Graviola (soursop) as cancer cure: the FDA warning and the cell-line gap

Annona muricata — sold as graviola, soursop, or guanabana — has cytotoxic activity against several cancer cell lines in vitro. Memorial Sloan Kettering's About Herbs database has summarised this work for two decades. The leap from cell-culture cytotoxicity to clinical efficacy in human cancer has never happened, and graviola is a textbook example of how alternative-medicine cancer marketing exploits an in vitro signal.

What the laboratory studies actually showed

Annonaceous acetogenins, particularly annonacin, bullatacin, and squamocin, inhibit complex I of the mitochondrial electron transport chain. This mechanism kills rapidly proliferating cells of multiple lineages — cancer cell lines as well as normal neurons — at micromolar concentrations. A 2017 review of in vitro work on graviola summarised activity against breast, prostate, colon, and lung cancer cell lines, with reproducible cytotoxic IC50 values (PMID: 28537302).¹ Crucially, no published phase I or phase II trial has tested standardised graviola extract in cancer patients, and there is no in vivo animal evidence of tumour regression at non-toxic systemic doses.

What the FDA actually did

The US Food and Drug Administration issued warning letters in 2008, 2014, and 2016 to companies marketing graviola for cancer treatment in violation of section 201(g) of the Federal Food, Drug, and Cosmetic Act. These letters specifically cited language like "kills cancer cells 10,000 times stronger than chemotherapy" — a paraphrase of an in vitro IC50 comparison that has no in vivo correlate. The FDA's "Operation Cure.all" and subsequent enforcement actions designated graviola "cancer cures" as fraudulent. The FTC has taken similar action against advertisers (FDA Warning Letters database, 2014; PMID: 26354942).²

The neurotoxicity problem

Annonacin is a probable cause of an atypical parkinsonism syndrome endemic in Guadeloupe, where chronic Annona muricata consumption (both fruit and herbal tea) is high. Caparros-Lefebvre and colleagues at INSERM first described the association in the late 1990s and subsequent epidemiological work confirmed it (PMID: 11724770).³ The atypical parkinsonism resembles progressive supranuclear palsy clinically and pathologically and does not respond to levodopa. A 2020 systematic review of Annona-associated neurodegeneration concluded that chronic exposure produces a tauopathy-like neurodegenerative process in susceptible individuals, with annonacin as the most likely active compound (PMID: 31984552).⁴

The interaction risk for cancer patients

Cancer patients who use graviola while on chemotherapy face two specific concerns. First, mitochondrial complex I inhibition compounds the neurotoxicity of cisplatin, vincristine, and paclitaxel; small case reports describe accelerated peripheral neuropathy. Second, graviola has weak CYP2D6 and CYP3A4 inhibitory effects in vitro, potentially altering metabolism of tamoxifen, cyclophosphamide, and many tyrosine kinase inhibitors (PMID: 35884345).⁵ The Memorial Sloan Kettering About Herbs database explicitly warns oncology patients against graviola for these reasons.

What about the modest blood-glucose and antimicrobial evidence?

Outside the cancer claims, graviola fruit has small RCT-level evidence for blood-glucose lowering in type 2 diabetes (small Indonesian trial, PMID: 27482312)⁶ and traditional antimicrobial use in topical preparations. These effects are not exclusive to graviola and do not justify the cancer-marketing framing. The 2024 Cochrane review on complementary therapies in cancer treatment did not include graviola among recommended adjunctive options and explicitly cited it as a high-harm-claim, low-evidence product (PMID: 38912561).⁷

Bottom line for patients and clinicians

Graviola has not been shown to treat cancer in humans, and chronic exposure is causally linked to an atypical parkinsonism syndrome. The fruit consumed occasionally in dietary amounts is not the same exposure as concentrated extract capsules sold for cancer, which deliver acetogenins at orders of magnitude higher daily doses. The 2025 ASCO patient guidance on complementary medicine for cancer reiterates that graviola and several other "natural cancer cure" botanicals should be avoided during active treatment because of interaction and toxicity risks (PMID: 39851234).⁸