Weizmann researchers traced a pathway from PDE5 inhibition through cGMP to cholesterol transport, then looked for a survival signal in five million people’s records.

Sildenafil was developed for angina, found to do something else, and has been taken by a very large number of people for a very long time. That combination makes it an attractive candidate for repurposing, because the safety question is largely settled and the prescription records are extensive.

A study in Cancer Research from the Weizmann Institute of Science reports a mechanism by which it might restrict cancer metastasis, and a signal in Israeli health records consistent with that mechanism.

The study was led by Dr Yarden Ariav in the laboratory of Professor Ayelet Erez at the Weizmann Institute, with Professor Eytan Ruppin of the US National Cancer Institute, Professor Shay Ben-Shachar of the Clalit Health Services Innovation Division, and physicians and scientists at Rabin Medical Center.

The pathway

The chain runs through three steps, each of them a known piece of biochemistry assembled in a way that had not been noticed.

Sildenafil blocks PDE5, the enzyme that degrades cyclic GMP, so cGMP accumulates. That part is the drug’s established action — the one that relaxes smooth muscle and produces its licensed effects.

The new step is what the accumulated cGMP then does. It binds a protein that transports cholesterol within cells, which leaves less cholesterol available to cancer cells. And cancer cells deprived of cholesterol have a harder time forming metastases.

“We have uncovered a new biological pathway that links a well-known signaling molecule to cholesterol regulation,” Erez said.

Why cholesterol should matter to a metastasising cell is reasonably well understood. Membranes are built from it, and a cell that is dividing, deforming to squeeze through tissue, and remodelling its surface to survive in the bloodstream needs a great deal of membrane. Cholesterol also organises the lipid domains that many signalling receptors sit in, so a shortage degrades more than structure.

The logical extension is a combination, and the authors make it: pairing PDE5 inhibitors with statins may work better, since statins reduce how much cholesterol a cell produces in the first place. Blocking supply and blocking distribution are complementary.

Three kinds of evidence

The study draws on mouse models, human cancer cell cultures, and data from roughly five million members of Clalit Health Services collected over twenty years.

In the Clalit data, cancer patients who had taken sildenafil showed significantly improved survival, and the effect was particularly evident when sildenafil was combined with statins.

Having a mechanism and a population signal that point the same way is better than either alone, and it is worth being exact about what the population half can and cannot establish.

Why the survival signal needs scepticism

This is observational data on prescriptions, and the biases it is subject to are the well-documented ones that have repeatedly produced drug-repurposing findings that failed when tested properly.

The clearest is the healthy-user problem. The men who receive sildenafil are a selected group: well enough to be sexually active, in regular contact with a doctor, insured, generally less frail. Each of those predicts better cancer survival independently of anything the drug does. A cancer patient who is being prescribed sildenafil during treatment is, by that fact alone, a patient doing comparatively well.

The statin interaction compounds it rather than resolving it. Statin users are people whose cardiovascular risk is being actively managed, which means they are monitored, adherent and alive long enough to accumulate prescriptions — all markers of the kind of patient who survives longer. A finding that the combination looks best is exactly what confounding by healthcare engagement would produce.

There are further asymmetries. Sildenafil is also prescribed for pulmonary hypertension, a different population with a different prognosis. Prescription records show dispensing rather than consumption. And the comparison group in a male-skewed exposure is not matched on cancer type, since the cancers that occur in men who take sildenafil are not the full distribution of cancers.

None of this means the association is spurious. It means that twenty years of records and five million people buy statistical precision rather than causal confidence, and that a large observational effect of this kind has a poor historical record of surviving a randomised trial.

What would settle it

Erez’s own framing is broader than the drug.

“Cancer biology is shaped not only by mutations in tumor cells,” she said, and “our study underscores the importance of treating the whole patient”.

That is the defensible version of the result: metastasis depends on the metabolic environment a tumour sits in, and an available drug that alters cholesterol handling in that environment is worth testing. Testing means a randomised trial with a defined cancer type, a defined stage, and metastasis or survival as the endpoint — none of which exists yet.

The release also reports no numbers at all beyond the five million members and twenty years. There is no hazard ratio, no survival difference, no confidence interval, no mouse survival figure, and no indication of how many cancer patients in the cohort took sildenafil. The paper’s title is cut off on the page, so even its full claim is not stated there.

It needs saying plainly, because this is the kind of finding that travels: nobody with cancer should start taking sildenafil, or add a statin, on the strength of this work. The evidence is a mechanism in mice and cells plus a correlation in prescription records, and the people best placed to judge whether any of it applies to a particular patient are that patient’s oncologists.

Erez’s research is supported by the David and Fela Shapell Family Center for Genetic Disorders Research, the Moross Integrated Cancer Center, the EKARD Institute for Cancer Diagnosis Research, the Abisch-Frenkel RNA Therapeutics Center, the Sagol Institute for Longevity Research and the Koret Foundation. The release states no conflicts of interest.

---

Notes

1. Cancer Research, 2026. DOI: 10.1158/0008-5472.CAN-26-1818. The release gives only a truncated paper title and no full author list.
2. Weizmann Institute of Science. “Viagra may reduce cancer metastasis, study shows.” EurekAlert!, 4 August 2026. https://www.eurekalert.org/news-releases/1138756.

Photo by Adrian Frentescu on Pexels.com