Breast cancer family history calculator: from pedigree to referral decision
A practical guide for primary care, breast clinic, and genetic counselling teams — the red flags in a family history, the referral criteria that matter, the main risk models in mainstream use, and how Evagene ties the pedigree to a modern hereditary breast cancer risk workflow.
Short version. Family history is one of the strongest risk factors for breast cancer. A carefully constructed three-generation pedigree, annotated with cancer type, age of diagnosis, bilaterality, and male-breast or ovarian cancer coexisting, is the input every hereditary breast cancer risk model needs. Red-flag patterns prompt referral under NICE CG164 and NCCN guidelines. Once in specialist care, several models support the decisions that follow — BRCAPRO and BOADICEA for carrier probability and testing eligibility, Tyrer-Cuzick for lifetime risk and screening intensity, Gail and Claus for specific secondary purposes. Evagene runs BRCAPRO directly on the pedigree you draw, with no re-entry, and exports pedigree data cleanly to external tools such as CanRisk and IBIS where those are needed.
Family history as a risk factor
Breast cancer clusters in families more than most common cancers. Part of that clustering reflects shared environmental factors and selective reporting biases; part reflects genuine inherited risk. Large epidemiological studies have consistently shown that the relative risk associated with a single affected first-degree relative is roughly doubled compared with no family history, that the risk rises with the number of affected first-degree relatives, and that younger age at diagnosis in the affected relative compounds the effect.
The clinical task is to distinguish the family histories that suggest a specific hereditary cancer syndrome — most commonly hereditary breast and ovarian cancer (HBOC) driven by BRCA1 or BRCA2, but also Lynch syndrome, Li-Fraumeni, Peutz-Jeghers, Cowden, and others — from those that reflect familial clustering without an identifiable single-gene driver. The first group benefits from targeted genetic testing and management. The second benefits from enhanced risk-based screening without necessarily pursuing testing.
Risk models exist because this distinction cannot reliably be made by eye. Clinical intuition systematically over- and under-identifies at-risk families in predictable ways. Quantitative models provide a reproducible, guideline-compatible answer to the referral and management questions that follow a family history.
Red-flag patterns
Certain features of a family history strongly suggest a hereditary cancer syndrome. These are the signals that should prompt specialist referral and formal risk calculation.
- Early-onset breast cancer. Diagnoses under age 50, and particularly under age 40, in first- or second-degree relatives.
- Bilateral breast cancer. Especially if the first diagnosis was at a young age.
- Triple-negative breast cancer under age 60. Associated with elevated BRCA1 carrier probability regardless of family history.
- Male breast cancer. A strong signal for BRCA2 in particular. Any male breast cancer in the family warrants consideration.
- Ovarian cancer in the same family. Breast and ovarian cancer co-occurring in a kindred is the classic HBOC pattern. Any ovarian cancer in a first- or second-degree relative is usually referable on its own.
- Multiple primary cancers in one individual. Particularly breast and ovarian in the same person.
- Ashkenazi Jewish heritage. The BRCA1 and BRCA2 founder variants increase the prior probability of carrier status substantially, and referral criteria are correspondingly more permissive in many jurisdictions.
- Pedigree pattern of autosomal dominant inheritance. Cancers across generations with a 50 percent transmission pattern suggest a dominant susceptibility gene.
- Pancreatic or prostate cancer in the same family. Pancreatic cancer is part of the BRCA2-associated spectrum; aggressive early-onset prostate cancer is also associated with BRCA2. See our pancreatic cancer family history page for detail.
- Specific syndromic clues. Features of Cowden syndrome (macrocephaly, hamartomas, trichilemmomas), Li-Fraumeni (childhood cancers, sarcomas, early-onset breast), or Peutz-Jeghers (mucocutaneous pigmentation, hamartomatous polyps) prompt referral regardless of breast family history.
The absence of red flags does not always mean an absence of risk — population-based screening exists for good reasons — but their presence materially changes the clinical posture.
Referral criteria: NICE CG164 and NCCN
Published criteria translate red-flag patterns into specific referral thresholds. The two most widely cited frameworks are NICE CG164 (UK) and NCCN Breast/Ovarian/Pancreatic guidelines (US). They are not identical but overlap substantially.
NICE CG164 in the UK categorises family histories into near-population risk, moderate risk (typically equivalent to lifetime breast cancer risk between 17 and 30 percent), and high risk (lifetime risk of 30 percent or more, or a 10 percent or higher BRCA1/2 carrier probability). Moderate-risk women are managed via the breast surveillance pathway with age-adjusted mammography; high-risk women are offered additional MRI surveillance and genetic testing. Specific family-history triggers for specialist referral include two first-degree relatives with breast cancer (one under 50), three first- or second-degree relatives with breast cancer, a male breast cancer, a first-degree relative with bilateral breast cancer, and combinations involving ovarian cancer.
NCCN publishes regularly updated criteria for BRCA1/2, Lynch, and other syndromes. Triggers include specific ancestry-enriched variants (Ashkenazi Jewish founder variants in particular), early-onset breast cancer, triple-negative breast cancer under 60, male breast cancer, ovarian cancer, pancreatic cancer in the family for BRCA2 consideration, and defined combinations of affected relatives.
Both frameworks increasingly reference quantitative model output — a carrier probability above a defined threshold on BRCAPRO, BOADICEA, or equivalent — as an additional route to testing eligibility. The model is not a substitute for the criteria; it is a complementary way to reach the same decision.
The main breast cancer risk models
Five breast cancer risk models come up repeatedly in this clinical space. A short characterisation of each is useful.
| Model | Primary output | Inputs | Typical use |
|---|---|---|---|
| BRCAPRO | BRCA1/2 carrier probability | Pedigree, breast/ovarian/male breast, ages, Ashkenazi status, prior tests | Testing eligibility |
| BOADICEA (CanRisk) | Multi-gene carrier + lifetime risk | Pedigree, hormonal/reproductive, density, lifestyle, PRS | Comprehensive assessment |
| Tyrer-Cuzick (IBIS) | 10-year and lifetime breast risk | Pedigree, hormonal/reproductive, benign breast disease, density | Screening intensity, chemoprevention |
| Gail (BCRAT) | 5-year and lifetime risk | Age, first-degree relatives, hormonal, prior biopsy | Screening-age women without strong FH |
| Claus | Cumulative risk from FH alone | First- and second-degree breast cancer | Older literature baseline; less used today |
BRCAPRO is the focused BRCA1/BRCA2 carrier probability calculator from the BayesMendel suite. See our BRCAPRO calculator guide for details.
BOADICEA, delivered via the CanRisk tool from the University of Cambridge, is the broadest mainstream model — multi-gene, polygenic risk, density, and lifestyle. It is endorsed in NICE and NCCN. See our BOADICEA alternative page.
Tyrer-Cuzick (IBIS) estimates lifetime breast cancer risk integrating hormonal and reproductive factors with family history. Widely used in the UK for screening intensity. See our Tyrer-Cuzick alternative page.
Gail (the NCI's Breast Cancer Risk Assessment Tool, BCRAT) estimates 5-year and lifetime breast cancer risk. It is designed for women without a strong family history and uses limited family-history inputs (first-degree relatives only). For hereditary cancer clinics, Gail is not the primary tool; for screening-age women in primary care, it remains in common use.
Claus is a historical family-history-only model that pre-dates BRCAPRO and BOADICEA. Its continued appearance is mostly as a legacy literature reference; most modern services use the newer models.
The honest summary is that different models answer different clinical questions, and a well-run service selects among them based on the question in front of it.
From pedigree to referral decision
A typical workflow for a breast cancer family history evaluation proceeds as follows. First, construct a three-generation pedigree on both parental lineages, annotated with every cancer diagnosis (type, age, laterality), current age or age at death, and Ashkenazi Jewish heritage. Second, check the pedigree against red-flag patterns and referral criteria. Third, if red flags are present, run a carrier probability model (BRCAPRO or BOADICEA) and a lifetime risk model (Tyrer-Cuzick) as appropriate to the clinical question. Fourth, combine model output with guideline criteria to make the referral, testing, and surveillance decisions.
Each step depends on the pedigree being accurate. A pedigree that omits the paternal line, records undifferentiated "cancer" without site, or fails to capture ages at diagnosis produces poor model output. Time spent on pedigree quality pays back later in model accuracy and referral precision.
How Evagene supports the workflow
Evagene is a browser-based clinical pedigree management platform built for exactly this workflow. The pedigree is drawn once on a canvas designed for live-consultation speed — gesture drawing, keyboard shortcuts, standard clinical notation — and is annotated with the cancer diagnoses, ages, and ethnicity information that risk models need. Disease annotation uses ICD-10 and OMIM codes across a 200-plus disease catalogue, with breast, ovarian, and related cancers handled natively.
From the same pedigree, BRCAPRO runs directly with no re-entry. The BayesMendel package runs in an R sidecar process that invokes the validated code, so the calculation is the same one clinical research relies on. Results appear alongside the pedigree. For comprehensive risk with polygenic and lifestyle integration, Evagene supports GEDCOM 5.5.1 export so the pedigree can be taken into CanRisk; for lifetime-risk decisions, the same export and structured JSON feed into IBIS.
Two additional capabilities are relevant to breast cancer workflows. Batch risk screening runs BRCAPRO alongside MMRpro, PancPRO, and Mendelian analyses across the full disease catalogue in one operation, flagging conditions where the pedigree crosses thresholds. A family presenting with breast cancer but also containing a pancreatic case or a colorectal case has those signals surfaced automatically. AI interpretation with bring-your-own-key large language models drafts structured clinical summaries combining pedigree findings with risk results, which a clinician reviews. The BYOK model means traffic goes to your contracted LLM provider, not through a Evagene-hosted model.
For programmatic use, risk results are available through the REST API and MCP server. Documentation is at evagene.net/help. Evagene is free during Alpha via the waiting list, with zero install required.
Frequently asked questions
How does family history affect breast cancer risk?
A first-degree relative with breast cancer roughly doubles baseline risk. Risk rises with more affected relatives and younger age at diagnosis. Combined with ovarian cancer or male breast cancer in the same family, a hereditary cancer syndrome becomes more likely.
What are the red-flag patterns for hereditary breast cancer?
Early-onset, bilateral, triple-negative under 60, male breast cancer, ovarian cancer in the same family, multiple primaries, Ashkenazi Jewish heritage, and autosomal dominant transmission across generations.
When should someone be referred for genetic testing?
Following NICE CG164 or NCCN criteria, which include specific family-history patterns and quantitative model thresholds. Local commissioning frameworks provide the operational cut-offs.
What are the main breast cancer risk models?
BRCAPRO (BRCA1/2 carrier probability), BOADICEA (comprehensive multi-gene + polygenic + lifestyle), Tyrer-Cuzick (lifetime breast cancer risk), Gail (screening-age short- and long-term risk), and Claus (historical FH-only).
What does a good family history include?
Three generations on both parental sides, with cancer site, age of diagnosis, laterality, vital status, ethnicity, and any prior genetic testing results for every relative.
How does Evagene support the workflow?
Pedigree on a single canvas, BRCAPRO running directly on it via BayesMendel, batch screening across all catalogued diseases, AI-drafted reports with BYOK LLMs, and GEDCOM export to external tools.
Which model should I run first?
Depends on the clinical question. Testing eligibility: BRCAPRO or BOADICEA. Screening intensity: Tyrer-Cuzick. General screening-age risk: Gail. Many services run more than one.