Jiang Huang

367 patients with primary knee osteoarthritis and a pain score of 5 or higher were randomized to ibuprofen 1200 mg/day or Curcuma domestica extracts 1500 mg/day for 4 weeks. Turmeric extract was non-inferior to ibuprofen on WOMAC total, pain and function scores at week 4, with the stiffness subscale showing only a trend. Adverse event rates were similar overall but abdominal pain and discomfort were significantly more frequent with ibuprofen. Note that Curcuma domestica is a synonym of Curcuma longa, so this trial is on the same species as Jiang Huang, though it uses a concentrated extract rather than a decoction.

A Bayesian network meta-analysis of randomized controlled trials of curcumin, alone and in combination, for knee osteoarthritis, searched to April 2023 and analysed with random-effects models. It pools the trial evidence for the one indication where curcuminoid preparations have been most extensively tested in patients.

Ten adjudicated cases of turmeric-associated liver injury from the prospective US DILIN cohort enrolled between 2004 and 2022, all since 2011 and six since 2017. Eight were women, median age 56. Injury was hepatocellular in nine and mixed in one; five patients were hospitalised and one died of acute liver failure. Turmeric was chemically confirmed in all seven products tested and three also contained piperine. Seven of ten carried HLA-B*35:01, two homozygously, an allele frequency of 0.450 versus 0.056 to 0.069 in controls. Latency was one to four months.

A randomized placebo-controlled six-way crossover in eight healthy volunteers given 4 g of standardised curcuminoids plus 24 mg piperine, or placebo, four times over two days before midazolam (CYP3A probe), flurbiprofen (CYP2C9 probe) or paracetamol (dual UGT and SULT probe). No meaningful changes were seen in Cmax, AUC, clearance, half-life or metabolite levels for any probe, and no effect on midazolam pharmacodynamics. Unconjugated curcuminoid concentrations stayed below 0.05 to 0.08 micromolar and piperine below 0.6 micromolar. The authors concluded short-term use of a piperine-enhanced curcuminoid preparation is unlikely to cause a clinically significant CYP3A, CYP2C9 or paracetamol-conjugation interaction.

Curcumin inhibited human BCRP in vitro with a Ki of 0.70 plus or minus 0.41 micromolar. In eight healthy Japanese subjects, 2 g of curcumin tablets increased the plasma AUC of sulfasalazine 2.0-fold after a 100 microgram microdose and 3.2-fold after a therapeutic 2 g dose. In mice curcumin raised sulfasalazine AUC eightfold in wild-type animals but not in Bcrp-null animals, confirming the transporter as the mechanism. This is the clearest demonstration that curcumin causes a clinically meaningful interaction at the intestinal wall despite its negligible systemic bioavailability.

A two-phase randomized single-blind crossover study in 18 healthy male volunteers genotyped for ABCB1 C3435T. Curcumin 1000 mg once daily for 14 days increased talinolol AUC(0-48h) by 67.0 percent and AUC(0-infinity) by 80.8 percent, significantly raised Cmax and reduced apparent oral clearance by 25.9 percent, without change in tmax or half-life. The effect was greatest in TT subjects, indicating that the interaction is mediated by P-glycoprotein and is genotype dependent.

Curcumin has iron-chelator properties, modulating proteins of iron metabolism and reducing spleen and liver iron content. In mice with DSS-induced colitis maintained on an iron-sufficient diet, curcumin induced mild anaemia. The authors flag that this may contribute to anaemia of inflammation and iron deficiency in inflammatory bowel disease, which is one of the settings where curcumin supplementation is most often recommended.

Curcumin inhibited platelet aggregation induced by epinephrine, ADP, platelet-activating factor, collagen and arachidonic acid, preferentially inhibiting the PAF- and arachidonic-acid-mediated pathways with IC50 values of 25 to 20 micromolar; much higher concentrations were needed against the other agonists. Inhibition of thromboxane formation and of calcium signalling were the proposed mechanisms. These concentrations are far above the unconjugated plasma levels achievable by oral dosing, which is why this widely cited antiplatelet effect has not translated into a clinical bleeding signal.