The population frequency of predicted pathogenic variants in the genes associated with Autosomal Dominant Polycystic Liver Disease (ADPLD) and kidney cysts

This study estimates that predicted pathogenic variants in genes associated with Autosomal Dominant Polycystic Liver Disease (ADPLD) occur in approximately 1 in 95 to 1 in 151 individuals within the general population, with higher frequencies observed in admixed American, Finnish, and African/African American groups compared to Europeans, despite the condition's variable penetrance and expressivity.

The Gist

The Big Picture: Finding the "Glitch" in the Liver Blueprint

Imagine your body is a massive, bustling city. In this city, there are specific construction crews responsible for building and maintaining the "pipes" (cysts) in your liver and kidneys. Usually, these crews work perfectly. But sometimes, there is a typo in the city's blueprint (your DNA). This typo causes the crews to build too many pipes, leading to Polycystic Liver Disease (PLD).

For a long time, doctors thought this "typo" was very rare—like finding a specific typo in a library of a million books. This study asked a simple question: "How common is this typo really in the general population?"

To find out, the researchers didn't look at sick patients in a hospital. Instead, they looked at a giant digital library of DNA from over 800,000 healthy people (called gnomAD). They were essentially scanning the "citizens" of this digital city to see how many of them carried the blueprint error, even if they didn't know it yet.

The Six Suspects (The Genes)

The researchers focused on six specific genes (the "construction manuals") known to cause this liver disease: GANAB, ALG8, LRP5, PRKCSH, SEC61B, and SEC63.

Think of these genes as six different foremen on a construction site. If any one of them has a bad instruction, the pipes start growing out of control.

• The "Heavy Hitters": Some foremen (like PRKCSH and SEC63) usually cause big, obvious problems.
• The "Quiet Ones": Others (like LRP5 and ALG8) usually cause milder issues, but the study found they are actually the most common culprits in the general population.

The Detective Work: Two Different Flashlights

The researchers used two different "flashlights" to scan the DNA library, because looking for a typo is tricky.

1. Flashlight A (The "Computer Guess" Strategy):
   Since they couldn't ask the people in the database if they had liver cysts (the data is anonymous), they used a computer program to predict which typos were dangerous. They looked for:

   • Missing pages: Big chunks of the manual deleted (Loss of Function).
   • Nonsense words: Sentences that stop abruptly (Null variants).
   • Typos in key words: Misspelled words that change the meaning of a sentence (Missense variants), but only if the word is very rare and looks like it breaks the sentence.
   • Result: This method found that 1 in 95 people carries a dangerous typo. That's like finding a glitch in the blueprint for every single bus stop in a medium-sized city!

2. Flashlight B (The "Expert Opinion" Strategy):
   This time, they only looked at typos that had already been flagged by medical experts in a database called ClinVar. This is like only counting the typos that a professional editor has already circled in red.

   • Result: This found 1 in 151 people. This number is lower because experts are stricter; they only flag the typos they are 100% sure about.

The Surprising Discoveries

1. The "Quiet" Foremen are Everywhere
The study found that the genes LRP5 and ALG8 were the most common sources of these typos. This is surprising because, in the past, doctors mostly saw these genes in people with severe disease. It turns out these genes are actually very common in the general population, but they often cause such mild symptoms that people never know they have them.

2. Geography Matters
Just like some languages have more common spelling mistakes than others, some populations have more of these typos. The study found that people of Finnish, Admixed American, and African/African-American ancestry had a higher frequency of these typos compared to people of European ancestry.

• Analogy: Imagine a specific typo in a recipe. In one town (Finland), almost every baker might have that specific typo in their recipe book, while in another town (Europe), it's much rarer.

3. The "Silent Carriers"
Here is the most important twist: Having the typo doesn't mean you will get sick.
The study estimates that 1 in 95 people has the "bad blueprint," but not everyone with that blueprint develops liver cysts.

• Analogy: Think of the typo as a loaded gun. Most people who carry the gun never pull the trigger. Some people have "safety locks" (other genes or environmental factors) that prevent the cysts from forming. This is called incomplete penetrance. So, while the risk is common, the disease might be less common.

Why Does This Matter?

1. Raising the Alarm
Doctors now know this condition is much more common than they thought. If a patient comes in with kidney cysts, the doctor should also check their liver, and vice versa. It's like realizing that a leak in the kitchen sink might actually be a sign of a problem in the basement plumbing.

2. Better Testing
Because these typos are so common, genetic testing panels (the tests doctors order to find the cause of kidney disease) need to include these liver genes. If a doctor only looks for kidney genes, they might miss the liver connection.

3. Future Treatments
If we know that 1 in 95 people carries these typos, pharmaceutical companies might be more interested in developing drugs to stop liver cysts. It's a bigger market than they thought!

The Bottom Line

This study is like a massive census of the human "instruction manual." It revealed that the "typos" causing Polycystic Liver Disease are far more common than we ever imagined (affecting roughly 1% of the population).

However, just because you have the typo doesn't mean your liver will explode with cysts. Many people carry these genetic "glitches" and live completely healthy lives, perhaps with just a few harmless bumps on their liver that they never notice. The goal now is to help doctors spot these connections early so they can monitor patients better and hopefully develop treatments for those who do get sick.

Technical Summary

1. Problem Statement

Autosomal Dominant Polycystic Liver Disease (ADPLD) is a genetic condition characterized by multiple liver cysts, often overlapping clinically and genetically with Autosomal Dominant Polycystic Kidney Disease (ADPKD). ADPLD is caused by pathogenic variants in six specific genes: GANAB, ALG8, LRP5, PRKCSH, SEC61B, and SEC63.

Despite the clinical relevance of these genes (which are included in cystic kidney panels), the true population frequency of ADPLD-causing variants remains uncertain. Previous estimates based on clinical diagnoses or small cohorts often underestimated the disease burden because:

• Many cases are asymptomatic.
• Formal diagnosis requires specific imaging criteria (e.g., $\ge$4 cysts) that may not be met in early stages.
• There is a lack of large-scale population data distinguishing between "predicted pathogenic" variants and confirmed clinical cases.

The study aims to determine the precise population frequency of predicted pathogenic variants in these six genes using large-scale genomic databases to better understand the genetic landscape of ADPLD.

2. Methodology

The researchers utilized two minimally overlapping datasets from the gnomAD (Genome Aggregation Database) consortium: v.2.1.1 (WES/WGS, $n \approx 141,456$) and v.4.1 (WES/WGS, $n \approx 807,162$).

Variant Filtering Strategy:
Since gnomAD lacks clinical phenotypic data, the authors could not apply standard ACMG/AMP criteria directly. Instead, they developed a computational strategy to identify "predicted pathogenic" variants:

1. Exclusion: Synonymous, intronic, and UTR variants were excluded.
2. Structural/Copy Number Variants (SVs/CNVs): Variants classified as "Loss of Function" (LoF) by gnomAD affecting exons were included.
3. Null Variants: Nonsense, frameshift, and splice-site variants (excluding those in the last exon/50bp of the penultimate exon to avoid NMD escape) were included.
4. Missense Variants: To be included, variants had to meet all of the following criteria:
   • Rare (allele count $\le$ 5).
   • Predicted damaging by three tools: PolyPhen-2 (score $\ge$ 0.95), SIFT4G (score $\le$ 0.05), and MutationTaster (disease-causing/probably deleterious).
   • Located on a conserved amino acid residue across vertebrates (chicken, mouse, human).
5. Validation: The missense filtering strategy was validated against ClinVar, showing 75% sensitivity and 90% specificity.

Comparative Approaches:

• Strategy A (Primary): Applied the computational filter above to gnomAD v.2.1.1.
• Strategy B (ClinVar-based): Used variants in gnomAD v.2.1.1 and v.4.1 that were already assessed as "Pathogenic," "Likely Pathogenic," or "Conflicting" (VUS + P/LP) in ClinVar, LOVD, or HGMD. This served as a replication dataset, acknowledging that ClinVar assessments rely on clinical data and ACMG criteria.

Statistical Analysis:
Frequencies were calculated by dividing the number of individuals with at least one predicted pathogenic variant by the total number of individuals sequenced for that gene. Chi-squared tests were used to compare frequencies across different ancestral groups.

3. Key Contributions

• First Large-Scale Population Estimate: Provides the most comprehensive estimate of ADPLD variant frequency to date, moving beyond clinical cohorts to the general population.
• Dual-Strategy Validation: Cross-referenced a rigorous computational prediction model (v.2.1.1) with ClinVar-annotated variants (v.4.1) to ensure robustness.
• Ancestral Disparities: Identified significant differences in variant frequencies across diverse ancestral populations, highlighting the need for diverse genomic databases.
• Gene-Specific Insights: Challenged previous literature suggesting PRKCSH and SEC63 are the most common causes, revealing LRP5 and ALG8 as the most frequent in the general population.

4. Key Results

Population Frequency Estimates:

• Strategy A (Computational Filter, gnomAD v.2.1.1): Found a frequency of 1 in 95 people (1,217 individuals out of 115,143).
• Strategy B (ClinVar-based, gnomAD v.4.1): Found a frequency of 1 in 151 people (5,174 individuals out of 781,274).
  • Note: The lower frequency in Strategy B is attributed to the fact that ClinVar has not assessed all variants in the larger v.4.1 dataset, leading to an underestimation.

Gene-Specific Findings:

• Most Common Genes: LRP5 (1 in 274 in v.2.1.1; 1 in 324 in v.4.1) and ALG8 (1 in 277 in v.2.1.1; 1 in 614 in v.4.1) were the most frequently affected genes.
• Least Common: SEC61B was the rarest (1 in 12,146 in v.2.1.1).
• Variant Types: Null variants were generally more common than missense variants in the ClinVar dataset, though missense variants were abundant in the computational filter.

Ancestral Differences (gnomAD v.4.1):
Significant variations were observed compared to the European reference population (1 in 197):

• Admixed American: 1 in 100 ($p < 0.0001$)
• Finnish: 1 in 107 ($p < 0.0001$) – driven largely by ALG8 variants.
• African/African American: 1 in 130 ($p < 0.0001$)
• Amish: 1 in 14 (driven by structural variants, though sample size was small).

Clinical Implications:

• The study suggests that the total burden of liver cysts (including those from ADPKD genes PKD1/PKD2 and ADPLD genes) affects at least 1 in 80 people.
• Not all carriers develop cysts due to incomplete penetrance and variable expressivity.
• LRP5 and ALG8 variants are associated with milder phenotypes and are more common in the general population than previously thought.

5. Significance

• Clinical Awareness: The high frequency (1 in 95–151) suggests ADPLD is far more common than clinical records indicate. This implies many individuals with liver cysts may be misdiagnosed or undiagnosed.
• Genetic Counseling: The finding of incomplete penetrance is crucial for counseling; carrying a pathogenic variant does not guarantee symptomatic disease.
• Therapeutic Development: Accurate prevalence data is essential for pharmaceutical companies to justify developing targeted treatments for liver cysts.
• Renal Practice: Since these genes are often tested in kidney cyst panels, nephrologists must be aware that finding these variants may indicate a risk for liver cysts, even if kidney failure is rare in ADPLD.
• Limitations & Future Directions: The study acknowledges that computational predictions may miss founder variants or have false positives/negatives. As tools improve and more genes are identified, these frequencies will likely increase further.

In conclusion, this study fundamentally shifts the understanding of ADPLD from a rare clinical entity to a relatively common genetic condition in the general population, with significant implications for screening, diagnosis, and future therapeutic strategies.