|Year : 2019 | Volume
| Issue : 1 | Page : 21-27
Quality-adjusted time without symptoms or toxicity (Q-TWiST) analysis of a Phase III randomized trial to compare the benefit of gefitinib versus pemetrexed/carboplatin for epidermal growth factor receptor-mutated non-small cell lung cancer
Vijay Patil1, Vanita Noronha1, Amit Joshi1, Anuradha Chougule1, Atanu Bhattacharjee2, Alok Goel1, Vikas Talreja1, Nandini Menon1, Anant Ramaswamy1, Ashay Karpe1, Nikhil Pande1, Arun Chandrasekharan1, Abhishek Mahajan3, Amit Janu3, Nilendu Purandare4, Kumar Prabhash1
1 Department of Medical Oncology, Tata Memorial Hospital, Mumbai, Maharashtra, India
2 Department of Cancer Epidemiology, Tata Memorial Hospital, Mumbai, Maharashtra, India
3 Department of Radiology, Tata Memorial Hospital, Mumbai, Maharashtra, India
4 Department of Nuclear Medicine, Tata Memorial Hospital, Mumbai, Maharashtra, India
|Date of Web Publication||9-Sep-2019|
Department of Medical Oncology, Tata Memorial Hospital, Parel, Mumbai - 400 012, Maharashtra
Source of Support: None, Conflict of Interest: None
Background:> In an open-label, Phase III randomized study, gefitinib was found to be superior to pemetrexed-platinum in terms of progression-free survival in patients with epidermal growth factor receptor (EGFR)-mutated non-small cell lung cancer. In this analysis, we have assessed the efficacy of gefitinib over pemetrexed-platinum using quality-adjusted time without symptoms or toxicity (Q-TWiST) of treatment analysis method.
Methods: The overall survival in each arm was partitioned into the following three health states: time spent in Grade 2 or above toxicity after randomization and before progression (TOX state), time spent after randomization and before progression without Grade 2 or above toxicity (TWiST state), and time spent after progression (REL state). The mean Q-TWiST was calculated for each arm using utility coefficients of 0.65 for TOX, 0.71 for TWiST, and 0.67 for REL states. The difference in Q-TWiST and the 95% confidence interval (CI) of the difference were calculated using a nonparametric bootstrap.P= 0.05 was considered statistically significant.
Results: The mean TOX duration (37.6 vs. 16.2 days,P < 0.001) and TWiST duration (211.7 vs. 162.2 days, P = 0.002) were higher in the gefitinib arm. The mean REL state duration was higher in the pemetrexed-carboplatin arm (164.4 vs. 74.3 days,P < 0.001). The mean Q-TWiST was 224.6 days in the gefitinib arm versus 235.9 days in the pemetrexed-carboplatin arm. The difference was 11.3 days; 95% CI, 6.920–29.564,P= 0.224.
Conclusion: The mean Q-TWiST of patients in the gefitinib arm was similar to that of patients in the pemetrexed-carboplatin arm. These results challenge the superiority of sequencing gefitinib followed by chemotherapy over pemetrexed-carboplatin followed by gefitinib in terms of Q-TWiST.
Keywords: Carboplatin, epidermal growth factor receptor, epidermal growth factor receptor mutation, gefitinib, non-small cell lung cancer, pemetrexed, quality-adjusted time without symptoms or toxicity
|How to cite this article:|
Patil V, Noronha V, Joshi A, Chougule A, Bhattacharjee A, Goel A, Talreja V, Menon N, Ramaswamy A, Karpe A, Pande N, Chandrasekharan A, Mahajan A, Janu A, Purandare N, Prabhash K. Quality-adjusted time without symptoms or toxicity (Q-TWiST) analysis of a Phase III randomized trial to compare the benefit of gefitinib versus pemetrexed/carboplatin for epidermal growth factor receptor-mutated non-small cell lung cancer. Cancer Res Stat Treat 2019;2:21-7
|How to cite this URL:|
Patil V, Noronha V, Joshi A, Chougule A, Bhattacharjee A, Goel A, Talreja V, Menon N, Ramaswamy A, Karpe A, Pande N, Chandrasekharan A, Mahajan A, Janu A, Purandare N, Prabhash K. Quality-adjusted time without symptoms or toxicity (Q-TWiST) analysis of a Phase III randomized trial to compare the benefit of gefitinib versus pemetrexed/carboplatin for epidermal growth factor receptor-mutated non-small cell lung cancer. Cancer Res Stat Treat [serial online] 2019 [cited 2020 Aug 3];2:21-7. Available from: http://www.crstonline.com/text.asp?2019/2/1/21/266442
| Introduction|| |
Epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) such as gefitinib, erlotinib, and afatinib are the current standard of care for the systemic first-line treatment of patients with activating EGFR-mutated non-small cell lung cancer (NSCLC).,,,,,, Multiple studies done across the globe have compared these agents to platinum-doublet chemotherapy. Oral EGFR TKIs have consistently and independently improved progression-free survival (PFS) without an improvement in overall survival (OS). Pemetrexed-platinum-doublet chemotherapy followed by maintenance pemetrexed is the preferred chemotherapy regimen in nonsquamous NSCLC; this regimen had not been compared to an EGFR TKI. We have previously reported a Phase III randomized study comparing pemetrexed-platinum-doublet chemotherapy followed by maintenance pemetrexed versus gefitinib (Pem-Gef). The results were consistent with the existing literature and showed that gefitinib improved PFS without improving OS.
Another finding consistent across all studies done with upfront TKI versus a platinum doublet is the nature of toxicity. The doublet chemotherapy leads to a higher incidence of hematological toxicity compared to TKIs, while the TKIs cause a higher incidence of skin rash, diarrhea, and fatigue.,, In randomized studies, the maximum grade of adverse events are reported; the temporal profile of the side effects is largely ignored. The hematological side effects that are caused by chemotherapy in these studies are frequently short-lived adverse events. Conversely, toxicities caused by the oral TKIs are of longer duration and are of a waxing and waning type, often hampering the quality of life of these patients. The simplicity of capturing the frequency of maximum-grade adverse events fails to capture the picture of recurrent adverse events and the duration of each event., The limitations of worst-grade adverse event reporting can be overcome by using the quality-adjusted time without symptoms or toxicity (Q-TWiST) of treatment analysis. Q-TWiST is a useful method to describe adverse events associated with targeted therapy which have a prolonged duration of administration.,, Whether the improvement in PFS seen with oral TKIs would still be significant if the temporal profile of toxicities is taken into account is a question which needs to be answered. Hence, we planned an exploratory post hoc Q-TWiST analysis of our previously published Phase III study to compare the Q-TWiST between the two arms (gefitinib and pemetrexed-carboplatin). The null hypothesis of this analysis was that the restricted mean duration of Q-TWiST health state in the gefitinib arm would not be higher than that in patients in the pemetrexed-carboplatin arm.
| Methods|| |
Pem-Gef was an open-label, Phase III randomized study, the details of which have already been published. The methodology of the trial, in brief, was as follows: patients with newly diagnosed lung adenocarcinoma harboring an activating EGFR mutation were randomly assigned in a 1:1 fashion to either gefitinib (250 mg orally daily) or chemotherapy consisting of pemetrexed (500 mg/m 2 intravenously [IV] 3-weekly) and carboplatin (area under the curve [AUC] 5 IV 3 weekly). Response assessment scans (axial imaging with contrast-enhanced computed tomography scans of the thorax and upper abdomen) were performed every 6 weeks for the first 12 weeks and then every 8 weeks until objective progressive disease (PD) was documented. Response assessment was done according to the Response Evaluation Criteria in Solid Tumors criteria, version 1.1. The adverse events were graded according to the Common Terminology Criteria for Adverse Events (CTCAE) version 4.02. Patients post progression in the gefitinib arm were offered pemetrexed-carboplatin if they were considered to be fit to receive the combination chemotherapy. Patients in the pemetrexed-carboplatin arm were offered gefitinib at progression. The patients were followed up till death.
The study protocol was approved by the institutional ethics committee (IEC-1) of Tata Memorial Hospital (Mumbai, Maharashtra, India). The trial was registered with the Clinical Trials Registry-India (CTRI/2015/08/006113). All patients provided a written informed consent prior to accrual, and the study was conducted in accordance with good clinical practice guidelines and the Declaration of Helsinki.
Health state definitions
OS time was partitioned into three health states for the purpose of the Q-TWiST analysis. These three health states included the time spent in Grade 2 or above toxicity after randomization and before progression (TOX state), time spent after randomization and before progression without Grade 2 or above toxicity (TWiST state), and time spent after progression (REL state). TOX state was defined as the number of days spent in toxicity after randomization, but before progression or before censoring for progression. The TOX state was the cumulative number of days spent in Grade 2 or above adverse events. The TOX duration was the cumulative sum of the time spent with nonoverlapping adverse events. For example, if a patient had Grade 3 rash for 10 days in the 1st month of treatment and then developed Grade 3 diarrhea for 4 days in the 3rd month of treatment, then the TOX duration for this patient was 14 days. When a patient had overlapping toxicities, the duration taken was from the start of the first adverse event till the end of the last adverse event. For example, if a patient had Grade 3 rash for 10 days and on the 8th day of the rash developed Grade 3 diarrhea for 5 days, then the duration of TOX for this patient was 13 days. Patients who did not have adverse events of Grade 2 or above were assigned a TOX duration of zero. TWiST state was defined as the duration in days spent after randomization, but before progression without toxicity of treatment. TWiST state was the duration of PFS without the time spent in TOX state. REL state was defined as the duration in days spent after the first progression until death. REL state was the interval between PFS and OS.
PFS was defined as the time from randomization to the first documentation of objective PD, change in treatment, or death from any cause. Patients who had not progressed at the time of analysis were censored at their last date of follow-up. OS was defined as the time from randomization to death from any cause. Patients who had not died at the time of analysis were censored at their last date of follow-up. For the current analysis, the database was locked on June 30, 2017.
Quality-adjusted time without symptoms or toxicity analysis
RStudio version 3.1.2 (R Core Team, Vienna, Austria) was used for this analysis. The Q-TWiST function written by Bogart et al. was used for this analysis. The restricted mean health state duration for the three states was calculated for both arms, and the 95% confidence interval (CI) for the difference between the two arms was calculated using a nonparametric bootstrap method. Five hundred bootstraps were used for this calculation. P≤ 0.05 was considered significant difference. The Kaplan–Meier time-to-event estimation method was used in estimating and plotting the graphs of TOX state, PFS, and OS for each arm. The graphs of the three states in each arm were overlaid on a single plot for each arm.
The mean Q-TWiST for each arm was calculated using the following formula:
Mean Q-TWiST= μTOX × restricted mean TOX + μTWiST × restricted mean TWIST + μREL × restricted mean REL.
μTOX, μTwiST, and μREL stand for the utility scores for TOX, TWiST, and REL health states, respectively. The values of μTWiST and μREL used for this analysis were 0.71 and 0.67, respectively. These values were chosen based on the work done by Chouaid et al. The mean utility score preprogression on first-line therapy (μTWiST) reported in their study was 0.71, and the mean utility score reported postprogression (μREL) was 0.67. The utility score for TOX was unfortunately not reported for lung cancer patients. We chose a utility score of 0.65 for the TOX state based on the work reported by Nafees et al.
The median follow-up in patients in the current study was 740 days. Hence, the time duration considered for the calculation of primary Q-TWiST analysis was taken as 750 days. The difference in mean Q-TWiST scores with the 95% CI and the corresponding P value between the two arms was calculated. P≤ 0.05 was considered statistically significant. In addition, the survival plot of Q-TWiST for both arms was plotted on a single graph. Additional exploratory analysis was also done with time duration taken at 400 days and 1100 days, to help understand the outcome at 1 year and 3 years, respectively.
Threshold utility analysis
Threshold utility analysis was done to calculate the Q-TWiST value as per the formula given above. The value of the threshold utility coefficient is unknown in the Indian population, and it varies from 0 to 1. Hence, the threshold utility analysis was performed using different weightage of utility values between 0 and 1 in 0.25 increments for μTOX and μREL. A score of 1 denoted time of perfect health, whereas a score of 0 denoted time period similar to death. This analysis was restricted to the time duration considered for the calculation of primary Q-TWiST analysis, i.e., 750 days.
The incidence plot analysis was planned post hoc. Two toxicities which were higher in the gefitinib arm, i.e., skin rash and diarrhea, were selected. The status of each of these toxicities in the first seven visits was documented, and an AUC was calculated. On the Y axis, the proportion of patients having the corresponding adverse event was shown, whereas on the X axis, the visits were plotted. The AUC between the two arms was compared for each toxicity. P≤ 0.05 was considered statistically significant.
| Results|| |
A total of 290 patients were enrolled in the study, with 145 patients in each arm. The mean ages were 54 and 53 years in the gefitinib and pemetrexed-carboplatin arms, respectively. The Eastern Cooperative Oncology Group Performance Status was 0–1 in 135 patients (93.1%) in the gefitinib arm and 138 patients (95.2%) in the pemetrexed-carboplatin arm. Any-grade adverse event occurred in 128 patients (88.3%) in the gefitinib arm versus 133 patients (91.7%) in the pemetrexed-carboplatin arm; P= 0.434. Grade 2 or above adverse events developed in 92 patients (62.3%) in the gefitinib arm and in 88 patients (60.7%) in the pemetrexed-carboplatin arm.
Health state duration
[Figure 1] and [Figure 2] show the partitioned OS curves for the gefitinib arm and the pemetrexed-carboplatin arm, respectively. The mean TOX duration was higher in the gefitinib arm (38 vs. 17.6 days, P= 0.003). Similarly, the mean TWiST duration was higher in the gefitinib arm (266.7 vs. 184.1 days, P < 0.001). Conversely, the mean REL state duration was higher in the pemetrexed-carboplatin arm (320.7 vs. 171.8 days, P < 0.001) [Table 1]. A similar pattern of higher mean TOX and TWiST with gefitinib and higher mean REL with pemetrexed-carboplatin was seen at 400 and 1100 days, respectively [Table 1]S, [Table 2]S, [Figure 1], [Figure 2], [Figure 3] and [Figure 1]S, [Figure 2]S, [Figure 3]S, [Figure 4]S.
|Figure 1: Partitioned overall survival curve for gefitinib arm with duration of follow-up limited to 750 days|
Click here to view
|Figure 2: Partitioned overall survival curve for pemetrexed carboplatin arm with duration of follow-up limited to 750 days|
Click here to view
|Table 1: The restricted mean health duration in days of the time spent in Grade 2 or above toxicity after randomization and before progression, time spent after randomization and before progression without Grade 2 or above toxicity, and time spent after progression until death states in patients treated in the gefitinib arm compared to patients treated in the pemetrexed-carboplatin arm|
Click here to view
|Figure 3: Quality-adjusted time without symptoms or toxicity curve of both arms with duration of follow-up till 750 days|
Click here to view
Quality-adjusted time without symptoms or toxicity
The mean Q-TWiST was 329.1 days (standard error, 14.5) in the gefitinib arm and 357 days in the pemetrexed-carboplatin arm (absolute difference, 27.9; 95% CI, 11.9–67.6, P= 0.169) in the analysis with a time duration of 750 days [Table 2]. [Figure 3] depicts the Q-TWiST curve for both arms. From this figure, it is evident that the curves are overlapping.
|Table 2: The mean quality-adjusted time without symptoms or toxicity in gefitinib and pemetrexed-carboplatin arms at different time durations|
Click here to view
The mean Q-TWiST differences at 400 and 1100 days were 11.3 days (95% CI, 6.92–29.56; P= 0.224) and 34.9 days (95% CI, 20–89.9; P= 0.213), respectively, both numerically favoring the pemetrexed-carboplatin arm [Table 2]. The results revealed a trend in favor of the pemetrexed-carboplatin arm, and the difference was more evident with extended time durations. [Figure 5]S and [Figure 6]S depict the Q-TWiST curves with extended time durations of 400 and 1100 days.
Threshold utility analysis
[Table 3] depicts the results of the threshold utility analysis. The Q-TwiST analysis was in favor of gefitinib when the utility coefficients were below 0.5 for REL.
Incidence plot analysis
The incidence plot analysis revealed that both rash and diarrhea were higher in the gefitinib arm. [Figure 7]S and [Figure 8]S show the AUC for diarrhea and skin rash, respectively. The difference in the AUC for rash and diarrhea was 10 units (P = 0.192) and 21.89 units (P = 0.09), respectively, signifying higher toxicity in gefitinib arm.
| Discussion|| |
To the best of our knowledge, this is the first study reporting on Q-TWiST analysis on palliative systemic therapy in activating EGFR-mutated NSCLC. The result of the Q-TWiST analysis was surprising. The mean Q-TWiST of the gefitinib arm was not superior to that of the pemetrexed-carboplatin arm. These results challenge the current practice of preferring an oral TKI as the first-line therapy due to the established prolongation of PFS.
However, the interpretation of these results needs to be carried out with the understanding that patients in the gefitinib arm received chemotherapy, preferentially pemetrexed and carboplatin at progression, and patients in the pemetrexed-carboplatin arm received gefitinib at progression. Q-TWiST is a cumulative function of time spent in TOX, TWiST, and REL states. Hence, in addition to the initial treatment, the subsequent treatments can also impact the Q-TwiST results as they affect the REL state. As seen from our analysis, the gefitinib arm had a higher mean TWiST and mean TOX duration, but the duration in REL state was lower. The lower REL state in gefitinib arm represents the outcomes provided by subsequent lines of chemotherapy, whereas the high REL state in pemetrexed-carboplatin arm represents the outcome provided by gefitinib therapy after progression. Hence, the results indicate that the sequence of gefitinib followed by chemotherapy is not superior in terms of Q-TWiST compared to the sequence of pemetrexed and carboplatin chemotherapy as first-line therapy, followed by gefitinib at progression. Similar suggestions about better outcome with sequencing of chemotherapy followed by oral TKI as compared to the sequence of oral TKI upfront and then chemotherapy were reported by An et al. at the World Lung Cancer Congress in 2016 (abstract number MA 16.05). In this analysis, the sequence of chemotherapy followed by EGFR TKI at progression appeared to have a favorable impact on OS as noted in the multivariate analysis (hazard ratio, 0.64, P= 0.065).
The TOX state duration in the gefitinib arm was nearly double of that seen in the pemetrexed-carboplatin arm. Overall, the rate of any-grade adverse events was similar in both arms. Certain toxicities such as skin rash and diarrhea were seen in a higher proportion of patients in the gefitinib arm, whereas hematological toxicities were higher in the chemotherapy arm.,,, Hence, we did an incidence plot analysis of the two most common gefitinib toxicities to understand the temporal profile of these toxicities. These toxicities, depicted in [Figure 7]S and [Figure 8]S, that develop early are seen in a higher proportion of patients and persist for a longer period. As a result, these toxicities contribute to a higher TOX duration in the gefitinib arm. In this study, for the calculation of TOX duration, we considered Grade 2 or above adverse events. We selected Grade 2 events as they are considered moderate, limit age-appropriate instrumental activities of daily living, and have a psychosocial impact on the quality of life of the patient.
The strengths of the study are based on the extensive capture of mature data. Over twenty adverse events, with date of appearance and date of resolution, were captured prospectively, providing adequate data for calculating the TOX duration. The analysis was done on mature data: 257 patients (88.6%) had progressed and 210 patients (72.4%) died. The surprising finding of higher TOX duration in gefitinib arm was reconfirmed by doing a longitudinal toxicity analysis.
The study has some limitations. The Q-TWiST adverse event data were recorded according to the National Cancer Institute's CTCAE, version 4.02, by the treating physicians. Perhaps, the use of patient-reported outcomes in addition to physician assessment of adverse events would further strengthen the analysis. In addition, the Q-TWiST analysis fails to take into account the higher number of visits required by the patient in the case of intravenous chemotherapy when compared to oral TKI treatment. Further, the utility scores for health states used in the study were reported from high-income countries, and hence their applicability to patients from low- and middle-income countries is unknown.,
| Conclusion|| |
The Q-TWiST of gefitinib followed by chemotherapy is not superior to the Q-TWiST of pemetrexed-carboplatin followed by gefitinib. These results question the sequencing of gefitinib followed by chemotherapy over pemetrexed-carboplatin followed by gefitinib in EGFR-mutated NSCLC. Higher TOX duration was seen in the gefitinib arm. Typical toxicities such as skin rash and diarrhea were seen early and persisted for a longer duration in the gefitinib arm.
The authors would like to thank Emily Bogart.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Rosell R, Carcereny E, Gervais R, Vergnenegre A, Massuti B, Felip E, et al.
Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non-small-cell lung cancer (EURTAC): A multicentre, open-label, randomised phase 3 trial. Lancet Oncol 2012;13:239-46.
Patil VM, Noronha V, Joshi A, Choughule AB, Bhattacharjee A, Kumar R, et al.
Phase III study of gefitinib or pemetrexed with carboplatin in EGFR-mutated advanced lung adenocarcinoma. ESMO Open 2017;2:e000168.
Zhou C, Wu YL, Chen G, Feng J, Liu XQ, Wang C, et al.
Erlotinib versus chemotherapy as first-line treatment for patients with advanced EGFR mutation-positive non-small-cell lung cancer (OPTIMAL, CTONG-0802): A multicentre, open-label, randomised, phase 3 study. Lancet Oncol 2011;12:735-42.
Maemondo M, Inoue A, Kobayashi K, Sugawara S, Oizumi S, Isobe H, et al.
Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR. N Engl J Med 2010;362:2380-8.
Mitsudomi T, Morita S, Yatabe Y, Negoro S, Okamoto I, Tsurutani J, et al.
Gefitinib versus cisplatin plus docetaxel in patients with non-small-cell lung cancer harbouring mutations of the epidermal growth factor receptor (WJTOG3405): An open label, randomised phase 3 trial. Lancet Oncol 2010;11:121-8.
Sequist LV, Yang JC, Yamamoto N, O'Byrne K, Hirsh V, Mok T, et al.
Phase III study of afatinib or cisplatin plus pemetrexed in patients with metastatic lung adenocarcinoma with EGFR mutations. J Clin Oncol 2013;31:3327-34.
Rajendra A, Noronha V, Joshi A, Patil VM, Menon N, Prabhash K. Epidermal growth factor receptor-mutated non-small-cell lung cancer: A primer on contemporary management. Cancer Res Stat Treat 2019;2:36-53. [Full text]
Cabarrou B, Boher JM, Bogart E, Tresch-Bruneel E, Penel N, Ravaud A, et al.
How to report toxicity associated with targeted therapies? Ann Oncol 2016;27:1633-8.
Thanarajasingam G, Atherton PJ, Novotny PJ, Loprinzi CL, Sloan JA, Grothey A, et al.
Longitudinal adverse event assessment in oncology clinical trials: The toxicity over time (ToxT) analysis of alliance trials NCCTG N9741 and 979254. Lancet Oncol 2016;17:663-70.
Cole BF, Gelber RD, Gelber S, Mukhopadhyay P. A quality-adjusted survival (Q-TWiST) model for evaluating treatments for advanced stage cancer. J Biopharm Stat 2004;14:111-24.
Revicki DA, Feeny D, Hunt TL, Cole BF. Analyzing oncology clinical trial data using the Q-TWiST method: Clinical importance and sources for health state preference data. Qual Life Res 2006;15:411-23.
Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al.
New response evaluation criteria in solid tumours: Revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228-47.
Bogart E, Jouin A, Béhal H, Duhamel A, Filleron T, Kramar A, et al.
Analysis of survival adjusted for quality of life using the Q-TWiST function: Interface in R. Comput Methods Programs Biomed 2016;125:79-87.
Chouaid C, Agulnik J, Goker E, Herder GJ, Lester JF, Vansteenkiste J, et al.
Health-related quality of life and utility in patients with advanced non-small-cell lung cancer: A prospective cross-sectional patient survey in a real-world setting. J Thorac Oncol 2013;8:997-1003.
Nafees B, Stafford M, Gavriel S, Bhalla S, Watkins J. Health state utilities for non-small cell lung cancer. Health Qual Life Outcomes 2008;6:84.
An HJ, Shin Y, Hong SH, Kim YK, Kim SJ, Lee KY, et al
. MA16.05 for EGFR mutant non-small cell lung cancer, treatment sequence matters? J Thorac Oncol 2017;12:S436-7.
Di Maio M, Basch E, Bryce J, Perrone F. Patient-reported outcomes in the evaluation of toxicity of anticancer treatments. Nat Rev Clin Oncol 2016;13:319-25.
Jhita T, Petrou S, Gumber A, Szczepura A, Raymond NT, Bellary S, et al.
Ethnic differences in health related quality of life for patients with type 2 diabetes. Health Qual Life Outcomes 2014;12:83.
Wee HL, Machin D, Loke WC, Li SC, Cheung YB, Luo N, et al.
Assessing differences in utility scores: A comparison of four widely used preference-based instruments. Value Health 2007;10:256-65.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]