Vitamin B6 (Pyridoxine) for Primary Hyperoxaluria

Vitamin B6 (pyridoxine) is the only vitamin with a guideline-backed role in primary hyperoxaluria (PH), but only in type 1 (PH1) and only for some patients. Pyridoxine lowers urine oxalate in a subgroup of people with PH1, and response, defined as a fall of more than 30%, is most often seen in carriers of the Gly170Arg and Phe152Ile variants of the AGXT gene (Groothoff 2023). In the first prospective trial, 6 of 12 patients reached that threshold (Hoyer-Kuhn 2014); estimates of responders range from about a third (Fargue 2013) to about half (Monico 2005, Am J Nephrol).

European expert guidelines recommend starting pyridoxine in everyone suspected to have PH and in everyone with genetically proven PH1, at a maximum of 5 mg/kg a day (Groothoff 2023). That is a medicine-level dose set by a specialist: for a child weighing 30 kg it is 150 mg a day, several times the upper limit for children (40 mg a day at 4-8 years, 60 mg at 9-13) and above the 100 mg adult limit; those limits don't apply to medical treatment, but such patients should be under a doctor's care (NIH ODS). The verdict is mixed because the answer depends on genotype, with strong biochemical responses in some people, little or none in others, and no randomised trial behind any of it.

PH1 is caused by a deficiency of alanine-glyoxylate aminotransferase (AGT), a liver enzyme that depends on pyridoxal phosphate, the active form of vitamin B6 (Fargue 2013). Without enough working AGT the liver can't clear glyoxylate properly and overproduces oxalate, which forms calcium oxalate stones and deposits in the kidneys (Gupta 2022).

In lab-grown cells carrying the most common responsive variant, Gly170Arg, pyridoxine raised the amount of AGT, its activity and its import into peroxisomes, the compartment where it works. It acted both as a cofactor and as a chemical chaperone that helps the faulty protein reach the right place; Phe152Ile showed a partial effect (Fargue 2013). In patients, Gly170Arg and Phe152Ile, two variants that mistarget the enzyme, are the ones linked to B6 response (Monico 2005, Am J Nephrol).

Genotype guides expectations but doesn't settle them. In the prospective trial, 2 of 4 patients without either common variant still responded (Hoyer-Kuhn 2014), and the guideline notes that other non-truncating variants such as Gly41Arg may respond partly (Groothoff 2023), even though Gly41Arg showed no effect in the cell model (Fargue 2013). That is why guidelines recommend testing every PH1 patient for response rather than relying on genotype alone (Groothoff 2023).

A second cell study found that pyridoxine can partly work against itself: its phosphate form competed with the true cofactor and formed an inactive enzyme complex, while two other B6 forms, pyridoxamine and pyridoxal, rescued the faulty variants better (Oppici 2015). These are laboratory findings, and the authors called for those forms to be investigated as treatments (Oppici 2015).

Doses in PH1 are weight-based and far above nutritional intakes. The guideline suggests a maximum of 5 mg/kg a day and higher doses only in selected patients under close monitoring, because long-term high-dose pyridoxine is potentially neurotoxic (Groothoff 2023). That ceiling rests on a retrospective review of 23 patients in which, among the 14 Gly170Arg carriers, oxalate reduction was similar at 1-4.9, 5-9.9 and 10-14.9 mg/kg a day; its authors concluded that 5 mg/kg appeared sufficient (Monico 2005, Kidney Int).

The prospective trial started at 5 mg/kg a day and rose by 5 mg/kg every 6 weeks to 20 mg/kg by week 24 (Hoyer-Kuhn 2014). Three patients who came off dialysis took a median of 8.8 mg/kg a day (Lorenz 2021). At the other end, two patients in a 1985 report normalised their urine oxalate on 25 mg a day (Yendt 1985). Once someone responds, the guideline recommends tapering gradually to the lowest dose that keeps urine oxalate optimally reduced, judging each dose on at least two urine samples (Groothoff 2023).

Vitamin B6 had been used in PH1 for more than 40 years before anyone tested it prospectively. In that trial, 12 patients with genetically confirmed PH1 and normal kidney function took 5 mg/kg a day, rising every 6 weeks to 20 mg/kg by week 24. Urine oxalate fell on average 25.5%, from 2.09 to 1.52 mmol/1.73 m² a day (P = 0.01), while blood B6 rose from 22.5 to 1,217 ng/mL (Hoyer-Kuhn 2014).

Six of 12 reached the 30% response threshold: all 3 homozygous for Gly170Arg, 1 of 5 heterozygous for Gly170Arg and/or Phe152Ile, and 2 of 4 with neither variant. No one reached complete biochemical remission, not even the homozygous patients, and there was no control group (Hoyer-Kuhn 2014).

In a retrospective review of 23 patients, urine oxalate on B6 became normal in the 6 with two copies of the c.508G>A (Gly170Arg) variant (P < 0.001), fell partly in the 8 with one copy (P < 0.001), and did not change significantly in the 9 without it (P = 0.06) (Monico 2005, Kidney Int). A second analysis of 23 patients by the same group linked response to both Gly170Arg and Phe152Ile (Monico 2005, Am J Nephrol).

In the Dutch national cohort, 33 genotyped patients showed that homozygous Gly170Arg and Phe152Ile were associated with pyridoxine responsiveness and preserved kidney function over time when treatment started on time (van Woerden 2004). In the Rare Kidney Stone Consortium registry, among 83 children with PH1, those homozygous for Gly170Arg had the lowest urine oxalate after starting B6 and those without Gly170Arg the highest, though the authors note homozygous children may simply have milder disease; kidney function over time did not differ between the groups (Sas 2024).

Kidney failure is usually irreversible in PH1, but three patients homozygous for Gly170Arg came off dialysis after a median 10 months (range 5-19) of haemodialysis once pharmacologic-dose pyridoxine (median 8.8 mg/kg a day) was started or resumed, and kept their kidney function over a median 3.2 years (Lorenz 2021). Five similar patients who received a kidney-only transplant stayed on pyridoxine; over a median follow-up of 8.5 years (range 0.2-13.9), four grafts were still working at last check, and one failed at 13.9 years after an acute illness (Lorenz 2014).

The guideline cites registry data from 267 transplanted PH1 patients: combined liver-kidney transplant gave better event-free survival than kidney-only transplant in pyridoxine-insensitive patients (P < 0.001) but not in pyridoxine-sensitive ones (P = 0.411). It supports kidney-only transplant in selected patients whose urine oxalate normalises or nearly normalises on pyridoxine (Groothoff 2023).

No randomised, placebo-controlled trial of B6 exists, the studies are small and mostly from a few specialist groups, and most measure urine oxalate rather than stones or kidney survival. A long-term study of 25 patients treated with orthophosphate plus pyridoxine found 74% free of end-stage kidney disease at 20 years (Milliner 1994), but with no control group and two treatments given together, pyridoxine's share can't be separated. The rationale and guideline advice concern PH1; in a 1985 report, two patients with another type of PH didn't respond to 2 mg a day and improved only moderately at 25 or 200 mg (Yendt 1985).

B6 sits within conservative care that also includes high fluid intake and crystallisation inhibitors such as citrate (Cochat 2012; Groothoff 2023). For genotypes that predict no B6 response, RNA interference therapy is indicated, and partial responders may also qualify if their oxalate stays high (Groothoff 2023). In a phase 3 trial, lumasiran cut urine oxalate by 65.4% over 6 months, 53.5 percentage points more than placebo (Garrelfs 2021), and in a trial in PH1 or PH2, 64.7% of the PH1 participants on nedosiran reached normal or near-normal urine oxalate versus none on placebo (Baum 2023). See the main primary hyperoxaluria page for the wider picture, and vitamin B6 for kidney stones for ordinary kidney stones.

At PH1 doses the main concern is nerve damage. Taking 1-6 g of pyridoxine a day for 12-40 months can cause severe, progressive sensory neuropathy with loss of coordination (NIH ODS). Seven adults on high doses developed ataxia and severe sensory dysfunction and all improved after stopping (Schaumburg 1983), but three people in their 80s who took 600 mg a day for 3-10 years had not improved 2 years after stopping (Kulkantrakorn 2014). Volunteers given 1 or 3 g a day developed symptoms sooner at the higher dose, and symptoms kept progressing for 2-3 weeks after stopping (Berger 1992). See peripheral neuropathy.

That is why the guideline caps the usual dose at 5 mg/kg a day and asks for close monitoring above it (Groothoff 2023). The US upper limits are 30-80 mg a day for children aged 1-18 and 100 mg for adults, the same in pregnancy and breastfeeding (80 mg under 19, 100 mg from 19); they don't apply to medical treatment, but people taking B6 as treatment should be under a doctor's care (NIH ODS). Weight-based PH1 doses usually exceed them, and anyone with PH1 who is pregnant or breastfeeding should review the dose with the PH team. The PH studies did not report on nerve side effects, so the safety picture at these doses comes from the general data above. PH1 often leads to kidney failure (Gupta 2022), so for anyone with chronic kidney disease or on dialysis, B6 decisions belong with the kidney team.

Pyridoxine at 200 mg a day for 12-120 days can lower blood levels of phenytoin and phenobarbital, and some antiepileptic drugs (valproic acid, carbamazepine, phenytoin) speed up B6 breakdown (NIH ODS). Anyone on these medicines needs their neurologist and PH team to coordinate. In PH, high-dose vitamin C is best avoided because the body turns it into oxalate (ERKNet). Count B6 from a B-complex or multivitamin towards the total so the prescribed dose isn't exceeded.