Gail vs Tyrer-Cuzick vs CanRisk Breast Risk Models
Compare common breast cancer risk calculators by use case, inputs, output, and limitations.
1. Why Comparing Breast Risk Models Matters
Breast cancer risk assessment is not a single calculation. A high-risk breast clinic, mammography program, genetics service, and primary prevention visit may each need a different answer: 5-year invasive breast cancer risk, lifetime risk, MRI eligibility, pathogenic variant carrier probability, or combined breast and ovarian risk.
The Gail Model, Tyrer-Cuzick/IBIS, and CanRisk/BOADICEA overlap, but they are not interchangeable. Each model reflects a different evidence base, set of inputs, and intended clinical use. The safest workflow is to choose the model that matches the decision being made, then interpret the number in guideline and patient context.
2. Gail vs Tyrer-Cuzick vs CanRisk: At-a-Glance Comparison
| Model | Best use | Key inputs | Outputs | OncoToolkit link |
|---|---|---|---|---|
| Gail / BCRAT | Simple risk estimate for invasive breast cancer; commonly used for chemoprevention thresholds. | Age, race/ethnicity, reproductive history, breast biopsy/atypia history, first-degree family history. | 5-year and lifetime invasive breast cancer risk. | Gail calculator |
| Tyrer-Cuzick / IBIS | Screening and high-risk clinic triage when broader family history and breast density are relevant. | Personal/reproductive factors, benign breast disease, extended family history, genetic context, breast density depending on version. | Often 10-year and lifetime breast cancer risk. | IBIS guide |
| CanRisk / BOADICEA | Genetics-oriented risk assessment, carrier probability, and combined breast/ovarian risk discussions. | Detailed pedigree, pathogenic variants, polygenic risk score, mammographic density, reproductive, hormonal, lifestyle, and tumor factors where available. | Future cancer risk and pathogenic variant carrier probabilities. | CanRisk guide |
2.1 What each model is trying to answer
- Gail/BCRAT asks: "What is this woman's absolute risk of invasive breast cancer over the next 5 years and up to age 90 using a compact set of common clinical variables?"
- Tyrer-Cuzick/IBIS asks: "What is the breast cancer risk when family history, benign breast disease, hormonal factors, and often breast density are brought into a richer screening-risk model?"
- CanRisk/BOADICEA asks: "What is the future cancer risk and probability of carrying a susceptibility gene variant when detailed family history and genetic information are central?"
Clinical caution: Do not average outputs from different models. If a Tyrer-Cuzick lifetime risk is above an MRI threshold but Gail is not, that does not mean the "true" answer is halfway between them. It means the models are answering different questions with different information.
3. Practical Workflow for Choosing a Breast Risk Model
- Define the clinical decision first. Is the visit about chemoprevention, MRI screening, genetic testing, risk-reducing surgery, or counseling an unaffected relative?
- Check whether a protocol names a model. Imaging centers and genetics clinics may specify Tyrer-Cuzick, CanRisk, BRCAPRO, or another model for eligibility documentation.
- Collect the minimum reliable data. For Gail, a brief reproductive and biopsy history may be enough. For CanRisk, a three-generation pedigree and test results may be necessary.
- Document the version and assumptions. Tyrer-Cuzick estimates can change when breast density is added; CanRisk estimates can change when genetic test results or polygenic risk scores are entered.
- Translate the estimate into a conversation. Risk models support shared decision-making; they do not replace individualized assessment by breast, genetics, radiology, or oncology teams.
4. Common Clinical Scenarios
Scenario A: Chemoprevention discussion in a woman with atypia
Start with Gail/BCRAT if the local prevention pathway uses 5-year invasive breast cancer risk. A history of atypical hyperplasia can move risk meaningfully, but the estimate should still be interpreted alongside age, thromboembolic risk, menopausal status, uterine history, bone health, and patient preferences.
Useful internal links: Gail calculator and Gail model guide.
Scenario B: Dense breasts and MRI eligibility
Tyrer-Cuzick/IBIS is commonly used in imaging workflows because it can incorporate broader family history and breast density in current implementations. MRI thresholds often refer to lifetime risk from family-history-based models, so confirm the model specified by the screening program.
Useful internal link: Tyrer-Cuzick / IBIS guide.
Scenario C: Multiple relatives with breast and ovarian cancer
CanRisk/BOADICEA is usually a better fit than Gail because the question is no longer just population-level invasive breast cancer risk. The key issues are pedigree structure, age at diagnosis in relatives, bilateral disease, ovarian/tubal/peritoneal cancer, known pathogenic variants, and carrier probability.
Useful internal link: CanRisk / BOADICEA guide.
Scenario D: Average-risk screening visit
A formal risk model may still be useful, but it should not distract from age-appropriate screening, symptoms, breast exam findings, prior imaging, and local population screening recommendations. A "low" modeled risk does not rule out cancer in a symptomatic patient.
5. Limitations and Safety Checks
| Issue | Why it matters | Practical check |
|---|---|---|
| Incomplete family history | CanRisk and Tyrer-Cuzick are sensitive to the number, age, and lineage of affected relatives. | Record maternal and paternal lineage separately and update the estimate when new information arrives. |
| Model-version drift | IBIS and BOADICEA have evolved to include additional predictors such as density or genetic data. | Document version, density category, and whether genetic results or PRS were included. |
| Known pathogenic variant | Gail is not designed to manage high-penetrance hereditary cancer syndromes. | Use genetics-specific tools and refer to a genetics professional where appropriate. |
| Personal cancer history | Some risk models estimate first primary breast cancer risk and are not intended after prior breast cancer. | Verify eligibility criteria before applying the result to surveillance or prevention decisions. |
6. Frequently Asked Questions
Is Tyrer-Cuzick better than Gail?
Not universally. Tyrer-Cuzick includes more family-history and risk-factor detail and is often more useful for MRI screening triage. Gail remains practical for 5-year invasive breast cancer risk and prevention-medication discussions. "Better" depends on the decision.
When should CanRisk replace Tyrer-Cuzick?
CanRisk is preferred when genetic susceptibility, carrier probability, ovarian cancer risk, polygenic risk score, or a complex pedigree is central to the question. Tyrer-Cuzick may remain appropriate for imaging workflows that specifically request an IBIS lifetime risk.
What should I do if the patient's risk crosses 20% lifetime risk in one model only?
First check inputs, version, breast density, and family-history details. Then follow the local protocol for which model determines MRI eligibility. Discordance should prompt careful documentation and, when needed, high-risk clinic or genetics review.
Start With the Model That Matches the Decision
Use Gail for compact 5-year/lifetime invasive risk, Tyrer-Cuzick for screening-risk triage, and CanRisk for genetics-oriented breast and ovarian risk assessment.
Open Breast Cancer ToolsReferences
- National Cancer Institute. Breast Cancer Risk Assessment Tool (BCRAT/Gail Model). Source
- Ikonopedia. IBIS Risk Assessment Tool v8.0b. Source
- CanRisk. Official CanRisk web tool. Source
- Lee A, et al. BOADICEA: a comprehensive breast cancer risk prediction model incorporating genetic and nongenetic risk factors. Genet Med. 2019;21:1708-1718. Source
- American Cancer Society. Breast cancer screening recommendations for women at high risk. Source
- Joint ABS-UKCGG-CanGene-CanVar consensus regarding clinical use of CanRisk. Br J Cancer. 2024. Source