Addressable population estimator
Market sizing is three numbers multiplied: how many people get the cancer, what share carry the target, and what share reach the treatment setting. All three already live in this corpus. This page multiplies them for 150 target and cancer pairs across 27 cancers, shows every input with its source, and reports a range rather than a point.
- 1New cases of Non-small-cell lung cancer in World, 2022GLOBOCAN site: Trachea, bronchus and lung. GLOBOCAN 2022, International Agency for Research on Cancer (IARC), Global Cancer Observatory. GLOBOCAN does not split lung cancer by histology. NSCLC is roughly 85% of lung cancer; figures are for all lung cancer.2.48m
- 2Share of that site which is Non-small-cell lung cancerNSCLC share of lung cancer. American Cancer Society, Cancer Facts & Figuresx 80 to 85%
- 3Patients whose tumour carries or expresses KRASRecorded as “25-30” (Adenocarcinoma, any KRAS mutation). G12C ~13% of adenocarcinoma. Sourcex 25 to 30%
- 4Advanced or metastatic (de novo plus relapsed)Roughly half of lung cancer is distant at diagnosis; most regional disease also recurs. Editorial range informed by SEER Cancer Stat Facts: Lung and bronchus (stage at diagnosis).x 60 to 75%
This is arithmetic on public inputs, not a forecast of treated patients or revenue. It ignores diagnosis rates, testing rates, access, competition, duration of therapy and price.
Prevalence figures come from the target pages and are mostly from Western cohorts. Stage at diagnosis (which drives the setting share) is later in most of the world, so the true advanced-disease share is higher outside high-income countries.
GLOBOCAN reports by organ site. Where a subtype share is applied, it is an editorial range with its source shown.
Incidence fetched 2026-09-08. Change a factor you disagree with by editing src/data/setting-shares.ts or the target’s prevalence row.
Every target with a prevalence figure in Non-small-cell lung cancer, World, advanced or metastatic (de novo plus relapsed)
| Target | Prevalence as recorded | Parsed range | Addressable patients per year |
|---|---|---|---|
| 80-90 (IHC, any expression) | 80 to 90% | 953k to 1.42m | |
| 70-80 (Adenocarcinoma, any expression) | 70 to 80% | 834k to 1.27m | |
| 60-70 (IHC, moderate-high) | 60 to 70% | 714k to 1.11m | |
| 20-45 (Response by PD-L1 stratum (proxy)) | 20 to 45% | 238k to 712k | |
| 25-30 (Adenocarcinoma, any KRAS mutation) | 25 to 30% | 298k to 474k | |
| 25-30 (TPS >=50%) | 25 to 30% | 298k to 474k | |
| 20-30 (Squamous enriched) | 20 to 30% | 238k to 474k | |
| 20-25 (High expression (>=50% cells 2+/3+)) | 20 to 25% | 238k to 395k | |
| 15 (MTAP deletion) | 15% | 179k to 237k | |
| 10-15 (Activating mutation (US/Europe)) | 10 to 15% | 119k to 237k | |
| 5-10 (ATM loss/mutation (sensitising context)) | 5 to 10% | 60k to 158k | |
| 3-5 (Rearrangement) | 3 to 5% | 36k to 79k | |
| 3-4 (Exon 14 skipping) | 3 to 4% | 36k to 63k | |
| 2-4 (V600E and non-V600) | 2 to 4% | 24k to 63k | |
| 2-3 (ERBB2 exon 20 mutation) | 2 to 3% | 24k to 47k | |
| 1-2 (Fusion) | 1 to 2% | 12k to 32k |
Inputs
Incidence is GLOBOCAN 2022 new cases by country, summed for continents, the same data behind cases by country. Where GLOBOCAN reports an organ site broader than the OnCo cancer (all breast cancer, all lung cancer, all leukaemia), a sourced subtype share range is applied.
Prevalence is the figure recorded on each target page for that cancer, parsed into a range (“15-20”, “>95”, “~25”). Setting share is an editorial range per cancer in src/data/setting-shares.ts, informed by the SEER stage-at-diagnosis distribution plus relapse after early-stage treatment; 27 cancers have one.
Limits
This is the size of the biological population, not the treated or revenue-generating one: it ignores diagnosis and testing rates, line of therapy beyond the setting chosen, access, competing options, treatment duration and price. Prevalence figures are mostly from Western cohorts and some vary by ethnicity (EGFR in lung cancer, for instance). Stage at diagnosis is later outside high-income countries, so the advanced-disease share is understated there.
Disagree with a factor? Every one is a single line in a data file with a source next to it. The crowding index uses the same arithmetic.