Ivermectin and mebendazole in cancer: what does 84.4% mean?

Ivermectin and mebendazole in cancer became the focus of intense debate after a paper reported an 84.4% “clinical benefit” rate. That number sounds impressive. The important question, however, is not simply what percentage was reported, but how the outcome was measured.

The Anticancer Research paper followed 197 people with a range of cancers who were prescribed ivermectin plus mebendazole off-label through a U.S. telemedicine platform. The compounded capsules contained 25 mg of ivermectin and 250 mg of mebendazole. Participants completed standardized digital surveys at baseline and again about six months later.

Only 122 people, or 61.9% of the original cohort, completed the six-month follow-up. The headline figure — an 84.4% clinical benefit rate — was calculated in this responding group.

What the 84.4% figure actually includes

The authors grouped several very different outcomes under “clinical benefit”: no current evidence of disease, tumor regression, and stable disease. Among the 122 respondents, 32.8% reported no current evidence of disease, 15.6% reported regression, and 36.1% reported stable disease. Another 15.6% reported progression.

The key limitation is how those outcomes were obtained. They were patient-reported. The study did not independently adjudicate cancer status through standardized imaging, blinded radiology review, or verified medical records. In other words, “tumor regression” in this dataset is not equivalent to an objective response documented using the usual criteria of an oncology trial.

Stable disease also cannot automatically be attributed to the two drugs. Some cancers progress slowly, and some participants were receiving established cancer treatments at the same time. At six months, 27.9% reported chemotherapy, 21.3% radiation therapy, and 19.7% surgery. With no control group, the study cannot separate any possible effect of ivermectin and mebendazole from the effects of conventional treatment, the natural course of disease, or differences between patients.

Attrition adds another major source of uncertainty. Seventy-five of the 197 enrolled patients did not provide six-month follow-up data — 38.1% of the starting cohort. If people whose disease worsened were less likely to remain in the program or answer the survey, the outcomes among responders could look substantially better than the outcomes of the full group. The design cannot reliably rule out that form of selection bias.

Why these drugs attracted interest in the first place

Drug repurposing is a legitimate research strategy. A medicine developed for one condition can sometimes turn out to have useful effects in another.

For mebendazole, there is a plausible laboratory mechanism: it interacts with tubulin and disrupts microtubules, structures cells need for division. In preclinical cancer models, that can interfere with tumor-cell proliferation. Ivermectin has also been reported to affect several cancer-related signaling pathways in cell and animal models.

But biological plausibility is only the beginning of the evidence chain. A concentration that inhibits cancer cells in a dish may not be achievable or safe in a person, and an effect in an animal model does not guarantee a meaningful clinical benefit in patients. Controlled clinical trials are what bridge that gap.

In this cohort, 25.4% of respondents reported adverse effects, described mainly as gastrointestinal, and most of those participants continued after dose adjustments. That provides some tolerability information for this particular group, but it is not a substitute for systematic safety assessment in a controlled trial.

Why the journal added a warning

The paper also has an important conflict-of-interest context. All authors disclosed affiliation with and/or salary support from The Wellness Company. The same company operated the telemedicine platform through which the ivermectin–mebendazole combination was prescribed and dispensed. Disclosure does not invalidate a study, but it increases the importance of independent verification.

In June 2026, Anticancer Research issued an official Expression of Concern linked to the paper. An Expression of Concern is not a retraction; it is a formal warning that unresolved questions require further scrutiny. The journal announced a post-publication audit addressing the verifiability and statistical reliability of the dataset as well as ethical oversight and IRB-related documentation.

The paper itself acknowledges several of its central weaknesses: observational design, reliance on self-reported outcomes, possible selection bias, and uncontrolled confounding. The authors describe the findings as hypothesis-generating and call for randomized controlled trials.

That is the appropriate level of interpretation. The 84.4% figure is not evidence that ivermectin and mebendazole treat cancer. It is a self-reported outcome from part of an uncontrolled, highly heterogeneous cohort. The result may justify a testable hypothesis, but establishing efficacy requires a properly designed clinical trial.