West Midlands Cancer Intelligence Unit Research Articles

Neuroendocrine Carcinoma of the Cervix: Review of a Series of Cases and Correlation With Outcome.

Introduction Neuroendocrine carcinoma (NEC) of the cervix is associated with a poor prognosis despite multimodal treatment. The correct diagnosis of this tumor type is imperative to provide clinicians and patients with prognostic information and ensure that appropriate treatment is provided. Methods A clinicopathological study was undertaken on all cervical tumors registered as NEC with the West Midlands Cancer Intelligence Unit between January 1, 1998 and December 31, 2009. Of the 45 cases diagnosed during the study period, the tumor samples of 41 cases were traced, anonymized, and then independently reviewed by 2 gynecological pathologists. Results The review confirmed 31/41 (78%) cases to be NEC, which overall, represented 1.3% of all the cervical cancers registered in the West Midlands over the period of the study. In the correct histological context, synaptophysin was the most sensitive and specific positive immunohistochemical marker of NEC differentiation. The cases that on review were confirmed as NEC had a significantly worse outcome than the non-NEC cases: median survival for NEC cases was 33.3 months versus 315.0 months for the non-NEC cases, P = .013. Conclusions Histological review of a series of NECs has shown significantly reduced survival in those patients with confirmed NEC in comparison with those patients where a diagnosis of NEC was not confirmed. We propose morphological and immunohistochemical criteria for the diagnosis of cervical NEC; and discourage unqualified use of the term "small cell carcinoma" as this does not accurately convey the diagnosis of SCNEC. We urge pathologists to use the 2014 World Health Organization classification when reporting these tumors.

PTH-016 Factors Associated with Upper Gastrointestinal Cancer Occurrence After OGD that Did Not Detect Cancer in The West Midlands

Introduction Up to 14% of upper gastrointestinal cancer (UGIC) subjects had an OGD that did not diagnose cancer up to 3 years prior to diagnosis. The rate of and associated risk factors for post-OGD upper gastrointestinal cancer (POUGIC) in the West Midlands were examined. Methods Computerised OGD records from 8 NHS trusts in the West Midlands between 1998 and 2010 were retrieved and submitted to the West Midlands Cancer Intelligence Unit (WMCIU) for UGIC registrations linkage. Subjects undergoing OGD 3 to 36 months before diagnosis were identified as POUGIC and subjects with no OGD 3 to 36 months prior to diagnosis served as controls. The influence of age, gender, indication, endoscopist specialty, trainee involvement, sedation, number of biopsies taken from focal abnormalities, site and histology of UGIC on POUGIC were examined by logistic regression analysis. Results 115,113 OGD records were submitted to WMCIU and 3870 UGIC were linked. After exclusions, 2909 UGIC subjects were analysed. There were 275 (9.5%) POUGIC subjects (154 oesophageal cancer (OC) and 121 gastric cancer (GC)). The POUGIC rate ranged from 7.6 to 11.8% between the trusts. Of the POUGIC subjects, 143 (52.0%) had OGD 3 to 12 months and 132 (48.0%) 12 to 36 months prior to UGIC diagnosis. POUGIC subjects were younger (69.6±11.7 yrs) compared with controls (72.0±11.6 yrs) (p = 0.001). There were no association between POUGIC and OC or GC (GC OR 1.15, 95%CI 0.89–1.48, p = 0.289) or subject gender (female 1.14, 0.88–1.48, p = 0.312). Subjects with alarm symptoms (0.33, 0.26–0.43, p Conclusion The POUGIC rate in the West Midlands was 9.5% between 1998 and 2010 and varied between 7.6 and 11.8% between trusts. POUGIC was associated with younger age, a lack of alarm symptoms and less biopsies from abnormalities. There was no significant survival difference between POUGIC subjects and controls at 1 yr. Disclosure of Interest None Declared

Contralateral breast cancer: incidence according to ductal or lobular phenotype of the primary

To identify differences in the incidence of contralateral breast cancer between patients with a primary tumour diagnosis of invasive ductal carcinoma (IDC) and those with a diagnosis of invasive lobular carcinoma (ILC). Data from two large cancer registries (registry A & B) the Northern and Yorkshire Cancer Registry Information Service (NYCRIS) and the West Midlands Cancer Intelligence Unit (WMCIU) from 1998-2003 for all cases of invasive breast cancer of either pure ductal or pure lobular reported histology were obtained. The invasive status of the contralateral tumour diagnosis and tumour morphology was collected. Chi-square tests were undertaken to examine the differences in contralateral rates for both registries and univariate analysis to ascertain which predictors affected contralateral breast cancer risk for registry A the WMCIU cases. A total of 38,132 patients were studied, 32,735 patients with IDC and 5397 (14.2%) patients with ILC over the 6-year period. There was no significant difference between the occurrence and time to occurrence of contralateral breast cancer according to original cancer histology, 901 (2.8%) patients with IDC versus 166 (3.1%) patients with ILC (p=0.169). The analysis of registry A cases showed no association between original histology (ductal versus lobular), age at diagnosis, tumour grade, use of radiotherapy for the primary cancer or use of systemic therapy (chemotherapy and/or endocrine therapy), and development of a contralateral breast cancer. There is no apparent increase in risk of developing a contralateral breast cancer according to the primary cancer histology either IDC or ILC. Standard mammographic follow-up does not need to take account of original tumour pathology. Increased intervention or post-treatment surveillance for the contralateral breast is not indicated in the context of ILC. The role of MRI should be restricted to those patients with ILC who are planning breast-conservation surgery, but not for assessing the contralateral breast.

P02 * PRIMARY CENTRAL NERVOUS SYSTEM LYMPHOMA (PCNSL) IN THE CEREBELLUM: REPORT OF TWO RARE CASES AND A TEN YEAR MULTI-CENTRE REVIEW OF PCNSL

INTRODUCTION: Primary central nervous system lymphoma (PCNSL) is a malignant tumour. Its incidence is 0.46-0.51/100000/year. Cerebellar PCNSL is rare. Previously documented cases included only snapshots of the radiological appearances in large PCNSL series; therefore the radiological progression of cerebellar PCNSL has not been described. We report 2 cases of solitary cerebellar lymphoma, followed by a 10-year review of PCNSLs in our region. METHOD: Report of 2 rare cases with radiological emphasis; retrospective multi-centre review. Patients diagnosed with PCNSL in Jan 2002 - Dec 2011 were identified in the West Midlands Cancer Intelligence Unit registry. Retrospective analyses of survival associated with age, sex and treatment were carried out. RESULTS: 153 patients were diagnosed. The incidence is 0.28/100000/year. None of them was cerebellar lesion. The 2 cases described occurred after Dec 2011. The mean age was 63 years. 101 patients (66%) were >60 years old. The M:F ratio was 1.25. Survival in patients >60 years was significantly less than ≤60 years (3.3 and 18 mth, respectively). Survival was longest in patients receiving both chemo- and radiotherapy (52 pt, 26.9 mth), followed by either chemo- or radiotherapy (28 pt, 3.1 mth and 38 pt, 3.0 mth) and least in those receiving none (35 pt, 1.3 mth). CONCLUSION: We report 2 rare cases of cerebellar PCNSL and described the rapid radiological transformation. The 10 year review of PCNSL confirmed its poor prognosis and showed that maximal therapy is associated with better survival.

Axillary ultrasound and FNAC/biopsy ‘Versus’ The West Midlands cancer intelligence unit prediction score: Which is more accurate at pre-operatively staging the axilla for breast cancer?

Introduction: Pre-operative staging of the axilla in breast cancer patients is essential for deciding initial surgical treatment and often avoids the need for a second operation. The aim of this study was to evaluate pre-operative axillary ultrasound and FNAC/biopsy and the West Midlands Cancer Intelligence Unit (WMCIU) prediction score in assessing axillary node status in screening and symptomatic patients.

OC-013 How often is upper Gastrointestinal Cancer Missed during Endoscopy?

<h3>Introduction</h3> Meta-analysis of published literature suggests that up to 14% of upper gastrointestinal cancer (UGIC) subjects had a negative oesophagogastroduodenoscopy (OGD) up to 3 years prior to diagnosis. We have examined how often UGIC is missed at OGD in our organisation and associated risk factors. <h3>Methods</h3> Computerised OGD records from Sandwell General and City Hospitals between 1999 and 2007 were retrieved and submitted to the West Midlands Cancer Intelligence Unit (WMCIU) for UGIC registrations linkage. Subjects undergoing OGD 3 months to 5 years before diagnosis were identified as potentially missed UGIC cases and those with no OGD 3 months to 5 years before diagnosis served as controls. The influence of age, gender, indication, endoscopist specialty, trainee involvement, sedation use, number of biopsies taken, site and histology of UGIC on missed UGIC were examined by logistic regression analysis. <h3>Results</h3> 36577 OGD records were submitted to WMCIU for cancer linkage and 524 UGIC were linked. 419 control UGIC: 388 (93%) diagnosed at initial OGD; 31 (7%) diagnosed at repeat OGD appropriately (within 90 days, with previous benign histology). 105 (20%) potentially missed UGIC cases: 39 (37%) oesophageal cancer (OC) and 66 (63%) gastric cancer (GC); 42 (40%) had OGD 3 months to 1 year prior to diagnosis; 33 (31%) had OGD 1–3 years prior and 30 (29%) had OGD 3–5 years prior. Furthermore, 30% of missed UGIC cases had OGD where abnormality at site of UGIC was seen but not biopsied or benign histology from an inadequate number of biopsies ( &lt; 4) within 3 months to 1 year prior to diagnosis. Lack of alarm symptoms (2.51, 95%CI 1.58–4.00, p = 0.0001) and female gender (1.79, 1.16–2.79, p = 0.009) were associated with missed UGIC. The number of biopsies taken was significantly lower in the missed UGIC group than in the controls (2.1 ± 0.2 vs 5.4 ± 0.1, p &lt; 0.05). In subjects with OC, mid-oesophageal OC appeared more likely to be missed than lower-oesophageal OC (2.04, 0.99–4.23, p = 0.05). Oesophageal squamous cell carcinoma was much more likely to be missed than oesophageal adenocarcinoma (4.47, 1.88–10.65, p &lt; 0.001). In GC subjects, there was no association between missed UGIC and tumour site (1.10, 0.61–1.97, p = 0.3) or histology subtype (1.10, 0.46–2.67, p = 0.8). Age (1.5, –0.8–3.8, p = 0.2), endoscopist specialty (1.39, 0.70–2.76, p = 0.34), trainee involvement (1.2, 0.78 – 1.86, p = 0.39) and sedation use (0.98, 0.64–1.51, p = 0.9) were not associated with increased risk of missing UGIC. <h3>Conclusion</h3> Missing UGIC at OGD was seen in 14.3% of subjects within 3 years of diagnosis. It was associated with lack of alarm symptoms, female gender, oesophageal squamous cell carcinoma and an insufficient number of biopsies from recognised abnormalities. <h3>Disclosure of Interest</h3> None Declared

P1-08-10: Invasive Lobular Breast Cancer – No Increased Risk of Contralateral Disease.

Abstract Background: Invasive lobular carcinoma (ILC) is the second most common type of breast cancer accounting for up to 14% of invasive breast cancers. Worldwide the incidence of breast cancer is increasing each year. Large population based studies using the Surveillance, Epidemiology, and End results (SEER) data have shown that the incidence of ILC has increased from 1977 to 1995. ILC has historically been thought to be associated with an increased risk of developing contralateral breast cancer and this belief may, in part, be responsible for the increasing trend towards contralateral prophylactic mastectomy. Methods: All female patients diagnosed with breast cancer during the time period 1998–2003 were identified from two large cancer registries, Northern and Yorkshire Cancer Registry and Information Service (NYCRIS) and West Midlands Cancer Intelligence Unit (WMCIU). All patients diagnosed with either Infiltrating duct(adeno) carcinoma, microinvasive ductal carcinoma, infiltrating ductular carcinoma (all classified as ‘ductal') or lobular (adeno)carcinoma were included. Patients with a mixed type of breast cancer were excluded from the analysis. Follow-up was complete until October 2010. Data were compiled for diagnosed contralateral breast cancer of any histological type. Results: Of the 32,735 patients with invasive ductal cancer 898 (2.74%) developed a contralateral breast cancer. In comparison 166 (3.1%) of 5397 patients with lobular cancer developed a contralateral breast cancer. The median time to first contralateral event was equivalent for both morphologies (ductal 40 months and lobular 39 months). Conclusion: This study suggests that there is no increased risk for developing a contralateral breast cancer in patients diagnosed with an invasive lobular breast cancer. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-08-10.

The risk of oesophageal cancer is not affected by a diagnosis of breast cancer

Oesophageal adenocarcinoma (OAC) is less common and develops at a later age in women compared with men. Endogenous oestrogen may therefore protect against OAC development. A cohort of women with breast cancer, a tumour commonly treated with oestrogen antagonists, was examined to identify the subsequent risk of developing OAC. Earlier studies have implicated radiotherapy in increasing oesophageal cancer (OC) risk among women with breast cancer. West Midlands Cancer Intelligence Unit data recording cancer diagnosis and treatment information was examined to identify patients with a first malignant primary breast cancer during 1977-2004. Patients were followed until diagnosis of a second primary cancer, death or end of the time period examined. Age-adjusted and period-adjusted standardized incidence ratios (SIR) were calculated as an estimate of relative risk for a second primary OC. Seventy-three thousand six hundred and thirteen women were eligible for the study, providing 486 679 person years at risk for analysis. One hundred and thirty-two second primary OCs were observed, compared with 121 expected (SIR 1.09; 95% confidence interval: 0.91-1.29). Radiotherapy treatment in 37 888 women did not affect the risk of a second primary OC (SIR 1.07; 95% confidence interval: 0.79-1.41). No difference was identified when examined by OC morphology.There was no association between breast cancer and a second primary OC. Radiotherapy that avoids deep irradiation in the treatment of breast cancer, local nodes or recurrence was not associated with an increased risk of developing a second primary OC.

PREDICT: a new UK prognostic model that predicts survival following surgery for invasive breast cancer

IntroductionThe aim of this study was to develop and validate a prognostication model to predict overall and breast cancer specific survival for women treated for early breast cancer in the UK.MethodsUsing the Eastern Cancer Registration and Information Centre (ECRIC) dataset, information was collated for 5,694 women who had surgery for invasive breast cancer in East Anglia from 1999 to 2003. Breast cancer mortality models for oestrogen receptor (ER) positive and ER negative tumours were derived from these data using Cox proportional hazards, adjusting for prognostic factors and mode of cancer detection (symptomatic versus screen-detected). An external dataset of 5,468 patients from the West Midlands Cancer Intelligence Unit (WMCIU) was used for validation.ResultsDifferences in overall actual and predicted mortality were <1% at eight years for ECRIC (18.9% vs. 19.0%) and WMCIU (17.5% vs. 18.3%) with area under receiver-operator-characteristic curves (AUC) of 0.81 and 0.79 respectively. Differences in breast cancer specific actual and predicted mortality were <1% at eight years for ECRIC (12.9% vs. 13.5%) and <1.5% at eight years for WMCIU (12.2% vs. 13.6%) with AUC of 0.84 and 0.82 respectively. Model calibration was good for both ER positive and negative models although the ER positive model provided better discrimination (AUC 0.82) than ER negative (AUC 0.75).ConclusionsWe have developed a prognostication model for early breast cancer based on UK cancer registry data that predicts breast cancer survival following surgery for invasive breast cancer and includes mode of detection for the first time. The model is well calibrated, provides a high degree of discrimination and has been validated in a second UK patient cohort.

Adjuvant UK: A UK Prognostic Model for Early Breast Cancer That Includes Mode of Detection.

Abstract AIM: To develop and validate a prognostication model to predict overall survival for women treated for early breast cancer in the UK based on cancer registry data. Unlike SEER, this dataset includes accurate information on mode of detection as well as systemic therapy.METHOD: Using the Eastern Cancer Registration &amp; Information Centre (ECRIC) dataset, information was collated for 5818 women diagnosed with invasive breast cancer in East Anglia from 1999-2003. All patients underwent surgery, had records of pathological staging (tumour size, grade, lymph node status, ER status), systemic treatment (chemotherapy, hormone therapy, both), mode of detection (screen-detected, symptomatic) and follow up, censored on 31 December 2007. A model was derived from these data using Cox proportional hazards, with ER positive and ER negative tumours modelled separately, and this was subsequently validated in an external dataset of 5468 patients from the West Midlands Cancer Intelligence Unit (WMCIU). Validation was performed by comparing (a) observed and expected mortality (overall &amp; breast cancer specific) at 5 &amp; 8 years and (b) receiver operating characteristic (ROC) curves in both ECRIC &amp; WMCIU datasets.RESULTS:ECRIC datasetDifference in overall observed/expected mortality &amp;lt;1% at 5 years (14.8% vs 14.9%) and 8 years (18.9% vs 18.9%). Area under ROC curve (AUC) was 0.81.Difference in breast cancer specific observed/expected mortality &amp;lt;1% at 5 years (10.6% vs 11.0%) and 8 years (12.9% vs 13.5%). AUC was 0.84.WMCIU dataset:Difference in overall observed/expected mortality &amp;lt; 1% at 5 years (15.8% vs 16.5%) and 8 years (17.5% vs 17.8%). AUC was 0.79.Difference in breast cancer specific observed/expected mortality &amp;lt;2% at 5 years (11.0% vs 12.6%) and 8 years (12.2% vs 13.6%). AUC was 0.82.Overall model fit was good across all subgroups although the ER positive model provided better discrimination (AUC 0.82) than ER negative (AUC 0.75). There was no significant difference between the ROC curves generated with ECRIC and WMCIU data (ER positive X2 = 0.17, p=0.68; ER negative X2 =0.00, P=0.95).CONCLUSION: We have developed a prognostication model for early breast cancer based on data from a UK cancer registry that has included mode of detection for the first time. The model is well calibrated, provides a high degree of discrimination and has been validated in a second patient cohort. This model will underpin a new web-based prognostication and treatment benefit tool for early breast cancer based on UK data (Adjuvant UK). Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4033.

Should the Age of Breast Cancer Screening for Ethnic Minorities Be Reduced?

Abstract IntroductionThe National Health Service Breast Screening Programme (NHSBSP) offers breast screening every three years for all women aged 50 and over in the United Kingdom (UK). Women aged between 50 and 70 are routinely invited. The NHSBSP will extend the age range of women invited for breast screening from ages 47 to 73 in the future. However, the ideal age for screening should take into consideration the relative risk of the population. The UK has a significant population of ethnic minority groups, mainly Asian and Afro-Caribbean, comprising of about 8% of the total population. The ethnic minority groups have remained geographically concentrated in specific regions especially London and West Midlands, where they comprise 30% and 14% of the population respectively. It has been reported that ethnic minorities present with relatively advanced breast cancer at a younger age. These groups could potentially benefit from early screening. We wanted to test this hypothesis by assessing the age and route of presentation of breast cancer across ethnic groups in the West Midlands region.MethodsThe West Midlands Cancer Intelligence Unit (WMCIU), a population based registry, collects prospective data on breast cancer diagnosed and treated in this region. Data on patients treated for breast cancer from 2001 to 2007 was obtained from the WMCIU. The age and route of presentation (screening or symptomatic) of Asian, Afro-Caribbean and Caucasian patients were analysed and compared.Results.A total of 27,444 female patients were diagnosed during this period. Data regarding ethnicity was available for 18,941 Caucasians, 528 Asians and 274 Afro-Caribbean's. Diagnosis through screening was 37%, 26% and 29% for Asian, Afro-Caribbean and the Caucasian population respectively. In the symptomatic group, 26.5% of Asians and 35% of Afro-Caribbeans presented below the age of 47 years compared to 13% of the Caucasian population. Furthermore, 13.5% of Asians, 16.5%of Afro-Caribbeans and 5.9% of Caucasians presented at or below the age of 40 years.ConclusionA significant percentage of breast cancers diagnosed within ethnic groups will still be missed despite revising the screening age to 47 years. Unless the participation of ethnic groups in screening is targeted, the benefits (detecting disease at an earlier stage and potentially reduced mortality) will be lost. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 4008.

Screening histories of invasive breast cancers diagnosed 1989–2006 in the West Midlands, UK: variation with time and impact on 10-year survival

To examine variations in screening status category with time for 20,862 invasive breast cancers diagnosed in the West Midlands between 1989 and 2006, and to provide 10-year relative survival rates for each category. Population-based breast screening was introduced in the West Midlands from 1988. Primary invasive breast cancers diagnosed from 1 April 1989 to 31 March 2006 in women eligible for breast screening were identified from the West Midlands Cancer Intelligence Unit's cancer registration database. A screening status was determined for each case using previously published methodology. Screening status rates are affected initially by the gradual rollout of the screening programme and more recently by the need to wait for three years before interval cancers and cancers in lapsed and non-attenders can be identified. In the steady-state period 1994-2003, 40.6% of invasive breast cancers were screen-detected, 36.3% were interval cancers and 11.2% were cancers in non-attenders (NA). Ten-year relative survival was 89.6% for screen-detected breast cancers, 73.3% for interval cancers and 51.9% for cancers in NA. The detection of invasive breast cancers by screening has improved over time, with an equivalent reduction in the numbers of interval cancers. This, together with a stable rate of cancers in NA and an improvement in the survival of women with interval cancers, suggests that further improvements in breast cancer mortality can be predicted. International comparisons of interval cancer rates are difficult due to interprogramme differences in the time between screens.

The influence of deprivation and ethnicity on the incidence of esophageal cancer in England

The incidence of esophageal cancer (EC), particularly esophageal adenocarcinoma (EAC), has been rising dramatically. In the USA, esophageal squamous cell carcinoma (ESCC) is associated with deprivation and black ethnicity, while EAC is more common among whites. The influence of social deprivation and ethnicity has not been studied in England. West Midlands Cancer Intelligence Unit data were used to study the incidence of ESCC and EAC, and the influence of age, sex, socioeconomic status (Townsend quintiles by postcode) and ethnicity (to individual patients from Hospital Episode Statistics). From 1977 to 2004, a total of 15,138 EC were identified. Five-year directly age standardized incidence rates per 100,000 (95% CI) for men increased from 8.6 (8.0-9.1) in 1977-1981 to 13.7 (13.1-14.3) in 2000-2004 and for women from 5.0 (4.7-5.4) to 6.3 (5.9-6.6). ESCC incidence did not alter, but EAC incidence rose rapidly in males [2.1 (1.9-2.4) to 8.5 (8.1-9.0)] and in females [0.5 (0.4-0.6) to 1.7 (1.5-1.9)]. ESCC was strongly associated with the most socially deprived quintile. EAC was not associated with differences in socioeconomic status. EAC was significantly more common in white men 7.3 (6.9-7.7) and women 1.5 (1.3-1.6) when compared with black and Asian populations. In England the incidence of EAC has rapidly risen, particularly in men over the last three decades. ESCC was strongly associated with social deprivation. EAC was more common in white populations, but no association with the socioeconomic status was found.

Adjuvant UK: Validation of a UK prognostic model for early breast cancer

11086 Background: To develop and validate a prognostication model to predict overall survival for women treated for early breast cancer in the UK based on cancer registry data. The main advantages are that unlike SEER, this dataset includes accurate information on mode of detection as well as systemic therapy. Methods: Using the Eastern Cancer Registration &amp; Information Centre (ECRIC) dataset, information was collated for 5,818 women diagnosed with invasive breast cancer in East Anglia from 1999–2003. All patients underwent surgery, had records of pathological staging (tumour size, grade, lymph node status, ER status) and systemic treatment (chemotherapy, hormone therapy, both), mode of detection (screen-detected, symptomatic) and were followed up to December 2007. A model was derived from these data using Cox proportional hazards and this was subsequently validated in an external dataset of 5,468 patients from the West Midlands Cancer Intelligence Unit (WMCIU). Validation was performed by comparing (a) observed and predicted mortality - overall and by risk group decile - and (b) receiver operating characteristic (ROC) curves in both ECRIC &amp; WMCIU datasets. Results: ECRIC dataset: Observed/predicted deaths were: 2 years (262/267), 5 years (868/867), 8 years (1108/1093). Area under ROC curve was 0.800WMCIU dataset: Observed/predicted deaths were: 2 years (317/298), 5 years (862/919), 8 years (955/999). Area under ROC curve was 0.783. The model and validation worked equally well at all time points (2, 5, 8 years) and in all risk group deciles. In the validation dataset the observed and predicted 8-year OS was within 1% (17.5% versus 18.3%). There was no significant statistical difference between the ROC curves (X2 = 2.35, P=.12) for prediction at 8 years in ECRIC and WMCIU data. Conclusions: We have developed a prognostication model for early breast cancer based on data from a UK cancer registry and validated the model using data from another UK registry. The model was well calibrated and provides a high degree of discrimination. This model will underpin the first web-based prognostication and treatment benefit tool for early breast cancer in the UK (Adjuvant UK). No significant financial relationships to disclose.

Patients with prostate cancer are less likely to develop oesophageal adenocarcinoma: could androgens have a role in the aetiology of oesophageal adenocarcinoma?

Oesophageal adenocarcinoma (OAC) is more common in men. Androgens may therefore contribute to the pathogenesis of OAC. Prostate cancer (PC), an androgen sensitive tumor with a long natural history, may allow insights into this putative association. West Midlands Cancer Intelligence Unit data from 1977 to 2004 were examined to identify patients with a first malignant primary of PC. Patients were followed until diagnosis of a second primary cancer, death or end of the time period. Age- and period-adjusted standardized incidence ratios (SIR) were calculated as an estimate of the relative risk of a second malignant primary of the oesophagus. Between 1977 and 2004, 44,819 men within the West Midlands developed PC as a first primary malignancy. After exclusion for lack of follow-up, 38,627 men were eligible, providing 143,526 person years at risk for analysis. 86 second primary oesophageal cancers were observed, compared with 110 expected, resulting in an SIR of 0.78 (95% CI 0.62-0.96). There was a reduced risk of OAC 0.7 (0.5-0.95) but not of oesophageal squamous cell carcinoma (OSCC) 1.03 (0.69-1.47). The risk of developing OAC, but not OSCC, is lower than expected in patients with PC. A diagnosis of PC may be associated with aetiological factors that are negatively associated with OAC, or anti-androgen therapy may influence the development of OAC.

Population estimates of survival in women with screen-detected and symptomatic breast cancer taking account of lead time and length bias

Evidence of the impact of breast screening is limited by biases inherent in non-randomised studies and often by lack of complete population data. We address this by estimating the effect of screen detection on cause-specific fatality in breast cancer, corrected for all potential biases, using population cancer registry data. Subjects (N = 26,766) comprised all breast cancers notified to the West Midlands Cancer Intelligence Unit and diagnosed in women aged 50-74, from 1988 to 2004. These included 10,100 screen-detected and 15,862 symptomatic breast cancers (6,009 women with interval cancers and 9,853 who had not attended screening). Our endpoint was survival to death from breast cancer. We estimated the relative risk (RR) of 10-year cause-specific fatality (screen-detected compared to symptomatic cancers) correcting for lead time bias and performing sensitivity analyses for length bias. To exclude self-selection bias, survival analyses were also performed with interval cancers as the comparator symptomatic women. Uncorrected RR associated with screen-detection was 0.34 (95% CI 0.31-0.37). Correcting for lead time, RR was 0.49 (95% CI 0.45-0.53); length bias analyses gave a range of RR corrected for both phenomena of 0.49-0.59, with a median of 0.51. Self-selection bias-corrected estimates yielded a median RR of 0.68. After adjusting for various potential biases, women with screen-detected breast cancer have a substantial survival advantage over those with symptomatic breast cancer.

Do delays between diagnosis and surgery in resectable oesophageal cancer affect survival? a study based on West Midlands cancer registration data

This retrospective study investigates if delays between the diagnosis of cancer of the oesophagus and surgical resection influence long-term survival. Data held by the West Midlands Cancer Intelligence Unit on 800 patients who underwent oesophagectomy for a diagnosis of cancer of the oesophagus or oesophagogastric junction between 1995 and 2000 were reviewed. Six hundred and thirty-two patients treated with curative intention and who had not received neo-adjuvant treatment in the form of radio- or chemotherapy were included in the analysis. The time interval between histological diagnosis and surgical resection was stratified into four groups: less than 3, 3–6, 6–9 and more than 9 weeks. The Cox proportional hazard model was used to test for the independent effect of delays. The results showed no difference in long-term survival according to the delay between histological diagnosis and surgical resection. On multivariate analysis adverse prognostic factors were advanced age, incomplete resection and lymph node involvement. Patients with a longer delay had a higher rate of complete tumour resection suggesting that they were more appropriately selected for the surgical treatment approach. In conclusion we have found no evidence that shorter delays from the date of histological diagnosis to surgical resection are beneficial to long-term survival.

Breast screening histories: variation with time and impact on 10-year survival

Every case of primary breast cancer diagnosed in the West Midlands since screening began until 31 March 2001 was assigned a screening status based on the woman's screening history. Eligible cases were identified from the West Midlands Cancer Intelligence Unit's cancer registration database. Screen-detected (SD) tumours were identified through regional breast screening units. The remaining tumours were allocated to one of eight screening status categories. A total 14,625 breast cancers were classified. The number of SD, interval cancers (IC), and cancers in lapsed attenders increased over time. There was a decrease over time in the number of cancers diagnosed in women before invitation and in nonattenders. IC rates were lower 24–36 months from screening to diagnosis compared with 12–24 months. Around one-third of cancers in lapsed attenders and nonattenders were diagnosed within 8 months of their last invitation, suggesting that they may have been screening provoked. There was a significant difference in the 10-year relative survival rates for attenders (SD, IC, lapsed attenders and nonattenders; 85.2% and 53.9%, respectively; RR = 1.58; P < 0.00001). Women with IC had survival rates above those of nonattenders, highlighting the benefits of screening. Table 1

The West Midlands breast cancer screening status algorithm – methodology and use as an audit tool

To illustrate the ability of the West Midlands breast screening status algorithm to assign a screening status to women with malignant breast cancer, and its uses as a quality assurance and audit tool. Breast cancers diagnosed between the introduction of the National Health Service [NHS] Breast Screening Programme and 31 March 2001 were obtained from the West Midlands Cancer Intelligence Unit (WMCIU). Screen-detected tumours were identified via breast screening units, and the remaining cancers were assigned to one of eight screening status categories. Multiple primaries and recurrences were excluded. A screening status was assigned to 14,680 women (96% of the cohort examined), 110 cancers were not registered at the WMCIU and the cohort included 120 screen-detected recurrences. The West Midlands breast screening status algorithm is a robust simple tool which can be used to derive data to evaluate the efficacy and impact of the NHS Breast Screening Programme.

West Midlands screening histories project: methodology and use as an audit tool

The Screening Histories Project aims to assign a screening status classification to all primary breast cancers in women aged 50–64, who were diagnosed in the West Midlands from 1 March 1988 to 31 March 2001. The possible classifications are: screen detected, interval cancer, nonattender, lapsed attender and diagnosed before invite. To obtain a screening status, screen-detected cancers are downloaded from the National Breast Screening Computer System (NBSS), matched against the West Midlands Cancer Intelligence Unit's cancer registration database, and classified as screen detected. All other eligible cases from the cancer registration database are manually checked against the NBSS. This involves assigning each case a screening unit and printing individual classification forms. The screening status is dependent upon whether the woman was invited prior to diagnosis and whether she attended her invitations. The first project phase is complete and all women diagnosed up to 31 March 1998 have been assigned a screening history. Thirty-seven per cent of women had a screen-detected cancer, 22% were interval cancers, 14% were diagnosed before invite, 11% were nonattenders and 2% were lapsed attenders. The dataset obtained by applying this algorithm is a valuable resource for the evaluation of breast screening in the West Midlands.