Lapacho (Pau d'Arco)

Handroanthus impetiginosus (Mart. ex DC.) Mattos belongs to Bignoniaceae, the trumpet-creeper or catalpa family. Plants of the World Online (POWO), the Kew taxonomic database, accepts the name and dates it to the journal Loefgrenia in 1970, where J. R. Mattos proposed Handroanthus as a new genus for the ipês of Brazil. For most of its history the tree was a Tabebuia. POWO records that Martius used the name Tecoma impetiginosa without a description in his 1843 work on Brazilian medicinal plants, that it was formally published in de Candolle’s Prodromus in 1845, and that Standley moved it to Tabebuia in 1936; it lists 18 synonyms in all, including Tabebuia avellanedae Lorentz ex Griseb. (1879), the name used in many lapacho studies. A 2022 CITES proposal explains that Tabebuia, created by de Candolle in 1838, was shown by the 2007 molecular work of Grose and Olmstead to contain three separate lineages, now the genera Tabebuia, Handroanthus, and Roseodendron, with 113 species between them. Alwyn Gentry’s 1992 Flora Neotropica treatment kept the species in Tabebuia, which is why both names remain in use. Zhang and colleagues (Molecules 2020) note that the epithet comes from impetigo, the contagious skin infection, because the plant was believed to treat it.

Zhang and colleagues describe a deciduous tree up to 30 m tall that sheds its leaves in the dry season. The leaves are palmately compound, with serrated, elliptic to oblong leaflets, and are arranged in opposite or nearly opposite pairs. Showy purple, dark pink, or pink trumpet-shaped flowers have a bell-shaped to tubular, five-lobed calyx, and the fruit is a long, cylindrical capsule holding thin seeds. The CITES proposal adds that the species can reach 50 m tall and 2 m in trunk diameter, growing as large trees in remnant Atlantic Forest but staying much smaller in the Brazilian cerrado savanna, that the trees need light, and that in Bolivia the seeds of Handroanthus species are dispersed by wind and stay viable for no more than three months. POWO gives the native range as central Mexico through Central America to Colombia, Venezuela, French Guiana, Suriname, Peru, Bolivia, Brazil, Paraguay, and northwestern Argentina. According to the CITES proposal, the species grows in tropical and subtropical climates with seasonal rainfall of 500 to 1,200 mm a year, from sea level to 1,400 m, on well-drained plateaus, slopes, and even steep terrain. It is naturally sparse: a 1970s survey of the Brazilian Amazon found 0.11 trees per hectare, against 2.5 per hectare in Bolivian forest inventories.

The wood is the tree’s main commercial product. The CITES proposal describes ipê wood as very hard, heavy, and durable, resistant to fungi and termites, with yellowish-brown to dark olive heartwood whose channels contain a greenish-yellow deposit known as lapachol or ipeina. Medicinal chemistry is less tidy. Gómez Castellanos and colleagues list flavonoids, cyclopentene dialdehydes, benzoic acid derivatives, furanonaphthoquinones, naphthoquinones, and anthraquinones in the bark and heartwood, and Zhang and colleagues report that beta-lapachone is found mainly in the heartwood, while 19 glycosides, along with oxalic acid and sugars, have been isolated from the bark. Park and colleagues (Journal of Ethnopharmacology 2006) isolated lapachol and two anthraquinones from dried inner bark sold as taheebo. Yet a Health Canada team working with Gentry (Awang and colleagues, Journal of Herbs, Spices and Medicinal Plants 1994) found no detectable lapachol in bark from three authenticated Tabebuia species and less than 0.001% in the wood, and the naphthoquinones they did find in bark differed from those usually reported. Several species are sold as lapacho: MSKCC lists Tabebuia heptaphylla alongside H. impetiginosus, and a Paraguayan crude drug called tayï pytá is the bark and trunkwood of that species (Schmeda-Hirschmann and Papastergiou, 2003).

Gómez Castellanos and colleagues describe a long tradition of local use in South America, with the bark peeled and rasped once or twice a year and boiled into a decoction. They give the Guarani and Tupi name as tajy, which they gloss as “to have strength and vigor,” and report that biomedical uses of pau d’arco were recorded as early as 1873. Local uses vary by region. Portillo and colleagues (Journal of Ethnopharmacology 2001) studied lapacho as one of 14 Paraguayan plants used in traditional medicine for skin diseases, and in Paraguay tayï pytá, from the related Tabebuia heptaphylla, is taken for cancer, wounds, and inflammation. De Miranda and colleagues (BMC Pharmacology 2001) report that in northeastern Brazil the inner bark is used as a painkiller, anti-inflammatory, anticancer remedy, and diuretic. A 2022 systematic review by Lima Bezerra and colleagues (Toxicon) found five Handroanthus species used for cancer in the traditional medicine of Brazil, Bolivia, and Argentina, with bark the part used in 88% of reported preparations. Even the species name comes from medicine: it first appeared in Martius’s 1843 account of Brazilian medicinal plants (POWO).

The modern cancer reputation began in Brazil. Gómez Castellanos and colleagues write that the bark came back to the attention of clinicians and the public in 1967 after Brazilian press reports, when the news magazine O Cruzeiro began describing “miraculous” cures in cancer patients at a hospital. Scientific interest grew in parallel. Between 1960 and 1990 the NCI screened plant extracts from around the world for anticancer compounds, and lapachol was identified as the compound responsible for the activity attributed to the tree. In an NCI-funded study, Rao, McBride, and Oleson (Cancer Research 1968) isolated lapachol from extracts of another tree, Stereospermum suaveolens, and found highly significant activity against the Walker 256 carcinosarcoma in rats by injection and by mouth, with the best results from oral doses twice a day. Block and colleagues reported early clinical studies of oral lapachol in 1974. MSKCC summarizes the outcome as no clinical effect in patients with chronic myelocytic leukemia, with nausea, vomiting, and urine discoloration. A decade later, Preusch and Suttie (Archives of Biochemistry and Biophysics 1984) showed that lapachol blocks vitamin K recycling, explaining anticoagulant effects that had already been seen in rats and in people. Lapachol never became an approved cancer drug, and later drug development focused on beta-lapachone.

By 2009 Gómez Castellanos and colleagues were describing red lapacho as a global commodity, sold as loose tea or tea bags and promoted as a wonder drug for cancer, and as a case study in how local knowledge is globalized and commercialized. Regulators have occasionally stepped in. In 2003 the US Federal Trade Commission (FTC) sued the marketers of Seasilver, a liquid supplement containing pau d’arco among other ingredients, which the FTC said was advertised as a cure for cancer and hundreds of other diseases; a 2004 settlement judgment barred them from claiming that any product containing pau d’arco treats cancer without reliable scientific evidence. In 2012 the European Food Safety Authority (EFSA) included Tabebuia bark in its compendium of botanicals containing substances of possible concern, citing lapachol and beta-lapachone. Meanwhile the trees were being felled for timber. After a proposal led by Colombia, the European Union, and Panama, parties to CITES voted in November 2022 to list all ipê species (Handroanthus, Tabebuia, and Roseodendron) in Appendix II, and the US Fish and Wildlife Service notes that the listing took effect on November 25, 2024.

The best-characterized action is on blood clotting. Preusch and Suttie showed that lapachol is a potent inhibitor of vitamin K epoxide reductase and vitamin K quinone reductase in rat liver, the enzymes that recycle vitamin K so it can keep activating clotting proteins. The inhibition resembled that of coumarin anticoagulants such as warfarin. It was not seen in liver from warfarin-resistant rats, and the authors proposed that lapachol takes on a shape similar to the 4-hydroxycoumarins. MSKCC describes this as an action similar to warfarin’s. A second route may involve platelets: Son and colleagues (Journal of Ethnopharmacology 2006) found that fractions of an inner-bark extract inhibited rabbit platelet clumping triggered by collagen and arachidonic acid, partly by suppressing the release of arachidonic acid. How much of either effect survives in a cup of tea is unknown, because lapachol content in bark products varies so widely.

Beta-lapachone is the compound behind most cancer research. Gómez Castellanos and colleagues call it the main antitumor compound, with cell-death effects seen in the test tube. Its mechanism depends on NQO1, an enzyme found at high levels in many solid tumors. Bey and colleagues (Molecular Cancer Therapeutics 2013) showed in breast cancer cells that NQO1 drives beta-lapachone through a futile redox cycle, generating about 120 moles of superoxide per mole of drug in two minutes. The resulting reactive oxygen species break DNA strands, overactivate the repair enzyme PARP1, drain the cell’s NAD+ and ATP, and trigger programmed necrosis; adding the antioxidant enzyme catalase protected the cells. MSKCC lists further laboratory effects, including reduced COX-2 and telomerase activity in prostate cancer cells and apoptosis through mitochondrial signaling. In people, the drug version ARQ 761 caused anemia in 79% of patients and possible methemoglobinemia, a form of hemoglobin that cannot carry oxygen, showing that the same oxidative chemistry that kills tumor cells can harm healthy ones.

Antimicrobial and anti-inflammatory effects are laboratory and animal findings. In Portillo’s screen, lapacho extracts were among the most active of 14 Paraguayan plants against 11 fungal strains, including yeasts. Park and colleagues found that lapachol and two anthraquinones from taheebo inner bark inhibited Helicobacter pylori at 2 to 8 micrograms per milliliter, stronger than metronidazole but far weaker than amoxicillin or tetracycline. Byeon and colleagues (Journal of Ethnopharmacology 2008) reported that a water extract suppressed prostaglandin E2 and nitric oxide in mouse immune cells and, at 100 mg/kg by mouth for a week, prevented ear swelling caused by arachidonic acid in mice, and de Miranda and colleagues found that an inner-bark extract reduced pain behavior in mice by 44% to 64% and paw swelling in rats by 13%. MSKCC adds antipsoriatic, immunomodulating, and antidepressant-like effects in laboratory and animal studies and stresses that human studies are needed. It also warns that effects are hard to predict because species, plant part, extraction method, absorption, dose, and route all vary.

Cancer: this is why most people buy pau d’arco, and it is the use with the least support. No clinical trial has tested lapacho bark as a cancer treatment. The only human data on its compounds come from lapachol, which MSKCC says had no clinical effect in chronic myelocytic leukemia, and from ARQ 761, an intravenous beta-lapachone drug. In that phase 1 trial, Gerber and colleagues (British Journal of Cancer 2018) treated 42 people with advanced solid tumors; among 32 who could be evaluated, the best response was stable disease in 12, with some tumor shrinkage in 6, and the authors described only modest activity. Lima Bezerra and colleagues’ 2022 review of 78 studies found that Handroanthus extracts are cytotoxic to many cancer cell lines, but that toxicology results were inconclusive and caution is needed. Laboratory results also depend on the preparation: Pires and colleagues (Molecules 2015) found that a methanol extract of the bark was toxic to four human cancer cell lines, while an infusion, pills, and syrup made from the same plant material were not.

Mouth sores during radiotherapy: Giacomelli and colleagues (Nutrition and Cancer 2015) gave Orasol Plus, a lapacho-based swallowable solution, three times a day to 40 consecutive adults throughout radiotherapy for head and neck cancer. Grade 2 oral mucositis developed in 27.5% and grade 3 in 10%, and no one reached the most severe grade, 4; no one had to stop radiotherapy because of mouth sores, and the solution was well tolerated. MSKCC counts this among preliminary findings of benefit. But the study was designed mainly to test tolerance and safety, it had no control group, and the abstract does not describe the product’s full composition, so it cannot show that lapacho prevented anything. The authors themselves called for a phase III trial. People having radiotherapy should follow their treatment team’s mouth-care plan and ask before adding any product.

Period pain: McClure and colleagues (Advances in Integrative Medicine 2022) gave 12 generally healthy women aged 18 to 45 with primary dysmenorrhea 1,050 mg a day of encapsulated pau d’arco (Tabebuia avellanedae) for eight weeks. The trial was single-arm and open-label, and its main purpose was safety. Nine of the 12 finished. Pain intensity fell significantly from baseline after the first dose and at four and eight weeks, while pain interference, quality of life, sexual function, and the inflammation marker hs-CRP improved but not significantly. The authors called the capsules generally safe with moderate tolerability and said efficacy studies are warranted. Without a placebo group, a fall in pain over two months cannot be separated from expectation or natural variation, so this is a starting point for research, not evidence that pau d’arco works for cramps.

Infections: lapacho is widely promoted for candida, other fungal infections, and bacterial infections, and traditionally for malaria and trypanosomiasis. Laboratory support is real: Portillo’s antifungal screen and Park’s Helicobacter pylori work both found activity, and MSKCC lists antibacterial and antifungal effects in laboratory studies. But MSKCC states plainly that these effects have not been tested in humans. No controlled trial has asked whether pau d’arco tea or capsules clear thrush, vaginal yeast infections, H. pylori, or any other infection, and concentrations that work on a lab plate may never be reached in the body. Proven antifungal, antibiotic, and antiparasitic treatments remain the standard, and an infection that is not improving needs medical review rather than more tea.

Inflammation, pain, and other claims: anti-inflammatory and pain-relieving effects are supported by mouse and rat studies (Byeon, de Miranda) and by cell work, and lapacho products are also promoted for arthritis, immunity, psoriasis, and mood. Zhang and colleagues’ 2020 review of the plant’s immune-related research concluded that most work is still in cell and mouse models and that further evidence from clinical trials on effects and safety is required. Gómez Castellanos and colleagues were blunter: the bioscientific evidence for products derived from the tree is insufficient, and progress depends first on establishing proper quality control. For now pau d’arco has no use supported by controlled human evidence.

Traditionally, lapacho is a decoction rather than a steeped tea: Gómez Castellanos and colleagues describe inner bark and heartwood, peeled and rasped, boiled in water, and sold internationally as loose material or in tea bags. Modern products include shredded bark, capsules and tablets of powdered bark, liquid extracts, and multi-ingredient tonics. There is no regulator-approved or clinically established dose. The only oral amounts tested in recent human studies are 1,050 mg a day of encapsulated bark for eight weeks in the period-pain trial and a lapacho-based solution taken three times a day during radiotherapy, neither of which established that the dose was effective. MSKCC does not publish a dose for the public. Label directions and traditional recipes are not based on dose-finding studies, and it is not known how much lapachol or beta-lapachone, if any, a given cup or capsule delivers.

Product quality is the central problem. Awang and colleagues analyzed 15 lapacho, pau d’arco, and taheebo products sold in Canada and 2 from Brazil. Naphthoquinones were present in all the plant samples except 2, but none were detected in the 3 Canadian concentrates; both Brazilian products contained lapachol, with about 1.0% in the woody material. The authors concluded that most products probably did come from Tabebuia species, but that naphthoquinone levels were highly variable, suggesting changes after harvest. Gómez Castellanos and colleagues add that lapacho products are generally not characterized chemically or pharmacologically and that quality assurance is lacking from crude drug to final product. Different species are sold under the same names, and the CITES proposal notes that the wood of Handroanthus, Tabebuia, and Roseodendron cannot be told apart even under a microscope.

If you decide to try pau d’arco despite the thin evidence, treat it as an unstandardized product with real risks. Avoid concentrated extracts and anything sold as isolated lapachol or beta-lapachone, and look for products that name the species and plant part and come from a manufacturer that tests identity. Do not take it if you are pregnant, could become pregnant, are breastfeeding, have a bleeding disorder, or take blood thinners, and tell your doctor, dentist, and pharmacist that you use it. Stop and seek medical advice if you notice unusual bruising, nosebleeds, bleeding gums, blood in urine or stool, or black stools. Never use it to delay or replace cancer treatment or treatment for an infection. Multi-ingredient tonics deserve extra caution: the Seasilver case involved a blend of pau d’arco with aloe vera and other ingredients sold with sweeping disease claims.