Statistics quick reference — colorectal adenocarcinoma¶
Last curated: 2026-10-26
Method. Every row states figure, source year, population and method. GLOBOCAN and GBD estimates are shown side by side and never averaged. Trial percentages are intention-to-treat unless stated. Legacy PMIDs were re-retrieved live on 2026-08-30; added PMIDs were retrieved live on 2026-10-12, 2026-10-19 or 2026-10-26. Rows added 2026-10-26 come from structured PubMed abstracts; their full texts were not opened.
Global burden and epidemiology¶
| Statistic | Figure | Source/year | Population | Method |
|---|---|---|---|---|
| Incident cancers, colorectum including anus | 2,041,007 | Sung 2026, PMID 42417444 | Global, 2024; ICD-10 C18–C21 | GLOBOCAN 2024 registry/model estimate, Table 1; not adenocarcinoma-only |
| Deaths, colorectum including anus | 917,895 | Sung 2026, PMID 42417444 | Global, 2024; ICD-10 C18–C21 | GLOBOCAN 2024 registry/model estimate, Table 1 |
| Share of all incident cancers, colorectum including anus | 9.9% | Sung 2026, PMID 42417444 | Global, 2024 | GLOBOCAN 2024, Table 1 |
| Share of all cancer deaths, colorectum including anus | 9.4% | Sung 2026, PMID 42417444 | Global, 2024 | GLOBOCAN 2024, Table 1 |
| Share of all incident cancers | 9.6% | Bray 2024, PMID 38572751 | 185 countries, 2022 | GLOBOCAN registry/model estimate |
| Share of all cancer deaths | 9.3% | Bray 2024, PMID 38572751 | 185 countries, 2022 | GLOBOCAN estimate |
| Incident CRC cases | >1.9 million | Morgan 2023, PMID 36604116 | 185 countries, 2020 | GLOBOCAN |
| CRC deaths | 930,000 | Morgan 2023, PMID 36604116 | 185 countries, 2020 | GLOBOCAN |
| Projected incident cases | 3.2 million | Morgan 2023, PMID 36604116 | Global, 2040 | Demographic projection |
| Projected deaths | 1.6 million | Morgan 2023, PMID 36604116 | Global, 2040 | Demographic projection |
| Male incidence, high region | 40.6/100,000 | Morgan 2023, PMID 36604116 | Australia/New Zealand + European regions, 2020 | Age-standardized GLOBOCAN |
| Female incidence, low region | 4.4/100,000 | Morgan 2023, PMID 36604116 | Several African regions/Southern Asia, 2020 | Age-standardized GLOBOCAN |
| Male mortality, high region | 20.2/100,000 | Morgan 2023, PMID 36604116 | Eastern Europe, 2020 | Age-standardized GLOBOCAN |
| Female mortality, low region | 2.5/100,000 | Morgan 2023, PMID 36604116 | Southern Asia, 2020 | Age-standardized GLOBOCAN |
| Incident CRC cases | 2.17 million (95% UI 2.00–2.34) | GBD Collaboration 2022, PMID 34967848 | 204 locations, 2019 | GBD model; not interchangeable with GLOBOCAN |
| CRC deaths | 1.09 million (95% UI 1.00–1.15) | GBD Collaboration 2022, PMID 34967848 | 204 locations, 2019 | GBD model |
| CRC DALYs | 24.3 million (95% UI 22.6–25.7) | GBD Collaboration 2022, PMID 34967848 | 204 locations, 2019 | GBD model |
| Projected young-onset share of colon cancer | 11% | Spaander 2023, PMID 37105987 | Review projection, 2030 | Literature synthesis/projection |
| Projected young-onset share of rectal cancer | 23% | Spaander 2023, PMID 37105987 | Review projection, 2030 | Literature synthesis/projection |
| Young-onset hereditary syndrome fraction | ~20% | Spaander 2023, PMID 37105987 | Age <50 cases | Review synthesis |
| US early-onset CRC incidence AAPC, 2019–2023 | 8.3% per year | Al Khaldi 2026, PMID 42691927 | SEER, age <50 | Joinpoint trend; females 9.1% vs males 7.6%; rectal 9.0% vs colon 4.5% |
| Nonmetro vs large-metro male CRC mortality | 23% higher | Islami 2024, PMID 37962495 | USA, recent registry period | Descriptive population analysis |
| Nonmetro vs large-metro female CRC mortality | 21% higher | Islami 2024, PMID 37962495 | USA | Descriptive population analysis |
Screening and prevention¶
| Statistic | Figure | Source/year | Population | Method |
|---|---|---|---|---|
| NordICC colonoscopy uptake | 42% | Bretthauer 2022, PMID 36214590 | 84,585 invitees, Europe | Randomized invitation trial |
| 10-y CRC incidence | 0.98% vs 1.20% | Bretthauer 2022, PMID 36214590 | Invitation vs usual care | Intention-to-screen |
| Incidence risk ratio | 0.82 (95% CI 0.70–0.93) | Bretthauer 2022, PMID 36214590 | Same | Randomized ITT |
| 10-y CRC mortality | 0.28% vs 0.31% | Bretthauer 2022, PMID 36214590 | Same | Randomized ITT |
| Mortality risk ratio | 0.90 (95% CI 0.64–1.16) | Bretthauer 2022, PMID 36214590 | Same | Randomized ITT |
| FIT pooled CRC sensitivity | 0.79 | Lee 2014, PMID 24658694 | Average-risk screening studies | Systematic review/meta-analysis |
| FIT pooled specificity | 0.94 | Lee 2014, PMID 24658694 | Same | Meta-analysis |
| Next-gen stool DNA CRC sensitivity | 93.9% | Imperiale 2024, PMID 38477986 | Average-risk screening cohort | Cross-sectional validation |
| Next-gen stool DNA advanced-precursor sensitivity | 43.4% | Imperiale 2024, PMID 38477986 | Same | Cross-sectional validation |
| Next-gen stool DNA specificity | 90.6% | Imperiale 2024, PMID 38477986 | No advanced neoplasia | Cross-sectional validation |
| Blood cfDNA CRC sensitivity | 83.1% | Chung 2024, PMID 38477985 | ECLIPSE average-risk participants | Prospective validation |
| Blood cfDNA specificity | 89.6% | Chung 2024, PMID 38477985 | No advanced neoplasia | Prospective validation |
| Blood cfDNA advanced-precursor sensitivity | 13.2% | Chung 2024, PMID 38477985 | Advanced precancerous lesions | Prospective validation |
| ADR effect on interval-cancer risk | HR 0.97 per 1-point ADR increase | Corley 2014, PMID 24693890 | 314,872 colonoscopies, 136 endoscopists | Observational provider-level analysis |
| Polypectomy CRC-mortality reduction | 53% vs expected | Zauber 2012, PMID 22356322 | National Polyp Study, median 15.8 y | Cohort vs population expected mortality |
| EPoS II 5-y CRC incidence, 5- vs 3-y first surveillance | 0.77% vs 0.82% | Jover 2026, PMID 42748427 | 10,799 high-risk adenoma patients | Interim 5.5-y analysis of 10-y NI trial |
| Navigation vs usual care, screening uptake RR | 1.58 (95% CI 1.23–2.02) | Ramai 2026, PMID 41932450 | 76 RCTs | Network meta-analysis |
| Mailed FIT vs usual care, screening uptake RR | 1.36 (95% CI 1.07–1.74) | Ramai 2026, PMID 41932450 | Same | Network meta-analysis |
| DENEB CRC sensitivity | 91% (95% CI 80–96) | Mannucci 2026, PMID 42575119 | Testing cohort, age ≥45 | Locked RT-qPCR miRNA assay; case-control |
| DENEB advanced-adenoma sensitivity | 81% (95% CI 69–89) | Mannucci 2026, PMID 42575119 | Same | Same; not a screening-program mortality trial |
Pathology and hereditary yield¶
| Statistic | Figure | Source/year | Population | Method |
|---|---|---|---|---|
| Clear CRM at rectal surgery | 87% (354/408) | MERCURY 2006, PMID 16984925 | 408 rectal cancers | Prospective MRI–pathology cohort |
| MRI specificity for clear CRM | 92% | MERCURY 2006, PMID 16984925 | Same | Prospective diagnostic accuracy |
| Clear CRM when MRI predicted clear | 94% (327/349; 95% CI 91–96) | MERCURY 2006, PMID 16984925 | Same | Prospective diagnostic accuracy |
| Technically satisfactory MRI | 93% (379/408) | MERCURY 2006, PMID 16984925 | Same | Multicenter quality assessment |
| Median node harvest | 12 (range 0–49) | MERCURY 2006, PMID 16984925 | Same | Resection pathology |
| Node-survival studies positive | 16/17 | Chang 2007, PMID 17374833 | 61,371 colon patients | Systematic review |
| Universal-screen MMRd prevalence | 6.22% (95% CI 5.08–7.61) | Eikenboom 2022, PMID 33887476 | 58,580 CRCs | Random-effects meta-analysis |
| Germline MMR pathogenic variant | 2.00% (95% CI 1.59–2.50) | Eikenboom 2022, PMID 33887476 | Same | Meta-analysis |
| Missing MMR IHC | 11.81% | Eikenboom 2022, PMID 33887476 | Same | Workflow meta-analysis |
| Germline completion among eligible | 76.30% | Eikenboom 2022, PMID 33887476 | Same | Workflow meta-analysis |
| Commercial-panel any PGV | 14.2% (4,864/34,244) | Coughlin 2022, PMID 36370464 | CRC patients undergoing MGPT | Retrospective commercial cohort |
| CRC/polyposis-gene PGV | 9.1% | Coughlin 2022, PMID 36370464 | Same | Retrospective cohort |
| Variant of uncertain significance | 38.2% | Coughlin 2022, PMID 36370464 | Same | Retrospective cohort |
| Prospective unselected PGV | 15.5% (56/361) | Uson 2022, PMID 33857637 | Mayo multicenter CRC | Prospective universal panel |
| Incremental actionable finding | 9.4% | Uson 2022, PMID 33857637 | Same | Guideline/panel comparison |
| Management changed by germline result | 11% | Uson 2022, PMID 33857637 | Same | Prospective follow-up |
| Relative cascade testing uptake | 16% | Uson 2022, PMID 33857637 | Families of positive patients | Prospective follow-up |
| Ohio universal/tiered PGV | 7.1% (234/3,310) | Pearlman 2021, PMID 34250417 | 51 Ohio hospitals | Prospective statewide study |
| MMR-gene PGV | 4.3% | Pearlman 2021, PMID 34250417 | Same | Prospective study |
| Syndromes missed by UTS alone | 38.6% | Pearlman 2021, PMID 34250417 | Same | Counterfactual testing analysis |
Localized colon cancer¶
| Statistic | Figure | Source/year | Population | Method |
|---|---|---|---|---|
| MOSAIC stage III 10-y OS | 67.1% vs 59.0% | André 2015, PMID 26527776 | Resected stage III | Randomized FOLFOX vs FL follow-up |
| Stage III absolute OS difference | 8.1 points | André 2015, PMID 26527776 | Same | Derived trial difference |
| Stage II OS benefit | Not demonstrated | André 2015, PMID 26527776 | Resected stage II subgroup | Randomized subgroup |
| DYNAMIC chemotherapy use | 15% vs 28% | Tie 2022, PMID 35657320 | 455 stage II patients | Randomized ctDNA strategy |
| DYNAMIC 2-y RFS | 93.5% vs 92.4% | Tie 2022, PMID 35657320 | ctDNA-guided vs standard | Noninferiority randomized trial |
| DYNAMIC absolute chemo reduction | 13 points | Tie 2022, PMID 35657320 | Same | Derived trial difference |
| QUASAR estimated 5-y survival gain | ~3.6% | QUASAR 2007, PMID 18083404 | Mostly stage II colorectal | Randomized chemo vs observation |
| TRIBE2 PFS after two treatments | 19.2 vs 16.4 mo | Cremolini 2020, PMID 32164906 | First-line mCRC; contextual comparator | Phase III randomized |
| IROCAS 3-y DFS, mFOLFIRINOX vs mFOLFOX6 | 65.6% vs 66.3%; HR 1.07 (95% CI 0.84–1.36) | Bennouna 2026, PMID 42927853 | pT4 and/or pN2 stage III colon, n=795 | Phase III randomized |
| NICHE-2 3-y DFS | 100% (97.5% CI 100–100) | Chalabi 2026, PMID 42906564 | Clinical stage II–III dMMR colon, per-protocol n=111 | Single-arm phase II; neoadjuvant nivo+ipi |
| NICHE-2 pathological complete response | 68% (75/111) | Chalabi 2024, PMID 38838311 | Same | Single-arm phase II |
Localized rectal cancer¶
| Statistic | Figure | Source/year | Population | Method |
|---|---|---|---|---|
| RAPIDO 3-y disease-related treatment failure | 23.7% vs 30.4% | Bahadoer 2021, PMID 33301740 | MRI high-risk LARC | Phase III randomized |
| RAPIDO failure HR | 0.75 (95% CI 0.60–0.95) | Bahadoer 2021, PMID 33301740 | Same | ITT |
| RAPIDO 5-y locoregional recurrence | 44/431 (10%) vs 26/428 (6%) | Dijkstra 2023, PMID 36661037 | Experimental TNT vs standard | Trial follow-up |
| PRODIGE 23 3-y DFS | 76% vs 69% | Conroy 2021, PMID 33862000 | LARC | Phase III randomized |
| PRODIGE 23 DFS HR | 0.69 (95% CI 0.49–0.97) | Conroy 2021, PMID 33862000 | Same | ITT |
| PROSPECT selective preop radiation | 9.1% | Schrag 2023, PMID 37272534 | Intermediate-risk rectal, FOLFOX arm | Phase III randomized |
| OPRA 5-y DFS | 71% vs 69% | Verheij 2024, PMID 37883738 | Induction vs consolidation TNT | Randomized phase II follow-up |
| OPRA 5-y TME-free survival | 39% vs 54% | Verheij 2024, PMID 37883738 | Same | Kaplan–Meier trial follow-up |
| Initial dostarlimab cCR | 12/12 | Cercek 2022, PMID 35660797 | dMMR stage II/III rectal | Single-arm phase II initial report |
| Major LARS prevalence | 53.1% (254/478) | Pieniowski 2020, PMID 32530135 | Population rectal-surgery cohort, mean 6.7 y | Cross-sectional population study |
| No LARS prevalence | 22.6% (108/478) | Pieniowski 2020, PMID 32530135 | Same | Derived complement of reported 77.4% any LARS |
| TNTCRT 3-y DFS | 74.8% vs 66.0% | Wang 2026, PMID 42492015 | High-risk LARC | Phase III TNT vs nCRT |
| TNTCRT DFS HR | 0.674 (95% CI 0.489–0.929) | Wang 2026, PMID 42492015 | Same | ITT |
| TNTCRT pCR | 26.37% vs 9.80% | Wang 2026, PMID 42492015 | Same | Pathology |
| STAR-TREC 12-mo TME-free, LCCRT vs SCRT | 78.5% vs 60.6% | Bach 2026, PMID 42633754 | Organ-preservation choosers, mrT1–T3bN0 | 12-month interim of 30-mo primary |
| UNION 3-y EFS (interim) | 91.3% vs 77.4% | Tao 2026, PMID 42927765 | cT3–4/N+ LARC, China, not MMR-selected | Phase III; SCRT→camrelizumab+CAPOX vs LCRT→CAPOX; investigator-assessed |
| UNION EFS HR (interim) | 0.33 (95% CI 0.15–0.71) | Tao 2026, PMID 42927765 | Same | Stratified; P=0.0029 vs boundary 0.0102 |
| AZUR-1 sustained 12-mo cCR (investigator) | 68% | Cercek 2026, PMID 42927858 | Stage II/III dMMR/MSI-H LARC, n=144 with ≥12 mo follow-up | Single-arm interim; dostarlimab alone |
Metastatic systemic therapy and precision oncology¶
| Statistic | Figure | Source/year | Population | Method |
|---|---|---|---|---|
| CALGB 80405 median OS | 30.0 vs 29.0 mo | Venook 2017, PMID 28632865 | KRAS-WT first-line mCRC | Randomized cetuximab vs bevacizumab |
| TRIBE2 PFS2 | 19.2 vs 16.4 mo | Cremolini 2020, PMID 32164906 | Fit untreated mCRC | Phase III randomized |
| TRIBE2 PFS2 HR | 0.74 (95% CI 0.63–0.88) | Cremolini 2020, PMID 32164906 | Same | ITT |
| KEYNOTE-177 median PFS | 16.5 vs 8.2 mo | André 2020, PMID 33264544 | Untreated MSI-H/dMMR mCRC | Phase III randomized |
| KEYNOTE-177 PFS HR | 0.60 (95% CI 0.45–0.80) | André 2020, PMID 33264544 | Same | ITT |
| KEYNOTE grade ≥3 treatment AE | 22% vs 66% | André 2020, PMID 33264544 | Pembrolizumab vs chemotherapy | Randomized safety |
| BEACON median OS | 9.3 vs 5.9 mo | Tabernero 2021, PMID 33503393 | Previously treated BRAF V600E mCRC | Phase III randomized |
| BEACON OS HR | 0.61 (95% CI 0.48–0.77) | Tabernero 2021, PMID 33503393 | Same | Updated ITT |
| BEACON response | 19.5% vs 1.8% | Tabernero 2021, PMID 33503393 | Encorafenib–cetuximab vs control | RECIST |
| MOUNTAINEER response | 38.1% (95% CI 27.7–49.3) | Strickler 2023, PMID 37142372 | HER2+, RAS-WT refractory mCRC | Single-arm phase II |
| CodeBreaK 300 median PFS | 5.6 vs 2.2 mo | Fakih 2023, PMID 37870968 | KRAS G12C refractory mCRC | Phase III randomized |
| CodeBreaK 300 PFS HR | 0.49 | Fakih 2023, PMID 37870968 | High-dose combination vs control | Randomized |
| CodeBreaK 300 response | 26.4% vs 0% | Fakih 2023, PMID 37870968 | Same | RECIST |
| SUNLIGHT median OS | 10.8 vs 7.5 mo | Prager 2023, PMID 37133585 | Refractory mCRC | Phase III TAS-102–bev vs TAS-102 |
| SUNLIGHT OS HR | 0.61 | Prager 2023, PMID 37133585 | Same | ITT |
| RECOURSE median OS | 7.1 vs 5.3 mo | Mayer 2015, PMID 25970050 | Refractory mCRC | Phase III TAS-102 vs placebo |
| RECOURSE OS HR | 0.68 | Mayer 2015, PMID 25970050 | Same | ITT |
| CORRECT median OS | 6.4 vs 5.0 mo | Grothey 2013, PMID 23177514 | Refractory mCRC | Phase III regorafenib vs placebo |
| CORRECT OS HR | 0.77 | Grothey 2013, PMID 23177514 | Same | ITT |
| FRESCO-2 median OS | 7.4 vs 4.8 mo | Dasari 2023, PMID 37331369 | Refractory mCRC | Phase III fruquintinib vs placebo |
| FRESCO-2 OS HR | 0.66 | Dasari 2023, PMID 37331369 | Same | ITT |
| SOLARIS median PFS | 11.8 vs 10.3 mo | Ng 2026, PMID 42545685 | Untreated mCRC, chemo + bevacizumab | Phase III high- vs standard-dose vitamin D3; PMC author manuscript |
| SOLARIS PFS HR | 0.92 (95% CI 0.73–1.16) | Ng 2026, PMID 42545685 | Same | Unstratified Cox; one-sided log-rank P = 0.25 |
| COMMIT PFS HR, combo vs atezo | 0.439 (95% CI 0.23–0.84) | Rocha Lima 2026, PMID 42525921 | First-line dMMR/MSI-H mCRC | Phase III; chemo-alone arm closed |
| COMMIT ORR | 86.1% vs 46% | Rocha Lima 2026, PMID 42525921 | Same | RECIST |
| CaPP3 ITT HR, 100 vs 600 mg aspirin | 0.97 (95% CI 0.67–1.42) | Burn 2026, PMID 42425127 | Lynch syndrome | Time to first Lynch-syndrome cancer |
| CheckMate 8HW OS HR, nivo+ipi vs nivo, all lines | 0.61 (95% CI 0.46–0.82) | Lonardi 2026, PMID 42926106 | Centrally confirmed MSI-H/dMMR mCRC | Phase III final analysis; median follow-up 68.0 mo |
| CheckMate 8HW OS HR, nivo+ipi vs chemo, first line | 0.37 (95% CI 0.24–0.57) | Lonardi 2026, PMID 42926106 | Same | Final analysis |
| CheckMate 8HW OS HR, nivo+ipi vs nivo, first line | 0.66 (95% CI 0.44–0.99); P=0.0432 | Lonardi 2026, PMID 42926106 | Same | Did not meet prespecified P<0.006 |
| PRABITAS median OS, 28-day vs 14-day TAS-102–bev | 10.3 vs 11.3 mo; HR 0.93 (95% CI 0.80–1.08) | Sakakida 2026, PMID 42925880 | Refractory mCRC, 178 Japanese hospitals, n=970 | Pragmatic phase III non-inferiority; margin 1.25 |
| PRABITAS grade ≥3 neutropenia | 39% vs 13% | Sakakida 2026, PMID 42925880 | Same | 28-day vs 14-day |
| Anti-EGFR resistant-clone half-life | ~4.4 mo | Parseghian 2019, PMID 30462160 | Serial plasma after EGFR withdrawal | Exponential ctDNA model |
Liver and lung metastasis¶
| Statistic | Figure | Source/year | Population | Method |
|---|---|---|---|---|
| EORTC 40983 3-y PFS absolute gain, randomized | 7.3 points | Nordlinger 2008, PMID 18358928 | ≤4 resectable liver metastases | Phase III ITT |
| EORTC 40983 PFS HR, randomized | 0.79 (95% CI 0.62–1.02) | Nordlinger 2008, PMID 18358928 | Same | ITT |
| EORTC 40983 median OS | 61.3 vs 54.3 mo | Nordlinger 2013, PMID 24120480 | Same | 8.5-y trial follow-up |
| EORTC 40983 OS HR | 0.88 (95% CI 0.68–1.14) | Nordlinger 2013, PMID 24120480 | Same | ITT |
| RFA local recurrence per lesion | 6.0% | Tanis 2014, PMID 24411080 | EORTC CLOCC selected lesions | Nonrandomized cross-trial analysis |
| Resection local recurrence per lesion | 5.5% | Tanis 2014, PMID 24411080 | EORTC EPOC selected lesions | Nonrandomized cross-trial analysis |
| RFA recurrence for lesions ≤30 mm | 2.9% | Tanis 2014, PMID 24411080 | Small ablated lesions | Lesion-level analysis |
| LAVA-CRLM one-year freedom from local progression | 77.7% MWA vs 83.0% SBRT; HR 0.87 (95% CI 0.43–1.74) | Risumlund 2026, PMID 42829779 | 92 randomized patients in modified ITT; lesions eligible for both | Two-centre phase II; Figure 2/Results |
| LAVA-CRLM grade ≥2 treatment-related toxicity | 13/52 (25%) MWA vs 1/40 (3%) SBRT | Risumlund 2026, PMID 42829779 | Treated safety population | Table 2/Results; trial not powered for toxicity comparison |
| ASAC DFS HR, aspirin vs placebo | 1.08 (95.44% CI 0.85–1.38) | Yaqub 2026, PMID 42777749 | 428 treated after curative-intent CRLM | Phase III; full analysis set |
| ASAC 36-mo OS | 76% vs 85% | Yaqub 2026, PMID 42777749 | Same | HR for death 1.64 (95% CI 1.01–2.65) |
| PulMiCC median survival | 3.5 vs 3.8 y | Milosevic 2020, PMID 32388895 | 93 randomized lung-metastasis patients | Updated randomized analysis |
| PulMiCC death HR | 0.93 (95% CI 0.56–1.56) | Milosevic 2020, PMID 32388895 | Metastasectomy vs control | Underpowered randomized trial |
| SECA-I transplant 5-y OS | 56% | Dueland 2015, PMID 24950280 | 21 transplant recipients | Nonrandomized comparison |
| Chemotherapy comparator 5-y OS | 9% | Dueland 2015, PMID 24950280 | 47 NORDIC VII liver-only patients | Nonrandomized comparator |
| SECA-II 5-y OS | 83% | Dueland 2020, PMID 31188200 | Strictly selected transplant recipients | Prospective single-arm |
| SECA-II 3-y DFS | 35% | Dueland 2020, PMID 31188200 | Same | Kaplan–Meier |
| Expanded-criteria transplant median OS | 18 mo | Smedman 2020, PMID 32333527 | 10 SECA-II arm D patients | Prospective single-arm |
| Expanded-criteria transplant median DFS | 4 mo | Smedman 2020, PMID 32333527 | Same | Prospective single-arm |
Known conflicts and caveats¶
- GBD versus GLOBOCAN: 2019 GBD (2.17 million cases) and 2020 GLOBOCAN (>1.9 million) differ by year, inputs and modeling; do not average. The 2024 GLOBOCAN table additionally groups colon, rectum and anus (ICD-10 C18–C21), so its 2,041,007 cases are not an adenocarcinoma-only estimate (Sung 2026, PMID 42417444).
- Screening efficacy versus strategy: NordICC intention-to-screen includes 42% uptake; per-protocol estimates lose randomization.
- Stool/blood accuracy versus mortality: cross-sectional sensitivity does not prove mortality reduction.
- Node count: association may reflect stage migration and care quality rather than a biological dose response.
- ctDNA: prognostic HRs are not treatment-effect HRs; DYNAMIC evaluates a specific algorithm.
- Rectal organ preservation: clinical complete response and TME-free survival are not synonymous with cure.
- Late-line medians: trial-eligible performance status limits generalization.
- Metastasectomy: observational survival is dominated by selection; PulMiCC control survival was unexpectedly high.
- Transplant: OS can be long despite early recurrence; graft opportunity cost is absent from Kaplan–Meier survival.
- Subgroup estimates: biomarker and regimen subgroups often lack powered interaction tests.