Loading
Lung cancer is where modern cancer epidemiology began: two studies in 1950 tied it to cigarettes, and a fifty-year cohort proved it. This roadmap follows the evidence from there through chemotherapy that barely worked, the scans that find the disease early, the drugs for cancers that depend on one broken gene, immunotherapy, and small-cell disease, to 2032.
Lung cancer kills more people than any other cancer, and it is the cancer whose cause is best understood. Wynder and Graham in the United States and Doll and Hill in Britain published within months of each other in 1950; the British Doctors Study then followed a cohort for fifty years and turned an association into a dose-response relationship with a cessation curve attached. Peto's 2000 analysis gave the number that tobacco control still runs on: stopping before middle age avoids more than 90 percent of the risk.
Treatment took much longer. The 1995 individual-patient meta-analysis of 52 trials showed cisplatin-based chemotherapy helped, by about 5 percent at five years after surgery; E1594 in 2002 showed that no combination of cytotoxic drugs beat any other, with a median survival of 7.9 months. The plateau broke twice. In 2004 Lynch and Paez independently found that the minority who responded dramatically to gefitinib had EGFR mutations, and IPASS in 2009 proved that selecting by that mutation was what made the drug work; in 2007 Soda found the EML4-ALK fusion, and crizotinib was approved four years later. In 2012 Topalian's phase 1 showed durable responses to PD-1 blockade in lung cancer, and by 2018 chemotherapy plus pembrolizumab was standard for most patients without a driver.
Early detection followed its own track. PLCO showed in 2011 that chest radiography does nothing; NLST that year and NELSON in 2020 showed that low-dose computed tomography cuts lung cancer deaths. Eligibility is the unsolved part: pack-year thresholds exclude high-risk groups systematically, never-smokers are outside the criteria entirely, and uptake among those who qualify remains a fraction of them.
Since 2017 the direction has been earlier and combined. PACIFIC put consolidation durvalumab after chemoradiotherapy in stage III disease and five-year survival reached 42.9 percent; CheckMate 816, KEYNOTE-671, AEGEAN and IMpower010 moved immunotherapy around the operation; ADAURA and ALINA did the same with targeted drugs for EGFR and ALK. Small-cell lung cancer, static since twice-daily radiotherapy in 1999, finally moved: ADRIATIC lifted median survival in limited-stage disease from 33.4 to 55.9 months, and tarlatamab, a DLL3 T-cell engager, became the first drug built for the disease. What has not moved is resistance, which arrives in one to three years for every targeted drug, and the gap between who is eligible for screening and who gets it.
UK and NHS specifics (Targeted Lung Health Check coverage and uptake, NICE positions and Cancer Drugs Fund status, molecular testing turnaround, thoracic surgery and radiotherapy capacity, audit indicators and trial access) are on the UK and NHS page for lung cancer and are not restated here.
Wynder and Graham's 684 proven cases appeared in JAMA in May 1950 and Doll and Hill's British case-control study in the BMJ in September. Both were attacked as artefacts of how cases were chosen, so Doll and Hill built a prospective cohort of British doctors in 1951 and followed it for fifty years. The method they used, exposure recorded before outcome with a measured dose-response, is the foundation of modern cancer epidemiology; the 1964 United States Surgeon General's report and everything in tobacco control since rests on it.
Half of the evidence base on which every tobacco control policy in the world rests. Lung cancer was a rare disease at the start of the twentieth century and the commonest cancer killer by its end; this paper and Doll and Hill's are where the cause was named.
Doll and Hill's 1950 study is where the evidence that smoking causes cancer begins. Everything from cigarette warnings and tax to smoke-free laws and lung screening eligibility descends from this study and the cohort that followed it.
The methodological ancestor of modern cancer epidemiology. Cohort design, exposure recorded before outcome, and a dose-response relationship measured rather than asserted: this is the template every later study of a cancer risk factor follows.
Smoking is the single largest preventable cause of cancer death, and quitting at any age helps, with the greatest gain from quitting young. Cessation support belongs in every cancer service, including lung screening programmes.
Therapeutic nihilism was the default until the Non-small Cell Lung Cancer Collaborative Group pooled individual data on 9,387 patients from 52 trials in 1995 and found a 27 percent reduction in the risk of death when chemotherapy was added to supportive care. E1594 then randomised 1,207 patients between four platinum doublets and found no difference: a 19 percent response rate and a median survival of 7.9 months, whichever drugs were used. LACE later pooled 4,584 resected patients and fixed the adjuvant rule, a 5.4 percent absolute survival gain concentrated in stage II and III disease. The first crack in the ceiling was biological rather than cytotoxic: ECOG 4599 added bevacizumab in 2006 and took median survival past twelve months for the first time, at 12.3 against 10.3 months, in the non-squamous patients without brain metastases who were allowed into the trial.
Gefitinib worked spectacularly in about one patient in ten and was close to being abandoned. In May 2004 Lynch in the NEJM and Paez in Science independently reported that the responders had activating EGFR mutations; Paez also explained the geography, finding the mutation in 15 of 58 Japanese tumours and 1 of 61 American ones. IPASS in 2009 randomised 1,217 clinically selected patients and showed the benefit lived entirely in the mutation-positive subgroup. In 2007 Soda found the EML4-ALK fusion in 5 of 75 tumours; Kwak screened 1,500 patients to enrol 82 and got a 57 percent response rate with crizotinib, a drug built for a different target. Resistance came with it: Kobayashi's single re-biopsied patient in 2005 carried T790M, and Sequist's 37 patients in 2011 showed resistance could be genetic, histological or even reversible.
PLCO randomised 154,901 people to four annual chest radiographs or usual care and after 13 years found 1,213 lung cancer deaths against 1,230: no effect at all. NLST the same year and NELSON in 2020 showed that low-dose computed tomography does reduce lung cancer mortality, and the United States task force widened eligibility in 2021 from age 55 and 30 pack-years to age 50 and 20. Aldrich had already shown what a pack-year threshold does: in a southern United States cohort, 31 percent of white smokers qualified against 17 percent of Black smokers, who develop the disease at lower cumulative exposure.
Topalian's 2012 phase 1 found durable responses to PD-1 blockade in 18 percent of lung cancer patients and the first hint that PD-L1 expression predicted them. CheckMate 017 and 057 and KEYNOTE-010 established second-line nivolumab and pembrolizumab; KEYNOTE-024 moved pembrolizumab in front of chemotherapy for PD-L1-high tumours; KEYNOTE-189 and KEYNOTE-407 combined it with chemotherapy for everybody else. IMpower110 reproduced the PD-L1-high result with atezolizumab and a different assay, which is where the assay problem became unavoidable: SP142, 22C3 and SP263 do not select the same patients.
PACIFIC put a year of durvalumab after chemoradiotherapy in unresectable stage III disease; the five-year update found 42.9 percent alive against 33.4 percent and a third still free of progression. Resectable disease followed: CheckMate 816 gave nivolumab before surgery, KEYNOTE-671, AEGEAN and CheckMate 77T on both sides, IMpower010 afterwards, and NADIM II showed 93 percent of patients reaching the operating theatre against 69 percent. Targeted drugs made the same move: ADAURA for EGFR, ALINA for ALK, and LAURA for EGFR-mutant stage III disease, where a hazard ratio of 0.16 replaced durvalumab consolidation that had never worked well in that genotype.
ALK moved from crizotinib to alectinib (ALEX) to lorlatinib (CROWN, 78 percent progression-free at 12 months against 39). MET exon 14 got capmatinib, RET got selpercatinib, ROS1 got crizotinib then entrectinib, BRAF V600E got dabrafenib and trametinib, HER2 got trastuzumab deruxtecan and zongertinib, and KRAS G12C, undruggable for thirty years, got sotorasib and adagrasib, with CodeBreaK 200 showing a 1.1-month gain over docetaxel. In EGFR-mutant disease, FLAURA2 added chemotherapy to osimertinib and MARIPOSA beat it outright with amivantamab and lazertinib, 23.7 months against 16.6, at the cost of ten times the treatment discontinuation.
Turrisi's twice-daily thoracic radiotherapy in 1999 raised five-year survival from 16 to 26 percent, and the Aupérin overview showed that irradiating a brain with no detectable disease in it extends life. Then nothing, for a quarter of a century. George's 110 genomes explained why: the disease is defined by losing both copies of TP53 and RB1, and a loss cannot be inhibited. Rudin's four transcription-factor subtypes gave the field something to design around, the NOTCH finding pointed at DLL3, and in 2023 tarlatamab produced a 40 percent response rate in twice-treated patients. ADRIATIC then lifted median survival in limited-stage disease from 33.4 to 55.9 months.
The standard of care in limited-stage small-cell lung cancer from 1999 until ADRIATIC added immunotherapy in 2024, and a rare example of a curative gain in a disease that has had almost none.
Treating the brain before the cancer gets there works in small-cell lung cancer, and it is the proof of principle behind every later attempt to prevent rather than treat brain metastases.
Why small-cell lung cancer has no targeted therapy in the conventional sense: it is built from the loss of TP53 and RB1, and the drugs that transformed non-small-cell disease inhibit gains rather than restore losses. The NOTCH result pointed at DLL3 and, eventually, at tarlatamab.
The organising framework for every small-cell lung cancer trial designed since. It is also why the slow progress in the disease is now attributed to treating four diseases as one rather than to the biology being intractable.
This is the paper Europe PMC returns for registry id NCT02763579 with the most citations, so it is the natural first reading for anyone following the IMpower133 trial. Read the abstract above alongside the trial page and the registry entry; the record was linked automatically and its figures have not been checked by hand.
Immunotherapy is now part of first-line treatment for extensive-stage small-cell lung cancer everywhere, on the strength of a gain measured in weeks. The size of that gain is the reason small-cell lung cancer remains the clearest unmet need in thoracic oncology.
Small-cell lung cancer got its first targeted drug, and the mechanism is a T-cell engager rather than a kinase inhibitor. It also brought cytokine release syndrome, and the inpatient monitoring that goes with it, into thoracic oncology for the first time.
The biggest single advance in small-cell lung cancer since twice-daily radiotherapy in 1999, and the strongest argument that the disease responds to immunotherapy when it is given at a lower tumour burden.
TRACERx sequenced 327 regions from 100 resected tumours and found that drivers are almost always clonal while later alterations are heterogeneous in more than three quarters of tumours, with copy-number heterogeneity carrying a hazard ratio of 4.9 for recurrence or death. Abbosh built a per-patient phylogenetic ctDNA assay from the same tumours and identified the patients who would relapse before any scan did. Neither is yet a clinical test in lung cancer: no randomised trial has shown that acting on a positive minimal residual disease result improves survival, which is the gap the ctDNA-guided adjuvant trials are designed to close.
About one lung cancer in five worldwide occurs in someone who never smoked, and the proportion is rising as smoking falls. Sherlock-Lung's 232 whole genomes found three subtypes with no tobacco signature, the dominant one slow-growing with drivers datable to years before diagnosis. Hill and Swanton then proposed a mechanism for at least part of it: fine particulate matter does not create the EGFR mutation, which is already present in 18 percent of histologically normal lungs, but inflames the tissue enough to let it grow. That makes air quality a cancer intervention with a measurable endpoint, and makes the screening eligibility criteria, which exclude never-smokers entirely, the wrong instrument for a growing share of the disease.
Four things, none of them a new drug class. Screening reaching the people at highest risk rather than the people easiest to enrol, which means risk-model eligibility and delivery into deprived areas rather than wider pack-year thresholds. Resistance treated as a diagnosis, with a re-biopsy or a ctDNA profile choosing the next line rather than a default. Brain metastasis prevention measured as a primary endpoint rather than counted as a secondary one. And small-cell lung cancer trialled as the four diseases Rudin's subtypes describe, in a platform with shared controls, rather than as one disease with a 62 percent grade 3 toxicity rate and a two-month gain.
Every era's records, trial outcomes and papers, and every watch item, as JSON.
Probability ranges are named estimates that the claim is borne out on roughly a five-year horizon. They are meant to be argued with: propose a revision with your name and reasoning via a pull request to src/data/confidence.ts.
Readouts, decisions and registry completion dates ahead. Each date is quoted from its source, not inferred; a missing date means no source states one.
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
One of the most cited trial reports Europe PMC returns for Tarlatamab in Small-cell lung cancer, so it is a natural first reading for anyone weighing the idea it is linked from. The record was linked automatically by title and abstract; read the abstract above and the paper itself before relying on any figure.
This is the paper Europe PMC returns for registry id NCT03425643 with the most citations, so it is the natural first reading for anyone following the KEYNOTE-671 trial. Read the abstract above alongside the trial page and the registry entry; the record was linked automatically and its figures have not been checked by hand.
Patients with operable stage II-III lung cancer without EGFR or ALK alterations should have chemo-immunotherapy discussed before surgery rather than only afterwards. Three pre-operative cycles do not compromise the operation and improve cure rates. Whether to continue immunotherapy after surgery, as the perioperative trials do, and whether patients with pCR need any further treatment, remain open questions.
One of the most cited trial reports Europe PMC returns for KRAS in Non-small-cell lung cancer, so it is a natural first reading for anyone weighing the idea it is linked from. The record was linked automatically by title and abstract; read the abstract above and the paper itself before relying on any figure.
Every resected non-squamous lung cancer should be tested for EGFR mutations, because patients who carry one live longer if they take osimertinib for three years after surgery. The trial does not tell us whether adjuvant chemotherapy can be omitted, nor what happens on relapse after osimertinib, and the three-year duration was chosen empirically.
Lung screening works when it uses volumetric nodule management, and it works against a no-screening control. The protocol underpins the UK Targeted Lung Health Check programme and European recommendations. Benefit in women remains less precisely estimated.
One of the most cited trial reports Europe PMC returns for DLL3 in Small-cell lung cancer, so it is a natural first reading for anyone weighing the idea it is linked from. The record was linked automatically by title and abstract; read the abstract above and the paper itself before relying on any figure.
This is the paper Europe PMC returns for registry id NCT02763579 with the most citations, so it is the natural first reading for anyone following the IMpower133 trial. Read the abstract above alongside the trial page and the registry entry; the record was linked automatically and its figures have not been checked by hand.
Shares Genomic and evolutionary classification of lung cancer in never smokers, Lung adenocarcinoma promotion by air pollutants, Capmatinib in MET exon 14-mutated or MET-amplified non-small-cell lung cancer, Comprehensive genomic profiles of small cell lung cancer.
Shares Close the United Kingdom lung cancer gaps the UK page names: screening rollout, diagnostic pathway, molecular testing and drug access, Capmatinib in MET exon 14-mutated or MET-amplified non-small-cell lung cancer, Phylogenetic ctDNA analysis depicts early-stage lung cancer evolution, Make resistance a diagnosis: sequence at every progression and choose the next line from what the tumour became.
Shares Pembrolizumab plus Chemotherapy for Squamous Non-Small-Cell Lung Cancer, Beamion LUNG-1: zongertinib in previously treated HER2-mutant non-small-cell lung cancer, First-Line Atezolizumab plus Chemotherapy in Extensive-Stage Small-Cell Lung Cancer, LIBRETTO-001: selpercatinib in RET fusion-positive non-small-cell lung cancer.
Shares Genomic and evolutionary classification of lung cancer in never smokers, Lung adenocarcinoma promotion by air pollutants, Paclitaxel-carboplatin alone or with bevacizumab for non-small-cell lung cancer, Treat lung cancer in never-smokers as its own disease, with its own detection programme.
Shares EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy, Osimertinib with or without chemotherapy in EGFR-mutated advanced NSCLC, ADAURA: three years of osimertinib after surgery for EGFR-mutated lung cancer, Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma.
Shares Take lung screening scanners to supermarket car parks in the poorest areas, Treat the deprivation gradient in lung cancer as a defect in delivery that can be fixed and measured, Close the United Kingdom lung cancer gaps the UK page names: screening rollout, diagnostic pathway, molecular testing and drug access, Screening for Lung Cancer: US Preventive Services Task Force Recommendation Statement.
Shares Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non-small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial, Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR, FLAURA: osimertinib as first treatment for EGFR-mutated lung cancer, Tumour heterogeneity and clonal evolution.
Shares Prophylactic cranial irradiation for patients with small-cell lung cancer in complete remission, Durvalumab after chemoradiotherapy in limited-stage small-cell lung cancer, Measure brain metastasis prevention as a primary endpoint, not as a secondary one, Durvalumab plus platinum-etoposide versus platinum-etoposide in first-line treatment of extensive-stage small-cell lung cancer (CASPIAN).