Rankings

Incidence per 100,000 person-years, age-standardized to the standard population declared by the source. Rates standardized to different standard populations are never compared in one ranking.

Deaths per 100,000 person-years, age-standardized to the standard population declared by the source.

Number of deaths with this cancer as underlying cause in the reference year for the population scope.

Number of new cases diagnosed in the reference year for the population scope. Registry counts are observed; national/global figures may be estimates and are labeled as such.

5-year relative survival for the diagnosis period and population declared by the source. Population survival statistics do not predict an individual outcome.

Deaths divided by new cases in the same population, year and sex. A crude proxy of lethality; it is not a survival probability and is affected by incidence trends and registration completeness.

Interventional studies registered on ClinicalTrials.gov whose overall status is RECRUITING, NOT_YET_RECRUITING, ENROLLING_BY_INVITATION or ACTIVE_NOT_RECRUITING and whose conditions map to this cancer or one of its descendants in the NCIt hierarchy.

Active interventional studies with PHASE3 among their phases, mapped to this cancer or its descendants.

Interventional studies currently RECRUITING whose conditions map to this cancer or its descendants.

Interventional studies with PHASE3 among their phases and overall status RECRUITING, mapped to this cancer or its descendants.

Herfindahl–Hirschman index of lead sponsors over active interventional studies (sum of squared sponsor shares; 1 = a single sponsor). Requires ≥ 10 active studies.

Share of interventional studies first posted since 2010 that ended TERMINATED or WITHDRAWN among those that reached a terminal status (COMPLETED, TERMINATED, WITHDRAWN). Registrant-reported statuses; reasons are shown as posted and never inferred. Requires ≥ 30 terminal studies.

PubMed records matching the cancer’s query with a publication date in the last 12 months.

PubMed records matching the cancer’s MeSH-anchored query (stored verbatim with the count) with a publication date in the last 5 full years plus the current year.

Ratio of publications in the last 12 months to the average annual publications over the preceding 5-year window. Values above 1 indicate accelerating literature.

Relative change in the number of interventional studies first posted in the last 12 months versus the preceding 12 months, over the cancer and its descendants. Requires at least 20 studies in the prior window so small denominators do not produce spurious growth.

Accepted CIViC evidence items whose disease maps to this cancer or its descendants.

Distinct genes linked to this cancer (or descendants) by an accepted CIViC evidence item or by a GDC cohort frequency of at least 5% with ≥ 20 affected cases.

Open-access genomic studies (GDC projects) whose disease maps to this cancer or its descendants.

Active interventional studies divided by annual deaths (thousands) in the same population scope. A burden-normalized measure of clinical-research intensity; it depends on the epidemiology source of the scope and on how registrants phrase conditions.

PubMed records over the last 5 years divided by annual deaths (thousands) in the same population scope. A burden-normalized measure of research output.

Burden percentile minus research-activity percentile (publications, last 5 years) within the same scope. Positive values flag cancers with high mortality burden but comparatively little literature. Caveat (Phase 1): literature counts are query-based per entity (query stored on each row); they are not aggregated over descendants.

log₂ of a cancer’s share of deaths divided by its share of publications (last 5 years), both over the eligible top-level cancers of the scope. 0 = literature share matches death share; positive values flag comparatively under-published cancers. Caveat: literature counts are query-based per entity and not aggregated over descendants.

Burden percentile minus active-trial percentile within the same scope. Positive values flag cancers with high mortality burden but comparatively few active trials. A quantitative signal, not an accusation (CLAUDE.md §324). Caveat (Phase 1): trial counts aggregate a cancer and its NCIt descendants only; registrations phrased at a broader level (e.g. "colorectal cancer" for colon and rectal cancer) are attributed to the broader entity, which can overstate the gap of narrower top-level sites.

log₂ of a cancer’s share of deaths divided by its share of active interventional trials, both computed over the eligible top-level cancers of the scope. 0 = trial share matches death share; +1 = twice as many deaths as trials would suggest; −1 = half. Scale-free, so it can be compared across scopes, unlike the percentile-based Trial Gap Index.