Skip to content

Statistics — retinoblastoma

Last curated: 2026-09-01 (independent audit, same date; every figure below was re-checked against a live PubMed record).

Every row names population, year/era and method. Conflicting estimates are shown side by side and are not averaged.

Global burden and presentation

Measure Estimate Population / era Method Source
Annual diagnoses ~8,000 Worldwide; estimate current to 2015 disease-primer synthesis Dimaras 2015, PMID 27189421
Global cohort size 4,064 149 countries; diagnosed 2017 prospective cluster cohort Global Study Group 2022, PMID 35839812
Participating centres 260 149 countries prospective recruitment PMID 35839812
Median age at diagnosis 23.2 months (IQR 11.0–36.5) 2017 global cohort patient-level median PMID 35839812
cT4, high-income 0.8% (5/636) 2017 cTNMH staging PMID 35839812
cT4, upper-middle-income 5.4% (55/1,027) 2017 cTNMH staging PMID 35839812
cT4, lower-middle-income 19.7% (342/1,738) 2017 cTNMH staging PMID 35839812
cT4, low-income 42.9% (196/457) 2017 cTNMH staging PMID 35839812
Enucleation availability 100% of enrolled children 2017 global centres centre report PMID 35839812
IV chemotherapy availability 98.8% (4,014/4,064) 2017 global centres centre report PMID 35839812
Incidence, Europe 4.0 per million children 0–14 (95% CI 3.9–4.1) 81 registries, 31 countries, 2000–2013 population-based registry Virgili 2024, PMID 39388193
5-year survival, Europe 97.8% (95.5–98.9) 3,180 cases population-based registry PMID 39388193
Incidence, Argentina 5.0 per million children 0–14 (95% CI 3.5–6.4) 438 patients, 2000–2009 national paediatric registry Moreno 2014, PMID 24729462
Incidence, Brazil 2.23 per million (world-standardised, 0–14) 675 patients, 28 registries, 2000–2018 population-based registry Barbosa 2022, PMID 36507141
Incidence, Ethiopia 1 per 52,156 live births 221 patients, 4 hospitals, 2017–2020 birth-cohort analysis; authors call it an underestimate Sherief 2023, PMID 37221479
Asia-Pacific share of global burden 43% (3,452 of 8,099) 2013 estimates, 6 countries modelled apportionment Jain 2019, PMID 30385881
Median age at diagnosis, high- vs low-income 14.1 vs 30.5 months 4,351 patients, 153 countries, 2017 cross-sectional Fabian 2020, PMID 32105305
Extraocular at diagnosis, low-income 49.1% (256/521) same cross-sectional PMID 32105305
Metastasis at diagnosis, low-income 18.9% (94/498) same cross-sectional PMID 32105305
Advanced disease, low- vs upper-middle/high-income OR 17.92 (95% CI 12.94–24.80) same, age-adjusted logistic regression PMID 32105305
Genetic testing available 74 of 145 countries (51.0%) 438 centres, 2024 cross-sectional service survey Global Study Group 2026, PMID 42463304
Intra-arterial chemotherapy available 69 of 145 countries (47.6%) same service survey PMID 42463304
Intravitreal chemotherapy available 101 of 145 countries (69.7%) same service survey PMID 42463304
Countries with no identified Rb centre 49 2024 service survey PMID 42463304

Income-stratified survival

Stratum 3-year survival Population Method Source
High income 99.5% (95% CI 98.8–100.0) diagnosed 2017 time-to-event PMID 35839812
Upper-middle income 91.2% (89.5–93.0) diagnosed 2017 time-to-event PMID 35839812
Lower-middle income 80.3% (78.3–82.3) diagnosed 2017 time-to-event PMID 35839812
Low income 57.3% (52.1–63.0) diagnosed 2017 time-to-event PMID 35839812
Low- vs high-income mortality HR 16.67 (95% CI 4.76–50.00) 2017 global cohort adjusted Cox model PMID 35839812
cT4 vs cT1 mortality HR 8.98 (4.44–18.18) 2017 global cohort adjusted Cox model PMID 35839812
Age effect to 3 years HR 1.38 per year (1.23–1.56) 2017 global cohort adjusted Cox model PMID 35839812

Delay and abandonment

Measure Estimate Population / era Method Source
Symptom-to-treatment lag mean 4.2 months; range 0.5–61.6 1,120 Indian children; prospective interval decomposition Das 2025, PMID 40719713
Parent-lag share 44% same component share PMID 40719713
Diagnosis-lag share 26% same component share PMID 40719713
Treatment-lag share 31% same component share PMID 40719713
Extraocular disease at diagnosis 25.2% same prospective staging PMID 40719713
Stage III abandonment 38.5% single resource-limited cohort retrospective Pant 2017, PMID 29337595
Stage IV abandonment 46.6% same retrospective PMID 29337595
Ethiopian lag mean 14.31 months; range 0.25–62.25 38 caregivers cross-sectional phone survey Sherief 2023, PMID 36803347
≥3-month first-care delay 76.3% same caregiver report PMID 36803347
≥1 additional facility before treatment centre 97.4% same referral history PMID 36803347

Classification and eye salvage

System / group Eye salvage Population / treatment Method Source
cT1 95% (139/147) 565 eyes; systemic chemotherapy + focal retrospective prediction Yousef 2021, PMID 33769386
cT2 56% (230/410) same same PMID 33769386
cT3 0% same same PMID 33769386
ICRB A 98% same same PMID 33769386
ICRB B 93% same same PMID 33769386
ICRB C 76% same same PMID 33769386
ICRB D 44% same same PMID 33769386
Focal seeds 62% same morphology subgroup PMID 33769386
Diffuse seeds 42% same morphology subgroup PMID 33769386
High-risk pathology after primary enucleation 55.1% (751/1,362) 16 centres, 11 countries retrospective Kurian 2025, PMID 39922380
PPV of group D for high-risk pathology 42.0% / 35.1% / 43.2% ICRB Philadelphia / ICRB LA / COG same cohort PMID 39922380
PPV of group E for high-risk pathology 58.5% / 59.0% / 59.5% same three schemes same cohort PMID 39922380
Group D eyes upstaged to E by ICRB vs IIRC 17% (57/150) 332 advanced eyes, single institution reclassification Singh 2024, PMID 38830602
12-month ocular survival, ICRB D / E1 / E2 / E3 79% / 59% / 49% / 1% same Kaplan-Meier PMID 38830602
Primary enucleation rate for group D, by IIRC version 8% (Philadelphia), 34% (COG), 37% (CHLA) 39 centres, 1,807 group D eyes survey Scelfo 2017, PMID 28730089
Curable stage (IRSS 0/I/II) by presenting sign strabismus 94.8%, leukocoria 90.1%, orbital mass 33.7% 4,578 patients, 121 countries cross-sectional Yanagihara 2025, PMID 40334716
5-year disease-specific survival by pT stage pT1 99.5%, pT2a 95.5%, pT3b 93.0%, pT3c/d 92.3%, pT4 40.9% 700 primarily enucleated, 29 centres retrospective Zhao 2021, PMID 34944860
Diffuse infiltrating retinoblastoma frequency 4.6% (132/2,854 eyes; 95% CI 3.9–5.5) international registry, 2001–2013 registry series Tomar 2025, PMID 40738737
5-year survival, cT3 diffuse infiltrating vs non-diffuse 82% (78–86) vs 94% (93–95), P<0.001 same Kaplan-Meier PMID 40738737

Intra-arterial chemotherapy

Measure Estimate Population / era Method Source
2-year progression-free globe salvage, IAC 53% (38/72) unilateral group D/E; six Chinese hospitals randomized trial Wen 2023, PMID 37536351
Same, IV chemotherapy 27% (19/71) same randomized trial PMID 37536351
Relative effect RR 1.97 (95% CI 1.27–3.07) same intention-to-treat PMID 37536351
Myelosuppression, IAC 51% (37/72) same prespecified safety PMID 37536351
Myelosuppression, IV 70% (50/71) same prespecified safety PMID 37536351
Ophthalmic-artery occlusion 3% (2/72) IAC arm safety PMID 37536351
Ophthalmic-artery stenosis 18% (13/72) IAC arm safety PMID 37536351
Group D OAC avoidance of enucleation/EBRT 78.6% (88/112) single-centre; mixed prior treatment retrospective Abramson 2016, PMID 26756643
Treatment-naive group D 85.1% (40/47) same subgroup PMID 26756643
Primary IAC group D salvage 94% 70-eye series retrospective Shields 2014, PMID 24656794
Primary IAC group E salvage 36% same retrospective PMID 24656794
Pooled globe salvage, IAC 73% (95% CI 68–77) 43 studies, 3,174 eyes random-effects meta-analysis Weinberger 2025, PMID 40541288
Pooled globe salvage, group D vs E 66% (60–71) vs 49% (39–59) same meta-analysis PMID 40541288
IAC vs IV overall success 75.7% (65.7–83.6) vs 69.5% (51.9–82.8), P<0.001 26 studies, 1,541 eyes meta-analysis Chen 2018, PMID 29703164
IAC vs IV, group D salvage 79.5% (71.8–85.4) vs 55.1% (45.6–64.2), P<0.001 same meta-analysis PMID 29703164
Multi-institutional IAC feasibility success 67% 14 patients, 9 institutions (COG ARET12P1) prospective feasibility Chintagumpala 2024, PMID 37817345
Catheterisation success, single high-volume centre 98.5% of procedures 95 eyes, 289 injections prospective registry Gobin 2011, PMID 21320950

Intravitreal chemotherapy

Measure Estimate Population Method Source
Injections in initial safety series 135 in 30 eyes selected eyes retrospective Munier 2012, PMID 22368262
Extraocular spread 0 detected same; median follow-up 13.5 months clinical/pathologic observation PMID 22368262
Vitreous haemorrhage 3/135 injections same adverse events PMID 22368262
Continuing seed regression 100% (40/40 eyes) median 3-year follow-up retrospective series Shields 2016, PMID 26630319
Globe salvage 88% (35/40) same eye-level outcome PMID 26630319
Complete regression 78% (21/27) recurrent/refractory seeds retrospective Yousef 2021, PMID 34322023
Focal-seed regression 100% (10/10) same subgroup PMID 34322023
Diffuse-seed regression 65% (11/17) same subgroup PMID 34322023
Any ocular adverse effect 59% (16/27) same adverse-event collection PMID 34322023

Enucleation pathology and adjuvant treatment

Measure Estimate Population Method Source
5-year DFS, 3 CEV cycles 90.4% 187 unilateral high-risk enucleated patients randomized noninferiority trial Ye 2024, PMID 39432296
5-year DFS, 6 cycles 89.2% same randomized trial PMID 39432296
Difference 1.2% (95% CI −7.5 to 9.8) same noninferiority margin 12% PMID 39432296
RFS with adjuvant chemotherapy for isolated massive choroidal invasion 97.2% 60-child observational cohort retrospective comparison Feng 2023, PMID 37603354
RFS without adjuvant chemotherapy 55.6% same retrospective comparison PMID 37603354
OS with vs without adjuvant 97.2% vs 66.7% same retrospective comparison PMID 37603354
Number treated to prevent one relapse/death 3 same derived estimate PMID 37603354
Massive choroidal invasion by continent (AS/AUS/EUR/NA/SA) 31% / 7% / 13% / 19% / 27% (P=0.001) 1,426 primarily enucleated eyes retrospective global Kaliki 2024, PMID 39151183
Postlaminar optic-nerve invasion by continent 27% / 0% / 16% / 21% / 19% (P=0.0006) same retrospective global PMID 39151183
High-risk features present after primary enucleation 36% (145/403) Asian Indian case-control retrospective Kaliki 2015, PMID 25841975
Orbital recurrence without vs with adjuvant chemotherapy (PLONI) 31% vs 2% (P<0.001) 292 patients, 9 countries retrospective cohort Lin 2026, PMID 41475682
Metastasis subhazard ratio, no adjuvant chemotherapy (PLONI) 5.39 (95% CI 1.75–16.60) same competing-risks model PMID 41475682
2-year EFS with vs without high-risk features (COG) 0.96 (0.89–0.98) vs 0.99 (0.96–1.0) 321 enucleated unilateral, central review prospective Chévez-Barrios 2019, PMID 31539297
Central-vs-local histopathology disagreement 23% of high-risk, 17% of non-high-risk cases same central review PMID 31539297
Unanimous high-risk features among specialists 3 of 15 listed features (100% agreement) 27 respondents, 16 countries survey Kaliki 2022, PMID 34762098

Extraocular and trilateral disease

Measure Estimate Population Method Source
1-year EFS stage II–III 88.1% (90% CI 66.6–96.2) ARET0321, 57 eligible total prospective trial Dunkel 2022, PMID 35820112
1-year EFS stage IVa 82.6% (61.0–92.9) same prospective trial PMID 35820112
1-year EFS stage IVb/trilateral 28.3% (12.7–46.2) same prospective trial PMID 35820112
Therapy-related deaths 2 same safety PMID 35820112
Heritable trilateral risk 3.78% 607-patient prospective cohort incidence de Bloeme 2024, PMID 38992673
Follow-up scans per pineal tumour detected 494 all cystic/suspicious glands followed diagnostic yield PMID 38992673
Scans per tumour if limited to suspicious glands 22 same modeled protocol restriction PMID 38992673
Baseline MRI missed later asynchronous pineal TRb 89% meta-analysis, 197 cases pooled case timing de Jong 2022, PMID 33939299
Scans per asymptomatic pineal TRb detected (6-monthly to 36 months) ≥311 meta-analysis, 138 pineal TRb cases modelled screening yield de Jong 2020, PMID 32061409
Scans per life saved, same schedule ≥776 same modelled PMID 32061409
Time from diagnosis to orbital disease (untreated) median 13.7 months 44 children who abandoned treatment retrospective natural history Zhao 2021, PMID 34359552
Time from metastasis to death (untreated) median 2.0 months same retrospective PMID 34359552
Mortality by 12 / 24 / 36 / 48 months (untreated) 36% / 77% / 95% / 100% same retrospective PMID 34359552

Subsequent malignancies — do not merge heritability groups

Measure Estimate Population / era Method Source
Any SMN SIR, heritable 11.9 (95% CI 10.4–13.5) 1,128 survivors; diagnosis 1914–2006 cohort vs general population Schonfeld 2021, PMID 33473166
Any SMN SIR, non-heritable 0.8 (0.5–1.2) 924 survivors same PMID 33473166
First SMN cumulative incidence at 50 years 33.1% (29.0–37.2) heritable competing-risk PMID 33473166
Second SMN cumulative incidence at 50 years 6.0% (3.8–8.2) heritable competing-risk PMID 33473166
Melanoma/CNS/oral/breast SIR range 3.1–17 across four sites heritable site-specific SIRs summarized as range PMID 33473166
Melanoma SIR, Danish cohort 26.78 (9.78–58.30) 133 heritable survivors; 1943–2013 registry cohort Gregersen 2020, PMID 33090227
Radiation-associated sarcoma RR at ≥60 Gy 10.7 nested case-control dose-response Wong 1997, PMID 9333268
Head/neck bone sarcoma SIR 2,213 (1,671–2,873) 952 irradiated heritable survivors cohort Kleinerman 2019, PMID 31622129
Body/extremity bone sarcoma SIR 169 (115–239) same cohort PMID 31622129
50-year SMN mortality, heritable 25.5% (20.8–30.2) US long-term survivors competing-risk Yu 2009, PMID 19351917
50-year SMN mortality, non-heritable 1.0% (0.2–1.8) same competing-risk PMID 19351917

Vision and psychosocial outcomes

Measure Estimate Population Method Source
Detailed visual-testing sample 11 survivors age 5–17 cross-sectional Reynolds 2023, PMID 37442536
Impaired contrast/saccade parameters 10/10 completing tests same objective testing PMID 37442536
AYA QOL sample 101 survivors adolescent/young adult cross-sectional Belson 2023, PMID 36520266
Variance in overall QOL explained by depression + social participation 38% same regression PMID 36520266
Additional variance from vision psychosocial domain 16% same regression PMID 36520266
School-age psychosocial sample 69 survivors mean age 10.89 years survivor/parent survey Morse 2023, PMID 36385462
Ethiopian qualitative sample 13 parents all care phases phenomenological interviews Sherief 2023, PMID 36545916
Strabismus after successful salvage 69% of 42 patients mean 93.7 months follow-up retrospective Fabian 2018, PMID 30009948
Deviation before vs after strabismus surgery (eso) 42.0 ± 10.4 → 8.5 ± 5.3 prism dioptres (P<0.001) 18 salvaged patients retrospective Masoomian 2024, PMID 38481156
Foveal atrophy after macular retinoblastoma 77% (36/47 eyes) globe salvaged in all retrospective Chandra 2022, PMID 37006899
Retinoblastoma-specific PROMs available 1 of 143 identified instruments 110 eligible studies scoping review Janic 2020, PMID 32770435

Known conflicts and caveats

  • The 51% 50-year second-cancer estimate in Wong 1997 and 33.1% in Schonfeld 2021 differ by cohort, era, endpoint and competing-risk methods; they must not be averaged (PMIDs: 9333268, 33473166).
  • The Danish melanoma SIR 26.78 is a small national cohort estimate with a wide CI; it is not the melanoma-specific figure from Schonfeld (PMIDs: 33090227, 33473166).
  • Globe-salvage estimates are treatment-, stage- and centre-selected and are not useful-vision probabilities.
  • The global income analysis uses national income categories; it cannot isolate individual household income or one causal delay mechanism.
  • Small zero-event procedural series cannot prove zero metastatic risk.
  • Percentages may not sum exactly because of rounding.
  • Incidence figures use incompatible denominators — per million children 0–14, per million 0–19, and per live birth — and differ in registry completeness (13% of Brazilian cases were death-certificate-only). They are listed side by side and must not be pooled (PMIDs: 39388193, 36507141, 37221479, 24729462).
  • Trilateral screening yield estimates of ≥311 scans per detected tumour (modelled, 2020), 494 scans (prospective, all cystic glands, 2024) and 22 scans (prospective, suspicious glands only, 2024) describe different protocols, not conflicting facts (PMIDs: 32061409, 38992673).
  • Melanoma SIR 26.78 (Danish) and the 3.1–17 multi-site range (Schonfeld) are different quantities from different cohorts; the Danish overall heritable SIR (11.39) and Schonfeld's (11.9) are close, but the non-heritable estimates differ (1.52 vs 0.8) (PMIDs: 33090227, 33473166).
  • Meta-analytic IAC-versus-IV comparisons carry unequal follow-up (mean 21.7 vs 49.4 months), which biases the metastasis comparison toward IAC (PMID 29703164).