The 2026 Ivermectin and Mebendazole Observational Study

Table of Contents

1. What the Presentation Said
2. Finding the Paper Behind the Figures
3. How the Study Was Done
4. What the Paper Reported
5. The Presentation's Figures Beside the Paper's
6. What “Stable Disease” and “No Evidence of Disease” Mean
7. What an Uncontrolled, Self-Selected Cohort Can and Cannot Show
8. Expression of Concern and Retraction
9. Other Human Evidence on Ivermectin and Mebendazole in Cancer
10. How the Claim Compares With the Research
11. What Is Not Known
12. Connections
13. Key Research Papers
14. Featured Videos

What the Presentation Said

In the part of the presentation about antiparasitic medicines, the speaker said that the clinic's favourite antiparasitic drugs were mebendazole and ivermectin, and then described a recent study of the two drugs together in people with cancer. As the speaker presented it:

• the study was linked to a school of public health in Connecticut (later in the talk, to Yale public health, together with a physician the speaker named);
• it was a prospective observational study in which ivermectin and mebendazole were prescribed for six months without altering the patients' current treatments or procedures;
• at six months, 48.4% of patients had tumour regression or no evidence of disease, 32.8% had no evidence of disease, and overall about 84% had no evidence of disease, regression or stable disease;
• nearly 87% of patients completed treatment, with only mild side effects.

The speaker said the clinic had used this combination for cancer for some time and that its own patients did not see even mild side effects, which the speaker attributed to the clinic's detoxification and nutrition programme. The speaker also said they did not know exactly what other treatments the study's patients were receiving. The machine transcript of the talk is garbled in this section: it also gives figures of “seven cancer patients”, “50 milligrams”, “0.4%” of patients seeing clinical benefit and “0.6%” progression, which are compared with the paper below.

The presentation did not say whether the paper had been peer reviewed, who funded it, or whether the journal had raised concerns about it. The transcript gives no date for the talk, so this page does not say whether it was given before or after the journal's notices.

Finding the Paper Behind the Figures

A PubMed and preprint search finds one study that matches the design and every legible percentage in the presentation: Real-world Clinical Outcomes of Ivermectin and Mebendazole in Cancer Patients: Results from a Prospective Observational Cohort. The 48.4%, 32.8%, 84.4% and 86.9% figures all appear in it.

Where and when it appeared

• Preprint: first posted on the open repository Zenodo on 7 April 2026 as a preprint (a paper shared before peer review).
• Journal: received by the journal Anticancer Research on 8 April 2026, revised 20 April, accepted 24 April, and printed in the June 2026 issue (volume 46, issue 6, pages 3243–3255). It was indexed in PubMed and MEDLINE, tagged as an observational study (Hulscher and colleagues, 2026).
• Notices: the journal issued an Expression of Concern on 9 June 2026, an update to it in September 2026, and a retraction dated October 2026 (published online 1 October 2026). These are described in their own section below.

The data were collected between August 2025 and March 2026, so the presentation's description of a “2025” study fits the start of data collection, but the paper itself is a 2026 publication.

Authors, affiliations, funding and conflicts

The paper lists twelve authors. Every author lists an affiliation with a U.S. telemedicine company; two also list a private foundation in Texas, and one lists the Department of Epidemiology and Public Health at the Yale School of Public Health in New Haven, Connecticut. That last affiliation is the likely source of the presentation's “school of public health in Connecticut”.

The paper's own statements, as printed in the preprint, are:

• Conflict of interest: all authors are affiliated with and/or receive salary support from the company that operated the telemedicine platform through which the combination was prescribed and dispensed; the statement adds that the company also offers compounded ivermectin and mebendazole formulations as part of its clinical services.
• Funding: “No external funding was received for this project.”
• Ethics review: the project was described as a retrospective analysis of a clinical programme evaluation that “did not meet the definition of human subjects research and did not require Institutional Review Board (IRB) review.”
• Data: the underlying data are “on file” with the company; aggregated summary data are in the article.

The full text of the journal version could not be opened for this page (the journal's site refused automated access), so the details of method above and below come from the preprint, whose abstract figures match the PubMed abstract of the journal version exactly.

How the Study Was Done

Who took part

Adults with cancer who had been prescribed the combination off-label (for a use not approved by regulators) by licensed U.S. providers on the company's telemedicine platform, after an online intake form and a consultation. Eligible patients were invited by email to answer two voluntary online surveys. 197 people answered the first survey, and 122 (61.9%) answered the second.

What they took

Compounded (pharmacy-made) capsules each containing 25 mg of ivermectin and 250 mg of mebendazole. The first prescription was typically 90 capsules. The dose and schedule were set individually by the prescriber and the patient: among the 122 who answered the follow-up, 44.3% took on average one capsule a day, 27.9% two, 11.5% three and 12.3% four, and many followed on-and-off cycles and reordered as they wished.

How outcomes were measured

• The first survey ran from August to September 2025 and the follow-up from January to March 2026, about six months apart (range four to seven months).
• Every answer was self-reported. At follow-up, participants chose whether their cancer showed “no current evidence of disease”, had “regressed”, had “stayed about the same” (counted as stable disease) or had “spread or progressed”.
• The paper's main measure, the “Clinical Benefit Ratio”, was defined as the share of respondents choosing no evidence of disease, regression or stable disease.
• The paper's limitations paragraph states that outcomes “were not clinically adjudicated or radiographically confirmed” (no doctor or scan checked them), that there was no control group, and that confounding by other treatments “cannot be excluded”.

Who was in the group

• Mean age 67; 52.3% men and 47.7% women; mostly White, with over half holding a college or graduate degree.
• Cancer types: prostate 27.9%, breast 18.3%, lung 8.6%, colon 5.1%, liver 2.5%, and 37.6% other or unspecified types.
• Median time since diagnosis 1.2 years; 47.7% were within their first year after diagnosis.
• At the first survey, 37.1% said their cancer was actively spreading or progressing and 62.9% said it was not currently spreading.
• Prior or current treatments reported at the first survey: surgery 42.1%, chemotherapy 31.5%, radiation 28.9%, immunotherapy 17.3%, hormone therapy 14.7%, targeted therapy 8.6%.

What the Paper Reported

Among the 122 people who answered the follow-up survey, the paper (now retracted) reported these self-reported results (Hulscher and colleagues, 2026):

• No current evidence of disease: 40 people (32.8%).
• Regression: 19 people (15.6%).
• Stayed about the same (stable disease): 44 people (36.1%).
• Spread or progressed: 19 people (15.6%).
• No evidence of disease or regression combined: 59 people (48.4%).
• “Clinical Benefit Ratio” (all three non-progression answers): 103 people (84.4%, 95% confidence interval 77.0–89.8%).

Use, side effects and dose

• 86.9% said they had finished the first 90-capsule prescription; 66.4% were still taking the combination at follow-up.
• 25.4% (31 people) reported side effects, most often digestive symptoms, fatigue or weakness, dizziness, skin reactions and neurological symptoms. Of those 31, 14 reduced or paused the medicine and 2 stopped it.
• In the dose comparison, the cancer answers did not differ by dose (p = 0.91), while side effects did (p = 0.0014); the highest side-effect share, 47.1%, was in the two-capsule group.

Other treatments at the same time

At follow-up, respondents reported cancer-related supplements (49.2%), dietary changes (37.7%), chemotherapy (27.9%), radiation (21.3%) and surgery (19.7%), and free-text answers mentioned fasting, ketogenic diets, hyperbaric oxygen, red-light therapy and supplements such as vitamin D, turmeric and berberine.

The authors' own conclusion called the findings “hypothesis-generating” and wrote that randomized controlled trials were needed. The preprint's limitations paragraph adds: “Given these threats to validity, therapeutic benefit cannot be inferred.”

The Presentation's Figures Beside the Paper's

Read side by side, several of the presentation's figures match the paper and several do not:

• 48.4% regression or no evidence of disease, 32.8% no evidence of disease, about 84% overall: these match the paper (48.4%, 32.8%, 84.4%).
• “Seven cancer patients”: the paper reports 197 people at the first survey and 122 at follow-up; every percentage above is a share of the 122.
• “0.4%” saw clinical benefit: no such figure is in the paper; its clinical-benefit figure is 84.4%.
• “0.6%” disease progression: the paper reports 15.6% progression.
• “50 milligrams”: the paper's capsule held 25 mg of ivermectin and 250 mg of mebendazole, taken one to four times a day.
• “Nearly 87% completed treatment for six months”: the paper's 86.9% is the share who finished the first 90-capsule prescription; 66.4% were still taking it at about six months.
• “Tumour shrinkage or no evidence of disease”: in the paper these were survey answers, not measured shrinkage on scans.
• “Without altering current treatments”: the paper reports that many respondents were also having chemotherapy, radiation or surgery, and about half were taking other cancer-related supplements.
• “Only mild side effects”: the paper describes side effects as predominantly mild, reported by 25.4%, with 2 of 31 stopping the medicine.
• Yale: one of twelve authors listed a Yale School of Public Health affiliation; all twelve listed the telemedicine company that prescribed and sold the capsules.

What “Stable Disease” and “No Evidence of Disease” Mean

Cancer studies normally judge whether a solid tumour has responded with a standard rulebook called RECIST (Response Evaluation Criteria in Solid Tumours). Its current version, RECIST 1.1, was published in 2009 by an international working group, drawing on a data warehouse of more than 6,500 patients (Eisenhauer and colleagues, 2009). Under RECIST, a radiologist measures up to five “target” tumours on scans (no more than two per organ) and adds up their diameters, then compares the sum over time.

The four standard categories

• Complete response: the US National Cancer Institute defines it as “the disappearance of all signs of cancer in response to treatment”, adding that it “does not always mean the cancer has been cured.”
• Partial response: “a decrease in the size of a tumor, or in the extent of cancer in the body, in response to treatment” (National Cancer Institute). RECIST sets a minimum amount of measured shrinkage before a decrease counts.
• Stable disease: “cancer that is neither decreasing nor increasing in extent or severity” (National Cancer Institute) — in RECIST terms, change too small to count as either response or progression.
• Progressive disease: “cancer that is growing, spreading, or getting worse” (National Cancer Institute). RECIST 1.1 requires at least a 20% increase in the summed diameters plus at least 5 mm in absolute size, or new tumours (Eisenhauer and colleagues, 2009).

“No evidence of disease” and “clinical benefit”

“No evidence of disease” is not a RECIST category. Doctors use the phrase when examinations, tests and scans find no sign of cancer, often after surgery or other treatment has removed the visible tumour; like complete response, it does not mean the cancer cannot return. In the 2026 study it was one of four answers a participant could pick on an online survey.

A “clinical benefit rate” or “disease control rate” in a trial usually adds complete response, partial response and stable disease together. Because stable disease is included, the figure can be high even when few tumours shrink — in the 2026 study, 36.1 of the 84.4 percentage points came from people answering that their cancer had “stayed about the same”.

Even scans need a rulebook

Response categories depend on the method used. In a Chinese study of 35 people with advanced lung cancer having chemotherapy, RECIST 1.1 and two scan criteria based on tumour metabolism (positron emission tomography (PET) scans) gave different response categories for 57% to 63% of patients (Shang and colleagues, 2016). If trained readers applying written rules to the same scans can disagree that often, a patient's own one-word answer on a survey is a much looser measure.

What an Uncontrolled, Self-Selected Cohort Can and Cannot Show

A cohort study follows a group of people over time. This one was uncontrolled (no comparison group that did not take the drugs) and self-selected (people chose to buy the medicine, chose to answer the surveys, and chose whether to answer again). Such a design can show:

• who used a treatment, how they took it, and how many kept taking it;
• which side effects users reported;
• whether results look interesting enough to justify a proper trial — which is how the authors described their findings.

It cannot show that the drugs caused the outcomes, for reasons that are general to this design:

• No comparison group. Without one, there is no way to know how many of the same people would have reported stable disease or no evidence of disease anyway. RECIST itself notes that a randomised control arm is what makes response data interpretable; it drops the requirement to confirm responses in randomised trials “since the control arm serves as appropriate means of interpretation of data” (Eisenhauer and colleagues, 2009).
• Other treatments. Many respondents were having chemotherapy, radiation or surgery at the same time. A tumour removed by surgery, or shrunk by chemotherapy, would show up as “no evidence of disease” or “regression” in the survey whatever the antiparasitic drugs did.
• Starting point. At the first survey, 62.9% already said their cancer was not spreading, and nearly half were within a year of diagnosis — the period when surgery and first treatments usually happen. The paper does not report how many already had no evidence of disease when they started.
• Missing people. 75 of the 197 (38.1%) did not answer the follow-up. People who are doing badly, or who have died, are less likely to answer a survey; the paper compared the two groups only on chemotherapy and radiation use.
• Self-report. No scans, pathology reports or medical records were checked against the answers, and the same survey could not distinguish slow-growing cancers, such as many prostate cancers, from fast-growing ones.
• Who asked. The surveys were sent by the company that prescribed and sold the capsules, which the paper states in its conflict-of-interest declaration.

The paper's own limitations paragraph names the main points: no control group, no clinical or radiographic confirmation, and possible confounding from concurrent therapies, supplements and lifestyle changes.

A published response

On 15 April 2026, a week after the preprint appeared, the fact-checking outlet Lead Stories published an article on it reporting that researchers it consulted pointed to the lack of a control group, the reliance on self-reported outcomes with no independent verification, and the loss of more than a third of participants before follow-up.

Expression of Concern and Retraction

An Expression of Concern is a formal notice in which a journal warns readers that there are questions about a paper while it investigates. A retraction withdraws the paper from the scientific record; the article usually stays online, marked as retracted. The journal's notices on this paper, all indexed in PubMed, are:

9 June 2026: Expression of Concern

The editorial board wrote that “serious scientific concerns were raised by the international medical community regarding the verifiability and statistical reliability, and ethical oversight of the underlying dataset”. It recorded that the authors had told the journal the cohort comprised “provider-screened cancer patients enrolled in a structured prospective cohort”, stated that the journal “does not endorse or condone the unvalidated off-label use of medications for unapproved oncological indications”, and announced an audit of ethics-board documentation and source records for the 197 participants' diagnoses and reported regressions (Anticancer Research, Expression of Concern, 2026).

September 2026: Update to the Expression of Concern

The board reported that the audit was concluded and its requests for ethics-board and source documents were withdrawn, that it drew no adverse inference from their non-production, and that “No allegation of fabrication, falsification, plagiarism, image manipulation, or other research misconduct is made against any author.” It also stated that it did not allege any count, percentage or test statistic was miscalculated. The continuing review instead concerned the study's design, analysis and reporting: the construction of the “Clinical Benefit Ratio” and the self-reported survey measure behind it; the timing of the baseline survey relative to when people started the drugs; the basis for the dose–side-effect relationship; the analysis offered for how representative the follow-up group was; the dose-group analysis; and how participants without follow-up, and concurrent conventional cancer therapy, were handled in attributing the outcomes. The authors had appealed, and the board asked readers to continue to exercise caution in interpreting the reported outcomes (Anticancer Research, Update to the Expression of Concern, 2026).

October 2026: Retraction

The retraction notice (volume 46, issue 10, page 5877) begins, as indexed in PubMed: “Following post-publication review, independent expert assessments, and consideration of the authors’ appeal, responses, and proposed correction, the Editor of Anticancer Research retracts the above article because methodological and analytical problems” … The full notice could not be opened for this page, so its remaining wording is not reported here. PubMed now lists the article with “RETRACTED:” before its title and the publication type “Retracted Publication” (Anticancer Research, Retraction, 2026).

Other Human Evidence on Ivermectin and Mebendazole in Cancer

With the 2026 cohort retracted, the human evidence on these two drugs in cancer rests on a small number of studies. The sibling pages Ivermectin in Cancer Research and Benzimidazoles in Cancer cover the laboratory and animal work in detail.

Mebendazole: one small randomised trial

In Egypt, 40 people with metastatic colorectal cancer were randomised, double-blind, to standard chemotherapy (bevacizumab with FOLFOX4) plus either placebo or mebendazole 500 mg twice daily for 12 weeks. On computed tomography (CT) scans at 12 weeks, the overall response rate was 65% (13 of 20) with mebendazole and 10% (2 of 20) with placebo, and median progression-free survival was 9.25 versus 3 months (Hegazy and colleagues, 2022). The trial was small, from one centre, and was registered retrospectively (after it began); no larger trial confirming it was found in this search. In this trial mebendazole was added to chemotherapy, not used in place of it.

Mebendazole and albendazole: reviews

A 2021 review in a parasitology journal found anticancer activity of mebendazole and albendazole in cell culture and in animals across liver, lung, ovary, prostate, colorectal, breast and head-and-neck cancers and melanoma, but only two clinical reports for albendazole and two case reports for mebendazole in people. It also notes that these drugs, though generally safe, can cause liver toxicity when used for prolonged periods (more than 14–28 days) or occasionally even after a single dose, and that high-dose prolonged albendazole can cause neutropenia (a fall in infection-fighting white cells) through bone-marrow suppression (Chai and colleagues, 2021).

Ivermectin: no controlled trial found in this search

No published randomised trial of ivermectin as a cancer treatment was found in this search. A mixed-methods study (observation, surveys and interviews) in rural Loja, Ecuador, found that 19% of participants diagnosed with cancer used ivermectin-based medicines alongside conventional treatment; participants said they felt better, while the specialists interviewed said there was no scientific knowledge supporting its use in human cancer and no authorisation to prescribe it (Jiménez-Gaona and colleagues, 2023).

Case reports with many treatments at once

• A 2026 case report describes one woman with stage IV hormone-receptor-positive breast cancer who had a near-complete response after chemotherapy given with fasting and induced low blood sugar, a ketogenic diet, hyperthermia, hyperbaric oxygen, and six repurposed drugs including mebendazole and ivermectin (Slocum and colleagues, 2026). With so many treatments together, including chemotherapy, a single report cannot separate their effects.
• A 2026 case series describes six people with stage IV cancers treated between 2011 and 2022 with combinations of ivermectin, mebendazole, praziquantel, niclosamide and antifungals chosen by an electro-acupuncture reading, all of whom were reported to have survived longer than expected (Guilford and Yu, 2026). It had no comparison group.

How the Claim Compares With the Research

What the record supports

• A prospective observational study of ivermectin and mebendazole in people with cancer was published in 2026, and it did report 48.4% regression or no evidence of disease, 32.8% no evidence of disease and an 84.4% “Clinical Benefit Ratio” at about six months.
• One of its authors listed a Yale School of Public Health affiliation.
• The paper describes side effects as mostly mild, in 25.4% of respondents.

What it supports only in part

• The percentages are the paper's, but they are answers to an online survey from 122 of 197 people, not measured tumour responses; the presentation described them as tumour shrinkage and disease status.
• The paper itself and its authors described the findings as hypothesis-generating and stated that therapeutic benefit cannot be inferred from them.

What the record contradicts

• The paper reports 197 enrolled people, not seven; a 25 mg ivermectin plus 250 mg mebendazole capsule, not 50 milligrams; and 15.6% progression, not 0.6%.
• The paper reports that many participants were also having chemotherapy, radiation, surgery and other supplements, so the outcomes cannot be attributed to the two drugs alone.
• The journal issued an Expression of Concern in June 2026 and retracted the paper in October 2026, citing methodological and analytical problems. A retracted paper is no longer part of the accepted scientific record.

What has no research at all

• No study was found testing the speaker's statement that patients on a detoxification and nutrition programme have no side effects from these drugs.
• No randomised trial of the ivermectin and mebendazole combination in any cancer was found.

Decisions about cancer treatment and antiparasitic medicines belong with a clinician.

What Is Not Known

• Whether ivermectin, mebendazole or the two together change cancer outcomes in people: no randomised trial of the combination has been published.
• Whether the single small mebendazole trial in colorectal cancer would hold up in a larger, multi-centre trial.
• What the 75 people who did not answer the follow-up survey experienced, and how many of the 122 who did had already had their tumours removed or treated before starting the capsules.
• How compounded capsules of these drugs compare with approved tablets in content and absorption; the paper does not report testing them.
• The long-term safety of taking these drugs daily or in cycles for months alongside chemotherapy, given the liver and blood-count effects reported with prolonged benzimidazole use.
• The full reasoning of the October 2026 retraction notice beyond its opening words, which this page could not open.

Connections

• Cancer and Parasites
• Ivermectin in Cancer Research
• Benzimidazoles: Mebendazole and Fenbendazole in Cancer
• Are Parasites Inside Every Tumor?
• Ivermectin
• Fenbendazole
• Fenbendazole: Off-Label Cancer Use
• Albendazole and Mebendazole for Roundworm
• Strongyloides
• Parasites
• Oncology
• Prostate Cancer
• Breast Cancer
• Colorectal Cancer

Key Research Papers

1. Hulscher N, Victory K, Thorp JA, Pinsky D, Diaz-Villalobos A, Gillooly P, Coulson F, Annazone M, Radesi C, Brooks J, McCullough PA, Risch H. RETRACTED: Real-world Clinical Outcomes of Ivermectin and Mebendazole in Cancer Patients: Results from a Prospective Observational Cohort. Anticancer Research. 2026;46(6):3243-3255. Retracted October 2026 (PubMed record 42203321, tagged Retracted Publication); listed here only as the retracted record the presentation described. doi:10.21873/anticanres.18194
2. Retraction. Anticancer Research. 2026;46(10):5877. PubMed PMID: 42823188
3. Expression of Concern: Real-world Clinical Outcomes of Ivermectin and Mebendazole in Cancer Patients: Results from a Prospective Observational Cohort. Anticancer Research. 2026;46(6):e3243. PubMed PMID: 42300708
4. Update to the Expression of Concern: Real-world Clinical Outcomes of Ivermectin and Mebendazole in Cancer Patients: Results from a Prospective Observational Cohort. Anticancer Research. 2026;46(9):e3243u. PubMed PMID: 42703869
5. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, Dancey J, Arbuck S, Gwyther S, Mooney M, Rubinstein L, Shankar L, Dodd L, Kaplan R, Lacombe D, Verweij J. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45(2):228-47. PubMed PMID: 19097774
6. Shang J, Ling X, Zhang L, Tang Y, Xiao Z, Cheng Y, Guo B, Gong J, Huang L, Xu H. Comparison of RECIST, EORTC criteria and PERCIST for evaluation of early response to chemotherapy in patients with non-small-cell lung cancer. Eur J Nucl Med Mol Imaging. 2016;43(11):1945-53. PubMed PMID: 27236466
7. Hegazy SK, El-Azab GA, Zakaria F, Mostafa MF, El-Ghoneimy RA. Mebendazole; from an anti-parasitic drug to a promising candidate for drug repurposing in colorectal cancer. Life Sci. 2022;299:120536. PubMed PMID: 35385794
8. Chai JY, Jung BK, Hong SJ. Albendazole and Mebendazole as Anti-Parasitic and Anti-Cancer Agents: an Update. Korean J Parasitol. 2021;59(3):189-225. PubMed PMID: 34218593
9. Jiménez-Gaona Y, Vivanco-Galván O, Morales-Larreategui G, Cabrera-Bejarano A, Lakshminarayanan V. Outcome of Ivermectin in Cancer Treatment: An Experience in Loja-Ecuador. Nurs Rep. 2023;13(1):315-326. PubMed PMID: 36976682
10. Slocum AK, Tastekin D, Duraj T, Seyfried TN. Management of advanced HR-positive breast cancer using metabolically supported chemotherapy and repurposed drugs: a case report. Front Oncol. 2026;16:1795402. PubMed PMID: 42100430
11. Guilford FT, Yu S. Cancer Therapy Using Antiparasitic Medications Guided by Acupuncture Meridian Assessment: A Case Series of Six Patients. Integr Med (Encinitas). 2026;25(3):27-33. PubMed PMID: 42491797

PubMed Topic Searches

1. Ivermectin and mebendazole in cancer
2. Mebendazole cancer randomised trials
3. Ivermectin cancer clinical trials
4. RECIST response criteria

Other Sources

1. Preprint version of the study, posted 7 April 2026, with its methods, conflict-of-interest, funding and ethics statements. Zenodo record 19455636. zenodo.org
2. National Cancer Institute, Dictionary of Cancer Terms: stable disease. cancer.gov
3. National Cancer Institute, Dictionary of Cancer Terms: complete response. cancer.gov
4. National Cancer Institute, Dictionary of Cancer Terms: partial response. cancer.gov
5. National Cancer Institute, Dictionary of Cancer Terms: progressive disease. cancer.gov
6. Lead Stories, fact check of the ivermectin and mebendazole preprint, 15 April 2026. leadstories.com

Featured Videos