Northern Ireland Cancer Registry Research Articles

25 year trends in cancer incidence and mortality among adults aged 35-69 years in the UK, 1993-2018: retrospective secondary analysis

ObjectiveTo examine and interpret trends in UK cancer incidence and mortality for all cancers combined and for the most common cancer sites in adults aged 35-69 years.DesignRetrospective secondary data analysis.Data...

Use of radiotherapy in patients with oesophageal, stomach, colon, rectal, liver, pancreatic, lung, and ovarian cancer: an International Cancer Benchmarking Partnership (ICBP) population-based study

There is little evidence on variation in radiotherapy use in different countries, although it is a key treatment modality for some patients with cancer. Here we aimed to examine such variation. This population-based study used data from Norway, the four UK nations (England, Northern Ireland, Scotland, and Wales), nine Canadian provinces (Alberta, British Columbia, Manitoba, New Brunswick, Newfoundland and Labrador, Nova Scotia, Ontario, Prince Edward Island, and Saskatchewan), and two Australian states (New South Wales and Victoria). Patients aged 15-99 years diagnosed with cancer in eight different sites (oesophageal, stomach, colon, rectal, liver, pancreatic, lung, or ovarian cancer), with no other primary cancer diagnosis occurring within the 5 years before to 1 year after the index cancer diagnosis or during the study period were included in the study. We examined variation in radiotherapy use from 31 days before to 365 days after diagnosis and time to its initiation, alongside related variation in patient group differences. Information was obtained from cancer registry records linked to clinical or patient management system data, or hospital administration data. Random-effects meta-analyses quantified interjurisdictional variation using 95% prediction intervals (95% PIs). Between Jan 1, 2012, and Dec 31, 2017, of 902 312 patients with a new diagnosis of one of the studied cancers, 115 357 (12·8%) did not meet inclusion criteria, and 786,955 were included in the analysis. There was large interjurisdictional variation in radiotherapy use, with wide 95% PIs: 17·8 to 82·4 (pooled estimate 50·2%) for oesophageal cancer, 35·5 to 55·2 (45·2%) for rectal cancer, 28·6 to 54·0 (40·6%) for lung cancer, and 4·6 to 53·6 (19·0%) for stomach cancer. For patients with stage 2-3 rectal cancer, interjurisdictional variation was greater than that for all patients with rectal cancer (95% PI 37·0 to 84·6; pooled estimate 64·2%). Radiotherapy use was infrequent but variable in patients with pancreatic (95% PI 1·7 to 16·5%), liver (1·8 to 11·2%), colon (1·6 to 5·0%), and ovarian (0·8 to 7·6%) cancer. Patients aged 85-99 years had three-times lower odds of radiotherapy use than those aged 65-74 years, with substantial interjurisdictional variation in this age difference (odds ratio [OR] 0·38; 95% PI 0·20-0·73). Women had slightly lower odds of radiotherapy use than men (OR 0·88, 95% PI 0·77-1·01). There was large variation in median time to first radiotherapy (from diagnosis date) by cancer site, with substantial interjurisdictional variation (eg, oesophageal 95% PI 11·3 days to 112·8 days; pooled estimate 62·0 days; rectal 95% PI 34·7 days to 77·3 days; pooled estimate 56·0 days). Older patients had shorter median time to radiotherapy with appreciable interjurisdictional variation (-9·5 days in patients aged 85-99 years vs 65-74 years, 95% PI -26·4 to 7·4). Large interjurisdictional variation in both use and time to radiotherapy initiation were observed, alongside large and variable age differences. To guide efforts to improve patient outcomes, underlying reasons for these differences need to be established. International Cancer Benchmarking Partnership (funded by the Canadian Partnership Against Cancer, Cancer Council Victoria, Cancer Institute New South Wales, Cancer Research UK, Danish Cancer Society, National Cancer Registry Ireland, The Cancer Society of New Zealand, National Health Service England, Norwegian Cancer Society, Public Health Agency Northern Ireland on behalf of the Northern Ireland Cancer Registry, DG Health and Social Care Scottish Government, Western Australia Department of Health, and Public Health Wales NHS Trust).

Use of chemotherapy in patients with oesophageal, stomach, colon, rectal, liver, pancreatic, lung, and ovarian cancer: an International Cancer Benchmarking Partnership (ICBP) population-based study

There are few data on international variation in chemotherapy use, despite it being a key treatment type for some patients with cancer. Here, we aimed to examine the presence and size of such variation. This population-based study used data from Norway, the four UK nations (England, Northern Ireland, Scotland, and Wales), eight Canadian provinces (Alberta, British Columbia, Manitoba, Newfoundland and Labrador, Nova Scotia, Ontario, Prince Edward Island, and Saskatchewan), and two Australian states (New South Wales and Victoria). Patients aged 15-99 years diagnosed with cancer in eight different sites (oesophageal, stomach, colon, rectal, liver, pancreatic, lung, or ovarian cancer), with no other primary cancer diagnosis occurring from within the 5 years before to 1 year after the index cancer diagnosis or during the study period were included in the study. We examined variation in chemotherapy use from 31 days before to 365 days after diagnosis and time to its initiation, alongside related variation in patient group differences. Information was obtained from cancer registry records linked to clinical or patient management system data or hospital administration data. Random-effects meta-analyses quantified interjurisdictional variation using 95% prediction intervals (95% PIs). Between Jan 1, 2012, and Dec 31, 2017, of 893 461 patients with a new diagnosis of one of the studied cancers, 111 569 (12·5%) did not meet the inclusion criteria, and 781 892 were included in the analysis. There was large interjurisdictional variation in chemotherapy use for all studied cancers, with wide 95% PIs: 47·5 to 81·2 (pooled estimate 66·4%) for ovarian cancer, 34·9 to 59·8 (47·2%) for oesophageal cancer, 22·3 to 62·3 (40·8%) for rectal cancer, 25·7 to 55·5 (39·6%) for stomach cancer, 17·2 to 56·3 (34·1%) for pancreatic cancer, 17·9 to 49·0 (31·4%) for lung cancer, 18·6 to 43·8 (29·7%) for colon cancer, and 3·5 to 50·7 (16·1%) for liver cancer. For patients with stage 3 colon cancer, the interjurisdictional variation was greater than that for all patients with colon cancer (95% PI 38·5 to 78·4; 60·1%). Patients aged 85-99 years had 20-times lower odds of chemotherapy use than those aged 65-74 years, with very large interjurisdictional variation in this age difference (odds ratio 0·05; 95% PI 0·01 to 0·19). There was large variation in median time to first chemotherapy (from diagnosis date) by cancer site, with substantial interjurisdictional variation, particularly for rectal cancer (95% PI -15·5 to 193·9 days; pooled estimate 89·2 days). Patients aged 85-99 years had slightly shorter median time to first chemotherapy compared with those aged 65-74 years, consistently between jurisdictions (-3·7 days, 95% PI -7·6 to 0·1). Large variation in use and time to chemotherapy initiation were observed between the participating jurisdictions, alongside large and variable age group differences in chemotherapy use. To guide efforts to improve patient outcomes, the underlying reasons for these patterns need to be established. International Cancer Benchmarking Partnership (funded by the Canadian Partnership Against Cancer, Cancer Council Victoria, Cancer Institute New South Wales, Cancer Research UK, Danish Cancer Society, National Cancer Registry Ireland, The Cancer Society of New Zealand, National Health Service England, Norwegian Cancer Society, Public Health Agency Northern Ireland on behalf of the Northern Ireland Cancer Registry, DG Health and Social Care Scottish Government, Western Australia Department of Health, and Public Health Wales NHS Trust).

Impact of COVID-19 on cancer incidence, presentation, diagnosis, treatment and survival in Northern Ireland.

The COVID-19 pandemic had a major impact on cancer patients and services but has been difficult to quantify. We examined how the entire cancer pathway-from incidence, presentation, diagnosis, stage, treatment and survival-was affected in Northern Ireland during April-December 2020 compared to equivalent 2018-2019 periods using retrospective, observational cancer registry data from the Northern Ireland Cancer Registry (NICR). There were 6748 cancer cases in April-December 2020 and an average 7724 patients in April-December 2018-2019. Incident cases decreased by 13% (almost 1000). Significant differences were found across age cohorts and deprivation quintiles, with reductions greatest for younger people (<55 years; 19% decrease) and less deprived (22% decrease). A higher proportion had emergency admission (16%-to-20%) with lower proportions diagnosed pathologically (85%-to-83%). There was a significant stage shift, with lower proportions of early stage (29%-to-25%) and higher late-stage (21%-to-23%). Lower proportions received surgery (41%-to-38%) and radiotherapy (24%-to-22%) with a higher proportion not receiving treatment (29%-to-33%). One-year observed-survival decreased from 73.7% to 69.8% and 1-year net-survival decreased from 76.1% to 72.9%, with differences driven by five tumours; Lung (40.3%-to-35.0%), Head-and-Neck (77.4%-to-68.4%), Oesophageal (53.5%-to-42.3%), Lymphoma (81.1%-to-75.2%) and Uterine cancer (87.4%-to-80.4%). Our study reveals profound adverse impact of COVID-19 on the entire cancer patient pathway, with 13% fewer cases, greater emergency admissions and significant stage-shift from early to more advanced-stage disease. There was major treatment impact with lower rates of surgery and radiotherapy and higher proportions receiving no treatment. There were significant reductions in 1-year survival. Our study will support service recovery and protect cancer services in future pandemics or disruptions.

H11 More cancer, better counting or better at diagnosis? The history of skin cancer registration

Abstract Cancer registries help us to understand skin cancer cause, prevalence, treatment and prevention. They provide evidence to gain new healthcare resources. This paper studies the evolution of cancer registries and their relevance to dermatology. In the 16th century, ‘Bills of Mortality’, the first English healthcare registry, catalogued the number and causes of death in London parishes. Originally set up to track bubonic plague, it expanded to cover other diseases, including cancer. The 1728 London ‘General Census of Cancer’ was an early systemic study of cancer. In the 19th and early 20th centuries, hospital censuses often monitored cancers. These were superseded by population-based cancer registries, allowing continuous data collection. An early example is the Hamburg Cancer Registry (1926), which recorded names on a card-based registry and compared them with death certificates. The first national cancer registry was the Danish national cancer registry, formed in 1942. In the UK, the Scottish Cancer Registry was formed in 1958, the Welsh Cancer Intelligence &amp; Surveillance Unit in 1972 and the Northern Ireland Cancer Registry in 1994. In England, eight regional registries were created, with the South West being the first in 1945. The Calman–Hine report (1995) highlighted a desire for a national cancer registry. In 2008, the National Cancer Intelligence Network enabled the assimilation and analysis of data sets from English registries. They merged in 2013, forming the National Cancer Registry and Analysis Service (NCRAS), England. Critically, the 2006 National Health Service act (section 251) allowed cancer data collection without formal consent. Although melanoma data were generally good quality, the inclusion and quality of nonmelanoma skin cancers (NMSCs) were historically less reliable, with improvement since 2013. Historically, the exclusion of NMSCs from published cancer data is common. Wales is the last UK registry to not routinely report NMSC. However, this is due to change in 2024. The ‘global burden of disease tool’ (1990) utilizes various sources, including cancer registry data, to model incidence, mortality and health implications of disease globally, including melanoma and NMSCs. In 2022, the British Association of Dermatologists joined with the NCRAS to develop the Get Data Out skin tool, an annually updated website featuring more granular epidemiological data on skin cancers. High-quality epidemiological data are essential to understanding skin cancer’s regional and chronological variations, and to ensure that resources for healthcare and research are allocated appropriately.

Incidence and survival of uveal melanoma in Northern Ireland: how incomplete data can skew results in rare cancers.

The majority of Northern Irish uveal melanoma (UM) patients are diagnosed in Sheffield. This study aims to present incidence and survival outcomes for UM patients from Northern Ireland (NI). Collaborative retrospective study between Sheffield and Northern Ireland Cancer Registry (NICR). For UM cases not on both databases, outcomes and survival rates (via Kaplan-Meier analysis) were compared. Anonymised NICR data were used to calculate whole-population incidence of UM for NI. In total, 161 patients from NI were diagnosed in Sheffield, 90 of which were not registered with NICR at the start of this study. Data-omissions were not consistent across patient groups, leading to significant differences between those patients registered and those not. Registered patients had an all-cause 5-year survival rate of only 68.9% compared to 92.5% of those not registered (p < 0.01) and were >17x more likely to have systemic metastases than those not registered (p < 0·001). Following rectification of data-omissions, the European age-standardised incidence rate of UM for NI was 8·6 per million. This study illustrates the impact of incomplete population-wide data, serving as a real-world lesson in case-identification bias. Rare cancers are at higher risk of omission due to systemic failures as the small numbers involved are not detected by system-wide validation procedures. Following this study, data-transfer agreements between England and NI were actioned, preventing future data-omissions. We present survival and incidence data for UM in NI for the first time, showing the incidence is amongst the highest in Europe, with good survival rates.

Impact of the COVID-19 pandemic on emergency hospital cancer admissions in a UK region

BackgroundThe pandemic disrupted society and health services through lockdowns and resource reallocation to care for COVID-19 patients. Reductions in numbers of cancer patients having surgery, being diagnosed pathologically or via 2-week wait, and screening programs pauses have been described. The effect on emergency presentation, which represents an acute episode with poor outcomes, has not been investigated. This study explored the pandemic’s impact on emergency hospital admissions for cancer patients in a UK region.MethodsHospital discharge data for cancer patients in Northern Ireland, which included route to admission, were analysed for the pandemic era in 2020 compared to averages for March to December 2017–2019, focusing on volume and route of emergency admissions by demography and tumour site.FindingsCompared with the pre-pandemic era, the number of cancer emergency admissions fell by 12·3% in 2020. Emergency admissions for cancer were significantly reduced when COVID-19 levels were highest (− 18·5% in April and − 16.8% in October). Females (− 15·8%), urban residents (− 13·2%), and age groups 0 to 49 and 65–74 years old (− 17%) experienced the largest decreases as did those with haematological (− 14·7%), brain and CNS (− 27·9%), and lung cancers(− 14·3%). Significant reductions in referrals from outpatient departments (− 51%) and primary care (− 43%) (p < 0·001) were counterbalanced by admissions from other routes including confirmed or suspected COVID-19 infection (increase 83·6%).InterpretationReductions in emergency admissions, and pathologically diagnosed cancers, as reported by the Northern Ireland Cancer Registry (NICR), indicate undiagnosed patients in the community which has implications for future workloads and survival. Data suggest undiagnosed cases may be higher for haematological, brain and CNS, and lung cancers and among females. Efforts should be made to encourage people with symptoms to present for diagnosis or reassurance.FundingThe NICR is funded by the Public Health Agency of Northern Ireland. This work was supported by Macmillan Cancer Support and uses data collected by health services as part of their care and support functions.

Survival of cancer patients with pre-existing heart disease

BackgroundWhile cancer outcomes have improved over time, in Northern Ireland they continue to lag behind those of many other developed economies. The role of comorbid conditions has been suggested as a potential contributory factor in this but issues of data comparability across jurisdictions has inhibited efforts to explore relationships. We use data from a single jurisdiction of the UK using data from - the Northern Ireland Cancer Registry (NICR), to examine the association between mortality (all-cause and cancer specific) and pre-existing cardiovascular diseases among patients with cancer.Materials and MethodsAll patients diagnosed with cancer (excluding non-melanoma skin cancer) between 2011 and 2014 were identified from Registry records. Those with a pre-existing diagnosis of cardiovascular diseases were identified by record linkage with patient hospital discharge data using ICD10 codes. Survival following diagnosis was examined using descriptive statistics and Cox proportional hazards regression analyses. Analyses examined all-cause mortality and cancer specific mortality for lung, colorectal, breast and prostate cancer. As well as cardiovascular diseases, regression models controlled for age, gender (where appropriate), deprivation (as quintiles), stage at diagnosis and other comorbidities.ResultsAlmost 35,000 incident cancer cases were diagnosed during the study period of which approximately 23% had a prior heart condition. The pan-cancer hazard ratio for death in the presence of pre-existing cardiovascular diseases was 1.28 (95% CI: 1.18-1.40). All-cause and cancer specific mortality was higher for patients with cardiovascular diseases across lung, female breast, prostate and colorectal cancer groups after controlling for age, gender (where appropriate), deprivation (as quintiles), stage at diagnosis and other comorbidities.ConclusionPre-existing morbidity may restrict the treatment of cancer for many patients. In this cohort, cancer patients with pre-existing cardiovascular diseases had poorer outcomes than those without cardiovascular diseases. A high prevalence of cardiovascular diseases may contribute to poorer cancer outcomes at a national level.

Risk factors and prognostic implications of diagnosis of cancer within 30 days after an emergency hospital admission (emergency presentation): an International Cancer Benchmarking Partnership (ICBP) population-based study.

SummaryBackgroundGreater understanding of international cancer survival differences is needed. We aimed to identify predictors and consequences of cancer diagnosis through emergency presentation in different international jurisdictions in six high-income countries.MethodsUsing a federated analysis model, in this cross-sectional population-based study, we analysed cancer registration and linked hospital admissions data from 14 jurisdictions in six countries (Australia, Canada, Denmark, New Zealand, Norway, and the UK), including patients with primary diagnosis of invasive oesophageal, stomach, colon, rectal, liver, pancreatic, lung, or ovarian cancer during study periods from Jan 1, 2012, to Dec 31, 2017. Data were collected on cancer site, age group, sex, year of diagnosis, and stage at diagnosis. Emergency presentation was defined as diagnosis of cancer within 30 days after an emergency hospital admission. Using logistic regression, we examined variables associated with emergency presentation and associations between emergency presentation and short-term mortality. We meta-analysed estimates across jurisdictions and explored jurisdiction-level associations between cancer survival and the percentage of patients diagnosed as emergencies.FindingsIn 857 068 patients across 14 jurisdictions, considering all of the eight cancer sites together, the percentage of diagnoses through emergency presentation ranged from 24·0% (9165 of 38 212 patients) to 42·5% (12 238 of 28 794 patients). There was consistently large variation in the percentage of emergency presentations by cancer site across jurisdictions. Pancreatic cancer diagnoses had the highest percentage of emergency presentations on average overall (46·1% [30 972 of 67 173 patients]), with the jurisdictional range being 34·1% (1083 of 3172 patients) to 60·4% (1317 of 2182 patients). Rectal cancer had the lowest percentage of emergency presentations on average overall (12·1% [10 051 of 83 325 patients]), with a jurisdictional range of 9·1% (403 of 4438 patients) to 19·8% (643 of 3247 patients). Across the jurisdictions, older age (ie, 75–84 years and 85 years or older, compared with younger patients) and advanced stage at diagnosis compared with non-advanced stage were consistently associated with increased emergency presentation risk, with the percentage of emergency presentations being highest in the oldest age group (85 years or older) for 110 (98%) of 112 jurisdiction-cancer site strata, and in the most advanced (distant spread) stage category for 98 (97%) of 101 jurisdiction-cancer site strata with available information. Across the jurisdictions, and despite heterogeneity in association size (I2=93%), emergency presenters consistently had substantially greater risk of 12-month mortality than non-emergency presenters (odds ratio >1·9 for 112 [100%] of 112 jurisdiction-cancer site strata, with the minimum lower bound of the related 95% CIs being 1·26). There were negative associations between jurisdiction-level percentage of emergency presentations and jurisdiction-level 1-year survival for colon, stomach, lung, liver, pancreatic, and ovarian cancer, with a 10% increase in percentage of emergency presentations in a jurisdiction being associated with a decrease in 1-year net survival of between 2·5% (95% CI 0·28–4·7) and 7·0% (1·2–13·0).InterpretationInternationally, notable proportions of patients with cancer are diagnosed through emergency presentation. Specific types of cancer, older age, and advanced stage at diagnosis are consistently associated with an increased risk of emergency presentation, which strongly predicts worse prognosis and probably contributes to international differences in cancer survival. Monitoring emergency presentations, and identifying and acting on contributing behavioural and health-care factors, is a global priority for cancer control.FundingCanadian Partnership Against Cancer; Cancer Council Victoria; Cancer Institute New South Wales; Cancer Research UK; Danish Cancer Society; National Cancer Registry Ireland; The Cancer Society of New Zealand; National Health Service England; Norwegian Cancer Society; Public Health Agency Northern Ireland, on behalf of the Northern Ireland Cancer Registry; the Scottish Government; Western Australia Department of Health; and Wales Cancer Network.

P-OGC38 The Impact of the COVID-19 Pandemic on Barrett’s Oesophagus and Oesophago-gastric Cancer

BackgroundThe COVID-19 pandemic has placed an inexorable strain on endoscopy services worldwide, affecting the diagnosis of oesophago-gastric (OG) cancer and Barrett’s oesophagus (BO). As coronavirus infection rates rose many professional bodies advised that all endoscopy, except emergency and essential procedures be stopped immediately. We sought to quantify the decline in OG cancer and BO diagnoses following implementation of British Society of Gastroenterology (BSG) guidance related to COVID-19 and the psychosocial effects on BO patients.MethodsWe examined OG cancer and BO diagnoses in Northern Ireland from March-September 2020 and compared them with the three-year average number of patients during the same time period (corresponding to weeks 10-37) between 2017-2019 by utilising Northern Ireland Cancer Registry (NICR) data. The psychosocial impact of COVID-19 was assessed using an online survey, which included validated WHOQOL-BREF and EQ-5D-5L quality of life measures, and was completed by 24 BO patients from April-May 2020.ResultsBetween March and September 2020 in Northern Ireland, the proportion of OG cancer and BO diagnoses declined by 26.6% and 59.3%, respectively, compared to expected levels. In April, BO diagnoses fell by 95.5% but by September, whilst OG cancer rates had returned to baseline, BO cases remained supressed by approximately 20%. We estimate that these declines in diagnosis represent 53 ‘missed’ OG cancer and 236 ‘missed’ BO diagnoses. In the online survey sample, BO patients reported consistently lower quality of life scores than population norms, and highlighted a number of concerns with regard to their health and care.ConclusionsThe COVID-19 pandemic has resulted in an abrupt decline in OG cancer and BO diagnoses and has profoundly impacted the wellbeing of BO patients. Our study represents the first report of the impact of COVID-19 on the diagnosis of BO. Strategies to mitigate the ongoing effects of the pandemic are urgently required to preserve the ability to rapidly detect and diagnose cancer and pre-malignant conditions.

Abstract S02-03: The impact of the COVID-19 pandemic on Barrett’s esophagus and esophago-gastric cancer

Abstract Objective The COVID-19 pandemic has placed an inexorable strain on endoscopy services worldwide, affecting the diagnosis of esophago-gastric (EG) cancer and Barrett’s esophagus (BE). As coronavirus infection rates rose many professional bodies advised that all endoscopy, except emergency and essential procedures, be stopped immediately. We sought to quantify the decline in EG cancer and BE diagnoses following implementation of British Society of Gastroenterology (BSG) guidance related to COVID-19 and the psychosocial effects on BE patients. Methods We examined EG cancer and BE diagnoses in Northern Ireland from March-September 2020 and compared them with the three-year average number of patients during the same time period between 2017-2019 by utilizing Northern Ireland Cancer Registry (NICR) data. The psychosocial impact of COVID-19 was assessed using an online survey, which included validated WHOQOL-BREF and EQ-5D-5L quality of life measures, and was completed by 24 BE patients from April-May 2020. Results During the first six months of the pandemic the proportion of EG cancer and BE diagnoses declined by 26.6% and 59.3%, respectively, compared to expected levels. In April, BE diagnoses fell by 95.5% but by September, whilst EG cancer rates had returned to baseline, BE cases remained suppressed by approximately 20%. We estimate that these declines in diagnosis represent 53 ‘missed’ EG cancer and 236 ‘missed’ BE diagnoses. In the online survey sample, BE patients reported consistently lower quality of life scores than population norms, and highlighted a number of concerns with regard to their health and care. Conclusion The COVID-19 pandemic has resulted in an abrupt decline in EG cancer and BE diagnoses and has profoundly impacted the wellbeing of BE patients. Our study represents the first report of the impact of COVID-19 on the diagnosis of BE. Strategies to mitigate the ongoing effects of the pandemic are urgently required to preserve the ability to rapidly detect and diagnose cancer and pre-malignant conditions. Citation Format: Richard C. Turkington, Anita Lavery, David Donnelly, Anna T. Gavin, Damian T. McManus, Victoria Cairnduff, Charlene M. McShane, Brian T. Johnston, Lesley A. Anderson, Helen G. Coleman. The impact of the COVID-19 pandemic on Barrett’s esophagus and esophago-gastric cancer [abstract]. In: Proceedings of the AACR Virtual Meeting: COVID-19 and Cancer; 2021 Feb 3-5. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(6_Suppl):Abstract nr S02-03.

The Impact of the COVID-19 Pandemic on Barrett’s Esophagus and Esophagogastric Cancer

The Impact of the COVID-19 Pandemic on Barrett’s Esophagus and Esophagogastric Cancer

Esophageal Columnar Metaplasia in Childhood: A Population-Based Case Series Analysis

Adults with Barrett's esophagus (BE) are often entered into surveillance for esophageal adenocarcinoma (EAC), although cancer risk is relatively low. BE can be detected in children (< 16years). Little is known about the epidemiology of pediatric BE, and it is unclear what the optimal surveillance regimes are in children. To evaluate the demographic and clinical characteristics, and future neoplastic progression risk in all pediatric BE patients diagnosed in Northern Ireland between 1993 and 2010. Data from the population-based Northern Ireland BE register were matched to the Northern Ireland Cancer Registry for EAC outcomes until end 2013. Age-adjusted incidence of pediatric BE was calculated, and characteristics between pediatric and adult BE patients compared using Chi-square tests. Over 18years, 42 pediatric BE patients (< 16years) were identified, equivalent to an age-adjusted incidence of < 2 per 100,000 children. There was a clear age differential, with BE incidence increasing with age within the pediatric population. The majority (85.7%) of patients were male, a significantly higher male/female ratio than adult BE patients (p < 0.001). No pediatric BE patients progressed to high-grade dysplasia (HGD) or EAC, although the eldest patient was aged 34years by the end of follow-up. This is the largest series of pediatric BE ever reported. It demonstrates that pediatric BE is rare. The male preponderance of this condition is more apparent in childhood compared with adult cases. No children developed HGD/EAC during follow-up, suggesting that regular surveillance is not required, at least until adulthood.

Prognosis following surgical resection versus local excision of stage pT1 colorectal cancer: A population-based cohort study

AimsTo evaluate patient management following stage pT1 colorectal cancer (CRC) diagnosis, and to determine if surgical resection improved outcome compared with local excision, within a population-based study. MethodsData were collected from the Northern Ireland Cancer Registry. Cases of stage pT1 CRC diagnosed from 2007 to 2012 were identified. Analyses were conducted using Cox proportional hazard models to calculate hazard ratios (HR) and 95% confidence intervals (CI) for cancer-specific and all-cause mortality for individuals undergoing formal surgery versus local excision. Results394 patients with pT1 CRC were included. Of these, 37.1% were treated by local resection, 36.8% had biopsy followed by surgery and 26.1% had local excision followed by surgery. There were 60 deaths over a mean 4.8 years of follow-up, including 10 CRC-specific deaths. An additional 12 patients had a CRC recurrence or metastases during follow-up. Of the CRC-specific deaths or recurrences, 27.3% had local excision only. Individuals treated by formal surgery did not have a reduced risk of CRC-specific death (adjusted HR = 1.51, 95% CI 0.29, 7.89), but did have a reduced risk of all-cause mortality (adjusted HR = 0.51 95% CI 0.30, 0.87) compared with those undergoing local excision only. ConclusionsPatients with stage pT1 CRC undergoing formal surgery had a reduced risk of all-cause mortality compared with those treated by local excision only. However, this was not explained by a reduced risk of recurrence/disease-free survival or CRC death, and suggests that the observed benefits may simply reflect selection of a healthier patient population in the formal surgery group.

Abstract P1-09-10: Molecular and genetic characterisation of contralateral breast cancers in Northern Ireland: Opportunities for personalised surgery

Abstract Contralateral breast cancer (CBC) incidence is reported as 0.4-0.7% per year, equating to an approximate 10-15% 20 year risk after primary cancer diagnosis. Young age at primary diagnosis and a strong family history are consistently cited as significant risk factors for CBC development. Requests for contralateral prophylactic mastectomy (CPM) are rising, despite a lack of evidence that CPM significantly improves breast cancer specific survival. Therefore, a method to stratify CBC prognosis and risk at an individual level is required to allow us to personalise surgical decision-making for individual patients. This study aims to: 1. Explore prognostic factors following CBC diagnosis; 2. Investigate metastatic CBC prevalence by determining clonal relationships between primary and CBC tumours; 3. Assess the rate of germline predisposition gene mutations in a large cohort of Northern Irish CBC cases. A cohort of 403 CBC patients and 1:1 matched unilateral breast cancer controls (matched for age, year of diagnosis, nodal positivity and tumour morphology) were identified from the Northern Ireland Cancer Registry. Archival primary, CBC and normal lymph node tissue from these CBC patients were obtained for somatic and germline sequencing of a custom 94-gene panel, comprising genes most commonly mutated in breast cancer and all known predisposition genes. CBC patients had a significantly increased risk of breast cancer specific mortality when compared with 1:1 matched unilateral breast cancer controls (HR 6.45, CI 4.27-9.77). Within the CBC cohort, a shorter time interval between primary and CBC diagnosis is associated with increased breast cancer mortality, whereby women with &amp;lt;5 years between diagnoses have the poorest survival (p=0.001). In terms of receptor status, the PR status of ER positive primary and CBC tumours differed significantly (p&amp;lt;0.001). 89.1% of ER positive primary tumours were PR positive, whilst PR positivity was only 54.4% in ER positive CBCs. Indeed, 36.6% of patients with an ER+/PR+ primary BC developed an ER+/PR- CBC, after exposure to anti-oestrogen therapy. Targeted next generation sequencing of 42 cases to date has provided evidence of shared somatic mutations suggestive of metastatic clonality between primary and CBC in four cases (9.5%). Germline pathogenic predisposition gene mutations were identified in five (11.9%) patients and a further ten (23.8%) patients were noted to have germline predisposition variants of unknown significance. Breast cancer specific mortality risk was significantly higher in CBC patients compared with matched unilateral cancer controls, in excess of what would be expected from two primary cancer events. This finding suggests that there are as yet unquantified factors contributing to this excess hazard. Poorer outcome with shorter inter-tumour diagnostic intervals, in addition to evidence of primary-CBC clonality in some cases, suggests that metastatic CBC may play a role in excess risk. Additionally, the significant reduction in PR positivity of ER positive CBC tumours after ER+/PR+ primary tumours, could suggest that acquired hormonal resistance following primary anti-oestrogen therapy exposure may contribute to this additional mortality hazard. Citation Format: McIlmunn CG, Bannon FJ, Fitzpatrick D, Ervine A, Boyd C, Cameron I, Harita T, James J, Gonzalez de Castro D, Savage KI, McIntosh SA. Molecular and genetic characterisation of contralateral breast cancers in Northern Ireland: Opportunities for personalised surgery [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P1-09-10.

'Missed' oesophageal adenocarcinoma and high-grade dysplasia in Barrett's oesophagus patients: A large population-based study.

A systematic review suggests that 25% of oesophageal adenocarcinomas (OAC) are 'missed' at index endoscopy for Barrett's oesophagus (BO); however, this included few population-based studies and may be an overestimate. The objective of this article is to quantify the 'missed' rates of high-grade dysplasia (HGD) and OAC at index BO endoscopy. Patients from the Northern Ireland BO register diagnosed between 1993 and 2010 (n = 13,159) were linked to the Northern Ireland Cancer Registry to identify patients who developed OAC or HGD. Logistic regression analysis compared characteristics of 'missed' vs 'incident' HGD/OAC, defined as diagnoses within 3-12 months vs >1 year after incident BO, respectively. A total of 267 patients were diagnosed with HGD/OAC ≥3 months after BO diagnosis, of whom 34 (12.7%) were potentially 'missed'. The proportion of 'missed' HGD/OAC was 25% among BO patients with low-grade dysplasia (LGD) and 9% among non-dysplastic BO patients. Older age and BO-LGD carried a higher risk of 'missed' HGD/OAC. Non-dysplastic BO patients were more often diagnosed with a 'missed' OAC (rather than HGD; 89%), compared with BO-LGD patients (40%). Approximately one in 10 HGD/OAC cases are 'missed' at incident BO diagnosis, which is significant but lower than previous reports. However, 'missed' HGD/OAC cases represent only 0.26% of all BO patients.

A population-level investigation of cancer clinical trials participation in a UK region.

The aim of this study was to document cancer trial participation since establishment of the Northern Ireland Cancer Trials Network and investigate population and disease factors associated with trial participation. An independent cohort of over 51 000 cancer patients from the Northern Ireland Cancer Registry covering the same population (2007-2012) was linked to a database of 1316 interventional cancer trial participants in a UK region. The primary outcome measure was participation in an intervention clinical trial. Patients were followed up until 31 March 2013. Kaplan-Meier tests and Cox proportional hazard models using person days at risk to allow for death were used to investigate factors associated with trial participation. Multivariate analysis assessed the impact of age, cancer type and stage, distance from the cancer centre (radiotherapy), marital status, deprivation quintile and rurality. Participation was analysed separately for children (<15 years) and young individuals (15-24 years). Trial recruitment increased three-fold with establishment of a network. Participation was the highest for children at 21%, but relatively low at 2.05% for adults, although higher for haematological malignancies (4.5%). Lower likelihood of trial participation in adults was associated with female sex, older age, distance from regional Cancer Centre and stage 1 disease. The introduction of a regional Cancer Trials Network was associated with increased participation; however, trial participation remains relatively low at the population level especially among elderly patients. Linkage of clinical trials and cancer registry database provide an easy mechanism to monitor trial representativeness at the population level.

19 Barrett’s oesophagus in childhood: a population-based case series analysis

Background Adults detected with Barrett’s oesophagus (BO) are often entered into surveillance for oesophageal adenocarcinoma (OAC), although cancer risk is relatively low. More rarely, BO can be detected in children. Little is known about the epidemiology of paediatric BO, and it is unclear what the optimal surveillance regimes are in these children. Aim To evaluate the demographic and clinical characteristics, and future neoplastic progression risk in all paediatric BO patients diagnosed in Northern Ireland between 1993 and 2010. Method Data was collected using the Northern Ireland BO register and matched to the Northern Ireland Cancer Registry for OAC outcomes until end 2013. Age-adjusted incidence of paediatric BO was calculated, and chi-squared tests performed to compare characteristics between paediatric and adult BO patients. Results Over 18 years, 42 paediatric BO patients were identified, equivalent to an age-adjusted incidence of Conclusions This is the largest series of paediatric BO ever to be reported. It demonstrates that paediatric BO is a rare disease. The male preponderance of this condition is even more apparent in childhood compared with adult cases. No children developed HGD/OAC during follow-up, suggesting that regular surveillance is not required, at least until early adulthood.

Machine learning classification of surgical pathology reports and chunk recognition for information extraction noise reduction.

Machine learning techniques for the text mining of cancer-related clinical documents have not been sufficiently explored. Here some techniques are presented for the pre-processing of free-text breast cancer pathology reports, with the aim of facilitating the extraction of information relevant to cancer staging. The first technique was implemented using the freely available software RapidMiner to classify the reports according to their general layout: 'semi-structured' and 'unstructured'. The second technique was developed using the open source language engineering framework GATE and aimed at the prediction of chunks of the report text containing information pertaining to the cancer morphology, the tumour size, its hormone receptor status and the number of positive nodes. The classifiers were trained and tested respectively on sets of 635 and 163 manually classified or annotated reports, from the Northern Ireland Cancer Registry. The best result of 99.4% accuracy - which included only one semi-structured report predicted as unstructured - was produced by the layout classifier with the k nearest algorithm, using the binary term occurrence word vector type with stopword filter and pruning. For chunk recognition, the best results were found using the PAUM algorithm with the same parameters for all cases, except for the prediction of chunks containing cancer morphology. For semi-structured reports the performance ranged from 0.97 to 0.94 and from 0.92 to 0.83 in precision and recall, while for unstructured reports performance ranged from 0.91 to 0.64 and from 0.68 to 0.41 in precision and recall. Poor results were found when the classifier was trained on semi-structured reports but tested on unstructured. These results show that it is possible and beneficial to predict the layout of reports and that the accuracy of prediction of which segments of a report may contain certain information is sensitive to the report layout and the type of information sought.

Colorectal Cancer Risk Following Adenoma Removal: A Large Prospective Population-Based Cohort Study.

Randomized controlled trials have demonstrated significant reductions in colorectal cancer incidence and mortality associated with polypectomy. However, little is known about whether polypectomy is effective at reducing colorectal cancer risk in routine clinical practice. The aim of this investigation was to quantify colorectal cancer risk following polypectomy in a large prospective population-based cohort study. Patients with incident colorectal polyps between 2000 and 2005 in Northern Ireland were identified via electronic pathology reports received to the Northern Ireland Cancer Registry. Patients were matched to the Northern Ireland Cancer Registry to detect colorectal cancer and deaths up to December 31, 2010. Colorectal cancer standardized incidence ratios (SIR) were calculated and Cox proportional hazards modeling applied to determine colorectal cancer risk. During 44,724 person-years of follow-up, 193 colorectal cancer cases were diagnosed among 6,972 adenoma patients, representing an annual progression rate of 0.43%. Colorectal cancer risk was significantly elevated in patients who had an adenoma removed (SIR, 2.85; 95% CI, 2.61-3.25) compared with the general population. Male sex, older age, rectal site, and villous architecture were associated with an increased colorectal cancer risk in adenoma patients. Further analysis suggested that not having a full colonoscopy performed at, or following, incident polypectomy contributed to the excess colorectal cancer risk. Colorectal cancer risk was elevated in individuals following polypectomy for adenoma, outside of screening programs. This finding emphasizes the need for full colonoscopy and adenoma clearance, and appropriate surveillance, after endoscopic diagnosis of adenoma.