Beyond the Anecdotes: Old Drugs, New Questions in Cancer
A retired veterinary internist looks at what the evidence really says about repurposed medications
Please read this first. I am a retired veterinarian, not an oncologist, and nothing here is a treatment recommendation. None of these drugs is proven to treat cancer in people. If you or someone you love has cancer, these are questions to bring to your oncologist, not substitutes for the treatment they offer.
Why I'm writing this
I spent my career as a veterinary internist, and cancer was one of the hardest parts of the job. I have lost far too many patients to it, and I have lost family members to it too. Once a cancer has spread through the body, our options narrow quickly, and anyone who has sat in that exam room knows how it feels to run out of good answers.
So I keep looking. Lately I have been following a group of old, inexpensive drugs that researchers are studying as possible cancer helpers: mebendazole and fenbendazole (dewormers), hydroxychloroquine (a malaria and arthritis drug), and ivermectin (an antiparasitic). The internet is full of claims about them, many of them far ahead of the evidence.
This article is my attempt to sort it out. I will explain, in plain terms, how each drug is thought to work against cancer, what the human studies actually show, and where the real risks are.
How to read the evidence
Not all evidence is equal, and most online claims rest on the weakest kind. Here is the ladder, from weakest to strongest:
Lab dish studies. Cancer cells in a dish are exposed to the drug. Many compounds kill cancer cells this way, often at doses a person could never safely reach.
Animal studies. Usually mice with implanted tumors. Better, but mice are not people.
Case reports. A doctor or patient describes what happened to one or a few people. Useful for raising questions, poor at answering them.
Early trials (phase 1 and 2). Small groups of patients, testing safety, dose, and early signs of benefit.
Randomized controlled trials (phase 3). Patients are assigned by chance to get the drug or not. This is the only reliable way to know whether a drug helps.
The problem with case reports. In practice, we remember the patient who did remarkably well. We forget the ten who took the same thing and declined on the usual schedule. Websites that collect success stories have no record of the failures, so there is no way to know how often the drug did nothing. Most of these patients were also on other treatments, so we cannot tell which one helped.
The funding problem is real, but it is not the whole story. These drugs are off-patent, so no company profits from proving they work, and large trials cost tens of millions of dollars. That is a genuine barrier. But some of these drugs have been tested by universities and cancer centers for more than a decade. The honest picture is that some have had their chance and shown modest results, while others have barely been studied at all.
The dewormers: mebendazole and fenbendazole
Mebendazole is approved for people. Fenbendazole is its close cousin, used in animals. Mebendazole has real human cancer trials behind it. Fenbendazole has mostly anecdotes, a retracted case series, and a growing list of safety warnings.
How they are thought to work
Microtubules, in plain terms. Inside every cell is a scaffolding of tiny tubes called microtubules. Think of them as the cell's girders and conveyor belts. They hold the cell's shape, move cargo around, and, most importantly, pull the chromosomes apart when a cell divides in two.
These drugs latch onto the building blocks of those tubes and stop them from assembling. In a worm, the cell's transport system fails and the parasite starves. In a cancer cell, which divides constantly, the cell gets stuck halfway through division and dies. Vincristine, an established chemotherapy drug, attacks this same machinery in the same direction. Paclitaxel attacks the same tubes from the other side: it freezes them in place instead of stopping them from forming. Same target, opposite grip.
Lab studies suggest two other effects as well:
Cutting off fuel. Many cancer cells burn sugar at a furious rate. In lab studies, these drugs reduce how much glucose cancer cells can take in and use, by lowering the sugar transporters on the cell surface and a key sugar-burning enzyme. [17]
Waking up a tumor brake. p53 is a protein that tells damaged cells to stop dividing or self-destruct. It is often called the guardian of the genome. Many cancers switch it off or break it down too quickly. In a 2018 study of human lung cancer cells and mice, fenbendazole stabilized p53 so it built up in the cell, and moved it into the mitochondria, the cell's power plants, where it helps trigger self-destruction. The same study tied this p53 activation to the drop in sugar uptake, so the two effects may be linked. [17]
One caution: the effect was strongest in cells with normal, working p53. Roughly half of human cancers carry a mutated p53, so this mechanism may not apply to every tumor. And this was lab and mouse work, not a human study.
What the human studies show
The mebendazole story began by accident. Lab mice given a dewormer stopped growing their implanted brain tumors, and later mouse studies of glioblastoma, an aggressive brain cancer, showed longer survival. Mebendazole also crosses into the brain, which most drugs cannot.
Johns Hopkins, phase 1 (published 2021). Twenty-four patients with newly diagnosed high-grade brain tumors (18 glioblastoma, 6 anaplastic glioma) received mebendazole with adjuvant temozolomide, after they had already finished radiation plus temozolomide. Doses ran up to 200 mg/kg/day. It was reasonably well tolerated. Four patients at the top dose developed grade 3 liver-enzyme elevations, which reversed when the dose was lowered or stopped. This trial measured safety and dose, not whether the drug works. The survival numbers it reports have no comparison group, so they cannot be read as proof of benefit. [1]
India, randomized phase 2 (published 2022). Eighty-eight patients with recurrent glioblastoma were randomized to temozolomide plus mebendazole, or lomustine plus mebendazole. This is a real randomized trial of a repurposed dewormer. It is not, however, a test of mebendazole against no mebendazole. Both arms got the drug. The investigators set a bar in advance: a 9-month overall survival of 55 percent or better in either arm would justify further study. They reached 36.6 percent and 45 percent. Median survival was about 6.5 to 6.7 months. The trial missed its own benchmark. The authors noted that more than a quarter of the patients were sicker than patients in the historical trials they used as a yardstick, and that a fitter subgroup looked better in a look-back analysis. That subgroup was not the question the trial was built to answer. [2]
Other work. Mebendazole is also in a phase 1 study for recurrent children's brain tumors, and it has been looked at in colon cancer. I have not seen results from those that change the picture above.
So far, the human results are modest, and the one randomized trial we have cannot tell us whether mebendazole added anything. That is not a reason to stop studying it. It is a reason not to treat anecdotes as an answer.
The fenbendazole problem
Fenbendazole has no controlled human trials. Its best-known supporting paper, a 2025 case series of three patients with breast, prostate, and melanoma cancers who were reported to go into remission, was retracted on January 21, 2026. The journal found that the lead author had been offering services related to the topic and had not disclosed it. The retraction was for that undisclosed conflict, not because the journal proved the patient stories false. The paper still cannot be cited as reviewed evidence. The patients were also on other treatments, and the cases came from a website that collects success stories. [3]
Fenbendazole is also not as harmless in people as a dog dewormer label makes it sound. Several medical reports describe serious liver injury in people who dosed themselves after reading about it online:
A woman with lung cancer on immunotherapy developed liver injury, which cleared when she stopped fenbendazole (2021). [4]
A 67-year-old woman developed severe, biopsy-confirmed liver injury (2024). [5]
A woman with colon cancer on immunotherapy developed severe liver injury after raising her dose (2026). [6]
A man with prostate cancer developed liver injury after combining veterinary fenbendazole and ivermectin (2026). [7]
In several of these cases, doctors first blamed the cancer treatment, because the patient had not mentioned the fenbendazole. That delay alone is a risk. If you take anything, tell your oncologist.
Hydroxychloroquine: blocking the cell's recycling plant
Hydroxychloroquine has the longest trial record of any drug here. The results make the clearest point of the whole article: a few encouraging signals alongside chemotherapy, a disappointing result in the largest randomized trial, and almost nothing on its own.
How it is thought to work
Autophagy, in plain terms. The word means "self-eating." Every cell has a recycling system. It wraps worn-out parts in a little bag, then delivers the bag to a digestive compartment called the lysosome, which breaks the contents down into raw materials the cell can reuse.
Cancer cells lean on this system hard, especially under attack. When chemotherapy or radiation damages them, they ramp up recycling to patch themselves up and scrape together fuel. It is one way tumors survive treatment.
Hydroxychloroquine blocks the last step. The bags of cellular junk pile up but never reach the digestive compartment. The idea is to take away the tumor's repair shop just when the chemotherapy is doing its damage.
What the human studies show
Pancreatic cancer, on its own (Dana-Farber, 2014). Twenty patients with advanced disease took hydroxychloroquine alone. Only 2 had no cancer growth at two months. Essentially a negative result. [8]
Pancreatic cancer, with chemotherapy before surgery (randomized phase 2, published 2020). In patients with potentially removable tumors, adding high-dose hydroxychloroquine to gemcitabine and nab-paclitaxel improved the tumor's response under the microscope and in a blood marker. Overall survival and relapse-free survival did not differ between the arms. A better-looking tumor at surgery is not the same thing as a longer life. [9]
Advanced pancreatic cancer, with chemotherapy (randomized phase 2, 2019). In 112 patients, adding hydroxychloroquine did not improve survival at one year: 41 percent with the drug, 49 percent without. Median survival was 11.1 months versus 12.1 months. The share of tumors that shrank did rise, from about 21 percent to about 38 percent. Response without a survival gain is a signal worth understanding. It is not a reason to add the drug outside a trial. [10]
Ovarian cancer (randomized phase 2, 2025). Fifty-nine patients with platinum-sensitive relapsed disease received chemotherapy with or without hydroxychloroquine. Response rates were 85 percent versus 80 percent. Median progression-free survival was 12 months versus 11. Overall survival was 16 months versus 21. None of those differences was meaningful, and autophagy markers and quality of life did not differ either. [11]
Colorectal cancer (single-arm phase 2, reported as an abstract). Hydroxychloroquine was added to FOLFOX chemotherapy and bevacizumab in 37 patients. The response rate looked high, and the investigators called the results promising. Many patients left the study for surgery or other treatment, so survival is hard to read, and there was no comparison group. A randomized trial was planned. An abstract is not a finished paper. [12]
Not side-effect free. In the colorectal study, the hydroxychloroquine-related side effects included insomnia in about a quarter of patients, anxiety in a fifth, and vision changes in about one in ten. [12] At the higher doses used in cancer research, heart rhythm effects also have to be watched.
Ivermectin: early days, and a lesson from COVID
Ivermectin's cancer research is the youngest of the group. There is lab evidence, one registered human trial that has not reported results, and new state funding. There are no human efficacy results yet.
How it is thought to work
Ivermectin's anticancer effects in lab studies are spread across several targets, which makes them harder to explain neatly:
Jamming growth signals. It appears to interfere with some of the internal switches, such as the WNT and PAK1 pathways, that tell cancer cells to keep growing.
Damaging the power plants. It can disrupt mitochondria, the parts of the cell that make energy.
Waking up the immune system. In mouse studies, ivermectin made tumors more visible to the immune system, so that immunotherapy drugs worked better. This is the idea now being tested in people.
The COVID lesson
During the pandemic, ivermectin killed the virus in lab dishes. But the doses needed were far higher than a person can reach, and large randomized trials (TOGETHER, ACTIV-6, and PRINCIPLE) found no benefit. A drug that works in a dish but cannot reach that level in the body does not work. The same question hangs over ivermectin and cancer.
What is happening now
Cedars-Sinai trial. A single-group phase 1/2 trial in Los Angeles is combining ivermectin with immunotherapy (balstilimab or pembrolizumab) in women with metastatic triple-negative breast cancer. It is designed to test safety and early signs of benefit. About 34 patients are planned. As of the August 2026 ClinicalTrials.gov update it was still recruiting, with completion estimated for October 2026. No results have been posted. [13]
Florida funding. In September 2025, Governor Ron DeSantis and First Lady Casey DeSantis, herself a breast cancer survivor, opened $60 million in state cancer research grants, with priority for nutrition work and for repurposing generic drugs such as ivermectin. [14] In August 2026 the state announced that round's awards and opened a new $70 million cycle. [15] The award list includes a Moffitt Cancer Center project, led by Chang Jiang, titled "Repurposing Ivermectin for the Treatment of KEAP1/NRF2-Mutant Non-Small Cell Lung Cancer," funded at about $1.21 million under a generic-drug repurposing category. [18] That is a grant to study the question. It is not a result.
That money is welcome. It is also an admission that nobody yet knows whether these drugs work. Grants are not results.
Helpers, not replacements
Where there is any human signal at all, it shows up when these drugs are added to standard treatment, not used instead of it. Even there, the signal is mixed. That is the single most important point in this article.
DrugMain proposed mechanismBest human evidence so farKey cautionHydroxychloroquineBlocks autophagy (cell recycling)Several randomized phase 2 trials. Mixed. One survival endpoint missed; one pathologic response improved.Insomnia, anxiety, vision, and heart-rhythm effects at research dosesMebendazoleDisrupts microtubulesPhase 1 safety study, and a randomized phase 2 that missed its survival bar and had no drug-free armReversible liver-enzyme rises at the top studied dose. Results modest.IvermectinGrowth signals, mitochondria, immune effectsOne phase 1/2 trial under way. No results posted. One state grant awarded.Lab doses may be unreachable in peopleFenbendazoleDisrupts microtubulesCase reports only. Key paper retracted for an undisclosed conflict.Documented severe liver injury
A few practical cautions:
Never delay or drop proven treatment. The real harm from these drugs is not usually the drug. It is a patient who skips effective therapy while waiting for a cheap one to work.
Tell your oncologist everything you take. Liver injury from fenbendazole has been mistaken for immunotherapy side effects, leading to wrong treatment decisions. [6]
Veterinary products are not made for people. Doses, concentrations, and inactive ingredients differ, and nobody has worked out safe human dosing for fenbendazole. [16]
Ask about clinical trials. If you want access to one of these drugs, a trial is the safest route, and it adds to what we know.
Where this leaves us
I started down this road hoping to find good news, and in a way I did. Old, cheap drugs are being taken seriously. Mebendazole and hydroxychloroquine have been through real trials. Ivermectin has one under way, and a state has put money behind the question, including a funded Moffitt project.
But the honest answer today is that none of these drugs is proven to help people with cancer. The ideas are reasonable, the lab work is interesting, and the human results so far are modest or missing. That is exactly why we need properly designed studies, not more testimonials.
My advice is the same advice I gave clients throughout my career: look for good evidence, be skeptical of anything that sounds too good, and make decisions together with the doctor who knows your case.
In a future article, I will look at how artificial intelligence may help speed up this kind of research, both for cancer and for chronic inflammatory conditions, by sifting through existing drugs to find the ones worth testing.
References
Gallia GL, et al. Mebendazole and temozolomide in patients with newly diagnosed high-grade gliomas: results of a phase 1 clinical trial. Neuro-Oncology Advances, 2021. https://pmc.ncbi.nlm.nih.gov/articles/PMC7817892/
Patil VM, et al. Mebendazole plus lomustine or temozolomide in patients with recurrent glioblastoma: a randomised open-label phase II trial. eClinicalMedicine, 2022. https://pmc.ncbi.nlm.nih.gov/articles/PMC9156991/
Retraction statement: Fenbendazole as an Anticancer Agent? A Case Series of Self-Administration in Three Patients. Case Reports in Oncology, January 21, 2026. https://karger.com/cro/article/19/1/169/941881/Retraction-StatementPaper-by-William-Makis-Ilyes
Drug-induced liver injury in a patient with nonsmall cell lung cancer after the self-administration of fenbendazole. Case Reports in Oncology, 2021. https://karger.com/cro/article/14/2/886/820730/Drug-Induced-Liver-Injury-in-a-Patient-with
Severe drug-induced liver injury due to self-administration of the veterinary anthelmintic fenbendazole. ACG Case Reports Journal, 2024. https://pmc.ncbi.nlm.nih.gov/articles/PMC11068125/
Differentiating fenbendazole-induced liver injury from immunotherapy hepatitis: a case report. World Journal of Clinical Cases, 2026. https://pmc.ncbi.nlm.nih.gov/articles/PMC12836008
Drug-induced liver injury following co-ingestion of veterinary fenbendazole and ivermectin for prostate cancer: a case report. Cureus, 2026. https://assets.cureus.com/uploads/case_report/pdf/478304/20260515-240863-g1che1.pdf
Wolpin BM, et al. Phase II and pharmacodynamic study of autophagy inhibition using hydroxychloroquine in patients with metastatic pancreatic adenocarcinoma. The Oncologist, 2014. https://pubmed.ncbi.nlm.nih.gov/24821822
Zeh HJ, et al. A randomized phase II preoperative study of autophagy inhibition with high-dose hydroxychloroquine and gemcitabine/nab-paclitaxel in pancreatic cancer. Clinical Cancer Research, 2020. https://pmc.ncbi.nlm.nih.gov/articles/PMC8086597
Karasic TB, et al. Effect of gemcitabine and nab-paclitaxel with or without hydroxychloroquine on patients with advanced pancreatic cancer: a phase 2 randomized clinical trial. JAMA Oncology, 2019. https://pmc.ncbi.nlm.nih.gov/articles/PMC6547080
Targeting autophagy in platinum-sensitive relapsed ovarian cancer: randomized phase II trial of hydroxychloroquine with chemotherapy. Discover Oncology, 2025. https://link.springer.com/article/10.1007/s12672-025-01904-w
O'Hara MH, et al. Phase II trial of the autophagy inhibitor hydroxychloroquine with FOLFOX and bevacizumab in front-line treatment of metastatic colorectal cancer. ASCO abstract, 2017. https://ascopubs.org/doi/10.1200/JCO.2017.35.15_suppl.3545
Ivermectin in combination with balstilimab or pembrolizumab in patients with metastatic triple-negative breast cancer. ClinicalTrials.gov NCT05318469. https://clinicaltrials.gov/study/NCT05318469
Ivermectin, from the Capitol to state-funded cancer research. Florida Phoenix, September 24, 2025. https://floridaphoenix.com/2025/09/24/ivermectin-from-the-capitol-to-state-funded-cancer-research-its-a-thing-in-florida/
Governor Ron DeSantis announces $60 million in cancer innovation awards and opens a $70 million application cycle. Executive Office of the Governor, August 19, 2026. https://www.flgov.com/eog/news/press/2026/governor-ron-desantis-announces-60-million-cancer-innovation-awards-and-opens-70
Oral fenbendazole for cancer therapy in humans and animals. Anticancer Research, 2024. https://ar.iiarjournals.org/content/44/9/3725.abstract
Dogra N, Kumar A, Mukhopadhyay T. Fenbendazole acts as a moderate microtubule destabilizing agent and causes cancer cell death by modulating multiple cellular pathways. Scientific Reports, 2018. https://pmc.ncbi.nlm.nih.gov/articles/PMC6085345/
Florida Cancer Innovation Fund awardees, Fiscal Year 2025-26, Period 1. Florida Department of Health, August 2026. https://www.floridahealth.gov/wp-content/uploads/2026/08/Florida-Cancer-Innovation-Fund-FY25-26-Period-1-Grantees.pdf
