Secondary Logo

Journal Logo

Original Articles: Hepatology

Use of a Comprehensive 66-Gene Cholestasis Sequencing Panel in 2171 Cholestatic Infants, Children, and Young Adults

Karpen, Saul J.; Kamath, Binita M.; Alexander, John J.‡,§; Ichetovkin, Ilia||; Rosenthal, Philip; Sokol, Ronald J.#; Dunn, Shelley||; Thompson, Richard J.∗∗; Heubi, James E.††

Author Information
Journal of Pediatric Gastroenterology and Nutrition: May 2021 - Volume 72 - Issue 5 - p 654-660
doi: 10.1097/MPG.0000000000003094

Abstract

An infographic is available for this article at:https://links.lww.com/MPG/C250.

What Is Known/What Is New

What Is Known

  • Cholestasis can be caused by a diverse group of hepatobiliary diseases or disorders, many of which are genetic in origin.
  • Diagnosis of the underlying cause of cholestasis, particularly in infants and children, can be complex and challenging due to the broad overlap of presentations.
  • Targeted testing using next-generation sequencing may assist in diagnosing patients with cholestasis.

What Is New

  • Variants in 32 genes in this panel were identified as the cause of cholestasis in >2100 tested patients, highlighting the utility of a broad multi-gene approach to the diagnosis of cholestasis. The simultaneous analysis of multiple genes in a panel reduces the time to identify, or eliminate, specific genes as a cause of cholestasis.
  • Timely application of a multigene panel in cholestasis has the potential to reshape the clinician's diagnostic algorithm, offering the opportunity, with appropriate analyses, to revise the role and utility of liver biopsy.

Cholestasis can be caused by a diverse group of hepatobiliary diseases or disorders, of which many are genetic in origin, regardless of age (1). In particular, genetic disorders account for ∼25% of all cases of neonatal cholestasis (1). These monogenic disorders (resulting from one or more pathogenic variants in a single gene) include multisystem diseases such as Alagille syndrome (ALGS [affected genes: JAG1, NOTCH2]) (2,3); mitochondrial hepatopathies (POLG, MPV17, DGUOK) (4); hepatocellular diseases such as alpha-1-antitrypsin deficiency (A1ATD [SERPINA1]) (5); bile acid synthesis defects (HSD3B7, AKR1D1, and CYP27A1, among others) (6); transporter defects, including progressive familial intrahepatic cholestasis (PFIC [ATP8B1, ABCB11, ABCB4]) (7); biliary disorders (PKHD1, CFTR, NPHP) (8,9); or inborn errors of metabolism such as citrin deficiency (SLC25A13) (10) and Niemann–Pick type C (NPC1, NPC2) (1,11).

Cholestatic liver diseases in older children or adults may also have a genetic basis, including several due to variants in genes often thought to be limited to a neonatal presentation of cholestasis. For example, variants in ABCB11 are associated with benign recurrent intrahepatic cholestasis (12,13), whereas ABCB4 variants have been associated with intrahepatic cholestasis of pregnancy, low phospholipid-associated cholelithiasis, and PFIC (14–16). Bile acid synthesis defects and ALGS also can present or be diagnosed in older children and adults (6,17–19). Genetic variants also may play roles in complex liver diseases, including forms of intrahepatic cholestasis, gallstone development, fatty liver disease, drug-induced liver injury, and liver disease progression (20,21).

Diagnosis of the underlying cause of cholestasis in the infant and child can be complex and challenging due to the broad overlap of presentations. Reliance upon clinical presentation, routine serum chemistries, and other diagnostic assessments is frequently insufficient to distinguish between disorders (1,22). Liver biopsies are often utilized but may not be definitive in all cases; occasionally, more than one biopsy is required due to nonspecific and evolving histologic features (1).

Recently published guidelines note the potential utility of genetic testing in evaluating patients with cholestasis (1). Historically, genetic testing was limited to analysis of a single gene to confirm a diagnosis based on clinical observation and other assessments (ie, “phenotype-first”); however, given the diverse etiology and a large number of potential genetic causes of cholestasis, single-gene testing is impractical, inefficient, and expensive (23,24). Next-generation sequencing (eg, multigene panels, whole exome sequencing) now enables testing of larger numbers of genes simultaneously, leading to a trend towards utilizing genetic testing earlier in the diagnostic process, suggesting a “sequence-first” approach (1,23,24).

Previous studies in both adults and pediatrics have indicated the utility of multigene panels in patients with cholestasis (23–27), and a similar approach of gene panel testing has been used in other settings, for example, neurology, skeletal disorders, hearing loss, coagulation disorders, inborn metabolism errors, and intellectual disability (28–33). Targeted testing using a multigene panel may prove useful in real-world settings to assist in diagnosing patients with cholestasis, including those with an active or history of idiopathic cholestasis, unexplained chronic liver disease, or children listed for liver transplant without a definitive diagnosis.

The objective of this study was to analyze the initial findings from a broad North American-based genetic testing program available to a diverse group of specialists, designed to help clinicians diagnose potential causes of cholestasis.

METHODS

Beginning in February 2016, the Genetic Cholestasis Panel Testing Program (funding supported by Retrophin, Inc) was launched and is offered at no cost to qualified patients through their healthcare providers (HCPs). The testing program is conducted through EGL Genetics (formerly Emory Genetics Laboratory), a CAP/CLIA accredited clinical diagnostic laboratory. The initial panel included 57 genes; an additional nine genes (ALDOB, AMACR, DCDC2, EHHADH, GPBAR1, HSD17B4, SCP2, SLC10A1, and SLC10A2) were included in March 2017 (Table 1).

TABLE 1 - Genes included in the cholestasis sequencing panel
Causes of bile acid synthesis disorders due to single enzyme defects and cerebrotendinous xanthomatosis
AKR1D1 AMACR BAAT CYP7A1 CYP7B1
CYP27A1 DHCR7 HSD3B7 SLC27A5
Causes of peroxisomal disorders including Zellweger spectrum disorders
PEX1 PEX2 PEX3 PEX5 PEX6
PEX7 PEX10 PEX11B PEX12 PEX13
PEX14 PEX16 PEX19 PEX26
Other genetic causes of cholestasis
ABCB11 ABCB4 ABCC2 ABCG5 ABCG8
ALDOB ATP8B1 CC2D2A CFTR CLDN1
DCDC2 DGUOK EHHADH FAH GPBAR1
HNF1B HSD17B4 INVS JAG1 LIPA
MKS1 MPV17 NOTCH2 NPC1 NPC2
NPHP1 NPHP3 NPHP4 NR1H4 PKHD1
POLG SCP2 SERPINA1 SLC10A1 SLC10A2
SLC25A13 SMPD1 TJP2 TMEM216 TRMU
UGT1A1 VIPAS39 VPS33B
Genes added to the panel in March of 2017.

Infants and children were eligible if HCPs determined that the subject either was cholestatic, had a history of cholestasis without an identified cause, or had unexplained chronic liver disease. Patients with extrahepatic obstruction (eg, biliary atresia, choledochal cyst, and primary sclerosing cholangitis) and total parenteral nutrition-associated cholestasis were excluded. All patients or parents/guardians provided written informed consent. Limited clinical information was collected at the time of sample submission. Data requested included patient demographic information, 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) codes, and HCP specialty. HCPs collected whole blood or saliva samples, which were submitted (overnight shipping via mail in a pre-addressed envelope) along with completed enrollment form and signed patient informed consent.

In-solution hybridization with a custom SureSelect capture library (Agilent) was used to sequence the genes on the panel. Next-generation sequencing (2 × 100 bp, paired-end) was carried out using an Illumina HiSeq 2500 in rapid run mode (Illumina), FASTQ files were generated, and duplicate reads were removed. The individual DNA sequence reads were aligned to the published human genome reference (hg19 build), and variants were called (NextGENe SoftGenetics). Further analysis was performed using the custom EGL bioinformatics pipeline, which also performs additional quality metric calculations and annotates the identified variants utilizing a variety of external and internal sources. The sequence was analyzed within the coding exons and at least ±10 bp into the introns (where feasible to sequence context).

Variants identified were reviewed and interpreted by board-certified clinical molecular geneticists (American Board of Medical Genetics and Genomics). Variants were classified according to the EGL clinical interpretation guidelines that incorporate components including, but not limited to, expertise, professional guidelines (eg, American College of Medical Genetics and Genomics [ACMG]), published literature, known mutational spectrum, in silico predictions, variant databases (both professional and public), and population frequency information (eg, gnomAD). Sequence variations were typically classified in one of five categories: benign, likely benign, variant of unknown significance (VOUS), likely pathogenic (LP), or pathogenic (P).

For purposes of these analyses, we characterized “diagnostic” biallelic gene variants as pathogenic or likely pathogenic (P/LP) for autosomal recessive diseases, or P/LP variants for the autosomal dominant disease ALGS (genes JAG1 and NOTCH2). We characterized as “potentially diagnostic” genes for autosomal recessive diseases with 1 P/LP + VOUS.

Deletion/Duplication Analysis

A limited subset of the individuals in this cohort (104 samples) were tested using a custom-designed (Oxford Gene Technology, Oxfordshire, UK) gene-targeted array comparative genomic hybridization (aCGH) to assess for single and multi-exon level deletions/duplications of the original 57 genes on the panel.

RESULTS

Samples and General Information

A total of 2433 samples were submitted between February 25, 2016 and December 31, 2017 (Fig. 1 and Table 1, Supplemental Digital Content, https://links.lww.com/MPG/C247): 2171 of the 2433 submitted samples were analyzed and 262 samples were untested for various reasons including duplicate samples from the same subject; the insufficient amount of sample (blood or saliva); unlabeled samples submitted and cancelled by the laboratory; cancellations by the provider; samples that failed DNA extraction; samples that failed NGS due to quality; and incomplete paperwork. Six hundred and eighty-five samples were analyzed using the 57-gene panel (between February 25, 2016 and March 2017) and 1486 using the updated 66-gene panel (March 2017 through December 31, 2017). Median turnaround time (from receipt of samples in the laboratory to return of reports to clinicians) was 21 days (mean ± SD: 21.6 ± 8.4). Information on diagnostic codes submitted with samples is included in Text, Supplemental Digital Content, https://links.lww.com/MPG/C247.

Patient Demographics

Of the 2171 subjects, 94% were <18 years old, and the majority (n = 1230, 57%) were <1 year old (Table 2); the youngest was 1 day old. For the entire cohort, 59% were male; racial and ethnic characteristics were provided in 79% of subjects. The entire cohort is generally reflective of the population of North America (Table 2, Supplemental Digital Content, https://links.lww.com/MPG/C247). Of note, clinicians sent samples more often from African-American children in the youngest age group, <3 months (21%), than those >3 months (11.6%).

TABLE 2 - Demographic characteristics
Characteristic TotalN = 2171 Characteristic TotalN = 2171
Age range (y) 0–84 Sex, n (%)
Age categories, n (%) Female 843 (38.8)
 <1 month 257 (11.8) Male 1289 (59.4)
 1–<2 months 300 (13.8) Ethnicity, n (%)
 2–<3 months 306 (14.1) Caucasian/Northwestern European 624 (28.7)
 3–<4 months 109 (5.0) Hispanic 430 (19.8)
 4–<5 months 71 (3.3) African American 333 (15.3)
 5–<6 months 61 (2.8) Asian 82 (3.8)
 6–<12 months 126 (5.8) Native American 28 (1.3)
 1 year 85 (3.9) Other 216 (9.9)
 2–5 years 258 (11.9) Unknown 458 (21.1)
 6–10 years 169 (7.8)
 11–15 years 181 (8.3)
 16–17 years 85 (3.9)
 >18 years 163 (7.5)

Variants

Analysis of the 2171 samples yielded 583 P variants (in 521 subjects), 79 LP variants (in 77 subjects), and 3117 VOUS (in 1636 subjects) (Table 3 and Tables 3 and 4 for specific variants, Supplemental Digital Content, https://links.lww.com/MPG/C247). Across all age subgroups (<3, 3–<12, >12 months), no significant differences overall were found in the percentages of subjects with P variants (23%, 21%, and 26%, respectively; P = 0.0858), LP variants (3%, 4%, and 4%, respectively; P = 0.6935), or VOUS (25%, 25%, and 29%, respectively; P = 0.0851). In addition, there were no significant differences between subjects <3 vs >3 months old in the percentage with >1 P variant (23% vs 25%; P = 0.2542), with >1 LP variant (3% vs 4%; P = 0.3922), or with >1 P/LP variant (25% vs 28%; P = 0.1491). One hundred and sixty-six novel P/LP variants and 415 VOUS (with 496 previously known P/LP variants and 2702 VOUS) were identified among the 66 genes. In sum, 658 P/LP variants were found in 580 of the 2171 subjects (27%), along with a mean of 1.4 genes with VOUS per subject.

TABLE 3 - Frequency of pathogenic and likely pathogenic variants and diagnostic yield (number of patients) of 66 gene cholestasis panel
Frequency of pathogenic and likely pathogenic variants
Pathogenic Likely pathogenic Pathogenic or likely pathogenic
Genes with >1 variant, n 583 79 658
Subjects with >1 variant, n 521 77 580
 One variant, n (%) 432 (83) 74 (96) 477 (82)
 Two variants, n (%) 81 (16) 3 (4) 88 (15)
 Three variants, n (%) 8 (2) 0 15 (3)
Patients with >1 pathogenic and/or likely pathogenic variant
Gene Definite diagnosis Potential diagnosis Monoallelic Gene Definite diagnosis Potential diagnosis Monoallelic
ABCB11 14 14 11 NPC1 6 4 6
ABCB4 4 9 18 NPHP1 0 0 2
ABCC2 11 8 22 NPHP3 0 1 3
ABCG5 0 0 2 NPHP4 0 0 5
ABCG8 1 1 14 NR1H4 0 0 3
AKR1D1 3 0 1 PEX1 3 1 4
ALDOB 2 0 9 PEX10 0 0 2
ATP8B1 4 2 4 PEX11B 0 0 1
CC2D2A 0 0 14 PEX12 1 1 2
CFTR 4 29 82 PEX19 0 0 1
CYP27A1 3 3 10 PEX2 0 0 1
CYP7B1 0 0 6 PEX26 5 0 0
DCDC2 6 0 1 PEX5 0 0 1
DGUOK 1 1 1 PEX6 1 0 5
DHCR7 0 0 28 PKHD1 0 2 9
FAH 0 0 1 POLG 10 1 13
HNF1B 0 0 2 SERPINA1 21 1 78
HSD3B7 6 0 3 SLC25A13 2 0 1
INVS 0 1 0 SMPD1 1 0 4
JAG1 53 n/a n/a TJP2 4 2 1
LIPA 1 1 11 TMEM216 0 0 1
MKS1 0 0 2 UGT1A1 1 1 4
MPV17 1 1 3 VIPAS39 1 0 1
NOTCH2 9 n/a n/a VPS33B 2 0 1
Genes with no pathogenic or likely pathogenic variants
AMACR CYP7A1 HSD17B4 PEX14 PEX7 SLC10A2
BAAT EHHADH NPC2 PEX16 SCP2 SLC27A5
CLDN1 GPBAR1 PEX13 PEX3 SLC10A1 TRMU
All genes are defined as contributing to disease in an autosomal recessive, biallelic fashion except JAG1 and NOTCH2 (which are considered autosomal dominant).
Definite diagnosis included genes with two alleles that were pathogenic or likely pathogenic (homozygous or heterozygous), or a single pathogenic or likely pathogenic allele for JAG1 or NOTCH2.
Potential diagnosis included genes with one pathogenic/likely pathogenic allele + one VOUS.

Of the 66 genes in the panel, 29 contributed to a “definite diagnosis” (Table 3). A total of 119 subjects with biallelic P/LP gene variants (autosomal recessive) and 62 with ALGS (autosomal dominant) were reported, totaling 181 (8%) with a definite diagnosis. An additional 84 subjects had a potential diagnosis (one P/LP allele + one VOUS [autosomal recessive]), leading to a combined definite + potential diagnosis in 265 (12%; encompassing 32 genes; details in Tables 3 and 4). The top five principal genetic diagnoses (definite [potential diagnoses]) provided by the panel were: ALGS/JAG1 + NOTCH2 (n = 62), BSEP deficiency/ABCB11 (n = 14 [n = 14]), A1ATD/SERPINA1 (n = 21 [n = 1]), MDR3 deficiency/ABCB4 (n = 4 [n = 9]), and POLG (n = 10 [n = 1]). Bile acid synthetic disorders with definite diagnoses were found in 12 subjects (potential diagnosis, n = 3) and included AKR1D1 (n = 3), CYP27A1 (n = 3 [potential diagnosis, n = 3]), and HSD3B7 (n = 6). Direct hyperbilirubinemia ± cholestasis from Dubin–Johnson syndrome (ABCC2) was present in 11 (potential diagnosis, n = 8). Interestingly, 10 definitive peroxisomal disorders were diagnosed—PEX1 (n = 3), PEX6 (n = 1), PEX26 (n = 5), and PEX12 (n = 1), and significant numbers of patients presented with Niemann-Pick C/NPC1 (n = 6 [n = 4]), TJP2 (n = 4 [n = 2]), and DCDC2 (n = 6 [n = 0]) deficiencies.

TABLE 4 - Combinations of pathogenic variants, likely pathogenic variants, and variants of unknown significance of interest
Gene Homozygous PATH/LP, n Two alleles PATH/LP, n One allele VOUS, 1 allele PATH/LP, n One allele PATH/LP only, n VOUS only, n Other reportable, n Reduced function, n
ABCB11 10 4 14 11 96 0 0
ABCB4 3 1 9 18 112 0 0
AKR1D1 3 0 0 1 15 0 0
AMACR 0 0 0 0 14 0 0
ATP8B1 3 1 2 4 65 0 0
BAAT 0 0 0 0 19 0 0
CFTR 2 2 29 82 184 0 145
CYP27A1 2 1 3 10 51 0 0
CYP7A1 0 0 0 0 25 0 0
CYP7B1 0 0 0 6 29 0 0
DCDC2 4 2 0 1 22 0 0
DGUOK 1 0 1 1 29 0 0
HSD17B4 0 0 0 0 27 0 0
HSD3B7 6 0 0 3 29 0 0
JAG1 0 0 1 52 56 0 0
MPV17 1 0 1 3 6 0 0
NOTCH2 0 0 0 9 100 0 0
NPC1 4 2 4 6 67 0 0
NR1H4 0 0 0 3 16 0 0
POLG 0 10 1 13 73 0 0
SERPINA1 21 0 1 78 35 65 0
SLC27A5 0 0 0 0 49 0 0
TJP2 3 1 2 1 82 0 0
VIPAS39 1 0 0 1 28 0 0
VPS33B 1 1 0 1 22 0 0
All Genes of Interest 65 25 68 304 1251 65 145
Autosomal dominant disorders.
Column for variants in the SERPINA1 gene such as the PI∗I allele.Genes of interest include: ABCB11, ABCB4, AKR1D1, AMACR, ATP8B1, BAAT, CFTR, CYP27A1, CYP7A1, CYP7B1, DCDC2, DGUOK, HSD17B4, HSD3B7, JAG1, MPV17, NOTCH2, NPC1, NR1H4, POLG, SERPINA1, SLC27A5, TJP2, VIPAS39, VPS33B.LP = likely pathogenic; PATH = pathogenic; VOUS = variant of unknown significance.

863 Subjects <3 Months Old

In the current cohort, 40% (863/2171) were <3 months old (including 257 ages <1 month). In this group, 59 (7%) had a definite diagnosis divided among 15 different genes; an additional 32 (4%) had a potential diagnosis, totaling 91 (11%) (Table 3, Supplemental Digital Content, https://links.lww.com/MPG/C247). The top five definite/potential diagnoses in this neonatal cohort, in order, were: ALGS (JAG1/NOTCH2), SERPINA1, CFTR, ABCC2, and ABCB11. There were 217 pathogenic and 27 LP variants in samples from these children. Thus, the overall diagnostic results in the youngest were not dissimilar from the entire age cohort (including >18 years of age, Table 4, Supplemental Digital Content, https://links.lww.com/MPG/C247), with a potential greater representation from CFTR and ABCC2 variants.

Monoallelic Pathogenic/Likely Pathogenic Variants

Single allele P/LP variants (excluding JAG1 and NOTCH2) were present in 394 subjects (18%) spread among 44 genes. The genes with the greatest number of monoallelic variants reported were CFTR (n = 82), SERPINA1 (n = 78), DHCR7 (n = 28), ABCC2 (n = 22), ABCB4 (n = 18), ABCG8 (n = 14), CC2D2A (n = 14), POLG (n = 13), ABCB11 (n = 11), and LIPA (n = 11). Moreover, 72 individuals had multiple monoallelic P/LP variants in two or three genes (including 31 ages <1 year [<3 months, n = 25; 3 –<12 months, n = 6]), with the largest number of individuals having SERPINA1 (n = 25) or CFTR (n = 23) variants in combination with other genes.

Deletion/Duplication Analysis

Of the 104 samples analyzed on the gene targeted aCGH, 103 had no reportable findings, whereas one sample had a deletion of the entire NPHP1 gene. This NPHP1 deletion is a relatively common deletion identified in the general population due to the gene being flanked by low copy repeats mediating recombination. There were no sequencing variants identified in the NPHP1 gene in the individual with the NPHP1 gene deletion.

DISCUSSION

This analysis of 2171 principally North American pediatric patients represents the largest collection of data from a large gene panel focused upon cholestasis. Overall, 8% of samples analyzed returned definitive diagnoses and another 4% potential diagnoses, providing an overall diagnostic yield of 12%. As a whole, the principal diagnoses were ALGS, A1ATD, and BSEP deficiency/ABCB11, MDR3 deficiency/ABCB4 and POLG. In those <3 months of age, CFTR and ABCC2 deficiencies were among the top five diagnoses. The majority of HCPs were gastroenterologists or hepatologists (63%) who utilized this panel primarily in patients <1 year old, of which nearly three-quarters were <3 months old, indicating a clinical need for utilization of genetic testing in the youngest cholestatic infants.

On average, patients had 1.4 genes with VOUS, and 30% of patients had a P/LP variant of one of the panel's genes. A total of 84 patients had one P/LP variant and one VOUS (autosomal recessive) in the same gene. Among the variants reported, 496 P/LP variants and 2702 VOUS had been previously reported; 166 novel P/LP variants and 415 novel VOUS were identified. These data underscore the evolving nature of genetic information available to clinicians and families and the challenges in interpreting large gene panel results.

Of interest is the presence of P/LP monoallelic variants in 394 subjects, of whom 72 had two to three genes with P/LP variants. Principal among these is the role for the A1ATD PI∗Z allele in SERPINA1, present in 78 (3.6%) of the cohort. In the US, this allele is present (heterozygous) in ∼1:48 individuals (1:35 Caucasians) (34), whereas in this cohort the frequency was ∼1:28. A single copy of the PI∗Z allele has been associated with risk for a variety of liver diseases, often as a modifier (35,36). Whether or not these monoallelic variants contributed to cholestasis, and if the cholestasis resolved is unknown, since follow-up clinical data are unavailable.

Also notable is that of the 66 genes in the cholestasis panel, 29 were thought to provide etiologies underlying definitive diagnoses in the 2171 subjects. This is intriguing, and indicates that testing a larger number of cholestatic patients may be required to determine which, if any, genes in this panel may ultimately be removed from consideration for contributing to cholestasis in this group. For example, among four nephronophthisis genes (NPHP1-4), only one potential diagnosis was found in this cohort, and thus, these genes can be considered very rare causes, or not a cause, of cholestasis.

It is intriguing that ABCC2 variants were found in a substantial number of participants—22 with definitive diagnoses and 11 monoallelic. Recent reports from several centers have identified significant ABCC2 variants in pediatric cholestatic patients, including neonatal cholestasis—one from France (37) and another from Taiwan (38). It is likely that functionally- relevant variants in ABCC2 may affect the capacity of the transporter for conjugated bilirubin in the neonate, as well as impacting roles for bile flow. This study indicates that ABCC2 variants may play roles in neonatal cholestasis in a North American cohort.

Using the strict definition of definite diagnosis, 8% of samples yielded results from this panel. This indicates that there are likely contributions from genes or intergenic regions yet to be identified, or may reflect the broad clinician-defined enrollment criteria. New cholestasis genes are constantly being identified, indicating the need to periodically update this panel with new information (eg, SLC51B and TMEM30A(39,40) added to this panel during the course of this program). A second intriguing consideration is contributions from monoallelic P/LP variants and VOUS alone or in specific multi-genic combinations; however, these two hypotheses cannot be determined without follow-up clinical information or pre-clinical validations of these variants (see below). A more complete assessment of the panel's clinical utility will require prospectively following patients to determine if some gene variants are related to a progressive rather than transient cholestasis, or if re-classification of pathogenicity is warranted (41).

The clinical utility of genetic testing relies on the professional interpretation and classification of the variants identified. According to ACMG recommendations (42), known pathogenic variants can be used in clinical decision making and in conjunction with other clinical information when possible. For example, certain ABCB11 variants may portend concern for the development of hepatocellular cancer (43). Variants classified as LP generally can be used in clinical decision making when used in combination with other evidence of the disease in question; however, all possible follow-up testing should be undertaken, as this may generate additional evidence that could permit the reclassification of the variant (42). Generally, VOUS should not be used in clinical decision making until further information about the variant becomes available and the classification is resolved (42). In some cases, a variant never before seen in the population or disease, or with limited reports in the literature, may warrant certain consideration of additional monitoring and/or testing (42).

Some limitations of this study should be noted. Of import, the broad enrollment criteria relied upon clinical labeling of patients as cholestatic, and it is not definitively known which subjects had subsequent laboratory or imaging criteria that lent further credence to an underlying genetic cause for cholestasis. Early use of this testing program may have led to a higher than expected utilization and enrollment of subjects whose cholestasis resolved. Without detailed follow-up data, it is not known if there is a subset of cholestatic patients who may have a higher yield with the use of this cholestasis gene panel, nor is it known if the combination of monoallelic pathogenic variants in more than one gene in select patients may truly be contributory to their cholestasis (eg, possible multigenic mechanism) (44). Further, degrees of phenotypic severity of cholestasis may be reflected by discoverable means (eg, gene variants) or non-discoverable means (eg, timed developmental expression and protein-protein interactions). Similarly, there is a spectrum with respect to genetic variants (from the complete absence of function on both alleles through a minor variation on a single allele). Finally, classification of variants has some inherent limitations, and the field is continuously evolving as new information about variants becomes apparent.

CONCLUSIONS

The development of new diagnostic technologies has led to a potential paradigm shift in evaluating cholestasis by incorporating genetic testing earlier in the diagnostic odyssey. Studies like these support engaging genotype before phenotype, especially because the phenotype of cholestasis in infants is generally non-unique (45). The use of multigene panels is not only beginning to help clinicians and families with timely definitive diagnoses (and rule-outs of potential diagnoses) but also is yielding newly identified genetic variants whose clinical significance awaits future analyses and longitudinal studies. Educating and training to assist HCPs’ understanding of genetic results and their clinical implications will assist modern subspecialists to embark upon the appropriate course of action for their patients. In sum, this study's findings support the utility of comprehensive and rapid multigene testing in diagnosing cholestasis. With continued implementation, the use of multigene panels has the potential to reshape the clinical algorithm of diagnosing cholestatic disorders, including the role and utility of liver biopsy.

Acknowledgments

This study was funded by Retrophin, Inc. Writing support and editorial assistance were provided by Sherri D. Jones, PharmD, of MedVal Scientific Communication Services, Inc, and was funded by Retrophin, Inc. This manuscript was prepared according to the International Society for Medical Publication Professionals’ “Good Publication Practice for Communicating Company-Sponsored Medical Research: GPP3.” Cristina da Silva, MS, of EGL Genetics, contributed to the collation of data.

REFERENCES

1. Fawaz R, Baumann U, Ekong U, et al. Guideline for the evaluation of cholestatic jaundice in infants: joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr 2017; 64:154–168.
2. McDaniell R, Warthen DM, Sanchez-Lara PA, et al. NOTCH2 mutations cause Alagille syndrome, a heterogeneous disorder of the notch signaling pathway. Am J Hum Genet 2006; 79:169–173.
3. Li L, Krantz ID, Deng Y, et al. Alagille syndrome is caused by mutations in human Jagged1, which encodes a ligand for Notch1. Nat Genet 1997; 16:243–251.
4. Lee WS, Sokol RJ. Mitochondrial hepatopathies: advances in genetics, therapeutic approaches, and outcomes. J Pediatr 2013; 163:942–948.
5. Abboud RT, Nelson TN, Jung B, et al. Alpha1-antitrypsin deficiency: a clinical-genetic overview. Appl Clin Genet 2011; 4:55–65.
6. Sundaram SS, Bove KE, Lovell MA, et al. Mechanisms of disease: inborn errors of bile acid synthesis. Nat Clin Pract Gastroenterol Hepatol 2008; 5:456–468.
7. Srivastava A. Progressive familial intrahepatic cholestasis. J Clin Exp Hepatol 2014; 4:25–36.
8. Masyuk T, Masyuk A, LaRusso N. Cholangiociliopathies: genetics, molecular mechanisms and potential therapies. Curr Opin Gastroenterol 2009; 25:265–271.
9. Lazaridis KN, LaRusso NF. The cholangiopathies. Mayo Clin Proc 2015; 90:791–800.
10. Lin WX, Zeng HS, Zhang ZH, et al. Molecular diagnosis of pediatric patients with citrin deficiency in China: SLC25A13 mutation spectrum and the geographic distribution. Sci Rep 2016; 6:29732 doi: 10.1038/srep29732.
11. Vanier MT, Gissen P, Bauer P, et al. Diagnostic tests for Niemann-Pick disease type C (NP-C): a critical review. Mol Genet Metab 2016; 118:244–254.
12. Tai Y, Xie Y, Tang CW. Compound heterozygous mutations of ABCB11 responsible for benign recurrent intrahepatic cholestasis. J Dig Dis 2015; 16:299–302.
13. van Mil SW, van der Woerd WL, van der Brugge G, et al. Benign recurrent intrahepatic cholestasis type 2 is caused by mutations in ABCB11. Gastroenterology 2004; 127:379–384.
14. Degiorgio D, Crosignani A, Colombo C, et al. ABCB4 mutations in adult patients with cholestatic liver disease: impact and phenotypic expression. J Gastroenterol 2016; 51:271–280.
15. Park HJ, Kim TH, Kim SW, et al. Functional characterization of ABCB4 mutations found in progressive familial intrahepatic cholestasis type 3. Sci Rep 2016; 6:26872 doi: 10.1038/srep26872.
16. Pasmant E, Goussard P, Baranes L, et al. First description of ABCB4 gene deletions in familial low phospholipid-associated cholelithiasis and oral contraceptives-induced cholestasis. Eur J Hum Genet 2012; 20:277–282.
17. Saleh M, Kamath BM. Chitayat D Alagille syndrome: clinical perspectives. Appl Clin Genet 2016; 9:75–82.
18. Molho-Pessach V, Rios JJ, Xing C, et al. Homozygosity mapping identifies a bile acid biosynthetic defect in an adult with cirrhosis of unknown etiology. Hepatology 2012; 55:1139–1145.
19. Fischler B, Bodin K, Stjernman H, et al. Cholestatic liver disease in adults may be due to an inherited defect in bile acid biosynthesis. J Intern Med 2007; 262:254–262.
20. Droge C, Haussinger D, Keitel V. Genetic variants in adult liver diseases. Z Gastroenterol 2015; 53:1436–1446.
21. Hirschfield GM, Chapman RW, Karlsen TH, et al. The genetics of complex cholestatic disorders. Gastroenterology 2013; 144:1357–1374.
22. Feldman AG, Sokol RJ. Neonatal cholestasis: emerging molecular diagnostics and potential novel therapeutics. Nat Rev Gastroenterol Hepatol 2019; 16:346–360.
23. Herbst SM, Schirmer S, Posovszky C, et al. Taking the next step forward—diagnosing inherited infantile cholestatic disorders with next generation sequencing. Mol Cell Probes 2015; 29:291–298.
24. Shagrani M, Burkholder J, Broering D, et al. Genetic profiling of children with advanced cholestatic liver disease. Clin Genet 2017; 92:52–61.
25. Togawa T, Sugiura T, Ito K, et al. Molecular genetic dissection and neonatal/infantile intrahepatic cholestasis using targeted next-generation sequencing. J Pediatr 2016; 171:171–177. e1–4.
26. Wang NL, Lu YL, Zhang P, et al. A specially designed multi-gene panel facilitates genetic diagnosis in children with intrahepatic cholestasis: simultaneous test of known large insertions/deletions. PLoS One 2016; 11:e0164058 10.1371/journal.pone.0164058.
27. Hakim A, Zhang X, DeLisle A, et al. Clinical utility of genomic analysis in adults with idiopathic liver disease. J Hepatol 2019; 70:1214–1221.
28. Klein CJ, Foroud TM. Neurology individualized medicine: when to use next-generation sequencing panels. Mayo Clin Proc 2017; 92:292–305.
29. Polla DL, Cardoso MT, Silva MC, et al. Use of targeted exome sequencing for molecular diagnosis of skeletal disorders. PLoS One 2015; 10:e0138314 doi: 10.1371/journal.pone.0138314.
30. Vona B, Muller T, Nanda I, et al. Targeted next-generation sequencing of deafness genes in hearing-impaired individuals uncovers informative mutations. Genet Med 2014; 16:945–953.
31. Bastida JM, Del Rey M, Lozano ML, et al. Design and application of a 23-gene panel by next-generation sequencing for inherited coagulation bleeding disorders. Haemophilia 2016; 22:590–597.
32. Yubero D, Brandi N, Ormazabal A, et al. Targeted next generation sequencing in patients with inborn errors of metabolism. PLoS One 2016; 11:e0156359 doi: 10.1371/journal.pone.0156359.
33. Grozeva D, Carss K, Spasic-Boskovic O, et al. Targeted next-generation sequencing analysis of 1000 individuals with intellectual disability. Hum Mutat 2015; 36:1197–1204.
34. de Serres FJ, Blanco I, Fernandez-Bustillo E. Ethnic differences in alpha-1 antitrypsin deficiency in the United States of America. Ther Adv Respir Dis 2010; 4:63–70.
35. Schaefer B, Mandorfer M, Viveiros A, et al. Heterozygosity for the alpha-1-antitrypsin Z allele in cirrhosis is associated with more advanced disease. Liver Transpl 2018; 24:744–751.
36. Kok KF, Wahab PJ, Houwen RH, et al. Heterozygous alpha-I antitrypsin deficiency as a co-factor in the development of chronic liver disease: a review. Neth J Med 2007; 65:160–166.
37. Corpechot C, Barbu V, Chazouillères O, et al. Genetic contribution of ABCC2 to Dubin-Johnson syndrome and inherited cholestatic disorders. Liver Int 2020; 40:163–174.
38. Chen HL, Li HY, Wu JF, et al. Panel-based next-generation sequencing for the diagnosis of cholestatic genetic liver diseases: clinical utility and challenges. J Pediatr 2019; 205:153.e6–159.e6.
39. Sultan M, Rao A, Elpeleg O, et al. Organic solute transporter-β (SLC51B) deficiency in two brothers with congenital diarrhea and features of cholestasis. Hepatology 2018; 68:590–598.
40. Liu L, Zhang L, Zhang L, et al. Hepatic tmem30a deficiency causes intrahepatic cholestasis by impairing expression and localization of bile salt transporters. Am J Pathol 2017; 187:2775–2787.
41. Droge C, Bonus M, Baumann U, et al. Sequencing of FIC1, BSEP and MDR3 in a large cohort of patients with cholestasis revealed a high number of different genetic variants. J Hepatol 2017; 67:1253–1264.
42. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med 2015; 17:405–424.
43. Knisely AS, Strautnieks SS, Meier Y, et al. Hepatocellular carcinoma in ten children under five years of age with bile salt export pump deficiency. Hepatology 2006; 44:478–486.
44. Posey JE, Harel T, Liu P, et al. Resolution of disease phenotypes resulting from multilocus genomic variation. N Engl J Med 2017; 376:21–31.
45. Lu JT, Campeau PM, Lee BH. Genotype–phenotype correlation—promiscuity in the era of next-generation sequencing. N Engl J Med 2014; 371:593–596.
Keywords:

cholestatic disorders; diagnostic yield; gene panel; hepatic disease; next-generation sequencing

Supplemental Digital Content

Copyright © 2021 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition