Pathways and documents

Rare Disease and Genetics

When common explanations have run out

Rare Disease and Genetics is for the case that has outrun the common explanations, or the genetic report that needs reading. Before reaching for anything rare, it asks whether a familiar disease in unusual form, a medicine, or a combination of ordinary conditions accounts for the findings. It then builds a precise phenotype and family picture, ranks candidate diagnoses by mechanism with treatable conditions first, and proposes testing in tiers, from targeted biochemistry to a panel, exome or genome.

Who it is for

  • Physicians facing a multisystem picture that crosses specialties and has not responded to standard treatment
  • Cardiologists, neurologists, nephrologists and other specialists who receive a gene panel or exome result with a variant to interpret
  • Residents and registrars preparing a referral to clinical genetics or a specialist centre

What it does

  • Common causes ruled out first

    Tests the ordinary explanations first, such as an unusual presentation of a common illness, an adverse drug effect or two everyday conditions at once, because labelling someone with a rare disease on weak grounds is a common and costly mistake.

  • A precise phenotype

    Restates each finding in standard phenotype vocabulary, in the manner of the Human Phenotype Ontology, notes when it began and how it has behaved, and marks which findings are distinctive and which are frequent in healthy people.

  • The family picture

    Draws out a pedigree over three generations, including parental relatedness, ancestry, early deaths and relatives with similar problems, and infers a likely mode of inheritance; an unremarkable family history is not taken to rule out a genetic condition.

  • Candidate diagnoses, treatable first

    Sets out a table of candidate conditions with gene and inheritance, the features that fit and those that do not, the confirmatory test and whether treatment exists, with treatable diagnoses at the top.

  • Testing in tiers

    Starts with specific biochemical or first-line tests and escalates to a panel, exome or genome as the differential widens, naming the blind spots of each, for instance repeat expansions, large structural changes and mitochondrial DNA.

  • Genetic results in ACMG/AMP terms

    Reads a reported variant against the five ACMG/AMP classes, checks zygosity against the gene's inheritance, and explains what could resolve a variant of uncertain significance: segregation, functional evidence, phenotype match or later reanalysis.

How it works

  1. List every abnormal finding, including those that seemed unrelated, with the age at which each began and how it has changed.
  2. Add the family history across three generations, consanguinity and ancestry, and any tests already done, including the exact variant wording from a genetic report.
  3. The engine defines the phenotype, looks for patterns that point to a group of disorders, such as attacks brought on by fasting or intercurrent illness, and ranks candidates by mechanism.
  4. It proposes a tiered testing plan and, where a genetic result is supplied, classifies it and explains what it does and does not establish.
  5. The report covers counselling, implications for relatives and referral, says when a statement rests only on case reports, and ends with verified references.

What to include

  • Every abnormal finding, including ones that seemed unrelated
  • Age of onset and how the picture has evolved
  • Family history across three generations and any consanguinity
  • Tests already done, including genetic results with the variant

Example cases

  • 19M with recurrent rhabdomyolysis after exercise and fasting, normal CK between episodes, a brother similarly affected. Differential and testing?
  • Angiokeratomas, burning pain in the hands and feet, proteinuria and left ventricular hypertrophy at 32. Diagnosis and confirmation?
  • Variant of uncertain significance in MYH7 found on a cardiomyopathy panel. What does it mean and what next?

What the report contains

  1. Conclusion and Recommendation
  2. Phenotype Summary
  3. Candidate Diagnoses
  4. Testing Strategy
  5. Information Needed
  6. Key Evidence & References

Limits

  • A variant of uncertain significance is not a diagnosis. The engine explains what could reclassify it, but it should not drive irreversible decisions, and interpretation belongs with clinical genetics.
  • Rare-disease evidence is thin, and some statements rest on case series or case reports; the report says when this is so, and specialist centres may know more than the published literature.
  • Genetic testing needs consent and counselling before and after, and its availability and cost differ widely between countries and hospitals. Those decisions rest with the treating team and the patient.

Frequently asked questions

What does a variant of uncertain significance mean?

It means the laboratory does not have enough evidence to call the variant either disease-causing or harmless. It should not be used to make a diagnosis, to guide irreversible treatment or to test relatives predictively. Evidence that can reclassify it includes segregation with disease in the family, functional or biochemical studies, a closer phenotype match and later reanalysis as databases grow.

When should a gene panel be chosen over exome or genome sequencing?

A panel suits a well-defined phenotype with a known group of genes, such as cardiomyopathy. Exome or genome sequencing suits a broad or unusual multisystem picture where the differential is wide. None of them catches every kind of change: repeat-expansion disorders, copy-number and other structural changes, imprinting disorders and mitochondrial DNA variants can need dedicated tests.

Does a negative family history rule out a genetic disease?

No. A condition can be inherited or genetic even when no relative is affected: the variant may be de novo, the inheritance recessive, penetrance or expression incomplete, the parent mosaic, the family small, or parentage different from that assumed. The engine records what the history suggests but does not use a negative history to exclude a genetic cause.

How does the engine avoid suggesting a rare disease too readily?

Ordinary explanations are considered before rare ones: an unusual form of a common illness, a drug, or two common problems together. Every rare candidate is then listed with the findings against it as well as those in favour, and with the single test most likely to settle the question.