🇫🇷 Cancer in France

The Panorama 2026 gives 433,136 new cases in 2023, 57 per cent in men, at world age-standardised rates of 354.9 per 100,000 in men and 274 in women; the median age at diagnosis was 70 in men and 68 in women. In women the leading cancers were breast (61,214 cases, 33 per cent), colon and rectum (21,370, 11 per cent), lung (19,339, 10 per cent) and pancreas (7,668, 4 per cent). In men they were prostate (59,885, 24 per cent, a 2018 figure), lung (33,438, 14 per cent), colon and rectum (26,212, 11 per cent) and pancreas (8,323, 3 per cent). Between 2019 and 2023 the age-standardised incidence of lung cancer in women rose by 3.6 per cent a year and of pancreatic cancer by 1.6 per cent a year, while the all-cancer rate in men was stable (plus 0.06 per cent a year). The estimate paper behind these figures finds that the annual number of new cancers has roughly doubled since 1990 in both sexes, through rising rates in women and the growth and ageing of the population.

The Panorama, quoting the Inserm cause-of-death centre CépiDc, gives 164,314 cancer deaths in 2023, 91,380 in men and 72,934 in women, at world age-standardised mortality rates of 127.9 and 80.2 per 100,000, falling by 1.61 and 0.76 per cent a year between 2019 and 2023. Among men the leading causes were lung (20,468 deaths), prostate (9,002), colon and rectum (8,923), pancreas (6,645) and liver (6,272); among women breast (12,765), lung (10,699), colon and rectum (8,007), pancreas (6,730) and ovary (3,420). The Panorama says the overall decline reflects changes in which cancers occur, earlier diagnosis and treatment advances, with pancreas, and lung in women, the exceptions.

The Panorama gives age-standardised five-year net survival for people diagnosed 2010 to 2015 of 93 per cent for prostate cancer and for skin melanoma, 88 per cent for breast, 63 per cent for colon and rectum and for cervix, 43 per cent for ovary, 26 per cent for the central nervous system, 20 per cent for lung (24 in women, 18 in men), 18 per cent for liver, 17 per cent for oesophagus and 11 per cent for pancreas. INCa classes cancers with five-year net survival above 65 per cent as having a favourable prognosis and between 33 and 65 per cent as intermediate. The underlying survival study covers metropolitan registries for 1989 to 2018 and was published by INCa in July 2021.

The Registre national des cancers de l'enfant recorded 18,468 cancers in children aged 0 to 14 in France between 2014 and 2023, an average of 1,847 a year: leukaemias 517 (28 per cent), central nervous system tumours 487 (26.4 per cent), lymphomas 195 (10.6 per cent), sympathetic nervous system tumours 138 (7.5 per cent) and soft-tissue sarcomas 112 (6.1 per cent). Over the same years 4,497 cancers were registered in adolescents aged 15 to 17, about 450 a year, led by lymphomas (135 a year). For children diagnosed between 2000 and 2016, 92 per cent were alive one year and 83 per cent five years after diagnosis; for adolescents, 94 and 82 per cent. A first national study of adolescents and young adults aged 15 to 39 counted 54,735 diagnoses between 2000 and 2020 in 19 départements, with incidence rising by 1.62 per cent a year to 2014 and falling by 0.79 per cent a year from 2015 to 2020.

Using a 2018 study of cancers in metropolitan France in 2015, which puts the share attributable to the risk factors it studied at 41 per cent, the Panorama attributes 20 per cent of cancers to tobacco (68,000 cases), 8 per cent to alcohol (28,000), 5.4 per cent each to unbalanced diet and to excess weight (18,600), 4 per cent to certain infections, 3 per cent each to occupational exposures and to ultraviolet radiation, 1.8 per cent to ionising radiation and smaller shares to physical inactivity, hormone treatments, not breastfeeding and fine particles. Daily smoking among 18 to 75 year olds was 18 per cent in 2024, and the number of daily smokers fell by 28 per cent between 2021 and 2024. In 2021, 22 per cent of adults exceeded the national low-risk drinking guidelines on at least one measure. The Panorama adds that 21 per cent of colorectal cancers in people over 30 were directly linked to alcohol in 2015.

The Panorama says incidence data come from local registries designated by INCa: 28 in all, covering about 20 to 24 per cent of the population depending on the cancer, made up of 14 general registries, 10 specialised registries (by organ or entity) and 4 registries in the overseas departments of Guadeloupe, Martinique, French Guiana and La Réunion, where coverage of the population is complete. National figures are produced by statistical modelling of these registries' data. The estimates are made by a partnership of the Francim registry network, Santé publique France, INCa and the biostatistics service of the Hospices Civils de Lyon, which publishes a detailed study of incidence and mortality every five years; the 2023 estimate was projected from data observed from 1985 to 2018, and the prostate and thyroid projections were not published because their recent trends had changed. Two national registries count childhood cancers across the whole territory, overseas included.

The Panorama calls 2025 a turning point. The registre national des cancers (RNC) is to collect all cancer cases nationally, using nine main sources named in the implementing decree, at least eight of them annually, including the local registries, the national health data system (SNDS) and other national databases such as causes of death, emergency care and disability, and screening, clinical and biological data specific to cancer. INCa is the sole data controller; data are pseudonymised, and researchers, public or private, will reach linked datasets after a favourable opinion from the registry's scientific and ethics committee and from CESREES, and CNIL authorisation where required. It builds on INCa's 2011 'cohorte cancer' extraction from SNDS and on its cancer data platform, which from 2019 to 2023 added the data of 27 local registries. The first national counts from it had not been published when this page was written.

The French national estimate counts invasive cancers only and excludes skin cancers other than melanoma, as its methods section states. GLOBOCAN's all-cancers figure for France, which the table on this page reproduces from the 2022 release, is its code 39, all cancers C00-C97, which includes them. Both describe metropolitan France only: GLOBOCAN labels its entry 'France (metropolitan)', and the French estimate is for metropolitan France. They also refer to different years and are built by different models, and both are projections rather than counts. In the same 2022 file, GLOBOCAN's own count of non-melanoma skin cancer in France (its code 17) is 48,454, so its all-cancers figure without them is 435,114. IARC's fact sheet for France now carries the GLOBOCAN 2024 release instead, with 421,655 new cases in 2024 including non-melanoma skin cancer and 418,158 without it; the two releases differ in model as well as year. Compare France with other countries on GLOBOCAN's figures from a single release, and read French trends from the French series.

Breast: a mammogram every two years from 50 to 74; participation fell from a peak of 52.7 per cent in 2010-2011 to 46.3 per cent in 2023-2024, to which the Panorama adds 11 per cent of the target population screened individually; the programme found 41,759 breast cancers in 2021-2022. Tomosynthesis is not authorised within the organised programme. Bowel: a faecal immunochemical test every two years from 50 to 74, handed out by pharmacists and orderable online since 2022; participation was 30.7 per cent in 2024-2025 (31.8 in women, 29.6 in men), up from 29.6 per cent, and INCa is trialling posting the kit with the invitation. Cervix: a test every three years from 25 to 29 after two normal tests a year apart, then every five years to 65; standardised participation was 59.5 per cent in 2020-2022, with only women not screened at the recommended interval sent an invitation, and about 32,000 precancerous lesions found a year.

After a call for projects in January 2025, INCa chose IMPULSION, coordinated by Professor Marie-Pierre Revel of the Paris public hospitals and Professor Sébastien Couraud of the Hospices Civils de Lyon, with 6 million euros. It aims to recruit 20,000 volunteers over 18 to 24 months in Île-de-France, Hauts-de-France, Pays de la Loire, Provence-Alpes-Côte d'Azur and Auvergne-Rhône-Alpes: smokers and people who stopped less than 15 years ago, aged 50 to 74, with at least 20 pack-years. Participants have two scans a year apart and then one every two years; the national health insurance covers the scans in full. INCa's 2026 to 2030 roadmap sets the goal of generalising organised lung screening by 2030.

Service-public.fr lists 'tumeur maligne (cancer), affection maligne du tissu lymphatique ou hématopoïétique' among the ALD 30 conditions, which are exempting: care related to the condition is reimbursed up to the ceiling with the ticket modérateur (patient co-payment) waived, once the care protocol opens the right to 100 per cent. A split prescription form separates care for the condition, paid at 100 per cent, from unrelated care, paid at the usual rates. What remains payable: excess fees above the agreed tariff (dépassements d'honoraires), the 2 euro flat contribution, the medical deductible (franchise médicale), the daily hospital charge (forfait hospitalier) and medicines rated of low clinical benefit (SMR faible). A complementary insurance can cover the hospital charge and excess fees. Transport linked to the condition is covered on prescription for patients who meet the criteria.

The Panorama, quoting the national health insurance fund (CNAM), gives 29.3 billion euros as the weight of cancer care in the national health insurance spending objective (ONDAM) for 2024, on a much wider scope than hospital data. For anticancer medicines in 2024 it gives 5.6 billion euros on the 'liste en sus', the list of costly innovative hospital medicines paid on top of the hospital tariff (82 per cent of it on immunotherapies, up 14.5 per cent on 2023), 4.6 billion dispensed in community pharmacies (66 per cent on targeted therapies, up 11.1 per cent) and 241 million dispensed by hospital pharmacies to outpatients (rétrocession, down 40.3 per cent). The same page prints a figure of 7.4 billion euros for 2024, up 6.9 per cent, whose label in the brochure does not match the breakdown beside it, so it is not used here. Hospital spending on cancer rose by nearly 7 per cent in 2024. About 1.4 million people were treated specifically for cancer in acute hospital care in 2024; 437,735 had cancer surgery, 403,400 systemic drug treatment, 137,470 radiotherapy in the public sector and 128,000 in the private sector. In 2023, 90,466 patients received checkpoint inhibitors and 1,173 CAR-T cells.

The HAS page says the Commission de la transparence (CT) gives an opinion on the service médical rendu (SMR), graded important, moderate, low or insufficient; an insufficient SMR means the committee recommends against reimbursement. The SMR rests on the severity of the disease, the medicine's efficacy and side effects and its place in treatment. The amélioration du service médical rendu (ASMR) is graded from I to V, V meaning no progress over what is available, and is one of the criteria used to set the price. Opinions are given per medicine and per indication, so one indication can be reimbursed and another not. The committee is made up of doctors, pharmacists and methodologists and includes members drawn from patient and user associations; associations can submit written contributions on each medicine, which are published with the opinion. Prices are then negotiated by the Comité économique des produits de santé (CEPS).

Under article R161-71-3 of the social security code and a HAS college decision of 23 June 2022, an economic evaluation by the Commission d'évaluation économique et de santé publique (CEESP) is required at listing when the company claims ASMR I, II or III and a significant impact on health insurance spending, defined as any advanced therapy medicinal product, or forecast second-year sales in the indication of 20 million euros or more, or a claimed impact on how care is organised. A budget impact analysis is also required under the 2021 framework agreement between CEPS and the industry body LEEM when forecast second-year sales exceed 50 million euros. The CEESP estimates an incremental cost-effectiveness ratio (ratio différentiel coût-résultat, RDCR) and sends its opinion to CEPS as an aid to negotiating the price. The HAS pages on economic evaluation describe the ratio and its uncertainty but state no threshold value.

The EMA states that the centralised procedure is compulsory for human medicines containing a new active substance to treat cancer, as well as for biotechnology products and advanced therapies, so the marketing authorisation for a new cancer drug is a European Commission decision valid across the EU. The ANSM describes itself as the public body that gives access to health products in France and keeps them safe throughout their life cycle, with about 1,000 staff in Saint-Denis, Lyon and Montpellier-Vendargues; it authorises clinical trials, runs pharmacovigilance and contributes scientific advice through the EMA's working party. After European authorisation, a medicine still needs a HAS opinion and a price negotiated with the CEPS before it is reimbursed in France. How long those steps take is not given on this page (see the gaps); the early-access route below exists to pay for promising medicines in the meantime.

The 2021 social security financing law replaced the earlier temporary authorisations and temporary recommendations for use (ATU and RTU) with two regimes: accès précoce, decided by the HAS, and accès compassionnel, run by the ANSM. Early access requires four conditions: a serious, rare or disabling disease; no appropriate treatment; treatment that cannot be deferred; and a medicine presumed innovative. It covers medicines authorised but not yet reimbursed, and medicines not yet authorised whose company commits to apply, in which case the ANSM must first give a binding opinion on efficacy and safety unless the EMA's CHMP has already given a favourable one. The legal time limit for a decision is 90 days. The HAS four-year report, for 1 July 2021 to 30 June 2025, counts 394 decisions, of which 223 on first requests (94 before and 129 after marketing authorisation): 152 favourable (68 per cent) and 71 unfavourable. Of the 152, 75 were in oncology and 12 in haematology. The median time to decide a first request was 80 days, and medicines stayed in early access for an average of 17 months (510 days) before ordinary reimbursement. The HAS reports more than 140,000 patients treated, and that 80 per cent of products later rated ASMR I to III in the ordinary procedure had already been available through early access. A decree of 3 June 2026 (no. 2026-448), in force from 1 September 2026, led the HAS to revise its guidance and application forms.

The ANSM issues autorisations d'accès compassionnel (AAC) at the request of a doctor or hospital for a named patient, for medicines mainly not intended to be marketed in that indication and not under development or review there, valid for one year and renewable; where the medicine is being developed commercially, compassionate supply is tied to the company's commitment to apply for early access. A cadre de prescription compassionnelle (CPC), for three renewable years, covers an authorised medicine used in practice for other situations, provided no other treatment meets the need and the data suggest a favourable balance of benefit and risk. The ANSM publishes a reference list of medicines under exceptional access.

INCa describes the regime, defined in 2007 and recast by decrees no. 2022-689 and 2022-693 of 26 April 2022, as resting on three pillars: cross-cutting quality criteria, criteria specific to each treatment modality (cancer surgery, external radiotherapy and brachytherapy, systemic drug treatment), and minimum annual activity thresholds for some treatments and cancers. Authorised establishments must discuss each patient's file in a réunion de concertation pluridisciplinaire (RCP), announce the diagnosis and the treatment proposal according to INCa's protocol and give the patient a personalised care programme, belong to a regional cancer network (DSRC), organise supportive and palliative care, favour access to innovative treatments and trials, and be, or cooperate with, a hospital molecular genetics platform recognised by INCa so that tumour genetic tests can be done. Every child under 18 must be discussed at an interregional paediatric tumour board, and suspected rare cancers must be operated on within the rare-cancer networks. INCa's argument for volume thresholds cites a 2014 study of 11,865 patients in which high-volume hospitals and surgeons improved survival in advanced ovarian cancer. The Panorama counts 826 authorised establishments in 2024.

The droit à l'oubli applies to any cancer, whatever its site or type, if the treatment protocol ended at least five years earlier (the end of active treatment by surgery, radiotherapy or drugs, even if hormone therapy or immunotherapy continues), there has been no relapse, and the insurance contract ends before the borrower's 71st birthday. The person then declares neither the cancer nor the long-term condition status and is insured with no surcharge or exclusion. For people who do not qualify, the AERAS reference grid lists conditions, cancerous or not, that carry no or little excess risk and should be insured at or near the standard rate. The AERAS convention is signed by the government, banks, insurers and patient and consumer associations.

The programme's own report, in The Lancet Regional Health - Europe in 2025, says France was one of the first countries to integrate genome sequencing into clinical practice nationwide, starting with rare diseases, cancer genetic predisposition and cancers. By 31 December 2023 it had returned 12,737 results for rare disease and predisposition patients (median 202 days, diagnostic yield 30.6 per cent) and 3,109 for cancer patients (median 45 days). A 2026 study of 120 people with liver cancers that had stopped responding to treatment, sequenced through the programme, found an actionable alteration in 67 of 103 interpretable genomes; 31 received a matched drug and the disease was controlled in 10.

Unicancer's site describes itself as the only French hospital network dedicated 100 per cent to cancer, bringing together the 18 CLCCs and two affiliated establishments, with more than 25,500 staff; the federation marked its sixtieth anniversary in 2024. It sponsors academic trials across the network and runs Harmonie, a health data warehouse that links real-world clinical data from its centres to the national health data system (SNDS), with twenty years of history. INCa's clinical research survey found that 33 per cent of patients included in cancer studies in France in 2024 were included in CLCCs.

INCa was created by the public health law of 9 August 2004 and implements the ten-year cancer control strategy 2021-2030, launched by the President of the Republic around four priorities: prevention, limiting after-effects and improving quality of life, poor-prognosis cancers, and making sure progress reaches everyone. INCa reports that 212 of the strategy's 237 actions had been started between 2021 and 2025, and more than 70 of them finished. The second roadmap, for 2026 to 2030, sets five priorities, including a 'prevention generation' starting in early childhood, screening for everyone with organised lung screening generalised by 2030, and an 'augmented' care pathway with faster diagnosis for poor-prognosis cancers. INCa also labels research sites: its page lists eight integrated cancer research sites (SIRIC): CARPEM, CURAMUS, Curie, EpiCURE, ILIAD, InSiTu, LYriCAN+ and Montpellier Cancer.

The survey, a declaration by university hospitals, CLCCs, general hospitals and private establishments, counts every cancer study open to inclusion or in follow-up. In 2024 it recorded 3,665 studies, 71 per cent multicentre and 44 per cent international, and 70,530 patients included: 44,690 in interventional trials, 60,455 in studies of solid tumours and 9,657 in haematology. Half the inclusions were in university hospitals (CHU), 33 per cent in CLCCs, 9 per cent in private establishments and 8 per cent in general hospitals. 1,366 children, adolescents and young adults were included, 97 per cent in academic trials, and the university hospitals of Martinique, Guadeloupe and La Réunion included 478 patients. INCa reports inclusions up by more than 22 per cent between 2021 and 2024.

The trial list below includes PRODIGE 4/ACCORD 11, PRODIGE 24, PRODIGE 23, PRODIGE 7, GETUG-AFU 15 and 17, GETUG 13, PEACE-1 and SAFIR-ABC10 (Unicancer), PRIMA, REMARC and AHL2011 (LYSA), IFM 2009 and CASSIOPEIA (Intergroupe Francophone du Myélome), ALFA-0701, MAPS (IFCT), PAOLA-1 (ARCAGY-GINECO), TPExtreme and NIVOPOSTOP (GORTEC), LAP07 (GERCOR), LMS-04 (French Sarcoma Group), and CARMENA and GRAALL-2005 (Paris public hospitals). Each record links to its registry entry and to its primary paper where the corpus has it.

ameli.fr and assurance-maladie.ameli.fr answered with a human-verification page, so the number of people in the cancer long-term condition scheme, and the CNAM report the Panorama quotes for 29.3 billion euros, were not read at source; the 29.3 billion figure is quoted from the Panorama. sante.gouv.fr served a JavaScript wall, so CEPS's annual report, which gives price-negotiation times, was not read, and this page gives no figure for the time from European authorisation to French reimbursement. pfmg2025.aviesan.fr also served a JavaScript wall; the genomics card relies on the programme's peer-reviewed report instead. The HAS early-access figures are from its own report and infographic; there is no French public figure on this page for how many of those decisions were for cancer medicines later rated ASMR I to III.

This page adds the ANSM, Santé publique France and the Hospices Civils de Lyon, the national incidence estimate paper, the PFMG2025 programme report and its liver cancer study, and the IMPULSION lung screening pilot. The largest remaining gap is the trial record: OnCo's trials come mainly from ClinicalTrials.gov, while INCa runs its own registry of cancer trials open in France, and EU trials since 2023 are registered in CTIS. The 2024 survey counted 3,665 open cancer studies in France; OnCo's own count of trials with a French sponsor, or linked to a French institution, company or cooperative group, is about 200. The people list is weighted towards those who lead institutions and groups, because they are the people the corpus reached through institutional lists.

The survival figures are for people diagnosed between 2010 and 2015 and are national averages; they say nothing about an individual's outlook. The incidence figures are modelled, and the national registry created in 2025 has not yet published a count. Participation figures for screening come from different periods for each programme, as the table shows.