{"paper_id":"f4c6f0aa-0721-4dce-a174-366e44645053","body_text":"Secular trends in the incidence, prevalence, and survival of primary \nliver cancer in the United Kingdom from 2000-2021: a population-\nbased cohort study \n \nAuthors: \nBerta Cuyàs1,2,3*, MD \nEdilmar Alvarado-Tapias1,2,3*, PhD \nEng Hooi Tan4 , PhD \nAsieh Golozar5,6 , PhD \nTalita Duarte-Salles7,8 , PhD \nAntonella Delmestri4 , PhD \nJosep Maria Argemí Ballbé, PhD3,9,10 \nWai Yi Man4, MSc \nEdward Burn4, PhD \nCarlos Guarner-Argente1, PhD \nDaniel Prieto Alhambra4,8 , PhD \nDanielle Newby4, PhD \n \nAffiliations: \n1. Department of Gastroenterology, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de \nBarcelona, Sant Quintí, 89, 08041, Barcelona, Spain. \n2. Medicine Department, Autonomous University of Barcelona (UAB), 08193, Barcelona, Spain.  \n3. Centre for Biomedical Research in Liver and Digestive Diseases Network (CIBERehd), Instituto de \nSalud Carlos III, 28029, Madrid, Spain. \n4. Centre for Statistics in Medicine, Nuffield Department of Orthopaedics, Rheumatology and \nMusculoskeletal Sciences, University of Oxford, Oxford, OX3 7LD, United Kingdom. \n5. Odysseus Data Services, Cambridge, MA 02142, United States.  \n6. OHDSI Center at the Roux Institute, Northeastern University, Boston, MA 02115, USA.  \n7. Fundació Institut Universitari per a la recerca a l'Atenció Primària de Salut Jordi Gol i Gurina \n(IDIAPJGol), Gran Via Corts Catalanes, 587 08007 Barcelona, Spain.  \n8. Department of Medical Informatics, Erasmus University Medical Centre, 3015 GD Rotterdam, The \nNetherlands. \n9. Liver Unit. Clinica Universidad de Navarra. DNA & RNA Medicine Program. CIMA University of Navarra. \nAv Pio XII 36, Pamplona 31008, Navarra, Spain. \n10. University of Pittsburgh. Division of Gastroenterology Hepatology and Nutrition. 200 Lothrop st \nPittsburgh PA 15260. \n \n*Joint first authors \n \nCorresponding author:  \nProfessor Daniel Prieto Alhambra \nCentre for Statistics in Medicine, Nuffield Department of Orthopaedics, Rheumatology and \nMusculoskeletal Sciences, University of Oxford, Oxford, UK.  \ndaniel.prietoalhambra@ndorms.ox.ac.uk\n \n \nKEYWORDS \nPrimary liver cancer, hepatocellular carcinoma, incidence, prevalence, cancer survival \nWORD COUNT \nWords: 6,831. Abstract: 249 \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \nNOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.\n\nTABLES AND FIGURES \nTables: 2.  Figures: 5. \n \nCONFLICTS OF INTEREST \nProfessor Daniel Prieto-Alhambra research gr oup has received research grants from the \nEuropean Medicines Agency, from the Innovative Medicines Initiative, from Amgen, Chiesi, \nand from UCB Biopharma; and consultancy or s peaker fees (paid to his department) from \nAstellas, Amgen, Astra Zeneca, and UCB Biopharma. All other authors declare no conflicts \nof interest. \nFUNDING  \nThis activity under the European Health Data & Evidence Network (EHDEN) and OPTIMA \nhas received funding from the Innovative Medicines Initiative 2 (IMI2) Joint Undertaking \nunder grant agreement No 806968 and No. 101034347 respectively. IMI2 receives support \nfrom the European Union’s Horizon 2020 research and innovation programme and European \nFederation of Pharmaceutical Industries and Associations (EFPIA). The sponsors of the \nstudy did not have any involvement in the writi ng of the manuscript or the decision to submit \nit for publication. Additionally, there was par tial support from the Oxford NIHR Biomedical \nResearch Centre. The corresponding author had full access to all the data in the study and \nhad final responsibility for the decision to submit for publication. EAT is a recipient of a Joan \nRodes award from the ISCII (JR20/00047) and the PI21/01995 grant from the ISCIII-Fondos \nFeder. \nAUTHORS CONTRIBUTIONS \nAll authors were involved in the study conception and design, interpretation of the results, \nand the preparation of the manuscript. DN carried out data analysis for the manuscript. AG  \nreviewed the clinical code list used in this study. BC, EAT and DN wrote the initial draft of the \nmanuscript with EB and DPA. AD and WYM implemented the data curation, data \nharmonisation, data quality tests and assessment. DN, EB, AD, WYM and DPA had access \nto the CPRD data. All authors were involved in the interpretation of the results, critically \nreviewed the final manuscript, and gave consent for publication. \n \n \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\nABSTRACT \nBackground: Primary liver cancer (PLC) remains a global health challenge. Understanding \ntrends in the disease burden and survival is crucial to inform decisions regarding screening, \nprevention and treatment.   \nMethods: Population-based cohort study using UK primary care data from the Clinical \nPractice Research Datalink (CPRD) GOLD (200 0 to 2021), replicated in CPRD Aurum. PLC \nincidence rates (IR), period prevalence (PP) and survival at one, five and ten years over the \nstudy period were calculated, and stratified by age, sex and diagnosis year. \nResults: The crude IR of PLC was 4.56 (95%CI 4.42-4.70) per 100,000 person-years \nbetween 2000 and 2021, with an increase over time across age and sex strata. Sex-specific \nIR for males was higher than females, 6.60 (95%CI 6.36-6.85) vs. 2.58 (95%CI 2.44-2.74) \nper 100,000 person-years. Crude PP showed a 7-fold  increase over the study period, with \nPP 0.02% (95%CI 0.019%-0.022%) in 2021, and a 2.8-fold higher PP in males. Survival at \none, five and ten years after diagnosis was 41.7%, 13.2% and 7.1%, respectively, for both \nsexes. One-year survival increased only in men, from 33.2% in 2005-2009 to 49.3% in 2015-\n2019. \nConclusion: Over the past two decades, there has been a significant increase in the \nnumber of patients diagnosed with PLC. Despite a slight improvement in median and one-\nyear survival in men, prognosis remains poor. To improve the survival of PLC patients, it is \nnecessary to understand the epidemiological changes and address the preventable risk \nfactors associated with liver disease and promote early detection and access to care.  \n \nAbstract word count: 249 words \n \n \nLAY SUMMARY \nThis population-based cohort study shows that t he incidence and prevalence of primary liver \ncancer in the UK has increased in the last  20 years across both sexes and age groups, with \na 7-fold increase in crude period prevalence ov er the study period. One-year survival has \nimproved only in males over the study period and, regrettably, no increases in long-term \nsurvival were observed. Our findings are a call  for awareness to stimulate further research \nand public health actions on liver cancer.  \n \n \n \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\nINTRODUCTION \nPrimary liver cancer (PLC) is the sixth most  common cancer and the third leading cause of \ncancer death worldwide (1). Hepatocellular carcinoma (HCC) is the most common form of \nPLC accounting for approximately 90% of cases (2). \nThe incidence and mortality of PLC are growing worldwide, with approximately 906,000 new \ncases and 830,000 deaths in 2020. Incidence remains highest in Eastern Asia and Northern \nAfrica, although it is increasing in different par ts of Europe, Americas, and Oceania (3). PLC \nwas also one of the top five causes of cancer mortality in some countries in Europe, such as \nBosnia and Herzegovina, France, Italy, Republic of Moldova, and Romania; and Western \nAsia (4). The incidence and mortality from PLC are predicted to rise by more than 50% over \nthe next 20 years (1) with the UK showing a rapid rate of incident cases and projected to \nhave one of the highest annual increases over the next decade (5).  \nImprovements in survival have been made over  previous decades due to advancements in \nchemotherapy, surgical techniques, and the shift towards multidisciplinary teams to manage \ncare. Despite this, survival is still poor ev en in high-income countries compared to other \ncancers with one and five-year relative survival estimates around 40% and 10% respectively \n(4,6). \nThe main risk factors linked to PLC are cirrhosis, Hepatitis B virus (HBV), Hepatitis C virus \n(HCV), harmful alcohol consumption and metabolic factors such as diabetes and obesity (7). \nWith the use of direct-acting antiviral therapy , the risk attributed to HCV has substantially \ndecreased globally, while alcohol-associated li ver disease (ALD) and metabolic dysfunction-\nassociated steatotic liver disease (MASLD) are becoming more prominent risk factors for \nPLC (8). Less prevalent risk factors include autoimmune hepatitis, hemochromatosis, \nα 1-\nantitrypsin deficiency and aflatoxin ingestion. Regarding sociodemographic characteristics, \nolder age, being male and some racial or ethni c minorities (in particular, Hispanics) as well \nas lifestyle factors such as cigarette smokin g have also been associated with HCC (2). The \ndistribution of these risk factors has gradually changed over time and between populations \n(2).  \nUnderstanding trends in the incidence, prevalence, and overall survival of liver cancer is an \nimportant aspect to inform decisions regar ding screening, prevention, treatment, and \ndisease management. Due to the increases in ri sk factors for PLC such as obesity, alcohol \nconsumption and diabetes, a comprehensive asse ssment of the trends and disease burden \nof PLC is lacking in the UK. We therefore set out to characterise the secular trends of PLC in \nterms of incidence, prevalence and survival in the UK.  \nMETHODS \nData sources and Study design \nWe carried out a population cohort study using routinely collected primary care data from the \nUK. People with a diagnosis of PLC and a background cohort (denominator population) were \nidentified from Clinical Practice Research Datalink (CPRD) GOLD (July 2022). We \nadditionally carried out this study using CPRD Aurum to compare the results with CPRD \nGOLD. These databases contain pseudo-anonymi sed patient-level information on \ndemographics, lifestyle data, clinical diagnoses, prescriptions, and preventive care \ncontributed by general practitioners from the UK. CPRD GOLD contains data from across \nthe UK whereas Aurum only contains data from England. The use of CPRD data was \napproved by CPRD’s Research Data Governance process (22_001843). GOLD and Aurum \nare established primary care databases broadl y representative of the UK population (9). \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\nBoth databases were mapped to the Observati onal Medical Outcomes Partnership (OMOP) \nCommon Data Model (CDM) (10,11).  \nStudy participants and time at risk \nAll individuals were required to be aged 18 years or older and have at least one year of prior \nhistory. For the incidence and prevalence analysis , the study cohort consisted of individuals \npresent in the databases from 1st January 2000.  For CPRD GOLD, these individuals were \nfollowed up to whichever came first: the practice stopped contributing to the database, the \npatient left the practice, date of death, or the 31st of December 2021 (the end of study \nperiod) whereas for Aurum, the end of the study period was 31st of December 2019. For the \nsurvival analysis, only individuals with a newly diagnosed cancer were included. These \nindividuals were followed up from the date of thei r diagnosis to either date of death, practice \nstopped contributing to the database, patient left the practice, or end of the study period. Any \npatients whose death and cancer diagnosis occurred on the same date were removed from \nthe survival analysis. \nPrimary liver cancer definitions \nWe used Systematized Nomenclature of Medicine - Clinical Terms (SNOMED CT) \ndiagnostic codes to identify PLC events. Diagnostic codes related to intrahepatic \ncholangiocarcinoma were excluded. Diagnostic co des indicative of either non-malignant \ncancer or metastasis were excluded as well as  diagnosis codes indicative of melanoma and \nlymphoma occurring in the organs of interest. The study outcome cancer definition was \nreviewed with the aid of the Cohort Diagnostics R package (12). This package was used to \nidentify additional codes of interest and to remo ve those highlighted as irrelevant based on \nfeedback from clinicians with oncology, primar y care, and real-world data expertise through \nan iterative process during the initial stages of analyses. The clinical code lists used to \ndefine PLC can be found in supplementary information S1. OMOP-based computable \nphenotypes are available, together with all analytical code on Github to enable reproducibility \n(https://github.com/oxford-pharmacoepi/EHDENCancerIncidencePrevalence\n). For overall \nand annual crude incidence rates (IR) and annual prevalence, all PLC events in the period \n2000-2021 were included. For survival analyses, mo rtality was defined as all-cause mortality \nbased on date of death records.  \nStatistical methods \nThe population characteristics of patients with  a diagnosis of PLC were summarised, with \nmedian and interquartile range (IQR) used for continuous variables and counts and \npercentages used for categorical variables.  \nFor incidence, the number of events, the observed time at risk, and the incidence rate per \n100,000 person years were summarised along with 95% confidence intervals (95% CI). \nAnnual crude incidence rates were calculated as the number of incident PLC cases as the \nnumerator and the recorded number of person-ye ars in the general population within that \nyear as the denominator whereas overall incidence was calculated from 2000 to 2021.  \nAge-standardized IRs were calculated using the 2013 European Standard Population \n(ESP2013) (21). The ESP2013 is a population standard with a predefined age distribution \nwhich accounts for differences in age structur es between different populations to ensure fair \ncomparisons. The ESP2013 provides predefi ned age distribution in five-year age bands; \ntherefore, we collapsed these to obtain dist ributions for ten-year age bands used in this \nstudy. We used the age distribution of 20-29 years from ESP2013 for age-standardization as \nage distributions were not available for 18-29 years age band used in this study.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\nPeriod prevalence was calculated on 1st January for the years 2000 to 2021, with the \nnumber of patients fulfilling the case definit ion for liver cancer as the numerator. The \ndenominator was the participants eligible on 1st January in the respective years for each \ndatabase. The number of events, and preval ence (%) were summarised along with 95% \nconfidence intervals.  \nFor survival analyses, we used the Kaplan-Mei er method to estimate the overall survival \nprobability from observed survival times with 95% confidence intervals. We estimated the \nmedian survival and survival probability, one, five, and ten years after diagnosis. Any \npatients whose death date and cancer diagnosis date occurred on the same date were \nremoved from the survival analysis. \nAll results were stratified by database, age (ten-year age bands apart from the first and last \nage bands which were 18-29 years and 90+ years, respectively) and sex. For survival \nanalysis, we additionally stratified by cal endar time of cancer diagnosis (2000-2004, 2005-\n2009, 2010-2014, 2015-2019 and 2020-2021) allowing a maximum of five years follow-up \nfrom cancer diagnosis. To avoid re-identificati on, we do not report results with fewer than \nfive cases.  \nFor Aurum, the same statistical analyses were performed using data from 1st January 2000 \nto 31st\n December 2019 to compare with results from GOLD apart from the calendar time \nstratification which was only performed in GOLD. \nThe statistical software R version 4.2.3 was used for analyses. For calculating incidence and \nprevalence, we used the Incidence Prevalence R package (13). \nRESULTS \nPatient Populations and characteristics \nOverall, there were 11,388,117 eligible patients,  with at least one year of prior history \nidentified from January 2000 to December 2021 from CPRD GOLD. Attrition tables for this \nstudy can be found in the supplementary information (Supplement S2). A summary of study \npatient characteristics of those with a diagnosis of PLC for GOLD is shown in Table 1.  \n  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\nTable 1:  Baseline characteristics of primary liver cancer patients at the time of \ndiagnosis stratified by sex from CPRD GOLD. \nSex Male  Female \nNumber of patients 2,848 1,151 \nAge in years, Median (IQR) 70 (62 to 77) 73 (64 to 80) \nAge groups in years, N (%)     \n18-29   \n30-39   \n40-49   \n50-59   \n60-69   \n70-79   \n80-89   \n90+   \n11 (0.4%) \n18 (0.6%) \n97 (3.4%) \n411 (14.4%) \n811 (28.5%) \n952 (33.4%) \n512 (18.0%) \n36 (1.3%) \n<5 \n9 (0.8%) \n41 (3.6%) \n149 (12.9%) \n260 (22.6%) \n385 (33.4%) \n271 (23.5%) \n33 (2.9%) \nPrior history in days, Median (IQR) 3,977 (2,189 to \n5,642) \n3,943 (2,110 to \n5,642) \nComorbid conditions (any time prior)       \n    Chronic liver disease, N (%)   663 (23.3%) 176 (15.3%) \n    Recorded risk factors for PLC, N (%) \n    Alcoholic liver disease (ALD) \n    Hepatitis C \n    Hepatitis B \n    Non-alcoholic fatty liver disease (NAFLD) \n    Hemochromatosis \n    Autoimmune hepatitis \n  \n385 (13.5%) \n91 (3.2%) \n20 (0.7%) \n80 (2.8%) \n90 (3.2%) \n14 (0.5%) \n \n52 (4.5%) \n25 (2.2%) \n<5 \n26 (2.3%) \n9 (0.8%) \n37 (3.2%) \n    Other recorded risk factors, N (%) \n    Hypertensive disorder \n    Diabetes  \n    Hyperlipidaemia \n    Obesity \n  \n850 (29.8%) \n851 (29.9%) \n209 (7.3%) \n313 (11.0%) \n \n349 (30.3%) \n215 (18.7%) \n99 (8.6%) \n102 (8.9%) \nSmoking Status (any time 5 years prior), N (%) \nCurrent/former smoker \nNonsmoker \nMissing/no records \n \n806 (28.3%) \n954 (33.5%) \n1,088 (38.2%) \n \n270 (23.0%) \n559 (48.6%) \n322 (28.0%) \n \nThere were 3,999 patients with PLC in CPRD GOLD. Overall, those diagnosed with PLC \nwere more likely to be male (71%), with a median age of 71 (IQR 62 to 77) years at \npresentation. The highest percentage of PLC patients were those aged 70-79 years old \ncontributing to 33.4% of diagnosed patients, for both males and females, with similar \nobservations in Aurum (Supplement S3). Ma les had higher prevalence of chronic liver \ndisease, ALD and hemochromatosis, diabetes and were more likely to be smokers \ncompared to females. However, females had higher proportions of autoimmune hepatitis \ncompared to males. Similar percentages r egarding HCV and NAFLD were observed in PLC \npatients across both sexes. \nIncidence rates stratified by calendar year, age and sex \nThe overall crude IR of liver cancer in 2000 to 2021 was 4.56 (95% CI 4.42 to 4.70) per \n100,000 person-years. Sex-specific IR for females was 2.58 (95% CI 2.44 to 2.74) and for \nmales was 6.60 (95% CI 6.36 to 6.85) per 100, 000 person-years, with similar results in \nAurum. Annualised IRs increased from 2000 to 2019 for the whole population and both \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\nsexes with males having higher rates (Figure 1). For GOLD, IR dropped in 2020 before \nbouncing back in 2021. Age standardized incidence rates for CPRD GOLD for both sexes \nshowed similar trends (Supplement S4). All study results for this study can be found and \ndownloaded in a user-friendly interactive web application: https://dpa-pde-\noxford.shinyapps.io/LiverCancerIncPrevSurvShiny/. \n  \n \nFigure 1: Annual incidence rates for PLC from 2000 to 2021 stratified by database and \nsex. \nOverall crude IRs were higher with increasi ng age up to 80–89 years. Those aged 18 to 29 \nhad the lowest overall IRs with an IR of 0.09 (95% CI 0.05 to 0.15) per 100,000 person \nyears, whereas those aged 80–89-years had the hi ghest IR of 17.7 (95% CI 16.4- 18.9) with \nsimilar or slightly lower IRs in Aurum (Supplement S5).  \nAnnualised IRs for each age group (Figure 2) show IRs have increased over the study \nperiod for those aged 40-89 years of age. Fo r other age groups there was not enough data \nto assess annualised IR trends. Stratification on both sex and age showed similar trends to \nFigure 2 for both sexes (Supplement S6). Males had higher IRs across the study period with \nthe differences in IR between males and females widening over the study period. \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \nFigure 2: Annualised incidence rates from 2000 to 2021 stratified by database and age \ngroup. \nPrevalence for study population with database, age, and sex stratifications \nFor the whole population in GOLD, crude PP for PLC in 2021 was 0.020% (0.018% to \n0.022%) and 2.72-fold higher for males compared to females. Since 2000, PP has increased \n6.66-fold across the study period with males s eeing a 10-fold increase in PP compared to \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n5.5-fold increase in females. Similar PP was observed when comparing PP from 2000 to \n2019 across both databases (Figure 3). \n \n \nFigure 3: Annual prevalence from 2000 to 2021 for whole population and stratified by \nsex. \nWhen stratifying by age group, PP in 2021 was highest in those aged between 70-89 years \nof age (0.07%). PP increased over the study period for all age groups across both databases \nexcept for those aged 40-49 in GOLD where there was a little change in PP over time \nbetween 2011-2021, whereas Aurum showed a gradual increase in PP over the study period \n(Figure 4). Stratification on both sex and age group showed similar trends with females \ndriving the increase in PP in those aged 40-49 years old in Aurum (Supplement S7). \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \nFigure 4: Annual prevalence from 2000 to 2021 stratified by database and age group. \nOverall survival rates for cancer population with age, sex, and calendar year \nstratification \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\nThere were 3,892 patients with 3,047 (78.3%) deat hs over the study period in GOLD with a \nmedian follow up of 0.58 (IQR 0.17-1.58) years. The median survival for the whole \npopulation was 0.70 (95%CI 0.65-0.74) years with a slightly higher median survival of \naround 0.80 years (95%CI 0.75-0.84) in Aurum (Supplement S8). Survival after one-, five- \nand ten-years after diagnosis was 41.7%, 13.2% and 7.1% for both sexes. Similar \nobservations although slightly higher were observed in Aurum (Supplement S9).  \nMedian survival was slightly lower for females compared to males across both databases. In \nGOLD, females had a median survival of 0.62 (95% CI 0.56-0.69) years whereas males had \nmedian survival of 0.74 years (95% CI 0.68-0.81). Stratifying by age group, the general trend \nindicates median survival decreased from  40 years onwards for both databases \n(Supplement S10). Short term survival decreas ed with age from 30-39 years of age with the \nlowest survival observed in those aged 90 years and older (Table 2). \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \nTable 2: Survival rates of PLC from 2000 to 2019 for CPRD Aurum and 2000-2021 for GOLD stratified by database and age group. \n One year survival (%) Five-year survival (%) Ten-year survival (%) \nAge \nGroup Aurum GOLD Aurum GOLD Aurum GOLD \n18-29 62.9 (47.6 - 83.0) 31.8 (14.2 - 70.8) 46 (30.3-69.9) - - - \n30-39 65.7 (55.7 - 77.3) 62.5 (45.7 - 85.5) 51.9 ( 41 - 65.7) 44.2 (26.7 - 73.3) 46.7 (34.1-63.9) - \n40-49 60.3 (54.9 - 66.3) 54.9 (47.1 - 64.1) 35.8 (30 - 42.7) 29.1 (21.5 - 39.3) 25.5 (19.5 - 33.3) 26.5 (19.1 - 36.8) \n50-59 53.3 (50.3 - 56.5) 48.4 (44.3 - 52.9)  22.4 (19.6-25.7) 20.3 (16.8 - 24.7) 17.1 (14.1 - 20.7) 12.5 (9.0 - 17.4) \n60-69 50.7 (48.4 - 53.1) 47.8 (44.9 - 51.0)  19.1 (17.0-21.4) 16.0 (13.6 - 18.9) 10.3 (8.3 - 12.8) 9.2 (6.9 - 12.4) \n70-79 42.4 (40.3 - 44.5) 37.7 (35.1 - 40.5)  12.4 (10.8-14.2) 10.1 (8.3 - 12.3) 5.3 (3.9 - 7.2) 3.6 (2.3 - 5.7) \n80-89 33.9 (31.2 - 36.8) 28.1 (25.0 - 31.6) 6.9 (5.21-9.11) 5.0 (3.3 - 7.6) 1.9 (0.7 - 5.2) 0.7 (0.1 - 4.3) \n90+ 21.6 (15.6 - 30) 30.3 (20.6 - 44.6) 1.6 (0.26 - 10.2) 3.1 (0.5 - 19.6) - - \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \nTo investigate if survival has changed over time in GOLD, we stratified by calendar time of \ncancer diagnosis in five-year age windows. Figure 5  shows the Kaplan-Meier survival curve \nfor the whole population, stratified by sex and calendar year for GOLD.  \nOverall, median survival increased from 0.45 years (95% CI 0.37 - 0.65) for those diagnosed \nin 2000-2004 to 0.912 (95% CI 0.783 - 0.997)  years for those diagnosed in 2015-2019. \nHowever, when stratifying by sex this trend was only observed in males with median survival \nincreasing from 0.427 (95% CI 0.287 - 0.597) to 0.975 (95% CI 0.862 - 1.098). There was no \ndifference in median survival when compar ing 2015-2019 with 2020-2021. Stratification by \nage group did not show any clear pattern in improvements in median survival over time apart \nfrom those aged 60-69 years where there wa s an increase in median survival comparing \nthose diagnosed between 2005-2009 with 2015-2019 (0.63 years to 1.33 years) \n(Supplement S11) \nIn general, survival at one-year was higher in those diagnosed between 2015-2019 \ncompared with those diagnosed between 2000-2004 with survival increasing from 35.6% \n(95% CI 30.5 - 41.5) to 46.7% (95% CI 43.8 - 49.9). Stratification by sex showed one-year \nsurvival increased in males only, from 33.2% (95% CI 27.2 - 40.4) to 49.3% (95% CI 45.9 - \n53.1). For long term survival, there were no in creases in five-year survival for the whole \npopulation and for each sex (Supplement S12). One-year survival also improved only in \nthose aged 60-69 years of age (35.52% (95% CI 26.6 - 47.5) to 58.2% (95% CI 52.7 - 64.3)). \nThere was no clear pattern in improvements in long term survival over calendar year for \nother age groups. \n \nFigure 5: Kaplan-Meier survival curve of PLC stratified by sex and calendar year of \ndiagnosis. \nDISCUSSION \nThis study provides a comprehensive descriptive  analysis of the trends in PLC epidemiology \nin the UK. The incidence and prevalence of pr imary liver cancer (PLC) in the UK has \nincreased from 2000 to 2021 across both sexes and age groups. Regarding survival, males \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \nhad slightly higher median survival compared to females, but one-, five- and ten-year \nsurvival were similar between sexes. Short-term survival has improved only in males over \nthe study period, no increases in long-term survival were observed overall or by sex and age \nstrata. \n \nIncidence rates for PLC reported here are lower but broadly in line with the National Cancer \nStatistics (NCS) (9.4 per 100,000 person-years 2016-2018) (6), and other global studies \nacross high socio-demographic index countries such as the US, Australia, the Netherlands, \nand the UK (14,15). Slight differences are likely due to differences in study periods (2000-\n2021 for our data vs. 2016-2018 for NCS), possible incompleteness/misclassification in \nprimary care records and inclusion of other subt ypes of PLC such as cholangiocarcinoma in \nnational statistics which were we excluded these subtypes in this study. \n \nIn terms of secular trends, incidence increased for both sexes across the study period with \nsimilar trends seen in other recent  works from the UK, where Burton et al showed that PLC \nincidence increased from 4.4 per 100,000 in 1997 to 9.6 in 2017 in the UK, reaching a \nplateau since 2014 (16), and Liao et al  revealed that age-standardised incidence rates \nincreased over time from 2008 to 2018 (17). The present work follows up on from several \nCommissions which showed that standardised mortality rates from liver disease had \nincreased 400% since 1970 and that liver diseas e was the third biggest cause of premature \nmortality in UK, while mortality by other organ diseases kept decreasing. They presented ten \nrecommendations to reduce the burden of liver disease in the UK (18). However, the final \nreport of the Lancet Commission in 2021 stressed the continuing increase in burden of liver \ndisease, especially from excess alcohol consumption and obesity, which is concerning as \n49% of primary liver cancers are preventable (19). \n \nSome of the possible reasons that explain the increase in incidence of PLC in UK could be \nthe following. Firstly, as most PLCs typically  develop in patients with chronic liver disease, \nthe changes in the comorbidity profiles of patient s over recent decades are likely to play an \nimportant role in increasing PLC rates (18,20) . In the last 10 years, the introduction of \nsuccessful new treatments for HCV infection has generated a reduction in the prevalence of \nHCV-related HCCs (21). Despite this, there is an alarming rise in the prevalence of non-viral \nrisk factors, such as high alcohol consumption and alcohol-associated liver disease (ALD) \n(22), which could explain the increase in PL C (23). Furthermore, data from the Health \nSurvey for England revealed that over 60% of adul ts in the UK were overweight or obese in \n2021, with a higher proportion of men than wo men (24). The increase in obesity and \ndiabetes are also directly related to the incr ease of MASLD (25), which could contribute to \nthe increase in PLC cases. The main risk factors and the underlying liver diseases (ALD, \nMASLD) related to PLC identified in this study  are consistent with the recent data (16) \nshowing the greater burden of non-viral risk factors in front of the classical viral etiological \nfactors (i.e. HCV). Similar results have been described recently and the perspective is that \nthis trend will continue over the next years (2). \n \nSeveral studies have reported higher incidence rates in males compared to females in line \nwith our results, in the UK (16,17,26) and worldwide (14,27). Higher incidence rates with \nincreasing age are also supported by this work  (28). A decline in IR in the year 2020 was \nobserved coinciding with the COVID-19 pandem ic, which bears resemblance to findings \nfrom other studies that showed a significant reduction in HCC diagnosis during the first year \nof the pandemic attributed to the disruption of  routine healthcare (29). The pandemic also \ninfluenced the access to HCV treatment in the UK, with studies reporting a decrease of \n40.2% between 2019 and 2021 (30), which could reverse some of the progress made in \nHCV control and may influence future PLC rates (14).  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \n \nOne-, five- and ten-year survival in our study are in line with UK cancer registry and Burton \net al (6,16) as well as other international studies (4). Many types of cancer such as breast, \nprostate, stomach and colon have seen improvements in survival over recent decades in \nhigh-income countries (1,31,32), but the same tendency has not been seen in PLC, which \nremains one of the lowest survival rates of any cancer in the UK. As with many other \ncancers, liver cancer survival is significantly better when the disease is diagnosed at an early \nstage, however, in UK only 3 in 10 PLC are diagnosed at early stages (33) which could \nexplain the low survival rates in this study . In this regard, since 2022, NHS England are \npromoting an early detection liver cancer pi lot programme to help transform outcomes for \nPLC patients by checking for advanced fibrosis in high-risk communities (34). \n \nMales have higher levels of disease burden fo r PLC compared to females. The same sex \ndisparity trend has been observed globally (14). It has been suggested that this tendency \nmight be driven by differences in the prevalenc e of HCC risk factors, a lack of adherence to \nfollow-up/screening programs as well as epi genetics and biological factors (14,35). Despite \nwell-established sex differences in PLC incidence, we observed slightly lower median \nsurvival and one-year survival for females in our study. Cancer Research UK statistics for \nthe period 2015-2019, are in line with our resu lts showing a one-year survival of 39.3% \nfemales vs 49% males in Scotland, 35.9% vs 41.8% in England, 32.5% vs 37.4% in Wales \nand 39.8% vs 41% in Northern Ireland, respective ly (36). However, studies in Asia and the \nU.S. demonstrated that women with HCC present better overall survival than men likely due \nto better adherence to HCC surveillance (37,38). Furthermore, a study in the U.S. reported \nthat younger women had significantly better survival than younger men; however, this \nsurvival difference was not observed in ol der men and women, hypothesising that sex \nhormones may have a role to explain sex differ ences in survival (39). It has also been \nsuggested that the patterns of alcohol co nsumption have been underreported particularly \namong women, in whom the alcohol burden has increased in recent years (40), and that \nfemales are more susceptible to alcohol-induced liver injury leading to cirrhosis and an \nincreasing risk of HCC, although the exact me chanisms are still unclear (41,42). Females \nare less likely to use prevention services than males, possibly related to more perceived \nstigma, conflicting child/family, personal needs, and financial barriers leading to delayed \ndiagnosis (43).  \n \nIncreasing age leads to a progressive physio logical and metabolic reprogramming to adapt \nto gradual deterioration of organs and functions that can play an important role in liver \ncarcinogenesis in the elderly (44). A UK study showed that patients with MASLD-associated \nHCC were older than those with other aetiologies (71.3 years vs 67.1 years) and their \ncancers less often detected by surveillance ( 45). However, elderly patients have similar \nsuccess to treatments than younger patients and should be considered for all treatments \nafter assessment of their baseline clinical status and cancer burden (46). \n \nEncouragingly, median survival in our study has doubled from 2000-2004 to 2015-2019, \nreaching almost eleven months in males. Ding et al also showed a significant improvement \nin survival for HCC patients over the pas t three decades in the U.S., which has been \nattributed to advances in early diagnosis an d therapeutic approaches, such as effective \nsystemic therapy (47).  \n \nWe did not observe significant improvements of survival in our study period, which could be \ndue to numerous reasons. Firstly, despite the impressive reduction in the rates of HCV \ninfections, the UKHSA 2023 report suggests that  nearly three-quarters of people still living \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \nwith chronic HCV are unaware of their infection ( 30). As HCV is a risk factor for PLC, this \ncould explain no improvements in survival due to HCV leading to PLC diagnosis delays. The \nslight improvement in survival observed in males could be due to HCC screening programs \nwhich are aimed mainly at populations with chronic liver disease (cirrhosis), where most \npatients are male, which favours earlier diagnosis of PLC with more curative treatment \noptions (4). It has been reported that liver  cancers diagnosed in asymptomatic subjects \nwithin these screening programs are associated wi th a slight increase in survival (48). The \nlate diagnosis of this disease, especially in  women, could be related to socio-demographic \nfactors such as scarce primary care consulta tion due to family and social limitations \ngenerating a delay in the diagnosis of PLC and limiting the treatment options with the \nconsequence of lower survival. Finally, the impac t of the primary prevention measures, such \nas hepatitis B vaccination, reducing alcohol  and tobacco use, combating obesity and better \ncontrol of cardio-metabolic factors and the early diagnosis using more sensitive radiologic \nmethods, above all, could be more clearly seen in better long-term survival in the next 10 \nyears. \n \nThe main strength of this study is the use of  two large representative data sources covering \nthe whole of the UK. CPRD GOLD covers primary care practices from England, Wales, \nScotland, and Northern Ireland whereas CPRD Aurum covers England. The similarity \nbetween the results in both databases provides increased generalizability across the UK and \ndemonstrates the robustness of our findings. Another strength of our study is the inclusion of \na complete study population database for the assessment of incidence and prevalence. In \ncontrast, cancer registry studies extrapolate th e registry data to the whole population using \nnational population statistics, potentially intr oducing inaccurate denominators (49). The high \nvalidity and completeness of mortality data wi th over 98% accuracy compared to national \nmortality records (50) allowed us to examine t he impact of calendar time on overall survival - \none of the key outcomes in cancer care.   \n \nOur study has some limitations. Firstly, primary care data without linkage to a cancer registry \nwas used, potentially leading to misclassification and delayed recording of cancer diagnoses. \nHowever, previous validation studies have shown high accuracy and completeness of cancer \ndiagnoses in primary care records (51). Sec ondly, our use of primary care records also \nprecluded us from studying tumour histology , genetic mutations, staging of tumour at \ndiagnosis, or cancer therapies, which can all impact PLC survival. Finally, the main risk \nfactors for PLC, including the percentage of  underlying chronic liver disease, may be \nunderrepresented in this study, as in large population primary care databases they may be \ninfrequently recorded, or they might be diagnosed at the same time of the PLC and therefore \nnot included. The patterns of alcohol c onsumption can be also underreported or \nunderrecognized, particularly among women, being treated for medical conditions related to \nalcohol use (liver disease) or unrelated to alcohol use (MASLD) (40).  \n \nIn summary, the present study shows the number of people diagnosed with PLC in the UK \nhas substantially increased in the last 20 y ears and that overall survival remains low. \nUnderstanding the shift in the risk factors for PLC from virus-related to non-viral liver disease \nrequires more research and resources to manage the care of patients with ALD and MASLD \namong others. Although important therapeutic advances have been made and survival has \nslightly improved over time, over half of the patients with PLC are not alive after one year. \nTherefore, further progress in  prevention, early detection, and public health interventions \nsuch as screening programmes and education campaigns are needed.  \n \nABBREVIATIONS \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \nALD, Alcohol-associated liver disease. \nCPRD, Clinical Practice Research Datalink \nHCC, Hepatocellular carcinoma \nHCV, Hepatitis C virus \nHBV, Hepatitis B virus \nIR, Incidence rate \nIQR, Interquartile range \nMASLD, Metabolic dysfunction-associated steatotic liver disease \nNAFLD, Non-alcoholic fatty liver disease  \nOMOP, Observational Medical Outcomes Partnership \nPLC, Primary Liver Cancer \nPP, Period prevalence \nSNOMED CT, Systematized Nomenclature of Medicine - Clinical Terms  \n \nACKNOWLEDGEMENTS \nNone. \n \nDATA AVAILIABILITY \nThis study is based in part on data from the Clinical Practice Research Datalink (CPRD) \nobtained under the University of Oxford mult i-study licence from the UK Medicines and \nHealthcare products Regulatory Agency. The data is provided by patients and collected by \nthe NHS as part of their care and support. The interpretation and conclusions contained in \nthis study are those of the author/s alone. Patient level data used in this study was obtained \nthrough an approved application to the CPRD (application number 22_001843) and is only \navailable following an approval process to saf eguard the confidentiality of patient data. \nDetails on how to apply for data access can be found at https://cprd.com/data-access. \n \nREFERENCES  \n1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global \nCancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide \nfor 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209–49.  \n2. Llovet JM, Kelley RK, Villanueva A, Singal AG, Pikarsky E, Roayaie S, et al. \nHepatocellular carcinoma. Nat Rev Dis Primers. 2021;21;7(1):6.  \n3. Petrick JL, Braunlin M, Laversanne M, Valery PC, Bray F, McGlynn KA. International \ntrends in liver cancer incidence, overal l and by histologic subtype, 1978-2007. Int J \nCancer. 2016;1;139(7):1534-45.  \n4. Lepage C, Capocaccia R, Hackl M, Lemmens V, Molina E, Pierannunzio D, et al. \nSurvival in patients with primary liver c ancer, gallbladder and extrahepatic biliary tract \ncancer and pancreatic cancer in Europe 1999–2007: Results of EUROCARE-5. Eur J \nCancer. 2015;51(15):2169–78.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \n5. Smittenaar CR, Petersen KA, Stewart K, Moitt N. Cancer incidence and mortality \nprojections in the UK until 2035. Br J Cancer. 2016;25;115(9):1147-1155.  \n6. Cancer Research UK. Liver Cancer Statistics. Accessed 14 March 2024. \nhttps://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-\ncancer-type/liver-cancer.  \n7. Yang JD, Hainaut P, Gores GJ, Amadou A, Plymoth A, Roberts LR. A global view of \nhepatocellular carcinoma: trends, risk, prevention and management. Nat Rev \nGastroenterol Hepatol. 2019;Oct;16(10):589-604.  \n8. Kim D, Li AA, Gadiparthi C, Khan MA, Cholankeril G, Glenn JS, et al. Changing \nTrends in Etiology-Based Annual Mortality From Chronic Liver Disease, From 2007 \nThrough 2016. Gastroenterology. 2016;1163(e3).  \n9. Herrett E, Gallagher AM, Bhaskaran K, Forbes H, Mathur R, Staa T, et al. Data \nResource Profile: Clinical Practice Research Datalink (CPRD). Int J Epidemiol. \n2015;44(3):827-36.  \n10. Voss EA, Makadia R, Matcho A, Ma Q, Knoll C, Schuemie M, et al. Feasibility and \nutility of applications of the common data model to multiple, disparate observational \nhealth databases. J Am Med Inform Assoc. 2015;22(3):553-64.  \n11. Hripcsak G, Duke JD, Shah NH, Reich CG, Huser V, Schuemie MJ, et al. \nObservational Health Data Sciences and Informatics (OHDSI): Opportunities for \nObservational Researchers. Stud Health Technol Inform. 2015;216:574-8.  \n12. Gilbert J, Rao G, Schuemie M, Ryan P, Weaver J. CohortDiagnostics: Diagnostics for \nOHDSI Cohorts. R package version 3.2.5, \nhttps://github.com/OHDSI/CohortDiagnostics\n, \nhttps://ohdsi.github.io/CohortDiagnostics. 2023.  \n13. Raventós B, Català M, Du M, Guo Y, Black A, Inberg G, et al. IncidencePrevalence: \nAn R package to calculate population-level incidence rates and prevalence using the \nOMOP common data model. Pharmacoepidemiol Drug Saf. 2024;33(1):e5717.  \n14. Rumgay H, Arnold M, Ferlay J, Lesi O, Cabasag CJ, Vignat J, et al. Global burden of \nprimary liver cancer in 2020 and predictions to 2040. J Hepatol. 2022;77(6):1598-\n1606.  \n15. Liu Z, Jiang Y, Yuan H, Fang Q, Cai N, Suo C, et al. The trends in incidence of \nprimary liver cancer caused by specific etiologies: Results from the Global Burden of \nDisease Study 2016 and implications for liv er cancer prevention. J Hepatol. \n2019;70(4):674-683.  \n16. Burton A, Tataru D, Driver RJ, Bird TG, Huws D, Wallace D, et al. Primary liver cancer \nin the UK: Incidence, incidence-based mortality, and survival by subtype, sex, and \nnation. JHEP Rep. 2021;19;3(2):100232.  \n17. Liao W, Coupland CAC, Innes H, Jepsen P, Matthews PC, Campbell C, et al. \nDisparities in care and outcomes for primary liver cancer in England during 2008-\n2018: a cohort study of 8.52 million primary care population using the QResearch \ndatabase. EClinicalMedicine. 2023;11;59:101969.  \n18. Williams R, Aspinall R, Bellis M, Camps-Walsh G, Cramp M, Dhawan A, et al. \nAddressing liver disease in the UK: a blueprint for attaining excellence in health care \nand reducing premature mortality from lifesty le issues of excess consumption of \nalcohol, obesity, and viral hepatitis. The Lancet. 2014;384(9958):1953–97.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \n19. Williams R, Aithal G, Alexander GJ, Allison M, Armstrong I, Aspinall R, et al. \nUnacceptable failures: the final report of the Lancet Commission into liver disease in \nthe UK. The Lancet. 2020;395(10219):226–39.  \n20. Singal AG, Kanwal F, Llovet JM. Global trends in hepatocellular carcinoma \nepidemiology: implications for screening, prevention and therapy. Nat Rev Clin Oncol. \n2023;20(12):864-884. \n21. Harris HE, Costella A, Croxford S. Hepatitis C in the UK, 2020: Working to Eliminate \nHepatitis C as a Major Public Health Threat. 2020.  \n22. Huang DQ, Mathurin P, Cortez-Pinto H, Loomba R. Global epidemiology of alcohol-\nassociated cirrhosis and HCC: trends, projections and risk factors. Nat Rev \nGastroenterol Hepatol. 2023;20(1):37-49.  \n23. Karlsen TH, Sheron N, Zelber-Sagi S, Carrieri P, Dusheiko G, Bugianesi E, et al. The \nEASL–Lancet Liver Commission: protecting the next generation of Europeans against \nliver disease complications and premature mortality. The Lancet. \n2022;399(10319):61–116.  \n24. NHS Digital, Health Survey for England, 2021 part 1. https://digital.nhs.uk/data-and-\ninformation/publications/statistical/health-survey-for-england/2021/overweight-and-\nobesity-in-adults. Accessed 11 January 2024.  \n25. Orci LA, Sanduzzi-Zamparelli M, Caballol B, Sapena V, Colucci N, Torres F, et al. \nIncidence of Hepatocellular Carcinoma in Patients With Nonalcoholic Fatty Liver \nDisease: A Systematic Review, Meta-analysis, and Meta-regression. Clin \nGastroenterol Hepatol. 2022;20(2):283-292.e10.  \n26. Burton A, Balachandrakumar VK, Driver RJ, Tataru D, Paley L, Marshall A, et al. \nRegional variations in hepatocellular ca rcinoma incidence, routes to diagnosis, \ntreatment and survival in England. Br J Cancer. 2022;126(5):804-814.  \n27. McGlynn KA, Petrick JL, El-Serag HB. Epidemiology of Hepatocellular Carcinoma. \nHepatology. 2021;73 Suppl 1(Suppl 1):4–13.  \n28. Petrick JL, Florio AA, Znaor A, Ruggieri D, Laversanne M, Alvarez CS, et al. \nInternational trends in hepatocellular ca rcinoma incidence, 1978-2012. Int J Cancer. \n2020;147(2):317–30.  \n29. Geh D, Watson R, Sen G, French JJ, Hammond J, Turner P, et al. COVID-19 and \nliver cancer: lost patients and larger  tumours. BMJ Open Gastroenterol. \n2022;9(1):e000794.  \n30. Hepatitis C in the UK 2023. Working to eliminate hepatitis C as a public health threat. \nData to end of 2021. \nhttps://assets.publishing.service.gov.uk/media/63da9c888fa8f5187bafd5e5/hepatitis-\nc-in-the-UK-2023.pdf.  \n31. Arnold M, Rutherford MJ, Bardot A, Ferlay J, Andersson TML, Myklebust TÅ, et al. \nProgress in cancer survival, mortality, and incidence in seven high-income countries \n1995–2014 (ICBP SURVMARK-2): a population-based study. Lancet Oncol. \n2019;20(11):1493–505.  \n32. Allemani C, Weir HK, Carreira H, Harewood R, Spika D, Wang XS, et al. Global \nsurveillance of cancer survival 1995–2009: analysis of individual data for \n25/i1676/i1887 patients from 279 population-based registries in 67 countries \n(CONCORD-2). The Lancet. 2015;385(9972):977–1010.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \n33. Liver Cancer UK. Accessed 14 May 2024. https://livercanceruk.org/liver-cancer-\ninformation/statistics/#latediagnosis.  \n34. NHS England. https://www.england.nhs.uk/2023/10/nhs-mobile-testing-scheme-finds-\nthousands-of-new-cases-of-liver-damage/ Accessed 05 May 2024.  \n35. Liou WL, Tan TJ, Chen K, Goh GB, Chang JP , Tan CK. Gender survival differences in \nhepatocellular carcinoma: Is it all due to adherence to surveillance? A study of 1716 \npatients over three decades. JGH Open. 2023;24;7(5):377-386.  \n36. Cancer Research UK. Cancer Statistics Data Hub. \nhttps://crukcancerintelligence.shinyapps.io/CancerStatsDataHub/ Accessed 05 May \n2024.  \n37. Liou W, Tan TJ, Chen K, Goh GB, Chang JP, Tan C. Gender survival differences in \nhepatocellular carcinoma: Is it all due to adherence to surveillance? A study of 1716 \npatients over three decades. JGH Open. 2023;7(5):377–86.  \n38. Phipps M, Livanos A, Guo A, Pomenti S, Yeh J, Dakhoul L, et al. Gender Matters: \nCharacteristics of Hepatocellular Carcinoma in Women From a Large, Multicenter \nStudy in the United States. American Journal of Gastroenterology. 2020;115(9):1486–\n95.  \n39. Rich NE, Murphy CC, Yopp AC, Tiro J, Marrero JA, Singal AG. Sex disparities in \npresentation and prognosis of 1110 patients wi th hepatocellular carcinoma. Aliment \nPharmacol Ther. 2020;52(4):701–9.  \n40. Szabo G. Women and alcoholic liver disease - warning of a silent danger. Nat Rev \nGastroenterol Hepatol. 2018;15(5):253-254.  \n41. Yin M, Ikejima K, Wheeler MD, Bradford BU, Seabra V, Forman DT, et al. Estrogen is \ninvolved in early alcohol-induced liver injury in a rat enteral feeding model. \nHepatology. 2000;31(1):117-23.  \n42. Bizzaro D, Becchetti C, Trapani S, Lave zzo B, Zanetto A, D’Arcangelo F, et al. \nInfluence of sex in alcohol\n‐ related liver disease: Pre ‐ clinical and clinical settings. \nUnited European Gastroenterol J. 2023;11(2):218–27.  \n43. Jophlin LL, Singal AK, Bataller R, Wong RJ, Sauer BG, Terrault NA, et al. ACG \nClinical Guideline: Alcohol-Associated Liver Disease. Am J Gastroenterol. \n2024;119(1):30–54.  \n44. Macias RIR, Monte MJ, Serrano MA, González-Santiago JM, Martín-Arribas I, Simão \nAL, et al. Impact of aging on primary liver  cancer: epidemiology, pathogenesis and \ntherapeutics. Aging (Albany NY). 2021;11;13(19):23416–34.  \n45. Dyson J, Jaques B, Chattopadyhay D, Lochan R, Graham J, Das D, et al. \nHepatocellular cancer: The impact of obesity,  type 2 diabetes and a multidisciplinary \nteam. J Hepatol. 2014 Jan;60(1):110–7.  \n46. Hung AK, Guy J. Hepatocellular carcinoma in the elderly: Meta-analysis and \nsystematic literature review. World J Gastroenterol. 2015;21(42):12197.  \n47. Ding J, Wen Z. Survival improvement and prognosis for hepatocellular carcinoma: \nanalysis of the SEER database. BMC Cancer. 2021;21(1):1157.  \n48. Choi DT, Kum HC, Park S, Ohsfeldt RL, Shen Y, Parikh ND, et al. Hepatocellular \nCarcinoma Screening Is Associated With Increased Survival of Patients With \nCirrhosis. Clinical Gastroenterology and Hepatology. 2019;17(5):976-987.e4.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \n49. Sarfati D, Blakely T, Pearce N. Measur ing cancer survival in populations: relative \nsurvival vs cancer-specific survival. Int J Epidemiol. 2010;39(2):598-610.  \n50. Gallagher AM, Dedman D, Padmanabhan S, Leufkens HGM, Vries F. The accuracy of \ndate of death recording in the Clinical Practice Research Datalink GOLD database in \nEngland compared with the Office for National Statistics death registrations. \nPharmacoepidemiol Drug Saf. 2019;28(5):563-569.  \n51. Strongman H, Williams R, Bhaskaran K. What are the implications of using individual \nand combined sources of routinely collected data to identify and characterise incident \nsite-specific cancers? a concordance and validation study using linked English \nelectronic health records data. BMJ Open. 2020;20;10(8):e037719.  \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint \n\n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 5, 2024. ; https://doi.org/10.1101/2024.08.05.24311466doi: medRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}