Comparing Survival Extrapolation within All-Cause and Relative Survival Frameworks by Standard Parametric Models and Flexible Parametric Spline Models Using the Swedish Cancer Registry.

BACKGROUND: In health technology assessment, restricted mean survival time and life expectancy are commonly evaluated. Parametric models are typically used for extrapolation. Spline models using a relative survival framework have been shown to estimate life expectancy of cancer patients more reliably; however, more research is needed to assess spline models using an all-cause survival framework and standard parametric models using a relative survival framework. AIM: To assess survival extrapolation using standard parametric models and spline models within relative survival and all-cause survival frameworks. METHODS: From the Swedish Cancer Registry, we identified patients diagnosed with 5 types of cancer (colon, breast, melanoma, prostate, and chronic myeloid leukemia) between 1981 and 1990 with follow-up until 2020. Patients were categorized into 15 cancer cohorts by cancer and age group (18-59, 60-69, and 70-99 y). We right-censored the follow-up at 2, 3, 5, and 10 y and fitted the parametric models within an all-cause and a relative survival framework to extrapolate to 10 y and lifetime in comparison with the observed Kaplan-Meier survival estimates. All cohorts were modeled with 6 standard parametric models (exponential, Weibull, Gompertz, log-logistic, log-normal, and generalized gamma) and 3 spline models (on hazard, odds, and normal scales). RESULTS: For predicting 10-y survival, spline models generally performed better than standard parametric models. However, using an all-cause or a relative survival framework did not show any distinct difference. For lifetime survival, extrapolating from a relative survival framework agreed better with the observed survival, particularly using spline models. CONCLUSIONS: For extrapolation to 10 y, we recommend spline models. For extrapolation to lifetime, we suggest extrapolating in a relative survival framework, especially using spline models. HIGHLIGHTS: For survival extrapolation to 10 y, spline models generally performed better than standard parametric models did. However, using an all-cause or a relative survival framework showed no distinct difference under the same parametric model.Survival extrapolation to lifetime within a relative survival framework agreed well with the observed data, especially using spline models.Extrapolating parametric models within an all-cause survival framework may overestimate survival proportions at lifetime; models for the relative survival approach may underestimate instead.

Vulvar cancer incidence and net survival in Sweden 1960 to 2019: A population-based national study.

INTRODUCTION: Vulvar cancer is a rare gynecological cancer affecting mostly older women. The aim of this population-based study was to investigate the incidence and net survival of vulvar cancer in Swedish women from 1960 to 2019. MATERIAL AND METHODS: Data were retrieved from the mandatory Swedish Cancer Registry consisting of all women diagnosed with vulvar cancer between 1960 and 2019. Only women with a morphologically verified diagnosis of vulvar cancer were included. The individuals were then further matched with the Swedish Death Registry up until May 31, 2020. RESULTS: In total, 8499 women were included with the following morphologies: squamous cell carcinoma 7250 (85.8%), malignant melanoma 539 (6.4%), adenocarcinoma 401 (4.8%) and other: 259 (3.1%). More than 50% of vulvar cancer cases occurred in women aged between 65 and 84 years of age. The 5-year age-standardized net survival increased from 53.0% (95% confidence interval [CI] 48.9-57.5) in 1960 to 72.1% (95% CI 68.8-75.5) in 2019. The proportion of adenocarcinoma among all cases increased from 2.0% to 8.7% between the 1960s and 2010s and an increase in age-standardized 5-year net survival was found for adenocarcinoma. CONCLUSIONS: The age-standardized incidence of vulvar cancer cases in Sweden was stable between 1960 and 2019. During the study period, an increase in adenocarcinoma and a decrease in malignant melanoma cases was found. Five-year net survival increased by 20 percent units during the study period. For squamous cell carcinoma, an increased age-specific 5-year net survival was observed for all age groups, apart for women aged ≥85.

Comparing the diagnostic accuracy, reading time, and inter-rater agreement of breast MRI abbreviated and full protocols: a multi-reader study.

BACKGROUND: Earlier studies have shown that abbreviated protocol magnetic resonance imaging (AB-MRI) has similar diagnostic accuracy as the full protocol (Full MRI). PURPOSE: To compare the diagnostic accuracy, reading time, and inter-rater agreement of AB-MRI to Full MRI among women without known increased familial risk of breast cancer or prior biopsy. MATERIAL AND METHODS: In total, 395 MRI examinations were included in this study. Three readers were blinded to all patient information. The AB-MRI and Full MRI were read separately and in a different random order for each of the readers. Scores 1-2 were considered test negative while scores 3-5 were test positive. A positive reference test was the diagnosis of malignancy; a negative reference test was the absence of a diagnosis of breast cancer within a two-year follow-up. We used a generalized estimating equations approach to compare sensitivity and specificity between the two protocols. We used t-tests to compare the average reading time and Krippendorff's alpha to compare inter-rater agreement. RESULTS: MRI examinations of 395 women (median age=56 years) were evaluated. For AB-MRI and Full MRI, respectively, the sensitivity was 93.0% (95% CI=90.6-95.0) vs. 92.0% (95% CI=89.4-94.1), the specificity was 91.7% (95% CI=90.3-92.9) vs. 94.3% (95% CI=93.2-95.3), average reading time was 67 vs. 126 s, and the inter-rater agreement 0.79 vs. 0.83. The difference in sensitivity was not statistically significant (P=0.840), but the difference in specificity was significant (P=0.003). CONCLUSION: AB-MRI has similar sensitivity, but somewhat lower specificity. The average reading time for the abbreviated protocol is lower, as is inter-rater agreement.

Improving communication of cancer survival statistics-feasibility of implementing model-based algorithms in routine publications.

BACKGROUND: Routine reporting of cancer patient survival is important, both to monitor the effectiveness of health care and to inform about prognosis following a cancer diagnosis. A range of different survival measures exist, each serving different purposes and targeting different audiences. It is important that routine publications expand on current practice and provide estimates on a wider range of survival measures. We examine the feasibility of automated production of such statistics. METHODS: We used data on 23 cancer sites obtained from the Cancer Registry of Norway (CRN). We propose an automated way of estimating flexible parametric relative survival models and calculating estimates of net survival, crude probabilities, and loss in life expectancy across many cancer sites and subgroups of patients. RESULTS: For 21 of 23 cancer sites, we were able to estimate survival models without assuming proportional hazards. Reliable estimates of all desired measures were obtained for all cancer sites. DISCUSSION: It may be challenging to implement new survival measures in routine publications as it can require the application of modeling techniques. We propose a way of automating the production of such statistics and show that we can obtain reliable estimates across a range of measures and subgroups of patients.

Increased incidence and improved survival in endometrial cancer in Sweden 1960-2014: a population-based registry survey.

BACKGROUND: An investigation of trends of incidence and net survival (NS) for endometrial cancer in Sweden. METHODS: Morphologically verified endometrial carcinoma diagnosed 1960 to 2014 were collected from the nation-wide Swedish Cancer Registry. Endometrial cancer patients were assessed with regards to time trends for incidence and 54,825 cases remained for survival analyses. Cases diagnosed 1995 to 2014 were categorized according to detailed morphology and from 2005 to 2014 FIGO stage was also categorized. RESULTS: There was a trend of increasing incidence of endometrial carcinoma for women above 55 years of age. NS was improved at 5- and 10-year follow-up. The 5-year net survival in 2010-2014 was 86%. The most prominent improvement in NS was found in the elderly women above 75 years of age. CONCLUSIONS: This study observed increased incidence of endometrial cancer in Sweden from 1960 to 2014. The progress in diagnostics and treatment, seem to have improved the net survival, especially in elderly women.

Current clinical practice in the management of phyllodes tumors of the breast: an international cross-sectional study among surgeons and oncologists.

PURPOSE: Phyllodes tumors of the breast are rare fibroepithelial lesions that are classified as benign, borderline or malignant. There is little consensus on best practice for the work-up, management, and follow-up of patients with phyllodes tumors of the breast, and evidence-based guidelines are lacking. METHODS: We conducted a cross-sectional survey of surgeons and oncologists with the aim to describe current clinical practice in the management of phyllodes tumors. The survey was constructed in REDCap and distributed between July 2021 and February 2022 through international collaborators in sixteen countries across four continents. RESULTS: A total of 419 responses were collected and analyzed. The majority of respondents were experienced and worked in a university hospital. Most agreed to recommend a tumor-free excision margin for benign tumors, increasing margins for borderline and malignant tumors. The multidisciplinary team meeting plays a major role in the treatment plan and follow-up. The vast majority did not consider axillary surgery. There were mixed opinions on adjuvant treatment, with a trend towards more liberal regiments in patients with locally advanced tumors. Most respondents preferred a five-year follow-up period for all phyllodes tumor types. CONCLUSIONS: This study shows considerable variation in clinical practice managing phyllodes tumors. This suggests the potential for overtreatment of many patients and the need for education and further research targeting appropriate surgical margins, follow-up time and a multidisciplinary approach. There is a need to develop guidelines that recognize the heterogeneity of phyllodes tumors.

Survival in Swedish patients with systemic sclerosis: a nationwide population-based matched cohort study.

OBJECTIVES: To conduct the first-ever nationwide, population-based cohort study investigating survival patterns of all patients with incident SSc in Sweden compared with matched individuals from the Swedish general population. METHODS: We used the National Patient Register to identify patients with incident SSc diagnosed between 2004 and 2015 and the Total Population Register to identify comparators (1:5), matched on sex, birth year and residential area. We followed them until death, emigration or the end of 2016. Follow-up of the general population comparators started the same date as their matched patients were included. We estimated all-cause survival using the Kaplan-Meier method, crude mortality rates and hazard ratios (HRs) using flexible parametric models. RESULTS: We identified 1139 incident patients with SSc and 5613 matched comparators. The median follow-up was 5.0 years in patients with SSc and 6.0 years for their comparators. During follow-up, 268 deaths occurred in patients with SSc and 554 in their comparators. The 5-year survival was 79.8% and the 10-year survival was 67.7% among patients with SSc vs 92.9% and 84.8%, respectively, for the comparators (P < 0.0001). The mortality rate in patients with SSc was 42.1 per 1000 person-years and 15.8 per 1000 person-years in their comparators, corresponding to an HR of 3.7 (95% CI 2.9, 4.7) at the end of the first year of follow-up and 2.0 (95% CI 1.4, 2.8) at the end of the follow-up period. CONCLUSION: Despite advances in understanding the disease and in diagnostic methods over the past decades, survival is still severely impacted in Swedish patients diagnosed with SSc between 2004 and 2015.

Modelling multiple time-scales with flexible parametric survival models.

BACKGROUND: There are situations when we need to model multiple time-scales in survival analysis. A usual approach in this setting would involve fitting Cox or Poisson models to a time-split dataset. However, this leads to large datasets and can be computationally intensive when model fitting, especially if interest lies in displaying how the estimated hazard rate or survival change along multiple time-scales continuously. METHODS: We propose to use flexible parametric survival models on the log hazard scale as an alternative method when modelling data with multiple time-scales. By choosing one of the time-scales as reference, and rewriting other time-scales as a function of this reference time-scale, users can avoid time-splitting of the data. RESULT: Through case-studies we demonstrate the usefulness of this method and provide examples of graphical representations of estimated hazard rates and survival proportions. The model gives nearly identical results to using a Poisson model, without requiring time-splitting. CONCLUSION: Flexible parametric survival models are a powerful tool for modelling multiple time-scales. This method does not require splitting the data into small time-intervals, and therefore saves time, helps avoid technological limitations and reduces room for error.

Time Trends for Incidence and Net Survival of Cervical Cancer in Sweden 1960-2014-A Nationwide Population-Based Study.

BACKGROUND: The aim was to investigate time trends for incidence and long-term net survival in the morphologic subtypes and stages of cervical cancer in Sweden during the period 1960 to 2014. METHODS: Women with invasive cervical cancer were identified through the Swedish Cancer Registry. Incidence and net survival were calculated according to morphology, age at diagnosis, and FIGO stage at diagnosis. RESULTS: In total, 29,579 cases of invasive cervical cancer between 1960 and 2014 were included. The age-standardized incidence for squamous cell carcinoma (SCC) decreased until 2000; thereafter, the incidence rate stagnated, and a small increase was found in 2014. The incidence of adenocarcinoma continuously increased. The age-standardized 5-year net survival increased. However, decreasing net survival with increasing age was found. A higher stage at diagnosis showed a worse net survival. SCC and adenocarcinoma did not statistically differ as regards net survival in the last years of the study. CONCLUSIONS: Age-standardized 5-year net survival improved between 1960 and 2014. A positive trend for short- and long-term net survival was seen for women ages 18 to 64 years but long-term net survival for women ≥75 years decreased. In this study, age and FIGO stage at diagnosis were found to be important prognostic factors in determining net survival. The morphologies, SCC, and adenocarcinoma did not statistically differ as regards net survival in the last years of the study. IMPACT: This study demonstrates longitudinal data on cervical cancer in Sweden for over 50 years with sub analyses on morphology, age, and stage at diagnosis.

Reproductive history, as measured by parity, age at first birth and sex of offspring, and cancer-specific survival after a haematological malignancy.

BACKGROUND: Overall, women have better cancer-specific survival than men following haematological malignancies. The effect of reproductive factors on prognosis in women remains unknown and population-based studies are needed. MATERIAL AND METHODS: A nationwide cohort of 21,237 Swedish women with a recorded haematological malignancy at ages 18-69 years was identified in the Swedish Cancer Register 1970-2018. Pre-diagnosis childbirths for each woman were linked to the Swedish Multigeneration Register. Net survival and excess hazard ratios for parity, age at first birth, time since the latest birth, and sex of offspring were estimated using flexible parametric models adjusted for age, year, and educational level. RESULTS: In unadjusted analyses, parity (p = 0.0012) and high age at first birth (p < 0.0001) were associated with better survival. After co-adjustments for reproductive factors and confounders, the associations were attenuated. The adjusted association with parity was mainly observed among women aged above 40 years at diagnosis (p = 0.0033). The associations with reproductive factors were non-significant across subtypes of haematological malignancy. There was a tendency of higher excess mortality for an increasing number of boys compared to girls, although only significant for women with three or more children (p = 0.0126). CONCLUSION: Reproductive factors were in part associated with survival following diagnosis of a haematological malignancy. However, the effect sizes were small with inconsistent association patterns, and thus reproductive factors may only partly contribute to the survival advantage of women over men.

Cancer outcome research - a European challenge Part II: Opportunities and priorities.

In Part I of our review of cancer outcome research, we analysed pros and cons of various measures relevant to quantifying the burden of cancer. Based on our recommendations in Part I, we now discuss in Part II opportunities and priorities in four areas of outcome research: primary prevention; early detection screening; treatment; and quality-of-life assessment. We recommend the establishment of an infrastructure that facilitates high-quality research in these areas: (a) progress in primary prevention can be assessed most directly by monitoring cancer incidence although the interpretation of temporal trends is notoriously confounded by numerous factors that complicate causal inference. (b) preventive screening, with the aim to prevent advanced disease, appears to work well in in some tumours but not in others. It will require randomized control trials (RCTs) to quantify benefits and harms although conclusive studies are increasingly difficult to undertake. We therefore propose learning screening programmes (randomization at the time of rolling out population-based programmes) as the most feasible approach. (c) New therapeutic interventions tailored to the individual patient often require assessment in RCTs with rather complex and dynamic structure, making their design and analyses increasingly challenging but also more suited to be executed as academic, PI-initiated trials. (d) We next discuss assessment of quality-of-life aspects. Quality of life is a neglected component in outcome research with an urgent need for development, validation and standardization. We finally recommend four initiatives that would pave the way for a valid and informative assessment of the goals for improved cancer control in Europe as defined by the European Academy of Cancer Sciences.

Survival, causes of death, and the prognostic role of comorbidities in chronic lymphocytic leukemia in the pre-ibrutinib era: A population-based study.

OBJECTIVE: To evaluate temporal trends in survival and causes of death in patients with chronic lymphocytic leukemia (CLL) in a nationwide study. METHODS: The cohort consisted of 13,009 Swedish CLL patients diagnosed 1982-2013. Relative survival (RS) and excess mortality rate ratios (EMRR) with 95% confidence intervals (95% CIs) were estimated using flexible parametric survival models. Cause-specific hazard ratios (HRs) were estimated for the linear effect of 10-year increase in year of diagnosis. RESULTS: The excess mortality decreased comparing 2003-2013 to 1982-1992 (EMRR = 0.53, 95% CI 0.48-0.58). The 5-year RS increased between 1982 and 2012 for patients >51 years at diagnosis and improved for patients ≤51 years after 2002. The rate of CLL-specific deaths decreased over time (HR = 0.78, 95% CI 0.75-0.81). Compared to patients with no comorbidity, patients with 1 and 2+ Charlson Comorbidity Index points had HR = 1.35 (95% CI 1.25-1.45) and HR = 1.47 (95% CI 1.37-1.57) for CLL-related mortality, respectively. CONCLUSION: Survival in CLL patients improved in the era of chemoimmunotherapy, and this was largely explained by reduced CLL-related mortality. The increased rate of CLL-related mortality in patients with comorbidities emphasizes the importance of the newer and better tolerated targeted therapy.

Comparison of Relative Survival and Cause-Specific Survival in Men With Prostate Cancer According to Age and Risk Category: A Nationwide, Population-Based Study.

Net survival, estimated in a relative survival (RS) or cause-specific survival (CSS) framework, is a key measure of the effectiveness of cancer management. We compared RS and CSS in men with prostate cancer (PCa) according to age and risk category, using Prostate Cancer data Base Sweden, including 168,793 men younger than age 90 years, diagnosed 1998-2016 with PCa. RS and CSS were compared according to age and risk category based on TNM (tumor, nodes, and metastases) stage, Gleason score, and prostate-specific antigen level. Each framework requires assumptions that are unlikely to be appropriate for PCa. Ten-year RS was substantially higher than CSS in men aged 80-89 with low-risk PCa: 125% (95% confidence interval: 113, 138) versus 85% (95% confidence interval: 82, 88). In contrast, RS and CSS were similar for men under age 70 and for all men with regional or distant metastases. Both RS and CSS produce biased estimates of net survival for men with low- and intermediate-risk PCa, in particular for men over 80. Due to biases, net survival is overestimated in analysis of RS but underestimated in analysis of CSS. These results highlight the importance of evaluating the underlying assumptions for each method, because the "true" net survival is expected to lie between the limits of RS and CSS.

Cancer outcomes research-a European challenge: measures of the cancer burden.

In a mission that aims to improve cancer control throughout Europe, the European Academy of Cancer Sciences has defined two key indicators of progress: within one to two decades, overall cancer-specific 10-year survival should reach 75%, and in each country, overall cancer mortality rates should be convincingly declining. To lay the ground for assessment of progress and to promote cancer outcomes research in general, we have reviewed the most common population-based measures of the cancer burden. We emphasize the complexities and complementary approaches to measure cancer survival and the novel opportunities for improved assessment of quality of life. We propose that: incidence and mortality rates are standardized to the European population; net survival is used as the measure of prognosis but with proper adjustments for confounding when temporal trends in overall cancer survival are assessed; and cancer-specific quality of life is measured by a combination of existing questionnaires and utilizes emerging communication technologies. We conclude that all measures are important and that a meaningful interpretation also requires a deep understanding of the larger clinical and public health context.

Incidence and survival of epithelial ovarian, fallopian tube, peritoneal, and undesignated abdominal/pelvic cancers in Sweden 1960-2014: A population-based cohort study.

BACKGROUND: Despite improved surgical and oncological treatment, ovarian cancer continues to be the most lethal of the gynecologic malignancies. We aimed to analyze survival trends in epithelial ovarian cancer with regard to age, tumor site, and morphology in Sweden 1960 to 2014. METHODS: A nationwide population-based study was conducted using data from the Swedish Cancer Registry on 46,350 women aged 18 or older with a diagnosis of epithelial ovarian, fallopian tube, peritoneal, or undesignated abdominal/pelvic cancer 1960 to 2014. Analyses of age-standardized incidence and relative survival (RS) were performed and time trends modelled according to age, tumor site, and morphology. RESULTS: Overall incidence of ovarian, tubal, peritoneal, and undesignated abdominal/pelvic cancers declined since 1980. Median age at diagnosis increased. Serous carcinoma increased in incidence. RS at 1, 2 and 5 years from diagnosis improved since 1960, although not for the youngest and the oldest patients. Ten-year RS did not improve. The best RS was found for fallopian tube cancer and the worst RS for undesignated abdominal/pelvic cancer. Among the morphologic subgroups, endometrioid carcinoma had the best RS. CONCLUSIONS: Survival in epithelial ovarian, tubal, peritoneal, and undesignated abdominal/pelvic cancers in Sweden has improved over the last six decades. Advances in epithelial ovarian cancer treatment have extended life for the first 5 years from diagnosis but 10-year survival remains poor.

Temporal improvements noted in life expectancy of patients with colorectal cancer; a Dutch population-based study.

OBJECTIVE: Specific survival estimates are needed for the increasing number of colorectal cancer (CRC) survivors. The aim of this population-based study was to determine conditional loss in expectation of life (LEL) due to CRC. STUDY DESIGN AND SETTING: All surgically treated patients with CRC registered in the Netherlands Cancer Registry with stage I-III between 1990-2016, were included (N = 203,216). Estimates of conditional LEL were predicted using flexible parametric models and the total life years lost due to cancer were estimated. RESULTS: LEL decreased with older age and patients with rectal cancer or higher disease stage had highest LEL. In 2010, LEL for sixty-year old male and female patients was 2 vs. 2, 4 vs. 4, and 7 vs. 8 years for colon cancer, and 2 vs. 2, 4 vs. 5 and 7 vs. 8 years for rectal cancer, respectively. Conditional LEL in patients with CRC decreased during follow-up. Patients with combined stage I-III colon and rectal cancer in 2010 lost an estimated 18,628 and 11,336 life years. CONCLUSION: This study quantified the impact of CRC on patient's life expectancy, both on individual and population level and demonstrated temporal improvements in CRC survival. These results provide meaningful information that can be used during follow-up.

A multistate model incorporating estimation of excess hazards and multiple time scales.

As cancer patient survival improves, late effects from treatment are becoming the next clinical challenge. Chemotherapy and radiotherapy, for example, potentially increase the risk of both morbidity and mortality from second malignancies and cardiovascular disease. To provide clinically relevant population-level measures of late effects, it is of importance to (1) simultaneously estimate the risks of both morbidity and mortality, (2) partition these risks into the component expected in the absence of cancer and the component due to the cancer and its treatment, and (3) incorporate the multiple time scales of attained age, calendar time, and time since diagnosis. Multistate models provide a framework for simultaneously studying morbidity and mortality, but do not solve the problem of partitioning the risks. However, this partitioning can be achieved by applying a relative survival framework, allowing us to directly quantify the excess risk. This article proposes a combination of these two frameworks, providing one approach to address (1) to (3). Using recently developed methods in multistate modeling, we incorporate estimation of excess hazards into a multistate model. Both intermediate and absorbing state risks can be partitioned and different transitions are allowed to have different and/or multiple time scales. We illustrate our approach using data on Hodgkin lymphoma patients and excess risk of diseases of the circulatory system, and provide user-friendly Stata software with accompanying example code.

Risk of infections in patients with myeloproliferative neoplasms-a population-based cohort study of 8363 patients.

Infections are a common complication in patients with many hematologic malignancies, however, whether patients with myeloproliferative neoplasms (MPN) also are at an increased risk of infections is largely unknown. To assess the risk of serious infections, we performed a large population-based matched cohort study in Sweden including 8 363 MPN patients and 32,405 controls using high-quality registers between the years 1992-2013 with follow-up until 2015. The hazard ratio (HR) of any infection was 2.0 (95% confidence interval 1.9-2.0), of bacterial infections 1.9 (1.8-2.0), and of viral infections 2.1 (1.9-2.3). One of the largest risk increases was that of sepsis, HR 2.6 (2.4-2.9). The HR of any infection was highest in primary myelofibrosis 3.7 (3.2-4.1), and significantly elevated in all MPN subtypes; 1.7 (1.6-1.8) in polycythemia vera and 1.7 (1.5-1.8) in essential thrombocythemia. There was no significant difference in risk of infections between untreated patients and patients treated with hydroxyurea or interferon-α during the years 2006-2013. These novel findings of an overall increased risk of infections in MPN patients, irrespective of common cytoreductive treatments, suggest the increased risk of infection is inherent to the MPN.