chemistry, not clinical evidence. Malík M, Malá L, Killi UK, Plucar B, Mika OJ, Navrátilová Z, Olekšák P, Patočka J. Agaritine: A Natural Toxic Amino Acid of Cultivated Mushroom Agaricus spp. and Its Potential Health Risk. Mil Med Sci Lett. 2025;94(3):115-121. doi:10.31482/mmsl.2024.008. — Hydrazine-derivative background relevant to the safety ledger; contributes no human adverse-event attribution.
Source chapter: Chapter 17, The Fungal Evidence Atlas