The Claim
Amygdalin or laetrile, including products derived from apricot kernels, is an effective treatment for cancer in humans.
Where It Comes From
Laetrile became a prominent alternative cancer treatment in the twentieth century and was often promoted as “vitamin B17,” although it is not a vitamin. Amygdalin occurs naturally in apricot kernels and other Prunus seeds. The claim persists through stories that a natural anticancer compound was ignored or suppressed.
Why It Sounds Convincing
There is a real biochemical hook: amygdalin can release cyanide, and laboratory studies have reported cytotoxic effects under some conditions. Case reports and survivor testimonials also make the treatment sound more established than the controlled human evidence is.
The Short Answer
The human treatment evidence does not support amygdalin or laetrile as an effective cancer therapy. The largest direct clinical study found no meaningful benefit, while oral use can produce cyanide exposure. Laboratory activity is not enough to overturn that clinical record.
How It Would Have to Work
For the treatment claim to hold, a tolerable dose would need to reach tumours, produce a selective anticancer effect stronger than its toxicity, and improve clinically meaningful outcomes such as tumour response, symptoms or survival. A mechanism demonstrated only in cultured cells would not be sufficient.
What Actually Happens
Amygdalin is a cyanogenic glycoside. Its metabolism differs by route of administration and can generate cyanide, especially after oral exposure. A compound can be biologically active and still fail as a useful drug because exposure, selectivity, dose and toxicity determine clinical value.
What the Evidence Shows
The 1982 New England Journal of Medicine study treated 178 patients with amygdalin plus diet, vitamins and enzymes, without a randomized concurrent comparison group. The U.S. National Cancer Institute’s clinical-study account records that 175 patients could be evaluated for tumour response: one had a partial response lasting 10 weeks, while all evaluable patients had disease progression within seven months after the intravenous treatment phase. These details neither show zero biological activity nor establish a lasting treatment benefit. Without a randomized comparator, individual tumour trajectories cannot be assigned causally to the therapy.
The later Cochrane review found no qualifying randomized or quasi-randomized trials. The combination of weak human efficacy evidence, very limited observed clinical responses and documented cyanide toxicity weighs against recommending amygdalin or laetrile as cancer treatment; laboratory cytotoxicity alone cannot establish improved survival.
The Strongest Evidence for the Claim
The strongest case combines biological plausibility with individual responses reported in uncontrolled clinical accounts and preclinical literature. The U.S. National Cancer Institute (NCI) historical summary records a small set of practitioner-submitted patients assessed as possible responders; those records were selected rather than a randomized or representative sample, and concomitant therapies complicate attribution. Cell and biochemical experiments can produce anticancer signals, but neither selected clinical anecdotes nor laboratory cytotoxicity establishes improved survival for patients.
The Strongest Evidence Against It
Direct human outcomes carry more weight for a treatment claim than cell assays. The main clinical dataset is negative across outcomes patients care about, and the intervention introduces a known cyanide hazard. No later high-quality clinical evidence has established a compensating benefit.
Common Viral Arguments
“Cancer cells release cyanide only inside the tumour” compresses a complex metabolic story into a selective-targeting claim not demonstrated clinically. “It is natural, so it is safer than chemotherapy” confuses origin with risk. “Positive cell studies prove it works” skips the translational steps between a dish and a patient.
The Reasoning Error
The recurring errors are testimonial evidence and mechanism substitution: a plausible biochemical story or an individual recovery is treated as if it established treatment efficacy. Appeal to nature can also make apricot-derived products sound intrinsically safer than they are.
What Would Change Our Conclusion?
Well-conducted randomized human trials showing reproducible improvements in tumour response, progression or survival at an acceptably safe dose would materially change the conclusion. A genuinely new delivery technology could reopen the question if it produced different human pharmacology and clinical outcomes.
Evidence Status
Status: Contradicted
Evidence quality: Moderate
The treatment proposition is directly at odds with the available human clinical evidence. Evidence quality is not rated High because the decisive clinical literature is limited and old rather than a large modern randomized program.
Evidence & Sources
The structured Evidence records and annotated Sources attached to this Investigation state what each source establishes, its relationship to the Claim and its limitations.
Corrections & Updates
October 10, 2026 — evidence clarification. Specified the 178 treated/175 evaluable uncontrolled regimen, one ten-week partial response, subsequent progression, 2022 US National Cancer Institute (NCI) PDQ (Physician Data Query) cancer-information context and cyanide hazard; no convincing human treatment benefit. Contradicted / Moderate unchanged.
Last Reviewed
October 10, 2026.