Development and validation of a prognostic nomogram for progression-free and overall survival in patients with advanced renal cell carcinoma (aRCC) treated with pazopanib.

2014 ◽  
Vol 32 (4_suppl) ◽  
pp. 405-405
Author(s):  
Michael W. Kattan ◽  
Cora N. Sternberg ◽  
Faisal Mehmud ◽  
Kamalnayan Bhatt ◽  
Debasis Chakrabarti ◽  
...  
Keyword(s):  
Overall Survival ◽  
Regression Model ◽  
Missing Values ◽  
Cox Regression ◽  
Cox Model ◽  
Cox Proportional Hazards ◽  
Phase Iii ◽  
Concordance Index ◽  
Karnofsky Score ◽  
Multikinase Inhibitor

405 Background: Pazopanib, an oral multikinase inhibitor, demonstrated noninferiority to sunitinib in patients with aRCC in COMPARZ (NCT00720941), a randomized, phase III trial (NEJM 2013;369:722). The purpose of this study was to develop and validate prognostic nomograms based on outcome data from COMPARZ for predicting the probability of 12-month progression-free survival (PFS) and 30-month overall survival (OS) for aRCC patients who received pazopanib. Methods: Statistical modeling was performed on a dataset consisting of 557 patients from the pazopanib arm of COMPARZ. A Cox proportional hazards regression model was fit using predictors thought to be prognostic. These predictors included neutrophil count, platelet count, LDH, and alkaline phosphatase, all relative to ULN; calcium; albumin; hemoglobin; Karnofsky score; months from diagnosis to treatment; number of metastatic sites; and presence of lung, liver, and bone metastases. Data from patients on the sunitinib arm of COMPARZ and the pazopanib arm of the VEG105192 trial (NCT00334282; J Clin Oncol 2010;28:1061) were used for validation. Missing values were imputed using chained equations. For validation with the VEG105192 dataset, bootstrap-corrected estimates of discrimination and calibration were calculated following 1000 resamples. The Cox model was plotted as a nomogram. Results: Prognostic nomograms were developed and validated for predicting the probability of 12-month PFS and 30-month OS in aRCC patients based on a Cox regression model. Calibration plots suggested reasonable correspondence between predicted probabilities and actual proportions of PFS and OS. The concordance index for 12-month PFS was 0.636 with sunitinib patients and 0.635 for pazopanib patients; the concordance index was 0.692 for 30-month OS (sunitinib patients). When examining the PFS and OS nomograms, albumin, calcium, and LDH levels appeared to be the most influential predictors of outcome. Conclusions: The nomograms predict, with reasonable accuracy, treatment outcomes in patients with aRCC receiving pazopanib, based on their baseline clinical characteristics.

Development and internal validation of a prognostic nomogram for overall survival in patients with advanced renal cell carcinoma (aRCC) treated with pazopanib (PAZ).

2013 ◽  
Vol 31 (6_suppl) ◽  
pp. 380-380
Author(s):  
Michael W. Kattan ◽  
Debasis Chakrabarti ◽  
Kamalnayan Bhatt ◽  
Faisal Mehmud ◽  
Cora N. Sternberg ◽  
...  
Keyword(s):  
Renal Cell Carcinoma ◽  
Regression Model ◽  
Cell Carcinoma ◽  
Renal Cell ◽  
Missing Values ◽  
Cox Regression ◽  
Cox Model ◽  
Phase Iii ◽  
Advanced Renal Cell Carcinoma ◽  
Phase Iii Trial

380 Background: PAZ, an oral multikinase inhibitor, demonstrated significant improvement in PFS over placebo in patients with aRCC in a randomized, phase III trial (J Clin Oncol 2009:29; 475). The purpose of this study was to develop and internally validate a prognostic nomogram based on the outcome data from phase III trial for predicting the probability of 12-month OS for treatment-naïve and cytokine-pretreated aRCC patients who received PAZ. Methods: Statistical modeling was performed on a dataset consisting of 281 patients from the PAZ arm of the phase III trial. Missing values were first imputed using multiple imputation with chained equations. A Cox proportional hazards regression model was fit using routinely available predictors thought to be prognostic based on clinical judgment. These predictors included the neutrophil count relative to ULN, platelet count relative to ULN, LDH relative to ULN, alkaline phosphatase relative to ULN, corrected calcium, albumin, hemoglobin, ECOG performance status, months from diagnosis to treatment, number of metastatic sites, and presence of lung, liver and bone metastases. Bootstrap-corrected estimates of discrimination and calibration in the small were calculated following 1,000 resamples. The Cox model was plotted as a nomogram. Results: A prognostic nomogram was developed and internally validated for predicting the probability of 12-month OS in aRCC patients based on a Cox regression model using 13 predictor variables. Calibration plots suggested reasonable correspondence between predicted probabilities and actual proportions of OS. The bootstrap-corrected concordance index, a measure of discrimination, was 0.70. When examining the nomogram, albumin appeared to have the greatest potential influence on predicted OS. Months from diagnosis to treatment and corrected calcium level seemed to be the next most potentially influential predictors. Conclusions: The nomogram predicts with reasonable accuracy and should facilitate treatment decision making for patients with advanced renal cell carcinoma. Validation in a separate dataset is planned.

Overall survival of patients with metastatic castrate-resistant prostate cancer (mCRPC) who have PTEN tumor suppressor gene loss of function.

2021 ◽  
Vol 39 (6_suppl) ◽  
pp. 58-58
Author(s):  
Shilpa Gupta ◽  
Ibrahim M. Abbass ◽  
Christopher Craggs ◽  
Sacha Satram ◽  
Tu My To ◽  
...  
Keyword(s):  
Overall Survival ◽  
Tumor Suppressor ◽  
Regression Model ◽  
Real World ◽  
Cox Regression ◽  
Cox Model ◽  
Suppressor Gene ◽  
Survival Outcomes ◽  
Left Truncation ◽  
Castration Resistance

58 Background: It is estimated that more than 40% of patients with mCRPC have functional loss of phosphatase and tensin homolog (PTEN) tumor suppressor gene, which is associated with unfavorable prognosis and reduced response to androgen receptor-targeting therapy. We describe patient characteristics and survival outcomes by PTEN LOF status among patients with mCRPC in real-world clinical practice. Methods: We conducted a retrospective cohort study using data from the nationwide Flatiron Health-Foundation Medicine mCRPC Clinico-Genomic database (FH-FMI CGDB), a de-identified database linking data derived from electronic health records with genomic data derived from FMI comprehensive genomic profiling (CGP) tests. The study included patients ≥18 years old, with a primary diagnosis of mCRPC between 1/1/2013 and 6/30/2019 who underwent FMI CGP testing and who had a valid PTEN LOF status. Patients were included if their PTEN report date and mCRPC diagnosis date occurred before death or censoring. PTEN LOF status was identified via FMI’s CGP testing. Kaplan-Meier (KM) methods assessed overall survival (OS) by PTEN LOF status from the date of mCRPC diagnosis (later of metastasis and castration resistance) until death or end of study follow-up. A stratified Cox regression model was used to estimate the hazard of death. The Cox model was adjusted for age, race and sequence of metastasis/CRPC diagnoses, and was stratified by the year of mCRPC diagnosis. Adjustments to account for left-truncation and survivorship bias were made in the KM analysis and the Cox regression model. Results: Among the 458 patients who met the eligibility criteria, 174 (38%) had PTEN LOF. The majority of the study sample (76%) was diagnosed with castration-resistance after metastasis. The PTEN LOF group had a higher percentage of white patients (80% vs. 68%; p= 0.01) compared to the PTEN non-LOF group. The mean age of the study sample was 68 years, and there was no difference in mean age at diagnosis by PTEN LOF status ( p= 0.17). Based on the KM estimates adjusted for left-truncation, the median OS was 14.3 months (95% confidence interval [CI]: 11.1-19.7) in the PTEN LOF group compared to 18.3 months (95%CI: 15.5-21.5) in the PTEN non-LOF group (log-rank p= 0.049). In the multivariable Cox model, the PTEN LOF group had numerically 30% higher risk of death compared to the PTEN non-LOF group (hazard ratio = 1.30; 95% CI: 0.99-1.71; p= 0.057). Conclusions: Among real-world patients with mCRPC in the CGDB, PTEN LOF could be associated with poorer survival outcomes, potentially highlighting the unmet need among these patients. Additional studies with larger cohorts are needed to better evaluate the survival outcomes of patients with PTEN LOF. Therapeutic agents acting on the PTEN/PI3K/AKT/mTOR pathway are being tested in clinical trials, and could potentially improve outcomes in this subgroup of patients with mCRPC.

scholarly journals Predictors of colorectal cancer survival using cox regression and random survival forests models based on gene expression data

PLoS ONE ◽  
2021 ◽  
Vol 16 (12) ◽  
pp. e0261625
Author(s):  
Mohanad Mohammed ◽  
Innocent B. Mboya ◽  
Henry Mwambi ◽  
Murtada K. Elbashir ◽  
Bernard Omolo
Keyword(s):  
Colorectal Cancer ◽  
Overall Survival ◽  
Missing Values ◽  
Proportional Hazards ◽  
Cox Regression ◽  
Predictive Performance ◽  
Cox Proportional Hazards ◽  
Rank Score ◽  
Detection And Diagnosis ◽  
Better Than

Understanding and identifying the markers and clinical information that are associated with colorectal cancer (CRC) patient survival is needed for early detection and diagnosis. In this work, we aimed to build a simple model using Cox proportional hazards (PH) and random survival forest (RSF) and find a robust signature for predicting CRC overall survival. We used stepwise regression to develop Cox PH model to analyse 54 common differentially expressed genes from three mutations. RSF is applied using log-rank and log-rank-score based on 5000 survival trees, and therefore, variables important obtained to find the genes that are most influential for CRC survival. We compared the predictive performance of the Cox PH model and RSF for early CRC detection and diagnosis. The results indicate that SLC9A8, IER5, ARSJ, ANKRD27, and PIPOX genes were significantly associated with the CRC overall survival. In addition, age, sex, and stages are also affecting the CRC overall survival. The RSF model using log-rank is better than log-rank-score, while log-rank-score needed more trees to stabilize. Overall, the imputation of missing values enhanced the model’s predictive performance. In addition, Cox PH predictive performance was better than RSF.

scholarly journals 312 Female sex independently predicts adjuvant immunotherapeutic benefit from CTLA4 immune checkpoint inhibition

2020 ◽  
Vol 8 (Suppl 3) ◽  
pp. A339-A339
Author(s):  
Ahmad Tarhini ◽  
Ni Kang ◽  
Sandra Lee ◽  
F Stephen Hodi ◽  
Gary Cohen ◽  
...  
Keyword(s):  
Overall Survival ◽  
Treatment Efficacy ◽  
Significant Interaction ◽  
Cox Regression ◽  
Performance Status ◽  
Phase Iii ◽  
High Dose ◽  
List Type ◽  
Female Sex ◽  
Interaction Term

BackgroundSex differences in tumor immunity and response to immunotherapy were shown in murine models and descriptive analyses from recent clinical trials. Female sex hormones have been implicated in melanoma development and response to systemic therapy. We hypothesized a gender difference in response to adjuvant immunotherapy with ipilimumab (3 or 10 mg/kg; ipi3 or ipi10) versus high dose IFNα (HDI) as tested in the E1609 trial.MethodsE1609 demonstrated significant overall survival (OS) benefit with ipi3 versus HDI.1 We investigated treatment efficacy between ipi and HDI in the subgroups by sex (female, male), age (< 55 or ≥55), stage at study entry (IIIB, IIIC, M1a/1b), ECOG performance status (PS 0, 1), ulceration (yes, no), primary tumor (known, unknown), number of lymph nodes involved (0, 1, 2–3, 4+). Forest plots were created to compare OS and RFS with ipi3 vs. HDI and ipi10 vs. HDI using the concurrently randomized ITT populations. For the estimated HRs, 95% confidence intervals were created for all subgroups.ResultsThe subgroups of female, stage IIIC, PS=1, ulcerated, in-transit without lymph node involvement demonstrated significant improvement in overall survival (OS) and/or relapse free survival (RFS) with ipi3 versus HDI as summarized in table 1. Female sex was significant for both OS and RFS and was further explored. In investigating RFS with ipi3 versus HDI, a multivariate Cox regression model including sex, treatment and interaction term of sex*treatment, indicated a significant interaction between sex and treatment (P = 0.026). Including sex, PS (0 vs. 1), age (<55 vs. 55+), ulceration (yes vs. no), stage (IIIB, IIIC, M1a, M1b), treatment and interaction term of sex*treatment, indicated a significant interaction between sex and treatment (P = 0.024). While similar trends were seen, no significant interactions between sex and treatment effect were found in the OS multivariate analysis or in the comparison of ipi10 versus HDI. When exploring age, in the univariate analyses in the ipi3 versus HDI comparison older women appeared to drive most of the difference (age ≥55: OS, P=0.02 and RFS, P=0.08; differences non-significant for women <55). Table 1.Abstract 312 Table 1Treatment efficacy between ipi3 and HDI by subgroupConclusionsFemale sex was independently associated with RFS adjuvant immunotherapeutic benefit from ipi3, supporting a potentially important role for female related factors in the immune response against melanoma, and these warrant further investigation.Trial RegistrationNCT01274338Ethics ApprovalThe study protocol was approved by the institutional review board (IRB) of each participating institution and conducted in accordance with Good Clinical Practice guidelines as defined by the International Conference on Harmonisation. This study was monitored by the ECOG-ACRIN DataSafety Monitoring Committee and the NCI.ConsentAll patients provided IRB-approved written informed consent.ReferenceTarhini AA, Lee SJ, Hodi FS, Rao UNM, Cohen GI, Hamid O, Hutchins LF, Sosman JA, Kluger HM, Eroglu Z, Koon HB, Lawrence DP, Kendra KL, Minor DR, Lee CB, Albertini MR, Flaherty LE, Petrella TM, Streicher H, Sondak VK, Kirkwood JM. Phase III Study of Adjuvant Ipilimumab (3 or 10 mg/kg) Versus High-Dose Interferon Alfa-2b for Resected High-Risk Melanoma: North American Intergroup E1609. J Clin Oncol. 2020 Feb 20;38(6):567–575. PMID: 31880964.

Hyaluronan heterogeneity in pancreatic ductal adenocarcinoma, primary tumors, and sites of metastasis.

2021 ◽  
Vol 39 (15_suppl) ◽  
pp. e16231-e16231
Author(s):  
Veronica Placencio-Hickok ◽  
Marie Lauzon ◽  
Natalie Moshayedi ◽  
Michelle Guan ◽  
Sungjin Kim ◽  
...  
Keyword(s):  
Overall Survival ◽  
Liver Metastasis ◽  
Regression Model ◽  
Cox Regression ◽  
Progression Free Survival ◽  
Ductal Adenocarcinoma ◽  
High Status ◽  
Free Survival ◽  
Pdac Tissue ◽  
Treatment Naïve

e16231 Background: Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive cancers with an estimated five-year survival rate of 10%. The dense desmoplastic stroma in PDAC contributes to its aggressive nature and treatment resistance. Among the components comprising the stroma, hyaluronan (HA) has been demonstrated to play a critical role in tumor progression and survival. Previous preliminary studies have suggested differences in HA expression in primary and metastatic foci in PDAC. However, the effects of treatment and location of HA expression as well as the role of CD44, a known receptor for HA, on HA as a biomarker signature remain unknown. Thus, we investigated the potential of HA as a biomarker in primary PDAC and metastases. Methods: PDAC tissue from primary (n = 43) and metastatic (n = 66) sites were obtained from Cedars-Sinai Medical Center along with associated clinical data. Tissue slides were stained with H&E, HA using a histochemical assay, and CD44 by immunohistochemistry. HA staining was scored according to the proportion of stromal staining at an intensity greater than the background stroma. HA status was defined as ≥ 50% staining being HA high and < 50% as being HA low. CD44 staining was recorded as an H-score (percentage of tumor cells staining multiplied by intensity of staining on a scale from 0 to 3). Associations between HA levels and the requested variables were examined with t-test, Wilcoxon rank-sum test, Chi-squared test, Fisher’s exact test, or Cox regression model where appropriate. Kaplan-Meier curves were created to assess progression free survival and overall survival. Analyses were performed using SAS 9.4 with two-sided tests and a significance level of 0.05. Results: HA score was significantly higher in primary PDAC tissue compared to sites of metastases (p = 0.0148). Within the metastases, HA score was significantly higher in liver metastasis compared to other sites of metastasis (p = 0.0478). In the liver metastasis tissue, HA score trended lower in patients with previously treated tissue compared to treatment naïve tissue (p = 0.0622). In the treatment naive liver metastasis cohort, patients with HA high status had decreased progression free survival and overall survival compared to patients with HA low status (p = 0.0032 and p = 0.0478, respectively). Using HA score and CD44 in a Cox regression model demonstrated that for every one unit increase in HA score, the risk for recurrence/progression increased by 4.4% at any fixed point in time, adjusting for CD44 score (p = 0.0049). Conclusions: HA score is variable between primary PDAC, PDAC metastatic to the liver, and PDAC metastatic to other sites. Within liver metastases, patients with HA high status had decreased progression free survival and overall survival compared to patients with HA low status. HA levels can serve as a potential biomarker to guide pancreatic cancer treatments and trial design for agents targeting the stroma.

scholarly journals Bioinformatics-Based Identification of Tumor Microenvironment-Related Prognostic Genes in Pancreatic Cancer

2021 ◽  
Vol 12 ◽  
Author(s):  
Shaojie Chen ◽  
Feifei Huang ◽  
Shangxiang Chen ◽  
Yinting Chen ◽  
Jiajia Li ◽  
...  
Keyword(s):  
Pancreatic Cancer ◽  
Overall Survival ◽  
Regression Analysis ◽  
Tumor Microenvironment ◽  
Operating Characteristic ◽  
Cox Regression ◽  
Time Dependent ◽  
Concordance Index ◽  
Cox Regression Analysis ◽  
Score Model

ObjectiveGrowing evidence has highlighted that the immune and stromal cells that infiltrate in pancreatic cancer microenvironment significantly influence tumor progression. However, reliable microenvironment-related prognostic gene signatures are yet to be established. The present study aimed to elucidate tumor microenvironment-related prognostic genes in pancreatic cancer.MethodsWe applied the ESTIMATE algorithm to categorize patients with pancreatic cancer from TCGA dataset into high and low immune/stromal score groups and determined their differentially expressed genes. Then, univariate and LASSO Cox regression was performed to identify overall survival-related differentially expressed genes (DEGs). And multivariate Cox regression analysis was used to screen independent prognostic genes and construct a risk score model. Finally, the performance of the risk score model was evaluated by Kaplan-Meier curve, time-dependent receiver operating characteristic and Harrell’s concordance index.ResultsThe overall survival analysis demonstrated that high immune/stromal score groups were closely associated with poor prognosis. The multivariate Cox regression analysis indicated that the signatures of four genes, including TRPC7, CXCL10, CUX2, and COL2A1, were independent prognostic factors. Subsequently, the risk prediction model constructed by those genes was superior to AJCC staging as evaluated by time-dependent receiver operating characteristic and Harrell’s concordance index, and both KRAS and TP53 mutations were closely associated with high risk scores. In addition, CXCL10 was predominantly expressed by tumor associated macrophages and its receptor CXCR3 was highly expressed in T cells at the single-cell level.ConclusionsThis study comprehensively investigated the tumor microenvironment and verified immune/stromal-related biomarkers for pancreatic cancer.

scholarly journals Relationship Between Creatinine and Body Weight Ratio and Diabetes Mellitus in a Chinese Cohort Study

2021 ◽  
Author(s):  
Xinyu Wang ◽  
Zhuangsen Chen ◽  
Fan Yang ◽  
Xiaohan Ding ◽  
Changchun Cao ◽  
...  
Keyword(s):  
Body Weight ◽  
Missing Values ◽  
Proportional Hazards ◽  
Cox Regression ◽  
Weight Ratio ◽  
Cox Proportional Hazards ◽  
Additive Models ◽  
Cox Regression Analysis ◽  
Non Linear ◽  
The Relationship

Abstract Background: Research on the relationship between Creatinine to Body Weight Ratios (Cre/BW ratios) and the prevalence of diabetes is still lacking. The aim of this study was to investigate the potential association between Cre/BW ratios and incident of diabetes in Chinese adults.Methods: This retrospective study was conducted in 199,526 patients from Rich Healthcare Group in China from 2010 to 2016. The participants were divided into quartiles of the Cre/BW ratios. Multivariate multiple imputation and dummy variables were used to handle missing values. Cox proportional-hazards regression was used to investigate the association of Cre/BW and diabetes. Generalized additive models(GAM) were used to identify non-linear relationships.Results: Of all participants,after handling missing values and adjustment for potential confounders, the multivariate Cox regression analysis results showed that Cre/BW ratios was inversely associated with diabetes risk( HR: 0.268; 95% CI:0.229 to 0.314, P < 0.00001).For men, the hazard ratios(HRs) of incident diabetes was 0.255(95%CI: 0.212-0.307);and for women HR= 0.297 (95%CI: 0.218-0.406).Moreover, sensitivity analysis confirmed the stability of the results. Furthermore, GAM revealed a saturation effect on the independent association between Cre/BW and incident of diabetes.Conclusions: This study demonstrated that increased Cre/BW is negatively correlated with incident of diabetes in Chinese for the first time. And we found that the relationship between Cre/BW and incident of diabetes was non-linear.

scholarly journals Nomogram Predicting Overall Survival After Surgical Resection For Retroperitoneal Leiomyosarcoma Patients

2021 ◽  
Author(s):  
Aobo Zhuang ◽  
Hanxing Tong ◽  
Yuan Fang ◽  
Lijie Ma ◽  
Weiqi Lu ◽  
...  
Keyword(s):  
Public Health ◽  
Overall Survival ◽  
Surgical Resection ◽  
General Surgery ◽  
Cox Regression ◽  
Concordance Index ◽  
Clinical Center ◽  
Surgery Department ◽  
Actual Survival ◽  
Good Agreement

Abstract Aim: To develop a survival nomogram for patients with retroperitoneal leiomyosarcoma (RLMS) after surgery.Methods: 118 patients with RLMS after surgical resection at the General Surgery Department, Shanghai Public Health Clinical Center, Fudan University were retrospectively analyzed. The nomogram was constructed based on COX regression model and discrimination was assessed using the concordance index (c-index). The predicted and actual survival was evaluated through calibration plots.Results: The c-index of the nomogram was 0.779 (95% CI, 0.659-0.898). The predicted and actual survival probabilities are in good agreement in all calibration curve.Conclusion: This study built the first survival nomogram for patients with surgical resected RLMS.

Body weight loss (BWL) as a prognostic/predictive factor in previously treated patients (pts) with metastatic gastric or gastroesophageal junction cancer (mGC/GEJC): Post-hoc analyses of the phase III tags trial.

2021 ◽  
Vol 39 (3_suppl) ◽  
pp. 476-476
Author(s):  
Michele Ghidini ◽  
Howard S. Hochster ◽  
Toshihiko Doi ◽  
Eric Van Cutsem ◽  
Lukas Makris ◽  
...  
Keyword(s):  
Body Weight ◽  
Prognostic Factors ◽  
Proportional Hazards ◽  
Cox Model ◽  
Gastroesophageal Junction ◽  
Body Weight Loss ◽  
Cox Proportional Hazards ◽  
Phase Iii ◽  
Body Weight Data ◽  
Grade 3

476 Background: Nutritional status is closely linked to cancer mortality, and BWL has been shown to be prognostic for survival in curative, first-, and second-line settings in mGC/GEJC. In the phase III TAGS trial, trifluridine/tipiracil (FTD/TPI) showed clinical benefit versus placebo (PBO) and manageable safety in pts with mGC/GEJC who had received ≥2 prior chemotherapy regimens. The association of early BWL with survival outcomes in TAGS was examined in retrospective analyses. Methods: The TAGS intent-to-treat (ITT) population was categorized into pts who experienced <3% or ≥3% BWL from the start of treatment until day 1 of cycle 2 (each cycle: 28 days). Overall survival (OS), and progression-free survival (PFS) were compared between subgroups within each treatment arm due to significant imbalances of early BWL between treatment arms. The effect of early BWL on OS was assessed by a univariate Cox proportional hazards (PH) model as well as a multivariate Cox PH model that adjusted for baseline prognostic factors identified in the original ITT analysis. Results: Body weight data were available for 451 of 507 (89%) pts in the study (n=304, FTD/TPI; n=147, PBO). In the FTD/TPI and PBO arms, respectively, 74% (224/304 pts) and 65% (95/147) experienced <3% BWL, whereas 26% (80/304) and 35% (52/147) experienced ≥3% BWL at the end of cycle 1. Pts with <3% BWL had longer OS than those with ≥3% BWL in both FTD/TPI (median [m] OS: 6.5 vs 4.9 months [mo]; hazard ratio [HR], 0.75; 95% CI, 0.55–1.02) and PBO arms (mOS: 6.0 vs 2.5 mo; HR, 0.32; 95% CI, 0.21–0.49). The PFS HR for pts with <3% BWL vs ≥3% BWL was 0.95 (95% CI, 0.71–1.25; mPFS, 2.1 vs 1.9 mo) in the FTD/TPI group and 0.49 (95% CI, 0.34–0.72; mPFS, 1.9 vs 1.7 mo) in the PBO group. In the pooled ITT population, the unadjusted HR for the <3% vs ≥3% BWL group calculated using a univariate Cox model was 0.58 (95% CI, 0.46–0.73), indicating a strong prognostic effect of early BWL. Results of multivariate analyses were consistent with univariate analyses and suggested that early BWL was both a prognostic ( P<0.0001) and predictive (interaction P=0.0003) factor for OS in pts with mGC/GEJC. Grade ≥3 adverse events (AEs) of any cause were reported in 77% and 82% of FTD/TPI-treated pts in the <3% and ≥3% BWL subgroups, respectively, and in 45% and 67% of placebo-treated pts in the <3% and ≥3% BWL subgroups. Conclusions: To our knowledge, this is the first analysis to show that BWL is negatively associated with survival in pts with mGC/GEJC receiving third- or later-line treatment. In TAGS, early BWL (≥3% BWL at the end of cycle 1) was a strong negative prognostic factor for OS regardless of FTD/TPI or PBO treatment. Grade ≥3 AE frequencies were similar in FTD/TPI-treated pts with <3% or ≥3% BWL. The relationship of BWL to other prognostic factors will be explored further. Clinical trial information: NCT02500043.

Imputing missing time-dependent covariate values for the discrete time Cox model

2019 ◽  
Vol 29 (8) ◽  
pp. 2074-2086
Author(s):  
Havi Murad ◽  
Rachel Dankner ◽  
Alla Berlin ◽  
Liraz Olmer ◽  
Laurence S Freedman
Keyword(s):  
Discrete Time ◽  
Missing Values ◽  
Cox Regression ◽  
Cox Model ◽  
Missing At Random ◽  
Time Dependent ◽  
Outcome Variable ◽  
Sequential Approach ◽  
Random Patterns ◽  
Event Times

We describe a procedure for imputing missing values of time-dependent covariates in a discrete time Cox model using the chained equations method. The procedure multiply imputes the missing values for each time-period in a time-sequential manner, using covariates from the current and previous time-periods as well as the survival outcome. The form of the outcome variable used in the imputation model depends on the functional form of the time-dependent covariate(s) and differs from the case of Cox regression with only baseline covariates. This time-sequential approach provides an approximation to a fully conditional approach. We illustrate the procedure with data on diabetics, evaluating the association of their glucose control with the risk of selected cancers. Using simulations we show that the suggested estimator performed well (in terms of bias and coverage) for completely missing at random, missing at random and moderate non-missing-at-random patterns. However, for very strong non-missing-at-random patterns, the estimator was seriously biased and the coverage was too low. The procedure can be implemented using multiple imputation with the Fully conditional Specification (FCS) method (MI procedure in SAS with FCS statement or similar packages in other software, e.g. MICE in R). For use with event times on a continuous scale, the events would need to be grouped into time-intervals.

Sign in / Sign up

Export Citation Format

Share Document