Inheritance & Family History
"How removed is the relative?" only has a clean answer in one of two situations. Separating them is the whole game — the same family history means very different things depending on whether a specific gene variant has already been identified.
Situation A
A pathogenic variant is known in the family
Clean Mendelian math. The question is simply "did this relative inherit that one variant?" — answered cheaply and definitively by cascade testing for that specific variant.
Situation B
Family history of cancer, no known variant
There is no 50/25/12.5 shortcut. Risk is estimated with a validated tool (Tyrer-Cuzick / BOADICEA), and the action is risk assessment → refer to genetic counseling.
Situation A — a known familial variant
When the family's pathogenic variant is identified, each relative's chance of carrying it follows autosomal-dominant inheritance and halves with each degree of separation:
| Relationship | Degree | Chance of carrying the variant |
|---|---|---|
| Parent, child, full sibling | 1st | 50% |
| Grandparent, aunt/uncle, niece/nephew, half-sibling | 2nd | 25% |
| First cousin, great-grandparent, great-aunt/uncle | 3rd | 12.5% |
The trap that voids the table
A negative test in an intervening relative breaks the chain. A first cousin's risk is only 12.5% if the connecting aunt/uncle is untested. If that aunt/uncle tested negative for the family variant, the cousin's risk drops to population baseline. Untested intervening relatives keep the probability alive; a tested-negative one collapses it.
Practical point: once a variant is known, testing a relative is single-site and inexpensive — it looks for one change, not a whole panel. This is the one scenario where the FP's message is concrete: "your family's variant is known; a targeted test tells you yes or no."
Situation B — family history, no known variant
Here the halving table does not apply — you don't know a variant exists, so there is nothing to halve. Most hereditary-looking families never have an identified gene. Risk is estimated from the pedigree with a validated tool, and meaningful histories are referred for formal risk assessment and possible panel testing.
Family-history features that should prompt risk assessment (USPSTF 2019)
- Breast cancer diagnosed before age 50
- Bilateral breast cancer
- Breast and ovarian cancer — in one person or one lineage
- Male breast cancer
- Multiple relatives with breast cancer
- A relative with two primary BRCA-related cancers
- Ashkenazi Jewish ancestry
USPSTF grade B: screen these women with a brief familial-risk tool (Ontario FHAT, Manchester, Referral Screening Tool, Pedigree Assessment Tool, 7-question FHS, Tyrer-Cuzick), and refer positives for counseling ± testing. Grade D: do not routinely assess/test average-risk women without these features.
Carrying the variant ≠ getting the cancer
The percentages above are the chance of inheriting the variant. The chance of actually developing cancer is that number × the gene's penetrance. A first-degree relative of a BRCA1 carrier has a 50% chance of carrying it, and if they do, a ~55–72% lifetime breast-cancer risk — not a 50% cancer risk. Keep the two numbers separate when you counsel.
Why a "reassuring" family history can mislead
- Paternal transmission is under-recognized. These genes pass through fathers just as often, but "no breast cancer on Dad's side" often means there were only one or two women on that side to get it. Ask about the paternal lineage explicitly.
- Small families and few female relatives can't exclude anything. An uninformative pedigree (few women, early deaths from other causes) is not a low-risk pedigree — it's a pedigree that can't answer the question.
- De novo and mosaic variants happen — TP53 especially. An absent family history does not rule out a hereditary syndrome, particularly with very-early-onset or bilateral disease.
- Ancestry shifts pretest probability (e.g., Ashkenazi Jewish founder variants), independent of how many relatives are affected.
The family-physician move
- Take a three-generation history, both sides, with ages at diagnosis.
- Run a brief validated risk tool on anyone with the red-flag features above.
- If a familial variant is already known → arrange targeted cascade testing.
- If not, but the history is concerning → refer to genetic counseling. Don't order a multigene panel and interpret it yourself.
Then use the Breast Cancer Genes reference for what a returned result means and what to expect from the counselor.
Sources: USPSTF, BRCA-Related Cancer: Risk Assessment, Genetic Counseling, and Genetic Testing (2019, grade B; AAFP-endorsed); Kuchenbaecker et al., JAMA 2017 (BRCA penetrance); Pal et al., ACMG "Inherited Cancer Risk on a Continuum" 2026. General education for clinicians — not a substitute for formal genetic counseling.