Genetics for dermatologists. Part 2: Clinical evaluation, sequencing technologies and interpretation

History and examination

Dermatologists must be aware of red flag signs, such as erythroderma, collodion baby, blisters on sites of friction (at birth), therapy-resistant eczema with infections and growth failure, palmoplantar keratoderma, photosensitivity, poikiloderma (in early childhood), multiple benign or malignant tumours, cutaneous lesions in a linear/ segmental distribution, multiple hypo- or hyper-melanotic macules, loose hanging skin, positive family history and presence of other extra-cutaneous manifestations involving the central nervous system, skeletal system, teeth, eyes and ears (at any age).¹ Salient points in history and examination of genodermatoses are shown in Table 1 and Figures 1a, b.⁵

Close

Figure 1a:

Work-flow of a suspected case of genodermatosis. (NGS: Next generation sequencing, CES: Clinical exome sequencing, WES: Whole exome sequencing, WGS: Whole genome sequencing, PIGD: Pre-implantation genetic diagnosis)

Export to PPT

Close

Figure 1b:

Pictorial representation of the work-flow of a suspected case of genodermatosis. (NGS: Next generation sequencing, HGMD: Human Gene Mutation Database, ACMG: American College of Medical Genetics and Genomics, HGNC: HUGO Gene Nomenclature Committee, SIFT: Sorting Intolerant from Tolerant).

Export to PPT

Pedigree charting is crucial in the history and examination of genodermatoses because it helps identify patterns of inheritance, such as autosomal dominant, autosomal recessive, or X-linked traits. By documenting the family history of affected individuals, clinicians can pinpoint potential genetic links, assess the risk of recurrence in future generations, and guide genetic testing and counselling.

The approach to genodermatoses may be ‘phenotype first’ or ‘genotype first’. A better approach is ‘phenotype first’. Here, the clinician considers medical and family history to understand patterns of inheritance. The clinician also performs examination, and does relevant screening investigations - all of which help to arrive at phenotype. This gives us possible clinical differential diagnoses, following which one can proceed for genotyping. It is better to analyse the clinical phenotype before testing for genotypes.

However, with increasing ease of access to sequencing technologies, it is often tempting to order a genetic test when one suspects a genetic problem without completely phenotyping the patient. We also see patients in our OPD directly walking in with a genetic report. This approach of doing genetic testing before phenotyping is called ‘genotype first’ and can be disadvantageous on many accounts – 1) genetic test may come negative (the yield of whole exome sequencing, the most commonly ordered test for genodermatoses, based on a genotype first approach, is only 50%)⁵ 2) it may throw up a variant of uncertain significance (VOUS) and good phenotyping is essential to narrow down the variants 3) phenotype is important for initiating therapy, which cannot be decided by the genotype alone 4) Phenotype first approach also helps us to select the right genetic test.

Hence, good, clinical, and detailed phenotyping and screening investigations are important. However, we often need both for the final diagnosis since it helps in genetic counselling and offering precision therapy.⁵

Most genodermatoses run in families, and the absence of a family history does not exclude a genetic skin disorder.

Factors influencing diagnosis may include asymptomatic female carriers of X-linked disorders, early death, incomplete family history, and false paternity.¹ Non-mendelian inheritance patterns that do not follow the classic Mendelian rules also add complexity to genetic inheritance. Indications for genetic testing are enumerated in Table 2.^1,2

Pre-test counselling

A genetic test is extensive and complicated, and requires families to undergo a prior consenting session. This crucial step educates the parents and patients about sample type (usually blood), the selection of family members for sample collection (proband vs. trio vs. other), turnaround time, test cost, possible results (including the possibility of finding variants of uncertain significance (VOUS)) and incidental findings unrelated to the suspected disease, and associated limitations and risks ( for e.g., privacy concerns and how the results could affect other family members). Many families believe that a genetic test will lead the way to a cure or treatment for their child. It must be emphasised to them that this isn’t guaranteed.² At present, genetic tests continue to be expensive and out of reach for most patients. They are often not covered under insurance. This may change in the future as the identification of patients with genodermatoses increases and specific treatments are developed.^4,6

Post-test counselling

This is performed after the clinician receives the results of genetic testing. These are discussed with the patient or family, including any identified variants and their clinical significance, if known. This session is crucial to help them understand the results, whether they are diagnostic, inconclusive, or require further testing. Patients are guided on the next steps for management or follow-up, and if no clear diagnosis is found, the possibility of future re-evaluation as more information becomes available is also discussed.

Types of next-generation sequencing

Clinical applications of NGS include targeted panel sequencing, clinical exome sequencing (CES), whole-exome sequencing (WES), whole-genome sequencing (WGS), and RNA-sequencing. These are compared in Table 4.^2,3,10–12

Important points to remember:

• Genetic testing should be done at certified labs with clear protocols for lab audits. For commercial labs doing NGS, it is good to have CAP (College of American Pathologists) and NABL (National Accreditation Board for Testing and Calibration Laboratories) accreditation, too, as it ensures internationally recognised standards of quality. Results must be interpreted by a qualified clinical molecular geneticist or molecular genetic pathologist. If the lab is doing prenatal genetic tests, then it must possess a PCPNDT (Pre-Conception Pre-Natal Diagnostic Technique) certificate. Sometimes, the decision of research labs actively involved in NGS, to not pursue accreditation stems from their unique operational needs.

• The physician who orders the genetic test must have sufficient knowledge of genetic mechanisms to order and correctly interpret the results.

• Turnaround time varies between 4-12 weeks, depending on the type of test ordered. The clinician needs to take this into account while ordering prenatal tests, as the window period for intervention, in case the foetus is inviable, is only up to 20 weeks.

• Clinical exome sequencing (CES) or Whole exome sequencing (WES) is preferably carried out in trio testing (mother, father, affected child), as it enhances the diagnostic yield compared with singleton analyses. This approach helps identify and interpret de-novo and compound heterozygous variations more easily. Trio testing is easier to perform in paediatric patients as compared to affected adults. In the absence of parents’ sample, the inclusion of siblings can increase the ability to diagnose recessive disorders.² Another cost-effective approach is to analyse the affected child sample by NGS followed by carrier screening by the Sanger method in parents or siblings.

• Depth and Coverage - In exome sequencing, “depth” and “coverage” refer to two important metrics that assess the quality and completeness of the sequencing data:

• Depth (also known as sequencing depth or read depth)- refers to the average number of times each base in the target regions of the exome are sequenced. It indicates how many times a specific nucleotide has been read during the sequencing process. A higher depth means more confidence in the accuracy of the sequencing results for that base. For example, if a target region has a depth of 20x, it means that, on average, each base in that region has been sequenced 20 times.

• Coverage- refers to the proportion of target regions in the exome that has been sequenced at a certain depth. It represents the completeness of the sequencing data across the regions of interest. For example, if a target region has 1000 base pairs, and the sequencing has covered only 900 base pairs, it means that the coverage is 90%.

• Both depth and coverage are essential for ensuring the accuracy and reliability of exome sequencing results. Higher depth and coverage generally lead to better detection of genetic variants. Insufficient depth or coverage can result in missed variants or inaccuracies in variant calling. For germline variants, a depth of 80-100x is desirable, whereas for somatic variants, with low variant frequency, deep sequencing (upto 250x) is desirable.

• Targeted approaches may require higher depth and coverage in specific regions of interest, whereas broader sequencing approaches like WES and Whole genome sequencing (WGS) aim for comprehensive coverage of exonic or genomic regions with possibly a lower but optimal depth to achieve reliable variant detection.

• Standard NGS has challenges, with some regions of the genome being poorly covered. Large genes, pseudogenes, specific regions with high GC (guanine-cytosine) content, small indels (small insertions and/ or deletions of nucleotides), and trinucleotide repeats are challenging to sequence and interpret, leading to a coverage of <100%. High GC content can make DNA regions difficult to sequence because the strong bonds between G and C bases are harder for sequencing machines to read accurately. These regions may have gaps or lower-quality data compared to areas with more balanced base content.

• There may be differences in coverage between different sequencing platforms (for e.g., Illumina vs Ion Torrent), leading to differences in results between labs. Illumina is highly accurate, particularly for detecting SNVs, whereas Ion Torrent has a rapid turnaround time but higher error rates.

• Sequencing errors are also common, therefore, validation of NGS data using Sanger sequencing is desirable.¹³

• We still do not understand large parts of the coding region (exome), hence it is difficult to interpret the results of sequencing. Pathogenic variants may be missed during variant filtration and variant calling. Moreover, predicting the pathogenic effect of novel variants/ variants of uncertain significance (VOUS/ VUS) is difficult despite prediction from in-silico tools.

• The overall approximate diagnostic rate for genodermatoses is around 50%. The diagnostic rates are higher and approach 95-100% for disorders that are phenotypically well characterised and nearly exclusively monogenic, such as epidermolysis bullosa. In such cases, small, targeted panels can be used as first-line diagnostic test.⁵

• In patients suspected for genetic disorders where DNA sequencing does not reveal any significant mutations, RNA-sequencing (RNA-seq) is emerging as a complementary method. Unlike DNA-based approaches, RNA-seq directly evaluates RNA sequences and can identify aberrant transcripts. For e.g., in a patient with splice site mutations, clinicians can identify abnormal splicing events, such as exon skipping or intron retention which are not obvious in DNA sequencing.

• Storage of vast amounts of data generated from NGS studies poses a significant challenge. Issues like hardware limitations, patient confidentiality, and data accessibility remain pressing and unsolved.^14,15

Reasons for non-contributory WES/ No variants found in the report:

• Cannot detect non-coding variants as the targeted regions include all protein-coding exons and approximately ±20 base pairs from the exon-intron boundary.⁹

• Can miss structural variants – insertions, deletions, duplications, inversions, translocations, copy number variants, owing to the alignment process.

• Variants may be interpreted differently among clinical laboratories, leading to disagreements in around 10% of cases.³

• Inadequate clinical data provided by the clinician leading to errors in variant calling and filtration.

In inconclusive cases one must preserve the genetic material along with a detailed description of clinical features and images. Many times, the responsible gene may be found in future analyses. Regular follow-up visits will help unearth new clinical features with time, eventually leading to a diagnosis.¹

Re-analysis of exome data, providing complete and correct phenotype to enable variant filtering at the time of doing bioinformatics, and consultation between the clinician and the lab will increase the diagnostic yield.

References

The references, if any, that contributed to the ACMG classification are listed at the end of the report.

Nomenclature

The Human Genome Variation Society (HGVS) maintains standard gene variant nomenclature available, at (http://www.hgvs.org/mutnomen). It is recommended as the primary guideline for variant nomenclature, with laboratories specifying the version used in their test methods. Tools (https://mutalyzer.nl) can be used to generate accurate HGVS nomenclature for variant descriptions. E.g., according to HGVS, nonsense variants are reported as “*” or “Ter”, or even “X”.¹⁶

Terminology and ACMG variant classification

Frequently used terms like “mutation” and “polymorphism” often lead to confusion due to incorrect assumptions of pathogenic and benign effects, respectively. Hence, it is recommended to replace both terms with “variant/variation”. Since the clinical significance of a variant can span a spectrum from pathogenic to benign, the ACMG guidelines prescribe standardised terms like ‘pathogenic,’ ‘likely pathogenic,’ ‘uncertain significance,’ ‘likely benign,’ and ‘benign’ for further classification. Broad criteria on which this classification is based are depicted in Table 5.

Variant of uncertain significance (VOUS/VUS)

In many cases, a variant’s pathogenic or benign status isn’t convincingly established, leading to its classification as a VOUS or VUS. A VOUS is uncertain; it might cause a disorder or be entirely benign, but insufficient data hinders definite determination. Functional validation in cell culture or animal models, beyond current clinical evaluation, is needed for VOUS. While every exome harbours dozens of VOUSs, only those linked to the patient’s phenotype are reported. It must be emphasised that it shouldn’t guide clinical decisions, and efforts should focus on determining the variant as “pathogenic” or “benign.” Over time, increased genetic data sharing, as well as collaboration between clinical and research labs will reclassify more VOUSs. Laboratories must suggest periodic inquiries by healthcare providers to stay updated on any changes in knowledge regarding VOUSs.

When a clinician encounters a genetic report with a VOUS, the following approach is recommended:

• Review clinical context: Correlate the patient’s clinical presentation with the variant in question and check the family history and inheritance pattern of the disorder.

• Evaluate parental testing: If an asymptomatic parent carries the same VOUS, it might suggest that the variant is less likely to be pathogenic, especially for autosomal dominant conditions. Other family members can be tested to see if the variant co-segregates with the disease phenotype.

• Population frequency: If the VOUS has a high frequency in the general population (>5%), it is less likely to be pathogenic, particularly for severe or rare conditions. Use databases such as gnomAD to check variant frequency.

• Functional studies: Look for existing in vitro or in vivo functional studies on the variant. If none exist, consider collaborating with research institutions to investigate its impact on protein function.

• In silico predictions: Utilise bioinformatics tools that predict the potential impact of the variant on the protein structure and function, though these should be interpreted with caution.

• Literature review: Search for published reports that describe the variant and its association with disease.

• Consultation with genetics professionals: Engage with genetic counsellors or clinical geneticists to interpret VOUS and determine next steps for patient management.

• Reclassification over time: Periodically re-evaluate the VOUS as new data becomes available. Genetic databases are continuously updated, which might lead to the reclassification of the variant from uncertain significance to being either benign or pathogenic in the future.

• Patient communication: Clearly communicate the uncertainty of the VOUS to the patient and discuss the potential for future reclassification.

• Management plan: Establish a plan that may include regular monitoring and follow-up, especially if the patient is at risk for developing symptoms.

Secondary/incidental findings

Secondary or incidental findings are medically actionable genetic variants discovered during genetic testing that are unrelated to the initial purpose of testing. The ACMG has identified 73 genes with pathogenic variants associated with an increased risk of cancer or sudden death, offering medical interventions like prophylactic mastectomy or pacemaker implantation.³ However, these interventions carry medical and psychological risks. Counselling before testing involves informing individuals about the chance of secondary findings and the option to opt out of receiving such results in the consent form.

In-silico software tools

Various in-silico tools or software algorithms, available publicly and commercially, assist in interpreting the effect of sequence variants at nucleotide, protein and amino acid levels. Combining predictions from different tools enhances interpretation robustness. In-silico tools, such as PolyPhen2, SIFT, and MutationTaster, are commonly used in clinical labs for missense variant interpretations. A list of variant prediction in-silico tools can be found in Table 6.

Databases

Newer discovered variants get continuously added to the databases [Table 7] and serve as useful sources of information for clinicians as well as geneticists. Databases can be of 2 types: Population database and Disease database.

• i.

  Population databases: provide variant frequencies in large populations. These include pathogenic variants as well as healthy individuals. However, they lack comprehensive information on variant functional effects or associated phenotypes.

• ii.

  Disease databases: focus on variants in patients with diseases, along with the assessment of variant pathogenicity.

Pharmacogenomics

Some labs offer additional information about variants in genes involved with drug metabolism that affect drug efficacy and confer increased risk for adverse events.

Common complex disorders

Variations in common, complex disease genes are sometimes identified while sequencing Mendelian genes. These are reported as “risk alleles” or under “other reportable” categories in the diagnostic report, instead of using the terms “pathogenic”, and “likely pathogenic”.

Genes of uncertain significance (GUS)

This is when the lab finds variation in a gene that has never been associated with any patient phenotype or if it’s linked to a different phenotype. Variants found in a GUS are reported as ‘Variants in a gene of uncertain significance’. These variants, if reported, should always be classified under “Uncertain Significance”. Further evidence is needed to confirm the gene’s association with a disease before any variant in it is considered pathogenic for that disease.