Explained variation in survival analysis

MASH: MEDIATION ANALYSIS OF SURVIVAL OUTCOME AND HIGH-DIMENSIONAL OMICS MEDIATORS WITH APPLICATION TO COMPLEX DISEASES

Environmental exposures such as cigarette smoking influence health outcomes through intermediate molecular phenotypes, such as the methylome, transcriptome, and metabolome. Mediation analysis is a useful tool for investigating the role of potentially high-dimensional intermediate phenotypes in the relationship between environmental exposures and health outcomes. However, little work has been done on mediation analysis when the mediators are high-dimensional and the outcome is a survival endpoint, and none of it has provided a robust measure of total mediation effect. To this end, we propose an estimation procedure for Mediation Analysis of Survival outcome and High-dimensional omics mediators (MASH) based on sure independence screening for putative mediator variable selection and a second-moment-based measure of total mediation effect for survival data analogous to the R 2 measure in a linear model. Extensive simulations showed good performance of MASH in estimating the total mediation effect and identifying true mediators. By applying MASH to the metabolomics data of 1919 subjects in the Framingham Heart Study, we identified five metabolites as mediators of the effect of cigarette smoking on coronary heart disease risk (total mediation effect, 51.1%) and two metabolites as mediators between smoking and risk of cancer (total mediation effect, 50.7%). Application of MASH to a diffuse large B-cell lymphoma genomics data set identified copy-number variations for eight genes as mediators between the baseline International Prognostic Index score and overall survival.

Follow-up strategy of radiofrequency ablation for papillary thyroid microcarcinoma: defining a response-to-ablation system

OBJECTIVE

To define a response-to-ablation system based on dynamic risk stratification proposed by the 2015 American Thyroid Association guidelines for predicting clinical outcomes and guiding follow-up strategies for patients with low-risk papillary thyroid microcarcinoma (PTMC) who underwent radiofrequency ablation (RFA).

METHODS

This retrospective study reviewed patients with low-risk PTMC who underwent RFA between 2014 and 2018. We classified patients into three groups based on their response to therapy at the 1-year follow-up: complete, indeterminate, and incomplete. The primary endpoints were local tumor progression (LTP) and disease-free survival (DFS).

RESULTS

Among the 748 patients (mean age, 43.7 years ± 9.8; 586 women), 4.0% (30/748) had LTP during a median follow-up of 5 years. The response was complete in 80.2% (600/748) of the patients, indeterminate in 18.1% (135/748), and incomplete in 1.7% (13/748). The LTP rate in the final follow-up was 1% (6/600), 8.1% (11/135), and 100% (13/13), respectively. The risk of LTP was significantly different in the incomplete response group (HR, 1825.82; 95% CI: 458.27, 7274.36; p < 0.001) and indeterminate response group (HR, 8.12; 95% CI: 2.99, 22.09; p < 0.001) than in the complete response group. There were significant differences in DFS among groups (p < 0.001). The proportion of variation explained and C-index of the system was high (27.66% and 0.79, respectively).

CONCLUSIONS

We defined a response-to-ablation system that provides a new paradigm for the management of patients with PTMC who underwent RFA. Our data confirm that the system can effectively predict the risk of LTP and guide ongoing follow-up recommendations.

KEY POINTS

• The response-to-ablation system can classify patients with low-risk PTMC who underwent RFA into complete, indeterminate, or incomplete response categories. • Results suggest that, in this population, this system can identify three separate cohorts of patients who have significantly different clinical outcomes. • The response-to-ablation system will help better tailor the ongoing follow-up recommendations.

A dynamic analysis of empirical survival outcomes after pancreatectomy for pancreatic ductal adenocarcinoma

BACKGROUND

Survival outcomes after pancreatectomy for pancreatic ductal adenocarcinoma may be biased by right-censoring. We herein analyzed a large dataset with no censored events for up to 5 years and dynamically investigated the impact of known prognostic factors, accounting for unobserved tumor characteristics.

METHODS

Consecutive patients undergoing pancreatectomy from 2000 to July 2015 were included. The 1- to 5-year empirical survival rates were calculated, and factors associated with long-term survival (≥5 years) were analyzed using multivariable models. Dynamic analyses of survival and recurrence were conducted through landmarking, and the contribution of unobserved heterogeneity was estimated using frailty models.

RESULTS

The study population included 1,048 patients. The median follow-up was 30.4 months in the whole cohort and 97.2 months in survivors. The median survival was 30.4 months, with empirical 1- to 5-year rates of 85.5%, 59.6%, 43.2%, 32.1%, and 27.5%. A favorable pathological profile was associated with 5-year survival, albeit 25.7% of long-survivors received an R1 resection, and 28.8% had N2 disease. The median recurrence-free survival was 17.2 months. At landmark analyses, baseline prognostic lost strength over time, with no independent predictors of survival being identified in the sets of patients alive at 4 and 5 years. There was a significant amount of unobserved heterogeneity in the early postoperative period.

CONCLUSION

The 5-year post-pancreatectomy empirical survival was 27.5%. Dynamic analyses showed a time-varying structure of prognostic variables and a substantial impact of unobserved tumor characteristics that may drive the disease course under the selective pressure of surgical resection and adjuvant chemotherapy.

Measures of explained variation under the mixture cure model for survival data

Explained variation is well understood under linear regression models and has been extended to models for survival data. In this article, we consider the mixture cure models. We propose two approaches to define explained variation under the mixture cure models, one based on the Kullback-Leibler information gain and the other based on residual sum of squares. We show that the proposed measures have desired properties as measures of explained variation, similar to those under other regression models. A simulation study is conducted to demonstrate the properties of the proposed measures. They are also applied to real data analyses to illustrate the use of explained variation.

The Impact of Image Acquisition Parameters and ComBat Harmonization on the Predictive Performance of Radiomics: A Renal Cell Carcinoma Model

Radiomics, one of the potential methods for developing clinical biomarker, is one of the exponentially growing research fields. In addition to its potential, several limitations have been identified in this field, and most importantly the effects of variations in imaging parameters on radiomic features (RFs). In this study, we investigate the potential of RFs to predict overall survival in patients with clear cell renal cell carcinoma, as well as the impact of ComBat harmonization on the performance of RF models. We assessed the robustness of the results by performing the analyses a thousand times. Publicly available CT scans of 179 patients were retrospectively collected and analyzed. The scans were acquired using different imaging vendors and parameters in different medical centers. The performance was calculated by averaging the metrics over all runs. On average, the clinical model significantly outperformed the radiomic models. The use of ComBat harmonization, on average, did not significantly improve the performance of radiomic models. Hence, the variability in image acquisition and reconstruction parameters significantly affect the performance of radiomic models. The development of radiomic specific harmonization techniques remain a necessity for the advancement of the field.

Perioperative Blood Transfusion in Head and Neck Cancer Revisited

OBJECTIVE

To reevaluate the frequency of perioperative blood transfusion, transfusion triggers, and survival impact in patients with incident, surgically treated head and neck cancer (HNC) in restrictive transfusion regimens.

METHODS

Retrospective analysis of surgically treated patients with incident HNC with and without perioperative blood transfusion between 2008 and 2019 at the Department of Otorhinolaryngology, Head and Neck Surgery, Medical University of Innsbruck, according to the department's clinical Head and Neck Tumor Registry.

RESULTS

Of the 590 patients included, perioperative transfusions were administered in 6.3% (n = 37, transfusion group). Following multivariable logistic regression, likelihood of blood transfusions was increased in patients with poorer general health conditions (ASA score III/IV; OR 3.7; 95% CI 1.9-8.6; p = 0.002), hemoglobin <12.5 g/dL (OR 2.7; 95% CI 1.1-6.4; p = 0.03), longer duration of surgery (OR 1.006 per minute of surgery time; 95% CI 1.003-1.008; p < 0.001), and negative p16 status (OR 5.3; 95% CI = 1.1-25; p = 0.03). Based on 14 matching variables related to survival and perioperative blood transfusion, a control group of 37 matching patients without perioperative transfusion was identified. Using univariate analysis, overall survival in transfusion and control groups did not differ significantly (p = 0.25). After adjusting for four parameters with limited matching accuracy (Chi square p < 0.2) in Cox regression analysis, a transfusion related hazard ratio close to 1 (HR 0.92; 95% CI 0.34-2.51; p = 0.87) was observed.

CONCLUSION

Considering current restrictive transfusion regimens and general transfusion risks, the administration of blood products in HNC patients during the perioperative period is not associated with additional oncologic hazard.

LEVEL OF EVIDENCE

III Laryngoscope, 2022.

C-index regression for recurrent event data

Recurrent event data analysis plays an important role in many fields, e.g., medicine, social science, and economics. While the existing approaches under the proportional rates or mean model yield poor performance when the underlying model is misspecified, we propose a novel model-free approach by introducing a lower bound on the concordance index (C-Index). We develop an estimation method through deriving a continuous lower bound on the C-Index based on the log-sigmoid function and also provide a variable selection procedure in high dimensional settings. Under both low and high dimensional settings, simulation results show that the proposed methods outperform the gamma frailty recurrent event model when the proportional mean assumption is violated. Moreover, an application to the hospital readmission dataset shows results in line with previous studies and a higher C-Index value further assures model decency.

Validation of dynamic risk stratification and impact of BRAF in risk assessment of thyroid cancer, a nation-wide multicenter study

INTRODUCTION

The dynamic risk stratification (DRS) is a relatively new system in thyroid cancer that considers the response to primary treatment to improve the initial risk of recurrence. We wanted to validate DRS system in a nationwide multicenter study and explore if the incorporation of BRAFV600E into DRS helps to better categorize and predict outcomes.

MATERIALS AND METHODS

Retrospective study of 685 patients from seven centers between 1991 and 2016, with a mean age of 48 years and a median follow-up time of 45 months (range 23-77). The overall BRAFV600E prevalence was 53.4%. We classified patients into four categories based on DRS ('excellent', 'indeterminate', 'biochemical incomplete', and 'structural incomplete' response). Cox regression was used to calculate adjusted hazard ratios (AHR) and proportions of variance explained (PVEs).

RESULTS

We found 21.6% recurrences and 2.3% cancer-related deaths. The proportion of patients that developed recurrence in excellent, indeterminate, biochemical incomplete and structural incomplete response to therapy was 1.8%, 54%, 91.7% and 96.2% respectively. Considering the outcome at the end of the follow up, patients showed no evidence of disease (NED) in 98.2, 52, 33.3 and 25.6% respectively. Patients in the structural incomplete category were the only who died (17.7%). Because they have similar outcomes in terms of NED and survival, we integrated the indeterminate and biochemical incomplete response into one unique category creating the 3-tiered DRS system. The PVEs of the AJCC/TNM staging, ATA risk classification, 4-tiered DRS, and 3-tiered DRS to predict recurrence at five years were 21%, 25%, 57% and 59% respectively. BRAFV600E was significantly associated with biochemical incomplete response (71.1 vs 28.9%) (HR 2.43; 95% CI, 1.21 to 5.23; p=0.016), but not with structural incomplete response or distant metastases. BRAF status slightly changes the AHR values of the DRS categories but is not useful for different risk grouping.

CONCLUSIONS

This is the first multicenter study to validate the 4-tiered DRS system. Our results also show that the 3-tiered DRS system, by integrating indeterminate and biochemical incomplete response into one unique category, may simplify response to therapy keeping the system accurate. BRAF status does not provide any additional benefit to DRS.

Molecular classification of follicular thyroid carcinoma based on TERT promoter mutations

Follicular thyroid carcinoma (FTC) has different clinicopathological characteristics than papillary thyroid carcinoma. However, there are no independent systems to predict cancer-specific survival (CSS) in FTC. Telomerase reverse transcriptase (TERT) promoter mutations are associated with tumor aggressiveness. Thus, it could be a potential prognostic marker. The aim of this study was to refine the CSS risk prediction using TERT promoter mutations in combination with the fourth edition of World Health Organization (WHO 2017) morphological classification. We investigated 77 FTC patients between August 1995 and November 2020. Cox regression was used to calculate hazard ratios to derive alternative groups. Disease-free survival (DFS) and CSS predictability were compared using Proportion of variation explained (PVE) and C-index. CSS was significantly different in encapsulated angioinvasive (EA)-FTC patients stratified by TERT promoter mutations [wild-type (WT-TERT) vs. mutant (M-TERT); P < 0.001] but not in minimally invasive (MI)-FTC and widely invasive (WI)-FTC patients (P = 0.691 and 0.176, respectively). We defined alternative groups as follows: Group 1 (MI-FTC with WT-TERT and M-TERT; EA-FTC with WT-TERT), Group 2 (WI-FTC with WT-TERT), and Group 3 (EA-FTC with M-TERT; WI-FTC with M-TERT). Both PVE (22.44 vs. 9.63, respectively) and C-index (0.831 vs. 0.731, respectively) for CSS were higher in the alternative groups than in the WHO 2017 groups. Likewise, both PVE (27.1 vs. 14.9, respectively) and C-index (0.846 vs. 0.794, respectively) for DFS were also higher in the alternative groups than in the WHO 2017 groups. Alternative group harmonizing of the WHO 2017 classification and TERT promoter mutations is effective in predicting CSS in FTC patients, thereby improving DFS predictability.

Microscopic Extrathyroidal Extension Results in Increased Rate of Tumor Recurrence and Is an Independent Predictor of Patient’s Outcome in Middle Eastern Papillary Thyroid Carcinoma

Background

Papillary Thyroid Cancer (PTC) is the most common endocrine malignancy, with recurrence rate as high as 30%. A great deal of controversy surrounds the significance of microscopic extrathyroidal extension (m-ETE) as a prognostic factor. The most recent edition (8^th) of American Joint Committee on Cancer (AJCC) staging system has removed m-ETE from the definition of pT3, which suggests that m-ETE may lack prognostic impact in PTC patients. Moreover, data about m-ETE prevalence and clinical impact on Middle Eastern PTC remains unknown. We therefore investigate the prevalence of m-ETE and its clinico-pathological correlation and prognostic impact in Middle Eastern PTC. We also compared the AJCC 7^th and 8^th staging systems and their prognostic performance.

Methods

PTCs from 1430 consecutive adult (> 18 years) patients from single tertiary care hospital were included in this study. A retrospective analysis of PTC patients’ survival and recurrence were compared between AJCC 8^th and AJCC 7^th staging systems using Proportion of Variation Explained (PVE) and Harrell’s C-index.

Results

Median follow up of the study cohort was 9.3 years. 31.2% (446/1430) of patients had m-ETE. In the overall cohort, m-ETE was associated with multiple adverse features such as older age (p < 0.0001), male sex (p = 0.0245), tall cell variant (p < 0.0001), bilateral tumors (p < 0.0001), multifocality (p < 0.0001), lymphovascular invasion (p < 0.0001), lymph node metastasis (p < 0.0001), distant metastasis (p = 0.0166), tumor recurrence (p < 0.0001), radioactive iodine refractoriness (p < 0.0001), BRAF mutation (p < 0.0001) and reduced recurrence-free survival (RFS; HR = 1.75; 95% CI = 1.30 – 2.35; p < 0.0001) irrespective of tumor size. Of the 611 patients with T3 disease based on AJCC 7^th edition, 359 (58.8%) were down-staged in AJCC 8^th edition classification. Overall, the prognostic performance of AJCC 8^th edition was inferior to AJCC 7^th on the basis of lower PVE (3.04% vs. 3.73%) and lower C-index (0.40 vs. 0.48).

Conclusions

In Middle Eastern PTC, m-ETE is significantly associated with compromised survival and acts as an independent predictor of RFS. Given these findings, m-ETE should be included in the thyroid cancer treatment guidelines.

Fitting and Cross-Validating Cox Models to Censored Big Data With Missing Values Using Extensions of Partial Least Squares Regression Models

Fitting Cox models in a big data context -on a massive scale in terms of volume, intensity, and complexity exceeding the capacity of usual analytic tools-is often challenging. If some data are missing, it is even more difficult. We proposed algorithms that were able to fit Cox models in high dimensional settings using extensions of partial least squares regression to the Cox models. Some of them were able to cope with missing data. We were recently able to extend our most recent algorithms to big data, thus allowing to fit Cox model for big data with missing values. When cross-validating standard or extended Cox models, the commonly used criterion is the cross-validated partial loglikelihood using a naive or a van Houwelingen scheme -to make efficient use of the death times of the left out data in relation to the death times of all the data. Quite astonishingly, we will show, using a strong simulation study involving three different data simulation algorithms, that these two cross-validation methods fail with the extensions, either straightforward or more involved ones, of partial least squares regression to the Cox model. This is quite an interesting result for at least two reasons. Firstly, several nice features of PLS based models, including regularization, interpretability of the components, missing data support, data visualization thanks to biplots of individuals and variables -and even parsimony or group parsimony for Sparse partial least squares or sparse group SPLS based models, account for a common use of these extensions by statisticians who usually select their hyperparameters using cross-validation. Secondly, they are almost always featured in benchmarking studies to assess the performance of a new estimation technique used in a high dimensional or big data context and often show poor statistical properties. We carried out a vast simulation study to evaluate more than a dozen of potential cross-validation criteria, either AUC or prediction error based. Several of them lead to the selection of a reasonable number of components. Using these newly found cross-validation criteria to fit extensions of partial least squares regression to the Cox model, we performed a benchmark reanalysis that showed enhanced performances of these techniques. In addition, we proposed sparse group extensions of our algorithms and defined a new robust measure based on the Schmid score and the R coefficient of determination for least absolute deviation: the integrated R Schmid Score weighted. The R-package used in this article is available on the CRAN, http://cran.r-project.org/web/packages/plsRcox/index.html. The R package bigPLS will soon be available on the CRAN and, until then, is available on Github https://github.com/fbertran/bigPLS.

Proposal of a New Prognostic Model for Differentiated Thyroid Cancer with TERT Promoter Mutations

The role of telomerase reverse transcriptase (TERT) promoter mutations as an independent poor prognostic factor in differentiated thyroid cancer (DTC) patients is well known, but there is no prognostic system that combines the TERT promoter mutation status with tumor-node-metastasis (TNM) stage to predict cancer-specific survival (CSS). A total of 393 patients with pathologically confirmed DTC after thyroidectomy were enrolled. After incorporating wild-type TERT and mutant TERT with stages I, II, and III/IV of the AJCC TNM system 8th edition (TNM-8), we generated six combinations and calculated 10-year and 15-year CSS and adjusted hazard ratios (HRs) for cancer-related death using Cox regression. Then, a new mortality prediction model termed TNM-8T was derived based on the CSS and HR of each combination in the four groups. Of the 393 patients, there were 27 (6.9%) thyroid cancer-related deaths during a median follow-up of 14 years. Patients with a more advanced stage had a lower survival rate (10-year CSS for TNM-8T stage 1, 2, 3, and 4: 98.7%, 93.5%, 77.3%, and 63.0%, respectively; p < 0.001). TNM-8T showed a better spread of CSS (p < 0.001) than TNM-8 (p = 0.002) in the adjusted survival curves. The C-index for mortality risk predictability was 0.880 (95% CI, 0.665-0.957) in TNM-8T and 0.827 (95% CI, 0.622-0.930) in TNM-8 (p < 0.001). TNM-8T, a new prognostic system that incorporates the TERT mutational status into TNM-8, showed superior predictability to TNM-8 in the long-term survival of DTC patients.

A critical analysis of the 8th edition TNM staging for head and neck cutaneous squamous cell carcinoma with lymph node metastases and comparison to N1S3 stage and ITEM risk score: A multicenter study

BACKGROUND AND OBJECTIVES

We performed a critical analysis of the 8th edition American Joint Committee on Cancer (AJCC) staging for head and neck cutaneous squamous cell carcinoma (HNcSCC) with nodal metastases and compared the performance to the N1S3 and ITEM systems.

METHODS

Multicenter study of 990 patients with metastatic HNcSCC treated with curative intent. The end points of interest were disease-specific (DSS) and overall survival (OS). Model fit was evaluated using Harrell's Concordance Index (C-index), proportion of variation explained (PVE), Akaike information criterion, and Bayesian information criterion.

RESULTS

N1S3 and ITEM demonstrated good distribution into risk categories in contrast to the AJCC system, which classified the majority (90.6%) of patients as N2-3 and Stage IV due to the high rate of extranodal extension. The N2c and N3a categories appeared redundant. There was considerable discordance between systems in risk allocation on an individual patient basis. N1S3 was the best performed (DSS: C-index 0.62, PVE 10.9%; OS: C-index 0.59, PVE 4.5%), albeit with relatively poor predictive value.

CONCLUSIONS

The AJCC N category and tumor node metastasis stage have poor patient distribution and predictive performance in HNcSCC. The AJCC stage, N1S3, and ITEM score all provide limited prognostic information based on objective measures highlighting the need to develop a staging system specific to HNcSCC.

Nodal metastasis size predicts disease-free survival in cutaneous head and neck squamous cell carcinoma involving the parotid but not cervical nodes

BACKGROUND

The 8th edition American Joint Committee on Cancer nodal (N) staging of cutaneous squamous cell carcinoma of the head and neck (cSCCHN) is largely based on lymph node metastasis size, despite conflicting data in the literature. This study aimed to investigate the prognostic significance of largest node size in cSCCHN.

METHODS

Retrospective analysis of 94 patients undergoing curative-intent treatment for nodal cSCCHN with surgery ± radiotherapy at Liverpool Hospital, Sydney, Australia was conducted. Survival outcomes were assessed using multivariate Cox regression. The primary end point was disease-free survival (DFS). Objective measures of model performance were used in exploratory analyses to identify optimal size thresholds for predicting survival.

RESULTS

Nodal metastasis size significantly predicted DFS on multivariate analysis (hazard ratio 1.24; 95% confidence interval 1.06-1.46; P = 0.008). This prognostic impact occurred predominantly in parotid metastases (hazard ratio 1.27; 95% confidence interval 1.07-1.51; P = 0.006); each 1 cm increase in size increased the risk of recurrence or death by 27%, irrespective of the number of involved nodes. In parotid metastases, size thresholds of ≤3, 3-4.5 and >4.5 cm optimized prognostic discrimination. Extranodal extension (ENE) was associated with decreased DFS in nodes ≤3 cm in size (P = 0.025), but not in those >3 cm (P = 0.744).

CONCLUSION

Size is an important prognostic factor in cSCCHN with parotid metastases, with optimal thresholds of ≤3, >3-4.5 and >4.5 cm. The prognostic impact of ENE was seen only in nodal metastases ≤3 cm in size. These results may have important implications for node size thresholds and inclusion of ENE in the American Joint Committee on Cancer N staging categories.

Prognostic value of the 8th edition American Joint Commission Cancer nodal staging system for patients with head and neck cutaneous squamous cell carcinoma: A multi-institutional study

BACKGROUND

The 8th edition American Joint Committee on Cancer staging manual (AJCC8) introduced a separate staging system for head and neck cutaneous squamous cell carcinoma (HNcSCC) which parallels mucosal SCC and incorporates extranodal extension (ENE). This study aims to evaluate its prognostic utility.

METHODS

Univariate analysis of 1146 patients with metastatic HNcSCC from four Australian cancer centers was performed according to both AJCC 7th (AJCC7) and the 8th editions.

RESULTS

AJCC8 increased classification of 924 (80.6%) patients to either pN2a or pN3b and 341 patients (29.8%) from stage III to IV compared to AJCC7. The disease-specific survival (DSS) was not significantly different between pN1, pN2 or pN3a categories per AJCC8. Estimates of model performance for the AJCC8 pN staging revealed modest predictive capacity (Harrell's C of 0.62 for DSS).

CONCLUSIONS

The risk stratification according to pN classification of AJCC8 staging system performed poorly as a prognostic indicator.

Strap muscle invasion in differentiated thyroid cancer does not impact disease-specific survival: a population-based study

The American Joint Committee on Cancer (AJCC) 8^th TNM staging system of differentiated thyroid cancer defines gross strap muscle invasion as T3b stage. However, the impact of strap muscle invasion on disease-specific survival (DSS) remains controversial. To elucidate the survival impact of strap muscle invasion of any degree in thyroid cancers, the Surveillance, Epidemiology, and End Results (SEER) database (1973–2018) was queried for thyroid cancer only patients on July 2019 (n = 19,914). The Cox proportional hazard analysis with multivariable adjustment revealed that strap muscle invasion was not a significant factor for DSS in tumors equal to or smaller than 40 mm (hazard ratio (HR) = 1.620 [confidence interval (CI) 0.917 – 2.860]; p = 0.097). The competing risk analysis with multivariable adjustment showed that strap muscle invasion did not significantly impact DSS regardless of tumor size or cause of death (cancer-caused death (Subdistribution HR (SDHR) = 1.567 [CI 0.984 – 2.495]; p = 0.059); deaths to other causes (SDHR = 1.155 [CI 0.842 – 1.585]; p = 0.370). A “modified” staging schema discarding strap muscle invasion as a T stage criterion showed better 10-year DSS distinction between T stages. The modified staging schema may better reflect cancer-caused death risk and may prevent potential overstaging.

Number of nodal metastases and the American Joint Committee on cancer staging of head and neck cutaneous squamous cell carcinoma: A multicenter study

OBJECTIVES

We aimed to determine if the number of nodal metastases is an independent predictor of survival in HNcSCC, whether it provides additional prognostic information to the AJCC N and TNM stage and identify optimal cut-points for risk stratification.

MATERIALS AND METHODS

Retrospective multi-institutional cohort study of patients with parotid and/or cervical nodal metastases from HNcSCC treated with curative intent by surgery ± adjuvant therapy. The impact of number of nodal metastases on disease-specific and overall survival was assessed using multivariate Cox regression. Optimal cut-points for prognostic discrimination modelled using the AIC, BIC, C-index and PVE.

RESULTS

The study cohort included 1128 patients, with 962 (85.3%) males, median age of 72.9 years (range: 18-100 years) and median follow-up 3.4 years. Adjuvant radiotherapy was administered to 946 (83.9%) patients. Based on objective measures of model performance, number of nodal metastases was classified as 1-2 (N = 816), 3-4 (N = 162) and ≥5 (N = 150) nodes. In multivariate analyses, the risk of disease-specific mortality progressively increased with 3-4 nodes (HR, 1.58; 95% CI: 1.03-2.42; p = 0.036) and ≥5 nodes (HR, 2.91; 95% CI: 1.99-4.25; p < 0.001) with similar results for all-cause mortality. This simple categorical variable provided superior prognostic information to the TNM stage.

CONCLUSION

Increasing number of nodal metastases is an independent predictor of mortality in HNcSCC, with categorization as 1-2, 3-4 and ≥5 nodes optimizing risk stratification and providing superior prognostic information to TNM stage. These findings may aid in the development of future staging systems as well as identification of high-risk patients in clinical trials.

Modification of the Tumor-Node-Metastasis Staging System for Differentiated Thyroid Carcinoma by Considering Extra-Thyroidal Extension and Lateral Cervical Lymph Node Metastasis

BACKGROUND

Concerns have arisen about the classification of extra-thyroidal extension (ETE) and lateral cervical lymph node metastasis (N1b) in the 8th edition of the tumor-node-metastasis staging system (TNM-8). This study evaluated the prognostic validity of a modified-TNM staging system, focusing on ETE and N1b, in differentiated thyroid carcinoma (DTC) patients.

METHODS

This multicenter retrospective cohort study included 4,878 DTC patients from five tertiary hospitals. In the modified-TNM, T3b in TNM-8 was down-staged to T2, and stage II was subdivided into stages IIA and IIB. Older patients with N1b were reclassified as stage IIB.

RESULTS

The modified-TNM resulted in staging migration in 540 patients (11%) classified as stage II according to the TNM-8, with 75 (14%), 381 (71%), and 84 patients (16%) classified as stages I, IIA, and IIB, respectively. The 10-year disease-specific survival (DSS) rates in patients classified as stages I, II, III, and IV by TNM-8 were 99.8%, 95.9%, 81.0%, and 41.6%, respectively. The DSS rates of patients classified as stages I, IIA, IIB, III, and IV according to the modified-TNM were 99.8%, 96.4%, 93.3%, 81.0%, and 41.6%, respectively. DSS curves between stages on TNM-8 (P<0.001) and modified-TNM (P<0.001) differed significantly, but the modified-TNM discriminated better than TNM-8. The proportions of variation explained values of TNM-8 and modified-TNM were 6.3% and 6.5%, respectively.

CONCLUSION

Modification of the TNM staging system focusing on ETE and N1b could improve the prediction of DSS in patients with DTC. Further researches are needed to validate the prognostic accuracy of this modified-TNM staging system.

The binary presence or absence of lymph node metastasis or extrathyroidal extension is not associated with survival in papillary thyroid cancers: Implications for staging systems

BACKGROUND

The characteristics of diagnosed papillary thyroid cancer (PTC) have changed over time with the increasing trend of early diagnosis, and the survival impact of conventional prognostic factors such as lymph node metastasis (LNM) and extrathyroidal extension (ETE) is controversial. We investigated PTC prognostic factors for overall survival (OS) and disease specific survival (DSS), focusing on LNM, ETE, and their implications for PTC staging systems.

METHODS

We assessed prognostic factors for OS and DSS in a nationwide sample of Korean PTC patients (N = 5192, median follow-up 121 months) using Cox regression. The binary presence or absence of LNM and ETE, as well as other measures of LNM and ETE, were examined for their survival impact. We also evaluated the relative performance of PTC staging systems before and after revising the staging criteria for LNM and ETE.

RESULTS

The binary presence of LNM or ETE was not a prognostic factor for OS or DSS, nor were other various measures of LNM. However, the extent of ETE as none, microscopic, or gross independently influenced survival (OS hazard ratio for gross vs. none: 3.28, 95% confidence interval (CI) 1.97-5.46; DSS hazard ratio for gross vs. none: 3.75, 95% CI 1.59-8.81). The performance of PTC staging systems improved when the extent of ETE and/or location of LNM were used as staging components.

CONCLUSION

The extent of ETE and/or location of LNM may be better survival indicators than their binary presence or absence, and we propose staging criteria revisions to pertinent staging systems to better reflect the contemporary PTC population.

Quantifying degrees of necessity and of sufficiency in cause-effect relationships with dichotomous and survival outcomes

We suggest measures to quantify the degrees of necessity and of sufficiency of prognostic factors for dichotomous and for survival outcomes. A cause, represented by certain values of prognostic factors, is considered necessary for an event if, without the cause, the event cannot develop. It is considered sufficient for an event if the event is unavoidable in the presence of the cause. Necessity and sufficiency can be seen as the two faces of causation, and this symmetry and equal relevance are reflected by the suggested measures. The measures provide an approximate, in some cases an exact, multiplicative decomposition of explained variation as defined by Schemper and Henderson for censored survival and for dichotomous outcomes.

The measures, ranging from zero to one, are simple, intuitive functions of unconditional and conditional probabilities of an event such as disease or death. These probabilities often will be derived from logistic or Cox regression models; the measures, however, do not require any particular model.

The measures of the degree of necessity implicitly generalize the established attributable fraction or risk for dichotomous prognostic factors and dichotomous outcomes to continuous prognostic factors and to survival outcomes. In a setting with multiple prognostic factors, they provide marginal and partial results akin to marginal and partial odds and hazard ratios from multiple logistic and Cox regression.

Properties of the measures are explored by an extensive simulation study. Their application is demonstrated by three typical real data examples.

Impact of the change from the seventh to eighth edition of the AJCC TNM classification of malignant tumours and comparison with the MACIS prognostic scoring system in non-medullary thyroid cancer

Background

In 2018, AJCC TNM staging changed for differentiated (DTC) and anaplastic (ATC) thyroid carcinoma. The impact of this change on mortality rates was investigated and compared with the MACIS prognostic score.

Methods

Analysis of a prospective database of DTC/ATC was undertaken. Patients were staged according to TNM7 and TNM8 criteria, and MACIS scores calculated. Five‐year disease‐specific mortality rates were determined. Proportions were compared with Fisher's exact and χ² goodness‐of‐fit tests.

Results

Between August 2002 and December 2016, 310 patients had primary surgery for thyroid cancer. After exclusions, 159 patients (154 DTC, 5 ATC) remained to be studied. The MACIS score was less than 6 in 105 patients (66·0 per cent), 6–6·99 in 19 (11·9 per cent), 7–7·99 in 14 (8·8 per cent) and 8 or more in 21 (13·2 per cent), with corresponding disease‐specific 5‐year mortality rates of 0, 5, 14 and 86 per cent. For TNM7 the distribution was stage I in 53·5 per cent (85 patients), stage II in 10·1 per cent (16), stage III in 14·5 per cent (23) and stage IV in 22·0 per cent (35), and differed from that for TNM8: 76·7 per cent (122), 10·7 per cent (17), 4·4 per cent (7) and 8·2 per cent (13) respectively (P < 0·001). Overall disease‐specific 5‐year mortality rates by stage for TNM7 versus TNM8 were: stage I, 0 of 85 versus 3 of 100 (P = 0·251); stage II, 0 of 16 versus 6 of 16 (P = 0·018); stage III, 3 of 23 versus 2 of 7 (P = 0·565); stage IV, 20 of 32 versus 11 of 11 (P = 0·020).

Conclusion

Compared with TNM7, TNM8 downstaged more patients to stage I and accurately reflected worse prognosis for stage IV disease. TNM8 is an inferior predictor of mortality compared with MACIS.

Refining the tumor-node-metastasis staging system for individualized treatment of differentiated thyroid carcinoma

BACKGROUND

Patients with differentiated thyroid carcinoma (DTC) are staged according to the single age cut point in addition to anatomic extent. A novel staging system is needed to properly show the character and prognosis of DTC by considering age as a continuous variable. We aimed to refine stage and prognostic groups of the eighth edition tumor-node-metastasis (TNM-8) staging system for DTC and to suggest a possible revision.

METHODS

We conducted a retrospective data abstraction study of patients with newly diagnosed DTC who were treated at one of two tertiary referral centres in Seoul, Korea between 1994 and 2005. We used recursive partitioning analysis to derive a new staging classification (TNM-RPA) and compared its prediction of cancer-specific survival with that of TNM-8.

RESULTS

The cohort comprised 6342 patients with DTC who were followed up for a median of 11.4 years. Higher TNM-RPA groups were associated with increased risk of death (10-year cancer-specific survival for stages IA, IB, IIA, IIB, III, and IV: 99.6%, 98.1%, 93.0%, 92.4%, 75.1%, and 56.6%, respectively; P < 0.001). The C-index values were 0.869 (95% CI, 0.833-0.905) for the TNM-RPA and 0.819 (0.789-0.850) for TNM-8. The proportions of variance explained for the ability of the TNM-RPA and TNM-8 stages to predict cancer-specific survival were 7.1% and 5.7%, respectively.

CONCLUSION

This study presents a RPA-based TNM stage groupings that incorporate multiple age cutoffs and essential anatomic information, which can be conveniently used to facilitate the individual prediction of long-term cancer-specific survival in patients with DTC.

Clinical Validation of the Prognostic Stage Groups of the Eighth-Edition TNM Staging for Medullary Thyroid Carcinoma

CONTEXT

Despite advances in thyroid cancer staging systems, considerable controversy about the current staging system for medullary thyroid carcinoma (MTC) continues.

OBJECTIVE

We aimed to evaluate the prognostic performance of the current eighth edition of the American Joint Committee on Cancer (AJCC)/Union for International Cancer Control TNM staging system (TNM-8) and the alternative proposed prognostic stage groups based on recursive partitioning analysis (TNM-RPA).

DESIGN, SETTING, AND PATIENTS

We retrospectively analyzed 182 patients with MTC treated at a single tertiary Korean hospital between 1995 and 2015.

INTERVENTIONS AND MAIN OUTCOME MEASURES

Survival analysis was conducted according to TNM-8 and TNM-RPA. The area under the receiver-operating characteristic curve (AUC), the proportion of variation explained (PVE), and the Harrell concordance index (C-index) were used to evaluate predictive performance.

RESULTS

Under TNM-8, only two (1.1%) patients were downstaged compared with the seventh edition of the AJCC TNM staging system (TNM-7). The AUC at 10 years, PVE, and C-index were 0.679, 8.7%, and 0.744 for TNM-7 and 0.681, 8.9%, and 0.747 for TNM-8, respectively. Under TNM-RPA, 104 (57.14%) patients were downstaged compared with TNM-8. TNM-RPA had better prognostic performance with respect to cancer-specific survival (AUC at 10 years, 0.750; PVE, 20.9%; C-index, 0.881).

CONCLUSIONS

The predictive performance of the revised TNM-8 in patients with MTC has not changed despite its modification from TNM-7. The proposed changes in TNM-RPA were statistically valid and may present a more reproducible system that better estimates cancer-specific survival of individual patients.

Survival in Differentiated Thyroid Cancer: Comparing the AJCC Cancer Staging Seventh and Eighth Editions

BACKGROUND

The American Joint Committee on Cancer Cancer Staging System for differentiated thyroid cancer has been recently revised. The purpose of this study was to compare survival prognostication between the seventh and eighth editions.

MATERIALS AND METHODS

We retrospectively reviewed 2579 differentiated thyroid cancer patients who underwent surgery at MD Anderson Cancer Center between 2000 and 2015. Disease-specific survival (DSS) and overall survival were estimated using the Kaplan-Meier method and compared using log rank test. The effect of potential predictor was estimated using Cox proportional hazards model. Power of survival prediction was estimated using Harrell's C concordance index (C-index), and predictive capacities for DSS were estimated using proportion of variance explained (PVE).

RESULTS

Revision of tumor-node-metastasis (TNM) categories, age cutoff, and stage reassignment in the 8^th edition caused reclassification of overall stage compared with the seventh edition. The proportion of patients in stage I and II increased from 62% to 83% and 5% to 12%, respectively, while the proportion of patients in stage III and IV decreased from 20% to 2% and 14% to 3%, respectively. Ten-year DSS for stages I-IV based on the seventh edition were 100%, 97.5%, 98.3%, and 82.6%, respectively, while 10-year DSS for the corresponding stage in the eighth edition were 99.8%, 88.3%, 72.4%, and 71.9%, respectively. In multivariate analysis for both seventh edition (C-index 0.94, PVE 4.6%) and eighth edition (C-index 0.94, PVE 4.8%), the factors predictive of worse outcome for DSS were older age, advanced tumor size category, and distant metastasis while cervical lymph node metastases did not predict worse survival. For the eighth edition, patients <55 years of age with stage II disease had significantly worse DSS (p < 0.001) than patients ≥55 years with stage II disease but appeared to be similar to patients ≥55 years with stage III (p = 0.742) and IV disease (p = 0.566). Patients ≥55 years old with T3a and T3b disease had 10-year DSS of 67% and 92%, respectively (p = 0.390).

CONCLUSION

The AJCC eighth edition is similar to the seventh edition in disease-specific survival prediction. Potential modifications that may improve disease-specific survival prediction in future renditions include reconsideration of T3b tumor category and upstaging classification of patients <55 years of age with distant metastases.

The Prognostic Impact of Tumor Size in Papillary Thyroid Carcinoma is Modified by Age

BACKGROUND

Although the importance of tumor size in papillary thyroid cancer (PTC) is well established, there is no research investigating whether age modifies the impact of tumor size, and there is conflicting evidence regarding optimal size thresholds for prognostic discrimination. We aimed to verify that tumor size is an independent prognostic factor in PTC, investigate the impact of patient age, and identify optimal size cutoffs for risk stratification using objective measures of model performance.

METHODS

A retrospective analysis of 574 patients with PTC, using multivariate Cox regression models to test the impact of tumor size on recurrence-free survival (RFS). Subgroup analyses were performed in patients aged <55 and ≥55 years. Exploratory analyses to identify optimal size cutoffs for prognostic discrimination were performed using the proportion of variation explained (PVE) and Harrell's C-index.

RESULTS

Tumor size predicted RFS on multivariate analysis in the overall study cohort (hazard ratio [HR] 1.16; [95% confidence interval (CI)1.01-1.34]; p = 0.038). In subgroup analysis, there was no association between tumor size and RFS in patients aged <55 years (HR 1.11; [CI 0.89-1.38]; p = 0.362). In contrast, size was an independent predictor of RFS in patients aged ≥55 years (HR 1.52; [CI 1.11-2.07]; p = 0.009). In this subgroup, an optimal size threshold of >2 cm versus ≤2 cm (HR 5.24; [CI 2.30-11.92]; p < 0.001; PVE: 36%; C-index: 0.66) provided the greatest prognostic discrimination. There was no incremental improvement in prognostic value by further stratification of size.

CONCLUSION

In our PTC cohort, the impact of tumor size on RFS was limited to patients aged ≥55 years. A single size threshold of 2 cm maximized prognostic discrimination with tumors >2 cm associated with a five times higher risk of recurrence than those ≤2 cm. These findings need to be validated in independent large cohorts and the potential management and staging implications further studied.

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.

An Analysis of The American Joint Committee on Cancer 8th Edition T Staging System for Papillary Thyroid Carcinoma

BACKGROUND

The American Joint Committee on Cancer (AJCC) removed microscopic extrathyroidal extension (ETE) from the 8th edition T staging for papillary thyroid cancer (PTC) based on increasing evidence that it is not an independent prognostic factor.

OBJECTIVES

We compared the prognostic performance of AJCC 7th (pT7) and 8th (pT8) edition T stage systems, particularly in patients ≥55 years old without macroscopic ETE or distant metastases in whom T classification affects AJCC Tumor Node Metastasis (TNM) stage.

METHOD

A retrospective analysis of disease-free survival (DFS) in 577 patients with PTC comparing pT8 vs pT7 using the Akaike information criterion (AIC), Harrell's C-index, and Proportion of Variation Explained (PVE).

RESULTS

Of 105 patients with AJCC7 T3 disease, 74 were down-staged. Overall, the prognostic performance of pT7 and pT8 was similar. However, in patients ≥55 years old without macroscopic ETE or distant metastases, pT8 was inferior to pT7 on the basis of higher AIC, lower C-index (0.67 vs 0.76), and lower PVE (30% vs 45%). In this subset, microscopic ETE was associated with multiple other adverse prognostic features and reduced DFS (hazard ratio, 2.8; 95% confidence interval, 1.5 to 5.2; P = 0.002), irrespective of tumor size.

DISCUSSION

In our cohort, pT8 was inferior to pT7 in patients ≥55 years old without macroscopic ETE or distant metastases in whom T classification affects TNM stage. Microscopic ETE was strongly associated with other adverse prognostic factors and reduced DFS in this patient subgroup and may be an effective surrogate for disease biology in PTC, irrespective of whether it is an independent prognostic factor.

Model for End-Stage Liver Disease and Seven Derivations to Prioritize Liver Transplant Candidates: Which Is the Winner?

OBJECTIVES

Under the present liver transplant policy, patients with the highest risk of death receive preference for organ placement. The aim of this study was to evaluate the Model for End-stage Liver Disease (MELD) and seven prognostic derivatives of this test for outcome prediction in cirrhotic patients on liver transplant wait lists.

MATERIAL AND METHODS

The study included 416 patients (65.9% male; age 49 ± 13.9 years) who were entered to liver transplant wait lists from January 2013 to October 2016. Study endpoints were 3-month, 6-month, and 1-year mortality.

RESULTS

All prognostic models had acceptable overall performances (0.12 < Brier score < 0.21). The MELD-to-serum sodium ratio test outperformed its counterparts at all 3 endpoints. Estimated C statistics ranged from 0.77 to 0.83. The largest value at 3 months was for the 5-variable MELD score (0.83), and the largest value at 6 months (0.82) and 1 year (0.83) was for the MELD-albumin score. The Hosmer-Lemeshow goodness-of-fit test and calibration plots revealed underestimation for the entire range of predicted risk (P < .001). With decision curve analysis, the MELD-to-serum sodium ratio and United Kingdom Model for End-Stage Liver Disease scoring tests covered the most extensive range of optimal threshold probabilities.

CONCLUSIONS

Although some derivations, including sodium and albumin, showed effective prioritization of liver transplant candidates, poor calibration statistics highlighted the need for a recalibration process as an inevitable prerequisite before daily clinical use of these tests at the individual level.

A Study of R2 Measure under the Accelerated Failure Time Models

For right-censored data the accelerated failure time (AFT) model is an alternative to the commonly used proportional hazards regression model. It is a linear model for the (log-transformed) outcome of interest, and is particularly useful for censored outcomes that are not time-to-event, such as laboratory measurements. We provide a general and easily computable definition of the R² measure of explained variation under the AFT model for right-censored data. We study its behavior under different censoring scenarios and under different error distributions; in particular, we also study its robustness when the parametric error distribution is misspecified. Based on Monte Carlo investigation results, we recommend the log-normal distribution as a robust error distribution to be used in practice for the parametric AFT model when the R² measure is of interest. We apply our methodology to an alcohol consumption during pregnancy data set from Ukraine.

Prediction errors for state occupation and transition probabilities in multi-state models

In this paper, we consider the estimation of prediction errors for state occupation probabilities and transition probabilities for multistate time-to-event data. We study prediction errors based on the Brier score and on the Kullback-Leibler score and prove their properness. In the presence of right-censored data, two classes of estimators, based on inverse probability weighting and pseudo-values, respectively, are proposed, and consistency properties of the proposed estimators are investigated. The second part of the paper is devoted to the estimation of dynamic prediction errors for state occupation probabilities for multistate models, conditional on being alive, and for transition probabilities. Cross-validated versions are proposed. Our methods are illustrated on the CSL1 randomized clinical trial comparing prednisone versus placebo for liver cirrhosis patients.

Comparison of the Seventh and Eighth Editions of the American Joint Committee on Cancer/Union for International Cancer Control Tumor-Node-Metastasis Staging System for Differentiated Thyroid Cancer

OBJECTIVE

To evaluate the efficacy and prognostic validity for disease-specific survival (DSS) of the eighth edition American Joint Committee on Cancer/Union for International Cancer Control tumor-node-metastasis (TNM) staging system (TNM-8) compared to the seventh edition (TNM-7) in patients with differentiated thyroid carcinoma (DTC).

METHODS

The seventh and eighth editions of the TNM staging system were applied to 1613 DTC patients who underwent thyroid surgery between 1996 and 2003. The proportion of variation explained and Harrell's c-index were evaluated to compare the predictive capability of DSS.

RESULTS

The mean age of the patients was 44.7 years, and the median follow-up period was 11.2 years. When TNM-8 was applied, 63% of T3 and 3% of N1b DTCs were downgraded to T1/T2 and N1a, respectively. About 38% of patients were downstaged according to TNM-8. The 10-year DSS rates in TNM-7 stages I, II, III, and IV were 99.7%, 98.2%, 98.8%, and 83.2%, respectively. Those in TNM-8 stages I, II, III, and IV were 99.6%, 95.4%, 72.3%, and 48.6%, respectively. The proportion of variation explained values of TNM-7 and TNM-8 were 6.0% and 7.0%, respectively. The Harrell's c-index of TNM-7 was 0.86 and that of TNM-8 was 0.88.

CONCLUSIONS

A significant number of patients were reclassified to lower stages with the application of TNM-8 compared to TNM-7. Applying TNM-8 could improve the accuracy of the staging system for predicting DSS in patients with DTC.

Dysmobility Syndrome and Risk of Mortality for Community-Dwelling Middle-Aged and Older Adults: The Nexus of Aging and Body Composition

Dysmobility syndrome is a newly proposed concept to comprehensively consider bone-muscle-adiposity as a whole to associate with mortality and other adverse outcomes in the older adults. Little was known in Asian populations since the body composition was highly related to ethnicity. The study aimed to evaluate the association between dysmobility syndrome and mortality and to explore the most optimal operational definition for dysmobility syndrome. The prevalence of dysmobility syndrome was 3.9–10.1% based on different operational definitions of adiposity and skeletal muscle index. Subjects with dysmobility syndrome were older, more often to be women, having higher adiposity, lower lean body mass and bone mineral density. Multivariate Cox proportional hazard model showed that dysmobility and pre-dysmobility syndrome had higher risk of mortality than the robust group (Hazard ratio (HR): 11.3, 95% confidence interval (CI): 1.2–109.1; and HR 8.7, 95% CI 1.1-67.3, respectively). Overall, the modified operational definition of dysmobility syndrome in Asian populations using FNIH-adjusted skeletal muscle mass and waist circumference-defined adiposity may be the most optimal model for mortality prediction. Taking the nexus of body composition as a whole to evaluate the mortality risk of older adults is an important improvement beyond sarcopenia and osteoporosis.

The association of positive emotion and first smoking lapse: An ecological momentary assessment study

OBJECTIVE

Individuals attempting to quit smoking typically have poor success rates, and the majority fail to maintain long-term abstinence. Although a large body of evidence documents the impact of negative affect on reducing abstinence, there is a much smaller body of research on positive emotions, which could be an important mechanism that is associated with successful cessation. As such, this study examined positive emotions in real-time via ecological momentary assessment (EMA) to determine whether discrete positive emotions were uniquely related to 2 cessation milestones: quit day lapse and first lapse.

METHOD

Participants were 391 smokers who received tobacco cessation treatment. EMAs were completed pre- and postquit, and positive emotion was assessed with 3 items (enthusiastic, happy, and relaxed) rated on 5-point Likert scales. Analyses examined the associations of the means and slopes of each emotion on the current day with the likelihood of lapse on the following day.

RESULTS

When controlling for relevant covariates, prequit positive emotions were not related to quit day lapse. However, postquit positive emotions were associated with first lapse. Specifically, high levels of happiness and relaxation, as well as increasing levels of enthusiasm, happiness, and relaxation were related to a lower likelihood of next day lapse.

CONCLUSIONS

These are some of the first real-time, real-world data to demonstrate that distinct positive emotions are associated with a lower risk of lapse during the postquit period among smokers attempting to quit. (PsycINFO Database Record

Comparison of prognostic scoring systems in follicular thyroid cancer

INTRODUCTION Many studies have addressed the accuracy of prognostic scoring systems in the treatment of differentiated thyroid cancers as a whole but few have addressed this issue in patients with follicular thyroid cancer (FTC) alone. The aim of this study was to establish the accuracy of the various scoring systems in determining the overall and disease free survival of FTC patients in Singapore. METHODS Retrospective review was undertaken of 82 patients with FTC treated at a single tertiary institution between January 2000 and December 2014. Demographic, clinical, pathological and treatment outcomes were analysed. Prognostic scoring systems evaluated for the cohort included TNM (Tumour, Nodes, Metastases), AGES (Age, Grade, Extent, Size), MACIS (Metastases, Age, Completeness of resection, Invasion, Size), AMES (Age, Metastases, Extent, Sex) and EORTC (European Organisation for Research and Treatment of Cancer). Statistical analysis was performed by plotting Kaplan-Meier survival curves and using the Cox proportional hazards model. RESULTS There were 29 male and 53 female patients with a mean age of 48 years. The mean follow-up duration was 88 months and there were 7 deaths (9%). The ten-year overall survival rate was 90%. Factors predictive of survival on univariate analysis were age, size of tumour, invasiveness, completeness of resection, metastasis, external beam radiotherapy, and risk scores using the AGES and MACIS scoring systems (p<0.05). On multivariate analysis, AGES and MACIS provided the best prognostic information. CONCLUSIONS MACIS is the best prognostic scoring system currently available for FTC and it is superior to other scoring systems in term of guiding management. The scoring systems require further development to accommodate variations in clinical practice globally and to improve the prognostic accuracy.

Prognostic value of the eighth edition AJCC TNM classification for differentiated thyroid carcinoma

BACKGROUND

The prognostic value of the proposed eighth edition of the American Joint Committee on Cancer (AJCC) tumor, node, and metastasis (TNM) classification is currently unclear. The aim of the study was to evaluate the prognostic value of the eighth edition of the AJCC TNM classification.

METHODS

We retrospectively assessed 3176 patients with differentiated thyroid carcinoma (DTC) who underwent thyroidectomy at a tertiary Korean hospital from 1996 to 2005. Cancer-specific survival (CSS) was analyzed using the Kaplan-Meier method and compared using the log-rank test. Performance of the eighth edition TNM with respect to prediction of CSS was assessed against the current seventh edition.

RESULTS

Upon reclassification according to the eighth edition, 37.6% of patients were down-staged. The proportions of stage I and II tumors increased from 61.9% to 81.1% and from 1.7% to 16.0%, respectively, whereas those of stage III and IVB (formerly IVC in the seventh edition) decreased from 27.6% to 2.3% and 0.8% to 0.5%, respectively. The proportions of variance explained (PVEs) for the ability of the eighth and the seventh edition to predict CSS were 3.9% and 2.9%, respectively. The C-index values were 0.765 (95% confidence interval 0.764-0.766) for the eighth edition and 0.736 (0.735-0.737) for the seventh edition.

CONCLUSION

Our results demonstrate that the eighth edition TNM more accurately predicts CSS for patients with DTC than does the seventh edition.

Refining Dynamic Risk Stratification and Prognostic Groups for Differentiated Thyroid Cancer With TERT Promoter Mutations

CONTEXT

Currently, no recurrence or mortality risk systems consider molecular testing when predicting thyroid cancer outcomes.

OBJECTIVE

We developed an integrative prognostic system that incorporates telomerase reverse transcription (TERT) promoter mutations into the recently proposed risk reclassification system after initial therapy [dynamic risk stratification (DRS)] to better categorize and predict outcomes.

DESIGN

A total of 357 differentiated thyroid cancer (DTC) patients without initial distant metastasis were enrolled. Among patients with mutated TERT and wild-type, recurrence-free survival (RFS) was compared according to DRS grouping. Cox regression was used to calculate adjusted hazard ratios (AHRs) to derive AHR groups. Performance of the AHR grouping system with respect to prediction of structural recurrence and cancer-specific survival (CSS) was assessed against the current DRS system and the tumor/node/metastasis (TNM) classification.

RESULTS

Among 357 patients, there were 90 recurrences and 15 cancer-related deaths during a median of 14 years of follow-up. Patients in higher AHR groups were at higher risk of recurrence (10-year RFS for AHR 1, 2, 3, and 4: 94.9%, 82.7%, 50.2%, and 23.1%; P < 0.001) and cancer-related death (10-year CSS: 100.0%. 98.7%, 94.2%, and 76.9%; P < 0.001). The proportions of variance explained (PVEs) for the ability of AHR and DRS grouping to predict recurrence were 22.4% and 18.5%. PVEs of AHR and TNM system to predict cancer-related deaths were 11.5% and 7.4%.

CONCLUSIONS

The AHR grouping system, a simple two-dimensional prognostic system, is as effective as DRS at predicting structural recurrence and provides clinical implication for long-term CSS in patients with nonmetastatic DTC.

Review and evaluation of performance measures for survival prediction models in external validation settings

Background

When developing a prediction model for survival data it is essential to validate its performance in external validation settings using appropriate performance measures. Although a number of such measures have been proposed, there is only limited guidance regarding their use in the context of model validation. This paper reviewed and evaluated a wide range of performance measures to provide some guidelines for their use in practice.

Methods

An extensive simulation study based on two clinical datasets was conducted to investigate the performance of the measures in external validation settings. Measures were selected from categories that assess the overall performance, discrimination and calibration of a survival prediction model. Some of these have been modified to allow their use with validation data, and a case study is provided to describe how these measures can be estimated in practice. The measures were evaluated with respect to their robustness to censoring and ease of interpretation. All measures are implemented, or are straightforward to implement, in statistical software.

Results

Most of the performance measures were reasonably robust to moderate levels of censoring. One exception was Harrell’s concordance measure which tended to increase as censoring increased.

Conclusions

We recommend that Uno’s concordance measure is used to quantify concordance when there are moderate levels of censoring. Alternatively, Gönen and Heller’s measure could be considered, especially if censoring is very high, but we suggest that the prediction model is re-calibrated first. We also recommend that Royston’s D is routinely reported to assess discrimination since it has an appealing interpretation. The calibration slope is useful for both internal and external validation settings and recommended to report routinely. Our recommendation would be to use any of the predictive accuracy measures and provide the corresponding predictive accuracy curves. In addition, we recommend to investigate the characteristics of the validation data such as the level of censoring and the distribution of the prognostic index derived in the validation setting before choosing the performance measures.

Optimal cut-off age in the TNM Staging system of differentiated thyroid cancer: is 55 years better than 45 years?

OBJECTIVE

Age >45 years is included as a variable in the tumor, node, metastases (TNM) staging of differentiated thyroid cancer (DTC), but a higher cut-off value has been suggested to be more clinically relevant and prevent over-staging. We evaluated the optimal age cut-off to predict disease-specific survival (DSS) in patients with DTC.

DESIGN AND PATIENTS

This cohort study included 6333 patients with DTC who underwent thyroid surgery at two tertiary referral centres between 1996 and 2005. The optimal age cut-off value between 45 and 65 years for prediction of DSS was assessed. The proportion of variation explained (PVE) and Harrell's c-index was calculated to compare the predictability of each model.

RESULTS

The median age of patients was 46·0 years (IQR 37·8-54·6), and 5498 (87%) were female. Median follow-up period was 10·0 years, and 10-year DSS rate was 98%. Using TNM staging with 45 years as the cut-off (TNM45), 10-year DSS rates of stage I-IV were 99·4%, 96·1%, 97·7% and 85·9%, respectively (PVE = 3·0%, Harrell's c-index = 0·693); and using 55 years as the cut-off (TNM55), 99·4%, 92·2%, 95·3% and 79·7%, respectively (PVE = 4·3%, Harrell's c-index = 0·776). On receiver operating characteristic curve analysis, the optimal age cut-off for prediction of DSS was 55·4 years (area under the curve = 0·837, P < 0·001). About 20% of patients were down-staged to stage I using TNM55 compared to that using TNM45.

CONCLUSIONS

The cut-off age of 55 years was more appropriate for TNM staging to achieve better predictability for DSS in patients with DTC. This change would prevent over-staging in low-risk patients and prevent over-aggressive treatment.

Semi-parametric regression model for survival data: graphical visualization with R

Cox proportional hazards model is a semi-parametric model that leaves its baseline hazard function unspecified. The rationale to use Cox proportional hazards model is that (I) the underlying form of hazard function is stringent and unrealistic, and (II) researchers are only interested in estimation of how the hazard changes with covariate (relative hazard). Cox regression model can be easily fit with coxph() function in survival package. Stratified Cox model may be used for covariate that violates the proportional hazards assumption. The relative importance of covariates in population can be examined with the rankhazard package in R. Hazard ratio curves for continuous covariates can be visualized using smoothHR package. This curve helps to better understand the effects that each continuous covariate has on the outcome. Population attributable fraction is a classic quantity in epidemiology to evaluate the impact of risk factor on the occurrence of event in the population. In survival analysis, the adjusted/unadjusted attributable fraction can be plotted against survival time to obtain attributable fraction function.

Application of molecular biology of differentiated thyroid cancer for clinical prognostication

Although cancer outcome results from the interplay between genetics and environment, researchers are making a great effort for applying molecular biology in the prognostication of differentiated thyroid cancer (DTC). Nevertheless, role of molecular characterisation in the prognostic setting of DTC is still nebulous. Among the most common and well-characterised genetic alterations related to DTC, including mutations of BRAF and RAS and RET rearrangements, BRAF^V600E is the only mutation showing unequivocal association with clinical outcome. Unfortunately, its accuracy is strongly limited by low specificity. Recently, the introduction of next-generation sequencing techniques led to the identification of TERT promoter and TP53 mutations in DTC. These genetic abnormalities may identify a small subgroup of tumours with highly aggressive behaviour, thus improving specificity of molecular prognostication. Although knowledge of prognostic significance of TP53 mutations is still anecdotal, mutations of the TERT promoter have showed clear association with clinical outcome. Nevertheless, this genetic marker needs to be analysed according to a multigenetic model, as its prognostic effect becomes negligible when present in isolation. Given that any genetic alteration has demonstrated, taken alone, enough specificity, the co-occurrence of driving mutations is emerging as an independent genetic signature of aggressiveness, with possible future application in clinical practice. DTC prognostication may be empowered in the near future by non-tissue molecular prognosticators, including circulating BRAF^V600E and miRNAs. Although promising, use of these markers needs to be refined by the technical sight, and the actual prognostic value is still yet to be validated.