Search

Robustly predicting outcome for cancer patients from gene expression is an important challenge on the road to better personalized treatment. Network-based outcome predictors (NOPs), which considers the cellular wiring diagram in the classification, hold much promise to improve performance, stability and interpretability of identified marker genes. Problematically, reports on the efficacy of NOPs are conflicting and for instance suggest that utilizing random networks performs on par to networks that describe biologically relevant interactions. In this paper we turn the prediction problem around: instead of using a given biological network in the NOP, we aim to identify the network of genes that truly improves outcome prediction. To this end, we propose SyNet, a gene network constructed ab initio from synergistic gene pairs derived from survival-labelled gene expression data. To obtain SyNet, we evaluate synergy for all 69 million pairwise combinations of genes resulting in a network that is specific to the dataset and phenotype under study and can be used to in a NOP model. We evaluated SyNet and 11 other networks on a compendium dataset of >4000 survival-labelled breast cancer samples. For this purpose, we used cross-study validation which more closely emulates real world application of these outcome predictors. We find that SyNet is the only network that truly improves performance, stability and interpretability in several existing NOPs. We show that SyNet overlaps significantly with existing gene networks, and can be confidently predicted (~85% AUC) from graph-topological descriptions of these networks, in particular the breast tissue-specific network. Due to its data-driven nature, SyNet is not biased to well-studied genes and thus facilitates post-hoc interpretation. We find that SyNet is highly enriched for known breast cancer genes and genes related to e.g. histological grade and tamoxifen resistance, suggestive of a role in determining breast cancer outcome. [ABSTRACT FROM AUTHOR]

Effective symptom management is a critical component of cancer treatment. Computational tools that predict the course and severity of these symptoms have the potential to assist oncology clinicians to personalize the patient’s treatment regimen more efficiently and provide more aggressive and timely interventions. Three common and inter-related symptoms in cancer patients are depression, anxiety, and sleep disturbance. In this paper, we elaborate on the efficiency of Support Vector Regression (SVR) and Non-linear Canonical Correlation Analysis by Neural Networks (n-CCA) to predict the severity of the aforementioned symptoms between two different time points during a cycle of chemotherapy (CTX). Our results demonstrate that these two methods produced equivalent results for all three symptoms. These types of predictive models can be used to identify high risk patients, educate patients about their symptom experience, and improve the timing of pre-emptive and personalized symptom management interventions. [ABSTRACT FROM AUTHOR]

How can we predict the occurrence of acute kidney injury (AKI) in cancer patients based on machine learning with serum creatinine data? Given irregular and heterogeneous clinical data, how can we make the most of it for accurate AKI prediction? AKI is a common and significant complication in cancer patients, and correlates with substantial morbidity and mortality. Since no effective treatment for AKI still exists, it is important to take timely preventive measures. While several approaches have been proposed for predicting AKI, their scope and applicability are limited as they either assume regular data measured over a short hospital stay, or do not fully utilize heterogeneous data. In this paper, we provide an AKI prediction model with a greater applicability, which relaxes the constraints of existing approaches, and fully utilizes irregular and heterogeneous data for learning the model. In a cohort of 21,022 cancer patients who were registered into Korea Central Cancer Registry (KCCR) in Seoul National University Hospital between January 1, 2004 and December 31, 2013, our method achieves 0.7892 precision, 0.7506 recall, and 0.7576 F-measure in predicting whether a patient will develop AKI during the next 14 days. [ABSTRACT FROM AUTHOR]

Background: LncRNA CCAT1 is significantly overexpressed in various types of cancers, suggesting that it might be associated with prognosis and clinicopathological features in patients with cancer. Methods: A comprehensive search was performed in Pubmed, Web of Science, OVID and CNKI databases. We also retrieved articles from other sources, such as retrieving from the reference lists of relevant articles. Eligible studies were included based on defined exclusion and inclusion criteria to perform a meta-analysis. STATA 14.0 was used to estimate pooled hazard ratios (HRs) with 95% confidence interval (95% CI), the heterogeneity among studies and publication bias to judge the prognostic value. Results: A total of 1587 patients from 11 eligible studies were included in the meta-analysis. The results showed that high expression level of CCAT1 was significantly associated with shorter overall survival in cancer patients (HR 2.335, 95% CI:1.551–3.517); in the subgroup analysis, region (China or UK), sample size (more or less than 100), type of cancer (digestive or non-digestive disease) and paper quality (score more or less than 7) did not alter the association between CCAT1 expression and cancer prognosis but preoperative treatment did. And CCAT1 expression was an independent prognostic marker for overall survival in patients with cancer (pooled HR 2.195, 95%CI:1.316–3.664) using Cox multivariate analyses. The clinicopathological parameters analysis further showed that increased expression level of CCAT1 was correlated with tumor size, lymph node metastasis, TNM stage, distant metastasis, microvascular invasion and capsular formation in relevant cancers. Conclusions: The meta-analysis results from present study suggested that increased expression level of CCAT1 was associated with poor prognosis and can serve as an independent biomarker. And the expression level of CCAT1 was associated with clinicopathological features in relevant cancers. [ABSTRACT FROM AUTHOR]

Background: Hydronephrosis is a common finding in patients with bladder cancer. The aim of the study was to appraise the prognostic value of preoperative hydronephrosis in bladder cancer patients undergoing radical cystectomy. Methods: We conducted a literature search using PubMed and Embase databases in Aug 2018. Summary hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated using fixed-effect or random-effects models. The primary endpoint was overall survival (OS). Secondary endpoints were cancer-specific survival (CSS) and recurrence-free survival (RFS). Results: Overall, 13 studies published between 2008 and 2018 including 4,820 patients were selected for the meta-analysis. The age of bladder cancer patients ranged from 27 to 90.4 years, and the overall proportion of males was 72.5%. Preoperative hydronephrosis was reported in 27.4% of patients. The pooled HR was statistically significant for OS (HR, 1.36; 95% CI [1.20–1.55]) and CSS (HR, 1.64; 95% CI [1.33–2.02]), with no heterogeneity among the enrolled studies. Patients with bilateral hydronephrosis showed a poorer CSS compared to those with no hydronephrosis (HR 5.43, 95% CI [3.14–9.40]). However, there was no difference in CSS between no hydronephrosis and unilateral hydronephrosis groups (HR 1.35, 95% CI [0.84–2.14]). Despite a tendency towards poorer RFS (HR, 1.27; 95% CI [0.96–1.96]), the results demonstrated no significant association between presence of preoperative hydronephrosis and RFS after radical cystectomy. Conclusion: This meta-analysis indicates that preoperative hydronephrosis is significantly associated with poorer OS and CSS after radical cystectomy for patients with bladder cancer. Preoperative hydronephrosis has a stronger effect on CSS in patients with bilateral hydronephrosis. The presence of preoperative hydronephrosis not only predicts prognosis, but may also help to identify patients who benefit the most from neoadjuvant chemotherapy. [ABSTRACT FROM AUTHOR]

Background: Recent literature has suggested that tumor sidedness in colorectal cancer (CRC) is an independent prognostic and potentially predictive marker of survival. The aims of the study were to determine the prognostic significance of tumor sidedness in colorectal cancer patients undergoing primary tumor resection and to assess associated tumor biology. Methods: A total of 3281 consecutive patients who underwent surgical resection of their primary CRC from January 1998 to December 2012 were analyzed for association with tumor biologic factors and with overall survival. Metastatic patients were excluded from analysis. Results: Left sided CRCs were associated with a number of additional key prognostic markers including BRAFV600E wildtype status (P<0.001), mismatch repair proficiency (p<0.001), absence of peritumoral lymphocytic response (p = 0.001), high lymphocyte-to-monocyte ratio (p<0.001) and low neutrophil-to-lymphocyte ratio (p<0.001). In primary analysis with 3067 patients, there was no statistical difference in sidedness in the univariate analysis (p = 0.291). Three further subgroup analyses were performed. In the first subgroup, only stage III patients were analyzed. In the second, patients with mismatch repair deficiency were removed. In the third, additional clinicopathologic variables known to be independently prognostic were added into analysis. In all three subgroup analyses tumor sidedness was not an independent prognostic marker. Conclusions: Tumor sidedness was not an independent prognostic marker of CRC. However, right sided CRCs were associated with several key independent prognostic markers supporting a hypothesis that tumor sidedness is a surrogate for other biomarkers. [ABSTRACT FROM AUTHOR]

Background: We evaluated the associations of handgrip strength and cognitive function in cancer survivors ≥ 60 years old using data from the National Health and Nutrition Examination Survey (NHANES). Methods: Data in two waves of NHANES (2011–2014) were aggregated. Handgrip strength in kilogram (kg) was defined as the maximum value achieved using either hand. Two cognitive function tests were conducted among adults 60 years and older. The Animal Fluency Test (AFT) examines categorical verbal fluency (a component of executive function), and the Digital Symbol Substitution test (DSST) assesses processing speed, sustained attention, and working memory. Survey analysis procedures were used to account for the complex sampling design of the NHANES. Multiple linear regression models were used to estimate associations of handgrip strength with cognitive test scores, adjusting for confounders (age, gender, race/ethnicity, education, marital status, smoking status, depressive symptoms and leisure time physical activity). Results: Among 383 cancer survivors (58.5% women, mean age = 70.9 years, mean BMI = 29.3 kg/m2), prevalent cancer types were breast (22.9%), prostate (16.4%), colon (6.9%) and cervix (6.2%). In women, each increase in kg of handgrip strength was associated with 0.20 (95% CI: 0.08 to 0.33) higher score on AFT and 0.83 (95% CI: 0.30 to 1.35) higher score on DSST. In men, we observed an inverted U-shape association where cognitive function peaked at handgrip strength of 40–42 kg. Conclusions: Handgrip strength, a modifiable factor, appears to be associated with aspects of cognitive functions in cancer survivors. Prospective studies are needed to address their causal relationship. [ABSTRACT FROM AUTHOR]

Background: Clinical prediction models are increasingly used to predict outcomes such as survival in cancer patients. The aim of this study was threefold. First, to perform a systematic review to identify available clinical prediction models for patients with esophageal and/or gastric cancer. Second, to evaluate sources of bias in the included studies. Third, to investigate the predictive performance of the prediction models using meta-analysis. Methods: MEDLINE, EMBASE, PsycINFO, CINAHL, and The Cochrane Library were searched for publications from the year 2000 onwards. Studies describing models predicting survival, adverse events and/or health-related quality of life (HRQoL) for esophageal or gastric cancer patients were included. Potential sources of bias were assessed and a meta-analysis, pooled per prediction model, was performed on the discriminative abilities (c-indices). Results: A total of 61 studies were included (45 development and 16 validation studies), describing 47 prediction models. Most models predicted survival after a curative resection. Nearly 75% of the studies exhibited bias in at least 3 areas and model calibration was rarely reported. The meta-analysis showed that the averaged c-index of the models is fair (0.75) and ranges from 0.65 to 0.85. Conclusion: Most available prediction models only focus on survival after a curative resection, which is only relevant to a limited patient population. Few models predicted adverse events after resection, and none focused on patient’s HRQoL, despite its relevance. Generally, the quality of reporting is poor and external model validation is limited. We conclude that there is a need for prediction models that better meet patients’ information needs, and provide information on both the benefits and harms of the various treatment options in terms of survival, adverse events and HRQoL. [ABSTRACT FROM AUTHOR]

Background: Neoadjuvant chemotherapy before radical gastrectomy is preferred for locally advanced gastric cancer. To avoid the problematic use of pTNM for patients after neoadjuvant chemotherapy, the Union for International Cancer Control (UICC) and the American Joint Committee on Cancer (AJCC) gastric cancer TNM staging system (8th edition) added ypTNM for the first time. But patients achieving pathological complete response were not covered by the new ypTNM staging system. To investigate whether pathological complete response is associated with better outcome in gastric cancer, as was reported in rectal, breast and bladder cancer. Methods: We systematically searched the databases of PubMed, EMBASE, Web of Science and Cochrane Collaboration’s Central register of controlled trials from January 1988 to April 2015 for publications which reported outcomes of patients with and without pathological complete response (pCR) (pT0N0M0) to investigate whether pCR after neoadjuvant chemotherapy in gastric or gastroesophageal junction (GEJ) treated with radical surgery is associated with better survival. The primary outcome was overall survival (OS). The secondary outcome was disease-free survival (DFS). Both were measured with a relative risk (RR). A meta-analysis was performed using the fixed effects model. Forest plots and the Q test was used to evaluate overall heterogeneity for OS and DFS. Results: A total of seven trials, 1143 patients were included and analyzed after neoadjuvant chemotherapy and radical surgery with no other preoperative treatment. The average rate of pCR was 6.74% (range: 3%-15%). The RR of patients who achieved pCR in the primary tumor and lymph nodes is 0.5 (95% confidence interval [CI], 0.25–0.98; p = 0.04), 0.34 (95% CI, 0.21–0.55; p<0.0001) and 0.44 (95% CI, 0.30–0.63; p<0.0001) for one-year-OS, three-year-OS and five-year-OS, respectively. The summary RR for three-year-DFS was 0.43 (95% CI, 0.25–0.72; p = 0.002). Conclusion: Patients with resectable gastric or GEJ cancer who achieved pCR after neoadjuvant chemotherapy can gain a better outcome than patients without pCR. [ABSTRACT FROM AUTHOR]

Objective: There is a heated debate on whether the prognostic value of NME1 is favorable or unfavorable. Thus, we carried out a meta-analysis to evaluate the relationship between NME1 expression and the prognosis of patients with digestive system neoplasms. Methods: We searched PubMed, EMBASE and Web of Science for relevant articles. The pooled odd ratios (ORs) and corresponding 95%CI were calculated to evaluate the prognostic value of NME1 expression in patients with digestive system neoplasms, and the association between NME1 expression and clinicopathological factors. We also performed subgroup analyses to find out the source of heterogeneity. Results: 2904 patients were pooled from 28 available studies in total. Neither the incorporative OR combined by 17 studies with overall survival (OR = 0.65, 95%CI:0.41–1.03, P = 0.07) nor the pooled OR with disease-free survival (OR = 0.75, 95%CI:0.17–3.36, P = 0.71) in statistics showed any significance. Although we couldn’t find any significance in TNM stage (OR = 0.78, 95%CI:0.44–1.36, P = 0.38), elevated NME1 expression was related to well tumor differentiation (OR = 0.59, 95%CI:0.47–0.73, P<0.00001), negative N status (OR = 0.54, 95%CI:0.36–0.82, P = 0.003) and Dukes’ stage (OR = 0.43, 95%CI:0.24–0.77, P = 0.004). And in the subgroup analyses, we only find the “years” which might be the source of heterogeneity of overall survival in gastric cancer. Conclusions: The results showed that statistically significant association was found between NME1 expression and the tumor differentiation, N status and Dukes’ stage of patients with digestive system cancers, while no significance was found in overall survival, disease-free survival and TNM stage. More and further researches should be conducted to reveal the prognostic value of NME1. [ABSTRACT FROM AUTHOR]

Background: Appetite and symptoms, conditions generally reported by the patients with cancer, are somewhat challenging for professionals to measure directly in clinical routine (latent conditions). Therefore, specific instruments are required for this purpose. This study aimed to perform a cultural adaptation of the Cancer Appetite and Symptom Questionnaire (CASQ), into Portuguese and evaluate its psychometric properties on a sample of Brazilian cancer patients. Methods: This is a validation study with Brazilian cancer patients. The face, content, and construct (factorial and convergent) validities of the Cancer Appetite and Symptom Questionnaire, the study tool, were estimated. Further, a confirmatory factor analysis (CFA) was conducted. The ratio of chi-square and degrees of freedom (χ2/df), comparative fit index (CFI), goodness of fit index (GFI) and root mean square error of approximation (RMSEA) were used for fit model assessment. In addition, the reliability of the instrument was estimated using the composite reliability (CR) and Cronbach’s alpha coefficient (α), and the invariance of the model in independent samples was estimated by a multigroup analysis (Δχ2). Results: Participants included 1,140 cancer patients with a mean age of 53.95 (SD = 13.25) years; 61.3% were women. After the CFA of the original CASQ structure, 2 items with inadequate factor weights were removed. Four correlations between errors were included to provide adequate fit to the sample (χ2/df = 8.532, CFI = .94, GFI = .95, and RMSEA = .08). The model exhibited a low convergent validity (AVE = .32). The reliability was adequate (CR = .82 α = .82). The refined model showed strong invariance in two independent samples (Δχ2: λ: p = .855; i: p = .824; Res: p = .390). A weak stability was obtained between patients undergoing chemotherapy and radiotherapy (Δχ2: λ: p = .155; i: p < .001; Res: p < .001), and between patients undergoing chemotherapy combined with radiotherapy and palliative care (Δχ2: λ: p = .058; i: p < .001; Res: p < .001). Conclusion: The Portuguese version of the CASQ had good face and construct validity and reliability. However, the CASQ still presented invariance in independent samples of Brazilian patients with cancer. However, the tool has low convergent validity and weak invariance in samples with different treatments. [ABSTRACT FROM AUTHOR]

Mid-study design modifications are becoming increasingly accepted in confirmatory clinical trials, so long as appropriate methods are applied such that error rates are controlled. It is therefore unfortunate that the important case of time-to-event endpoints is not easily handled by the standard theory. We analyze current methods that allow design modifications to be based on the full interim data, i.e., not only the observed event times but also secondary endpoint and safety data from patients who are yet to have an event. We show that the final test statistic may ignore a substantial subset of the observed event times. An alternative test incorporating all event times is found, where a conservative assumption must be made in order to guarantee type I error control. We examine the power of this approach using the example of a clinical trial comparing two cancer therapies. [ABSTRACT FROM AUTHOR]

Purpose: This research conducted a face validation study of patient responses to the application of an HRQOL assessment research tool in a comprehensive community cancer program setting across a heterogeneous cohort of cancer patients throughout the natural history of diagnosed malignant disease, many of whom would not be considered candidates for clinical research trial participation. Methods: Cancer registries at two regional cancer treatment centers identified 11072 cancer patients over a period of nine years. The EORTC QLQ-C30 was administered to patients at the time of their initial clinical presentation to these centers. To determine the significance of differences between patient subgroups, two analytic criteria were used. The Mann-Whitney test was used to determine statistical significance; clinical relevance defined a range of point differences that could be perceived by patients with different health states. Results: Univariate analyses were conducted across stratification variables for population, disease severity and demographic characteristics. The largest differences were associated with cancer diagnosis and recurrence of disease. Large differences were also found for site of origin, mortality and stage; minimal differences were observed for gender and age. Consistently sensitive QoL scales were appetite loss, fatigue and pain symptoms, and role (work-related), social and physical functions. Conclusions: 1) The EORTC QLQ-C30 collected meaningful patient health assessments in the context of non-research based clinical care, 2) patient assessment differences are manifested disparately across 15 QoL domains, and 3) in addition to indicating how a patient may feel at a point in time, QoL indicators may also reveal information about underlying biological responses to disease progression, treatments, and prospective survival. [ABSTRACT FROM AUTHOR]