The Patient's Journey With Chronic Hepatitis C
The Patient's Journey With Chronic Hepatitis C
Data were analyzed from nine multicenter, multinational phase 3 clinical trials of sofosbuvir-based regimens for the treatment of chronic hepatitis C: POSITRON, FISSION, FUSION, NEUTRINO, VALENCE, PHOTON-1, ION-1, ION-2 and ION-3. HRQL was assessed as the secondary endpoint in these trials. For the purpose of this study, only active treatment arms were included.
From the medical history collected at screening, we identified patients with a pre-treatment history of depression, clinically overt fatigue, anxiety, insomnia, as well as type 2 diabetes or hyperglycemia. Baseline haemoglobin, HCV RNA load, ALT, HCV genotype, the presence of cirrhosis, and treatment-related adverse events were recorded as described previously.
Patients with detectable HCV RNA at post-treatment week 4 were not followed-up at subsequent visits. Patients were considered to have achieved sustained virologic response (SVR-12) if they had undetectable HCV RNA at post-treatment week 12.
In all clinical trials, the Short Form-36 version 2 (SF-36v2) questionnaire was administered at each time point to the study participants in their native language. The QualityMetric Health Outcome Scoring Software v.4.5 (QualityMetric, Lincoln, RI, USA) with the maximum data recovery and 2009 U.S. norms was used.
The SF-36 HRQL instrument includes eight individual scales whose scores range between across 0–100 with higher scores indicating better health. The eight scales are: physical functioning (PF), role physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role emotional (RE), and mental health (MH). The two summary scores, physical component summary score (PCS) and mental component summary score (MCS), summarise the physical and mental health components of HRQL as measured by SF-36. The summary scores were calculated using the individual scores linearly transformed using the population norms to the mean of 50.0 and a standard deviation of 10.0.
From the SF-36 instrument, we also calculated health utility scores. In general, health utility is a generic health status classification system which reflects patients' preference for certain state of health. Health utility scores are typically used for calculation of quality-adjusted years of life in cost-utility and other economic analyses of healthcare interventions. In this study, we calculated SF-6D utility scores using the individual domains of SF-36 and a non-parametric Bayesian algorithm as previously described. For the purpose of uniform representation along with the other HRQL scores, SF-6D scores were transformed from a 0–1 to a 0–100 scale by multiplying the score by 100.
For the purpose of this study, we only included HRQL and health utility scores collected at the time points which were universally used across all the original trials: baseline (day 1), the last day of treatment, post-treatment week 4 and post-treatment week 12 for (in subjects with SVR only).
From the nine trials, we grouped treatment regimens based on the drugs used regardless of treatment duration. Therefore, we had five different treatment regimens: IFN + SOF + RBV, IFN + RBV, SOF + RBV, SOF + RBV + LDV, and SOF + LDV.
Clinico-demographic parameters, baseline HRQL scores and health utilities were summarised and compared across the treatment regimens using chi-square test or Kruskal-Wallis non-parametric test. The changes (decrements or improvements) in HRQL and utilities from patients' own baseline levels were calculated for each patient at each studied time point. The median changes were further compared to zero by a sign rank test for matched pairs. Due to multiplicity of measured parameters and time points, only P-values of 0.001 or less were considered statistically significant. The minimal clinically important difference (MCID) was set to 5% of the range size which is 5 for the individual HRQL domains and approximately 2 for the summary scores.
Independent predictors of baseline HRQL summary scores and SF-6D utilities, treatment-emergent and post-treatment changes in those were assessed using multiple linear regression with the treatment regimens and treatment duration being tested as potential predictors. Bidirectional stepwise selection of predictors with the significance level of 0.001 for stay was used. The list of potential HRQL predictors used for the selection procedure included age, gender, ethnicity, BMI, location, history of psychiatric disorders, type 2 diabetes, baseline haemoglobin, HCV viral load and ALT (at baseline only), cirrhosis, history of prior anti-HCV treatment, and having achieved SVR (at post-treatment week 4 only). All analyses were run in SAS 9.1 (SAS Institute, Cary, NC, USA).
The study was separately approved by each site's Institutional Review Board.
Methods
Data were analyzed from nine multicenter, multinational phase 3 clinical trials of sofosbuvir-based regimens for the treatment of chronic hepatitis C: POSITRON, FISSION, FUSION, NEUTRINO, VALENCE, PHOTON-1, ION-1, ION-2 and ION-3. HRQL was assessed as the secondary endpoint in these trials. For the purpose of this study, only active treatment arms were included.
From the medical history collected at screening, we identified patients with a pre-treatment history of depression, clinically overt fatigue, anxiety, insomnia, as well as type 2 diabetes or hyperglycemia. Baseline haemoglobin, HCV RNA load, ALT, HCV genotype, the presence of cirrhosis, and treatment-related adverse events were recorded as described previously.
Patients with detectable HCV RNA at post-treatment week 4 were not followed-up at subsequent visits. Patients were considered to have achieved sustained virologic response (SVR-12) if they had undetectable HCV RNA at post-treatment week 12.
Health-related Quality of Life
In all clinical trials, the Short Form-36 version 2 (SF-36v2) questionnaire was administered at each time point to the study participants in their native language. The QualityMetric Health Outcome Scoring Software v.4.5 (QualityMetric, Lincoln, RI, USA) with the maximum data recovery and 2009 U.S. norms was used.
The SF-36 HRQL instrument includes eight individual scales whose scores range between across 0–100 with higher scores indicating better health. The eight scales are: physical functioning (PF), role physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role emotional (RE), and mental health (MH). The two summary scores, physical component summary score (PCS) and mental component summary score (MCS), summarise the physical and mental health components of HRQL as measured by SF-36. The summary scores were calculated using the individual scores linearly transformed using the population norms to the mean of 50.0 and a standard deviation of 10.0.
From the SF-36 instrument, we also calculated health utility scores. In general, health utility is a generic health status classification system which reflects patients' preference for certain state of health. Health utility scores are typically used for calculation of quality-adjusted years of life in cost-utility and other economic analyses of healthcare interventions. In this study, we calculated SF-6D utility scores using the individual domains of SF-36 and a non-parametric Bayesian algorithm as previously described. For the purpose of uniform representation along with the other HRQL scores, SF-6D scores were transformed from a 0–1 to a 0–100 scale by multiplying the score by 100.
For the purpose of this study, we only included HRQL and health utility scores collected at the time points which were universally used across all the original trials: baseline (day 1), the last day of treatment, post-treatment week 4 and post-treatment week 12 for (in subjects with SVR only).
Statistical Analysis
From the nine trials, we grouped treatment regimens based on the drugs used regardless of treatment duration. Therefore, we had five different treatment regimens: IFN + SOF + RBV, IFN + RBV, SOF + RBV, SOF + RBV + LDV, and SOF + LDV.
Clinico-demographic parameters, baseline HRQL scores and health utilities were summarised and compared across the treatment regimens using chi-square test or Kruskal-Wallis non-parametric test. The changes (decrements or improvements) in HRQL and utilities from patients' own baseline levels were calculated for each patient at each studied time point. The median changes were further compared to zero by a sign rank test for matched pairs. Due to multiplicity of measured parameters and time points, only P-values of 0.001 or less were considered statistically significant. The minimal clinically important difference (MCID) was set to 5% of the range size which is 5 for the individual HRQL domains and approximately 2 for the summary scores.
Independent predictors of baseline HRQL summary scores and SF-6D utilities, treatment-emergent and post-treatment changes in those were assessed using multiple linear regression with the treatment regimens and treatment duration being tested as potential predictors. Bidirectional stepwise selection of predictors with the significance level of 0.001 for stay was used. The list of potential HRQL predictors used for the selection procedure included age, gender, ethnicity, BMI, location, history of psychiatric disorders, type 2 diabetes, baseline haemoglobin, HCV viral load and ALT (at baseline only), cirrhosis, history of prior anti-HCV treatment, and having achieved SVR (at post-treatment week 4 only). All analyses were run in SAS 9.1 (SAS Institute, Cary, NC, USA).
The study was separately approved by each site's Institutional Review Board.