Skip to content
Bending the Curve Trial by Trial
?

Oesophagus (upper gastrointestinal tract)

Oesophageal cancer

How chemoradiotherapy, preoperative strategies and immunotherapy shaped care for oesophageal cancer.

5-year relative survival · United States, SEER 8 registries

3.95%→21.79%

1975 (observed value; first year in the table) to 2018 (observed value; last year with an observed value in the table). Read these with the caveats under “Then and now”.Source

8 milestones, by decade

  • In the CROSS trial, 368 patients with resectable oesophageal or junctional cancer were enrolled (366 analysed: 75% adenocarcinoma, 23% squamous-cell carcinoma, 2% large-cell undifferentiated carcinoma) and assigned to surgery alone or to weekly carboplatin and paclitaxel with 41.4 Gy radiotherapy before surgery. Complete resection with no tumour within 1 mm of the margins (R0) was achieved in 92% versus 69%, and median overall survival was 49.4 versus 24.0 months (hazard ratio 0.657, 95% CI 0.495 to 0.871; P=0.003). A ten-year report (median follow-up 147 months) gave a hazard ratio for overall survival of 0.70 (95% CI 0.55 to 0.89), with 10-year overall survival of 38% versus 25%. The comparator was surgery alone, so these figures say nothing about preoperative chemotherapy.
  • CheckMate 577 randomised 794 patients with resected stage II or III oesophageal or gastro-oesophageal junction cancer and residual pathological disease after neoadjuvant chemoradiotherapy, 2:1 and double-blind, to nivolumab for up to 1 year or placebo. Median disease-free survival was 22.4 versus 11.0 months (hazard ratio 0.69; 96.4% CI 0.56 to 0.86; P<0.001), with a median follow-up of 24.4 months. Disease-free survival was the primary endpoint; the abstract gives no overall survival result.

Evidence current to October 2026 · Checked against primary sources to the extent possible

A guided tour of 4 milestones. You can pause or step through it yourself.

Filter milestonesshowing all 8
Treatment type
Setting

Molecular origins

  1. 1992 · Treatment · ChemotherapyRTOG 85-01: chemoradiotherapy versus radiotherapy aloneIn a US trial, people with localised oesophageal cancer had radiotherapy alone or chemotherapy plus radiotherapy. The trial stopped early because survival was longer with chemotherapy plus radiotherapy, at the cost of more severe side effects. Later trials found no significant survival gain from a higher radiation dose, and shorter median survival when cetuximab was added.RTOG 85-01 (N=121 at stopping): fluorouracil and cisplatin with 50 Gy vs 64 Gy alone; median survival 12.5 vs 8.9 months, 24-month survival 38% vs 10% (P<0.001, printed after the 12- and 24-month rates); 5-year OS 26% (95% CI 15 to 37) vs 0% in the long-term report. INT 0123 (218 eligible): 64.8 Gy vs 50.4 Gy, median survival 13.0 vs 18.1 months (not significant). SCOPE1 (N=258): added cetuximab, median OS 22.1 vs 25.4 months (adjusted HR 1.53).

Targeted therapy

  1. 1999 · Understanding · UnderstandingReflux as a risk factor for oesophageal adenocarcinomaA Swedish population-based study interviewed people with oesophageal adenocarcinoma and a comparison group. People with frequent, long-standing reflux symptoms had a much higher risk of this cancer. A later US registry analysis argued that the rise in oesophageal adenocarcinoma since the 1970s reflects a real increase in disease, not just more diagnosis.Lagergren et al. (Sweden, population-based case-control study, 1995 to 1997): 189 oesophageal adenocarcinomas, 262 cardia adenocarcinomas, 167 squamous-cell carcinomas, 820 controls. Recurrent reflux symptoms: odds ratio 7.7 (95% CI 5.3 to 11.4) for oesophageal adenocarcinoma, 2.0 (1.4 to 2.9) for cardia adenocarcinoma, 1.1 (0.7 to 1.9) for squamous-cell carcinoma. Pohl and Welch (SEER): incidence rose about sixfold, 4 to 23 per million, 1975 to 2001.

Immunotherapy

  1. 2012 · Treatment · ChemotherapyCROSS: neoadjuvant chemoradiotherapy for oesophageal cancerIn a Dutch trial, people with cancer of the oesophagus or the junction with the stomach had either surgery alone or five weeks of chemotherapy with radiotherapy followed by surgery. Median survival was about 49 months with chemoradiotherapy and 24 months with surgery alone, and the survival gain was still present ten years later.CROSS (366 analysed; 75% adenocarcinoma, 23% squamous): weekly carboplatin and paclitaxel with 41.4 Gy, then surgery, versus surgery alone. R0 resection 92% vs 69%; median OS 49.4 vs 24.0 months (HR 0.657; 95% CI 0.495 to 0.871; P=0.003). Long-term: median OS 48.6 vs 24.0 months (HR 0.68); squamous HR 0.48, adenocarcinoma HR 0.73. Ten-year OS 38% vs 25% for chemoradiotherapy vs surgery alone (HR 0.70; 95% CI 0.55 to 0.89).
  2. 2018 · Detection · Detection & preventionBarrett oesophagus: AspECT, BEST3 and BOSS randomised trialsThree randomised trials, run mainly in the UK, looked at Barrett oesophagus, a change in the lining of the gullet linked to oesophageal adenocarcinoma. One tested acid-suppressing medicine with or without aspirin, one tested a swallowable sponge test for finding Barrett oesophagus, and one found that regular surveillance endoscopy did not improve survival compared with endoscopy only when needed.AspECT (N=2557, 2x2 factorial, median follow-up 8.9 years): high-dose vs low-dose esomeprazole time ratio 1.27 (95% CI 1.01 to 1.58) for death, adenocarcinoma or high-grade dysplasia; aspirin 1.24 (0.98 to 1.57; P=0.068). BEST3 (13,514 randomised): Barrett oesophagus diagnosed in 140 of 6834 vs 13 of 6388 within 12 months (rate ratio 10.6). BOSS (3453): overall survival HR 0.95 (95% CI 0.82 to 1.10), surveillance vs endoscopy at need.

Precision & engineered

  1. 2021 · Treatment · ImmunotherapyCheckMate 577: adjuvant nivolumab after chemoradiotherapy and surgerySome people with oesophageal or junction cancer still have cancer cells in the removed tissue after chemoradiotherapy and surgery. In this trial they received nivolumab or a placebo for up to a year. The median time without return of cancer or death was about 22 months with nivolumab and 11 months with placebo. FDA approved the use in 2021.CheckMate 577: randomised 2:1, double-blind, placebo-controlled phase 3; resected stage II or III oesophageal or GEJ cancer with residual pathological disease after neoadjuvant chemoradiotherapy; nivolumab for up to 1 year (N=794). Median DFS 22.4 vs 11.0 months (HR 0.69; 96.4% CI 0.56 to 0.86; P<0.001); median follow-up 24.4 months. Grade 3 or 4 treatment-related adverse events 13% vs 6%.
  2. 2021 · Treatment · ImmunotherapyFirst-line PD-1 therapy: KEYNOTE-590 and CheckMate 648Two large trials added an immune checkpoint inhibitor to chemotherapy as the first treatment for advanced oesophageal cancer. In both, median survival was longer with pembrolizumab or nivolumab added than with chemotherapy alone, and the survival difference was larger in the subgroups with higher PD-L1 expression. FDA approvals followed in 2021 and 2022.KEYNOTE-590 (N=749; squamous-cell carcinoma and adenocarcinoma): pembrolizumab plus fluorouracil and cisplatin vs placebo plus chemotherapy; median OS in squamous-cell carcinoma 12.6 vs 9.8 months (HR 0.72; 95% CI 0.60 to 0.88), and 13.9 vs 8.8 months in squamous-cell carcinoma with CPS 10 or more. CheckMate 648 (N=970; squamous only): median OS 15.4 vs 9.1 months with tumour-cell PD-L1 1% or more (nivolumab plus chemotherapy vs chemotherapy; HR 0.54; 99.5% CI 0.37 to 0.80); nivolumab plus ipilimumab also improved OS.
  3. 2025 · Treatment · ChemotherapyESOPEC: perioperative FLOT versus CROSS in adenocarcinomaTwo randomised trials compared the two main approaches before surgery for oesophageal adenocarcinoma: chemotherapy before and after surgery, or chemotherapy with radiotherapy before surgery. ESOPEC, using the FLOT regimen, reported longer survival with chemotherapy alone; Neo-AEGIS, mostly using an older regimen, found similar survival but closed early.ESOPEC (N=438, resectable oesophageal adenocarcinoma): perioperative FLOT vs preoperative carboplatin-paclitaxel with 41.4 Gy, plus surgery; 3-year OS 57.4% vs 50.7% (HR 0.70; 95% CI 0.53 to 0.92; P=0.01); 3-year PFS 51.6% vs 35.0% (HR 0.66). Neo-AEGIS (362 analysed; closed early, underpowered): perioperative chemotherapy, predominantly a modified MAGIC regimen, vs CROSS-type trimodality therapy; median OS 48.0 vs 49.2 months (HR 1.03; 95% CI 0.77 to 1.38).
  4. 2025 · De-escalation · SurgerySANO: active surveillance versus surgery after complete responseIn the Dutch SANO trial, people whose oesophageal cancer could no longer be detected after chemotherapy with radiotherapy either had surgery straight away or were closely monitored, with surgery later if needed. Two-year survival with monitoring met the trial's non-inferiority definition, but the authors say longer follow-up is needed.SANO (stepped-wedge cluster-randomised non-inferiority phase 3, 12 Dutch hospitals; 309 patients with clinical complete response after neoadjuvant chemoradiotherapy): 2-year OS 74% (95% CI 69 to 78) with active surveillance vs 71% (62 to 78) with standard surgery, modified intention-to-treat (HR 1.14; 95% CI 0.74 to 1.78); intention-to-treat 75% vs 70% (HR 0.83). Median follow-up 38 months.

Then and now

  • 5-year relative survival · United States, SEER 8 registries

    • 3.95%

      1975 (observed value; first year in the table)

    All races, both sexes; SEER Cancer Stat Facts category 'esophageal cancer' (the page does not state the histologies included, does not report squamous cell carcinoma and adenocarcinoma separately, and does not say whether gastro-oesophageal junction tumours are counted). SEER 8 five-year relative survival series, 1975-2018. Age range is not stated for this series.

    Modeled trend value for 1975: 4.32%. This is the lowest observed value in the table; the highest is 24.74% in 2016. Relative survival excludes deaths from other causes. 'Observed' is the table's label for the actual yearly value, as opposed to the Joinpoint modeled-trend value in the adjacent column. The column is headed 'Year' and the chart note says only 'SEER 8 5-Year Relative Survival Percent from 1975-2018'; the page does not say the year is the year of diagnosis. SEER 8 covers a limited set of registries. The page does not describe changes over time in histology mix, subsite, staging, imaging or endoscopy practice and does not apportion any change in this series to a cause. Read beside the mortality values. Compare only with other values from this same SEER 8 series, not with the SEER 21 value.

    Source

  • 5-year relative survival · United States, SEER 8 registries

    • 21.79%

      2018 (observed value; last year with an observed value in the table)

    All races, both sexes; SEER Cancer Stat Facts category 'esophageal cancer' (the page does not state the histologies included, does not report squamous cell carcinoma and adenocarcinoma separately, and does not say whether gastro-oesophageal junction tumours are counted). SEER 8 five-year relative survival series, 1975-2018. Age range is not stated for this series.

    Modeled trend value for 2018: 23.94%. The observed value is lower than the 2016 and 2017 values (24.74% and 24.09%); the table shows no observed value after 2018 (the modeled-trend column continues beyond 2018). Relative survival excludes deaths from other causes. 'Observed' is the table's label for the actual yearly value, as opposed to the Joinpoint modeled-trend value in the adjacent column. The column is headed 'Year' and the chart note says only 'SEER 8 5-Year Relative Survival Percent from 1975-2018'; the page does not say the year is the year of diagnosis. SEER 8 covers a limited set of registries. The page does not describe changes over time in histology mix, subsite, staging, imaging or endoscopy practice and does not apportion any change in this series to a cause. Read beside the mortality values. Compare only with other values from this same SEER 8 series, not with the SEER 21 value.

    Source

  • 5-year relative survival · United States, SEER 21 registries (excluding IL)

    • 22.2%

      2016-2022 (headline value on the page)

    All races, both sexes; oesophageal cancer, all stages together. By stage at diagnosis in the same table (SEER Combined Summary Stage): localized 48.6% (19% of cases), regional 29.1% (32%), distant 5.3% (39%), unknown stage 15.9% (10%). The page states no age range, histology or subsite for these figures.

    A different registry set (SEER 21 excluding Illinois) and period from the SEER 8 series above: keep it separate and do not subtract it from, or chain it with, the 1975 or 2018 values. Relative survival excludes deaths from other causes. The page states that 2016-2022 is the data period for this figure and does not say whether these are diagnosis years. The page text gives 19.1% of cases diagnosed at the local stage; the table rounds to 19%.

    Source

  • Age-standardised mortality · United States (U.S. Mortality data shown on the SEER page; not a SEER registry series)

    • 3.7per 100,000

      1975 (observed value; first year in the table)

    All races, both sexes; oesophageal cancer deaths; age-adjusted rate per 100,000 (the page text read does not name the standard population).

    The table prints 3.70 for 1975; modeled trend value 3.67. The series then rises to its highest observed value (4.44 in 2005) before falling to 3.57 in 2024. The page gives 3.7 per 100,000 for 2020-2024 deaths combined. Histologies are combined; the page does not report squamous cell carcinoma and adenocarcinoma separately.

    Source

  • Age-standardised mortality · United States (U.S. Mortality data shown on the SEER page; not a SEER registry series)

    • 4.44per 100,000

      2005 (observed value; highest value in the table)

    All races, both sexes; oesophageal cancer deaths; age-adjusted rate per 100,000 (the page text read does not name the standard population).

    Modeled trend value for 2005: 4.39. Same series and caveats as the 1975 value above. The page's SEER 8 incidence series is 3.74 (1975), 5.21 at its highest (2004) and 4.16 (2023) per 100,000.

    Source

  • Age-standardised mortality · United States (U.S. Mortality data shown on the SEER page; not a SEER registry series)

    • 3.57per 100,000

      2024 (observed value; last year in the table)

    All races, both sexes; oesophageal cancer deaths; age-adjusted rate per 100,000 (the page text read does not name the standard population).

    Modeled trend value for 2024: 3.59; 3.57 is the lowest observed value in the table. The page text reports age-adjusted death rates falling on average 1.1% each year over 2015-2024 and new cases falling on average 0.6% each year over 2014-2023. Same series and caveats as the 1975 value above. Mortality reflects incidence as well as survival, and the page attributes no part of the change to treatment.

    Source

Survival figures are affected by earlier detection (lead time) and overdiagnosis, and by changes in who is diagnosed. For screened cancers, read them alongside mortality.

Evidence chains

A lane follows one defined population and setting, where each new trial tests against what was then standard. Compare only within a lane, and read every comparison as illustrative.

Oesophageal squamous cell carcinoma, first-line

Adults with previously untreated, unresectable advanced, recurrent or metastatic oesophageal squamous cell carcinoma (KEYNOTE-590 also enrolled adenocarcinoma and Siewert type 1 junction cancer; its squamous-cell result is plotted) · Metastatic · endpoint: OS

Results are not comparable across trials.

Both trials added a PD-1 inhibitor to first-line chemotherapy; KEYNOTE-590 used pembrolizumab with fluorouracil and cisplatin in a double-blind design (plotted: squamous-cell carcinoma, first interim analysis), and CheckMate 648 was open-label (plotted: nivolumab plus chemotherapy in all randomised patients, who all had squamous-cell carcinoma). PD-L1 is scored differently and the cut-offs differ: KEYNOTE-590 used a combined positive score (CPS) of 10 or more and CheckMate 648 used tumour-cell PD-L1 expression of 1% or more; those subgroup results are not drawn. CheckMate 648's nivolumab-plus-ipilimumab arm is not drawn. Eras, regions and follow-up differ. Compare only each trial's own two arms; the chart is not a pooled or network estimate.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

05101520OS, median (months)20212021KEYNOTE-590CheckMate 648
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 2 trials in this lane. KEYNOTE-590 (2021): experimental 12.6 months vs control 9.8 months; CheckMate 648 (2021): experimental 13.2 months vs control 10.7 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
KEYNOTE-5902021OS9.8 months12.6 months
CheckMate 6482021OS10.7 months13.2 months

Resectable oesophageal adenocarcinoma, preoperative therapy

Adults with resectable oesophageal or oesophagogastric-junction adenocarcinoma (CROSS enrolled 75% adenocarcinoma and 23% squamous cell carcinoma; its adenocarcinoma subgroup is plotted) · Neoadjuvant · endpoint: OS

Results are not comparable across trials.

The comparisons differ in direction and measure. CROSS compared chemoradiotherapy followed by surgery with surgery alone (the plotted adenocarcinoma result is a subgroup analysis, reported as medians in a long-term update). Neo-AEGIS compared perioperative chemotherapy (control; mostly an older epirubicin-platinum regimen) with chemoradiotherapy followed by surgery (the experimental arm in the record), and non-inferiority of perioperative chemotherapy was not established. ESOPEC compared perioperative FLOT (experimental) with CROSS-type chemoradiotherapy (control) and gives only a 36-month survival rate, which is listed beside the chart, not on the same axis. Because the control and experimental roles are not the same across the three trials, read each pair separately. Squamous cell carcinoma, where CROSS reported a different effect size, is outside this lane. Eras and chemotherapy regimens differ. Compare only each trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

0102030405060OS, median (months)20122025CROSSNeo-AEGIS
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 2 trials in this lane. CROSS (2012): experimental 43.2 months vs control 27.1 months; Neo-AEGIS (2025): experimental 49.2 months vs control 48 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
CROSS2012OS27.1 months43.2 months
Neo-AEGIS2025OS48 months49.2 months
ESOPEC2025OS (not plotted: different measure or time point)50.7 %57.4 %

Today and tomorrow

Standard of care today

Adjuvant
Resected oesophageal or gastro-oesophageal junction cancer after neoadjuvant chemoradiotherapy, FDA letter read. FDA, 20 May 2021: nivolumab (BLA 125554/S-092) for adjuvant treatment of completely resected oesophageal or gastroesophageal junction cancer with residual pathologic disease in patients who have received neoadjuvant chemoradiotherapy (the FDA file for this supplement also contains a later general-advice letter, signed 28 May 2021, that corrects the submission date quoted in the approval letter and states that the approval date is unchanged). CheckMate 577 is covered in the milestone timeline. EU documents for this indication were not read.
U 2021
Neoadjuvant
Resectable locally advanced oesophageal and junctional cancer: no FDA or EMA action and no guideline text was read for neoadjuvant chemoradiotherapy, perioperative chemotherapy or surveillance strategies, so none is described as current practice here. The milestone timeline records CROSS (preoperative carboplatin and paclitaxel with radiotherapy plus surgery versus surgery alone; primary report 2012 and ten-year report 2021), ESOPEC (perioperative FLOT versus preoperative chemoradiotherapy in resectable oesophageal adenocarcinoma, NEJM 2025) and SANO (active surveillance versus standard surgery after a clinical complete response to neoadjuvant chemoradiotherapy, reported as non-inferior for 2-year overall survival, with the authors noting that extended follow-up is required). NEEDS, a randomised comparison of two strategies in oesophageal squamous cell carcinoma, is listed under horizon.
van Hagen P, Hulshof MC, van Lanschot JJ, Steyerberg EW, van Berge Henegouwen MI, Wijnhoven BP, et al 2012Eyck BM, van Lanschot JJB, Hulshof MCCM, van der Wilk BJ, Shapiro J, van Hagen P, et al 2021Hoeppner J, Brunner T, Schmoor C, Bronsert P, Kulemann B, Claus R, et al 2025van der Wilk BJ, Eyck BM, Wijnhoven BPL, Lagarde SM, Rosman C, Noordman BJ, et al 2025
Metastatic

Advanced or metastatic oesophageal and junctional cancer, FDA letters read. Squamous cell carcinoma after prior therapy:

  • pembrolizumab, 30 July 2019 (BLA 125514/S-55 and S-56), for recurrent locally advanced or metastatic disease with PD-L1 CPS 10 or more after one or more prior lines of systemic therapy
  • nivolumab, 10 June 2020 (BLA 125554/S-081), for unresectable advanced, recurrent or metastatic disease after prior fluoropyrimidine- and platinum-based chemotherapy
  • tislelizumab-jsgr as a single agent (BLA 761232, approval letter e-signed 13 and 14 March 2024) for unresectable or metastatic disease after prior systemic chemotherapy that did not include a PD-(L)1 inhibitor. First line: pembrolizumab, 22 March 2021 (S-096), with platinum- and fluoropyrimidine-based chemotherapy for locally advanced or metastatic oesophageal or gastroesophageal junction carcinoma (tumours with epicentre 1 to 5 cm above the junction) not amenable to surgical resection or definitive chemoradiation
  • nivolumab, 27 May 2022, for unresectable advanced or metastatic squamous cell carcinoma with fluoropyrimidine- and platinum-based chemotherapy (S-106) and with ipilimumab (S-105)
  • tislelizumab-jsgr, 3 March 2025 (BLA 761380), with platinum-containing chemotherapy for unresectable or metastatic squamous cell carcinoma whose tumours express PD-L1 (1 or more). Labelling revisions: pembrolizumab, 22 May 2025 (oesophageal or junctional carcinoma with chemotherapy for PD-L1 CPS 1 or more
  • single agent after one or more prior lines for squamous histology with CPS 10 or more) and nivolumab, 23 May 2025 (the first-line squamous cell carcinoma indications for PD-L1 1 or more). KEYNOTE-590 and CheckMate 648 are in the milestone timeline. EU documents for oesophageal indications were not read
FDA 2019FDA 2020FDA 2024U 2021U 2022U 2022FDA 2025U 2025U 2025
Any setting

Guideline documents are cited by identifier only; their text was not read, so no recommendation is attributed to them, and whether they are the current editions was not checked beyond a PubMed search on 21 October 2026:

  • the ESMO Clinical Practice Guideline on gastric and oesophageal cancer (Ann Oncol, published online 15 October 2026, PMID 42880714)
  • the ASCO guideline update on immunotherapy and targeted therapy for advanced gastroesophageal cancer (J Clin Oncol 2026;44(12):1145-1165, PMID 41747202) and its living guideline version 2026.1.2 (J Clin Oncol, PMID 42607281)
  • and the NCCN Guidelines Insights on oesophageal and oesophagogastric junction cancers, version 4.2026 (J Natl Compr Canc Netw 2026;24(9), PMID 42753809)
Guideline: Al-Batran SE, Bencivenga M, Crosby T, Dinis-Ribeiro M, Huguet F, Lordick F, et al 2026Guideline: Shah MA, Kennedy EB, Deighton D, Jhawer M, Kadakia KC, Mukherjee S, et al 2026Guideline: Shah MA, Marzalik JS, Kennedy EB, Deighton D, Hamedi Z, Jhawer M, et al 2026Guideline: Ajani JA, D'Amico TA, Bentrem DJ, Chakrabarti S, Corvera CU, Das P, et al 2026

Trials to watch

  • KUNLUN NCT04550260

    In locally advanced, unresectable oesophageal squamous cell carcinoma, does durvalumab given with definitive chemoradiotherapy differ from placebo with definitive chemoradiotherapy in progression-free survival? Registry record: randomised, double-blind, placebo-controlled phase 3 (AstraZeneca); status active, not recruiting; enrolment 640 (actual); the primary outcome listed is progression-free survival per RECIST 1.1 by blinded independent central review; estimated primary completion June 2027; no results posted.

  • KEYNOTE-975 NCT04210115

    In oesophageal carcinoma (conditions listed: squamous cell carcinoma, gastroesophageal junction carcinoma and oesophageal adenocarcinoma) treated with concurrent definitive chemoradiotherapy, does pembrolizumab differ from placebo? Registry record: randomised, double-blind, placebo-controlled phase 3 (Merck Sharp & Dohme); status completed, with primary completion January 2026 and study completion June 2026 (both actual); enrolment 703 (actual); the primary outcomes listed are event-free survival and overall survival; no results posted. A PubMed search on 21 October 2026 found a study-design paper but no primary results paper.

  • NEEDS NCT04460352

    In oesophageal squamous cell carcinoma, does definitive chemoradiotherapy followed by surveillance, with salvage surgery only when needed, differ from neoadjuvant chemoradiotherapy followed by planned surgery? Registry record: randomised, open-label phase 3 (Karolinska University Hospital); status recruiting (status last verified March 2025; the record has not been updated since 6 March 2025, so the status may be out of date); enrolment 1,020 (estimated; the 2022 protocol paper states a target of 1,200 randomised patients); the primary outcomes listed are overall survival, global health-related quality of life and eating restrictions; estimated primary completion December 2026; no results posted.

References

Show all 42 references

Loading references… Open the list as a page

Every figure was checked against its primary source to the extent possible. Items that could not be fully confirmed are marked “needs review”. Educational history, not medical advice.

Report an error

Links open the primary source. A flag marked ⚑ needs review means a value or claim has not yet been independently confirmed.

2012TreatmentChemotherapy

landmarkNeoadjuvant

CROSS: neoadjuvant chemoradiotherapy for oesophageal cancer

In a Dutch trial, people with cancer of the oesophagus or the junction with the stomach had either surgery alone or five weeks of chemotherapy with radiotherapy followed by surgery. Median survival was about 49 months with chemoradiotherapy and 24 months with surgery alone, and the survival gain was still present ten years later.

CROSS (366 analysed; 75% adenocarcinoma, 23% squamous): weekly carboplatin and paclitaxel with 41.4 Gy, then surgery, versus surgery alone. R0 resection 92% vs 69%; median OS 49.4 vs 24.0 months (HR 0.657; 95% CI 0.495 to 0.871; P=0.003). Long-term: median OS 48.6 vs 24.0 months (HR 0.68); squamous HR 0.48, adenocarcinoma HR 0.73. Ten-year OS 38% vs 25% for chemoradiotherapy vs surgery alone (HR 0.70; 95% CI 0.55 to 0.89).

The role of neoadjuvant chemoradiotherapy in cancer of the oesophagus or the oesophagogastric junction was described, at the time of the trial, as not well established. [1] The Dutch CROSS investigators compared surgery alone with five weeks of weekly carboplatin and paclitaxel given alongside radiotherapy (41.4 Gy in 23 fractions) before surgery. [1]

Between March 2004 and December 2008, 368 patients were enrolled and 366 analysed; 75% had adenocarcinoma and 23% squamous-cell carcinoma. [1] In the first report, a complete resection with no tumour within 1 mm of the margins was achieved in 92% of patients after chemoradiotherapy and 69% after surgery alone, and median overall survival was 49.4 months versus 24.0 months (hazard ratio 0.657). [1]

Long-term follow-up reported median overall survival of 48.6 versus 24.0 months, with hazard ratios of 0.48 in squamous-cell carcinoma and 0.73 in adenocarcinoma. [2] At a median follow-up of 147 months, ten-year overall survival was 38% versus 25%, and the effect on survival did not change over time. [3]

The authors of the long-term report concluded that this regimen followed by resection should be regarded as a standard of care for resectable locally advanced disease. [2]

Standard before

The comparator in CROSS was surgery alone. The NEJM abstract describes the role of neoadjuvant chemoradiotherapy for oesophageal or junctional cancer as not well established.

What changed

The authors of the long-term report conclude that neoadjuvant chemoradiotherapy according to CROSS, followed by surgical resection, should be regarded as a standard of care for resectable locally advanced oesophageal or oesophagogastric-junctional cancer; the ten-year report describes it as having become a standard of care. Guideline text was not read.

Why it mattered

CROSS is a randomised comparison of preoperative chemoradiotherapy plus surgery with surgery alone, in both squamous-cell carcinoma and adenocarcinoma. Overall survival was longer with chemoradiotherapy at the first report, at a minimum of five years and at ten years. The same regimen is the chemoradiotherapy arm of the later ESOPEC and Neo-AEGIS trials in adenocarcinoma.

Caveats

  • Histology: 75% of the analysed patients had adenocarcinoma and 23% squamous-cell carcinoma. The histology-specific hazard ratios (squamous 0.48; adenocarcinoma 0.73) come from subgroup analyses in the long-term report.
  • CROSS compared chemoradiotherapy with surgery alone, not with perioperative chemotherapy. The head-to-head question in adenocarcinoma is addressed by the separate ESOPEC and Neo-AEGIS trials.
  • The median for the chemoradiotherapy arm differs between reports (49.4 months at the first report, 48.6 months at the long-term analysis) because they come from different follow-up analyses.
  • Enrolment ran from March 2004 to December 2008. The long-term report states that no adverse-event data were collected beyond the initial report.