Dihexa and cancer risk: separating lab data from TikTok hype

What did @moistbreadcrumbs2.0 actually say?

The creator argued that Dihexa's cancer risk is overblown, rooted in people misunderstanding what an oncogene actually is. Their core point: C-Met being an oncogene does not mean activating it causes cancer, and Dihexa's mechanism as a potentiator of HGF limits how much it can push that receptor anyway. That argument is more sophisticated than most peptide TikTok content, and it deserves a real evaluation rather than dismissal.

The claim that "the oncogene has to mutate before it even has the potential to cause" cancer is the centerpiece of their reasoning. They also pointed to phase 2 clinical trial data on Dihexa's prodrug showing low toxicity as supporting evidence. These are legitimate scientific points worth examining carefully, because some of them hold up, and some require significant qualification.

Does the science back this up?

Partially, yes. The oncogene framing is roughly correct, but it glosses over real complexity in the HGF/C-Met pathway that researchers take seriously. C-Met is not just an oncogene in mutated form. Overactivation of wild-type C-Met through excessive ligand signaling has been documented as a driver of tumor progression in several cancer types.

A 2012 review by Gherardi et al. in Nature Reviews Cancer documented that HGF/C-Met signaling promotes tumor invasion, angiogenesis, and metastasis even without receptor mutation, through mechanisms including transcriptional upregulation and autocrine signaling loops. The idea that you need a mutation before C-Met becomes oncologically relevant is an oversimplification. In established tumors especially, adding HGF potentiation is not necessarily a neutral act. Separately, the prodrug they reference, likely PNB-0408 or a related compound, does show early-phase tolerability data, but phase 2 trials in humans with cognitive conditions are not a cancer clearance study. That framing matters.

What did they get wrong (or right)?

They got the basic oncogene biology directionally right. Most oncogenes do require gain-of-function mutations or chromosomal alterations to become active cancer drivers, and the multi-hit model of oncogenesis (Knudson, 1971, PNAS) is well-established. Dismissing C-Met as dangerous purely because it appears on an oncogene list is genuinely bad reasoning, and the creator is correct to push back on that.

What they got wrong is the HGF bottleneck argument. They claim Dihexa is "bottlenecked by HGF itself" because it potentiates the ligand rather than acting directly on the receptor. But this framing assumes endogenous HGF levels are always the limiting factor. In inflammatory states, post-injury tissue, or in individuals with certain cancers, HGF is already elevated. Potentiating an already-elevated ligand is not the same as having a hard ceiling on receptor activation. The phase 2 trial citation also needs context: early-phase trials are powered to detect acute toxicity, not to assess cancer risk over years of use.

What should you actually know?

Dihexa is a synthetic peptide originally developed at Washington State University, investigated primarily for cognitive function. It is not FDA-approved and has not completed large-scale human safety trials. The HGF/C-Met pathway is genuinely complex, and the question of whether long-term potentiation of that pathway in healthy humans carries cancer risk remains unanswered, not because researchers are uninformed, but because the studies have not been done.

The creator frames cancer concern as coming from "uneducated individuals doing very light research." That is too dismissive. Researchers at major cancer centers actively study HGF/C-Met as a therapeutic target in oncology, including efforts to inhibit it. The fact that C-Met inhibitors are in clinical development for cancer treatment should at minimum prompt humility about the consequences of chronically amplifying that same pathway. This does not mean Dihexa causes cancer. It means we do not know, and intellectual honesty requires saying so.

• Dihexa has shown cognitive effects in rodent models (McCoy et al., 2013, Journal of Pharmacology and Experimental Therapeutics)
• HGF/C-Met overactivation without mutation has been linked to tumor progression in breast, lung, and gastric cancers
• No long-term human cancer surveillance data exists for Dihexa specifically
• The multi-hit oncogenesis model supports the creator's point that C-Met alone is not sufficient for cancer
• Early-phase clinical trials on related compounds show tolerability, not long-term oncological safety