Background
With increasing incidence worldwide and prolonged survival in the presence of active cancer because of unique tumor biology, neuroendocrine tumors (NETs) are now more prevalent than pancreatic, esophageal, and gastric cancers combined. 1 –5 There have been more clinical trials, new drug development studies, and epidemiology and healthcare delivery analyses that have improved the knowledge of and therapeutic options for NETs. 4 –9 However, information on the burden of cancer on mortality for patients with NETs remains scarce.
NETs are a heterogeneous malignancy with variable biologic, clinical, and prognostic characteristics. There is an evolving need to adapt treatment plans to their unique biology and chronic behavior, which is generally atypical for malignancies. With prolonged survival and more frequent incidental findings, patients with NETs may not die of their disease. 4,5 Understanding the cancer and noncancer prognosis of NETs is critical in guiding decisions regarding care, monitoring strategies, aggressiveness of therapy, and patient counseling to enable shared decision-making. 10 Yet, although overall survival, recurrence-free survival, and associated prognostic factors have been described in several studies, few have reported cancer-specific survival. 4,5,11 –15 Existing studies are limited to a single NET primary tumor site, consider a specific treatment modality, and do not address competing risks. 11,13 –15
Thus, we conducted a population-based study of patients diagnosed with NETs to examine cancer and noncancer deaths after diagnosis and identify prognostic factors associated with cause-specific death.
Methods
Study Design
We conducted a population-based retrospective cohort study using prospectively collected administrative datasets linked to administrative health databases stored at ICES in Toronto, Canada. The study was approved by Sunnybrook Health Sciences Centre Research Ethics Board, and was conducted and reported following the RECORD statement 16 and the PROGnosis RESearch Strategy (PROGRESS). 10
Data Sources
Datasets used included the Ontario Cancer Registry (OCR), Registered Persons Database (RPDB), Ontario Health Insurance Plan (OHIP), and Ontario Registrar General Death database (ORGD). 17 –19 Detailed information on all the databases used are provided in supplemental eTable 1 (available with this article at JNCCN.org). Datasets were linked using unique encoded identifiers at ICES.
Study Population and Cohort
Under the Canada Health Act, Ontario’s 13.5 million residents benefit from universally accessible and publicly funded healthcare though OHIP. 20 Adult patients (aged ≥18 years) diagnosed with NETs from January 1, 2001, through December 31, 2015, were identified in the OCR with ICD-O-3 codes using a previously reported strategy (supplemental eTable 2). 4,21
Outcomes
Our primary outcome was death after NET diagnosis, classified as cancer-specific death or noncancer death. Time to death was calculated from date of diagnosis to date of death. Cancer-specific death was defined as ICD-9 codes 140 through 239 (topography codes for any cancer death whether from NET or other histology) and noncancer death as other codes in the ORGD. The primary cause of death was defined as the antecedent cause of death when available or the immediate cause of death when antecedent causes were not captured. High agreement in cause of death has been reported between ORGD and clinical follow-up in a prospective cohort of patients with cancer. 22 Noncancer cause of death was further described using a previously described strategy (supplemental eTable 3). 23 The category “other cause of death” included stomach and duodenal ulcers; complications of pregnancy, childbirth, and the perinatal period; congenital anomalies; homicide and legal intervention; and other causes of death. Proportion of patients for each cause of death was reported for 2 time periods: ≤5 and >5 years from date of NET diagnosis. Participants were observed until date of death, date of last clinical contact, or December 31, 2016, allowing for a minimum of 12 months of follow-up for each patient.
Covariates
Baseline characteristics were measured at the time of diagnosis. Rural living was determined using the rurality index of Ontario based on the postal code of the patients’ primary residence, with index ≥40 representing rural residence. 24 Material deprivation is a composite index of the inability for individuals or households to afford consumption goods and activities typical in a society at a given point in time. 25,26 Baseline comorbidity burden was measured using the Johns Hopkins Adjusted Clinical Group system score based on health services use with a 24-month look-back window before the date of NET diagnosis. 27,28 The 32 aggregated diagnosis groups (ADGs) were summed and dichotomized, with a cutoff of 10 for high comorbidity burden. 28,29
NETs were subdivided according to primary tumor site into bronchopulmonary, gastroenteric, pancreatic, and others using ICD-O-3 topography codes. Gastroenteric NETs were further subdivided as gastric, small intestinal, colonic, and rectal. Metastatic disease was defined using ICD-9 and ICD-10 codes and divided into synchronous metastases (metastasis code ≤6 months from the date of NET diagnosis) and metachronous metastases (metastasis code >6 months after the date of NET diagnosis) (supplemental eTable 4).
Statistical Analysis
Treating time to death as a time-to-event variable, we estimated the cumulative incidence of death after NET diagnosis using the cumulative incidence function for cancer death, treating noncancer death as a competing risk. 30,31 We reported estimates for probability of death as percentages with 95% confidence intervals at 5 and 10 years. Because use of ICD-9 codes to define cancer-specific death relies on topography and not histology, it is possible that cancer-specific deaths would have been related to non-NET cancer. To explore the potential contribution of non-NET cancer-specific death to the results, we conducted a sensitivity analysis that excluded patients with a non-NET cancer diagnosis after NET diagnosis.
To explore the independent association of potential prognostic factors on the probability of noncancer death, we constructed multivariable subdistribution hazard models.32,33 We selected potential prognostic factors a priori based on clinical relevance and literature review, including age, sex, primary tumor site, material deprivation quintile, rural residence, year of diagnosis (dichotomous), and metastatic status. 4,5,32 We reported the subdistribution hazard ratios (sHRs) with 95% confidence interval.
We examined missing data for key variables. Data were missing for rural residency in 1.2% and material deprivation in 1.1% of the cohort. We performed a complete case analysis whereby patients with missing data were excluded for analyses using these variables. Statistical significance was set at P≤.05. All analyses were conducted using SAS Enterprise Guide 7.1 (SAS Institute Inc.).
Results
Of 8,707 identified patients with NET diagnosis, 91 were excluded for being aged <18 or >105 years and 9 for having a date of death recorded before the date of NET diagnosis. The final cohort comprised 8,607 patients diagnosed with NETs (Table 1 and supplemental eTable 5). Median follow-up was 42 months (interquartile range [IQR], 17–82 months) and there were 3,121 (36.3%) death events during follow-up, of which 2,487 were because of cancer. The most common primary tumor site was bronchopulmonary (22.8%), followed by small intestine (19.3%) and rectum (14.4%). A total of 42.2% of patients had metastases, including 32.0% synchronous metastases. This varied by primary tumor sites. For rectal and pancreatic NET, 90.8% and 70.7% of patients, respectively, did not have metastases. A non-NET cancer was diagnosed after NET diagnosis in 765 patients.
Characteristics of Patient Cohort
Overall survival for the entire cohort was 67.1% (95% CI, 66.0%–68.1%) at 5 years and 55.1% (95% CI, 53.6%–56.6%) at 10 years. The cumulative incidences of cancer-specific deaths along with the competing risk of noncancer death for the entire cohort are presented in Figure 1. The risk of cancer-specific death for all patients was higher than that of noncancer death, with 27.3% (95% CI, 26.3%–28.4%) and 5.6% (95% CI, 5.1%–6.1%), respectively, at 5 years, and 34.5% (95% CI, 33.2%–35.8%) and 10.3% (95% CI, 9.4%–11.3%), respectively, at 10 years. In patients with metastatic disease, both synchronous and metachronous, the risk of cancer-specific death largely exceeded that of noncancer death immediately after diagnosis. In patients without metastases, the risk of cancer-specific and noncancer deaths paralleled each other. Cancer-specific death was initially slightly higher than noncancer death with 7.7% (95% CI, 7.0%–8.6%) and 6.2% (95% CI, 5.4%–6.9%) at 5 years, respectively. Only after 8 years from diagnosis did the risk of noncancer death (11.7%; 95% CI, 10.4%–13.2%, at 10 years) become slightly higher than that of cancer-specific death (10.2%; 95% CI, 9.1%–11.3%, at 10 years). Sensitivity analysis excluding patients with non-NET cancer diagnosis after NET diagnosis did not substantially alter the results (supplemental eFigure 1).
Cancer-specific death and noncancer death cumulative incidence after neuroendocrine tumor diagnosis for (A) all patients and (B) stratified by metastatic status.
Citation: Journal of the National Comprehensive Cancer Network 19, 8; 10.6004/jnccn.2020.7666
Different patterns of cancer and noncancer deaths were observed depending on the primary NET site. When considering all patients, regardless of metastatic status, the risk of cancer-specific death was consistently higher than that of noncancer death; however, the magnitude of the cancer-specific death varied (Figures 2 and 3, and supplemental eFigure 2). The highest risks of cancer-specific death were observed for bronchopulmonary NETs, with 36.4% (95% CI, 34.2%–38.7%) at 5 years and 42.7% (95% CI, 0.1%–45.3%) at 10 years (Figure 3). This was followed by pancreatic NETs (34.8% [95% CI, 31.5%–38.2%] at 5 years and 48.4% [95% CI, 43.5%–53.0%] at 10 years) (Figure 2) and colonic NETs (21.4% [95% CI, 18.9%–24.0%] at 5 years and 26.6% [95% CI, 23.4%–29.9%] at 10 years) (Figure 4). When stratifying by metastatic status, the risk of cancer-specific death largely exceeded that of noncancer death across all primary NET sites, with most of the increase in cancer-specific death occurring in the first 2 years after diagnosis.
Cancer-specific death and noncancer death cumulative incidence after pancreatic neuroendocrine tumor diagnosis for (A) all patients and (B) stratified by metastatic status.
Citation: Journal of the National Comprehensive Cancer Network 19, 8; 10.6004/jnccn.2020.7666
Cancer-specific death and noncancer death cumulative incidence after bronchopulmonary neuroendocrine tumor diagnosis for (A) all patients and (B) stratified by metastatic status.
Citation: Journal of the National Comprehensive Cancer Network 19, 8; 10.6004/jnccn.2020.7666
Cancer-specific death and noncancer death cumulative incidence after gastroenteric NET diagnosis stratified by metastatic status for (A) gastric, (B) small intestine, (C) colonic, and (D) rectal NETs.
Abbreviation: NET, neuroendocrine tumor.
Citation: Journal of the National Comprehensive Cancer Network 19, 8; 10.6004/jnccn.2020.7666
In certain scenarios of nonmetastatic NET, the cumulative incidence of cancer-specific death was lower than that of noncancer death (Figure 4). For nonmetastatic gastric NETs, incidence of noncancer death became higher than cancer-specific death starting at 1 year after diagnosis, with 10.4% (95% CI, 7.1%–14.4%) at 5 years and 18.9% (95% CI, 13.4%–25.2%) at 10 years versus 5.8% (95% CI, 3.5%–8.9%) at 5 years and 9.0% (95% CI, 5.1%–14.3%) at 10 years, respectively (Figure 4). The same was observed for nonmetastatic small intestine NETs from the time of diagnosis, with 13.0% (95% CI, 10.2%–16.0%) at 5 years and 22.4% (95% CI, 17.7%–27.4%) at 10 years for noncancer death versus 4.7% (95% CI, 3.1%–6.7%) at 5 years and 8.6% (95% CI, 5.7%–12.3%) at 10 years for cancer-specific death, as well as for colonic NETs starting at 3 years after diagnosis, with noncancer death of 6.4% (95% CI, 4.5%–8.8%) at 5 years and 11.4% (95% CI, 7.9%–15.6%) at 10 years, and cancer-specific death of 4.1% (95% CI, 2.7%–5.9%) at 5 years and 5.9% (95% CI, 3.8%–8.6%) at 10 years, and finally rectal NETs starting at 2 years after diagnosis, with noncancer death of 2.7% (95% CI, 1.8%–3.9%) at 5 years and 6.6% (95% CI, 4.7%–9.0%) at 10 years, and cancer-specific death of 7.4% (95% CI, 5.9%–8.9%) at 5 years and 9.3% (95% CI, 7.4%–11.4%) at 10 years. This was not observed for nonmetastatic pancreatic and bronchopulmonary NETs, for which cancer-specific death exceeded noncancer death from the time of diagnosis (Figures 2 and 3).
We performed multivariable analyses to identify factors associated with the risk of cancer-specific death, accounting for the competing risk of noncancer death (Table 2 and supplemental eTable 6). Advancing age was consistently associated with increased hazards of cancer-specific death in all NETs and when stratifying by NET group of bronchopulmonary, pancreatic, and gastroenteric. Female sex was associated with lower hazards of cancer-specific death compared with male sex for all NETs and stratified by NET group, although the association was not significant for pancreatic NETs. Higher material deprivation, indicating lower socioeconomic status, was independently associated with increased hazards of cancer-specific death. Higher comorbidity burden was associated with lower hazards of cancer-specific death, likely because of the observed higher hazards of noncancer death (sHR, 1.73; 95% CI, 1.46–2.04), compared with patients with lower comorbidity burden. Both synchronous and metachronous metastases were independently associated with increased hazards of cancer-specific death in all NETs (sHR, 9.19 [95% CI, 8.16–10.36] vs 6.55 [95% CI, 5.84–7.60], respectively). This was also observed in stratified categories of NETs for bronchopulmonary, pancreatic, and gastroenteric NETs. Among all NETs, gastroenteric (sHR, 0.33; 95% CI, 0.29–036) and pancreatic NETs (sHR, 0.74; 95% CI, 0.65–0.85) presented independently lower hazards of cancer-specific death compared with bronchopulmonary NETs.
Multivariable Fine-Gray Model Showing Factors Associated With Cancer-Specific Death
Factors associated with noncancer death were also examined (supplemental eTable 7). Advancing age, higher comorbidity burden, and increased material deprivation were associated with increased hazards of noncancer death. Both synchronous (sHR, 0.64; 95% CI, 0.53–0.77) and metachronous (sHR, 0.68; 95% CI, 0.52–0.89) metastatic disease was associated with lower hazards of noncancer death, compared with no metastases.
Finally, looking at noncancer cause of death by time period, 420 deaths (16.0% of deaths over the time period) were caused by cancer in ≤5 years and 214 deaths (8.2% of deaths over the time period) in >5 years from NET diagnosis. The breakdown of noncancer cause of death is presented in Figure 5. The most common causes of noncancer death were diseases of the heart and noncardiac vascular diseases.
Causes of noncancer death after NET diagnosis.
Abbreviation: NET, neuroendocrine tumor.
Citation: Journal of the National Comprehensive Cancer Network 19, 8; 10.6004/jnccn.2020.7666
Discussion
This population-based cohort study is the first to describe patterns of and factors associated with cancer-specific death after a NET diagnosis. We showed that patients with NETs experienced prolonged survival, and quantified how the risk of cancer-specific and noncancer death contributed to this. Patients with metastatic disease consistently had a largely higher risk of cancer-specific versus noncancer deaths. On the other hand, the risk of noncancer death exceeded that of cancer-specific death in some nonmetastatic NETs, namely gastric, small intestine, colonic, and rectal.
Data on cancer-specific mortality for NETs are scarce. 11,13 –15 Two studies developed nomograms to predict cancer-specific survival in all pancreatic NETs and in metastatic pancreatic NETs, but there was no specific information regarding risk estimates for cancer-specific survival or noncancer death. 14,15 Two studies reported on cancer-specific survival for pancreatic NETs using Kaplan-Meier methods, which do not account for the competing risk of noncancer death. 13 Competing-risk methods and assessment of noncancer death along with cancer-specific death are recommended to improve the reporting of cancer-specific mortality. 33 A SEER study reported on NET recurrence patterns in the Medicare population (aged ≥65 years), but although it indicated that patients with localized NETs at diagnosis have low risk of recurrence and therefore low likelihood of dying from NETs, it did not examine cancer-specific death. 34 Our study provided long-term cancer and noncancer mortality outcomes for a contemporary cohort of NETs, further stratified by primary NET site and metastatic status, accounted for competing risks to avoid overestimation of the risk of cancer death, and offered a detailed analysis of prognostic factors for cancer-specific and noncancer death. 30,31,35
Because of the unique biology of NETs, most patients experience prolonged survival. This potentially comes with associated longitudinal risk for noncancer death, which can complicate decisions regarding goals, modality, and sequencing of therapy, as well as individual patient counseling and supportive care. Thus, it is important to have data-driven insights for estimates of cancer-specific death across all patients and NET sites, which we have provided herein using population-based data representative of the full spectrum of patients encountered in contemporary clinical practice. Our results indicate that the probability of dying of cancer was greater than that of dying of noncancer causes overall. Those findings are modulated by NET type and metastatic status.
Patients with metastatic NETs had a much higher risk of dying from cancer than from other causes, consistent with the more aggressive nature of their disease. Accordingly, metastatic status was a strong independent prognostic factor of cancer death, but not of noncancer death. The curves for all NETs and those stratified by primary NET site showed an initial steep increase in cancer death in the first 2 years after diagnosis, indicating potentially more-aggressive NETs, consistent with known biology of advanced gastric, pancreatic, colonic, rectal, and bronchopulmonary NETs. Metastatic small intestine NETs presented a unique pattern of risk of cancer-specific death, with a gradual increase over 10 years, along with more favorable risk estimates than other primary NET sites. This can indicate a different and more chronic disease behavior, the availability of more treatment options, or higher levels of comfort among providers in treating those tumors. Nevertheless, the slow pace and lower incidence of cancer death in small intestine NETs should be taken into consideration. It supports longitudinal sequencing of therapy for metastatic small bowel NETs over years, as well as treatment strategies that consider long-term sequelae and quality of life, such as early resection of primary small intestinal NETs to prevent abdominal complications. 36,37
In certain circumstances, cancer-specific death did not exceed noncancer death. The increase in incidence of NETs has previously been hypothesized as being related to diagnosis at an earlier stage and even possible overdiagnosis. 4,5 There is ongoing debate regarding the best management of incidental NETs; although some support monitoring and nonoperative management, others propose a more aggressive surgical approach capitalizing on the opportunity to achieve cure and avoid subsequent NET spread and its associated mortality. In this setting, patterns of cancer-specific and noncancer death can provide important information to avoid overtreatment. Patients with nonmetastatic gastric, small intestine, colonic, and rectal NETs are less likely to die of cancer-specific versus other causes. However, the current analysis could not ascertain the contribution of therapy patterns to the risk of cancer-specific death. It is important that data on cancer-specific mortality be used along with specific patient and tumor information, such as age, comorbidity, endocrinopathy, tumor grade and size, and individual preferences, to optimize the design of personalized treatment strategies. It is acknowledged that various types of NETs do not have the same opportunity for therapy, ranging from potentially curative resection for localized small bowel NETs to multiple lines of systemic therapy for metastatic pancreatic NETs. Our observations warrant further investigation into cancer-specific mortality according to therapy and management patterns. The unique patterns observed also suggest that cancer-specific mortality should be used as an outcome when assessing treatment approaches for those types of NETs.
Finally, examination of factors associated with cancer-specific and noncancer-related death showed that efforts to address cancer-specific death in NETs should include special considerations for older adults and socioeconomically deprived patients to ensure they can access and receive care throughout their cancer journey. The association of advancing age with an increased risk of cancer death may be related to underlying increased vulnerability and frailty, but also risks of de-escalating cancer care in older adults who are underrepresented in clinical studies. 38,39 Patients with a higher level of deprivation are known to exhibit different health-seeking behaviors and experience delayed diagnosis and access to care, which impact outcomes. 32,40,41
Despite the inherent challenges associated with retrospective population-based studies, this examination provides a long-term detailed and robust appraisal of cancer-specific mortality among patients with NETs. However, the administrative data used was not collected specifically to address the current research question. We lacked some information regarding tumor characteristics, including stage and grade. We used metastases as a surrogate for advanced disease due to the unique biology of NETs and previous observations of metastatic status being the most significant prognostic factor. 3 Using grade classification for the multivariable analysis and to stratify the results could have provided even more detailed information for clinical use, as could further stratification of primary NET site in models to examine factors associated with cancer-specific death. However, despite the overall large cohort, this resulted in too many levels for the covariates, with loss of levels of freedom and appropriateness of the regression model, such that primary NET sites were grouped into gastroenteric, pancreatic, bronchopulmonary, and others for the main model. Nevertheless, the population-based design of this study is a strength that allowed for a novel real-world assessment of the prognosis of patients diagnosed with NETs. These data also represent outcomes within a universal healthcare system, avoiding bias by insurance status–limited access. We used high-quality linked administrative databases and definitions with known accuracy and reliability, which included comprehensive availability of cause of death and had limited missing data and measurement error.
Conclusions
In this population-based analysis, we reported cancer-specific mortality for patients diagnosed with NETs. Among all patients diagnosed with NETs, the risk of dying of cancer is higher than that of dying of other causes, especially in those with metastatic disease. Heterogeneity exists by primary NET site in patterns of cancer and noncancer death. This initial description of cancer-specific mortality for NETs can support patient counseling, clinical decision-making, discussion regarding clinical practice guidelines, and future trial designs.
References
- 1.↑
National Cancer Institute. Surveillance, Epidemiology, and End Results Program. Cancer Stat Facts. Accessed August 21, 2019. Available at: https://seer.cancer.gov/statfacts/
- 3.↑
Yao JC , Hassan M , Phan A , et al. One hundred years after “carcinoid”: epidemiology of and prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol 2008;26:3063–3072.
- 4.↑
Hallet J , Cukier M , Saskin R , Liu N . Exploring the rising incidence of neuroendocrine tumors: a population-based analysis of epidemiology, metastatic presentation, and outcomes. Cancer 2015;121:589–597.
- 5.↑
Dasari A , Shen C , Halperin D , et al. Trends in the incidence, prevalence, and survival outcomes in patients with neuroendocrine tumors in the United States. JAMA Oncol 2017;3:1335–1338.
- 6.↑
Yao JC , Fazio N , Singh S , et al. Everolimus for the treatment of advanced, non-functional neuroendocrine tumours of the lung or gastrointestinal tract (RADIANT-4): a randomised, placebo-controlled, phase 3 study. Lancet 2016;387:968–977.
- 7.↑
Singh S , Moody L , Chan DL , et al. Follow-up recommendations for completely resected gastroenteropancreatic neuroendocrine tumors. JAMA Oncol 2018;4:1597–1604.
- 8.↑
Halperin DM , Shen C , Dasari A , et al. Frequency of carcinoid syndrome at neuroendocrine tumour diagnosis: a population-based study. Lancet Oncol 2017;18:525–534.
- 9.↑
Pavel M , Gross DJ , Benavent M , et al. Telotristat ethyl in carcinoid syndrome: safety and efficacy in the TELECAST phase 3 trial. Endocr Relat Cancer 2018;25:309–322.
- 10.↑
Hemingway H , Croft P , Perel P , et al. Prognosis research strategy (PROGRESS) 1: a framework for researching clinical outcomes. BMJ 2013;346:e5595.
- 11.↑
Ye H , Xu HL , Shen Q , et al. Palliative resection of primary tumor in metastatic nonfunctioning pancreatic neuroendocrine tumors. J Surg Res 2019;243:578–587.
- 12.↑
Fraenkel M , Kim MK , Faggiano A , Valk GD . Epidemiology of gastroenteropancreatic neuroendocrine tumours. Best Pract Res Clin Gastroenterol 2012;26:691–703.
- 13.↑
Powers BD , Rothermel LD , Fleming JB , et al. A survival analysis of patients with localized, asymptomatic pancreatic neuroendocrine tumors: no surgical survival benefit when examining appropriately selected outcomes. J Gastrointest Surg 2020;24:2773–2779
- 14.↑
Li Z , Du S , Feng W , et al. Competing risks and cause-specific mortality in patients with pancreatic neuroendocrine tumors. Eur J Gastroenterol Hepatol 2019;31:749–755.
- 15.↑
He CB , Zhang Y , Cai ZY , et al. The impact of surgery in metastatic pancreatic neuroendocrine tumors: a competing risk analysis. Endocr Connect 2019;8:239–251.
- 16.↑
Benchimol EI , Smeeth L , Guttmann A , et al. The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. PLOS Med 2015;12:e1001885.
- 17.↑
Iron Z , Zagorski BM , Sykora K . Living and Dying in Ontario: An Opportunity for Improved Health Information. ICES Investigative Report. Toronto, Ontario, Canada: Institute for Clinical Evaluative Sciences (ICES); 2008.
- 18.↑
Clarke EA , Marrett LD , Kreiger N . Cancer registration in Ontario: a computer approach. IARC Sci Publ 1991;(95):246–257.
- 19.↑
Robles SC , Marrett LD , Clarke EA , et al. An application of capture-recapture methods to the estimation of completeness of cancer registration. J Clin Epidemiol 1988;41:495–501.
- 20.↑
Government of Canada. Canada Health Act. Accessed August 21, 2019. Available at: https://www.canada.ca/en/health-canada/services/health-care-system/canada-health-care-system-medicare/canada-health-act.html
- 21.↑
Hallet J , Law CHL , Saskin R , Liu N . Rural-urban disparities in incidence and outcomes of neuroendocrine tumors: a population-based analysis of 6271 cases. Cancer 2015;121:2214–2221.
- 22.↑
Brenner DR , Tammemägi MC , Bull SB , et al. Using cancer registry data: agreement in cause-of-death data between the Ontario Cancer Registry and a longitudinal study of breast cancer patients. Chronic Dis Can 2009;30:16–19.
- 23.↑
Afifi AM , Saad AM , Al-Husseini MJ , et al. Causes of death after breast cancer diagnosis: a US population-based analysis. Cancer 2020;126:1559–1567.
- 24.↑
Karlj B . Measuring “rurality” for purposes of health-care planning: an empirical measure for Ontario. Ont Med Rev 2009;33:52.
- 25.↑
Krieger N . Overcoming the absence of socioeconomic data in medical records: validation and application of a census-based methodology. Am J Public Health 1992;82:703–710.
- 26.↑
Matheson FI , Dunn JR , Smith KLW , et al. Development of the Canadian Marginalization Index: a new tool for the study of inequality. Can J Pub Health 2012;103:S12–16.
- 27.↑
Reid RJ , Roos NP , MacWilliam L , et al. Assessing population health care need using a claims-based ACG morbidity measure: a validation analysis in the Province of Manitoba. Health Serv Res 2002;37:1345–1364.
- 28.↑
Reid RJ , MacWilliam L , Verhulst L , et al. Performance of the ACG case-mix system in two Canadian provinces. Med Care 2001;39:86–99.
- 29.↑
Weiner JP , Starfield BH , Steinwachs DM , et al. Development and application of a population-oriented measure of ambulatory care case-mix. Med Care 1991;29:452–472.
- 30.↑
Austin PC , Lee DS , Fine JP . Introduction to the analysis of survival data in the presence of competing risks. Circulation 2016;133:601–609.
- 31.↑
Austin PC , Fine JP . Practical recommendations for reporting Fine-Gray model analyses for competing risk data. Stat Med 2017;36:4391–4400.
- 32.↑
Hallet J , Coburn NG , Singh S , et al. Access to care and outcomes for neuroendocrine tumours: does socioeconomic status matter? Curr Oncol 2018;25:e356–364.
- 33.↑
Wissing MD , Greenwald ZR , Franco EL . Improving the reporting of cancer-specific mortality and survival in research using cancer registry data. Cancer Epidemiol 2019;59:232–235.
- 34.↑
Shen C , Dasari A , Chu Y , et al. Clinical, pathological, and demographic factors associated with development of recurrences after surgical resection in elderly patients with neuroendocrine tumors. Ann Oncol 2017;28:1582–1589.
- 35.↑
Putter H , Fiocco M , Geskus RB . Tutorial in biostatistics: competing risks and multi-state models. Stat Med 2007;26:2389–2430.
- 36.↑
Citterio D , Pusceddu S , Facciorusso A , et al. Primary tumour resection may improve survival in functional well-differentiated neuroendocrine tumours metastatic to the liver. Eur J Surg Oncol 2017;43:380–387.
- 37.↑
Daskalakis K , Karakatsanis A , Stålberg P , et al. Clinical signs of fibrosis in small intestinal neuroendocrine tumours. Br J Surg 2016;104:69–75.
- 38.↑
Townsley CA , Selby R , Siu LL . Systematic review of barriers to the recruitment of older patients with cancer onto clinical trials. J Clin Oncol 2005;23:3112–3124.
- 39.↑
Hurria A , Dale W , Mooney M , et al. Designing therapeutic clinical trials for older and frail adults with cancer: U13 conference recommendations. J Clin Oncol 2014;32:2587–2594.
- 40.↑
Booth CM , Li G , Zhang-Salomons J , Mackillop WJ . The impact of socioeconomic status on stage of cancer at diagnosis and survival: a population-based study in Ontario, Canada. Cancer 2010;116:4160–4167.
- 41.↑
Woods LM , Rachet B , Coleman MP . Origins of socio-economic inequalities in cancer survival: a review. Ann Oncol 2006;17:5–19.
