Adenine Phosphoribosyltransferase (APRT) Deficiency

2,8-dihydroxyadeninuria: radiolucent stones and 2,8-dihydroxyadenine crystal nephropathy, a treatable cause of chronic kidney disease.

Literature and regulatory information updated: 6 November 2026.

0. Reference Summary Table

1. Introduction

1.1. Definition

Adenine phosphoribosyltransferase (APRT) deficiency is an autosomal recessive disorder of purine metabolism in which the absence of functional activity of the enzyme that salvages adenine diverts this base to oxidation by xanthine oxidoreductase (XOR), with the formation of 2,8-dihydroxyadenine (2,8-DHA). This compound is very poorly soluble in urine and precipitates as crystals and stones, causing radiolucent lithiasis, episodes of acute kidney injury (AKI) and a crystalline tubulointerstitial nephropathy that progresses to chronic kidney disease (CKD) and kidney failure when untreated [2, 3, 8]. The disease is expressed only in the kidney and urinary tract, with normal intelligence and immunity [2, 5]. Because there is an effective and accessible pharmacological treatment, XOR inhibition, diagnostic delay is the main avoidable determinant of loss of kidney function [8, 10].

1.2. Historical milestones

The first observations were biochemical. In 1968, Kelley and colleagues described partial APRT deficiency in four asymptomatic individuals from three generations of one family, with 21–37% of normal activity, and Henderson and colleagues confirmed in 1969 the autosomal transmission of the enzyme variants [1, 8]. In the following years, Fox (1973), Delbarre (1974) and Emmerson (1975) reported families with partial deficiency associated with gout or hyperuricaemia, but segregation analysis showed that hyperuricaemia and partial deficiency were transmitted independently, and these associations were not upheld as causal [1]. Complete deficiency with 2,8-DHA lithiasis was described in France by Cartier, Hamet and Hamburger in 1974, and in 1976 came the work of Simmonds and colleagues, who identified 2,8-DHA as a new component of urinary stones, and of Debray and colleagues, in a child with lithiasis whose parents had partial deficiency [1, 5]. Van Acker and colleagues published in 1977 the family of two siblings with complete deficiency, one of whom had passed 2,8-DHA gravel since birth; allopurinol combined with a low-purine diet stopped stone formation, and the authors showed that homozygotes are identified by elevated urinary adenine and absent erythrocyte activity [1]. The third paediatric case, in a child of consanguineous Arab parents, was published by Barratt and colleagues in 1979 [1, 5].

Recognition in adults came with Gault and colleagues (1981), in a woman from Newfoundland with first symptoms at 42 years, who drew attention to stones reported as uric acid by routine chemical tests and correctly identified by infrared spectroscopy or X-ray diffraction, and to the possibility of initial presentation with kidney failure [1]. In 1984, Kishi and colleagues counted only ten published cases of complete deficiency and documented that the enzyme deficiency in leucocytes does not alter immune or phagocytic function [1]. In 1988, Glicklich and colleagues reported a patient from Bermuda diagnosed 23 years after the first colic and only after recurrence of 2,8-DHA stones in the kidney graft [1]. In Japan, Kamatani and colleagues (1987) studied 19 families and found that in 15 (79%) the patients had only partial deficiency in lysates, unlike non-Japanese cases, all with complete deficiency [1]. Iceland contributed 8 of the 62 type I homozygotes known in 1992, all identified through the observation of round reddish-brown urinary crystals by attentive laboratory technicians [1].

1.3. Molecular discovery

The APRT locus was assigned to chromosome 16 by somatic cell hybridisation in 1974, with confirmation by dosage effect in trisomy 16 cells, and localised by Fratini and colleagues (1986) to 16q24, distal to haptoglobin and the fragile site FRA16D [4, 5]. The protein sequence was determined in 1986 and the complete genomic sequence in 1987 [4, 8]. Also in 1987, Hidaka and colleagues identified the first pathogenic variants, in a Belgian patient compound heterozygous for a three-nucleotide deletion and for the insertion of a thymine at the intron 4 donor site, which turned out to be the most frequent European variant [4]. The molecular basis of the Japanese allele APRT*J (p.Met136Thr) was described in 1988, the Icelandic variant p.Asp65Val in 1990 and, in 1992, Kamatani and colleagues showed that three alleles account for almost all defective Japanese alleles [1, 4].

1.4. Nomenclature and classifications

The distinction between type I (no activity in erythrocyte lysates) and type II (residual activity of 10–25% in lysates, with sigmoid kinetics and lower affinity for phosphoribosyl pyrophosphate, PRPP) was proposed on the basis of in vitro assays [5]. When it was shown that activity in intact cells is below 1% in both types and that the clinical manifestations are identical, the division became of historical interest only [1, 2]. Type II essentially corresponds to the Japanese allele p.Met136Thr, and type II homozygotes have only been found in Japan [5]. The current designations are “APRT deficiency” and “2,8-dihydroxyadeninuria”; Orphanet classifies the entity as a rare genetic nephropathy secondary to a disorder of purine metabolism [3]. The disease belongs to the group of purine disorders with lithiasis and CKD, alongside PRPS1 superactivity, the xanthinurias, HPRT deficiency, renal hypouricaemia and autosomal dominant tubulointerstitial kidney disease associated with UMOD [14]. There are no clinical subtypes or allelic forms: no other phenotype is associated with APRT variants [2].

2. Molecular Genetics and Pathophysiology

2.1. The APRT gene

The APRT gene spans about 2.5 kb (2466 bp) at 16q24.3, at GRCh38 coordinates 16:88,809,339–88,811,928, and is the most telomeric gene on chromosome 16; it contains five exons and a 540-bp coding region [4, 5, 8]. The order of markers on the long arm is centromere–FRA16B–HP–FRA16D–APRT–telomere, and the GALNS gene, whose variants cause mucopolysaccharidosis IVA, lies immediately centromeric, with the same transcriptional orientation [4]. The promoter has the features of a housekeeping gene: it lacks TATA and CCAAT boxes and has five GC boxes, potential binding sites for the Sp1 factor; CpG dinucleotides are conserved between human, mouse, hamster and Escherichia coli beyond what would be expected by chance [4, 5]. Expression is constitutive in all tissues and subject to little regulation [4, 8]. The MANE Select reference transcript is NM_000485.3, which encodes NP_000476.1 [6].

2.2. The enzyme

APRT catalyses the magnesium-dependent reaction between adenine and 5-phosphoribosyl-1-pyrophosphate (PRPP), which forms adenosine 5'-monophosphate (AMP); together with hypoxanthine-guanine phosphoribosyltransferase (HPRT) it is one of the enzymes of the purine salvage pathway, and it is the only salvage route for adenine [2, 5]. The protein is a homodimer with subunits of about 19.5 kDa; the gene encodes 180 amino acids and the sequenced mature protein has 179 residues [4, 5]. The crystal structure of the recombinant enzyme bound to AMP shows nine β-sheets and six α-helices organised into three domains: a core containing the PRPP-binding motif, an adjacent flexible loop with a probable catalytic role and a variable region dedicated to base recognition [2]. Sequence identity with mouse and hamster is 82% and 90% [4]. The enzyme accepts analogues such as 2,6-diaminopurine and 8-azaadenine, which become toxic after conversion to nucleotides; deficient cells are therefore resistant to 2,6-diaminopurine and 6-methylpurine and do not grow in azaserine-alanosine-adenine medium, a property used in cellular studies [5]. Normal activity in erythrocyte lysate is 16–32 nmol/h per mg of haemoglobin [2, 5].

2.3. Biochemical mechanism

Body adenine derives mainly from endogenous synthesis, in particular the polyamine pathway, and only a small fraction is of dietary origin [4, 8, 14]. Without APRT, unsalvaged adenine is oxidised by XOR, first to 8-hydroxyadenine and then to 2,8-DHA [8]. The solubility of 2,8-DHA in water is below 3 mg/L, but urine can become supersaturated; adenine and 2,8-DHA are excreted in a ratio of 1:1.5, 8-hydroxyadenine in smaller amounts, and urinary 2,8-DHA concentrations of up to 80 mg/L (0.5 mmol/L) have been measured [5]. Unlike uric acid, the solubility of 2,8-DHA does not increase with alkalinisation across the whole physiological pH range, which explains the ineffectiveness of citrate and bicarbonate and the frequency of high urinary pH in patients with radiolucent stones [2, 5, 14]. This is the step that allopurinol and febuxostat block: by inhibiting XOR they reduce 2,8-DHA formation, while adenine continues to accumulate and is excreted in larger amounts, without known toxicity [5, 15]. In heterozygotes and in people without variants, 2,8-DHA is not detectable in urine or plasma [8, 16].

2.4. Mechanism of kidney injury

2,8-DHA forms stones in the urinary tract and is deposited in the renal parenchyma, in the tubular lumens, in the cytoplasm of epithelial cells and in the interstitium, with variable degrees of inflammation and fibrosis [2, 8]. The injury is not only mechanical. In mice fed an adenine-rich diet, which reproduces 2,8-DHA nephropathy, the injury depends on crystal size: the smallest are endocytosed by tubular cells, large ones obstruct entire tubules, and medium-sized ones trigger a repair process called “extratubulation”, in which tubular cells and macrophages engulf the crystals and transfer them to the interstitium, where they are degraded by granulomatous inflammation [17]. Deletion of Tnfr1 reduced tubular expression of CD44 and annexin 2 and inflammation, attenuating the disease, whereas deletion of Tnfr2, Cd44 or Ahsg had no effect; patient biopsies showed similar morphology, crystal clearance and injury [17]. The NLRP3 inflammasome, implicated in oxalate nephropathy, has also been associated with 2,8-DHA injury [8]. Kidney tissue studies from five Icelandic patients showed that inflammation, tubular atrophy and fibrosis accompany the crystal burden, with absence of APRT on immunohistochemistry, reduced N-cadherin and increased collagen III [18]. In vitro, 2,8-DHA reduced the viability and migration of tubular cells in a dose-dependent manner, decreased transepithelial resistance and activated the TNF-α and mTORC1 pathways, with increased CD44 adjacent to the crystals [19]. These data are preclinical and point to possible targets, with no therapeutic translation to date. The Aprt-null mouse obtained by homologous recombination excretes adenine and 2,8-DHA crystals and develops tubular dilatation, necrosis, inflammation and fibrosis with strain-dependent severity [4].

2.5. Inheritance

Transmission is autosomal recessive: patients are homozygous or compound heterozygous, and heterozygotes, with 21–37% of normal activity in lysates (about 25% in haemolysate and 46% in lymphoblasts or fibroblasts), have no disease or biochemical abnormalities throughout life [1, 5, 8]. Older reports of disease in heterozygotes were not confirmed; consanguinity is frequent in some series, and pseudodominance was described in a Japanese family in which the affected father was married to a first cousin [1, 5]. A single exception has been reported: 2,8-DHA lithiasis in a Japanese patient heterozygous for p.Met136Thr, with reduced overall activity and affinity for PRPP; there has been no subsequent confirmation of disease in heterozygotes [4]. For counselling purposes, inheritance should be treated as strictly recessive.

2.6. Variant spectrum

The systematic search by Runolfsdottir and colleagues, up to February 2020, identified 62 confirmed pathogenic variants in 482 cases from 33 countries, of which 311 (64.5%) had a molecular diagnosis; 28 were missense and the rest nonsense, insertions or deletions, frameshift, start codon loss or splice variants; 37 variants were seen in a single patient [7]. In 2026, the same group reported 68 pathogenic variants, all abolishing enzyme function in vivo [8]. Three variants account for 368 of the 576 causal alleles (63.8%) and 58% of the cases with a molecular diagnosis [7, 8]. The order of frequency and the geography of each are given in Table 2.1; conversions between historical and HGVS nomenclature were confirmed in VariantValidator.

Table 7 of GeneReviews contains two misalignments: the designation “Arg87PfsTer23” corresponds to the CCGA duplication (c.258_261dup), not to c.400+2dup, and the comment on reduced affinity for PRPP refers to p.Met136Thr, not to p.Trp98Ter [2, 6]. Historical nomenclature numbered nucleotides according to the old genomic sequence, so the 4-digit numbers cannot be converted to c. by hand [4, 6].

2.7. Variants of uncertain significance and hypomorphic variants

Two missense variants have been described in the heterozygous state in individuals with partial activity similar to that of carriers, c.266G>A p.(Arg89Gln) in Australia and c.346G>A p.(Ala116Thr) in China; both are relatively frequent in gnomAD (0.41% in South Asians and 0.23% in East Asians) and, if they were pathogenic, would imply about 20,000 and 5000 homozygotes, none of whom has been diagnosed, so they are not classified as causal and, in the homozygous state, would be expected to give a mild phenotype or none [7]. The variant c.362A>G p.(Gln121Arg) has a frequency of 0.5% in Africans in gnomAD, with two homozygotes without known disease, and requires functional study before any classification [7]. For novel variants in the homozygous or compound heterozygous state, measurement of enzyme activity or 2,8-DHA is recommended, because measurement in extracts may not suffice to establish functional significance [2, 8].

2.8. GALNS–APRT contiguous gene syndrome

The physical proximity between GALNS and APRT allows deletions that remove both genes. In a Czech patient with Morquio A syndrome (mucopolysaccharidosis IVA) and 2,8-DHA lithiasis, Wang and colleagues identified a deletion of about 100 kb with breakpoints in intron 2 of GALNS and intron 2 of APRT, sparing CYBA; and Fukuda and colleagues described a Japanese patient with a submicroscopic deletion of GALNS and APRT on one chromosome and the GALNS variant p.Arg386Cys on the other [4, 5]. The text of Nyhan's Atlas attributes the R386C variant to APRT, but the title of the original article and OMIM place it in GALNS [4, 5]. When the contiguous deletion is on one allele and there is a point variant in APRT on the other, the patient will have APRT deficiency with or without mucopolysaccharidosis, depending on the status of the second GALNS allele.

2.9. Genotype–phenotype correlations, penetrance and modifiers

There are no established genotype–phenotype correlations: patients with the same variant, including siblings, differ greatly in age at onset and severity [2, 8]. The two sisters homozygous for p.(Leu110Pro), one with lithiasis and the other without disease, and the 15–20% of asymptomatic homozygotes with abnormal 2,8-DHA excretion illustrate the incomplete penetrance of the clinical manifestations, although 2,8-DHA excretion is present in all [2, 4]. The modifiers identified are extrinsic: hydration, volume depletion, diagnostic delay, and adherence to and dose of the XOR inhibitor [10, 21]. A curious observation results from genetic linkage: 15 of the first 23 Icelandic patients (65%) were red-haired, compared with 4 of 27 unaffected siblings (15%), which is attributed to the proximity of MC1R, also at 16q24.3, and not to an effect of the disease [8]. No allelic disorders have been described [2].

3. Clinical Spectrum and Manifestations by System

3.1. Overview

The clinical picture depends entirely on 2,8-DHA excretion and its tendency to precipitate in the urinary tract and renal parenchyma [5]. Severity ranges from absence of symptoms to kidney failure, and symptom onset has been recorded from birth to 74 years [5]. In the GeneReviews synthesis, with more than 400 published patients, the renal presentation includes lithiasis in 60–90%, CKD in more than 50% of adults, AKI in about 30% and end-stage kidney disease in 15% already at diagnosis [2]. At least half of patients only have symptoms in adulthood, about 35% are diagnosed before 18 years of age and, without adequate treatment, 20–25% reach end-stage kidney disease, generally in adulthood [2]. About 15% of patients are asymptomatic and identified through family screening or crystals observed on routine urinalysis [2, 8]. Table 3.1 summarises the main cohorts; all are retrospective or registry-based, with heterogeneous treatment and different populations, so the figures are not directly comparable.

3.2. Urinary lithiasis

Lithiasis is the most common manifestation, in children and adults [2, 8]. Patients present with renal colic, haematuria, dysuria, passage of gravel or small stones, complicated urinary tract infection with fever and, sometimes, urinary retention [5, 14]. The stones are radiolucent, not visible on plain radiography, although occasional admixture with calcium makes them radiopaque; they are reddish-brown, round and friable when wet and greyish when dry [1, 5]. They are frequently bilateral and recurrent [25]. Without treatment, most symptomatic patients have recurrences, often multiple; in the RKSC registry, 18 of the 33 patients with stones (55%) had a recurrence, and five patients who only started allopurinol in adulthood had accumulated 28 episodes [10, 21]. Lower urinary tract symptoms (straining, hesitancy, dribbling, incomplete emptying) and abdominal pain are part of the complaints [2].

3.3. Acute kidney injury

AKI occurred in 34% of patients in the RKSC registry, a figure the authors considered new compared with previous series [8, 10]. The causes were biopsy-proven crystal nephropathy (7 patients), obstruction by stones (9) and volume depletion (2); two of the 16 patients with AKI before diagnosis required transient haemodialysis [10]. More than half (10 of 18) progressed to stage 3–5 CKD at a median of 1.6 years after the episode, which makes AKI a marker of poor prognosis and an opportunity for diagnosis [10]. In children, AKI due to bilateral obstruction by 2,8-DHA stones is well documented, even in infants, and may reverse [1, 5, 8].

3.4. Crystal nephropathy and chronic kidney disease

CKD results from tubulointerstitial nephropathy due to 2,8-DHA crystals and is present in more than half of adults at diagnosis [2]. One of the most relevant findings of recent cohorts is the frequency of advanced CKD without a history of lithiasis: in the RKSC registry, advanced CKD without stones was more prevalent than had been assumed, and the diagnosis sometimes comes only with biopsy [8, 10]. Progression without treatment is rapid: the median eGFR slope was −5.74 mL/min/1.73 m² per year in patients untreated before stage 5, compared with −0.38 in those treated with an XOR inhibitor (P=0.001) [10]. None of the 11 patients already at stage 5 at diagnosis had received an inhibitor [10]. The prevalence of advanced CKD was lower in Icelandic patients, diagnosed earlier, than in those from other countries, which the authors attribute to lower familiarity with the disease outside Iceland [10]. On histology, 2,8-DHA nephropathy is easily confused with oxalate nephropathy, particularly in primary hyperoxaluria, and with urate nephropathy [2, 8].

3.5. Manifestations in children

In the two paediatric studies, French and Icelandic, lithiasis was the presenting form in about 80% and 50% of children, respectively, and 3 children in each study (14%) had asymptomatic crystalluria [8, 21, 22]. In the Icelandic series no child had stage 3 or higher CKD at presentation, but, owing to diagnostic delay, 4 of 21 had already reached that stage when the diagnosis was made, one of them on kidney replacement therapy [8]. Reddish-brown stains in the nappy, recurrent urinary tract infections and haematuria are characteristic manifestations of early childhood [8, 23]. In Iceland, 3 of the first 23 patients were infants with a history of nappy stains [23]. Figure 66.2 of Nyhan's Atlas shows an 18-month-old child who had been passing stones since birth and who, at 24 years, was a healthy adult [5].

3.6. Manifestations in adults

In adults, CKD and AKI predominate, frequently with a long history of misinterpreted colic [8, 10]. The median diagnostic delay was 7.5 years (0.4–47.9) after the first symptomatic stone episode or after detection of elevated creatinine; contributing to this delay were radiolucent stones taken to be uric acid (4 patients), histological confusion with other crystal nephropathies (6) and unidentified urinary crystals (17) [10]. The median age of patients with stage 3–5 CKD at diagnosis was 44.5 years (11.9–67.9) and six were already on kidney replacement therapy before diagnosis, four with graft recurrence [10]. The case described in section 10, a 67-year-old man with an eGFR decline from 100 to 20 mL/min/1.73 m² over three and a half years without any colic, illustrates this form [20].

3.7. Recurrence in the kidney graft

2,8-DHA nephropathy recurs in the transplanted kidney and is frequently the circumstance in which the diagnosis is made [2, 8]. In the Franco-Icelandic series of Zaidan and colleagues, in nine patients whose diagnosis had been missed before transplantation, recurrence was diagnosed at a median of 5 weeks (1.5–312) after transplantation, with delayed graft function in 2, acute-on-chronic dysfunction in 5, acute dysfunction in 1 and normal function in 1; with allopurinol, function improved in 7, remained stable in 1 and worsened in 1, but at last follow-up two had lost the graft and five had chronic dysfunction [26]. In the international series of Runolfsdottir and colleagues (17 patients, 22 grafts), all 10 graft biopsies from patients without pretransplant treatment showed recurrence, compared with 3 of 8 in those treated (p=0.004) [27]. Even with an XOR inhibitor started early after transplantation, chronic injury may develop: in a 49-year-old Japanese man with recurrence at three months, high-dose febuxostat reduced crystals and inflammation, but the one-year biopsy showed expansion of the scarred area [28]. Two North American patients with early recurrence, one without a history of lithiasis, maintained stable function at two years with an XOR inhibitor and conversion of immunosuppression to belatacept, with a reduction in crystals on the one-year biopsies [29]. A Korean case of post-transplant recurrence was also published in 2026 [30].

3.8. Extrarenal manifestations

The disease has no neurological, immunological or growth manifestations; intelligence and immunity are normal [1, 5]. The Icelandic group observed various ocular complaints in a significant proportion of its patients, but only in patients on an XOR inhibitor, and the ocular symptoms led to discontinuation of treatment in some, without it being established whether they belong to the disease or to the drug [8, 10]. In 1986, two Belgian cases of corneal dystrophy were described in patients with APRT deficiency, attributed to crystalline deposits without histological confirmation [10]. GeneReviews recommends an ophthalmology consultation only in the presence of ocular symptoms [2]. In 2025, a 68-year-old man was presented at a congress in the United Kingdom with recurrent asymmetric dactylitis of the hands and feet treated as gout despite normal serum urate, who went on to have bilateral lithiasis, an eGFR of 30 mL/min and 2,8-DHA crystals on kidney biopsy; the abstract describes the joint picture as 2,8-DHA arthropathy, but does not mention demonstration of crystals in synovial fluid and reports the genetic result as heterozygous, so this is a preliminary observation, without confirmation that the joints are a site of deposition [31].

3.9. Natural history and treated disease

The natural history without treatment is one of stone recurrence, AKI and progressive loss of function, with 20–25% reaching end-stage kidney disease [2, 10]. With an XOR inhibitor started in time and at an adequate dose, the course is different: in children treated from 2.6 years of age for 18.9 years, only 2 had stones and the final eGFR was 114 mL/min/1.73 m²; all patients with stage 3–5 CKD at the end of follow-up had started treatment as adults [21]. In the French paediatric series, all treated children stabilised or improved their kidney function [22]. Non-adherence and insufficient dosing explained most recurrences on treatment [21].

4. Clinical Recognition and Red Flags

4.1. When to suspect

APRT deficiency should be considered in all children with renal colic, radiolucent stones or AKI, and in all infants with reddish-brown stains in the nappy [2, 8]. Recent reviews stress that the first manifestations vary even within a family with the same variants, from urinary tract infections and nappy stains to kidney failure, which explains late recognition [32]. In young and middle-aged adults, suspicion is mandatory in the presence of renal colic with radiolucent stones, unexplained progressive CKD and, above all, tubulointerstitial nephropathy of unknown cause on biopsy, even without a history of lithiasis [2, 8]. Two scenarios deserve particular attention. The first is the child or adolescent with a supposed uric acid stone, because these are very rare in childhood [1, 8]. The second is early dysfunction of a kidney graft in a patient with CKD of undetermined cause [2, 26]. A high urinary pH in a patient with radiolucent stones is a clue, because uric acid only crystallises in acid urine [2, 8]. The French PNDS (French National Diagnosis and Care Protocol) on lithiasis in children lists early onset, bilaterality, consanguinity and a history in siblings or parents as factors pointing to a genetic cause [11].

4.2. Common pitfalls

The first pitfall is to take radiolucency as proof of uric acid: the routine colorimetric reaction for uric acid and chemical stone analyses do not distinguish uric acid from 2,8-DHA, and the patient treated empirically with allopurinol improves, which perpetuates the error [1, 5]. The second is not requesting urine microscopy with specialist examination: in the RKSC registry, the crystals were not identified in 17 patients who had them [10]. The third is histological confusion with oxalate, which occurred in 6 patients of the same cohort [10]. The fourth is relying without confirmation on automated Fourier transform infrared spectroscopy (FTIR), which can give false-positive 2,8-DHA results [8]. The fifth is advanced CKD, in which urinary crystals become scarce owing to reduced renal clearance of 2,8-DHA [2, 8].

4.3. Urgent signs

Anuria or oliguria with bilateral loin pain, in a child or an adult with radiolucent stones, suggests bilateral obstruction by 2,8-DHA stones and requires immediate ultrasound, urological decompression and hydration [5, 8]. Acute dilatation of the urinary tract, pain that does not respond to the usual measures and signs of infection (fever, pyonephrosis) are severity criteria that justify urgent urological assessment [11]. A rise in creatinine without obstruction, with a sediment rich in brown crystals, suggests acute crystal nephropathy and is an indication to start an XOR inhibitor without waiting for molecular confirmation [8, 10]. In a kidney graft, any dysfunction in the first weeks in a patient with CKD of unknown cause should prompt a search for 2,8-DHA crystals in the urine and on biopsy [26, 27].

4.4. Presymptomatic screening

The siblings of a patient have a 25% probability of being affected and should all be tested, including those who are asymptomatic; the Icelandic group considers this screening mandatory [2, 8]. In the first Icelandic study, 5 of 19 patients were asymptomatic, three identified by routine urinalysis and two by family screening [8]. Newborn screening is not implemented in any country in the sources consulted; the Icelandic authors suggest that it could be feasible in Iceland and in other populations with a high carrier frequency [8].

5. Diagnosis: Criteria, Testing and Algorithms

5.1. Diagnostic criteria

The diagnosis is established in a proband with absent APRT activity in erythrocyte lysate or with biallelic pathogenic variants in APRT [2]. Enzyme activity was the original gold standard; today confirmation is usually by genetic testing, by a multigene lithiasis panel or by single-gene sequencing [8]. As 2,8-DHA is not detected in heterozygotes or healthy individuals, its demonstration in urine or plasma by ultra-performance liquid chromatography coupled with tandem mass spectrometry (UPLC-MS/MS) may also be considered diagnostic [8, 16]. Stone analysis, crystalluria and biopsy are suggestive but do not suffice to establish the diagnosis according to GeneReviews and the 2026 review, which require confirmation in all cases [2, 8]. Some authors consider 2,8-DHA crystals pathognomonic and confirmation unnecessary; this is not the position of the reference sources, especially since the description of false positives with FTIR [8, 11, 20].

5.2. Urine microscopy

Sediment microscopy is the most useful bedside test and should be requested explicitly. 2,8-DHA crystals are round, reddish-brown, with a dark outline and central spicules, and small and medium-sized ones show a central Maltese cross under polarised light; large, thick aggregates may not show it, as they are impermeable to light [2, 8]. Figure 2 of the review by Edvardsson and colleagues shows the same field under conventional and polarised light (https://pmc.ncbi.nlm.nih.gov/articles/PMC13481733/figure/Fig2/) [8]. Crystals are abundant in untreated non-oliguric patients, but may be scarce in advanced CKD and kidney failure [2, 8]. The French PNDS recommends fresh urine, preferably the whole first or second morning void; it notes that, unlike other crystals, 2,8-DHA crystals are always pathological regardless of sample storage, and that, when APRT deficiency is suspected, the crystalluria sample may be sent with a delay to a specialist laboratory [11]. Crystalluria assessed on a routine bacteriological examination does not replace a crystalluria study [11].

5.3. Imaging

2,8-DHA stones are not seen on plain abdominal radiography; the diagnosis of lithiasis requires ultrasound or computed tomography (CT) [2, 8]. Ultrasound frequently shows hyperechogenic kidneys, the “bright” kidney pattern [2, 14]. In children, ultrasound is the diagnostic and follow-up examination; CT is not indicated first-line, because its radiation dose is 2 to 10 times that of plain radiography, except in obese children [11]. Ultrasound or CT performed for another reason sometimes reveals stones in asymptomatic patients [2].

5.4. Stone analysis

Infrared or ultraviolet spectrophotometry, at acid and alkaline pH, and X-ray crystallography distinguish 2,8-DHA from uric acid and xanthine, the other radiolucent components; chemical and thermogravimetric methods do not and are not recommended [2]. Morphoconstitutional analysis, which combines stone morphology with FTIR, is the method advocated in the French PNDS, which follows the European Association of Urology (EAU) recommendation to analyse all stones from the first episode and at any age, and to repeat the analysis at each recurrence [11]. FTIR is now the most widely used method, but automated spectral interpretation has produced published false-positive 2,8-DHA results, so the result is only suggestive [8]. Stones may contain traces of other minerals [2].

5.5. Kidney biopsy

Biopsy shows diffuse deposits of 2,8-DHA crystals in the tubular lumens, in the cytoplasm of epithelial cells and in the interstitium, with inflammation and fibrosis, even in patients without lithiasis; the crystals are birefringent under polarised light [2, 8]. On haematoxylin-eosin and periodic acid-Schiff they appear greenish-brown, light blue on acid fuchsin-orange G staining and grey-black on silver [20]. Figure 3 of the 2026 review documents the intratubular, intracellular and interstitial distribution (https://pmc.ncbi.nlm.nih.gov/articles/PMC13481733/figure/Fig3/) [8]. The most important histological differential diagnosis is oxalate, also birefringent and in the same locations, and uric acid; as treatment differs, histological suspicion should be verified [2, 8]. In the graft, FTIR applied to the biopsy confirmed the diagnosis in all cases of the Zaidan series [26].

5.6. Enzyme activity

APRT activity in erythrocyte lysate is normally 16–32 nmol/h per mg of haemoglobin and is absent in almost all patients [2, 5]. There are exceptions with substantial activity in lysates: p.(Val150Phe), in some patients of Northern European ancestry, and p.(Met136Thr), present in more than 70% of Japanese patients; in these cases an intact-cell assay is needed, such as adenine uptake by erythrocytes or leucocytes [2]. After a recent transfusion, a normal result or one in the heterozygote range should be repeated three months later [2]. The assay in extracts does not reliably identify heterozygotes, whose activity overlaps with that of controls [2]. Nyhan's Atlas, with older data, recommended the enzyme assay as the test because of the fraction of alleles not identified by sequencing, and indicated that, in general, activity of at least 25% is not accompanied by symptoms [5].

5.7. Quantification of 2,8-DHA and adenine by UPLC-MS/MS

The Icelandic group developed a urinary 2,8-DHA assay by UPLC-MS/MS with an isotopically labelled internal standard, linear between 100 and 5000 ng/mL, with intra-day and inter-day coefficients of variation within ±15% and analysis in 6.5 minutes [33]. In untreated patients, 24-hour urinary excretion was 138 mg (64–292) and the 2,8-DHA/creatinine ratio in the first morning urine 13 mg/mmol (4–37), with a strong correlation between the two measures (rs=0.84, p<0.001); 2,8-DHA was detected in all samples from untreated patients and in none from heterozygotes or controls [16]. In 2024 a plasma assay that simultaneously measures 2,8-DHA, adenine, allopurinol, oxipurinol and febuxostat was validated according to FDA standards for bioanalytical methods [34]. In 122 paired plasma and urine samples from 26 patients, untreated plasma 2,8-DHA was 249 ng/mL (123–1315) and fell below the limit of detection with the highest inhibitor doses; in untreated patients, it correlated strongly with eGFR (rs=−0.74, p<0.0001) and not with the urinary ratio (rs=−0.26, p=0.14), which suggests that the plasma value better reflects the systemic burden, especially in CKD [35]. These assays are available in few laboratories; their diagnostic sensitivity and specificity in unselected populations have not been established in the sources consulted [2, 8]. The urinary purine profile by liquid chromatography with diode-array or mass spectrometric detection, performed in metabolic disease laboratories, shows increased adenine and 2,8-DHA [14].

5.8. Genetic testing

The recommended approach is APRT sequencing followed by targeted deletion/duplication analysis if only one variant or none is found, or a multigene lithiasis panel that includes APRT [2]. In the French series, sequencing of the coding regions and exon–intron junctions in 31 families identified 54 of the 62 alleles (87%); the remaining 13% may lie outside the sequenced regions or correspond to undetected mechanisms [2, 9]. Deletions are rare and include the contiguous GALNS–APRT deletion and the complex 254-bp deletion with an 8-bp insertion [2]. The proportion of alleles detected by sequencing in a series of patients with complete deficiency is not equivalent to the clinical sensitivity of the test in a population with suspected disease. In the RKSC cohort with suspected monogenic lithiasis (426 families, 657 patients), a panel of up to 160 genes solved 34% of families in 22 genes, with AGXT, HOGA1, SLC34A3, CYP24A1, SLC3A1 and CLCN5 the genes found in ten or more families; the diagnostic yield was higher below 16 years of age [36]. The 2026 review by the EAU Section of Endourology includes APRT among the monogenic lithiasis genes with targeted therapy and among the genetic causes with an increased risk of CKD [37]. The report should state the gene, the transcript with version (NM_000485.3), the HGVS nomenclature, the zygosity, the ACMG/AMP class and, where applicable, the phase of the variants confirmed by parental testing. For variants of uncertain significance in the homozygous or compound heterozygous state, enzyme activity or 2,8-DHA resolves the functional significance [8].

5.9. Algorithm

5.10. Initial assessment after diagnosis

GeneReviews recommends, if not already done: creatinine and/or cystatin C; testing for 2,8-DHA crystalluria and for albuminuria or proteinuria; renal ultrasound or CT to assess stone burden; kidney biopsy to be considered if there is reduced function and/or proteinuria; ophthalmology consultation if there are ocular symptoms; and clinical genetics or genetic counselling consultation [2]. In children, the PNDS adds growth assessment and screening for associated urological anomalies [11].

5.11. Prenatal and preimplantation diagnosis

Once the familial variants are known, prenatal diagnosis and preimplantation genetic testing are possible [2]. The existence of an effective treatment and the exclusively renal nature of the disease weigh in the discussion of these options, which is personal and should take place before pregnancy (section 8) [2].

6. Differential Diagnosis

6.1. Comparison table

6.2. Radiolucent stones: uric acid, xanthine and 2,8-DHA

The three main components of radiolucent stones are uric acid, xanthine and 2,8-DHA, and only spectrophotometry (or crystallography) reliably distinguishes them [2]. Normal serum and urinary uric acid and high urinary pH point to 2,8-DHA; hypouricaemia with hypouricosuria points to xanthinuria; and persistently acid urine, with or without hyperuricosuria, to uric acid [2, 11, 14]. The distinction is therapeutic: uric acid dissolves with alkalinisation, 2,8-DHA does not; and in xanthinuria allopurinol is not indicated, whereas in APRT deficiency it is the treatment [11]. The diseases with hyperuricaemia of enzymatic cause (HPRT deficiency and PRPS1 superactivity) and xanthinuria type III, due to molybdenum cofactor deficiency, are part of the differential.

6.3. Crystal nephropathies on biopsy

2,8-DHA nephropathy should be included in the differential diagnosis of any CKD or kidney failure with crystals on biopsy, even without lithiasis [2]. The most frequent confusion is with primary hyperoxalurias types 1, 2 and 3, because oxalate and 2,8-DHA crystals are both birefringent and are deposited in the same locations; the treatments currently available differ between the two diseases, and XOR inhibition is effective only in APRT deficiency [2, 8]. A brown or greenish-brown colour on haematoxylin-eosin, the colour on silver staining and concomitant crystalluria point to 2,8-DHA [20]. Urate nephropathy can also mimic 2,8-DHA nephropathy [1, 2]. APRT deficiency is on the list of diseases to look for in the genetic approach to adult CKD of undetermined cause.

6.4. Other monogenic causes of lithiasis in children

In a child with recurrent lithiasis or nephrocalcinosis, the hereditary causes to be excluded in parallel are APRT deficiency, cystinuria, Dent disease, familial hypomagnesaemia with hypercalciuria and nephrocalcinosis, and the primary hyperoxalurias [39]. A lithiasis gene panel allows all of them to be studied simultaneously [2, 36]. 2,8-DHA stones are radiolucent, unlike calcium oxalate stones, and cystine gives characteristic hexagonal crystals [8, 38].

7. Treatment and Risk Management Strategies

7.1. Goals and principles

The aim of treatment is to suppress 2,8-DHA formation continuously and for life, in order to prevent new stones, crystal deposition in the parenchyma, AKI and progression of CKD, and to protect the graft when there is a transplant [2, 8]. The central measure is XOR inhibition with allopurinol or febuxostat, which should be maintained in all patients, symptomatic or not, including after transplantation [2]. Abundant hydration is an adjunct; urinary alkalinisation has no place; dietary purine restriction has shown no additional benefit when inhibition is adequate [2, 8, 11]. Treatment is effective even when started late: allopurinol and febuxostat can improve kidney function in patients with advanced CKD [2, 23]. No medicine has an approved indication for APRT deficiency, in the USA, the European Union, Portugal or Brazil; all pharmacological treatment is off-label use [40–43].

7.2. Xanthine oxidoreductase inhibitors

Allopurinol, a hypoxanthine analogue, and febuxostat, a selective non-purine inhibitor, block XOR, the enzyme that oxidises adenine to 8-hydroxyadenine and 2,8-DHA [8, 15]. Inhibition reduces 2,8-DHA in urine and plasma, often below the limit of detection, and increases the excretion of adenine, which is soluble: in one trial, urinary adenine rose from 32 mg/24 h to 96 mg with allopurinol 400 mg/day and 112 mg with febuxostat 80 mg/day, without adverse effects [15]. Treatment is monitored by urinary 2,8-DHA or, failing that, by crystalluria [2, 15].

7.3. Allopurinol

Allopurinol is the first-line drug, because of accumulated experience, cost and the regulatory restriction of febuxostat to second line in the USA [2, 41]. In Portugal it is marketed as Zyloric® (100 and 300 mg, Faes Farma), Zurim®, Ikynol® and generics; in the USA as Zyloprim® and generics; in Brazil as Zyloric® (Aspen Pharma) and generics [40, 43, 44].

Reference dose. GeneReviews recommends 5–10 mg/kg/day, up to a maximum of 800 mg/day, in one or two daily doses, without routine reduction in renal impairment [2]. This is the reference dose in this article. The 2026 review by the Icelandic group does not change this range and adds that, before transplantation, the dose should be at least 400 mg/day (section 7.11) [8]. The French PNDS recommends allopurinol combined with hyperdiuresis, without specifying a dose [11].

Evolution of doses and evidence. Older sources cite 5–10 mg/kg/day up to 200–300 mg/day in adults as sufficient to eliminate 2,8-DHA from the urine, and later 300 mg twice a day [5]. In the French paediatric series the median dose was 9 mg/kg/day [22]; in the children of the RKSC registry, 100 mg/day (25–200) and in adults 200 mg/day (100–300) [21]. In the RKSC registry, adults treated with 100–300 mg/day had 35 stone events in 18 patients (34%), most of them on 300 mg/day, which led the authors to propose doses of 600–800 mg/day and to note that the former recommendation not to exceed 200–300 mg/day in CKD, for fear of hypersensitivity syndrome, was contradicted by later studies [10]. In the transplant series, the authors consider that doses of 400–600 mg/day will be needed to minimise recurrence [27]. The 2025 trial showed that 800 mg/day reduces plasma and urinary 2,8-DHA more than 400 mg/day (section 7.5) [45]. There is no dose-finding trial with clinical outcomes; titration is based on suppression of 2,8-DHA or, where the assay is unavailable, on the disappearance of urinary crystals [2, 15].

Comparison with the labels. The United States allopurinol label, approved in 1966 (Zyloprim®), indicates it for gout, hyperuricaemia of cancer therapy and recurrent calcium oxalate stones with uricosuria above 800 mg/day (men) or 750 mg/day (women); in gout with normal kidney function the starting dose is 100 mg/day, increased by 100 mg per week up to a maximum of 800 mg/day; in renal impairment the starting dose is 50 mg/day, increased by 50 mg every 2–4 weeks; in paediatric oncology the dose is 100 mg/m² every 8–12 hours (10 mg/kg/day, maximum 800 mg/day); and for calcium oxalate stones 200–300 mg/day [40]. The Portuguese SmPC of Zyloric® recommends in adults starting with 100 mg/day and using 100–200 mg/day in mild situations, 300–600 mg in moderately severe situations and 700–800 mg in severe situations, with a maximum of 800 mg/day or, by weight, 2–10 mg/kg/day; under 15 years, 10–20 mg/kg/day up to 400 mg/day; in severe renal impairment, less than 100 mg/day or 100 mg at intervals longer than one day, aiming to keep plasma oxipurinol below 100 µmol/L; and, in haemodialysis two to three times a week, 300–400 mg immediately after each session [43]. The Saudubray drug table, for the diseases with hyperuricaemia including APRT deficiency, indicates 10–20 mg/kg/day in children and 2–10 mg/kg/day in adults, adjusting to the minimum dose that normalises uric acid, a criterion designed for hyperuricaemia and not for 2,8-DHA [46]. The GeneReviews dose lies within the weight-based range of the labels, but the recommendation not to reduce the dose in renal impairment diverges from the labels, which reduce it; in these patients the decision should weigh the risk of progressive crystal nephropathy against the risk of toxicity and hypersensitivity, with a low starting dose and slow titration guided by 2,8-DHA or crystalluria [2, 40, 43].

Adverse effects and precautions. The most frequent adverse effects (more than 1%) are nausea, diarrhoea and elevated liver enzymes [40]. The most serious is allopurinol hypersensitivity syndrome, with Stevens-Johnson syndrome, toxic epidermal necrolysis or drug reaction with eosinophilia and systemic symptoms (DRESS), concentrated in the first months of treatment; treatment should be stopped immediately in the presence of rash or any other sign of hypersensitivity [40, 43]. The HLA-B*58:01 allele increases the risk; its frequency is up to 20% in Han Chinese, 8–15% in Thai, about 12% in Koreans and 1–2% in Japanese and Europeans, and it is also more frequent in people of African and Pacific Islander ancestry [40, 43]. The FDA label does not recommend allopurinol in carriers unless the benefit is clear, and both labels advise considering genotyping before initiation in high-prevalence populations; CKD and thiazide diuretics increase the risk [40, 43]. In the Icelandic registry, ocular complaints led to discontinuation of the inhibitor in some patients; the relationship with the drug has not been established [8, 10].

7.4. Febuxostat

Febuxostat is the alternative when allopurinol is not tolerated, is contraindicated or does not suppress 2,8-DHA, and it was more effective than allopurinol in reducing 2,8-DHA in the two available trials [2, 15, 45]. It is marketed as Uloric® in the USA and Adenuric® in the European Union and Portugal (80 mg) [41, 42]. It has no registration in Brazil in the ANVISA open database consulted, so allopurinol is the only registered option there [44].

Reference dose. GeneReviews recommends 80 mg/day, without routine reduction in renal impairment, and there are no paediatric dose data [2]. The Icelandic group recommends at least 80 mg/day before transplantation [8]. In a report of paediatric transplantation at 12 years, switching from allopurinol to febuxostat after transplantation was followed by absence of recurrence on biopsy at four months; this is an isolated observation, outside the label indications [47]. In a patient allergic to allopurinol, 80 mg/day reduced crystalluria [5].

Comparison with the labels. In the USA, Uloric®, approved on 13 February 2009, is indicated only for chronic hyperuricaemia of gout in adults with an inadequate response to allopurinol titrated to the maximum dose, intolerance or contraindication; the dose is 40 or 80 mg once daily, limited to 40 mg in severe renal impairment [41]. The label has a boxed warning: in the CARES study, patients with gout and established cardiovascular disease treated with febuxostat had more cardiovascular deaths than those treated with allopurinol, sudden death being the most frequent cause (2.7%, 83 of 3098, compared with 1.8%, 56 of 3092) [41]. The United States label advises against use in secondary hyperuricaemia, for example in Lesch-Nyhan or in transplant recipients, owing to lack of studies, and states that paediatric safety and efficacy have not been established [41]. In the European Union, Adenuric® has been authorised since 21 April 2008, for chronic hyperuricaemia with urate deposition and for the prevention and treatment of hyperuricaemia in adults undergoing chemotherapy for haematological malignancies at intermediate to high risk of tumour lysis syndrome; in gout, the dose is 80 mg daily, with 120 mg if uric acid remains above 6 mg/dL after 2–4 weeks; no adjustment is needed in mild or moderate renal impairment and experience is insufficient with creatinine clearance below 30 mL/min; it is not established below 18 years of age [42]. The European SmPC mentions the excess of fatal cardiovascular events in CARES and non-inferiority to allopurinol in the FAST study, recommending caution and monitoring in patients with major cardiovascular disease, without the second-line restriction of the United States label [42]. The GeneReviews dose (80 mg) is compatible with both labels, but maintenance without reduction in severe renal impairment diverges from the FDA label [2, 41].

Adverse effects. Liver function test abnormalities, nausea, arthralgia and rash are frequent; hepatic failure, sometimes fatal, and severe skin reactions (Stevens-Johnson syndrome, DRESS, toxic epidermal necrolysis) have been described, mainly in the first month; febuxostat should not be restarted after a severe reaction [41, 42]. The United States label recommends monitoring for cardiovascular events and interrupting the drug in the event of liver injury, discontinuing it permanently if this is confirmed without another cause [41].

7.5. Clinical trials

Both trials are small, open-label, of short duration and with biochemical outcomes; neither measured stones, AKI, kidney function or graft survival [15, 45]. The comparisons refer to fixed doses and the dose equivalence between the two drugs is not known, so the superiority of febuxostat 80 mg over allopurinol 400 mg does not imply superiority over allopurinol 800 mg [8, 15]. The withdrawal of 5 of 12 participants in the 2025 trial limits interpretation [45]. The authors themselves consider that the clinical relevance of the difference remains to be demonstrated [8]. Evidence of clinical benefit comes from observational cohorts (section 3.9), subject to indication and selection bias [10, 21, 22]. No negative, suspended or withdrawn therapeutic trials were found on ClinicalTrials.gov; the ongoing registered studies are registries and biobanks (RKSC registry, NCT00588562, until 2028; biobank, NCT02026388; ProRKS, NCT02780297; natural history study of purine and pyrimidine disorders, NCT06092346) [48].

7.6. Hydration

Abundant fluid intake reduces supersaturation and crystallisation of 2,8-DHA and is a complement to pharmacotherapy [2]. GeneReviews does not set a volume [2]. For children with lithiasis of any cause, the French PNDS recommends average hydration of 2–3 L/m²/day distributed over 24 hours, with the aim of keeping urine specific gravity below 1010 on dipstick or osmolality below 300 mOsm/L, especially in the first morning urine; in the youngest children with very active lithiasis, nocturnal hydration via nasogastric tube or gastrostomy may be needed [11]. This target is an extrapolation of the general PNDS measures for paediatric lithiasis, applied to APRT deficiency, for which the PNDS also recommends hyperdiuresis [11]. Situations of volume depletion (gastroenteritis, heat, exercise, perioperative fasting) increase the risk of crystalline AKI and require increased hydration [10].

7.7. Diet

There are no data to support purine restriction, especially when the XOR inhibitor has already reduced urinary 2,8-DHA to very low values [2, 8]. As only a small fraction of body adenine comes from the diet, the expected effect is small [8]. Some sources, including Saudubray, Nyhan's Atlas and case reports, still recommend a low-purine diet [5, 14, 20, 29]. This divergence is resolved in favour of the position of GeneReviews and the Icelandic registry group: diet does not replace pharmacotherapy and is not necessary with adequate suppression; it may be considered, without evidence of benefit, when XOR inhibition is not possible.

7.8. Alkalinisation

Urinary alkalinisation with citrate or bicarbonate is not effective and is not indicated, because the solubility of 2,8-DHA does not increase within the physiological pH range, unlike that of uric acid [2, 8, 11, 14]. This is a frequent error when the stone is taken for uric acid [5].

7.9. Urological treatment

Surgical treatment of 2,8-DHA stones follows the general rules for lithiasis [2]. Extracorporeal shock wave lithotripsy (ESWL) is effective: the stones fragment easily and leave few residues; open surgery is reserved for stones that do not respond to less invasive methods [5, 8]. In children, the PNDS considers ESWL the reference option for most stones, with an efficacy of 60–90% and low morbidity, under general anaesthesia as a day case; ureteroscopy, percutaneous nephrolithotomy, laparoscopy and open surgery are reserved for poor indications for, or failures of, ESWL; and urological intervention should be considered above 7 mm, because of the risk of obstruction [11]. Fragments should be collected for analysis before they are pulverised [11]. Acute bilateral obstruction or obstruction of a single kidney is an emergency requiring decompression (section 4.3).

7.10. Chronic kidney disease

Management of CKD follows the general KDIGO recommendations cited by GeneReviews: intensive blood pressure control, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers in the presence of proteinuria, and cardiovascular risk reduction [2]. The XOR inhibitor should be maintained at all stages, because suppression of 2,8-DHA can recover part of the lost function: in Iceland, allopurinol significantly improved kidney function in most patients with renal impairment, and in the case in section 10 the eGFR rose from 31 to 50 mL/min/1.73 m² in 10 months [20, 23]. In advanced CKD, plasma 2,8-DHA tracks the loss of eGFR and may be the best marker of the systemic burden [35].

7.11. Dialysis and kidney transplantation

It is not known whether patients on dialysis benefit from an XOR inhibitor, except when transplantation is planned [2]. Before transplantation, GeneReviews recommends the inhibitor for at least six weeks, if possible, based on the authors' unpublished experience; the 2026 review by the same group indicates at least four weeks, with allopurinol at least 400 mg/day or febuxostat at least 80 mg/day, regardless of kidney function [2, 8]. Both sources agree that pretransplant and lifelong post-transplant treatment is essential; an interval of six weeks, when the schedule allows, meets both recommendations. The evidence is observational: in the international series of 17 patients and 22 grafts, two-year graft survival was 91% in patients treated before transplantation and 55% in those untreated (P=0.16), and eGFR at two years was 61.3 versus 16.2 mL/min/1.73 m² (P=0.009); four grafts were lost in two patients who only started allopurinol weeks or months after transplantation [27]. Higher doses, especially of allopurinol, were associated with less graft recurrence, and delayed graft function in treated patients stabilised after a dose increase [8, 27]. With treatment instituted before surgery, transplant outcomes are comparable to those for other causes of kidney failure [8]. The choice of immunosuppression should exclude azathioprine, which is contraindicated with febuxostat and has a major interaction with allopurinol (section 7.12) [40, 41]. In a report of two patients, conversion to belatacept was part of a strategy to reduce chronic tubulointerstitial injury, without any effect being attributable to it [29]. Any graft dysfunction should prompt measurement of 2,8-DHA or a search for crystals and consideration of biopsy [26, 28].

7.12. Agents to avoid and interactions

Azathioprine and mercaptopurine should be avoided in patients on allopurinol or febuxostat, because XOR inhibition increases their concentrations and myelotoxicity [2]. Febuxostat is contraindicated with these drugs in the USA and not recommended in the European Union, where, if the combination is unavoidable, reducing the dose of mercaptopurine or azathioprine to 20% or less is recommended [41, 42]. With allopurinol at 300–600 mg/day, the United States label requires the dose of mercaptopurine or azathioprine to be reduced to about one third to one quarter [40]. The allopurinol label also advises avoiding capecitabine, mentions the interaction with pegloticase and the increased risk of rash with thiazides, ampicillin and amoxicillin, especially with renal impairment; febuxostat requires caution with theophylline [40, 41]. The PNDS notes that certain medicines crystallise in the tubules and cause acute renal failure, which matters for patients who already have crystal nephropathy [11].

7.13. Pregnancy and breastfeeding

The safety of allopurinol and febuxostat in human pregnancy has not been studied systematically [2]. With allopurinol, high doses caused fetal effects in mice, but not in rabbits or rats; in the literature, among about 50 exposed pregnancies, there were two newborns with major malformations, without a defined pattern [2, 40]. The Portuguese SmPC allows use only in the absence of a safer alternative or when the disease poses a risk to the mother or fetus, and advises against breastfeeding, because allopurinol and oxipurinol pass into milk [43]. With febuxostat, studies in rats and rabbits showed no teratogenicity at exposures up to 40–51 times the maximum human dose, but very high doses with maternal toxicity increased neonatal mortality; the European SmPC advises against use in pregnancy and breastfeeding [2, 41, 42]. GeneReviews recommends considering allopurinol in pregnancy in women with CKD with reduced GFR or with a kidney transplant, when the benefit outweighs the risk, and discussing medication before conception, including transplant immunosuppressants [2]. Febuxostat should be avoided in pregnancy and breastfeeding [42].

7.14. Children and transition to adulthood

Allopurinol is the paediatric drug, with doses of 5–10 mg/kg/day in the reference recommendation and a median of 9 mg/kg/day in the French series [2, 22]. Febuxostat has no paediatric dose data or authorisation below 18 years of age [2, 41, 42]. Children treated early maintained normal kidney function into adulthood, and recurrences on treatment were associated with poor adherence and insufficient dosing, frequent in adolescence [21]. Transition to adult nephrology should ensure continuity of the inhibitor, adjustment of the dose to adult weight, maintenance of hydration and information about the interaction with azathioprine and about pregnancy planning [2, 21]. Paediatric follow-up should include growth [11].

7.15. Regulatory status

7.16. Developments and investigational therapies

Diagnosis and monitoring (published human studies). Between 2024 and 2026, the Icelandic group validated the simultaneous plasma assay for 2,8-DHA, adenine and drugs, showed the strong correlation between plasma 2,8-DHA and eGFR in untreated patients and published the second randomised trial comparing doses of allopurinol and febuxostat [34, 35, 45]. These assays allow the dose to be individualised instead of fixed empirically, but their availability is limited to specialist centres [8].

Mechanisms of injury (preclinical). In 2026, studies of human tissue and cell lines were published linking the injury to crystal burden, to the TNF-α and mTORC1 inflammatory pathways and to the adhesion molecule CD44, proposed as a therapeutic target; the murine model had already shown that deletion of Tnfr1 attenuates the disease [17–19]. None of these targets has reached a human trial. The stated aim is to find alternatives for patients who do not tolerate XOR inhibitors and to act on crystal injury, inflammation and fibrosis, which current treatment does not control [8, 19].

Gene silencing of XOR (phase 1 trial in another indication). RNA interference (siRNA) therapies silence the expression of a gene, predominantly in the liver, and have changed the treatment of primary hyperoxaluria [8]. An siRNA targeting XOR, HZN-457, was studied in a phase 1 trial in healthy volunteers (NCT05565768, Horizon Therapeutics, completed in August 2023), with a view to hyperuricaemia and gout, and is also being studied in murine models [8, 48]. The strategy would, in theory, replace the daily dose of an inhibitor with a lasting reduction of the enzyme that generates 2,8-DHA; the results of the phase 1 trial were not in the sources consulted, no trial has included patients with APRT deficiency and the risk of silencing an enzyme that also generates uric acid and participates in drug catabolism is unknown in this population; XOR metabolises, for example, azathioprine and mercaptopurine [8, 41, 48]. The level of evidence for APRT deficiency is therefore only that of a hypothesis based on a shared mechanism.

Gene editing and gene therapy (no clinical programmes). No gene therapy or gene editing programmes for APRT deficiency were found, not even at the preclinical stage, in the sources consulted. The authors of the 2026 review point out that CRISPR correction would face the difficulty of delivering the components to all tissues, because APRT is ubiquitously expressed; unlike the hyperoxalurias, in which the target organ is the liver, there is no single organ whose correction would suffice [8]. For this reason, the strategy closest to the clinic is modulation of XOR, not replacement of APRT [8]. A negative search does not prove the absence of ongoing projects.

Screening. Newborn screening has been proposed for Iceland and other populations with a high carrier frequency, with no programme implemented in the sources consulted [8].

7.17. Recommendations of the reference sources and divergences

8. Reproductive Options and Family Counselling

8.1. Recurrence risks

Inheritance is autosomal recessive [2]. The parents of an affected individual are obligate, asymptomatic heterozygotes: they have no 2,8-DHA crystals in the urine and no risk of disease [2, 8]. Each sibling of an affected individual has, at conception, a 25% probability of being affected, a 50% probability of being an asymptomatic carrier and a 25% probability of inheriting neither variant [2]. The children of an affected individual are all obligate carriers, unless the other parent is a carrier or affected; in that case the risk to each child is 50% [2]. The parents' siblings have a 50% probability of being carriers [2]. In consanguineous couples and in populations with founder alleles, the probability that the partner is a carrier is higher, as shown by the pseudodominance described in a Japanese family [1]. Germline mosaicism has not been described in the sources consulted; in a recessive disease with carrier parents it has little practical relevance.

8.2. Risk that the partner is a carrier

The carrier frequency estimated from known pathogenic variants differs greatly between populations: 1 in 42 in Iceland, 1 in 274 in Ireland, 1 in 388 in Japan (Japanese HGVD database), 1 in 650 in the UK Biobank, about 1 in 1000 in non-Finnish Europeans and Latinos in gnomAD, and 1 in 1386 in Africans [7]. These frequencies underestimate the true frequency, because they only count variants already described [7]. Older estimates, based on enzyme assays, gave 0.4–1.2% heterozygotes in several European populations and about 1% in Japan [1, 2, 5]. Carrier detection in the family requires knowing the proband's variants, because the enzyme assay in extracts does not distinguish heterozygotes from controls [2].

8.3. Cascade testing and testing of minors

All siblings of an affected individual should be tested, even if asymptomatic, with molecular testing if the familial variants are known or with enzyme activity, 2,8-DHA quantification or experienced urine microscopy if they are not [2, 8]. As there is an effective treatment that prevents kidney injury, testing is appropriate at any age, including in children, in the context of formal genetic counselling [2]. The result does not allow prediction of age at onset, type or severity of manifestations, and the 15–20% of asymptomatic homozygotes excrete 2,8-DHA, so they should also be treated [2]. In siblings with biallelic variants, assessment of kidney function and urine follows [2].

8.4. Prenatal and preimplantation diagnosis

Once the familial variants are known, prenatal diagnosis and preimplantation genetic testing are possible [2]. The ideal time to determine risk, clarify carrier status and discuss these options is before pregnancy; information should be given about the efficacy of preventive treatment, which weighs in the decision, and it should be recognised that views on the use of prenatal diagnosis differ between professionals and between families [2]. GeneReviews recommends banking DNA from probands without a molecular diagnosis, because testing methods evolve [2].

8.5. Counselling on medication and pregnancy

Counselling of a woman of reproductive age with APRT deficiency should include the safety data on XOR inhibitors in pregnancy and breastfeeding (section 7.13), the interaction with immunosuppressants if there is a transplant, and the importance of hydration [2, 42, 43].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence: two approaches with different results

The prevalence of APRT deficiency has been estimated in two ways, with results that do not coincide. The first starts from the heterozygote frequency measured by enzyme assays (0.4–1.2%) and, by Hardy-Weinberg equilibrium, arrives at a homozygote prevalence of at least 1 in 50,000 to 1 in 100,000; GeneReviews deduced from this 70,000–80,000 affected individuals worldwide, of whom 40,000 in Asia, 9000 in Europe and 8000 in the Americas, with at least 3000 in the USA, most of them undiagnosed [2]. The second, by Runolfsdottir and colleagues (2021), sums the frequencies of confirmed pathogenic variants in six large genomic databases and arrives at much lower values outside founder populations, about 1 in 4 million in non-Finnish Europeans in gnomAD, with no evidence of widespread underdiagnosis [7]. By 2026, the same group had only been able to identify about 500 cases worldwide, combining the literature, deCODE genetics data, public databases and the Icelandic registry [8]. Both approaches have limitations: the first depends on enzyme assays in which heterozygotes and controls overlap and on small samples; the second only counts variants already described, the databases poorly represent some populations and consanguineous unions increase homozygotes above expectation [7, 8]. Orphanet records point values validated by experts (Table 9.1) [3].

These figures measure different things. The Orphanet values result from diagnosed cases and expert assessment; those of Runolfsdottir and colleagues are theoretical estimates of expected genotypes from known variants; the French value is an allele frequency in a small sample of newborns, with a wide confidence interval. None should be read as an observed national prevalence.

9.2. Founder populations

Japan, Iceland and France account for most of the published cases [7]. In Japan, about 70 families had been described by 1992, compared with about 36 outside Japan; the estimated gene frequency was 1.2%, and three alleles (p.Met136Thr 68%, p.Trp98Ter 21%, CCGA duplication 7%) accounted for almost all defective alleles [1]. The p.Met136Thr allele has a unique haplotype, is evenly distributed across the four main islands and Okinawa and is present in Korea, which places its origin before 300 BC [4]. In Iceland, all patients are homozygous for p.(Asp65Val), whose allele frequency of 1.2% is about 100 times higher than in Scandinavia and the United Kingdom, from which the Icelandic population derives, which confirms a founder effect [7]. In France, the c.400+2dup variant was found in 54% of patients and accounts for 40% of alleles; the French patients came almost all from metropolitan France [7, 9]. In Lebanon, the variant c.188G>A p.(Gly63Asp) was found in several patients, in a context of frequent consanguinity (about 35% of marriages) [7].

9.3. United States and Europe

In the USA, the second case of homozygous deficiency was only published in 1988, and the APRT registry of the Rare Kidney Stone Consortium, based at the National University Hospital of Iceland (Landspitali), includes patients from Iceland and other countries, among them the USA [1, 2, 7]. In the registry, six patients from the USA had c.400+2dup and novel variants were described in North American patients [7]. gnomAD does not allow the North American population to be isolated, so there is no specific genetic estimate for the USA [7]. In the United Kingdom, the purine research laboratory identified 20 patients since 1974, with diagnosis at a mean age of 25 years and high mortality among late diagnoses [24]. In continental Europe, cases with c.400+2dup come from France, Germany, Austria, Belgium, Italy and Poland; p.(Asp65Val) also appeared in Spain and the United Kingdom [7]. The sources consulted do not include data from Portugal or Brazil; the frequency of the c.400+2dup variant in gnomAD is 0.043% in Southern Europeans, and the gnomAD value for Latinos does not distinguish Brazil [7]. The absence of published cases in these countries does not demonstrate absence of the disease.

9.4. Special populations

In children, the disease more often presents with lithiasis and obstructive AKI, and early diagnosis is associated with normal long-term kidney function [21, 22]. In older adults, CKD without stones is frequent and the diagnosis sometimes comes with biopsy or after transplantation [10, 20]. In transplant recipients, recurrence is the rule without prior treatment [27]. In people of East Asian ancestry, the higher frequency of HLA-B*58:01 increases the risk of allopurinol hypersensitivity, which justifies considering genotyping before starting allopurinol [40, 43].

9.5. European networks and patient resources

The relevant European Reference Network is ERKNet (European Rare Kidney Disease Reference Network), whose “Metabolic & Stone Disorders” working group covers hereditary lithiasis (https://www.erknet.org/) [12]. MetabERN (https://metab.ern-net.eu/) covers hereditary metabolic diseases [13]. In France, the rare kidney disease reference centres of the ORKiD network are responsible for the PNDS on lithiasis in children [11]. The Rare Kidney Stone Consortium maintains the APRT Deficiency Registry and the APRT Deficiency Support Network (www.rarekidneystones.org/dha/), based at Landspitali, in Reykjavik [2]. The RKSC registry (NCT00588562) and its biobank (NCT02026388) continue to recruit [48].

10. Illustrative Clinical Case

10.1. Presentation

Alder, Gaspert and Fuster published in 2025 the case of a 67-year-old man, the son of non-consanguineous Tamil parents, referred for progressive loss of kidney function, with an eGFR decline from 100 to 20 mL/min/1.73 m² over three and a half years [20]. He felt well. He had controlled hypertension and type 2 diabetes and urinary tract infections in childhood, but denied renal colic, haematuria or passage of stones; the family history was unremarkable. He was taking olmesartan, amlodipine, hydrochlorothiazide and metformin [20].

10.2. Findings

At the first visit, plasma creatinine was 266 µmol/L (3.0 mg/dL), with an eGFR of 21 mL/min/1.73 m², potassium of 7.1 mmol/L and bicarbonate of 14 mmol/L. The urine sediment was reported as normal, with a protein/creatinine ratio of 257 mg/g and an albumin/creatinine ratio of 81 mg/g. Ultrasound showed a coarse 9 mm parenchymal calcification at the lower pole of the left kidney, without stones [20]. After withdrawal of metformin, hydrochlorothiazide and olmesartan and initiation of sodium polystyrene sulfonate and sodium bicarbonate, potassium normalised, but eGFR only recovered to 31 mL/min/1.73 m² [20]. Kidney biopsy showed numerous birefringent crystals, mainly intratubular and focally interstitial, greenish-brown on haematoxylin-eosin and periodic acid-Schiff, light blue on acid fuchsin-orange G staining and grey-black on silver, suggestive of 2,8-DHA. A new urine microscopy, this time targeted, showed round reddish-brown crystals with a central Maltese cross under polarised light [20]. Figure 1 of the article brings together the four panels, H&E, silver, conventional and polarised urine microscopy (https://pmc.ncbi.nlm.nih.gov/articles/PMC12630293/figure/Fig1/) [20].

10.3. Molecular diagnosis

Genetic testing identified homozygosity for the variant NM_000485.3:c.200G>A, p.(Arg67Gln), previously described in patients with APRT deficiency; the nomenclature was confirmed in VariantValidator [6, 20].

10.4. Management and course

Allopurinol was started, later replaced by febuxostat 80 mg/day. The crystalluria disappeared completely and eGFR improved steadily to 50 mL/min/1.73 m² in 10 months [20].

10.5. Comment

The case brings together the features of APRT deficiency in adults: rapidly progressive CKD without clinical lithiasis, a sediment reported as normal until the correct crystal was looked for, diagnosis by biopsy and partial recovery of function with XOR inhibition [10, 20]. The initial sediment “without abnormalities” shows how 2,8-DHA crystals go unnoticed without an explicit request [10]. Two aspects of management in the report diverge from the reference sources. The authors consider enzyme activity, genetics or 2,8-DHA unnecessary when crystals are present, yet they performed genetic testing; the 2026 review requires confirmation in all cases [8, 20]. And they recommend a low-purine diet, which GeneReviews and the Icelandic group consider of no demonstrated benefit under adequate inhibition [2, 8, 20]. The case does not document testing of siblings, which would have been indicated [2].

11. Prognosis and Long-Term Follow-up

11.1. Prognosis

The prognosis depends on kidney function at diagnosis and on the earliness, dose and adherence of treatment [14, 21]. Without adequate treatment, 20–25% of patients reach end-stage kidney disease, generally in adulthood, and up to 15% are already there when the diagnosis is made [2]. In the RKSC registry, the eGFR slope was −0.38 mL/min/1.73 m² per year in treated patients compared with −5.74 in those untreated before stage 5 [10]. In children treated early, eGFR after almost 19 years was 114 mL/min/1.73 m², compared with 62 in those who only started treatment as adults [21]. The French series recorded 20% progression to end-stage kidney disease over a median follow-up of 74 months, in a cohort with frequent late diagnosis, and the British series five deaths among 20 patients [9, 24]. In transplantation, two-year graft survival was 91% with prior treatment and 55% without it [27]. Life expectancy and long-term quality of life have not been quantified in the sources consulted; the RKSC quality of life study (NCT02124395) was completed in 2022 without results in the sources [48].

11.2. Surveillance

GeneReviews notes that there are no consensus surveillance guidelines and proposes the scheme in Table 11.1, which is complemented by the PNDS recommendations for children and by the biochemical monitoring developed by the Icelandic group [2, 8, 11].

Absence of crystals on microscopy is an imperfect indicator: some patients without crystalluria continue to form stones and some with persistent crystalluria do not, which motivated the development of quantitative assays [15, 33]. In patients with advanced CKD, crystalluria is unreliable and plasma 2,8-DHA may be preferable [2, 35].

12. Summary and Final Considerations

APRT deficiency is an autosomal recessive disease caused by biallelic variants of APRT, at 16q24.3, in which unsalvaged adenine is converted by XOR into 2,8-DHA, a compound insoluble at any physiological urinary pH that forms radiolucent stones and crystals in the renal tubules and interstitium. The disease manifests at any age, with recurrent lithiasis, obstructive or crystalline AKI, reddish-brown stains in the nappy, CKD often without lithiasis and recurrence in the kidney graft, and does not affect other organs in any established way. Three founder variants, p.(Met136Thr) in Japan, p.(Asp65Val) in Iceland and c.400+2dup in Europe, account for most known alleles, and there is no genotype–phenotype correlation. The observed prevalence is low outside these populations, with about 500 cases identified worldwide, but the median diagnostic delay of more than seven years and the proportion of patients diagnosed already in kidney failure show that many cases are recognised late.

Diagnosis begins with an explicitly requested urine microscopy, which shows round brown crystals with a Maltese cross, and is confirmed by genetics, by absent erythrocyte APRT activity or by quantification of 2,8-DHA; stone analysis by FTIR and biopsy are suggestive but require confirmation, and the biopsy is easily confused with oxalate nephropathy. Radiolucent stones in a child, or with alkaline urine in an adult, should not be assumed to be uric acid.

Treatment is lifelong XOR inhibition. Allopurinol, at 5–10 mg/kg/day up to 800 mg/day, is first line and febuxostat, 80 mg/day, is the alternative in adults, both used outside the label indications and without routine reduction in renal impairment according to GeneReviews, in divergence from the labels. Two small trials showed that febuxostat suppresses 2,8-DHA more than allopurinol at the doses studied, without clinical outcomes. Hydration is an adjunct, alkalinisation is useless and purine restriction adds no benefit under adequate inhibition. Before transplantation, the inhibitor should be started weeks in advance and at full dose, and azathioprine should be avoided. The evidence of benefit comes from cohorts, which show preserved kidney function in children treated early and functioning grafts in patients treated before transplantation. All siblings of an affected individual should be tested, at any age, because treatment prevents kidney injury. Lines of research include plasma monitoring of 2,8-DHA, the inflammatory targets of crystal injury and silencing of XOR by RNA interference, as yet without studies in this disease.

13. References

1. Online Mendelian Inheritance in Man (OMIM). #614723 Adenine phosphoribosyltransferase deficiency; APRTD. Baltimore: Johns Hopkins University; created 2012 Jul 17, edited 2022 Mar 25. Available from: https://www.omim.org/entry/614723
2. Edvardsson VO, Sahota A, Palsson R. Adenine phosphoribosyltransferase deficiency. 2012 Aug 30 [updated 2019 Sep 26]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK100238/
3. Orphanet. Adenine phosphoribosyltransferase deficiency (ORPHA:976). Available from: https://www.orpha.net/en/disease/detail/976
4. Online Mendelian Inheritance in Man (OMIM). *102600 Adenine phosphoribosyltransferase; APRT. Baltimore: Johns Hopkins University; created 1986 Jun 4, edited 2023 Aug 6. Available from: https://www.omim.org/entry/102600
5. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Adenine phosphoribosyltransferase (APRT) deficiency. In: Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 511–515.
6. VariantValidator. Validation of APRT variants against NM_000485.3 (MANE Select), GRCh38. Available from: https://variantvalidator.org/
7. Runolfsdottir HL, Sayer JA, Indridason OS, Edvardsson VO, Jensson BO, Arnadottir GA, et al. Allele frequency of variants reported to cause adenine phosphoribosyltransferase deficiency. Eur J Hum Genet. 2021;29(7):1061–1070. doi:10.1038/s41431-020-00805-6
8. Edvardsson VO, Runolfsdottir HL, Palsson R. Adenine phosphoribosyltransferase deficiency and 2,8-dihydroxyadeninuria. Pediatr Nephrol. 2026;41(10):3263–3273. doi:10.1007/s00467-026-07187-9
9. Bollée G, Dollinger C, Boutaud L, Guillemot D, Bensman A, Harambat J, et al. Phenotype and genotype characterization of adenine phosphoribosyltransferase deficiency. J Am Soc Nephrol. 2010;21(4):679–688. doi:10.1681/ASN.2009080808
10. Runolfsdottir HL, Palsson R, Agustsdottir IM, Indridason OS, Edvardsson VO. Kidney disease in adenine phosphoribosyltransferase deficiency. Am J Kidney Dis. 2016;67(3):431–438. doi:10.1053/j.ajkd.2015.10.023
11. Centres de Référence des Maladies Rénales Rares, filière ORKiD. Protocole national de diagnostic et de soins (PNDS): Lithiase urinaire de l'enfant. Saint-Denis: Haute Autorité de Santé; septembre 2021. Available from: https://www.has-sante.fr/jcms/p_3296532
12. European Rare Kidney Disease Reference Network (ERKNet). Metabolic & Stone Disorders. Available from: https://www.erknet.org/
13. European Reference Network for Hereditary Metabolic Disorders (MetabERN). Available from: https://metab.ern-net.eu/
14. Marie S, Dewulf JP, Nassogne MC. Disorders of purine and pyrimidine metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 32, sections 32.1–32.2. doi:10.1007/978-3-662-63123-2_32
15. Edvardsson VO, Runolfsdottir HL, Thorsteinsdottir UA, Sch Agustsdottir IM, Oddsdottir GS, Eiriksson F, et al. Comparison of the effect of allopurinol and febuxostat on urinary 2,8-dihydroxyadenine excretion in patients with adenine phosphoribosyltransferase deficiency (APRTd): a clinical trial. Eur J Intern Med. 2018;48:75–79. doi:10.1016/j.ejim.2017.10.007
16. Runolfsdottir HL, Palsson R, Thorsteinsdottir UA, Indridason OS, Agustsdottir IMS, Oddsdottir GS, et al. Urinary 2,8-dihydroxyadenine excretion in patients with adenine phosphoribosyltransferase deficiency, carriers and healthy control subjects. Mol Genet Metab. 2019;128(1–2):144–150. doi:10.1016/j.ymgme.2019.05.015
17. Klinkhammer BM, Djudjaj S, Kunter U, Palsson R, Edvardsson VO, Wiech T, et al. Cellular and molecular mechanisms of kidney injury in 2,8-dihydroxyadenine nephropathy. J Am Soc Nephrol. 2020;31(4):799–816. doi:10.1681/ASN.2019080827
18. Helgudottir HR, Runolfsdottir HL, Hardarson S, Palsson R, Edvardsson VO, Gudjonsson T. 2,8-Dihydroxyadenine crystal nephropathy and mediators of kidney injury in adenine phosphoribosyltransferase deficiency. Kidney360. 2026. doi:10.34067/KID.0000001299
19. Helgudottir HR, Palsson R, Edvardsson VO, Gudjonsson T. 2,8-Dihydroxyadenine disrupts epithelial integrity and alters kidney cell phenotype in vitro. J Mol Med (Berl). 2026;104:23. doi:10.1007/s00109-025-02625-x
20. Alder H, Gaspert A, Fuster DG. Adenine phosphoribosyltransferase deficiency as a rare cause of rapidly progressive chronic kidney disease. J Nephrol. 2025;38(8):2449–2451. doi:10.1007/s40620-025-02304-7
21. Runolfsdottir HL, Palsson R, Agustsdottir IM, Indridason OS, Edvardsson VO. Long-term renal outcomes of APRT deficiency presenting in childhood. Pediatr Nephrol. 2019;34(3):435–442. doi:10.1007/s00467-018-4109-x
22. Harambat J, Bollée G, Daudon M, Ceballos-Picot I, Bensman A; APRT Study Group. Adenine phosphoribosyltransferase deficiency in children. Pediatr Nephrol. 2012;27(4):571–579. doi:10.1007/s00467-011-2037-0
23. Edvardsson V, Palsson R, Olafsson I, Hjaltadottir G, Laxdal T. Clinical features and genotype of adenine phosphoribosyltransferase deficiency in Iceland. Am J Kidney Dis. 2001;38(3):473–480. doi:10.1053/ajkd.2001.26826
24. Balasubramaniam GS, Arenas-Hernandez M, Escuredo E, Fairbanks L, Marinaki T, Mapplebeck S, et al. Adenine phosphoribosyltransferase deficiency in the United Kingdom: two novel mutations and a cross-sectional survey. Clin Kidney J. 2016;9(6):800–806. doi:10.1093/ckj/sfw093
25. Lajmi K, Bouzid K, Blaiech N, Mabrouk S, Nouir S, Kechiche N. Urolithiasis due to adenine phosphoribosyl transferase (APRT) deficiency: a case report and practical recommendations in children. Arch Esp Urol. 2025;78(9):1323–1328. doi:10.56434/j.arch.esp.urol.20257809.171
26. Zaidan M, Palsson R, Merieau E, Cornec-Le Gall E, Garstka A, Maggiore U, et al. Recurrent 2,8-dihydroxyadenine nephropathy: a rare but preventable cause of renal allograft failure. Am J Transplant. 2014;14(11):2623–2632. doi:10.1111/ajt.12926
27. Runolfsdottir HL, Palsson R, Agustsdottir IMS, Indridason OS, Li J, Dao M, et al. Kidney transplant outcomes in patients with adenine phosphoribosyltransferase deficiency. Transplantation. 2020;104(10):2120–2128. doi:10.1097/TP.0000000000003088
28. Uesugi N, Miura Y, Nakafusa Y, Ikeda N, Masutani K. Chronic allograft injury by recurrent crystalline nephropathy of adenine phosphoribosyl transferase deficiency, even after early initiation of xanthine oxidase inhibitor. Nephron. 2025;149(2):92–96. doi:10.1159/000546495
29. Raza HMA, Ibrahim A, Talwar M, Balaraman V, Bhalla A, Wall BM, et al. Crystalline nephropathy due to 2,8-dihydroxyadeninuria in a transplanted kidney: 2 case reports. Am J Case Rep. 2025;26:e946972. doi:10.12659/AJCR.946972
30. Kim B, Lee H, Yoon HE, Shin SJ, Kim HS, Choi YJ, et al. Recurrent 2,8-dihydroxyadenine crystalline nephropathy after kidney transplantation due to adenine phosphoribosyltransferase deficiency. Kidney Res Clin Pract. 2026. doi:10.23876/j.krcp.26.166
31. Soliman E, Amin S, Nagra D, Soe NN, Rosa C, Banerjee S. P021 Crystal clear diagnosis: 2,8-dihydroxyadenine crystalline arthropathy mimicking gout. Rheumatol Adv Pract. 2025;9(Suppl 1):rkaf111.050. doi:10.1093/rap/rkaf111.050
32. Leow EH, Ganesan I, Chong SL, Yap CJY, Chao SM, Wang F, et al. Adenine phosphoribosyltransferase (APRT) deficiency: an increasingly recognized disease. Int Urol Nephrol. 2025;57(8):2519–2529. doi:10.1007/s11255-025-04420-6
33. Thorsteinsdottir M, Thorsteinsdottir UA, Eiriksson FF, Runolfsdottir HL, Agustsdottir IM, Oddsdottir S, et al. Quantitative UPLC-MS/MS assay of urinary 2,8-dihydroxyadenine for diagnosis and management of adenine phosphoribosyltransferase deficiency. J Chromatogr B Analyt Technol Biomed Life Sci. 2016;1036–1037:170–177. doi:10.1016/j.jchromb.2016.09.018
34. Thorsteinsdottir UA, Runolfsdottir HL, Eiriksson FF, Agustsdottir IMS, Edvardsson VO, Palsson R, et al. Optimization and validation of a UPLC-MS/MS assay for simultaneous quantification of 2,8-dihydroxyadenine, adenine, allopurinol, oxypurinol and febuxostat in human plasma. J Chromatogr B Analyt Technol Biomed Life Sci. 2024;1235:124041. doi:10.1016/j.jchromb.2024.124041
35. Thorsteinsdottir UA, Runolfsdottir HL, Edvardsson VO, Thorsteinsdottir M, Palsson R. Correlation of plasma and urine 2,8-dihydroxyadenine and adenine and clinical characteristics in individuals with adenine phosphoribosyltransferase deficiency. J Inherit Metab Dis. 2025;48(4):e70054. doi:10.1002/jimd.70054
36. Cogal AG, Ali AE, Arnous MG, Alhadi A, Zhou LT, Arroyo J, et al. Genetic and clinical characterization of a large cohort with suspected monogenic stone disease. Clin J Am Soc Nephrol. 2025;20(12):1670–1682. doi:10.2215/CJN.0000000817
37. Wagner CA, Sayer JA, Straub M. Role and relevance of genetic testing in patients with kidney stones: a review from EAU Section of Endourology. Curr Opin Urol. 2026. doi:10.1097/MOU.0000000000001420
38. Saudubray JM, García-Cazorla Á. Clinical approach to inborn errors of metabolism in paediatrics. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, section 1.6.9, Table 1.40. doi:10.1007/978-3-662-63123-2_1
39. Edvardsson VO, Goldfarb DS, Lieske JC, Beara-Lasic L, Anglani F, Milliner DS, et al. Hereditary causes of kidney stones and chronic kidney disease. Pediatr Nephrol. 2013;28(10):1923–1942. doi:10.1007/s00467-012-2329-z
40. Allopurinol tablets — prescribing information. ANDA211820; label effective 2026 Jun 3 (reference product Zyloprim, NDA016084, approved 1966). Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=00b27899-e68f-e9d9-e063-6394a90a88cb
41. Takeda Pharmaceuticals America. ULORIC (febuxostat) tablets — prescribing information. NDA021856; initial US approval 2009; label effective 2025 Dec 24. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=54de10ef-fe5f-4930-b91d-6bbb04c664bd
42. European Medicines Agency. Adenuric (febuxostat): summary of product characteristics (EPAR product information). First authorisation 2008 Apr 21. Available from: https://www.ema.europa.eu/en/documents/product-information/adenuric-epar-product-information_en.pdf
43. INFARMED. Zyloric 300 mg tablets (allopurinol) — summary of product characteristics. Faes Farma Portugal. Available from: https://extranet.infarmed.pt/INFOMED-fo/ (med_id 9523)
44. Agência Nacional de Vigilância Sanitária (ANVISA). Open data on registered medicines (DADOS_ABERTOS_MEDICAMENTOS.csv). Available from: https://dados.anvisa.gov.br/dados/DADOS_ABERTOS_MEDICAMENTOS.csv
45. Runolfsdottir HL, Thorsteinsdottir UA, Johannesdottir S, Oskarsdottir T, Agustsdottir IMS, Thorsteinsdottir M, et al. Comparison of the effect of allopurinol and febuxostat on 2,8-dihydroxyadenine in plasma and urine: a clinical trial. J Nephrol. 2025;38:2831–2839. doi:10.1007/s40620-025-02392-5
46. Morris AAM, Jones S. Medications used in the treatment of inborn errors of metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 45. doi:10.1007/978-3-662-63123-2_45
47. Aoki Y, Maeda M, Hashimoto J, Zaitsu A, Ueno K, Furukawa S, et al. Experience with early treatment using febuxostat in a patient with adenine phosphoribosyltransferase deficiency after pediatric kidney transplantation: a case report. CEN Case Rep. 2026;15:136. doi:10.1007/s13730-026-01176-5
48. ClinicalTrials.gov. Search: adenine phosphoribosyltransferase / 2,8-dihydroxyadenine; NCT02752633, NCT05565768, NCT00588562. Available from: https://clinicaltrials.gov/