The proband (I-2) was a 67-year-old Han Chinese woman from a non-consanguineous family. She was diagnosed with hepatic WD in 2013 at 54 years of age after presenting with abdominal distension. At diagnosis, she had cirrhosis, splenomegaly, portal-vein widening, Kayser-Fleischer (K-F) rings, and a serum ceruloplasmin concentration of 0.03 g/L, without ascites or neurological signs. She improved after approximately 2 weeks of intravenous sodium dimercaptopropanesulfonate and subsequently received penicillamine 750 mg/day, zinc, and dimercaptosuccinic acid but discontinued treatment after 8 months. In March 2026, she returned with decompensated cirrhosis, ascites, esophageal varices, albumin 30.7 g/L, platelets 86 × 10⁹/L, and a 24-h urinary copper of 186.1 μg/24 h before chelation therapy was restarted. Ascites resolved after diuretics, penicillamine 500 mg/day, and two further courses of intravenous sodium dimercaptopropanesulfonate, each consisting of 500 mg/day for 5 days. Her retrospective Leipzig score was 10^[7].

The proband’s son (II-1) was a 36-year-old man who was asymptomatic when diagnosed with hepatic WD in 2013 at 23 years of age. At diagnosis, he had splenomegaly and serum ceruloplasmin of 0.03 g/L but no K-F rings or neurological signs. Cirrhosis and portal hypertension were documented in 2016, and intermittent hand tremor developed in 2024. He had previously taken penicillamine and zinc gluconate intermittently, with poor treatment adherence. Before a March 2026 urine collection, he self-administered penicillamine 250 mg/day for 2 weeks; 24-h urinary copper was 3,726.6 μg/24 h and was not used for diagnostic Leipzig scoring because it was treatment-exposed. In May 2026, he had decompensated cirrhosis with small-volume ascites, esophageal and gastric varices, hypersplenism, and portal hypertensive gastropathy. During treatment in 2026, he received penicillamine 750 mg/day, elemental zinc 150 mg/day (as zinc gluconate), and two 5-day courses of intravenous sodium dimercaptopropanesulfonate at 500 mg/day. His tremor and biochemical abnormalities partially improved. His documented retrospective Leipzig subtotal was 7 after excluding treatment-exposed urinary copper, which still supported an established diagnosis.

The proband’s daughter (II-2) was a 46-year-old woman diagnosed with predominantly hepatic WD in 2009 at 29 years of age after abdominal symptoms, anorexia, tremor, and headache. K-F rings were repeatedly absent, and brain magnetic resonance imaging (MRI) in 2020 was normal. Serial imaging showed cirrhosis, splenomegaly, portal hypertension, and gastric fundal varices. Progressive hypersplenism led to total splenectomy with pericardial devascularization in December 2025, after which the platelet count increased from 51 × 10⁹/L to more than 290 × 10⁹/L. In 2026, serum ceruloplasmin was 0.03 g/L. The 24-h urinary copper was 1,670.1 μg/24 h on March 21 before the 2026 chelation course was started and 1,887.7 μg/24 h on July 1 during ongoing penicillamine 750 mg/day. The pretreatment March value was used for retrospective Leipzig scoring, giving a total score of 9. Earlier intermittent penicillamine treatment had generally been 500 mg/day, with poor adherence. Abdominal distension improved after treatment intensification to 750 mg/day in 2026, whereas intermittent tremor and headache persisted.

The proband’s grandson (III-1) was a 22-year-old man who had never developed clear hepatic or neurological symptoms. Approximately 10 years earlier, before chelation therapy, serum ceruloplasmin was 0.13 g/L and 24-h urinary copper was 103 μg/24 h; together with a heterozygous ATP7B finding, these results led to a clinical diagnosis of WD and initiation of penicillamine. At recent reassessment, he remained clinically asymptomatic; K-F rings were absent, ultrasound showed fatty liver, serum ceruloplasmin was 0.143 g/L, and 24-h urinary copper had decreased to 24.6 μg/24 h during ongoing penicillamine 375 mg/day. No clinical symptomatic response could be assessed because he was asymptomatic, and the on-treatment fall in urinary copper was not considered diagnostic evidence. However, liver biochemistry was assessed on two consecutive occasions, with mild elevations of both alanine aminotransferase (ALT) (90-98 U/L) and aspartate aminotransferase (AST) (41-47 U/L). WES identified only monoallelic c.2621C>T (p.Ala874Val), with no second pathogenic ATP7B sequence variant or phenotype-related copy-number variant (CNV). Using the historical pretreatment values, his retrospective Leipzig score was 3: ceruloplasmin 0.13 g/L, 1 point; urinary copper 103 μg/24 h [1.72 × the upper limit of normal (ULN) of 60 μg/24 h], 1 point; and one pathogenic ATP7B variant, 1 point. This corresponds to a possible diagnosis requiring additional evidence; his molecular diagnosis therefore remains unresolved.

Genomic DNA from the five tested family members was analyzed by phenotype-driven WES. Hybrid-capture libraries covered the coding exons of approximately 20,700 genes and 20-bp flanking intronic regions, together with the mitochondrial genome. Sequence reads were aligned to GRCh37/hg19 and analyzed using ISOgenetic v1.2. ATP7B variants were described using transcript NM_000053.4 and classified according to the 2015 American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) framework^[8]. WES-based CNV analysis assessed deletions or duplications spanning at least two adjacent exons; lower-confidence findings required orthogonal confirmation by quantitative polymerase chain reaction or multiplex ligation-dependent probe amplification (MLPA) in accordance with laboratory procedures and CNV interpretation standards^[9]. No phenotype-related CNV was reported in any tested individual.

Overall target-region mean sequencing depth and coverage at ≥ 20× were 167.84×/98.22% for the proband, 170.35×/98.26% for the son, 160.08×/98.17% for the daughter, 159.49×/98.31% for the grandson, and 195.60×/98.41% for the son-in-law. The capture design included all ATP7B coding exons and adjacent ±20-bp intronic sequence. Sanger sequencing was used for orthogonal confirmation and segregation analysis of c.2333G>T, c.2621C>T, and c.2975C>T [Figure 2].

The proband carried ATP7B c.2621C>T (p.Ala874Val) in exon 11 and c.2333G>T (p.Arg778Leu) in exon 8. Sanger segregation analysis showed transmission of c.2621C>T without c.2333G>T to her children, thereby demonstrating that the two variants in the proband were in trans. The son and daughter each carried c.2621C>T (p.Ala874Val) and c.2975C>T (p.Pro992Leu). Their trans configuration was inferred rather than directly demonstrated: c.2621C>T was maternally transmitted, whereas c.2975C>T was absent from the proband and present in both siblings, strongly supporting inheritance from the deceased father, who was unavailable for testing and was therefore classified as a presumed heterozygous carrier.

The son-in-law had no pathogenic ATP7B sequence variant or phenotype-related CNV detected by WES. The grandson carried only the maternal c.2621C>T (p.Ala874Val) variant. Thus, current molecular findings establish biallelic ATP7B-related WD in the proband and her two children but not in the grandson.

All three reported ATP7B variants were classified as pathogenic by the diagnostic laboratory under ACMG/AMP criteria. For c.2621C>T (p.Ala874Val), the applied evidence included PM3_VeryStrong, PP4, PS3_Supporting, PM2_Supporting, and PP3; functional work has demonstrated markedly impaired copper transport^[10]. For c.2333G>T (p.Arg778Leu), evidence included PM3_VeryStrong, PP4, PS3_Moderate, PM5_Strong, PM2_Supporting, and PP3. For c.2975C>T (p.Pro992Leu), evidence included PM3_VeryStrong, PS3_Moderate, PM2_Supporting, and PP3, with PP4 additionally applicable in an affected family member. Functional studies have shown that p.Arg778Leu and p.Pro992Leu impair ATP7B localization, trafficking, and/or copper excretion^[11]. ClinVar was broadly concordant with these laboratory interpretations, although isolated discordant submissions remained at the time of access (28 September 2026). For c.2621C>T (p.Ala874Val), 18 contributing submissions classified the variant as pathogenic, 4 as likely pathogenic, and 1 as a variant of uncertain significance (VUS)^[12]. For c.2333G>T (p.Arg778Leu), 23 classified it as pathogenic and 1 as VUS^[13]. For c.2975C>T (p.Pro992Leu), 15 classified it as pathogenic, 1 as likely pathogenic, and 1 as VUS^[14]. Accordingly, the current aggregate ClinVar records are formally labeled as having conflicting classifications, but the large majority of contributing submissions support pathogenic or likely pathogenic interpretations.