An approach to estimation in relative survival regression

Incidence, Stage, Treatment, and Survival of Noncardia Gastric Cancer

Importance

Gastric cancer is the fifth most common cancer worldwide, and investigating its incidence, characteristics, treatment, and outcomes over the past decades can help in selecting clinical strategies and future research directions.

Objective

To analyze the trends in incidence, staging, and treatment of gastric cancer.

Design, Setting, and Participants

This nationwide, population-based cohort study included patients diagnosed with noncardia gastric cancer (NCGC) between 1989 and 2021 in the Netherlands.

Main Outcomes and Measures

Differences in tumor characteristics, treatment, and survival were analyzed per fixed time periods (1989-1993, 1994-1998, 1999-2003, 2004-2008, 2009-2013, 2014-2018, and 2019-2021).

Results

In total, 47 014 patients (median [IQR] age, 73 [64-80] years; 28 032 [60%] male patients) were identified with mostly adenocarcinomas of the antrum region (when location was known). Age-standardized incidence decreased from 20.3 to 6.1 per 100 000 person-years between 1989 and 2021. During the study period, unknown T and N stages were recorded less frequently, and metastatic disease was diagnosed more frequently (1989-1993: 2633 of 9493 patients [28%]; 2019-2021: 1503 of 3200 patients [47%] in 2019-2021). Over time, fewer patients with metastatic disease underwent surgery with or without other treatment modalities (68% in 1989-1993 vs 64% in 2019-2021), and palliative chemotherapy in metastatic NCGC increased from 9% to 40%. For patients with nonmetastatic disease, 5-year relative survival improved from 28% (95% CI, 26.5%-29.2%) to 36% (95% CI, 33.5%-37.6%) between 1989 and 2021. For patients with nonmetastatic disease undergoing a resection, 5-year survival increased from 40% (95% CI, 38.3%-41.8%) to 51% (95% CI, 47.9%-53.3%). For patients with metastatic disease, 1-year relative survival increased from 10% (95% CI, 8.7%-11.1%) to 19% (95% CI, 17.2%-21.6%), but 3-year relative survival remained poor at 5% (95% CI, 3.6%-7.5%).

Conclusions and Relevance

In this nationwide cohort study involving 47 014 patients diagnosed with NCGC (1989-2021), the results showed a decrease in incidence, more accurate staging, a shift in treatment modalities, and improved patient survival.

Sex differences in treatment allocation and survival of potentially curable gastroesophageal cancer: A population-based study

BACKGROUND

Although curative treatment options are identical for male and female gastroesophageal cancer patients, access to care and survival may vary. This study aimed to compare treatment allocation and survival between male and female patients with potentially curable gastroesophageal cancer.

METHODS

Nationwide cohort study including all patients with potentially curable gastroesophageal squamous cell or adenocarcinoma diagnosed between 2006 and 2018 registered in the Netherlands Cancer Registry. The main outcome, treatment allocation, was compared between male and female patients with oesophageal adenocarcinoma (EAC), oesophageal squamous cell carcinoma (ESCC), and gastric adenocarcinoma (GAC). Additionally, 5-year relative survival with relative excess risk (RER), that is, adjusted for the normal life expectancy, was compared.

RESULTS

Among 27,496 patients (68.8% men), most were allocated to curative treatment (62.8%), although rates dropped to 45.6%>70 years. Curative treatment rates were comparable among younger male and female patients (≤70 years) with gastroesophageal adenocarcinoma, while older females with EAC were less frequently allocated to curative treatment than males (OR = 0.85, 95% confidence interval [CI] 0.73-0.99). For those allocated to curative treatment, relative survival was superior for female patients with EAC (RER = 0.88, 95% CI 0.80-0.96) and ESCC (RER = 0.82, 95% CI 0.75-0.91), and comparable for males and females with GAC (RER = 1.02, 95% CI 0.94-1.11).

CONCLUSIONS

While curative treatment rates were comparable between younger male and female patients with gastroesophageal adenocarcinoma, treatment disparities were present between older patients. When treated, the survival of females with EAC and ESCC was superior to males. The treatment and survival gaps between male and female patients with gastroesophageal cancer warrant further exploration and could potentially improve treatment strategies and survival.

A censored quantile regression approach for relative survival analysis: Relative survival quantile regression

We propose a censored quantile regression model for the analysis of relative survival data. We create a hybrid data set consisting of the study observations and counterpart randomly sampled pseudopopulation observations imputed from population life tables that adjust for expected mortality. We then fit a censored quantile regression model to the hybrid data incorporating demographic variables (e.g., age, biologic sex, calendar time) corresponding to the population life tables of demographically-similar individuals, a population versus study covariate, and its interactions with the variables of interest. These latter variables can be interpreted as relative survival parameters that depict the differences in failure quantiles between the study participants and their population counterparts.

What is relative survival and what is its role in haematology?

In many haematological diseases, the survival probability is the key outcome. However, when the population of patients is rather old and the follow-up long, a significant proportion of deaths cannot be attributed to the studied disease. This lessens the importance of common survival analysis measures like overall survival and shows the need for other outcome measures requiring more complex methodology. When disease-specific information is of interest but the cause of death is not available in the data, relative survival methodology becomes crucial. The idea of relative survival is to merge the observed data set with the mortality data in the general population and thus allow for an indirect estimation of the burden of the disease. In this work, an overview of different measures that can be of interest in the field of haematology is given. We introduce the crude mortality that reports the probability of dying due to the disease of interest; the net survival that focuses on excess hazard alone and presents the key measure in comparing the disease burden of patients from populations with different general population mortality; and the relative survival ratio which gives a simple comparison of the patients' and the general population survival. We explain the properties of each measure, and some brief notes are given on estimation. Furthermore, we describe how association with covariates can be studied. All the methods and their estimators are illustrated on a sub-cohort of older patients who received a first allogeneic hematopoietic stem cell transplantation for myelodysplastic syndromes or secondary acute myeloid leukemia, to show how different methods can provide different insights into the data.

Measuring population-level cure patterns for cancer patients in the United States

While the life expectancy of cancer survivors has substantially improved over time in the United States, the extent to which cancer patients are cured is not known. Population-level cure patterns are important indicators to quantify cancer survivorships. This population-based cohort study included 8978,721 cancer patients registered in the Surveillance, Epidemiology and End Results (SEER) databases between 1975 and 2018. The primary outcome was cure fractions. Five-year cure probability, time to cure and median survival time of uncured cases were also assessed. All four measures were calculated using flexible parametric models, according to 46 cancer sites, three summary stages, individual age and calendar year at diagnosis. In 2018, cure fractions ranged from 2.7% for distant liver cancer to 100.0% for localized/regional prostate cancer. Localized cancer had the highest cure fraction, followed by regional cancer and distant cancer. Except for localized breast cancer, older patients generally had lower cure fractions. There were 38 cancer site and stage combinations (31.2%) that achieved 95% of cure within 5 years. Median survival time of the uncured cases ranged from 0.3 years for distant liver cancer to 10.9 years for localized urinary bladder cancer. A total of 117 cancer site and stage combinations (93.6%) had increased cure fraction over time. A considerable proportion of cancer patients were cured at the population-level, and the cure patterns varied substantially across cancer site, stage and age at diagnosis. Increases in cure fractions over time likely reflected advances in cancer treatment and early detection.

Survival after Ivor Lewis versus McKeown esophagectomy for cancer: propensity score matched analysis

It is unknown whether Ivor Lewis (IL) or McKeown (McK) esophagectomy is preferred in patients with potentially curable esophageal or gastro-esophageal junction (GEJ) cancer. Patients with mid- and distal esophageal and GEJ cancer without distant metastases who underwent IL or McK esophagectomy in the Netherlands between 2015 and 2017, were selected from the Netherlands Cancer Registry. Patients were propensity score matched for sex, age, American Society of Anesthesiologist classification, comorbidity, tumor type, tumor location, clinical stage, neoadjuvant treatment and year of diagnosis. The primary outcome was a 3-year relative survival (RS). Secondary outcome parameters were number of lymph nodes examined, number of positive lymph nodes, radical resection rate, tumor regression grade, post-operative complications and mortality. A total of 1627 patients who underwent IL (n = 1094) or McK (n = 533) esophagectomy were included. Patient and tumor characteristics were balanced after propensity score matching, leaving 658 patients to be compared. The 3-year RS was 54% after IL and 50% after McK esophagectomy, P = 0.140. The median number of lymph nodes examined, median number of positive lymph nodes, radical resection rate and tumor regression grade were comparable between both groups. Recurrent laryngeal nerve palsy (2 vs. 5%, P = 0.006) occurred less frequently after IL esophagectomy. No differences were observed in post-operative anastomotic leakage rate, pulmonary complication rate and mortality rates. There was no statistically significant difference in the 3-year RS between IL and McK esophagectomy. Based on these results, both IL and McK esophagectomy can be performed in patients with mid to distal esophageal and GEJ cancer.

Sex differences in tumor characteristics, treatment, and outcomes of gastric and esophageal cancer surgery: nationwide cohort data from the Dutch Upper GI Cancer Audit

BACKGROUND

Sex differences in clinicopathological characteristics, treatment, and postoperative outcomes of gastric and esophageal cancer are largely undefined. This study aimed to compare tumor and treatment characteristics and outcomes of gastric and esophageal cancer surgery between male and female patients.

METHODS

Patients after elective surgery for primary esophageal (EAC) or gastric adenocarcinoma (GAC) registered in the Dutch Upper GI Cancer Audit between 2011 and 2016 were included. The primary endpoint, 5-year relative survival with relative excess risk (RER), i.e., adjusted for the normal life expectancy, was compared between male and female patients with EAC and GAC.

RESULTS

In total, 4937 patients were included (75% male) with a mean age of 66 years. cT and cN-stages showed a similar distribution in male and female patients. In females, antrum GAC was more frequent (47% vs. 38%, p < 0.001). Female patients with EAC less frequently received neo-adjuvant treatment (OR = 0.60, 95% CI 0.38-0.96, p = 0.033). For GAC, less postoperative morbidity (33% vs. 38% p = 0.017) and less re-interventions (12% vs. 16%, p = 0.008) were observed in females, although they had inferior 5-year relative survival (49% vs. 56%, RER = 1.31, 95% CI 1.09-1.58, p = 0.004). No differences in relative survival of EAC were observed.

CONCLUSIONS

In addition to significant sex differences in tumor location, female patients with esophageal adenocarcinoma less frequently received neo-adjuvant therapy, and female patients with gastric adenocarcinoma had inferior relative survival. Further consideration and exploration of sex differences in surgical treatment and outcomes are necessary to improve tailored treatment and outcomes.

Inverse probability weighting and doubly robust standardization in the relative survival framework

A commonly reported measure when interested in the survival of cancer patients is relative survival. Relative survival circumvents issues with inaccurate cause of death information by incorporating the expected mortality rates of cancer individuals from population lifetables of the general population. A summary of the cancer population prognosis can be obtained using the marginal relative survival. To explore differences between exposure groups, such as socioeconomic groups, the difference in marginal relative survival between exposed and unexposed can be obtained and under assumptions is interpreted as the average causal effect of exposure to survival. In a modeling context, this is usually estimated by applying regression standardization as the average of the individual-specific estimates after fitting a relative survival model. Regression standardization yields an estimator that consistently estimates the causal effect under standard causal inference assumptions and if the relative survival model is correctly specified. We extend inverse probability weighting (IPW) and doubly robust standardization methods in the relative survival framework as additional valuable tools for obtaining average causal effects when correct model specification might not hold for the relative survival model. IPW yields an unbiased estimate of the average causal effect if a correctly specified model has been fitted for the exposure (propensity score) whereas doubly robust standardization requires that at least one of the propensity score model or the relative survival model is correctly specified. An example using data on melanoma is provided and a simulation study is conducted to investigate how sensitive are the methods to model misspecification, including different ways for obtaining standard errors.

Real‐world data on treatment concepts in classical Hodgkin lymphoma in Sweden 2000–2014, focusing on patients aged >60 years

Treatment for patients > 60 years with classical Hodgkin lymphoma (cHL) is problematic; there is no gold standard, and outcome is poor. Using the Swedish Lymphoma Registry, we analysed all Swedish patients diagnosed with cHL between 2000 and 2014 (N = 2345; median age 42 years; 691 patients were >60 years). The median follow‐up time was 6.7 years. Treatment for elderly patients consisted mainly of ABVD or CHOP, and the younger patients were treated with ABVD or BEACOPP (with no survival difference). In multivariable analysis of patients > 60 years, ABVD correlated with better survival than CHOP (p = 0.027), and ABVD became more common over time among patients aged 61–70 years (p = 0.0206). Coinciding with the implementation of FDG‐PET/CT, the fraction of advanced‐stage disease increased in later calendar periods, also in the older patient group. Survival has improved in cHL patients > 60 years (p = 0.027), for whom ABVD seems superior to CHOP.

On models for the estimation of the excess mortality hazard in case of insufficiently stratified life tables

In cancer epidemiology using population-based data, regression models for the excess mortality hazard is a useful method to estimate cancer survival and to describe the association between prognosis factors and excess mortality. This method requires expected mortality rates from general population life tables: each cancer patient is assigned an expected (background) mortality rate obtained from the life tables, typically at least according to their age and sex, from the population they belong to. However, those life tables may be insufficiently stratified, as some characteristics such as deprivation, ethnicity, and comorbidities, are not available in the life tables for a number of countries. This may affect the background mortality rate allocated to each patient, and it has been shown that not including relevant information for assigning an expected mortality rate to each patient induces a bias in the estimation of the regression parameters of the excess hazard model. We propose two parametric corrections in excess hazard regression models, including a single-parameter or a random effect (frailty), to account for possible mismatches in the life table and thus misspecification of the background mortality rate. In an extensive simulation study, the good statistical performance of the proposed approach is demonstrated, and we illustrate their use on real population-based data of lung cancer patients. We present conditions and limitations of these methods and provide some recommendations for their use in practice.

Prediction of cancer survival for cohorts of patients most recently diagnosed using multi-model inference

Despite a large choice of models, functional forms and types of effects, the selection of excess hazard models for prediction of population cancer survival is not widespread in the literature. We propose multi-model inference based on excess hazard model(s) selected using Akaike information criteria or Bayesian information criteria for prediction and projection of cancer survival. We evaluate the properties of this approach using empirical data of patients diagnosed with breast, colon or lung cancer in 1990-2011. We artificially censor the data on 31 December 2010 and predict five-year survival for the 2010 and 2011 cohorts. We compare these predictions to the observed five-year cohort estimates of cancer survival and contrast them to predictions from an a priori selected simple model, and from the period approach. We illustrate the approach by replicating it for cohorts of patients for which stage at diagnosis and other important prognosis factors are available. We find that model-averaged predictions and projections of survival have close to minimal differences with the Pohar-Perme estimation of survival in many instances, particularly in subgroups of the population. Advantages of information-criterion based model selection include (i) transparent model-building strategy, (ii) accounting for model selection uncertainty, (iii) no a priori assumption for effects, and (iv) projections for patients outside of the sample.

Semiparametric estimation of the cure fraction in population-based cancer survival analysis

With rapid development in medical research, the treatment of diseases including cancer has progressed dramatically and those survivors may die from causes other than the one under study, especially among elderly patients. Motivated by the Surveillance, Epidemiology, and End Results (SEER) female breast cancer study, background mortality is incorporated into the mixture cure proportional hazards (MCPH) model to improve the cure fraction estimation in population-based cancer studies. Here, that patients are "cured" is defined as when the mortality rate of the individuals in diseased group returns to the same level as that expected in the general population, where the population level mortality is presented by the mortality table of the United States. The semiparametric estimation method based on the EM algorithm for the MCPH model with background mortality (MCPH+BM) is further developed and validated via comprehensive simulation studies. Real data analysis shows that the proposed semiparametric MCPH+BM model may provide more accurate estimation in population-level cancer study.

Using pseudo-observations for estimation in relative survival

A common goal in the analysis of the long-term survival related to a specific disease is to estimate a measure that is comparable between populations with different general population mortality. When cause of death is unavailable or unreliable, as for example in cancer registry studies, relative survival methodology is used-in addition to the mortality data of the patients, we use the data on the mortality of the general population. In this article, we focus on the marginal relative survival measure that summarizes the information about the disease-specific hazard. Under additional assumptions about latent times to death of each cause, this measure equals net survival. We propose a new approach to estimation based on pseudo-observations and derive two estimators of its variance. The properties of the new approach are assessed both theoretically and with simulations, showing practically no bias and a close to nominal coverage of the confidence intervals with the precise formula for the variance. The approximate formula for the variance has sufficiently good performance in large samples where the precise formula calculation becomes computationally intensive. Using bladder cancer data and simulations, we show that the behavior of the new approach is very close to that of the Pohar Perme estimator but has the important advantage of a simpler formula that does not require numerical integration and therefore lends itself more naturally to further extensions.

Trends in incidence, mortality and survival of testicular cancer patients in Belarus

INTRODUCTION

The objective of this study was to assess recent trends in incidence, mortality and relative survival (RS) in testicular cancer (TC) patients in Belarus and to provide international comparisons of our figures.

MATERIAL AND METHODS

We surveyed the Belarusian Cancer Registry for all male cases diagnosed with International Classification of Diseases for Oncology, third edition (ICD-O-3) topography code C62 between 1990 and 2015. Trends for incidence and mortality rates per 100,000 of the world standard population and annual percentage changes (APCs) were calculated. We also estimated the 1- and 5-year RS rates for the 1990-1998, 1999-2007 and 2008-2015 periods according to the Ederer II method. The RS estimates for the 2008-2015 period were age-standardized and compared with the published EUROCARE-5 data and SEER-18 database analysis.

RESULTS

A total of 2,500 and 2,439 cases were included into incidence and survival analyses, respectively. We found a significant increase in the TC age-standardized incidence rate (APC 2.6%) and a decline in the age-standardized mortality (APC -3.0%) over the study period. RS significantly increased in all patients` strata; a relative increase was more pronounced in advanced stages of seminoma and younger age groups. Nevertheless, the most recent figures of age-standardized RS including stage-specific estimates were generally worse than the European and SEER data.

CONCLUSIONS

We have found a significant increase in TC incidence in Belarus in recent years. Mortality has significantly declined with a corresponding increase in RS which, however, did not reach European or North American figures. Continued effort is required to improve the quality of management of TC patients in our country.

Comparison of model-building strategies for excess hazard regression models in the context of cancer epidemiology

BACKGROUND

Large and complex population-based cancer data are becoming broadly available, thanks to purposeful linkage between cancer registry data and health electronic records. Aiming at understanding the explanatory power of factors on cancer survival, the modelling and selection of variables need to be understood and exploited properly for improving model-based estimates of cancer survival.

METHOD

We assess the performances of well-known model selection strategies developed by Royston and Sauerbrei and Wynant and Abrahamowicz that we adapt to the relative survival data setting and to test for interaction terms.

RESULTS

We apply these to all male patients diagnosed with lung cancer in England in 2012 (N = 15,688), and followed-up until 31/12/2015. We model the effects of age at diagnosis, tumour stage, deprivation, comorbidity and emergency presentation, as well as interactions between age and all of the above. Given the size of the dataset, all model selection strategies favoured virtually the same model, except for a non-linear effect of age at diagnosis selected by the backward-based selection strategies (versus a linear effect selected otherwise).

CONCLUSION

The results from extensive simulations evaluating varying model complexity and sample sizes provide guidelines on a model selection strategy in the context of excess hazard modelling.

Net survival of patients with colorectal cancer: a comparison of two periods

BACKGROUND

The purpose of our analysis was to compare the results of treatment in patients who underwent resection for colorectal carcinoma.

METHODS

In the period from 1/1/1991 to 31/12/2000 1478 patients with colorectal carcinoma underwent potentially curative resection. We divided them into two 5-year period groups according to different treatment regimes. The 5-year net survival rate was estimated, where the net survival is the probability of survival derived solely from the cancer-specific hazard.

RESULTS

In a 10-year period, we resected 1478 patients. The 5-year net survival rate for R0-resected patients with colon cancer increased from 76.3 to 85.2% between the periods 1991-1995 and 1995-2000. The 5-year net survival rate for R0-resected patients with rectal cancer also increased from 67.5 to 73% in the same period.

CONCLUSION

A comparison of the 5-year net survival rate for R0-resected patients with colorectal cancer increased in the last period from 1995 to 2000 compared with the period from 1991 to 1995. In multivariate analysis, early stage at diagnosis and adjuvant chemotherapy was both associated with better net survival after surgery with curative intent. The improvement of net survival is potentially the result of combination of better surgical and adjuvant therapy.

Meta-analysis of long-term survival after elective endovascular or open repair of abdominal aortic aneurysm

BACKGROUND

Endovascular aneurysm repair (EVAR) has become the preferred strategy for elective repair of abdominal aortic aneurysm (AAA) for many patients. However, the superiority of the endovascular procedure has recently been challenged by reports of impaired long-term survival in patients who underwent EVAR. A systematic review of long-term survival following AAA repair was therefore undertaken.

METHODS

A systematic review was performed according to PRISMA guidelines. Articles reporting short- and/or long-term mortality of EVAR and open surgical repair (OSR) of AAA were identified. Pooled overall survival estimates (hazard ratios (HRs) with corresponding 95 per cent c.i. for EVAR versus OSR) were calculated using a random-effects model. Possible confounding owing to age differences between patients receiving EVAR or OSR was addressed by estimating relative survival.

RESULTS

Some 53 studies were identified. The 30-day mortality rate was lower for EVAR compared with OSR: 1·16 (95 per cent c.i. 0·92 to 1·39) versus 3·27 (2·71 to 3·83) per cent. Long-term survival rates were similar for EVAR versus OSR (HRs 1·01, 1·00 and 0·98 for 3, 5 and 10 years respectively; P = 0·721, P = 0·912 and P = 0·777). Correction of age inequality by means of relative survival analysis showed equal long-term survival: 0·94, 0·91 and 0·76 at 3, 5 and 10 years for EVAR, and 0·96, 0·91 and 0·76 respectively for OSR.

CONCLUSION

Long-term overall survival rates were similar for EVAR and OSR. Available data do not allow extension beyond the 10-year survival window or analysis of specific subgroups.

Incidence and survival in sinonasal carcinoma: a Danish population-based, nationwide study from 1980 to 2014

BACKGROUND

Sinonasal cancers are rare and comprise <1% of all malignancies. This study describes incidence and survival in sinonasal carcinomas in Denmark from 1980 to 2014.

METHODS

All patients registered in the Danish Cancer Registry in the period were included. Age-adjusted incidence rate, average annual percentage change, and relative survival were calculated. Age-period-cohort models were constructed.

RESULTS

1,720 patients with sinonasal carcinoma (median age 67 years, 63% males) were identified. There was no significant change in age-adjusted incidence; 0.70 in 1980 to 0.43 per 100,000 in 2014 (p > .05). Relative 5- and 10-year survival were 52% and 40% for men, 58% and 42% for women. An increase in 5-year survival from 1980 to 2014 from 46% to 65% (p < .05) was found. Nasal carcinomas had a significantly better relative survival compared to sinus carcinoma, as did squamous cell carcinomas when compared to neuroendocrine malignancies.

CONCLUSION

In Denmark between 1980 and 2014, the incidence of sinonasal carcinomas has been stable and the relative survival has increased significantly.

Explained variation of excess hazard models

The availability of longstanding collection of detailed cancer patient information makes multivariable modelling of cancer-specific hazard of death appealing. We propose to report variation in survival explained by each variable that constitutes these models. We adapted the ranks explained (RE) measure to the relative survival data setting, ie, when competing risks of death are accounted for through life tables from the general population. RE is calculated at each event time. We introduce weights for each death reflecting its probability to be a cancer death. RE varies between -1 and +1 and can be reported at given times in the follow-up and as a time-varying measure from diagnosis onward. We present an application for patients diagnosed with colon or lung cancer in England. The RE measure shows reasonable properties and is comparable in both relative and cause-specific settings. One year after diagnosis, RE for the most complex excess hazard models reaches 0.56, 95% CI: 0.54 to 0.58 (0.58 95% CI: 0.56-0.60) and 0.69, 95% CI: 0.68 to 0.70 (0.67, 95% CI: 0.66-0.69) for lung and colon cancer men (women), respectively. Stage at diagnosis accounts for 12.4% (10.8%) of the overall variation in survival among lung cancer patients whereas it carries 61.8% (53.5%) of the survival variation in colon cancer patients. Variables other than performance status for lung cancer (10%) contribute very little to the overall explained variation. The proportion of the variation in survival explained by key prognostic factors is a crucial information toward understanding the mechanisms underpinning cancer survival. The time-varying RE provides insights into patterns of influence for strong predictors.

Variation in treatment and survival of older patients with non-metastatic breast cancer in five European countries: a population-based cohort study from the EURECCA Breast Cancer Group

BACKGROUND

Older patients are poorly represented in breast cancer research and guidelines do not provide evidence based recommendations for this specific group. We compared treatment strategies and survival outcomes between European countries and assessed whether variance in treatment patterns may be associated with variation in survival.

METHODS

Population-based study including patients aged ≥ 70 with non-metastatic BC from cancer registries from the Netherlands, Belgium, Ireland, England and Greater Poland. Proportions of local and systemic treatments, five-year relative survival and relative excess risks (RER) between countries were calculated.

RESULTS

In total, 236,015 patients were included. The proportion of stage I BC receiving endocrine therapy ranged from 19.6% (Netherlands) to 84.6% (Belgium). The proportion of stage III BC receiving no breast surgery varied between 22.0% (Belgium) and 50.8% (Ireland). For stage I BC, relative survival was lower in England compared with Belgium (RER 2.96, 95%CI 1.30-6.72, P < .001). For stage III BC, England, Ireland and Greater Poland showed significantly worse relative survival compared with Belgium.

CONCLUSIONS

There is substantial variation in treatment strategies and survival outcomes in elderly with BC in Europe. For early-stage BC, we observed large variation in endocrine therapy but no variation in relative survival, suggesting potential overtreatment. For advanced BC, we observed higher survival in countries with lower proportions of omission of surgery, suggesting potential undertreatment.

On comparison of net survival curves

Background

Relative survival analysis is a subfield of survival analysis where competing risks data are observed, but the causes of death are unknown. A first step in the analysis of such data is usually the estimation of a net survival curve, possibly followed by regression modelling. Recently, a log-rank type test for comparison of net survival curves has been introduced and the goal of this paper is to explore its properties and put this methodological advance into the context of the field.

Methods

We build on the association between the log-rank test and the univariate or stratified Cox model and show the analogy in the relative survival setting. We study the properties of the methods using both the theoretical arguments as well as simulations. We provide an R function to enable practical usage of the log-rank type test.

Results

Both the log-rank type test and its model alternatives perform satisfactory under the null, even if the correlation between their p-values is rather low, implying that both approaches cannot be used simultaneously. The stratified version has a higher power in case of non-homogeneous hazards, but also carries a different interpretation.

Conclusions

The log-rank type test and its stratified version can be interpreted in the same way as the results of an analogous semi-parametric additive regression model despite the fact that no direct theoretical link can be established between the test statistics.

Electronic supplementary material

The online version of this article (doi:10.1186/s12874-017-0351-3) contains supplementary material, which is available to authorized users.

Worldwide comparison of survival from childhood leukaemia for 1995-2009, by subtype, age, and sex (CONCORD-2): a population-based study of individual data for 89 828 children from 198 registries in 53 countries

BACKGROUND

Global inequalities in access to health care are reflected in differences in cancer survival. The CONCORD programme was designed to assess worldwide differences and trends in population-based cancer survival. In this population-based study, we aimed to estimate survival inequalities globally for several subtypes of childhood leukaemia.

METHODS

Cancer registries participating in CONCORD were asked to submit tumour registrations for all children aged 0-14 years who were diagnosed with leukaemia between Jan 1, 1995, and Dec 31, 2009, and followed up until Dec 31, 2009. Haematological malignancies were defined by morphology codes in the International Classification of Diseases for Oncology, third revision. We excluded data from registries from which the data were judged to be less reliable, or included only lymphomas, and data from countries in which data for fewer than ten children were available for analysis. We also excluded records because of a missing date of birth, diagnosis, or last known vital status. We estimated 5-year net survival (ie, the probability of surviving at least 5 years after diagnosis, after controlling for deaths from other causes [background mortality]) for children by calendar period of diagnosis (1995-99, 2000-04, and 2005-09), sex, and age at diagnosis (<1, 1-4, 5-9, and 10-14 years, inclusive) using appropriate life tables. We estimated age-standardised net survival for international comparison of survival trends for precursor-cell acute lymphoblastic leukaemia (ALL) and acute myeloid leukaemia (AML).

FINDINGS

We analysed data from 89 828 children from 198 registries in 53 countries. During 1995-99, 5-year age-standardised net survival for all lymphoid leukaemias combined ranged from 10·6% (95% CI 3·1-18·2) in the Chinese registries to 86·8% (81·6-92·0) in Austria. International differences in 5-year survival for childhood leukaemia were still large as recently as 2005-09, when age-standardised survival for lymphoid leukaemias ranged from 52·4% (95% CI 42·8-61·9) in Cali, Colombia, to 91·6% (89·5-93·6) in the German registries, and for AML ranged from 33·3% (18·9-47·7) in Bulgaria to 78·2% (72·0-84·3) in German registries. Survival from precursor-cell ALL was very close to that of all lymphoid leukaemias combined, with similar variation. In most countries, survival from AML improved more than survival from ALL between 2000-04 and 2005-09. Survival for each type of leukaemia varied markedly with age: survival was highest for children aged 1-4 and 5-9 years, and lowest for infants (younger than 1 year). There was no systematic difference in survival between boys and girls.

INTERPRETATION

Global inequalities in survival from childhood leukaemia have narrowed with time but remain very wide for both ALL and AML. These results provide useful information for health policy makers on the effectiveness of health-care systems and for cancer policy makers to reduce inequalities in childhood cancer survival.

FUNDING

Canadian Partnership Against Cancer, Cancer Focus Northern Ireland, Cancer Institute New South Wales, Cancer Research UK, US Centers for Disease Control and Prevention, Swiss Re, Swiss Cancer Research foundation, Swiss Cancer League, and the University of Kentucky.

Worldwide comparison of ovarian cancer survival: Histological group and stage at diagnosis (CONCORD-2)

OBJECTIVE

Ovarian cancer comprises several histological groups with widely differing levels of survival. We aimed to explore international variation in survival for each group to help interpret international differences in survival from all ovarian cancers combined. We also examined differences in stage-specific survival.

METHODS

The CONCORD programme is the largest population-based study of global trends in cancer survival, including data from 60 countries for 695,932 women (aged 15-99years) diagnosed with ovarian cancer during 1995-2009. We defined six histological groups: type I epithelial, type II epithelial, germ cell, sex cord-stromal, other specific non-epithelial and non-specific morphology, and estimated age-standardised 5-year net survival for each country by histological group. We also analysed data from 67 cancer registries for 233,659 women diagnosed from 2001 to 2009, for whom information on stage at diagnosis was available. We estimated age-standardised 5-year net survival by stage at diagnosis (localised or advanced).

RESULTS

Survival from type I epithelial ovarian tumours for women diagnosed during 2005-09 ranged from 40 to 70%. Survival from type II epithelial tumours was much lower (20-45%). Survival from germ cell tumours was higher than that of type II epithelial tumours, but also varied widely between countries. Survival for sex-cord stromal tumours was higher than for the five other groups. Survival from localised tumours was much higher than for advanced disease (80% vs. 30%).

CONCLUSIONS

There is wide variation in survival between histological groups, and stage at diagnosis remains an important factor in ovarian cancer survival. International comparisons of ovarian cancer survival should incorporate histology.

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.

A reference relative time-scale as an alternative to chronological age for cohorts with long follow-up

Background

Epidemiologists have debated the appropriate time-scale for cohort survival studies; chronological age or time-on-study being two such time-scales. Importantly, assessment of risk factors may depend on the choice of time-scale. Recently, chronological or attained age has gained support but a case can be made for a ‘reference relative time-scale’ as an alternative which circumvents difficulties that arise with this and other scales. The reference relative time of an individual participant is the integral of a reference population hazard function between time of entry and time of exit of the individual. The objective here is to describe the reference relative time-scale, illustrate its use, make comparison with attained age by simulation and explain its relationship to modern and traditional epidemiologic methods.

Results

A comparison was made between two models; a stratified Cox model with age as the time-scale versus an un-stratified Cox model using the reference relative time-scale. The illustrative comparison used a UK cohort of cotton workers, with differing ages at entry to the study, with accrual over a time period and with long follow-up. Additionally, exponential and Weibull models were fitted since the reference relative time-scale analysis need not be restricted to the Cox model. A simulation study showed that analysis using the reference relative time-scale and analysis using chronological age had very similar power to detect a significant risk factor and both were equally unbiased. Further, the analysis using the reference relative time-scale supported fully-parametric survival modelling and allowed percentile predictions and mortality curves to be constructed.

Conclusions

The reference relative time-scale was a viable alternative to chronological age, led to simplification of the modelling process and possessed the defined features of a good time-scale as defined in reliability theory. The reference relative time-scale has several interpretations and provides a unifying concept that links contemporary approaches in survival and reliability analysis to the traditional epidemiologic methods of Poisson regression and standardised mortality ratios. The community of practitioners has not previously made this connection.

Global surveillance of cancer survival 1995-2009: analysis of individual data for 25,676,887 patients from 279 population-based registries in 67 countries (CONCORD-2)

BACKGROUND

Worldwide data for cancer survival are scarce. We aimed to initiate worldwide surveillance of cancer survival by central analysis of population-based registry data, as a metric of the effectiveness of health systems, and to inform global policy on cancer control.

METHODS

Individual tumour records were submitted by 279 population-based cancer registries in 67 countries for 25·7 million adults (age 15-99 years) and 75,000 children (age 0-14 years) diagnosed with cancer during 1995-2009 and followed up to Dec 31, 2009, or later. We looked at cancers of the stomach, colon, rectum, liver, lung, breast (women), cervix, ovary, and prostate in adults, and adult and childhood leukaemia. Standardised quality control procedures were applied; errors were corrected by the registry concerned. We estimated 5-year net survival, adjusted for background mortality in every country or region by age (single year), sex, and calendar year, and by race or ethnic origin in some countries. Estimates were age-standardised with the International Cancer Survival Standard weights.

FINDINGS

5-year survival from colon, rectal, and breast cancers has increased steadily in most developed countries. For patients diagnosed during 2005-09, survival for colon and rectal cancer reached 60% or more in 22 countries around the world; for breast cancer, 5-year survival rose to 85% or higher in 17 countries worldwide. Liver and lung cancer remain lethal in all nations: for both cancers, 5-year survival is below 20% everywhere in Europe, in the range 15-19% in North America, and as low as 7-9% in Mongolia and Thailand. Striking rises in 5-year survival from prostate cancer have occurred in many countries: survival rose by 10-20% between 1995-99 and 2005-09 in 22 countries in South America, Asia, and Europe, but survival still varies widely around the world, from less than 60% in Bulgaria and Thailand to 95% or more in Brazil, Puerto Rico, and the USA. For cervical cancer, national estimates of 5-year survival range from less than 50% to more than 70%; regional variations are much wider, and improvements between 1995-99 and 2005-09 have generally been slight. For women diagnosed with ovarian cancer in 2005-09, 5-year survival was 40% or higher only in Ecuador, the USA, and 17 countries in Asia and Europe. 5-year survival for stomach cancer in 2005-09 was high (54-58%) in Japan and South Korea, compared with less than 40% in other countries. By contrast, 5-year survival from adult leukaemia in Japan and South Korea (18-23%) is lower than in most other countries. 5-year survival from childhood acute lymphoblastic leukaemia is less than 60% in several countries, but as high as 90% in Canada and four European countries, which suggests major deficiencies in the management of a largely curable disease.

INTERPRETATION

International comparison of survival trends reveals very wide differences that are likely to be attributable to differences in access to early diagnosis and optimum treatment. Continuous worldwide surveillance of cancer survival should become an indispensable source of information for cancer patients and researchers and a stimulus for politicians to improve health policy and health-care systems.

FUNDING

Canadian Partnership Against Cancer (Toronto, Canada), Cancer Focus Northern Ireland (Belfast, UK), Cancer Institute New South Wales (Sydney, Australia), Cancer Research UK (London, UK), Centers for Disease Control and Prevention (Atlanta, GA, USA), Swiss Re (London, UK), Swiss Cancer Research foundation (Bern, Switzerland), Swiss Cancer League (Bern, Switzerland), and University of Kentucky (Lexington, KY, USA).

Follicular lymphoma in Sweden: nationwide improved survival in the rituximab era, particularly in elderly women: a Swedish Lymphoma Registry study

Treatment for follicular lymphoma (FL) improved with rituximab. In Sweden, first-line rituximab was gradually introduced between 2003 and 2007, with regional differences. The first national guidelines for FL were published in November 2007, recommending rituximab in first-line therapy. Using the population-based Swedish Lymphoma Registry, 2641 patients diagnosed with FL from 2000 to 2010 were identified and characterized by year and region of diagnosis, age (median, 65 years), gender (50% men), first-line therapy and clinical risk factors. Overall and relative survivals were estimated by calendar periods (2000-2002, 2003-2007 and 2008-2010) and region of diagnosis. With each period, first-line rituximab use and survival increased. Survival was superior in regions where rituximab was quickly adopted and inferior where slowly adopted. These differences were independent in multivariable analyses. Ten-year relative survival for patients diagnosed 2003-2010 was 92%, 83%, 78% and 64% in the age groups 18-49, 50-59, 60-69 and ⩾70, respectively. With increasing rituximab use, male sex emerged as an adverse factor. Survival improved in all patient categories, particularly in elderly women. The introduction and the establishment of rituximab have led to a nationwide improvement in FL survival. However, rituximab might be inadequately dosed in younger women and men of all ages.

Can cancer registries show whether treatment is contributing to survival increases for melanoma of the skin at a population level?

RATIONALE, AIMS AND OBJECTIVES

It is uncertain whether survival increases from melanoma recorded by some population registries include a treatment effect. The US Surveillance, Epidemiology and End Results (SEER) programme has good data quality control, large numbers of cases enabling high statistical precision and summary stage plus thickness, which we consider to be a best-case population registry scenario to investigate potential for a treatment effect. We have investigated SEER data to indicate whether survivals increases are fully attributable to earlier diagnosis and other non-treatment factors.

METHODS

Through relative survival regression, the effects of diagnostic period on 5-year excess mortality were investigated, adjusting for socio-demographic factors, lesion sub-site, histology, thickness and stage at diagnosis in 1990-2009 (n = 99 690 cases).

RESULTS

The reduction in excess mortality (95% confidence interval) between 1990-1999 and 2000-2009 was 31 (20-41)% for localised melanoma, 18 (12-22)% for regional melanoma and 3 (-5-10)% for melanomas with distant spread. Younger age was predictive of a greater percentage reduction. Treatment benefits are inferred from the higher survivals in 2000-2009 but uncertainty remains due to incomplete data to adjust for non-treatment factors and a lack of treatment data.

CONCLUSIONS

Registries should use new information systems to collect more complete data on stage, other prognostic indicators, co-morbidities and treatment, to provide more definitive and detailed information on population effects of cancer control.

Light smoking at base-line predicts a higher mortality risk to women than to men; evidence from a cohort with long follow-up

Background

There is conflicting evidence as to whether smoking is more harmful to women than to men. The UK Cotton Workers’ Cohort was recruited in the 1960s and contained a high proportion of men and women smokers who were well matched in terms of age, job and length of time in job. The cohort has been followed up for 42 years.

Methods

Mortality in the cohort was analysed using an individual relative survival method and Cox regression. Whether smoking, ascertained at baseline in the 1960s, was more hazardous to women than to men was examined by estimating the relative risk ratio women to men, smokers to never smoked, for light (1–14), medium (15–24), heavy (25+ cigarettes per day) and former smoking.

Results

For all-cause mortality relative risk ratios were 1.35 for light smoking at baseline (95% CI 1.07-1.70), 1.15 for medium smoking (95% CI 0.89-1.49) and 1.00 for heavy smoking (95% CI 0.63-1.61). Relative risk ratios for light smoking at baseline for circulatory system disease was 1.42 (95% CI 1.01 to 1.98) and for respiratory disease was 1.89 (95% CI 0.99 to 3.63). Heights of participants provided no explanation for the gender difference.

Conclusions

Light smoking at baseline was shown to be significantly more hazardous to women than to men but the effect decreased as consumption increased indicating a dose response relationship. Heavy smoking was equally hazardous to both genders. This result may help explain the conflicting evidence seen elsewhere. However gender differences in smoking cessation may provide an alternative explanation.

Comparison of the risk factors effects between two populations: two alternative approaches illustrated by the analysis of first and second kidney transplant recipients

BACKGROUND

Whereas the prognosis of second kidney transplant recipients (STR) compared to the first ones has been frequently analyzed, no study has addressed the issue of comparing the risk factor effects on graft failure between both groups.

METHODS

Here, we propose two alternative strategies to study the heterogeneity of risk factors between two groups of patients: (i) a multiplicative-regression model for relative survival (MRS) and (ii) a stratified Cox model (SCM) specifying the graft rank as strata and assuming subvectors of the explicatives variables. These developments were motivated by the analysis of factors associated with time to graft failure (return-to-dialysis or patient death) in second kidney transplant recipients (STR) compared to the first ones. Estimation of the parameters was based on partial likelihood maximization. Monte-Carlo simulations associated with bootstrap re-sampling was performed to calculate the standard deviations for the MRS.

RESULTS

We demonstrate, for the first time in renal transplantation, that: (i) male donor gender is a specific risk factor for STR, (ii) the adverse effect of recipient age is enhanced for STR and (iii) the graft failure risk related to donor age is attenuated for STR.

CONCLUSION

While the traditional Cox model did not provide original results based on the renal transplantation literature, the proposed relative and stratified models revealed new findings that are useful for clinicians. These methodologies may be of interest in other medical fields when the principal objective is the comparison of risk factors between two populations.

Excess mortality after curative surgery for colorectal cancer changes over time and differs for patients with colon versus rectal cancer

BACKGROUND

Improved management of colorectal cancer patients has resulted in better five-year survival for rectal cancer compared with colon cancer. We compared excess mortality rates in various time intervals after surgery in patients with colon and rectal cancer.

MATERIAL AND METHODS

We analysed all patients with curative resection of colorectal cancers reported in the Cancer Registry of Norway before (1994-1996) and after (2001-2003) national treatment guidelines were introduced. Excess mortality was analysed in different postoperative time intervals within the five-year follow-up periods for patients treated in 1994-1996 vs. 2001-2003.

RESULTS

A total of 11 437 patients that underwent curative resection were included. For patients treated from 1994 to 1996, excess mortality was similar in colon and rectal cancer patients in all time intervals. For those treated from 2001 to 2003, excess mortality was significantly lower in rectal cancer patients than in colon cancer patients perioperatively (in the first 60 days: excess mortality ratio = 0.46, p = 0.007) and during the first two postoperative years (2-12 months: excess mortality ratio = 0.54, p = 0.010; 1-2 years: excess mortality ratio = 0.60, p = 0.009). Excess mortality in rectal cancer patients was significantly greater than in colon cancer patients 4-5 years postoperatively (excess mortality ratio = 2.18, p = 0.003).

CONCLUSION

Excess mortality for colon and rectal cancer changed substantially after the introduction of national treatment guidelines. Short-term excess mortality rates was higher in colon cancer compared to rectal cancer for patients treated in 2001-2003, while excess mortality rates for rectal cancer patients was significantly higher later in the follow-up period. This suggests that future research should focus on these differences of excess mortality in patients curatively treated for cancer of the colon and rectum.

A class of transformation covariate regression models for estimating the excess hazard in relative survival analysis

Relative survival is the standard measure of excess mortality due to cancer in population-based cancer survival studies. In relative survival analysis, the observed hazard for cancer patients is the sum of the expected hazard for the general cancer-free population and the excess hazard associated with a cancer diagnosis. Previous models for relative survival analysis have assumed that the excess hazard rate is related to covariates by additive or multiplicative regression models. In this paper, a transformation covariate regression model is developed for estimation of the excess hazard rate, which includes both the additive and the multiplicative regression models as special cases. The baseline excess hazard rate and time-dependent hazard ratios can be approximated by means of regression splines, and the parameter estimates can be obtained using a standard statistical package. As is demonstrated through simulation, the proposed transformation hazards model provides a reasonably good fit to typical relative survival data. For illustration purposes, the sex difference in relative survival for lung and bronchus cancer patients is examined using data from population-based cancer registries (1973-2003).

Partitioning of excess mortality in population-based cancer patient survival studies using flexible parametric survival models

Background

Relative survival is commonly used for studying survival of cancer patients as it captures both the direct and indirect contribution of a cancer diagnosis on mortality by comparing the observed survival of the patients to the expected survival in a comparable cancer-free population. However, existing methods do not allow estimation of the impact of isolated conditions (e.g., excess cardiovascular mortality) on the total excess mortality. For this purpose we extend flexible parametric survival models for relative survival, which use restricted cubic splines for the baseline cumulative excess hazard and for any time-dependent effects.

Methods

In the extended model we partition the excess mortality associated with a diagnosis of cancer through estimating a separate baseline excess hazard function for the outcomes under investigation. This is done by incorporating mutually exclusive background mortality rates, stratified by the underlying causes of death reported in the Swedish population, and by introducing cause of death as a time-dependent effect in the extended model. This approach thereby enables modeling of temporal trends in e.g., excess cardiovascular mortality and remaining cancer excess mortality simultaneously. Furthermore, we illustrate how the results from the proposed model can be used to derive crude probabilities of death due to the component parts, i.e., probabilities estimated in the presence of competing causes of death.

Results

The method is illustrated with examples where the total excess mortality experienced by patients diagnosed with breast cancer is partitioned into excess cardiovascular mortality and remaining cancer excess mortality.

Conclusions

The proposed method can be used to simultaneously study disease patterns and temporal trends for various causes of cancer-consequent deaths. Such information should be of interest for patients and clinicians as one way of improving prognosis after cancer is through adapting treatment strategies and follow-up of patients towards reducing the excess mortality caused by side effects of the treatment.