Analysis Of Cancer Registry Data Research Articles

Asymptotic justification of maximum likelihood estimation for the proportional excess hazard model in analysis of cancer registry data

Asymptotic justification of maximum likelihood estimation for the proportional excess hazard model in analysis of cancer registry data

Spatio-temporally smoothed deep survival neural network

The analysis of registry data has important implications for cancer monitoring, control, and treatment. In such analysis, (semi)parametric models, such as the Cox Proportional Hazards model, have been routinely adopted. In recent years, deep neural network (DNN) has been shown to excel in many fields with its flexibility and superior prediction performance, and it has been applied to the analysis of cancer survival data. Cancer registry data usually has a broad spatial and temporal coverage, leading to significant heterogeneity. Published studies have suggested that it is not sensible to fit one model for all spatial and temporal locations combined. On the other hand, it is inefficient to fit one model for each spatial/temporal location separately. Motivated by such considerations, in this study, we develop a spatio-temporally smoothed DNN approach for the analysis of cancer registry data with a (censored) survival outcome. This approach can accommodate the significant differences across time and space, while recognizing that the spatial and temporal changes are smooth. It is effectively realized via cutting-edge optimization techniques. To draw more definitive conclusions, we also develop an approach for assessing the importance of each individual input variable. Data on head and neck cancer (HNC) and pancreatic cancer from the Surveillance, Epidemiology, and End Results (SEER) database is analyzed. Compared to direct competitors, the proposed approach leads to network architectures that are smoother. Evaluated using the time-dependent Concordance-Index, it has a better prediction performance. The important variables are also biomedically sensible. Overall, this study can deliver a new and effective tool for deciphering cancer survival at the population level.

Abstract LB160: Oral cancer incidence among smokers and smokeless tobacco users compared to nonusers of tobacco products - contemporary analysis of cancer registry data

Abstract Tobacco use has been identified as a risk factor for oral cancer (oral cavity and pharynx). However, the differential oral cancer risk among users of cigarettes and smokeless tobacco (ST) products has not been well studied. In this study, we estimate oral cancer incidence among adult males who were current smokers that had never used ST products (SM), ST users who had never smoked (STU), and nonusers of cigarettes or ST products (NON) using data from state-wide cancer registries and national surveys. Oral cancer incidence rates were estimated using Colorado, Florida, North Carolina, and Texas cancer registries from the years 2014 through 2017 (the most recent 4 years of available data). State-specific population counts by tobacco use status were estimated from the Tobacco Use Supplement to the Current Population Survey. Missing values in tobacco use status in cancer registry data were imputed either by assuming missing at random (MR) or using multiple imputation (MI). Poisson regression was used to estimate incidence rate ratios (IRR) by state and age group. A random-effect meta-analysis approach was used to summarize estimates by state and age groups while considering potential heterogeneity using the Mantel-Haenszel model. Cases of oral cancer among adult males 35 years and older with tobacco use data numbered 19,536 in the geographies studied over the covered period. Assuming MR, the pooled oral cancer IRR relative to NON were 1.0 (95% CI = 0.7 to 1.3) for STU and 3.4 (95% CI = 3.0 to 3.9) for SM, resulting in an IRR of 3.5 (95% CI = 2.5 to 4.9) for SM relative to STU. The corresponding MI-based estimates relative to NON were 1.4 (95% CI = 1.1 to 1.9) for STU and 3.6 (95% CI = 3.2 to 4.1) for SM, resulting in an IRR of 2.6 (95% CI = 2.0 to 3.3) comparing SM relative to STU. These estimates should be considered in the context of certain limitations, such as the estimates were not adjusted for other potential confounding variables (e.g., alcohol intake and other lifestyle factors, duration of tobacco use, and other tobacco product use) due to the lack of such information in data sources. Our population estimates based on state cancer registry data demonstrate that current ST users had a substantially lower oral cancer risk than current smokers. While causal relationships cannot be concluded, the large sample size in this contemporary analysis suggests that oral cancer risks among smokers may be reduced by switching to ST products; however, quitting all tobacco use is the optimal path for reducing oral cancer risks. Citation Format: Brendan Noggle, Hui Cheng, Mohamadi Sarkar. Oral cancer incidence among smokers and smokeless tobacco users compared to nonusers of tobacco products - contemporary analysis of cancer registry data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr LB160.

Primary care and cancer: an analysis of the impact and inequalities of the COVID-19 pandemic on patient pathways

ObjectivesWe explore the routes to cancer diagnosis to further undertanding of the inequality in the reduction in detection of new cancers since the start of the pandemic. We use different...

Heterogeneity in head and neck cancer incidence among black populations from Africa, the Caribbean and the USA: Analysis of cancer registry data by the AC3

BackgroundAfrica and the Caribbean are projected to have greater increases in Head and neck cancer (HNC) burden in comparison to North America and Europe. The knowledge needed to reinforce prevention in these populations is limited. We compared for the first time, incidence rates of HNC in black populations from African, the Caribbean and USA. MethodsAnnual age-standardized incidence rates (IR) and 95% confidence intervals (95%CI) per 100,000 were calculated for 2013–2015 using population-based cancer registry data for 14,911 HNC cases from the Caribbean (Barbados, Guadeloupe, Trinidad & Tobago, N = 443), Africa (Kenya, Nigeria, N = 772) and the United States (SEER, Florida, N = 13,696). We compared rates by sub-sites and sex among countries using data from registries with high quality and completeness. ResultsIn 2013–2015, compared to other countries, HNC incidence was highest among SEER states (IR: 18.2, 95%CI = 17.6–18.8) among men, and highest in Kenya (IR: 7.5, 95%CI = 6.3–8.7) among women. Nasopharyngeal cancer IR was higher in Kenya for men (IR: 3.1, 95%CI = 2.5–3.7) and women (IR: 1.5, 95%CI = 1.0–1.9). Female oral cavity cancer was also notably higher in Kenya (IR = 3.9, 95%CI = 3.0–4.9). Blacks from SEER states had higher incidence of laryngeal cancer (IR: 5.5, 95%CI = 5.2–5.8) compared to other countries and even Florida blacks (IR: 4.4, 95%CI = 3.9–5.0). ConclusionWe found heterogeneity in IRs for HNC among these diverse black populations; notably, Kenya which had distinctively higher incidence of nasopharyngeal and female oral cavity cancer. Targeted etiological investigations are warranted considering the low consumption of tobacco and alcohol among Kenyan women. Overall, our findings suggest that behavioral and environmental factors are more important determinants of HNC than race.

English

Breast cancer (BC) is a public health concern in sub-Saharan Africa with rising incidence and mortality. In Niger, BC is identified as the leading cancer morbidity and mortality within female population. Its epidemiological literature remains limited or not structured. The aim of this work was to summarize evidence in relation to epidemiology of female breast cancer in Niger. The idea was to know how much was published and available on the topic and what the gaps are. This review was conducted by considering all articles published on the epidemiology of breast cancer in Niger using The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) model. A comprehensive search of material was performed using PubMed/Medline, African Journals Online, (AJOL), Web of Science, Google Scholar and Scopus/Elsevier. In total, 16 articles were initially screened, 14 full text articles were accessed for eligibility and 6 were included in the final review. None of the screened papers had clinical trial methodology or survival analysis. Moreover, no peer reviewed paper was found on breast cancer screening. The output of this review pointed out sustainable gaps in the field of epidemiology and public health research on BC in Niger. The country should have a national screening program and an updated analysis of cancer registry data. Effort is to be done on studies about cancer survival, feasibility and acceptability of alternative screening methods adapted to low-income countries, such as Niger. Key words: Niger, breast cancer, public health, screening.

Socioeconomic deprivation and cancer survival in a metropolitan area: An analysis of cancer registry data from Hamburg, Germany.

BackgroundFew studies have investigated socioeconomic inequalities within cities. Yet, such analyses are particularly important given the increasing international trend to urbanization. Here we investigated area-based socioeconomic inequalities in cancer survival in Hamburg, a port city in the North of Germany (population: 1.84 million people).MethodsPatients with a diagnosis of colorectal, lung, female breast, and prostate cancer in 2004–2018 (follow-up until 31.12.2018) and registered in the Hamburg cancer registry were included. Area-based socioeconomic deprivation on urban district level was assigned to the patients and grouped in five quintiles. Relative survival in 2014–2018 was calculated using the period approach. Trend analyses between 2004 and 2018 were conducted. Relative excess risks adjusted for age and stage were computed with model-based period analyses.FindingsFor the 73,106 included patients, age-standardized 5-year relative survival in 2014–2018 decreased with increasing deprivation with significant differences between the most and least deprived group of 14·7 (prostate), 10·8 (colorectal), 8·0 (breast), and 2·5 (lung) percent units. Standardization by cancer stage decreased the difference for prostate cancer to 8·5 percent units and for breast cancer to 3·6 percent units but had only a minor effect for colorectal and lung cancer. Similar socioeconomic inequalities were already present in 2004–08.InterpretationStrong socioeconomic inequalities in cancer survival were observed in Hamburg, which could be partly explained by differences in the stage distribution. Further research including information on screening participation as well as information on cancer care are important to further understand and finally overcome these inequalities.Funding10.13039/501100005972German Cancer Aid.

The Role of Postoperative Radiotherapy for Carcinosarcoma of the Uterus.

Simple SummaryThe role of radiotherapy on carcinosarcoma, a rare malignant tumor, of the uterus is unclear. We reviewed data published from 2010 on the effects of radiotherapy on tumor control and survival in this patient group. Available data were mainly from cancer registries and suggested that radiotherapy, given either as vaginal brachytherapy (contact radiotherapy of the vagina) or external-beam radiotherapy or a combination of both, reduces the risk of recurrence and improves survival in patients with all stages of carcinosarcoma of the uterus without metastases in other organs.The role of postoperative radiotherapy delivered as external-beam radiotherapy (EBRT), vaginal brachytherapy (VBT) or a combination of both, in the management of carcinosarcoma of the uterus is not clearly defined, as only limited randomized trial data are available, indicating a reduction in locoregional recurrences after EBRT. We performed a structured review of data published from 2010. Although no relevant new data from prospective trials or meta-analyses were identified, 14 analyses of cancer registry data from the United States or Europe, focusing predominantly on the endpoint for overall survival, were identified, four of them using propensity-score matching to compare subgroups treated with vs. without radiotherapy. Although stage-by-stage data are rare, the registry analyses support the idea of a beneficial effect, especially of VBT, on overall survival in International Federation of Gynecology and Obstetrics (FIGO) stage IA patients (to a lesser extent in stage IB). For stages II to III, the data sets indicate the largest effects on overall survival for the combination of EBRT and VBT. In all stages, survival effects of radiotherapy apparently persist when given in addition to chemotherapy. Whereas some studies see the strongest survival effects in patients with positive lymph nodes, propensity-score matched data indicate an overall survival effect of radiotherapy (EBRT + VBT or VBT alone) in FIGO stages I to III regardless of lymph node surgery.

RADT-21. NATIONAL PRACTICE PATTERNS AND OUTCOMES OF STEREOTACTIC BODY RADIOTHERAPY VS. CONVENTIONAL EXTERNAL BEAM RADIOTHERAPY FOR SPINAL METASTASES

Abstract INTRODUCTION Up to 10% of cancer patients experience spinal cord compression from metastatic disease. Palliation and local control were traditionally pursued with conventional external beam radiotherapy (cEBRT), but advancements in image-guidance and intensity-modulation for stereotactic body radiotherapy (SBRT) have dramatically changed the management of these lesions. Herein we evaluate the national practice patterns and outcomes associated with cEBRT vs. SBRT. METHODS U.S. patients newly diagnosed with metastatic cancer necessitating RT to the spine were identified from the National Cancer Database (2004-2016), stratified by RT modality and cancer type, and evaluated using multivariable logistic regression and Cox proportional hazards. RESULTS 34,759 U.S. patients required spinal RT within 3 months of initial stage 4 cancer presentation, primarily for lung adenocarcinoma (25%), lung small cell carcinoma (14%), and prostatic (12%) metastases. Patients overwhelming received cEBRT (30Gy/10; 50%), followed by hypo-fractionated SBRT (15-30Gy/2-6; 11%) and single-fraction SBRT (i.e. stereotactic radiosurgery, SRS; 15-24Gy/1; 0.9%); whereas 38% received another regimen (e.g. 30-37.5Gy/12-15 or 40Gy/20). From 2004→2016, the rates of single-fraction SRS (0.4→1.9%) and hypo-fractionated SBRT (13.1→23.6%) increased, whereas cEBRT (86.5→74.4%) decreased. SBRT was significantly more likely to be utilized at academic hospitals as compared to cEBRT (OR 0.57; 95% CI: 0.49-0.66; p< 0.01). SBRT was more likely utilized for elderly or high comorbidity patients and varied across cancer types. Survival analysis indicated that across all cancer types, single-fraction SRS, was independently associated with improved overall survival compared to cEBRT (HR 1.51; 95%CI: 1.31-1.74; p< 0.01) after adjusting for patient characteristics, care setting, tumor type and systemic treatment. CONCLUSIONS Through analysis of cancer registry data, we found that practice patterns of RT for spinal metastases have been evolving nationally, with an increase in the use of SBRT. Single-fraction SBRT was associated with improved adjusted OS. Notably, we found that utilization of SBRT lags in the community setting.

Abstract PR14: Do Latinas with breast cancer who live in ethnic enclaves have better or worse survival? Analysis of cancer registry data from California and Texas

Abstract Introduction: Many Latinos live in ethnic enclaves—culturally distinct neighborhoods with high concentrations of individuals of the same ethnic origin, high linguistic isolation, a large share of recent immigrants, and ethnic specific businesses and resources. Prior studies on the association of enclave residence and cancer mortality were often limited to a single state and demonstrated mixed results. It is unclear whether the mixed results are due to widely varying measures and analytic methods, or true regional differences in enclave effects. Methods: We conducted parallel analyses of California and Texas cancer registry data from adult (≥18 years of age) Latinas diagnosed with invasive breast cancer from 1996 to 2005, with follow-up through 2014. Our key focus was to use the same measures and methods. We linked 2000 U.S. Census data to measure Latino enclaves and neighborhood socioeconomic status (nSES). We defined enclaves using an established multidimensional index of seven census tract measures (percent of residents who are Latino, foreign-born, recent immigrants, and linguistically isolated (general and of those who speak Spanish), with limited English proficiency (general and of those who speak Spanish) split into statewide quintiles. We fitted Cox proportional hazard models for all-cause and breast cancer specific mortality adjusted for year of diagnosis, patient age, birthplace (with multiple imputation), tumor stage, histology, grade, and size, and clustering by census tract. We explored interactions of enclave residence with nSES and patient birthplace. Results: Among 40,716 Latinas, the majority (61.3% in CA, 70.5% in TX) lived in ethnic enclaves. Ethnic enclave residence was more common among foreign- vs U.S.-born Latinas (72.0% vs 52.3% in CA and 82.4% vs. 68.5% in TX). In fully adjusted models for both states, foreign- vs. US-born women were more likely to die from breast cancer and all causes. Living in an ethnic enclave and in neighborhoods with higher SES was independently associated with decreased mortality. Patterns were consistent in terms of direction and significance of associations across states for all-cause and breast cancer-specific mortality. Future analyses will further explore the statistically significant interactions (p<.05) observed between enclave residence, nSES, and patient birthplace. Discussion: Applying the same methods across two states eliminated previously published inconsistent results about the association of enclave residence and mortality among Latinas with breast cancer. Future studies should consider pooling cancer registry data across states and using the same methods to better understand the true impact of ethnic enclave residence across the cancer continuum. Future studies should also focus on identifying the specific protective effects of enclave residence to inform interventions. This abstract is also being presented as Poster C073. Citation Format: Salma Shariff-Marco, Scarlett Lin Gomez, Alison Canchola, Hannah Fullington, Amy E. Hughes, Sandi L. Pruitt. Do Latinas with breast cancer who live in ethnic enclaves have better or worse survival? Analysis of cancer registry data from California and Texas [abstract]. In: Proceedings of the Eleventh AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2018 Nov 2-5; New Orleans, LA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl):Abstract nr PR14.

Development of a patient-oriented navigation model for lung cancer patients.

e18033 Background: Patient navigation is seen to support and enable patient-oriented, optimal care both in palliative and in screening settings. However, the evidence remains inconclusive on what patient groups are best targeted by navigation and what may be improved by such a care model. Lung cancer patients are at particular risk for sub-optimal care because they face complex care trajectories due to severe and rapid disease progression and accompanying comorbidity. Methods: To develop a navigation model for lung cancer, we conducted a mixed-methods study to investigate who may be at risk of receiving sub-optimal care in the German health care setting. To capture the patient perspective a longitudinal qualitative component was included with patients (N = 20) assessed at three dtime points. In addition, a secondary data analysis of cancer registry data of a comprehensive cancer center was conducted and a repository of patient support offers gathered. Results of the study components were integrated to develop a patient-oriented navigation model. Results: Secondary data analysis showed that medical care functioned according to tumor board recommendations. Patient data revealed institutional barriers that conflict with individual needs and preferences. A lack of contact persons, information provision as well as bureaucratic difficulties were identified. Patients without a social network seem particularly in need for support. Identification of regional support offers shows that there are resources available to meet some of these needs. However, knowledge on such offers was not common among patients and caregivers. Navigators should provide practical support, give advice on social care issues and refer to existing support offers. Conclusions: Social networks crucial. Patients lack knowledge to use available resources. Navigation needs to be implemented within existing care structures to reach patients.

Impact of screening on cervical cancer incidence in England: a time trend analysis

ObjectivesTo better model underlying trends in cervical cancer incidence so as to model past trends, to estimate the impact of cervical screening on cervical cancer rates at different ages and...

Comparative Effectiveness of Prostate Cancer Treatment Options: Limitations of Retrospective Analysis of Cancer Registry Data

Retrospective analyses of cancer registry and institutional data have consistently found better survival after radical prostatectomy versus radiation therapy, which contrasts with findings from a randomized trial. This is likely because of the inability of retrospective studies to fully account for comorbidity differences across treatment groups because of the lack of detailed data in the registries. We use a unique population-based data set with detailed data regarding comorbidities and functional limitations to assess whether this can provide valid comparisons of survival across prostate cancer treatment groups. The Surveillance, Epidemiology, and End Results (SEER)-Medicare Health Outcomes Survey (MHOS) data set results from a linkage between the SEER database and the MHOS database, which includes detailed information regarding patient-reported comorbidity and functional limitations. We analyzed 3102 patients with prostate cancer in SEER-MHOS and used latent class analysis to identify the healthiest group with minimal comorbidity burden and functional limitations. Among the healthiest group, we examined overall survival across treatments using the Kaplan-Meier method. Three distinct health groups were identified using latent class analysis; the healthiest group comprised 57% of the cohort and had a 10-year overall survival of 67%. Other health groups had higher rates of comorbidities or functional limitations. Among the healthiest group, 10-year overall survival differed across treatment groups: no local treatment (55%), external beam radiation therapy (69%), brachytherapy (76%), and radical prostatectomy (85%). Survival curves for the 3 treated groups separated at 4years of follow-up. Despite the detailed health status information available in SEER-MHOS, our retrospective analysis could not fully account for patient selection biases across prostate cancer treatment groups. These findings highlight an important limitation of retrospective studies using population-based data sets and serve as a reminder to interpret results with caution.

Iranian high risk regions due to esophageal cancer: spatial analysis of cancer registry data.

The aim of this study was to identify the esophageal cancer (EC) high risk regions to evaluate changes of relative risks (RRs) for both genders by time in Iranian provinces. EC is one of the public health problems in Iran. In spite of this fact, there is not comprehensive study estimating RRs across the Iranian provinces. In this cross-sectional study the data for EC cases were extracted from Ministry of Health and Medical Education (MOHME) including 30 provinces from 2004 to 2010. For estimating the model parameters, we used Bayesian approach by regarding spatial correlations of adjacent provinces. The Northern half of Iran has high risk and other half has low risk. During the time, the range of RRs has decreased for both gender and also the dispersion of EC is decreasing for women but nearly is fixed for men. While RR has declined during the study, focusing on the Northern half of Iran as high risk regions is a considerable fact for policymakers.

Breast cancer survival by molecular subtype: a population-based analysis of cancer registry data.

The relation between breast cancer molecular subtype and survival has been studied in several jurisdictions, but limited information is available for Ontario. The aim of this study was to determine breast cancer survival by molecular subtype and to assess the effect on survival of selected demographic and tumour-based characteristics. We extracted 29 833 breast cancer cases (in 26 538 girls and women aged ≥ 15 yr) diagnosed between 2010 and 2012 from the Ontario Cancer Registry. Cancers were categorized into 4 molecular subtypes: 1) luminal A (estrogen-receptor-positive and/or progesterone-receptor-positive [ER+ and/or PR+] and negative for human epidermal growth factor receptor 2 [HER2-]), 2) luminal B (ER+ and/or PR+/HER2+), 3) HER2-enriched (ER- and PR-/HER2+) and 4) triple-negative (ER- and PR-/HER2-). We estimated associations with predictor variables (age, stage at diagnosis, histologic type, comorbidity and place of residence [urban or rural]) using a multivariate Cox proportional hazards model. Likelihood ratio testing was used to evaluate differences in risk of death. Luminal A was the most commonly diagnosed subtype (59.0%) and had the greatest survival, whereas triple-negative had the poorest survival. For all subtypes, a dose-response effect was observed between the hazard of death and age and stage at diagnosis, with the greatest effect found for the HER2-enriched subtype (age: hazard ratio [HR] 7.87 [95% confidence interval (CI) 3.68-11.81]; stage at diagnosis: HR 37.71 [95% CI 34.64-41.27]). Moderate comorbidity (Charlson Comorbidity Index score 1 or 2) was associated with increased risk of death for triple-negative cancers (HR 2.42 [95% CI 1.36-4.31]), and severe comorbidity (Charlson Comorbidity Index score ≥ 3) increased the risk for all molecular subtypes. The results indicate the importance of including molecular subtype, along with age, stage at diagnosis and comorbidity, in assessing breast cancer survival. They highlight the need to address outcomes related to hormone-receptor-negative cancers, for which survival lags behind that for hormone-receptor-positive cancers.

Abstract A30: Affordability of screening, race and marital status predict early detection of breast cancer: Analysis of cancer registry data

Abstract Background: Breast cancer is the most common cancer diagnosis among women in Nebraska. Early diagnosis of breast cancer provides opportunities for better prognosis and treatment options and thereby improves chances of survival. Nebraska is predominantly a rural state, and hence, in addition to problems of affordability of health care due to lack of insurance or under-insurance, the people of rural Nebraska face unique challenges with regards to the availability and accessibility of cancer screening services. The purpose of this study was to examine whether access to cancer screening services, in terms of three dimensions: affordability, availability and accessibility, predict the stage of diagnosis for women diagnosed with breast cancer in Nebraska. Methods: Data on breast cancer cases in Nebraska were obtained from the Nebraska Cancer Registry for the years 2008 to 2012. Proximity to the nearest screening center was calculated as the shortest distance between the population weighted centroid of each census tract in Nebraska and the nearest screening facility, using geocoded facility locations available from the United States Food and Drug Administration website. Spatial accessibility to primary care provider (PCP) was calculated using the two-step floating catchment area method using geocoded PCP locations, using data from the University of Nebraska Medical Center Health Professions Tracking Service annual survey database. Hierarchical logistic regression models adjusted for age, race, ethnicity, marital status, rurality of residence and county poverty level were examined to assess the association of type of insurance: Medicare, Medicaid, Other (including TRICARE, Military, Veterans Affairs Indian/Public Health Service) and Private, supply of screening centers, proximity to screening center and spatial accessibility to primary care physicians with the stage at diagnosis of breast cancer. Breast cancer stage at diagnosis was categorized as early (in-situ or localized stage) or late (regional or distant stage). Geocoding and proximity distance calculations were done using ArcGIS 10.3.2 and statistical analyses were conducted using STATA 14 software. Results: Among 4,975 women aged 40 years or older and diagnosed with breast cancer in Nebraska between 2008 and 2012, 72.3% were diagnosed at an early stage (in-situ or localized). The results from the hierarchical logistic regression found that women who were uninsured were less likely (Odds Ratio [O.R]: 0.42; 95% Confidence Interval [C.I]: 0.25-0.73) to be diagnosed early and those women who had Medicaid coverage were also less likely (O.R: 0.56; 95% C.I: 0.40-0.78) to be diagnosed early, as compared to women having private insurance. Further, married women were 1.3 times more likely (O.R: 1.25; 95% C.I: 1.10-1.44) to be diagnosed early, and white women were 1.4 times more likely (O.R: 1.36; 95% C.I: 1.04-1.77) to be diagnosed early. Conclusion: Affordability of cancer screening services plays an important role in early detection of breast cancer. Hence, the coverage of preventive services through the Affordable Care Act is likely to be a positive policy change leading to higher screening rates and thereby, higher rates of early detection of breast cancer. Further, there exist racial disparities in early diagnosis of breast cancer that may be due to differential access to screening services. Targeted policy efforts will be needed to address these disparities to improve outcomes of breast cancer for all race groups. Citation Format: Aastha Chandak, Preethy Nayar, Ge Kan, Niodita Gupta. Affordability of screening, race and marital status predict early detection of breast cancer: Analysis of cancer registry data. [abstract]. In: Proceedings of the AACR Special Conference: Improving Cancer Risk Prediction for Prevention and Early Detection; Nov 16-19, 2016; Orlando, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2017;26(5 Suppl):Abstract nr A30.

Doubly robust estimator for net survival rate in analyses of cancer registry data.

Cancer population studies based on cancer registry databases are widely conducted to address various research questions. In general, cancer registry databases do not collect information on cause of death. The net survival rate is defined as the survival rate if a subject would not die for any causes other than cancer. This counterfactual concept is widely used for the analyses of cancer registry data. Perme, Stare, and Estève (2012) proposed a nonparametric estimator of the net survival rate under the assumption that the censoring time is independent of the survival time and covariates. Kodre and Perme (2013) proposed an inverse weighting estimator for the net survival rate under the covariate-dependent censoring. An alternative approach to estimating the net survival rate under covariate-dependent censoring is to apply a regression model for the conditional net survival rate given covariates. In this article, we propose a new estimator for the net survival rate. The proposed estimator is shown to be doubly robust in the sense that it is consistent at least one of the regression models for survival time and for censoring time. We examine the theoretical and empirical properties of our proposed estimator by asymptotic theory and simulation studies. We also apply the proposed method to cancer registry data for gastric cancer patients in Osaka, Japan.

Incidence of esophageal cancer in Sri Lanka: Analysis of cancer registry data and comparison with other South Asian populations.

The objectives of this study were to report the incidence of Carcinoma of Esophagus (CaE) in Sri Lanka and to compare these values with other cancer registry data of the region and with migrant populations. We compared the data published by the National Cancer Control Program over the last two decades with data from the National Cancer Registry Programme of the Indian Council of Medical Research and Karachi Cancer Registry. SEERstat was used to analyze the surveillance, epidemiology and end results database to analyze data on Indian migrant population. CaE was the fourth most common cancer overall and among females and third most common cancer among males. The incidence of CaE rises with age in both sexes, with a peak in the 70-74 year age group. There was a disproportionately higher number of CaE in the Tamil population (chi-square test, P < 0.00001). The commonest type of CaE in Sri Lanka was squamous cell carcinoma, Not otherwise specified (NOS) (n = 750, 70.5%), followed by adenocarcinoma, NOS (n = 83, 7.8%). India, Pakistan and Sri Lanka have comparable age-adjusted incidence and age distribution of CaE. All migrant populations had lower incidence of CaE than original population or population in their present country. Both cigarette smoking and alcohol consumption are more prevalent in Sri Lankan males than females. The incidence of CaE and its distribution among age groups in Sri Lanka was comparable to other countries of the region. Persons of Tamil ethnicity have a higher risk of developing CaE.

A multilevel excess hazard model to estimate net survival on hierarchical data allowing for non-linear and non-proportional effects of covariates.

The excess hazard regression model is an approach developed for the analysis of cancer registry data to estimate net survival, that is, the survival of cancer patients that would be observed if cancer was the only cause of death. Cancer registry data typically possess a hierarchical structure: individuals from the same geographical unit share common characteristics such as proximity to a large hospital that may influence access to and quality of health care, so that their survival times might be correlated. As a consequence, correct statistical inference regarding the estimation of net survival and the effect of covariates should take this hierarchical structure into account. It becomes particularly important as many studies in cancer epidemiology aim at studying the effect on the excess mortality hazard of variables, such as deprivation indexes, often available only at the ecological level rather than at the individual level. We developed here an approach to fit a flexible excess hazard model including a random effect to describe the unobserved heterogeneity existing between different clusters of individuals, and with the possibility to estimate non-linear and time-dependent effects of covariates. We demonstrated the overall good performance of the proposed approach in a simulation study that assessed the impact on parameter estimates of the number of clusters, their size and their level of unbalance. We then used this multilevel model to describe the effect of a deprivation index defined at the geographical level on the excess mortality hazard of patients diagnosed with cancer of the oral cavity. Copyright © 2016 John Wiley & Sons, Ltd.

L20 - Geographical spread of gastric cancer incidence in cremona: spatial analysis of cancer registry data

L20 - Geographical spread of gastric cancer incidence in cremona: spatial analysis of cancer registry data