Developing a personalized risk prediction model of death is fundamental for improving patient care and touches on the realm of personalized medicine. The increasing availability of genomic information and large-scale meta-analytic data sets for clinicians has motivated the extension of traditional survival prediction based on the Cox proportional hazards model. The aim of our paper is to develop a personalized risk prediction formula for death according to genetic factors and dynamic tumour progression status based on meta-analytic data. To this end, we extend the existing joint frailty-copula model to a model allowing for high-dimensional genetic factors. In addition, we propose a dynamic prediction formula to predict death given tumour progression events possibly occurring after treatment or surgery. For clinical use, we implement the computation software of the prediction formula in the joint.Cox R package. We also develop a tool to validate the performance of the prediction formula by assessing the prediction error. We illustrate the method with the meta-analysis of individual patient data on ovarian cancer patients.
GrafESchmoorCSauerbreiWet al.Assessment and comparison of prognostic classification schemes for survival data. Stat Med1999; 18: 2529–2545.
2.
GerdsTASchumacherM. Consistent estimation of the expected Brier score in general survival models with right-censored event times. Biometrical J2006; 48: 1029–1040.
3.
CoxDR. Regression models and life-tables (with discussion). J R Stat Soc Ser B1972; 34: 187–220.
4.
JenssenTKKuoWPStokkeTet al.Association between gene expressions in breast cancer and patient survival. Hum Genet2002; 111: 411–420.
5.
SabatierRFinettiPAdelaideJet al.Down-regulation of ECRG4, a candidate tumor suppressor gene, in human breast cancer. PLoS One2011; 6: e27656–e27656.
6.
LossosISCzerwinskiDKAlizadehAAet al.Prediction of survival in diffuse large-B-cell lymphoma based on the expression of six genes. N Engl J Med2004; 350: 1828–1837.
7.
AlizadehAAGentlesAJAlencarAJet al.Prediction of survival in diffuse large B-cell lymphoma based on the expression of 2 genes reflecting tumor and microenvironment. Blood2011; 118: 1350–1358.
8.
BeerDGKardiaSLRHuangCCet al.Gene-expression profiles predict survival of patients with lung adenocarcinoma. Nat Med2002; 8: 816–824.
9.
ChenHYYuSLChenCHet al.A five-gene signature and clinical outcome in non-small-cell lung cancer. N Engl J Med2007; 356: 11–20.
10.
SheddenKTaylorJMGEnkemannSAet al.Gene expression-based survival prediction in lung adenocarcinoma: a multi-site, blinded validation study. Nat Med2008; 14: 822–827.
11.
PoppleADurrantLGSpendloveIet al.The chemokine, CXCL12, is an independent predictor of poor survival in ovarian cancer. Br J Canc2012; 106: 1306–1313.
12.
YoshiharaKTsunodaTShigemizuDet al.High-risk ovarian cancer based on 126-gene expression signature is uniquely characterized by downregulation of antigen presentation pathway. Clin Canc Res2012; 18: 1374–1385.
13.
WaldronLHaibe-KainsBCulhaneACet al.Comparative meta-analysis of prognostic gene signatures for late-stage ovarian cancer. J Natl Canc Inst2014; 106: dju049–dju049.
14.
MichielsSKoscielnySHillC. Prediction of cancer outcome with microarrays: a multiple random validation strategy. Lancet2005; 365: 488–492.
15.
MatsuiS. Predicting survival outcomes using subsets of significant genes in prognostic marker studies with microarrays. BMC Bioinformatics2006; 7: 156–156.
16.
SchumacherMBinderHGerdsT. Assessment of survival prediction models based on microarray data. Bioinformatics2007; 23: 1768–1774.
17.
BøvelstadHMNygårdSStorvoldHLet al.Predicting survival from microarray data – a comparative study. Bioinformatics2007; 23: 2080–2087.
18.
BinderHSchumacherM. Allowing for mandatory covariates in boosting estimation of sparse high-dimensional survival models. BMC Bioinformatics2008; 9: 14–14.
19.
BøvelstadHMNygårdSBorganØ. Survival prediction from clinico-genomic models – a comparative study. BMC Bioinformatics2009; 10: 1–1.
20.
SimonRMSubramanianJLiMCet al.Using cross-validation to evaluate predictive accuracy of survival risk classifiers based on high-dimensional data. Brief Bioinformatics2011; 12: 203–214.
21.
GoemanJMeijerRChaturvediN. R penalized: L1 (lasso and fused lasso) and L2 (ridge) penalized estimation in GLMs and in the Cox model. CRAN2016. version 0.9-47.
22.
Van HouwelingenH. Dynamic prediction by landmarking in event history analysis. Scand J Stat2007; 34: 70–85.
23.
Van HouwelingenHCPutterH. Dynamic prediction in clinical survival analysis, Boca Raton: CRC Press, 2011.
24.
MauguenARachetBMathoulin-PélissierSet al.Dynamic prediction of risk of death using history of cancer recurrences in joint frailty models. Stat Med2013; 32: 5366–5380.
25.
MauguenARachetBMathoulin-PélissierSet al.Validation of death prediction after breast cancer relapses using joint models. BMC Med Res Methodol2015; 15: 27–27.
26.
RizopoulosD. Dynamic predictions and prospective accuracy in joint models for longitudinal and time-to-event data. Biometrics2011; 67: 819–829.
27.
TaylorJMParkYAnkerstDPet al.Real‐time individual predictions of prostate cancer recurrence using joint models. Biometrics2013; 69: 206–213.
28.
SèneMTaylorJMDignamJJet al.Individualized dynamic prediction of prostate cancer recurrence with and without the initiation of a second treatment: development and validation. Stat Methods in Med Res2016; 25: 2972–2991.
29.
Proust-LimaCSéneMTaylorJMet al.Joint latent class models for longitudinal and time-to-event data: a review. Stat Methods Med Res2014; 23: 74–90.
30.
RondeauVMauguenALaurentAet al.Dynamic prediction models for clustered and interval-censored outcomes: investigating the intra-couple correlation in the risk of dementia. Stat Methods Med Res2015; 26: 2168–2183.
31.
KrólAFerrerLPignonJPet al.Joint model for left-censored longitudinal data, recurrent events and terminal event: predictive abilities of tumor burden for cancer evolution with application to the FFCD 2000-05 trial. Biometrics. 2016. DOI: 10.1111/biom.12490.
32.
EmuraTNakatochiMMurotaniKet al.A joint frailty-copula model between tumour progression and death for meta-analysis. Stat Methods Med Res2015; 26: 2649–2666.
33.
GanzfriedBFRiesterMHaibe-KainsBet al.Curated ovarian data: clinically annotated data for the ovarian cancer transcriptome. Database2013. Article ID bat013. DOI:10.1093/database/bat013.
34.
RondeauVPignonJPMichielsS. A joint model for dependence between clustered times to tumour progression and deaths: a meta-analysis of chemotherapy in head and neck cancer. Stat Methods Med Res2015; 24: 711–729.
35.
HaneuseSLeeKH. Semi-competing risks data analysis, accounting for death as a competing risk when the outcome of interest is nonterminal. Circ Cardiovasc Qual Outcomes2016; 9: 322–331.
36.
JolyPCommengesDLetenneurL. A penalized likelihood approach for arbitrary censored and truncated data: application to age-specific incidence of dementia. Biometrics1998; 54: 185–194.
37.
EmuraTChenYH. Gene selection for survival data under dependent censoring, a copula-based approach. Stat Methods Med Res2016; 25: 2840–2857.
38.
NelsenRB. An introduction to copulas, 2nd ed. New York: Springer, 2006.
39.
TibshiraniR. The lasso method for variable selection in the Cox model. Stat Med1997; 16: 385–395.
40.
Van HouwelingenHCBruinsmaTHartAAet al.Cross-validated Cox regression on microarray gene expression data. Stat Med2006; 25: 3201–3216.
41.
TukeyJW. Tightening the clinical trial. Contr Clin Trials1993; 14: 266–285.
42.
EmuraTChenYHChenHY. Survival prediction based on compound covariate under Cox proportional hazard models. PLoS One2012; 7: e47627–e47627.
43.
RadmacherMDMcshaneLMSimonRM. A paradigm for class prediction using gene expression profiles. J Comput Biol2002; 9: 505–511.
44.
MatsuiSSimonRMQuPet al.Developing and validating continuous genomic signatures in randomized clinical trials for predictive medicine. Clin Canc Res2012; 18: 6065–6073.
45.
SimonRM. Design and analysis of DNA microarray investigations, New York: Springer Science & Business Media, 2003.
46.
EmuraT. R joint.Cox: penalized likelihood estimation and dynamic prediction under the joint frailty-copula models between tumour progression and death for meta-analysis. CRAN. version 2.10 2016-10-30.
47.
RondeauVGonzalezJRMazrouiYet al.R frailtypack: general frailty models: shared, joint and nested frailty models with prediction. CRAN. version 2.8.3 2016-01-13.
48.
HastieTTibshiraniRFriedmanJ. The elements of statistical learning, New York: Springer, 2009.
49.
SchleidgenSKlinglerCBertramTet al.What is personalized medicine: sharpening a vague term based on a systematic literature review. BMC Med Ethic2013; 14: 55–55.
50.
CommengesDProust-LimaCSamieriCet al.A universal approximate cross-validation criterion for regular risk functions. Int J Biostat2015; 11: 51–67.
51.
BurzykowskiTMolenberghsGBuyseMet al.Validation of surrogate end points in multiple randomized clinical trials with failure time end points. Appl Stat2001; 50: 405–422.
52.
BurzykowskiTMolenberghsGBuyseM. The evaluation of surrogate endpoints, New York: Springer, 2005.
53.
MichielsSLe MaîtreABuyseMet al.Surrogate endpoints for overall survival in locally advanced head and neck cancer: meta-analyses of individual patient data. Lancet Oncol2009; 10: 341–350.
54.
BuyseMSargentDJSaadED. Survival is not a good outcome for randomized trials with effective subsequent therapies. J Clin Oncol2011; 29: 4719–4720.
55.
ObaKPaolettiXAlbertsSet al.Disease-free survival as a surrogate for overall survival in adjuvant trials of gastric cancer: a meta-analysis. J Natl Canc Inst2013; 105: 1600–1607.
56.
BelleraCAMacGroganGDebledMet al.Variables with time-varying effects and the Cox model: some statistical concepts illustrated with a prognostic factor study in breast cancer. BMC Med Res Methodol2010; 10: 1–1.
57.
MazrouiYMauguenAMathoulin-PélissierSet al.Time-varying coefficients in a multivariate frailty model: application to breast cancer recurrences of several types and death. Lifetime Data Anal2016; 22: 191–215.
58.
BauliesSBelinLMallonPet al.Time-varying effect and long-term survival analysis in breast cancer patients treated with neoadjuvant chemotherapy. Br J Canc2015; 113: 30–36.
59.
PutterHSasakoMHartgrinkHet al.Long-term survival with non-proportional hazards: results from the Dutch Gastric Cancer Trial. Stat Med2005; 24: 2807–2821.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
