Clival chordoma research
Home › Diet, supplements and claims

Nine "anticancer" plants

Prepared 4 September 2026 by Kya at a friend's request, for her partner and the patient. Her partner remembered a plant said to hold "some of the strongest cancer-fighting compounds known" but not its name, so all nine plants people usually mean were researched to the same standard: what the compound does, whether it is a real drug and what that means for the plant, the evidence in people, anything in chordoma or chondrosarcoma, safety with the patient's medicines and around surgery and protons, and what can legally be bought in the UK. Every claim is sourced in research thread 11. It is not medical advice and nothing here argues for changing the standard plan of biopsy, surgery and proton therapy.

The short answer

The ranking

Scored 0 to 5 on human evidence of benefit in the form the family could use, relevance to chordoma or chondrosarcoma, safety in the patient, and UK practicality.

  1. Turmeric (curcumin). Some laboratory data in chondrosarcoma, the lookalike, none in chordoma. A few small, unconfirmed human signals in other cancers. No effect on thinking. Risks are known and manageable: keep to culinary or modest doses; avoid piperine (black pepper extract) and "high absorption" products, which are behind a run of liver injury cases; stop two weeks before surgery and during proton therapy because it is an antioxidant. An acceptable adjunct, not a treatment.
  2. Black seed (thymoquinone). Safe as a food. Almost no human anticancer evidence and none for chordoma. At supplement doses it inhibits the liver enzymes that clear opioids and other drugs, and animal studies show sedation, so nothing beyond culinary amounts on top of opioids or diazepam.
  3. Mistletoe (Iscador, Helixor). The most-trialled plant medicine, which is the lesson: when finally tested double-blind at scale in 2024 (290 patients) it did not extend survival. Weaker trials suggest some quality of life benefit. Injectable, private, about 3,000 pounds a year, no chordoma data, and it causes fevers, which are unwelcome around skull base surgery. Only for wellbeing, with the oncologist informed, and not before recovery from surgery.
  4. Sweet wormwood (artemisinin, artesunate). The most scientifically interesting: artesunate, the malaria drug, is in real cancer trials, with a 23-patient bowel cancer pilot and a larger trial not yet reported. But the shop "artemisinin" supplement is not the trialled drug, and artesunate switches on a liver enzyme that would destabilise her opioid, dexamethasone and carbamazepine levels. Oncologist or trial only, and there is no chordoma trial. 5 to 8. Happy tree, mayapple, yew, periwinkle. Sources of irinotecan and topotecan, etoposide, paclitaxel and docetaxel, vincristine and vinblastine. Poisonous as plants; the drugs belong to oncologists; chordoma cells resist them in the dish and the one trial gave one response in fifteen. Etoposide and the camptothecins sit above the others only because they have thin paediatric chordoma data that would matter if histology showed a poorly differentiated tumour.
  5. Graviola or soursop. Last, and separately. No human evidence. Its acetogenins poison mitochondria and the plant is linked to an atypical parkinsonism in Guadeloupe. Sold without standards. Specifically unsuited to a patient whose brainstem and cranial nerves will be examined closely for the next year. The likely source of the claim, and the one to say no to.

Why chordoma is different

Conventional chordoma barely responds to chemotherapy of any kind, which is why surgery and proton therapy are the treatment. That is the reason the drug-origin plants score so low despite being real medicines. It also means that "strongest compound" claims from cell studies in other cancers do not transfer.

What would change the picture

Questions for the oncology team

  1. Which of the patient's medicines are cleared by the liver enzymes CYP3A4, CYP2D6 or CYP2C9, and do you want her to avoid herbal supplements that affect them?
  2. What is your policy on supplements, especially antioxidants, in the two weeks before surgery and during proton therapy? Is culinary turmeric or black seed acceptable?
  3. Will you record any supplement she takes and tell the anaesthetist and the proton centre?
  4. Once histology is back, does the subtype change whether any chemotherapy has a role?
  5. Is there any trial for skull base chordoma or chondrosarcoma she could be considered for, in the UK or Europe?
  6. Given her weight loss, would you rather she saw a dietitian than added supplements?
  7. If the family want a prescribed complementary product later, would you object, and would you want the product, dose and prescriber recorded?

Key sources

Addendum, 12 September 2026: NMN and NAD+ precursors

Her partner asked where I stand on NMN (nicotinamide mononucleotide). Answer given: no, not now, and specifically not before radiotherapy. The clearest no given on any supplement so far, and the reasoning is mechanistic and specific rather than generic caution.

The case against, for the patient specifically.

NMN is an NAD+ precursor. NAMPT, the rate limiting enzyme of the NAD salvage pathway that produces NMN intracellularly, is upregulated in most cancers and associated with tumour progression. NAD+ is the substrate for PARP1 (DNA repair), the sirtuins SIRT1 and SIRT6, and CD38, which between them support genomic stability, pro-inflammatory cytokine secretion, angiogenesis, motility and invasion. An entire anticancer drug programme is built on depleting NAD+ in tumours: FK866, CHS828, STF-118804, and OT-82, the last developed in Ewing sarcoma, the nearest bone sarcoma neighbour to chordoma. Supplementing NMN pushes in the opposite direction to that programme.

The timing argument, which is the strongest one. She is heading for surgery and then almost certainly proton radiotherapy. Radiotherapy kills by DNA damage; PARP1 repairs DNA damage and consumes NAD+ to do it. Supplying NAD+ precursor to a tumour about to be irradiated is, on mechanism, supporting its capacity to survive treatment. Untested in humans, which is precisely the objection.

The honest counter-case, given to her partner. An NAD+ precursor (nicotinamide riboside) suppressed hepatocellular carcinoma progression in mice, so the preclinical biology is not unidirectional. NAMPT inhibitors that performed well in animal models failed to translate in patients, implying alternative NAD+ biosynthetic routes and an incompletely understood pathway. Human NMN trials are small, short, and report surrogate outcomes: NAD+ levels, walking distance, insulin sensitivity, sleep quality. None reports cancer outcomes. Nothing exists in chordoma.

Regulatory. NMN is not an authorised novel food in the UK. The FDA excluded it from the dietary supplement category in 2022 after it was investigated as a drug.

Redirect given. her partner is likely asking because the patient was markedly tired on 12 Sep. The tiredness workup is the bloods asked for that morning, cortisol, sodium, calcium, magnesium and vitamin D level, plus the sedation load of morphine MR 40 mg/day, gabapentin 900 mg/day and melatonin 4 mg. Two of those causes are correctable this week; NMN addresses none of them.

Levels in a friend's terms: the NAD+ and cancer mechanism is established; the tumour promotion risk from supplementation is frontier; the human benefit claims are anecdotal. Position: revisit after treatment if she wants to, not while there is an untreated tumour.

Sources: PMC10889166 (inhibitors of NAD+ production in cancer treatment), Nature Oncogenesis s41389-020-00264-0 (OT-82 in Ewing sarcoma), PMC5689593 (STF-118804 in pancreatic cancer), PMC10055624 (NR suppressing hepatocellular cancer in mice, the counter-case).

Follow-up 12 Sep 13:54: "are there risks to her taking it right now". Answered in three parts, because the question deserved separation rather than a blanket no.

  1. Direct toxicity: low. Human NMN trials to roughly 1,250 mg/day over weeks to months report mild adverse effects and no serious toxicity signal. No known interaction with morphine, gabapentin or diclofenac.
  2. The risk that matters is unquantifiable. The tumour biology argument is mechanistic; nobody has given NMN to a patient with an active skull base tumour and followed the outcome. Said plainly to her partner that I cannot claim the risk is small because no one can, only that the direction is wrong.
  3. The practical risk, which is live this week: three new symptoms are under active interpretation (hiccups, phantosmia, jaw and occipital sensitivity). Introducing a new agent now confounds the symptom record in the week before the neurosurgical appointment.

Position given: if already taking it, no alarm and no rush to stop, just tell me so it is recorded. If not started, not this week. Standing rules restated: stop anything not prescribed a week before any procedure, and the team must be told what she takes.

Action raised: a complete non-prescription list is missing. Asked her partner for everything she takes that is not on prescription, supplements included, to go on the Wednesday sheet for Mr Syrris. Currently recorded: vitamin D3 1,000 IU x2 and vitamin C 950 mg. That is very unlikely to be the whole picture, and it should have been asked for before now.

Addendum, 13 September 2026: Essiac tea

Asked by her partner in em-research at 23:10 on 13 September. Position: no. This is the first item in this file where the objection is not merely absence of evidence; there is a laboratory finding pointing the wrong way, and three specific interactions with medicines the patient is actually taking.

What it is. A four herb decoction: burdock root (Arctium lappa), Indian rhubarb root (Rheum palmatum), sheep sorrel (Rumex acetosella) and the inner bark of slippery elm (Ulmus rubra). Popularised by Rene Caisse, an Ontario nurse, from the 1920s; "Essiac" is her surname reversed. Sold also as Flor-Essence, an eight herb variant.

Efficacy: nothing, repeatedly. No human trial has shown an effect on any cancer. Tested and found ineffective by Memorial Sloan Kettering and the US National Cancer Institute between the 1950s and 1980s, by the University of British Columbia in 2005 and by the University of Michigan in 2006. The Natural Standard systematic review searched ten databases, twenty additional journals and fifty secondary bibliographies with no language or quality restriction and found no high quality clinical trial supporting any traditional use (PMID 19476742, L1 review of an empty field). Cancer Research UK and the NCI both state there is no evidence to support its use.

The one substantial human study. TEA-BC, a retrospective cohort of 510 women randomly drawn from the Ontario Cancer Tumour Registry with primary breast cancer diagnosed in 2003, comparing new Essiac users with never-users (PMID 17212569, L3). No significant effect on health related quality of life or mood. On the physical wellbeing and doctor relationship subscales the users did worse. The authors attribute that honestly to confounding: users were younger and had more advanced disease. Typical daily dose 43.6 ± 30.8 mL, per label directions. Friends were the commonest source of information.

The finding that decides it. Kulp and colleagues, Breast Cancer Research and Treatment 2006 (PMID 16541326, L5). Flor-Essence and Essiac at 1, 2, 4 and 8 percent stimulated proliferation of all four human breast cancer cell lines tested (MCF-7, T47D, MDA-MB-231, MDA-MB-436), in both oestrogen receptor positive and receptor negative lines, and produced a dose dependent increase in ER dependent luciferase activity in MCF-7. ICI 182,780 blocked the reporter signal but not the proliferation, so the growth stimulation runs through both ER mediated and ER independent mechanisms. Earlier in vitro comparison work is equivocal (PMID 14719086, L5); chemistry of the constituent herbs reviewed at PMID 10641040.

Honest limits: cell culture, breast cancer rather than chordoma, whole tonic rather than a defined constituent. But it is the only directional signal anyone has produced, and it is the wrong direction.

Three hazards specific to the patient, which weigh more than the efficacy question.

  1. Oxalate load. Sheep sorrel and rhubarb root are both rich in oxalic acid. She is 51 kg, taking diclofenac daily, and no creatinine or eGFR has ever been shown to the family despite bloods being ordered on 8 September. Oxalate plus an NSAID plus unknown renal function is not a combination to add blind. Calcium, magnesium and kidney function are already on the Wednesday bloods ask.
  2. Additive anthraquinone laxative. Rhubarb root is an anthraquinone, the same pharmacological class as the senna she takes two tablets of every night. She passed fresh red blood with mucus from the back passage at 21:41 on 12 September, attributed to NSAID proctitis.
  3. Absorption interference. Slippery elm is a mucilage that coats the gastrointestinal lining and is conventionally advised to be separated from other medicines. She takes modified release morphine, modified release diclofenac and gabapentin. Blunting absorption of her background opioid is a real risk with nothing on the other side of the ledger.

Where it lands. No, without qualification. Told her partner plainly, and told him equally plainly that he has been right to push on other items here, vitamin D and the NAD+ objection in particular, and that this is not one of them.

Brief for the family · source file em-docs/em-plant-compounds.md · last edited 14 September 2026