Ivermectin
Ivermectin is a worm and parasite medicine that is being tested as an add-on to immunotherapy in two small early trials. No trial has shown that it treats any cancer, and people who have dosed themselves outside a trial have ended up in hospital with seizures or liver damage.
Overview
Ivermectin is approved in the United States as Stromectol for intestinal strongyloidiasis and onchocerciasis, two parasitic worm infections, and for nothing else (Stromectol label on DailyMed). The approval dates from 22 November 1996 under NDA 050742 (Drugs@FDA NDA 050742). The World Health Organization lists it as an essential medicine for onchocerciasis, strongyloidiasis, other soil-transmitted worm infections and scabies (WHO Model List of Essential Medicines).
Reviews collect cell-culture and mouse experiments in which ivermectin slowed cancer cells or shrank tumours in animals through several pathways at once, and note that no large randomised trial has confirmed any benefit in people (Patel review 2025). The most cited animal experiment reported that ivermectin plus an anti-PD-1 antibody limited tumour growth in mouse models of breast cancer where neither agent alone worked (Draganov mouse study 2021). That paper now carries an editorial expression of concern because several images in one figure were found to be duplicates and the authors said the original data were no longer available (Expression of concern 2026).
All 239 ivermectin studies registered on ClinicalTrials.gov were screened for this page, and two are cancer treatment trials: a phase 1/2 study at Cedars-Sinai giving ivermectin with balstilimab or pembrolizumab to about 34 patients with metastatic triple-negative breast cancer, recruiting since October 2023 with no results posted (ClinicalTrials.gov NCT05318469), and ICONIC, a University of Florida phase 2 comparing two ivermectin doses alongside a standard checkpoint inhibitor in 80 patients with solid tumours, not yet recruiting and measuring a change in immune cells rather than tumour response (ClinicalTrials.gov NCT07487805). The ICONIC registry entry says that among the first nine patients treated in the Cedars-Sinai study no treatment-related serious adverse events were observed, which is the only human safety information from either trial so far (ClinicalTrials.gov NCT07487805). Two other registered studies are not tests of ivermectin against cancer: a São Paulo phase 2 of ivermectin plus losartan for COVID-19 in cancer patients, stopped for futility after 77 of a planned 176 patients (ClinicalTrials.gov NCT04447235), and a Mexican clinic's 'atavistic chemotherapy' protocol that lists ivermectin among eight antiprotozoal drugs, last updated in 2022 with no results (ClinicalTrials.gov NCT02366884). No phase 3 trial exists, and neither cancer trial has reported whether ivermectin shrinks tumours or lengthens life (ClinicalTrials.gov NCT05318469).
Outside trials the human evidence is survey and anecdote: 19 percent of cancer patients interviewed in Loja, Ecuador said they took ivermectin alongside chemotherapy, radiotherapy or immunotherapy (Loja survey 2023). A telemedicine cohort of 197 patients prescribed ivermectin with mebendazole reported self-assessed benefit in 84 percent of the 122 who answered a six-month survey (Hulscher cohort 2026). The journal has placed an expression of concern on that paper while it audits ethical approval and whether the diagnoses and reported regressions can be verified (Anticancer Research expression of concern 2026). A review for gynaecological cancers finds the clinical data limited to cell lines and 'strongly cautions' against use (Mujumdar review 2025), and the accompanying editorial notes that no major oncology organisation, including ASCO, SGO or NCCN, endorses ivermectin for cancer (Straughn editorial 2025).
Harm from self-dosing is documented: a 73-year-old woman with metastatic breast cancer who took high doses on the strength of online information developed seizures and respiratory failure and needed a ventilator, recovering within 48 hours (Saperstein case report 2026). A 65-year-old man with prostate cancer who alternated veterinary fenbendazole and ivermectin for three months developed severe liver injury that resolved within six weeks of stopping (Powderly case report 2026). An Oregon poison-centre series of 37 people who took ivermectin against COVID-19, most men over 60, recorded neurotoxicity in 30, hospital admission in 21 and one death (Hoang poison centre series 2022). The label itself records drowsiness, stupor, coma, confusion and death at recommended doses and in overdose, most often after veterinary products (Stromectol label on DailyMed). After a January 2025 podcast, combined ivermectin and benzimidazole prescribing to US cancer patients ran 2.6 times higher than in the same months a year earlier (Rockwell JAMA Network Open 2026).
The bottom line for a patient who has read about this drug online: it is being studied in two early trials, no trial has shown that it treats any cancer, and dosing outside a trial has caused documented harm (Stromectol label on DailyMed). The one trial open to patients is the Cedars-Sinai study in metastatic triple-negative breast cancer (ClinicalTrials.gov NCT05318469). The FDA's guidance on products promoted as cancer cures without trial evidence applies here (FDA on products claiming to cure cancer).
- A 2016 personal story, shared worldwide
- Microtubule disruption in cell lines, as for many compounds that never become drugs
- Dish concentration versus human antiparasitic dosing
- Real repurposing runs through registered trials: aspirin, metformin, statins, mebendazole
In parasites ivermectin opens glutamate-gated chloride channels and paralyses the worm. In cancer cell lines and mice it has been reported to act on several things at once, including the ATP-gated P2X4 and P2X7 receptors, WNT-TCF and Akt/mTOR signalling and PAK1, and the trials pair it with an anti-PD-1 antibody on the mouse finding that the pair drew T cells into tumours where neither drug alone worked. All of this is laboratory evidence; none of it has been shown in a patient.
- A 2016 personal story, shared worldwide
- Microtubule disruption in cell lines, as for many compounds that never become drugs
- Dish concentration versus human antiparasitic dosing
- Real repurposing runs through registered trials: aspirin, metformin, statins, mebendazole
1.Laboratory: in breast cancer cell lines, ivermectin triggers a form of cell death that releases ATP and other signals which attract immune cells (Draganov mouse study 2021).
- Route
- Oral tablet, 3 mg (Stromectol)
- Schedule
- Label: a single dose of about 200 micrograms per kg for strongyloidiasis and about 150 micrograms per kg for onchocerciasis, taken on an empty stomach with water; no cancer dose exists. Trials: the Cedars-Sinai study gives an assigned dose on days 1 to 3 of each week of a 21-day cycle with the antibody, and ICONIC compares 200 and 400 micrograms per kg on days 1 to 3 weekly for four weeks.
- Monitoring
- The label's warnings and overdosage sections describe drowsiness, stupor, coma, confusion, seizures and death with recommended doses and with overdose, most often after veterinary formulations; poisoning is managed with supportive care.
Source: US prescribing information (DailyMed). Doses are for orientation; the current label governs.
| Adverse event | Any grade | Grade 3+ |
|---|---|---|
| Pruritus (itching) Onchocerciasis trials, 963 adults, 100 to 200 micrograms per kg; Mazzotti reactions to dying parasites, not expected outside onchocerciasis | 27.5% | - |
| Skin oedema, papular, pustular or urticarial rash Onchocerciasis trials, 963 adults, 100 to 200 micrograms per kg; Mazzotti reactions to dying parasites, not expected outside onchocerciasis | 22.7% | - |
| Fever Onchocerciasis trials, 963 adults, 100 to 200 micrograms per kg; Mazzotti reactions to dying parasites, not expected outside onchocerciasis | 22.6% | - |
| Headache (regardless of cause) Onchocerciasis trials, 963 adults; the label judged 0.2 percent drug-related | 22.3% | - |
| Myalgia (regardless of cause) Onchocerciasis trials, 963 adults; the label judged 0.4 percent drug-related | 19.7% | - |
| Arthralgia or synovitis Onchocerciasis trials, 963 adults, 100 to 200 micrograms per kg; Mazzotti reactions to dying parasites, not expected outside onchocerciasis | 9.3% | - |
| Tachycardia (arrhythmia: fast heart rate) Onchocerciasis trials, 963 adults; judged possibly, probably or definitely drug-related | 3.5% | - |
| Peripheral oedema Onchocerciasis trials, 963 adults; judged possibly, probably or definitely drug-related | 3.2% | - |
| Facial oedema Onchocerciasis trials, 963 adults; judged possibly, probably or definitely drug-related | 1.2% | - |
| Orthostatic hypotension Onchocerciasis trials, 963 adults; judged possibly, probably or definitely drug-related | 1.1% | - |
| Decrease in white cell count Strongyloidiasis trials, 109 patients, one or two doses of 170 to 200 micrograms per kg; laboratory finding regardless of drug relationship | 3% | - |
| Dizziness (neurological, per label) Strongyloidiasis trials, 109 patients, one or two doses of 170 to 200 micrograms per kg; judged drug-related | 2.8% | - |
| ALT or AST elevation Strongyloidiasis trials, 109 patients, one or two doses of 170 to 200 micrograms per kg; laboratory finding regardless of drug relationship | 2% | - |
| Diarrhoea Strongyloidiasis trials, 109 patients, one or two doses of 170 to 200 micrograms per kg; judged drug-related | 1.8% | - |
| Nausea Strongyloidiasis trials, 109 patients, one or two doses of 170 to 200 micrograms per kg; judged drug-related | 1.8% | - |
| Somnolence, vertigo or tremor (neurological, per label) Strongyloidiasis trials, 109 patients, one or two doses of 170 to 200 micrograms per kg; 0.9 percent each, judged drug-related | 0.9% | - |
| Neurotoxicity: altered consciousness, stupor, coma, confusion, seizures, death Warnings and overdosage sections; reported at recommended doses and in overdose, most often after veterinary formulations; no rate is given | - | - |
Rates read from the US prescribing information. Blank cells mean the figure was not sourced, not that it is zero.
Trials
topTrials recruiting now
Country and place are remembered in this browser only. A postcode is sent to OpenStreetMap's Nominatim service to find coordinates when you press the button; nothing else leaves your device.
Landmark trials
Notes
top- Being studied: two early trials, one recruiting in metastatic triple-negative breast cancer (ClinicalTrials.gov NCT05318469) and one not yet open, ICONIC, which measures an immune-cell change rather than tumour shrinkage (ClinicalTrials.gov NCT07487805).
- Not shown to treat any cancer: neither trial has reported a result and no phase 3 exists (ClinicalTrials.gov NCT05318469). The only published human data are a survey (Loja survey 2023) and a self-reported cohort under an expression of concern (Anticancer Research expression of concern 2026).
- Harm outside trials: seizures and mechanical ventilation after high-dose self-use (Saperstein case report 2026), severe liver injury with veterinary fenbendazole (Powderly case report 2026), a poison-centre series with one death (Hoang poison centre series 2022), and the label's own warning of coma and death with overdose (Stromectol label on DailyMed).
- Tell your oncologist if you are taking it: the toxicity review names P-glycoprotein, the pump that keeps ivermectin out of the brain, as the central factor in neurotoxicity, so anything that blocks it raises the risk (Yilmaz toxicity review 2026). A paediatric case of severe neurotoxicity in a patient on regorafenib was attributed to a CYP3A4 interaction (ClinicalTrials.gov NCT07487805).
A case series selected by its authors from a decade of practice cannot show that any drug worked, and the selection method has no scientific basis.
Veterinary paste is dosed for animals by weight, and 'one squirt' is not a human dose. Liver injury of this severity can be fatal and this patient was fortunate to present in time.
Explains why a patient on cancer drugs that inhibit P-glycoprotein or CYP3A4 is at higher risk from ivermectin than a healthy volunteer, and why veterinary doses are dangerous.
Measures the reach of the claim rather than the drug: tens of thousands of prescriptions written on the strength of a podcast, with no trial behind them.
The clearest documented case of what the label's neurotoxicity warning looks like in a cancer patient dosing herself.
Shows why single stories cannot settle the question: a melanoma this immunogenic can wax and wane on its own, and the improvement did not last.
This is the paper most often quoted online as proof. Patients reporting how they feel to the company that sold them the treatment, with 38 percent not answering, cannot show whether a cancer shrank, and the journal is now checking whether the underlying records exist.
For a patient with a gynaecological cancer, the specialists who would run such a trial say the evidence does not exist yet.
The clearest recent statement of where the evidence stands: promising in a dish, untested in people, and already causing harm through self-dosing.
A measured position a patient can take to their oncologist: the drug is being tested properly, and until those tests report, it is not a treatment.
A reminder that ivermectin interacts with cancer drugs: the dose that a healthy adult tolerates can poison a patient whose liver enzymes are already occupied by a kinase inhibitor.
Population-level evidence that when a drug is promoted online for an unproven use, poison centres see the result.
Shows how widespread use already is where cancer care is expensive, and why the ICONIC investigators cite it as a reason to run a proper trial.
The pattern of harm that cancer self-dosing now repeats: mostly older men, veterinary products, and neurological effects that need hospital care.
An anecdote from the clinic that prescribed the regimen; it cannot show benefit and has not led to a trial.
This is the scientific basis for pairing ivermectin with checkpoint antibodies in the Cedars-Sinai and ICONIC trials. Its key figure is now in doubt, which makes those trials more important, not less, as the only way to find out whether the effect is real.
Useful as a map of the laboratory story; it contains no patient outcomes.
This is the paper the online claims usually trace back to. It is a call for trials, not evidence that the drug works in people, and the trials it called for are still running.
Latest papers
topQuery for this drug: (TITLE:"Ivermectin" OR ABSTRACT:"Ivermectin" OR TITLE:"Stromectol" OR ABSTRACT:"Stromectol" OR TITLE:"IVM" OR ABSTRACT:"IVM" OR TITLE:"MK-0933" OR ABSTRACT:"MK-0933") AND (cancer OR tumor OR tumour OR oncology OR carcinoma OR lymphoma OR leukemia OR leukaemia OR myeloma OR sarcoma OR melanoma OR glioma). Results are unfiltered search hits about Ivermectin, not a curated reading list.
Similar pages
not linked directly; found by shared links- TermModel organism
Shares Cell line, In vitro and in vivo, Preclinical.
- Key paperReverse swing-M, phase 1 study of repurposing mebendazole in recurrent high-grade glioma
Shares Recommended phase 2 dose (RP2D), Mebendazole, Systematic drug repurposing.
- Key paperSevere drug-induced liver injury due to self-administration of the veterinary anthelmintic medication fenbendazole
Shares Fenbendazole (veterinary anthelmintic), Hepatotoxicity (liver enzyme elevation), Misinformation and unproven therapies.
- TermOrganoid
Shares Cell line, In vitro and in vivo, Preclinical.
- Key paperRETRACTED: Fenbendazole as an anticancer agent? A case series of self-administration in three patients
Shares Fenbendazole (veterinary anthelmintic), Fenbendazole, ivermectin and other internet 'repurposed cures', Misinformation and unproven therapies.
- Key paperA phase 2a clinical study on the safety and efficacy of individualized dosed mebendazole in patients with advanced gastrointestinal cancer
Shares Mebendazole, Systematic drug repurposing, Single-arm trial.
- TrialVorbipiprant (CR6086) / Balstilimab (AGEN2034) Combination in Stage IV Refractory pMMR - MSS CRC, and Other Metastatic GI Cancers
Shares Balstilimab, Agenus.