Annals of Oncology Advance Access originally published online on September 12, 2006
Annals of Oncology 2006 17(11):1698-1704; doi:10.1093/annonc/mdl183
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2006 European Society for Medical Oncology
quality of life and supportive care |
Is a patient's self-reported health-related quality of life a prognostic factor for survival in non-small-cell lung cancer patients? A multivariate analysis of prognostic factors of EORTC study 08975
1 European Organisation for Research and Treatment of Cancer (EORTC), EORTC Data Center, Brussels, Belgium
2 Vrije Universiteit Medical Center, Amsterdam, The Netherlands
3 Hospital de Mataro, Spain
4 St. Antonius Hospital, Nieuwegein, The Netherlands
5 Rijnstate Hospital Arnhem, The Netherlands
6 National Cancer Institute, Cairo, Egypt
7 Jeroen Bosch Ziekenhuis's-Hertogenbosch, The Netherlands
8 University Medical Center, Mannheim, Germany
9 University Hospital, Ghent, Belgium
*Correspondence to: Dr F. Efficace, European Organization for Research and Treatment of Cancer (EORTC), Quality of Life Unit, Avenue E. Mounier, 83, 1200 Brussels, Belgium. Tel: +32 2 7741680; Fax: +32 2 7794568; E-mail: f.efficace{at}fondazioneime.org
 |
Abstract |
|---|
 Top Abstract introductionpatients and methodsresultsdiscussionappendixAcknowledgementsReferences |
|---|
Background: The aim of this prognostic factor analysis was to investigate if a patient's self-reported health-related quality of life (HRQOL) provided independent prognostic information for survival in non-small cell lung cancer (NSCLC) patients.
Patients and methods: Pretreatment HRQOL was measured in 391 advanced NSCLC patients using the EORTC QLQ-C30 and the EORTC Lung Cancer module (QLQ-LC13). The Cox proportional hazards regression model was used for both univariate and multivariate analyses of survival. In addition, a bootstrap validation technique was used to assess the stability of the outcomes.
Results: The final multivariate Cox regression model retained four parameters as independent prognostic factors for survival: male gender with a hazard ratio (HR) = 1.32 (95% CI 1.03–1.69; P = 0.03); performance status (0 to 1 versus 2) with HR = 1.63 (95% CI 1.04–2.54; P = 0.032); patient's self-reported score of pain with HR= 1.11 (95% CI 1.07–1.16; P < 0.001) and dysphagia with HR = 1.12 (95% CI 1.04–1.21; P = 0.003). A 10-point shift worse in the scale measuring pain and dysphagia translated into an 11% and 12% increased in the likelihood of death respectively. A risk group categorization was also developed.
Conclusion: The results suggest that patients' self-reported HRQOL provide independent prognostic information for survival. This finding supports the collection of such data in routine clinical practice.
Key words: lung cancer, prognostic factor, quality of life, survival
|
introduction |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionappendixAcknowledgementsReferences |
|---|
It is now generally accepted that, in addition to the traditional assessment of clinical outcomes, such as overall survival or tumor response, patients' self reported health status data plays a key role in cancer research and recent examples provide a strong rationale for measuring it in a clinical trial setting [1]. An additional area of investigation which has become increasingly important is that of patients' self-reported health status and/or of heath related quality of life (HRQOL) parameters that may provide independent prognostic information. This evidence has been recently replicated in various studies using a variety of questionnaires such as the Functional Living Index-Cancer [2], the Rotterdam Symptom Checklist [3] and the European Organisation for Research and Treatment of Cancer, Quality of Life Questionnaire-Core30 (EORTC QLQ-C30) [4–6]. Overall, the identification of independent prognostic factors could lead to a better interpretation of clinical trial results and might give individual guidance for the clinicians in the decision-making process of choice of treatment options. In advanced-disease settings, this information could also assist clinicians to recalibrate clinical prediction of survival and optimize the use of palliative care [7].
The majority of lung cancer patients are diagnosed with non-small-cell lung cancer (NSCLC). In addition, most patients with NSCLC present with locally advanced or metastatic disease that is incurable with existing treatment modalities [8]. Treatment is palliative and the aim is to prolong survival with improvement or maintenance of HRQOL aspects [9]. Previous studies have investigated the prognostic value of HRQOL parameters in this population. Whilst they have clearly identified a robust relation between survival and HRQOL parameters in the multivariate analysis, conflicting results have been found with the specific HRQOL domains [10–12]. In addition, one major limitation of previous prognostic factor studies using the EORTC QLQ-C30 with advanced NSCLC patients is that they have failed to take into account the possible harmful multicollinearity amongst HRQOL variables which may affect the stability of the final model predicting survival [13]. This phenomenon is relevant in multiple regression analyses when two or more predictor variables are so highly correlated that odd results could be obtained (e.g. parameter estimate of incorrect magnitude or incorrect sign) possibly also leading to incorrect model selection.
Given the above evidence, the main objective of this study was to evaluate whether pretreatment patients' reported HRQOL data in NSCLC patients independently predicted overall survival by also controlling for key socio-demographic and biomedical data. In addition, a previously developed bootstrap model averaging technique was applied to obtain better insight into the stability of the final model predicting survival [13]. We used data obtained from a randomized clinical trial comparing three different chemotherapy regimens in patients with advanced NSCLC (EORTC 08975).
|
patients and methods |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionappendixAcknowledgementsReferences |
|---|
study design
This study is based on data from an international prospective multicenter randomized controlled trial (RCT) in advanced NSCLC patients, conducted by the EORTC Lung Cancer Group (EORTC Study 08975). Overall 480 patients were randomly assigned from 29 institutions. Patients were randomized to receive paclitaxel 175 mg/m2 on day 1 followed by cisplatin 80 mg/m2 on day 1 (arm A), gemcitabine 1250 mg/m2 on days 1 and 8 and cisplatin 80 mg/m2 on day 1 after gemcitabine (arm B), or paclitaxel 175 mg/m2 on day 1 followed by gemcitabine 1250 mg/m2 on days 1 and 8 (arm C). All treatment cycles were repeated every 3 weeks. Full details of treatment schedule and treatment-related clinical outcomes have been previously reported [14].
patients
To be eligible for inclusion in the original trial, patients had to be diagnosed with histologically or cytologically confirmed NSCLC stage IIIB and stage IV. Additional eligibility criteria included age between 18 and 76 years, WHO performance status 
2, measurable disease and no previous chemotherapy. The study, approved by the EORTC protocol review committee and the ethics committee of each participating center, was conducted in compliance with the Helsinki declaration. All patients provided written informed consent.
HRQOL pretreatment assessment and variables examined
Pretreatment HRQOL data were measured by the EORTC QLQ-C30 (version 3.0) [15], which is a measure validated for a generic cancer population and also by the EORTC Lung Cancer module (QLQ-LC13) [16]. The QLQ-LC13 is a measure specifically validated on lung cancer patients. They both have robust psychometric properties resulting from their use in many international cancer clinical trials [17]. In these questionnaires higher scores on the functioning scales represent higher level of functioning, whilst higher scores on the symptom scales represent higher subjective perception of the symptoms. The EORTC QLQ-C30 is a core measure designed to be supplemented with the disease specific QLQ-LC13 developed and validated specifically in patients with lung cancer. Both instruments were available in the language of all participating patients (in translation) following EORTC procedures [18]. Assessments were performed at baseline within a time window of 14 days before or after randomization, but in any case before treatment start.
Wherever possible, the questionnaires were administered at the hospital, in a room where the patient would not be disturbed. EORTC guidelines for administering questionnaires were provided, ensuring a standard approach to the collection of HRQOL data. The EORTC QLQ-C30 scores were calculated using the recommended EORTC procedures [17]. These involved transformation of raw scores into a linear scale ranging from 0–100 with a higher score representing a higher level of functioning or higher level of symptoms. In order to reduce the risk due to multiple testing, we investigated the correlation matrix among all variables and excluded from the analysis, a priori, the HRQOL variables which were expected to have no prognostic value and had high intercorrelation with other scales. The following scale from the EORTC QLQ-C30 and QLQ-LC13 were then included: global quality of life scale, emotional, social and physical functioning, nausea and vomiting, pain, appetite loss, dyspnea (combined with the dyspnea scale of the QLQ-LC13), coughing and dysphagia. The HRQOL variables described above were all included as continuous factors, using data from baseline assessments. Key socio-demographic and biomedical variables were also included: age (continuous), gender (male versus female), stage of disease (IIIB versus IV), histological subtype (squamous versus non-squamous), and performance status (0 to 1 versus 2). These were based on previous clinical evidence [14]. Treatment was taken as a stratification factor.
statistical analysis
All the following analyses were conducted on 391 patients, of the pooled sample of 480, having baseline HRQOL data enrolled in arm A (134), arm B (132) and arm C (125). Overall survival (OS) was measured from the date of randomization to the date of death (due to any cause). Patients still alive at the time of analysis were censored at the last date known to be alive. Survival curves and probabilities were estimated using the Kaplan-Meier technique. Differences between survival curves were assessed using the log-rank test. The Cox proportional hazards regression model was used for both univariate and multivariate analyses of survival [19]. For the analysis of prognostic factors for survival the proportionality assumption was checked for each of the variables under study by testing the dependency of their hazard ratio over time [20]. Pearson's correlation coefficients were used to investigate the association between different covariates. When using a stepwise variable selection procedure to identify independent factors prognostic for survival, variables were added using forward selection according to a selection entry criterion of 0.05 and removed using backward elimination according to a selection stay criterion of 0.05. The importance of a prognostic factor was assessed via Wald-type test statistics, the hazard ratio and its 95% confidence interval for survival. A level of 5% of significance was used for both biomedical and HRQOL variables. The replication stability of the final model predicting OS was also investigated, using a bootstrap re-sampling procedure as proposed by Sauerbrei et al. [21] applied in the context of HRQOL by Van Steen et al. [13]. This technique generates a number of simulation datasets (each the same size as the original data set), by randomly selecting patients and replacing them before selecting the next patient (i.e. bootstrap resampling). The frequency of inclusion of the component variables in the Cox PH regression models, including all the selected covariates and stratified for treatment, fitted to each of these data sets using automatic forward stepwise selection (entry level of 
= 0.05), can be considered to be indicative for the importance of the factors [21]. However, as we are also interested in the best set of prognostic factors rather than in the best independent prognostic factor, we need to account for the correlation structure of the potential prognostic factors under consideration. Therefore, we calculated the model selection probabilities based on how many times a permissible model was selected in the bootstrap samples. These probabilities were then used as weights to obtain weighted averaged parameters [22]. All data analyses were performed using SAS version 9.1.
|
results |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionappendixAcknowledgementsReferences |
|---|
Clinical results have been reported separately [14], however, median survival times for the treatment arms were as follows arm A, 8.1 months; arm B, 8.9 months; arm C, 6.7 months. Response rates were 32%, 37% and 28% for arm A, B and C, respectively.
Characteristics of the 391 patients analyzed with HRQOL baseline assessment are reported in Table 1. Patient characteristics at baseline in terms of stage, histology, treatment and performance status were well balanced amongst patients for whom HRQOL data were available or not. Baseline HRQOL scores were overall similar to other groups of patients with the same disease [23]. There was a significant difference in median survival between patients with and without baseline HRQOL data (data not shown). There were 302 deaths reported in the 391 patients with information available for all baseline variables considered and a valid baseline HRQOL form. This allows for an adequately powered analysis.
|
univariate analysis of survival
Amongst the sociodemographic and biomedical variables only performance status independently predicted survival (P = 0.013). Only emotional functioning, nausea/vomiting and coughing were not of prognostic value in the univariate analysis. Details are reported in Table 2.
|
multivariate analysis of survival
The final Cox regression model retained performance status (P = 0.032), gender (P = 0.03), patients' self reported pain (P < 0.001) and dysphagia (P = 0.003). Details are reported in Table 3.
|
bootstrap Model Averaging (MA)
To have greater insight into the stability of the final Cox multivariate model and evaluate the importance of a single variable being included as an independent factor, a state of the art bootstrap model averaging technique based on 5000 bootstrap-generated simulation datasets was run. Due to the missing data on the physical functioning scale, this was not included in additional bootstrap analysis. The results of the inclusion frequencies are listed in Table 4. This table shows the weighted averaged parameters as well as estimates obtained from the most likely model and the full model containing all variables. Interestingly the highest inclusion frequencies were: pain (97.9%), dysphagia (78.5%), gender (62.2%) and performance status (52.2%). The recorded inclusion frequencies highlight the importance of a single variable being included as an independent factor in the model. However, model selection probabilities do provide information about the joint occurrence of variables. Inspection of Table 5, reporting the top ten selected models out of the 5000 bootstrap generated simulation datasets, confirms that the one containing pain, dysphagia, gender and performance status is the most adequate. This evidence further strengthens the results obtained with the classical Cox regression analysis.
|
|
risk groups categorization and survival
Based on the final multivariate model, a prognostic group categorization was also developed. For pain and dysphagia we took the median as cut off values, namely a score of 40 and 10 points respectively. The 1 year survival rate for the patients in the better, intermediate and worse prognostic group was 50%, 30.6% and 18.3% respectively. Details are reported in Figure 1.
|
|
discussion |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionappendixAcknowledgementsReferences |
|---|
The main finding of this analysis is that patients' reported pain and dysphagia provide independent prognostic information for survival in advanced NSCLC patients. We observed an 11% and 12% increase in the hazard of death for every 10-point shift worse (i.e. higher score) in the scales measuring pain and dysphagia, respectively (Table 3).
The large patient sample (n = 391) consisted of a homogeneous population of NSCLC from a prospective multicenter randomized trial with sufficient follow-up and mature survival data allowing for an adequately powered analysis. Data collection procedure in this prospective study was of high quality, employing two HRQOL measures with robust psychometric properties. Furthermore, a recently developed state of the art statistical approach has been used to support and confirm the results obtained with the classical Cox analysis which is typically used in HRQOL prognostic factor studies. This statistical approach has been used in recent HRQOL prognostic factor studies providing insights into the stability of final predictive model [6, 24].
The present findings have been observed whilst also taking into account key previously known socio-demographic and biomedical prognostic data. The original large randomized trial of this dataset previously identified (in the multivariate analysis) performance status as the only key factor predicting survival in this population [14]. Smit et al. [14], showed that patients with performance status (PS) 0 to 1 had a median survival of 8.5 months, whereas those with PS 2 had a median survival of 3.3 months (P < 0.0001). In order to test the prognostic value of HRQOL parameters, we used the same starting sociodemographic and biomedical variables (as well as cut-off criteria) of the ones previously reported [14]. In addition to PS, the present analysis also identified patients' self-reported pain and dysphagia as well as gender as further independent prognostic information for duration of survival. Our finding of female gender and good PS as favorable independent prognostic factors of survival are also supported by a previously conducted large prognostic factor study involving 2,531 advanced NSCLC patients [25].
Based on the variables in the multivariate model, three prognostic groups were also identified (Figure 1). This risk group categorization, once further validated, may help health care providers to make more accurate prognosis in this population.
Research in NSCLC has provided interesting, but often conflicting results. Some studies previously used the EORTC QLQ-C30 to assess the prognostic value of HRQOL parameters. Herndon et al. [10] in a baseline sample of 206 advanced NSCLC patients, found that in the multivariate analysis only pain was predictive of survival with an increase of 13% in the hazard of death for 20 point increase in the scale. This finding is in line with our results, confirming patients' self reported pain as a key prognostic factor in this population. In our study, pain was selected as an independent prognostic variable in 97.9% of the 5000 simulation multivariate analyses (Table 4) and was also the only consistent factor being included in all the top ten models (Table 5). Langendijk et al. [26] analyzing 129 inoperable NSCLC patients of mixed stages (I, II, IIIa, IIIb) before undergoing treatment of high dose radiotherapy, found that the global QOL scale was the stronger prognostic factor in those who had pathological lymph nodes. Maione et al. [27] although measured HRQOL using the EORTC QLQ-C30 and the QLQ-LC13, only included in the prognostic factor analysis the global QOL scale of the EORTC QLQ-C30 without taking into account any other possible relevant prognostic scales. They found the pretreatment global QOL score was an independent predictor of survival in a large population of elderly advanced NSCLC patients. Brown et al. [28] identified global QOL, role functioning, fatigue, appetite loss and constipation as independent predictors of survival in a population of 239 NSCLC patients. Dysphagia was found to be prognostic in the univariate analysis but not in the final model. However, as a HRQOL prognostic factor investigation was not the main objective of their study, few details are provided regarding the analysis. Yet Montazeri et al. [11] found that only global QOL scale was significantly associated with survival in the multivariate analysis of 129 patients. However, due to possible harmful multicollinearity, previous evidence raised concerns about the interpretation of the global QOL scale of the EORTC QLQ-C30 being included in the final multivariate model [13]. Given this, it is difficult to clearly interpret the results of the above-mentioned studies which found this global QOL variable being prognostic.
The conflicting results in the literature could reflect the different methodologies used to analyze the data (as an example by selecting different cut off values for variables) or, alternatively, different selection of parameters to be included in the regression analysis. Nevertheless, the published studies do provide complementary evidence of the significant association between patients' self reported status using HRQOL parameters and survival. Further, other methodologically sound studies that have used different HRQOL self reported measures in NSCLC patients also confirm this robust association. Eton et al. [12] using the Functional Assessment of Cancer Therapy-Lung questionnaire found that a higher baseline score on the physical well being scale was independently associated to a lower risk of death (risk ratio of 0.95, P < 0.001).
This research has limitations, one being that the median survival of the patients who provided HRQOL baseline data, in the original trial, was higher than patients who did not. Given this, it is possible that the sample of our study might reflect patients with better baseline health condition. Hence, some caution is needed as our sample might not be representative of the entire study population.
Whilst investigating the reasons underlying the association between HRQOL parameters and survival lies beyond the scope of this research, it is possible to speculate that, at least, patients' self assessment of their own health status provides a good and strong indicator of their prognosis independent of previously known traditional biomedical parameters. Further research is needed to identify specific HRQOL parameters being prognostic in various cancer disease sites; however, current evidence confirms the strong and independent link between patients' self reported health status and survival in advanced disease.
In conclusion, our findings indicate that patients' self-reported pain and dysphagia independently predict overall survival in advanced NSCLC. This analysis provides an evidence-based rationale for collecting HRQOL data in routine clinical practice as these could offer additional useful information for clinical decision-making.
|
appendix |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionappendixAcknowledgementsReferences |
|---|
The following investigators contributed patients to this study: M. Koolen (Academisch Medisch Centrum, Amsterdam, The Netherlands); F. Wilschut (Ziekenhuis Gelderse Vallei, Bennekom, The Netherlands); J. Stigt (Sophia Ziekenhuis, Zwolle, The Netherlands); A. Van Bochove (Ziekenhuis De Heel, Zaandam, The Netherlands); W. Pieters (Elkerliek Ziekenhuis, Helmond, The Netherlands); N.J.J. Schlosser (Universitair Medisch Centrum, Utrecht, The Netherlands); A. Price (Western General Hospital, Edinburgh, United Kingdom); R. Schipper (Catharina Ziekenhuis, Eindhoven, The Netherlands); M.-A. Haller (CHRU De Nancy–Hopitaux de Brabois, Nancy, France); A. Lukker (St. Maartens Gasthuis, Venlo); J. Bozzino (Newcastle General Hospital, Newcastle-on-Tyne, United Kingdom); N. Van Walree (Ziekenhuis De Baronie, Breda, The Netherlands); H. Belderbos (St. Ignatius Ziekenhuis, Breda, The Netherlands); P. Zatloukal (University Hospital Bulovka, Prague, Czech Republic); M. Möllers (Gelre Ziekenhuizen–Lukas locatie, Apeldoorn, The Netherlands), H. Dik (Rijnland Ziekenhuis, Leiderdorp, The Netherlands); V. Spataro (Ospedale San Giovanni, Bellinzona, Switzerland); H.B. Kwa (Onze Lieve Vrouw Gasthuis, Amsterdam, The Netherlands); D. Galdermans (Algemeen Ziekenhuis Middelheim, Antwerpen, Belgium); L. Willems (Leiden University Medical Centre, Leiden, the Netherlands); and B. Rapoport (The Medical Oncology Centre of Rosebank, Rosebank, South Africa).
|
Acknowledgements |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionappendixAcknowledgementsReferences |
|---|
This publication is supported in part by grants number 2U10 CA11488–31 through 5U10 CA11488–35 from the National Cancer Institute (Bethesda, Maryland, USA). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute. Bristol Myers Squibb and Ely Lilly provided paclitaxel and gemcitabine as investigational agents free of charge in the original trial. We also thank all the EORTC Lung Cancer Group members for enrolling patients into the study (see Appendix).
Received for publication May 19, 2006. Revision received June 26, 2006. Accepted for publication June 27, 2006.
|
References |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionappendixAcknowledgementsReferences |
|---|
1. Taphoorn MJ, Stupp R, Coens C, et al. (2005) Health-related quality of life in patients with glioblastoma: a randomised controlled trial. Lancet Oncol 6:937–944.[CrossRef][Web of Science][Medline]
2. Seidman AD, Portenoy R, Yao TJ, et al. (1995) Quality of life in phase II trials: a study of methodology and predictive value in patients with advanced breast cancer treated with paclitaxel plus granulocyte colony-stimulating factor. J Natl Cancer 87:1316–1322.
3. Chiarion-Sileni V, Del Bianco P, De Salvo GL, et al. (2003) Quality of life evaluation in a randomised trial of chemotherapy versus bio-chemotherapy in advanced melanoma patients. Eur J Cancer 39:1577–1585.[CrossRef][Web of Science][Medline]
4. Maisey NR, Norman A, Watson M, et al. (2002) Baseline quality of life predicts survival in patients with advanced colorectal cancer. Eur J Cancer 38:1351–1357.[CrossRef][Web of Science][Medline]
5. Fang FM, Liu YT, Tang Y, et al. (2004) Quality of life as a survival predictor for patients with advanced head and neck carcinoma treated with radiotherapy. Cancer 100:425–432.[CrossRef][Web of Science][Medline]
6. Efficace F, Therasse P, Piccart M, et al. (2004) Health-related quality of life parameters as prognostic factors in a non-metastatic breast cancer population. An international multicenter study. J Clin Oncol 22:3381–3388.
7. Glare P, Virik K, Jones M, et al. (2003) A systematic review of physicians' survival predictions in terminally ill cancer patients. BMJ 327:195.
8. Greenlee RT, Murray T, Bolden S, et al. (2000) Cancer Statistics, 2000. CA Cancer J Clin 50:7–33.[Abstract]
9. Dancey J, Shepherd FA, Gralla RJ, et al. (2004) Quality of life assessment of second-line docetaxel versus best supportive care in patients with non-small-cell lung cancer previously treated with platinum-based chemotherapy: Results of a prospective, randomized phase III trial. Lung Cancer 43:183–194.[CrossRef][Web of Science][Medline]
10. Herndon JE, Fleishman S, Kornblith AB, et al. (1999) Is quality of life predictive of the survival of patients with advanced non-small cell lung carcinoma? Cancer 85:333–340.[CrossRef][Web of Science][Medline]
11. Montazeri A, Milroy R, Hole D, et al. (2001) Quality of life in lung cancer patients: as an important prognostic factor. Lung Cancer 31:233–240.[CrossRef][Web of Science][Medline]
12. Eton DT, Fairclough DL, Cella D, et al. (2003) Early change in patient-reported health during lung cancer chemotherapy predicts clinical outcomes beyond those predicted by baseline report: results from Eastern Cooperative Oncology Group Study 5592. J Clin Oncol 21:1536–1543.
13. Van Steen K, Curran D, Kramer J, et al. (2002) Multicollinearity in prognostic factor analyses using the EORTC QLQ-C30: identification and impact on model selection. Stat Med 21:3865–3884.[CrossRef][Web of Science][Medline]
14. Smit EF, van Meerbeeck JP, Lianes P, et al. (2003) Three-arm randomized study of two cisplatin-based regimens and paclitaxel plus gemcitabine in advanced non-small-cell lung cancer: a phase III trial of the European Organization for Research and Treatment of Cancer Lung Cancer Group–EORTC 08975. J Clin Oncol 21:3909–3917.
15. Aaronson NK, Ahmedzai S, Bergman B, et al. (1993) The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 85:365–376.
16. Bergman B, Aaronson NK, Ahmedzai S, et al. (1994) The EORTC QLQ-LC13: a modular supplement to the EORTC Core Quality of Life Questionnaire (QLQ-C30) for use in lung cancer clinical trials. EORTC Study Group on Quality of Life. Eur J Cancer 30A:635–642.
17. Fayers P, Aaronson N, Bjordal K, et al. (2001) EORTC QLQ-C30 Scoring Manual. 3rd Edition (EORTC Publications, Brussels, Belgium).
18. Cull A, Sprangers M, Bjordal K, et al. (2002) EORTC Quality of Life Group Translation Procedure. (EORTC Publications, Brussels).
19. Cox DR. (1972) Regression models and life tables. J Royal Stat Soc 4:187–220.
20. Hess KR. (1995) Graphical methods for assessing violations of the proportional hazards assumption in Cox regression. Stat Med 14:1707–1723.[Web of Science][Medline]
21. Sauerbrei W and Schumacher M. (1992) A bootstrap resampling procedure for model building: application to the Cox regression model. Stat Med 11:2093–2109.[Web of Science][Medline]
22. Augustin NH, Sauerbrei W, Schumacher M. (2002) Incorporating model selection into prognostic factor model predictions. Freiburg Center for Data Analysis and Modeling, Preprint no. 76. (http://www.fdm.uni-freiburg.de/pub/preprints.php).
23. Fayers P, Weeden S, Curran D. (1998) EORTC QLQ-C30 Reference Values. (EORTC Publications, Brussels, Belgium).
24. Efficace F, Bottomley A, Coens C, et al. (2006) Does a patient's self-reported health-related quality of life predict survival beyond key biomedical data in advanced colorectal cancer? Eur J Cancer 42:42–49.[CrossRef][Web of Science][Medline]
25. Albain KS, Crowley JJ, LeBlanc M, et al. (1991) Survival determinants in extensive-stage non-small-cell lung cancer: the Southwest Oncology Group experience. J Clin Oncol 9:1618–1626.[Abstract]
26. Langendijk H, Aaronson NK, de Jong JM, et al. (2000) The prognostic impact of quality of life assessed with the EORTC QLQ-C30 in inoperable non-small cell lung carcinoma treated with radiotherapy. Radiother Oncol 55:19–25.[CrossRef][Web of Science][Medline]
27. Maione P, Perrone F, Gallo C, et al. (2005) Pretreatment quality of life and functional status assessment significantly predict survival of elderly patients with advanced non-small-cell lung cancer receiving chemotherapy: a prognostic analysis of the multicenter Italian lung cancer in the elderly study. J Clin Oncol 23:6865–6872.
28. Brown J, Thorpe H, Napp V, et al. (2005) Assessment of quality of life in the supportive care setting of the big lung trial in non-small-cell lung cancer. J Clin Oncol 23:7417–7427.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Basch, X. Jia, G. Heller, A. Barz, L. Sit, M. Fruscione, M. Appawu, A. Iasonos, T. Atkinson, S. Goldfarb, et al. Adverse Symptom Event Reporting by Patients vs Clinicians: Relationships With Clinical Outcomes J Natl Cancer Inst, November 17, 2009; (2009) djp386v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. G. Blanco, I. S. Toste, R. F. Alvarez, G. R. Cuadrado, A. M. Gonzalvez, and I. J. G. Martin Age, comorbidity, treatment decision and prognosis in lung cancer Age Ageing, November 1, 2008; 37(6): 715 - 718. [Full Text] [PDF]
|
|
|
|
 |
|
 F. Efficace, P. F. Innominato, G. Bjarnason, C. Coens, Y. Humblet, S. Tumolo, D. Genet, M. Tampellini, A. Bottomley, C. Garufi, et al. Validation of Patient's Self-Reported Social Functioning As an Independent Prognostic Factor for Survival in Metastatic Colorectal Cancer Patients: Results of an International Study by the Chronotherapy Group of the European Organisation for Research and Treatment of Cancer J. Clin. Oncol., April 20, 2008; 26(12): 2020 - 2026. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Gotay, C. T. Kawamoto, A. Bottomley, and F. Efficace The Prognostic Significance of Patient-Reported Outcomes in Cancer Clinical Trials J. Clin. Oncol., March 10, 2008; 26(8): 1355 - 1363. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. E. Jones and C. C. Doebbeling Beyond the Traditional Prognostic Indicators: The Impact of Primary Care Utilization on Cancer Survival J. Clin. Oncol., December 20, 2007; 25(36): 5793 - 5799. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||