Clinical and Molecular Analysis of ATP7B Variants Identified by Next-Generation Sequencing in Iraqi Adults With Wilson Disease

Summary

Objectives:

This study aimed to identify and analyse ATP7B variants in Iraqi adults with Wilson disease (WD) by long-read next-generation sequencing.

Methods:

This cross-sectional study was conducted at the Poisoning Consultation Center at Ghazy Al-Hariri Hospital for Surgical Specialties and the Gastroenterology Consultation Clinic at Baghdad Teaching Hospital, Medical City in Baghdad, Iraq. Unrelated patients with clinical and biochemical features suggestive of WD were recruited between October 2022 and October 2023. DNA was extracted from peripheral blood samples. Variants in the ATP7B gene were identified using long-read next-generation sequencing and then analysed by in-silico tools.

Results:

A total of 45 patients were recruited in which 59 unique variants were detected; of them, 47 were deleterious, 9 were variants of uncertain significance (VUS) and 3 had a conflicting interpretation of pathogenicity. Those variants were detected in 80 out of 90 alleles of the ATP7B gene. Of the participants, 23 (51.1%) patients had 2 deleterious variants (8 in homozygous and 15 in compound heterozygous state); 12 (26.7%) patients had 1 deleterious variant plus 1 VUS or 1 with conflicting pathogenicity; and 10 (22.2%) patients were carriers of a single disease-causing variant. The most frequent variant, c.4021G>A (p.Gly1341Ser), was detected in 5 alleles, while c.3191A>C (p.Glu1064Ala) was detected in 4 alleles, followed by c.2165dupT (p.Arg723GlufsTer32) and c.3247C>T (p.Leu1083Phe), each detected in 3 alleles. Among the 59 variants, 42 were missense, 9 were frameshift, 6 were stop-gain, 2 were splice-donors and 1 was an in-frame deletion. The variant H1069Q, which is common worldwide, was not detected in this study.

Conclusions:

The ATP7B mutational spectrum in Iraqi patients with WD is significantly diverse, despite high rates of consanguinity. Evidence was provided for 8 variants to be considered for reclassification as deleterious. The diagnostic criteria for those with high Leipzig scores with only a single deleterious variant remain questionable.

Advances in Knowledge

• This study identified 47 unique deleterious ATP7B variants in Iraqi adults with Wilson disease (WD), highlighting a highly diverse mutational spectrum.

• A total of 9 variants of uncertain significance (VUS) and 3 variants with conflicting pathogenicity were found.

• Most patients (51.1%) had 2 deleterious variants, while 26.7% had 1 plus a VUS/conflicting variant and 22.2% carried a single disease-causing variant.

• The most frequent variants were c.4021G>A (p.Gly1341Ser) and c.3191A>C (p.Glu1064Ala).

• This study provides further evidence to suggest reclassification of 8 variants as deleterious and 4 as benign.

• The results raised questions about the diagnostic criteria for patients with high modified Leipzig scores and only a single deleterious variant.

Application to Patient Care

• This study enables more precise genetic diagnosis of WD in Iraqi adults by expanding the known spectrum of ATP7B variants in the Middle East and the region.

• Patients should receive personalised management by identifying specific mutations, which may inform prognosis and therapeutic strategies.

• This study assists in distinguishing between pathogenic and benign variants, reducing uncertainty in patient care.

• The results enhance genetic counselling, family screening and planning.

• Certain variants should be reclassified to improve the accuracy of genetic counselling and risk assessment.

• There is a need to re-evaluate diagnostic criteria for patients with strong clinical evidence but only 1 identified deleterious variant, potentially impacting treatment decisions.

1. Introduction

Wilson disease (WD), also known as progressive hepatolenticular degeneration (OMIM #277900), is a rare autosomal recessive disorder caused by biallelic pathogenic variants in the ATP7B gene.1 The ATP7B gene encodes a P-type copper-transporting ATPase that facilitates copper incorporation into ceruloplasmin (CP) and mediates hepatic copper excretion. Located on chromosome 13q14.3, the ATP7B gene comprises 21 exons and 20 introns encoding a protein of 1,465 amino acids.2 Mutations in ATP7B impair protein synthesis or function, leading to defective copper excretion and subsequent accumulation in organs such as the liver, eyes and brain. This results in a spectrum of clinical manifestations, including hepatic, neurological and systemic symptoms.3

The global prevalence of WD is estimated at 1 in 30,000 to 1 in 50,000 individuals, with a carrier frequency of approximately 1 in 90.1 Most patients present between the ages of 5 and 35, although onset can occur as early as 8 months or as late as the ninth decade.4 Diagnosis in children is typically more straightforward due to predominant hepatic involvement, although subtle neurological signs may already be present. In adolescents and adults, the initial presentation may involve neurological or psychiatric symptoms, which can delay diagnosis, especially when hepatic signs are absent.5

Traditionally, WD diagnosis has relied on clinical presentation and conventional biochemical markers, such as elevated 24-hour urinary copper, low serum CP levels and increased hepatic copper content. These markers, however, may sometimes yield inconclusive or misleading results, making the diagnosis of WD challenging, especially in adults. Therefore, molecular testing is recommended for a definitive diagnosis, particularly in asymptomatic individuals and first-degree relatives of affected patients.6

Screening of first-degree relatives using biochemical and/or molecular methods is strongly recommended by both the American Association for the Study of Liver Diseases and the European Association for the Study of the Liver (EASL).7 To date, over 1,275 distinct variants in the ATP7B gene have been identified.7 These include common hotspot mutations as well as a broad spectrum of rare variants, displaying considerable ethnic and regional variation.6 A clear genotype–phenotype correlation in WD has not been established, suggesting that epigenetic factors, including environmental and nutritional influences, may modulate clinical outcomes.8

In 2001, the Working Group at the 8^th International Meeting on Wilson Disease in Leipzig, Germany, developed a diagnostic scoring system that has since been incorporated into the EASL guidelines. A score of 4 or higher confirms WD; a score of 3 indicates probable disease requiring further testing, while a score of 2 or less makes WD unlikely.9 The system was later updated to include additional criteria such as family history and serum CP levels below or above 5 mg/dL. The most recent revision to the Leipzig criteria, incorporated in the 2025 EASL-European Reference Network Clinical Practice Guidelines, includes relative exchangeable copper measurements and emphasises the use of genetic testing alongside refined clinical criteria to enhance diagnostic accuracy and reduce diagnostic delays.10

Next-generation sequencing (NGS) of the ATP7B gene is now considered the gold standard for identifying both known and novel mutations. Mutational analysis by sequencing methods can distinguish healthy heterozygote carriers from asymptomatic individuals with biallelic variants and confirm the diagnosis of WD in symptomatic patients. Although NGS detects mutations in approximately 95% of affected individuals, access to this technology is limited in many low-resource settings. Furthermore, short-read NGS may fail to identify large structural variants such as deletions, duplications or insertions.11

Long-read NGS, capable of sequencing fragments ranging from 10,000 to 100,000 base pairs, offers more comprehensive genomic coverage and higher accuracy in detecting complex variants. It can detect single-nucleotide polymorphisms as well as small and large structural alterations (deletions, insertions, indels, and complex rearrangements). However, interpretation of molecular findings can be challenging, particularly when variants are classified as conflicting or of uncertain significance (VUS).12

This study aimed to detect and analyse ATP7B variants in Iraqi adults with a diagnosis or clinical and biochemical features suggestive of Wilson disease using direct-gene long-read NGS (LR-NGS).

2. Methods

2.1. Data collection

This cross-sectional study was conducted between October 2022 and October 2023. Patients were enrolled from two centres within the Medical City complex in Baghdad, Iraq: The Poisoning Consultation Center at Ghazy Al-Hariri Hospital for Surgical Specialties and the Gastroenterology Consultation Clinic at Baghdad Teaching Hospital. The study included unrelated adults aged 18 years or more, who had either a confirmed diagnosis or clinical and biochemical findings suggestive of WD. Clinical data, including patient history, physical examination findings and diagnostic test results, were obtained through direct interviews and hospital records.

Initial evaluation included laboratory and imaging studies, which were consistent with WD. These comprised elevated 24-hour urinary copper excretion (>100 μg/day), decreased serum ceruloplasmin levels (<20 mg/dL) and raised liver enzyme levels. The presence of Kayser-Fleischer (KF) rings and a positive family history of WD or a similar disorder were also documented. Patients diagnosed with viral or autoimmune hepatitis were excluded from the study.

All patients' data were translated into numerical scores using the modified Leipzig scoring (MLS) system. For sample size calculation, the following equation was used:

$$\text{Finite}\hspace{0.17em}\text{population}:\hspace{0.17em}n\prime =\frac{n}{1+\hspace{0.17em}\frac{z^2\times \hat{p}(1-\hat{p})}{{\epsilon }^{2}N}}$$

Where z is the z score, ε is the margin of error, N is the population size and ^p is the population proportion. Assuming the confidence level of 95%, a margin of error of 1%, a population size of Iraq of 40+ millions and a population proportion of 1%, the representative sample size was 43.

2.2. Molecular analysis

Comprehensive sequencing of the entire ATP7B gene—including coding regions, non-coding regions, promoter, 3'-UTR, and 5'-UTR—was performed using direct gene LR-NGS.

For molecular analysis, 2 mL of peripheral blood was collected from each patient in K₂EDTA tubes and stored at 4 °C. Genomic DNA was extracted using the Reliaprep Blood gDNA Miniprep System (Promega, USA) and DNA quantification was carried out using the Qubit™dsDNA HS Assay Kit (Invitrogen-Thermo Fisher®, USA).

PCR was conducted using the KAPA HiFi PCR Kit (Roche®, Switzerland) to amplify the entire ATP7B gene; 10 gene-specific primers from Macrogen® (Korea) were used in 14 different combinations (forward and reverse), placed into two tubes for each combination. Amplified products were purified using Agencourt AMPure XP PCR Purification (Beckman Coulter, USA).

2.3. Next-generation sequencing

Sequencing was performed on the MinION Mk1C platform (Oxford Nanopore Technologies, UK). Data acquisition and processing were handled via MinKNOW software, which captures raw signals and transfers them to a base-calling pipeline according to the manufacturer's protocols.

Bidirectional sequence reads were assembled and aligned to the reference human genome (GRCh37/UCSC hg19) and NCBI Reference Sequence Transcripts. Gene-specific filters were applied to identify single-nucleotide variants, deletions, insertions and duplications across the entire gene. Regions with insufficient coverage or ambiguous copy number data were further evaluated using alternative sequencing or copy number detection methods.

Data analysis begins with real-time basecalling and methylation calling using MinKNOW™, the software that controls Oxford Nanopore sequencing devices. The high accuracy basecalling mode was used as recommended. Following basecalling, the EPI2ME™ wf-human-variation workflow provides all-in-one calling of SNVs, SVs, copy number variants, STR expansions and methylation, covering both 5-methylcytosine and 5-hydroxymethylcytosine. This workflow also enables phasing of these variants. The EPI2ME platform provides data analysis solutions for all levels of experience. The workflow outputs a series of intuitive reports, VCF files, listing variants, a BEDmethyl file containing methylation information and QC metrics. The results from the wf-human-variation workflow were further explored by viewing in a track-based genome browser, such as Integrative Genomics Viewer (IGV), and can be assessed for known pathogenicity through tertiary analysis software.

2.4. Bioinformatic analysis

Variant classification and annotation were carried out using tools such as IGV, Mutation Taster and other bioinformatics software. Variants were reported if categorised as pathogenic, likely pathogenic or VUS. In-silico analyses were performed using PolyPhen-2 and SIFT to predict the impact of each variant on protein function.

3. Results

A total of 45 unrelated Iraqi adults with features suggestive of WD were recruited. Their ages ranged between 19 and 68 years (mean ± standard deviation [SD] = 37.1 ± 13.6). There were 29 (64.4%) males and 16 (35.6%) females, with a male-to-female ratio of 1.8:1 [Table 1].

Table 1.

Demographic and clinical characteristics of the recruited Iraqi patients with Wilson disease (N = 45).

Characteristic	n (%)
Age in years
19–30	20 (44.4)
31–50	16 (35.6)
>50	9 (20)
Sex
Male	29 (64.4)
Female	16 (35.6)
Parental consanguinity
Positive	34 (75.6)
Negative	11 (24.4)
Family history of WD or a similar condition
Positive	17 (37.8)
Negative	28 (62.2)
Ethnicity
Arabs	42 (93.3)
Kurds	3 (6.7)
Kayser-Fleischer ring
Positive	8 (17.8)
Negative	37 (82.2)
Clinical presentation
Hepatic	35 (77.8)
Neurologic	6 (13.3)
Mixed	2 (4.4)
Pure psychiatric	1 (2.2)
Asymptomatic	1 (2.2)
Main clinical presentation, mean ± SD
Mean age in years ± SD	Hepatic (n = 35)	Mixed (n = 2)	Neuropsychiatric (n = 7)
*At presentation: 32 ± 15.2	33.9 ± 15.6	16.0 ± 2.8	31.42 ± 13.52
At diagnosis 37.1 ± 13.6	38.8 ± 13.7	23.5 ± 5.6	36.14 ± 12.97

WD = Wilson disease; SD = standard deviation.

^*P = 0.119 using student's t-test.

Direct ATP7B gene sequencing using LR-NGS detected a wide range of disease-causing variants and a group of VUS, some with conflicting interpretations of their probable deleterious or damaging effects on protein function. Overall, molecular confirmation of WD diagnosis was achieved in 23 of the 45 examined patients, yielding a diagnostic rate of 51.1%.

A total of 13 patients had an MLS score of ≥4 prior to NGS and the diagnosis was molecularly confirmed in 8 patients (homozygous or compound heterozygous); 4 had 1 VUS in addition to 1 disease-causing variant. The remaining patient had only 1 deleterious variant. An MLS score of 3 prior to NGS was found in 13 patients and the diagnosis was confirmed in 7 patients; 2 had 1 VUS in addition to a disease-causing variant and 4 had 1 deleterious variant only. An MLS score of 2 prior to NGS was found in 17 patients and the diagnosis was confirmed in 9 patients; 3 had a VUS in addition to the disease-causing variant and the remaining 5 had only 1 deleterious variant. A single patient had an MLS score of 1 with hepatic manifestations and was found to be a heterozygote with 1 additional conflicting variant. Another patient was asymptomatic with a positive family history (a deceased sister) and was recruited as part of family screening [Supplementary Table 1].

A total of 59 unique variants were identified in 80 out of 90 alleles of the ATP7B gene, with a detection rate of 88.9% [Table 2]. The most frequently detected variant was c.4021G>A (p.Gly1341Ser), which was detected in 5 alleles, followed by c.3191A>C (p.Glu1064Ala) detected in 4 alleles; 2 (3.4%) variants were detected in 3 alleles each, 12(20.3%) variants were detected in 2 alleles each and 43 (72.9%) variants were detected in 1 allele only. Of all detected 59 variants, 42 (71.2%) were missense, 9 (15.2%) were frameshift, 5 (8.5%) were stop-gain, 2 (3.4%) were splice-donor and 1 (1.7%) was an in-frame deletion.

Table 2.

Variants identified by direct ATP7B gene sequencing using long-read next-generation sequencing.

#	Nucleotide change	Amino acid change	Variant effect	ClinVar classification	Region	Location on chromosome	dbSNP code	No. of alleles
1	c.4021G>A	p.Gly1341Ser	Missense	LP	Exon 19	52511412	587783317	5
2	c.3191A>C	p.Glu1064Ala	Missense	LP	Exon 14	52518297	374094065	4
3	c.2165dupT	p.Arg723GlufsTer32	Frameshift	Pathogenic	Exon 8	52532636	587783302	3
4	c.3247C>T	p.Leu1083Phe	Missense	Pathogenic	Exon 15	52516687	1286080173	3
5	c.3646G>A	p.Val1216Met	Missense	Pathogenic	Exon 17	52513240	776280797	2
6	c.2827G>A	p.Gly943Ser	Missense	Pathogenic	Exon 12	52523836	28942076	2
7	c.2828G>A	p.Gly943Asp	Missense	Pathogenic	Exon 12	52523835	779323689	2
8	c.3317T>A	p.Val1106Asp	Missense	LP	Exon 15	52516617	775541743	2
9	c.2182A>G	p.Asn728Asp	Missense	LP	Exon 8	52532620	1459023925	2
10	c.3007G>A	p.Ala1003Thr	Missense	Pathogenic	Exon 13	52520473	201497300	2
11	c.2128G>A	p.Gly710Ser	Missense	Pathogenic	Exon 8	52532674	137853285	2
12	c.4148G>C	p.Arg1383Thr	Missense	LP	Exon 21	52509142	769604207	2
13	c.3529C>T	p.Gln1177Ter	Stop-gain	Pathogenic	Exon 16	52515244	1057516479	2
14	c.2924C>A	p.Ser975Tyr	Missense	Pathogenic	Exon 13	52520556	778163447	2
15	c.2558A>g	p.Asp853Gly	Missense	LP	Exon 10	52524425	752634617	2
16	c.3263T>A	p.Leu1088Ter	Stop-gain	Pathogenic	Exon 15	52516671	753250853	2
17	c.2810delT	p.Val937GlyfsTer5	Frameshift	LP	Exon 12	52523852	770532598	1
18	c.3895C>T	p.Leu1299Phe	Stop-gain	Pathogenic	Exon 18	52511620	749472361	1
19	c.22A>G	p.Ile8Val	Missense	VUS	Exon 1	52585452	1322672752	1
20	c.3809A>G	p.Asn1270Ser	Missense	Pathogenic	Exon 18	52511706	121907990	1
21	c.2975C>T	p.P992Leu	Missense	Pathogenic	Exon 13	52520505	201038679	1
22	c.3140A>T	p.Asp1047Val	Missense	LP	Exon 14	52518348	1395504465	1
23	c.1840G>T	p.Gly614Cys	Missense	VUS	Exon 5	52539037	376565432	1
24	c.2930C>T	p.Thr977Met	Missense	Pathogenic	Exon 13	52520550	72552255	1
25	c.2131G>T	p.Gly711Trp	Missense	Pathogenic	Exon 8	52532671	1394999756	1
26	c.2549C>T	p.Thr850Ile	Missense	Pathogenic	Exon 10	52524434	777629392	1
27	c.956delC	p.Pro319HisfsTer44	Frameshift	Pathogenic	Exon 2	52548399	753674382	1
28	c.4166A>G	p.His1389Arg	Missense	VUS	Exon 21	52509124	1170033995	1
29	c.3053C>T	p.Ala1018Val	Missense	LP	Exon 13	52520427	371840514	1
30	c.2215A>G	p.Ile739Val	Missense	VUS	Exon 8	52532587	1469592954	1
31	c.4022G>A	p.Gly1341Asp	Missense	Pathogenic	Exon 20	52509831	779494870	1
32	c.4094C>T	p.Ser1365Phe	Missense	VUS	Exon 20	52509759	747301758	1
33	c.2363C>T	p.Thr788Ile	Missense	Pathogenic	Exon 9	52531736	541408630	1
34	c.1741A>T	p.Asn851Tyr	Missense	VUS	Exon 5	52539136	756435919	1
35	c.3556+1G>A	–	Splice-donor	Pathogenic	Intron 16	52515216	184388696	1
36	c.2297C>T	p.Thr766Met	Missense	LP	Exon 8	52532505	121907997	1
37	c.3716T>G	p.Val1239Gly	Missense	LP	Exon 18	52511799	374628199	1
38	c.1349delA	p.Asp450ValfsTer48	Frameshift	LP	Exon 3	52544821	764128100	1
39	c.3243+1G>A	–	Splice-donor	Pathogenic	Intron 14	52518244	748819198	1
40	c.2804C>T	p.Thr935Met	Missense	Pathogenic	Exon 12	52523859	750019452	1
41	c.1877G>C	p.Gly626Ala	Missense	LP	Exon 6	52536042	587783299	1
42	c.3964C>T	p.Arg1322Cys	Missense	VUS	Exon 19	52511469	534984209	1
43	c.2332C>T	P.Arg778Trp	Missense	Pathogenic	Exon 8	52532470	137853284	1
44	c.2790-2792delCAT	p.Ile930del	Inframe deletion	Pathogenic	Exon 12	52523870	1228359983	1
45	c.2532delA	p.Val845SerfsTer28	Frameshift	Pathogenic	Exon 10	52524450	755709270	1
46	c.3818C>T	p.Pro1273Leu	Missense	Pathogenic	Exon 18	52511697	758355520	1
47	c.3451C>T	p.Arg1151Cys	Missense	Pathogenic	Exon 16	52515322	755554442	1
48	c.3688A>G	p.Ile1230Val	Missense	VUS/LP	Exon 17	52513198	200911496	1
49	c.1520-1523delAAAG	p.Glu507GlyfsTer16	Frameshift	Pathogenic	Exon 3	52544647	769128344	1
50	c.3741-3742dupCA	p.Lys1248ThrfsTer83	Frameshift	Pathogenic	Exon 18	52511772	1462451206	1
51	c.1145-1151delCCCAACT	p.Ser382TrpfsTer24	Frameshift	Pathogenic	Exon 2	52548204	1176709391	1
52	c.3008C>T	p.Ala1003Val	Missense	Pathogenic	Exon 13	52520472	775055397	1
53	c.2675A>G	p.Asn892Ser	Missense	VUS	Exon 11	52524198	779694030	1
54	c.314C>A	p.Ser105Ter	Stop-gain	Pathogenic	Exon 2	52549042	753236073	1
55	c.1670A>G	p.Glu557Gly	Missense	VUS	Exon 4	52542617	769605349	1
56	c.2998G>A	p.Gly1000Arg	Missense	Pathogenic	Exon 13	52520482	751078884	1
57	c.1847G>A	p.Arg616Gln	Missense	Pathogenic	Exon 5	52539030	752850609	1
58	c.1531C>T	p.Gln511Ter	Stop-gain	Pathogenic	Exon 3	52544640	1449610384	1
59	c.2513delA	p.Lys838SerfsTer35	Frameshift	Pathogenic	Exon 10	52524469	777362050	1

LP = likely pathogenic; VUS = variant of uncertain significance.

Exons 8 and 13 harbour 7 (11.9%) variants each; exons 12 and 18 harbour 5 (8.5%) variants each; exon 10 harbours 4 (6.8%) variants; exons 2, 3, 5 and 15 harbour 3 (5.1%) variants each; exons 14, 16, 17, 19, 20 and 21 harbour 2 (3.4%) variants each; exons 4, 6, 9 and 11 harbours 1 (1.7%), while introns 14 and 16 harbour 1 (1.7%) variant each. No variants were detected in exon 7 [Fig. 1].

Fig. 1.

Variant distribution on all exons and intron of the ATP7B gene in patients with Wilson disease.

Based on the zygosity status and the presence of additional variants, the 45 recruited patients with WD were grouped into 3 groups. Group 1 had 23 (51.1%) patients with biallelic deleterious variants and a definitive diagnosis of WD. It included 8 homozygotes, 1 of whom had 1 additional pathogenic variant and 1 had an additional variant with conflicting pathogenicity and 15 compound heterozygotes [Table 3]. Group 2 included 12 (26.7%) patients who were heterozygotes for 1 deleterious variant with an additional variant. Of them, 9/12 had VUS and 3/12 had a variant with a conflicting interpretation by ClinVar. These three variants were shown to affect protein function and/or RNA splicing or amino acid residue. The other 9/12 were VUS; 5 of them are expected to disrupt protein function, while 4 are not expected to disrupt protein function [Table 4,Table 5]. Group 3 included 10 (22.2%) heterozygote patients [Table 6]. The molecular status of the 45 enrolled patients was also captured [Supplementary Table 2].

Table 3.

Variants detected in homozygous and compound heterozygous state in 23 patients with Wilson disease.

							MLS score
Pt. No.	Pts' IDs	First variant	Effect/classification	Second variant	Effect/ClinVar classification	Presentation	Before NGS	After NGS
Patients with homozygous variants (n = 8)
1	2	c.4021G>A (p.Gly1341Ser)	Missense/ pathogenic	c.4021G>A (p.Gly1341Ser)	Missense/ pathogenic	Hepatic	3	7
2	6	c.3191A>C (p.Glu1064Ala)	Missense/LP	c.3191A>C (p.Glu1064Ala)	Missense/LP	Mixed	5	9
3	10	c.2182A>G (p.Asn728Asp)	Missense/LP	c.2182A>G (p.Asn728Asp)	Missense/LP	Hepatic	2	6
4	11	c.3247C>T (p.Leu1083Phe)	Missense/ pathogenic	c.3243+1G>A	Splice-donor/LP	Neurologic	4	8
5	17	c.2128G>A (p.Gly710Ser)	Missense/ pathogenic	c.2128G>A (p.Gly710Ser)	Missense/ pathogenic	Hepatic	3	7
6	19	c.4148G>C (p.Arg1383Thr)	Missense/LP	c.4148G>C (p.Arg1383Thr)	Missense/LP	Mixed	3	7
7	20	c.3529C>T (p.Gln1177Ter)	Stop-gain/ pathogenic	c.3529C>T (p.Gln1177Ter)	Stop-gain/ pathogenic	Hepatic	3	7
8	29	c.2165dupT (p.Arg723GlufsTer32)	Single base duplication/ Pathogenic	c.2558A>G (p.Asp853Gly)	Missense/ conflicting	Hepatic	2	6
Patients with compound heterozygous variants (n = 15)
9	1	c.4021G>A (p.Gly1341Ser)	Missense/LP	c.3809A>G (p.Asn1270Ser)	Missense/ pathogenic	Hepatic	3	7
10	3	c.3646G>A (p.Val1216Met)	Missense/ pathogenic	c.2975C>T (p.Pro992Leu)	Missense/ pathogenic	Hepatic	4	8
11	4	c.2930C>T (p.Thr977Met)	Missense/ pathogenic	c.2131G>T (p.Gly711Trp)	Missense/ pathogenic	Hepatic	2	6
12	5	c.2549C>T (p.Thr850Ile)	Missense/ pathogenic	c.956delC (p.Pro319HisfsTer44)	Frameshift del./ pathogenic	Hepatic	2	6
13	7	c.4021G>A (p.Gly1341Ser)	Missense/LP	c.2810delT (p.Val937GlyfsTer5)	Frameshift/LP	Hepatic	2	6
14	8	c.2828G>A (p.Gly943Asp)	Missense/ pathogenic	c.2513delA (p.Lys838SerfsTer35)	Frameshift/ pathogenic	Hepatic	2	6
15	9	c.3556+1G>A	Splice-donor/ pathogenic	c.3191A>C (p.Glu1064Ala)	Missense/LP	Hepatic	2	6
16	12	c.3191A>C (p.Glu1064Ala)	Missense/LP	c.2332C>T (p.Arg778Trp)	Missense/ pathogenic	Hepatic	2	6
17	13	c.2828G>A (p.Gly943Asp)	Missense/ pathogenic	c.2790-2792delCAT (p.Ile930del)	In frame-del./LP	Neurologic	4	8
18	14	c.3007G>A (p.Ala1003Thr)	Missense/ pathogenic	c.2532delA (p.Val845SerfsTer28)	Frameshift/ pathogenic	Hepatic	7	11
19	15	c.3818C>T (p.Pro1273Leu)	Missense/ pathogenic	c.3451C>T (p.Arg1151Cys)	Missense/ pathogenic	Psychiatric	3	7
20	16	c.3741-3742dupCA (p.Lys1248ThrfsTer83)	Frameshift/ pathogenic	c.2827G>A (p.Gly943Ser)	Missense/ pathogenic	Hepatic	4	8
21	18	c.3247C>T (p.Leu1083Phe)	Missense/ pathogenic	c.3008C>T (p.Ala1003Val)	Missense/ pathogenic	Hepatic	2	6
22	21	c.4021G>A (p.Gly1341Ser)	Missense/LP	c.3263T>A (p.Leu1088Ter)	Stop-gain/ pathogenic	Hepatic	6	10
23	22	c.2998G>A (p.Gly1000Arg)	Missense/ pathogenic	c.1847G>A (p.Arg616Gln)	Missense/ pathogenic	Hepatic	3	7

Patient ID 11 had two variants, the first (c.3247C>T) was in a homozygous state and the second (c.3243+1G>A) was in a heterozygous state. Patient ID 29 had two variants, the first (c.2165dupT) was in a homozygous state and the second (c.2558A>G) was in heterozygous state.

NGS = next-generation sequencing; MLS = modified Leipzig score; LP = likely pathogenic.

Table 4.

Variant characteristics in 12 patients with Wilson disease with 2 variants (a disease-causing variant plus a variant of uncertain significance or a variant with conflicting pathogenicity).

	First variant (disease-causing)		Second variant			MLS score
Patient's ID	Type	Effect/classification	Type	Effect/ClinVar classification	Presentation	Before NGS	After NGS
23	c.3895C>T (p.Leu1299Phe)	Stop-gain/ pathogenic	c.22A>G (p.Ile8Val)	Missense/VUS	Hepatic	2	3
24	c.3140A>T (p.Asp1047Val)	Missense/LP	c.1840G>T (p.Gly614Cys)	Missense/VUS	Neurologic	4	5
25	c.2827G>A (p.Gly943Ser)	Missense/ pathogenic	c.4166A>G (p.His1389Arg)	Missense/VUS	Hepatic	2	3
26	c.2924C>A (p.Ser975Tyr)	Missense/ pathogenic	c.2558A>G (p.Asp853Gly)	Missense/VUS/LP	Hepatic	1	5
27	c.3053C>T (p.Ala1018Val)	Missense/LP	c.2215A>G (p.Ile739Val)	Missense/VUS	Hepatic	3	4
28	c.4022G>A (p.Gly1341Asp)	Missense/ pathogenic	c.4094C>T (p.Ser1365Phe)	Missense/VUS	Hepatic	3	4
30	c.3317T>A (p.Val1106Asp)	Missense/LP	c.1741A>T (p.Asn581Tyr)	Missense/VUS	Hepatic	2	3
31	c.1349delA (p.Asp450ValfsTer48)	Frameshift/LP	c.3716T>G (p.Val1239Gly)	Missense/VUS LP	Asymptomatic	1	5
32	c.1877G>C (p.Gly626Ala)	Missense/LP	c.3964C>T (p.Arg1322Cys)	Missense/VUS	Hepatic	4	5
33	c.3646G>A (p.Val1216Met)	Missense/ pathogenic	c.3688A>G (p.Ile1230Val)	Missense/VUS/LP	Hepatic	5	6
34	c.314C>A (p.Ser105Ter)	Stop-gain/ pathogenic	c.2675a>G (p.Asn892Ser)	Missense/VUS	Hepatic	7	8
45	c.3263T>A (p.Leu1088Ter)	Stop-gain/pathogenic	c.1670A>G (p.Glu557Gly)	Missense/VUS	Neurologic	5	6

NGS = next-generation sequencing; VUS = variant of uncertain significance; MLS = modified Leipzig score; LP = likely pathogenic.

Table 5.

Detailed analysis for the second variants (variant of uncertain significance and conflicting pathogenicity) from ClinVar, dbSNP and Franklin databases.

	Variants of uncertain significance/variants with conflicting pathogenicity
Patient's ID	Type	Effect/ClinVar classification	SIFT Score	SIFT prediction/aggregated prediction	Franklin ACMG classification
23	c.22A>G (p.Ile8Val)	Missense/VUS	0.5	Benign supporting/uncertain	VUS (PM2, PP2)
24	c.1840G>T (p.Gly614Cys)	Missense/VUS	0.04	Uncertain/deleterious	LP (PP3, PM2, PM5, PP2)
25	c.4166A>G (p.His1389Arg)	Missense/VUS	0.2	Benign moderate/benign	VUS (PM2, PP2, BP4)
26	c.2558A>G (p.Asp853Gly)	Missense/VUS/LP	0.04	Uncertain/deleterious	LP (PP3, PM2, PM1, PP2)
27	c.2215A>G (p.Ile739Val)	Missense/VUS	0.4	Uncertain/uncertain	VUS (PM2, PM1, PP2)
28	c.4094C>T (p.Ser1365Phe)	Missense/VUS	0.2	Deleterious supporting/deleterious	LP (PM2, PM1, PP3, PP2)
30	c.1741A>T (p.Asn581Tyr)	Missense/VUS	0.07	Uncertain/deleterious	VUS (PM2, PP3, PP2)
31	c.3716T>G (p.Val1239Gly)	Missense/VUS/LP	0.02	Deleterious supporting/deleterious	Pathogenic (PM3, PP3, PM2,PP2, PS3, PP5)
32	c.3964C>T (p.Arg1322Cys)	Missense/VUS	0.03	Deleterious supporting/deleterious	LP (PM2, PM1,PP3, PM5, PP2)
33	c.3688A>G (p.Ile1230Val)	Missense/VUS/LP	0.03	Uncertain/deleterious	Pathogenic (PM3, PM2, PM5, PM1, PP3, PP2,PP5)
34	c.2675A>G (p.Asn892Ser)	Missense/VUS	0.04	Benign moderate/benign	VUS (PM2, PM1,PP2, BP4)
45	c.1670A>G (p.Glu557Gly)	Missense/VUS	0.3	Uncertain/deleterious	VUS (PM2, PP3, PP2)

VUS = variant of uncertain significance; PM = pathogenic moderate; PP = pathogenic supporting; BP = benign supporting; LP = likely pathogenic.

Table 6.

Variant characteristics, clinical presentation and modified Leipzig scores of 10 patients with suggestive features of Wilson disease and a single heterozygous pathogenic variant.

						MLS score
Patients' IDs	Nucleotide change	Codon change	Variant effect	Variant classification	Presentation	Before NGS	After NGS
35	c.2165dupT	p.Arg723GlufsTer32	Frameshift	Pathogenic	Hepatic	2	3
36	c.2363C>T	p.Thr788Ile	Missense	Pathogenic	Hepatic	3	4
37	c.2297C>T	p.Thr766Met	Missense	LP	Hepatic	6	7
38	c.3007G>A	p.Ala1003Thr	Missense	Pathogenic	Hepatic	2	3
39	c.2804C>T	p.Thr935Met	Missense	Pathogenic	Hepatic	3	4
40	c.1531C>T	p.Gln511Ter	Stop-gain	Pathogenic	Neurologic	2	3
41	c.1520-1523delAAAG	p.Glu507GlyfsTer16	Frameshift	Pathogenic	Hepatic	2	3
42	c.1145-1151delCCCAACT	p.Ser382TrpfsTer24	Frameshift	Pathogenic	Hepatic	3	4
43	c.3317T>A	p.Val1106Asp	Missense	LP	Hepatic	2	3
44	c.2924C>A	p.Ser975Tyr	Missense	LP	Neurologic	3	4

MLS = modified Leipzig score; LP = likely pathogenic.

Pure neurologic manifestations were present in 6 (13.3%) patients; 2 were compound heterozygotes, 2 with a disease-causing variant plus a VUS and 2 were carriers for a single pathogenic variant. A patient (ID = 15) had a pure psychiatric manifestation and was found to be compound heterozygous for missense pathogenic variants (p.Pro1273Leu / p.Arg1151Cys). Another patient (ID = 6) presented at the age of 14 with mixed hepatic and neurologic manifestations plus pancytopenia. She was molecularly tested at the age of 19 years and shown to be a homozygote for a likely pathogenic variant c.3191A>C (p.Glu1064Ala).

4. Discussion

The diagnosis of WD is primarily based on clinical presentation, supported by characteristic biochemical abnormalities.5 In Iraq, molecular diagnostics, measurement of hepatic total copper content and assessment of serum free exchangeable copper are not routinely performed in suspected cases. Furthermore, molecular data on ATP7B gene variants in this region remain limited.

In the current study, male patients slightly outnumbered females, consistent with findings from previous local studies.12,13,14 A notable majority of patients (75.6%) were offspring of consanguineous unions. The overall consanguinity rate in Iraq is high, ranging between 47% and 60%, as reported in various local studies, and between 40% and 50% in Arab Gulf countries.14,15,16 Given the autosomal recessive inheritance of WD, a high prevalence of consanguinity among affected individuals is expected and has been documented in Iraqi and Turkish WD cohorts.12,13,16,17,18

In alignment with global data, most patients in this study (77.8%) presented with hepatic manifestations only.5

In terms of ethnicity, all patients enrolled in this study were Iraqi Arabs except for 3 Kurdish patients. The first patient (ID = 1) was a 24-year-old male, born to non-consanguineous parents who had a family history (his cousin) of WD. He presented with hepatic manifestations; molecular analysis showed a compound heterozygous state with two pathogenic variants (c.4021G>A and c.3809A>G). The second patient (ID = 33) was a 23-year-old male, born to consanguineous parents with a family history (deceased sister) of WD. At the age of 15 years, he presented with hepatic manifestations; molecular analysis showed the presence of 2 variants, the first one (c.3645G>A) was pathogenic and the second one (c.3688A>G) had conflicting pathogenicity. The third patient (ID = 44) was a 30-year-old male, had consanguineous parents and a negative family history. At the age of 20 years, he presented with neurologic signs without KF ring. Molecular study revealed the presence of a single pathogenic variant (c.2924C>A).

The 2 variants, c.3809A>g and c.3688A>G, were only found in Kurds; whether these variants are unique to Kurds or not requires a more comprehensive analysis. Previous research in Iraq has investigated ATP7B variants in children and another study documented the clinical and epidemiological features of WD across all age groups.14,19

In the current adult cohort, 59 variants were identified across 80 of 90 alleles, with 43 out of the 58 variants observed in only 1 allele. This high degree of allelic heterogeneity, even among consanguineous individuals, aligns with findings from both local and international studies.13,14 These findings emphasise the extensive variability of ATP7B variants, which now exceed 1,275 globally, and illustrates that many heterozygous carriers remain asymptomatic. Compared with the previous Iraqi paediatric cohort, which reported 76 variants among 55 children, this study on adults identified 17 new variants.14 Together, these studies have uncovered 118 unique variants among 100 Iraqi patients, reinforcing the genetic diversity of ATP7B mutations in this population.

Missense mutations were the most frequently observed variant type, followed by frameshift, stop-gain, splice-donor and in-frame deletion mutations, mirroring local and global findings.7,14 Despite the recurrence of some variants in multiple alleles, the globally common H1069Q variant was not observed, consistent with prior Iraqi studies.17

The most frequently identified variant in this study, p.Gly1341Ser, was previously reported in a homozygous state in an Iraqi child as well as in Saudi Arabia and South India.14,21,22 The second most common variant, p.Glu1064Ala, has been documented in the United Kingdom.22 Other variants, such as p.Leu1083Phe and p.Arg723GlufsTer32 have been previously reported; the former in Korea and China while the latter in Poland.24,25 Other remaining variants have been identified globally.

Based on genetic findings, patients were grouped into 3 categories: Group 1 patients with biallelic variants (n = 23) included both 8 homozygous and 15 compound heterozygous individuals. In all cases, the diagnosis was clear-cut. Detected variants were either pathogenic or likely pathogenic per ACMG guidelines or had been previously reported in symptomatic patients with WD. Notably, 1 patient (ID = 15), a 20-year-old female, presented with isolated psychiatric symptoms and a positive KF ring. She had compound heterozygous pathogenic variants: p.Pro1273Leu and p.Arg1151Cys. Additionally, 2 patients (ID = 11 and ID = 13) exhibited purely neurological symptoms without hepatic involvement. Furthermore, a 57-year-old female (ID = 11), had no history of consanguinity but had affected family members. She had compound heterozygous variants: p.Leu1083Phe and splice-donor variant c.3243+1G>A. Another patient (ID = 13), a 24-year-old female, had consanguineous parents and presented with dementia and weakness. She carried p.Gly943Asp and p.Ile930del variants.

Group 2 included 11 symptomatic and one asymptomatic patient. All have 1 known pathogenic variant and an additional VUS or conflicting interpretation; 10 had hepatic and 2 had pure neurologic manifestations and 5 of them had an MLS score of ≥4 prior to sequencing.

ACMG criteria for variant pathogenicity emphasises minor allele frequency, co-segregation and in silico predicitions as important factors in assessment.26 Tools such as SIFT and PolyPhen-2 were used in this study. Based on Franklin's variant classification system, 8 of 12 secondary variants were predicted to be deleterious and 4 to be benign. Given their clinical presentation, biochemical profile and high Leipzig scores (≥4), these 8 VUSs are proposed for reclassification as likely pathogenic. For example, a 41-year-old pregnant woman (ID = 32), who had jaundice and a Leipzig score of 4, responded well to trientine. She had p.Gly626Ala (likely pathogenic) and p.Arg1322Cys (VUS). In silico analysis suggested p.Arg1322Cys is deleterious and may disrupt protein function. A 41-year-old male (ID = 34) had liver cirrhosis and underwent transplantation. He had p.Ser105Ter (pathogenic) and p.Asn892Ser (VUS), the latter showed potential splicing disruption in computational models. A 23-year-old male (ID = 33) had a family history of WD. He carried p.Val1216Met (pathogenic) and p.Ile1230Val (VUS/likely pathogenic), also previously reported in an asymptomatic adult with a unilateral KF ring and affected offspring.27 It must be stated clearly that accumulated evidence from patients' clinical data, allele frequency, in silico tools as well as functional studies must be considered together before variant pathogenicity is confirmed.

Group 3 included 10 heterozygous patients with no additional variant; 8 patients in this group had hepatic and 2 had neurologic manifestations. They all showed biochemical abnormalities consistent with WD. A total of 5 had an MLS ≥4 and the remaining 5 had a score of 3. For example, a 56-year-old male (ID = 37) with fatty liver and suggestive biochemistry, had a family history of liver disease and an MLS score of 7 after sequencing. A diagnosis of a manifesting carrier was considered. Further testing with whole-exome sequencing (WES) or epigenetic profiling may be required. MLS scores of 4 were found in 4 other patients; 3 had hepatic and 1 had neurologic manifestations. These cases warrant further molecular analysis, including WES, to examine other related genes. The remaining 5 patients had Leipzig scores <4, making WD unlikely. Investigation for other metabolic liver disorders is recommended.

Identification of a single pathogenic variant confirms WD only in the presence of definitive clinical and biochemical abnormalities.5 In this group, such criteria were not consistently met.

ATP7B remains the only gene known to cause WD.3 Recent studies suggest that heterozygous ATP7B variants are more common than previously believed.28,29 Manifesting heterozygotes have been reported in studies from Iraq, Paris and China, though these were not population-based.14,30,31 Some evidence suggests carriers may exhibit intermediate phenotypes with only biochemical disturbances.20 However, in this study, group 3 patients were symptomatic.

Reduced diagnostic yield for WD occurs due to the inherent complexities of the disease, including incomplete penetrance and variable expressivity of genetic mutations, making the typical clinical and biochemical features inconsistent or absent. Additionally, suboptimal or incomplete diagnostic tests, the lack of definitive biomarkers, difficulties in interpreting genetic findings and potential influences from genetic modifiers, epigenetics and lifestyle habits contribute to the underdiagnosis of WD, even when clinical suspicion is high.32

Based on these findings, when biallelic pathogenic variants are identified, the diagnosis is confirmed. In contrast, the presence of a single heterozygous variant—particularly with a high Leipzig score—requires further exploration for variants in regulatory regions or other genes.

This study was subject to certain limitations. Functional studies were not performed for the detected VUS or those with conflicting pathogenicity. Other related genes that may explain the manifesting heterozygotes were also not studied. Additionally, non-manifesting parents, siblings and other close family members were not tested.

5. Conclusion

The mutational spectrum of ATP7B is markedly heterogeneous, even in the context of high consanguinity. This study suggests that 8 variants merit reclassification as deleterious. Despite the use of direct gene sequencing, diagnostic uncertainty remains, particularly in patients with Leipzig scores ≥4 who possess only 1 deleterious variant. The distinction between WD and manifesting carriers in such cases is still unresolved. Routine measurement of free exchangeable copper, further molecular testing (e.g., WES or whole-genome sequencing), functional studies to validate VUS pathogenicity and family-based molecular screening are recommended. These steps can enhance diagnostic accuracy and guide more effective clinical management of patients with WD.

Authors' Contribution

Ruqayah Gheyath Yaseen Al-Obaidi: Investigation, Formal analysis, Resources, Data curation, Writing – Original Draft, Funding acquisition. Bassam Musa Sadik Al-Musawi: Conceptualization, Methodology, Design, Data curation, Writing - Review & Editing, Supervision.

Ethics Statement

This study adhered to the ethical standards of the Declaration of Helsinki and was approved by the Research Ethics Committee of the Department of Pathology & Forensic Medicine, College of Medicine, University of Baghdad (Approval No. 18, dated January 19, 2023). Informed consent for the use and publication of medical data was obtained from all participants.

Conflict of Interest

The authors declare no conflicts of interest.

Funding

The first author has fully and personally funded the costs of this study.

Data Availability

Data is available upon reasonable request from the corresponding author.