Incorporating direct and indirect evidence using bayesian methods: an applied case study in ovarian cancer

GetReal in network meta-analysis: a review of the methodology

Pairwise meta-analysis is an established statistical tool for synthesizing evidence from multiple trials, but it is informative only about the relative efficacy of two specific interventions. The usefulness of pairwise meta-analysis is thus limited in real-life medical practice, where many competing interventions may be available for a certain condition and studies informing some of the pairwise comparisons may be lacking. This commonly encountered scenario has led to the development of network meta-analysis (NMA). In the last decade, several applications, methodological developments, and empirical studies in NMA have been published, and the area is thriving as its relevance to public health is increasingly recognized. This article presents a review of the relevant literature on NMA methodology aiming to pinpoint the developments that have appeared in the field. Copyright © 2016 John Wiley & Sons, Ltd.

Topotecan, pegylated liposomal doxorubicin hydrochloride, paclitaxel, trabectedin and gemcitabine for advanced recurrent or refractory ovarian cancer: a systematic review and economic evaluation

BACKGROUND

Ovarian cancer is the fifth most common cancer in the UK, and the fourth most common cause of cancer death. Of those people successfully treated with first-line chemotherapy, 55-75% will relapse within 2 years. At this time, it is uncertain which chemotherapy regimen is more clinically effective and cost-effective for the treatment of recurrent, advanced ovarian cancer.

OBJECTIVES

To determine the comparative clinical effectiveness and cost-effectiveness of topotecan (Hycamtin(®), GlaxoSmithKline), pegylated liposomal doxorubicin hydrochloride (PLDH; Caelyx(®), Schering-Plough), paclitaxel (Taxol(®), Bristol-Myers Squibb), trabectedin (Yondelis(®), PharmaMar) and gemcitabine (Gemzar(®), Eli Lilly and Company) for the treatment of advanced, recurrent ovarian cancer.

DATA SOURCES

Electronic databases (MEDLINE(®), EMBASE, Cochrane Central Register of Controlled Trials, Health Technology Assessment database, NHS Economic Evaluations Database) and trial registries were searched, and company submissions were reviewed. Databases were searched from inception to May 2013.

METHODS

A systematic review of the clinical and economic literature was carried out following standard methodological principles. Double-blind, randomised, placebo-controlled trials, evaluating topotecan, PLDH, paclitaxel, trabectedin and gemcitabine, and economic evaluations were included. A network meta-analysis (NMA) was carried out. A de novo economic model was developed.

RESULTS

For most outcomes measuring clinical response, two networks were constructed: one evaluating platinum-based regimens and one evaluating non-platinum-based regimens. In people with platinum-sensitive disease, NMA found statistically significant benefits for PLDH plus platinum, and paclitaxel plus platinum for overall survival (OS) compared with platinum monotherapy. PLDH plus platinum significantly prolonged progression-free survival (PFS) compared with paclitaxel plus platinum. Of the non-platinum-based treatments, PLDH monotherapy and trabectedin plus PLDH were found to significantly increase OS, but not PFS, compared with topotecan monotherapy. In people with platinum-resistant/-refractory (PRR) disease, NMA found no statistically significant differences for any treatment compared with alternative regimens in OS and PFS. Economic modelling indicated that, for people with platinum-sensitive disease and receiving platinum-based therapy, the estimated probabilistic incremental cost-effectiveness ratio [ICER; incremental cost per additional quality-adjusted life-year (QALY)] for paclitaxel plus platinum compared with platinum was £24,539. Gemcitabine plus carboplatin was extendedly dominated, and PLDH plus platinum was strictly dominated. For people with platinum-sensitive disease and receiving non-platinum-based therapy, the probabilistic ICERs associated with PLDH compared with paclitaxel, and trabectedin plus PLDH compared with PLDH, were estimated to be £25,931 and £81,353, respectively. Topotecan was strictly dominated. For people with PRR disease, the probabilistic ICER associated with topotecan compared with PLDH was estimated to be £324,188. Paclitaxel was strictly dominated.

LIMITATIONS

As platinum- and non-platinum-based treatments were evaluated separately, the comparative clinical effectiveness and cost-effectiveness of these regimens is uncertain in patients with platinum-sensitive disease.

CONCLUSIONS

For platinum-sensitive disease, it was not possible to compare the clinical effectiveness and cost-effectiveness of platinum-based therapies with non-platinum-based therapies. For people with platinum-sensitive disease and treated with platinum-based therapies, paclitaxel plus platinum could be considered cost-effective compared with platinum at a threshold of £30,000 per additional QALY. For people with platinum-sensitive disease and treated with non-platinum-based therapies, it is unclear whether PLDH would be considered cost-effective compared with paclitaxel at a threshold of £30,000 per additional QALY; trabectedin plus PLDH is unlikely to be considered cost-effective compared with PLDH. For patients with PRR disease, it is unlikely that topotecan would be considered cost-effective compared with PLDH. Randomised controlled trials comparing platinum with non-platinum-based treatments might help to verify the comparative effectiveness of these regimens.

STUDY REGISTRATION

This study is registered as PROSPERO CRD42013003555.

FUNDING

The National Institute for Health Research Health Technology Assessment programme.

Assessing key assumptions of network meta-analysis: a review of methods

BACKGROUND

Homogeneity and consistency assumptions underlie network meta-analysis (NMA). Methods exist to assess the assumptions but they are rarely and poorly applied. We review and illustrate methods to assess homogeneity and consistency.

METHODS

Eligible articles focussed on indirect comparison or NMA methodology. Articles were sought by hand-searching and scanning references (March 2013). Assumption assessment methods described in the articles were reviewed, and applied to compare anti-malarial drugs.

RESULTS

116 articles were included. Methods to assess homogeneity were: comparing characteristics across trials; comparing trial-specific treatment effects; using hypothesis tests or statistical measures; applying fixed-effect and random-effects pair-wise meta-analysis; and investigating treatment effect-modifiers. Methods to assess consistency were: comparing characteristics; investigating treatment effect-modifiers; comparing outcome measurements in the referent group; node-splitting; inconsistency modelling; hypothesis tests; back transformation; multidimensional scaling; a two-stage approach; and a graph-theoretical method. For the malaria example, heterogeneity existed for some comparisons that was unexplained by investigating treatment effect-modifiers. Inconsistency was detected using node-splitting and inconsistency modelling. It was unclear whether the covariates explained the inconsistency.

CONCLUSIONS

Presently, we advocate applying existing assessment methods collectively to gain the best understanding possible regarding whether assumptions are reasonable. In our example, consistency was questionable; therefore the NMA results may be unreliable.

Modeling and simulation as a tool to bridge efficacy and safety data in special populations

The registration and approval of novel medicines have traditionally been based on evidence arising from large prospective trials. Such an approach is often not possible or unsuitable to evaluate the benefit-risk balance in special populations (e.g., children, ethnic groups, rare diseases). Inferences by modeling and simulation can play a major role in evidence synthesis. A framework is proposed that promotes its acceptability and the basis for decision making during development, registration, and therapeutic use of drugs.

Multiple treatment comparisons in a series of anti-malarial trials with an ordinal primary outcome and repeated treatment evaluations

BACKGROUND

Artemisinin-based combination therapies (ACT) are widely used in African countries, including Cameroon. Between 2005 and 2007, five randomized studies comparing different treatment arms among artesunate-amodiaquine and other ACT were conducted in Cameroonian children aged two to 60 months who had uncomplicated Plasmodium falciparum malaria. In these studies, the categorical criterion proposed by the World Health Organization (WHO) to assess the relative effectiveness of anti-malarial drugs was repeatedly evaluated on Days 14, 21 and 28 after treatment initiation. The aim of the present study was to compare the effects of different treatments on this repeated ordinal outcome, hence using the fully available information.

METHODS

The quantitative synthesis was based on individual patient data. Due to the incomplete block design concerning treatment arms between different trials, a mixed treatment comparison (MTC) meta-analysis approach was adopted. The repeated ordinal outcome was modelled through a latent variable, as a proportional odds mixed model with trial, period and treatment arms as covariates. The model was further complexified to account for the variance heterogeneity, and the individual log-residual variance was modelled as a linear mixed model, as well. The effects of individual covariates at inclusion, such as parasitaemia, fever, gender and weight, were also tested. Model parameters were estimated using a Bayesian approach via the WinBUGS software. After selecting the best model using Deviance Information Criterion (DIC), mixed treatment comparisons were based on the estimated treatment effects.

RESULTS

Modeling the residual variance improved the model ability to adjust the data. The results showed that, compared to artesunate-amodiaquine (ASAQ), dihydroartemisinin-piperaquine (DHPP) was significantly more efficacious. Artesunate-chlorproguanil-dapsone (ASCD) was less efficacious than artesunate-sulphadoxine-pyrimethamine (ASSP), artemether-lumefantrine (AMLM) and DHPP, the difference with the latter being significant. No difference in efficacy was found between ASAQ and AMLM.

CONCLUSIONS

Bayesian mixed treatment comparisons of a network of connected randomized trials with repeated measurements of the primary categorical outcome allowed to take into account both the individual- and between- studies sources of heterogeneity. The results of the present study complete the previous quantitative review based on a binary outcome at a fixed time point, suggesting that DHPP represents an alternative for the treatment of uncomplicated P. falciparum malaria in Cameroonian children.

Targeted therapy for patients with chronic myeloid leukemia: clinical trial experience and challenges in inter-trial comparisons

The treatment of chronic myeloid leukemia (CML) was revolutionized by the introduction of the targeted tyrosine kinase inhibitor (TKI) imatinib mesylate. Later, to improve efficacy and tolerability, the more potent TKIs dasatinib and nilotinib were evaluated in CML. Clinicians comparing the clinical efficacy of TKIs face considerable challenges, including the variable treatment histories of patients receiving second-line therapy. The aim of this review is to highlight the pitfalls and possible solutions for comparing efficacy across disparate CML trials. Comparison of efficacy across trials is aided by careful consideration of possible confounding factors, including treatment history, definitions of imatinib intolerance or resistance, and BCR-ABL mutational status at baseline. However, methods exist to improve the comparability of data from different trials, yielding a more clinically and statistically meaningful inter-trial comparison.

Health technology assessment: Research trends and future priorities in Europe

Objective:

To provide an overview of health services research related to health technology assessment (HTA) and to identify research priorities from a European perspective.

Methods:

Several methods were used: systematic review of articles indexed with the MeSH term ‘technology assessment’ in PubMed from February 1999-2009; online survey among experts; and conference workshop discussions.

Results:

Research activity in HTA varies considerably across Europe. The research was categorised into six areas: (1) the breadth of analysis in HTA (such as economic, organizational and social aspects); (2) HTA products developed to meet the needs of policy-makers (such as horizon scanning, mini-HTA, and core HTA); (3) handling life-cycle perspectives in relation to technologies; (4) topics that challenge existing methods and for which HTA should be developed to address the themes more comprehensively (such as public health interventions and organizational interventions); (5) development of HTA capacity and programmes; and (6) links between policy and HTA. An online survey showed that the three areas that were given priority were the relationship between HTA and policy-making (71%), the impact of HTA (62%) and incorporating patient aspects in HTA (50%). Policy-makers highlighted HTA and innovation processes as their main research priority (42%). Areas that the systematic review identified as future priorities include issues within the six existing research areas such as disinvestment, developing evidence for new technologies, assessing the wider effects of technology use, and determining how HTA affects decision-making. In addition, relative effectiveness and individualized treatments are areas of growing interest.

Conclusions:

The research priorities identified are important for obtaining high quality and cost-effective health care in Europe. Managing the introduction, use and phasing out of technologies challenges health services throughout Europe, and these processes need to be improved to successfully manage future more general challenges. An ageing population and a diminishing workforce both require strong efforts to ensure effective and well-organized use of human resources and technologies. Furthermore, Europe needs to focus on innovation. This is closely linked to use of technologies and calls for future research.

Multiple treatment comparison meta-analyses: a step forward into complexity

The use of meta-analysis has become increasingly useful for clinical and policy decision making. A recent development in meta-analysis, multiple treatment comparison (MTC) meta-analysis, provides inferences on the comparative effectiveness of interventions that may have never been directly evaluated in clinical trials. This new approach may be confusing for clinicians and methodologists and raises specific challenges relevant to certain areas of medicine. This article addresses the methodological concepts of MTC meta-analysis, including issues of heterogeneity, choice of model, and adequacy of sample sizes. We address domain-specific challenges relevant to disciplines of medicine, including baseline risks of patient populations. We conclude that MTC meta-analysis is a useful tool in the context of comparative effectiveness and requires further study, as its utility and transparency will likely predict its uptake by the research and clinical community.

Economic evaluation of sorafenib in unresectable hepatocellular carcinoma

BACKGROUND AND AIM

A double-blind, randomized phase III trial of sorafenib in advanced hepatocellular carcinoma demonstrated that sorafenib significantly prolonged overall survival compared to placebo (median overall survival = 10.7 months vs 7.9 months, P < 0.001). Sorafenib is the first and only systemic agent demonstrating survival benefit in these patients. The aim of this study was to assess the cost-effectiveness of sorafenib versus best supportive care in the treatment of advanced hepatocellular carcinoma in the USA.

METHODS

A Markov model was developed following time-to-progression and survival using phase III trial data. Health effects are expressed as life-years gained. Resource utilization included drugs, physician visits, laboratory tests, scans, and hospitalizations. Unit costs, expressed in 2007 $US, came from diagnosis-related groupings, fee schedules, and the Red Book. Costs and effects were evaluated over a patient's lifetime and discounted at 3%.

RESULTS

Results are presented as incremental cost/life-year gained. Deterministic and probabilistic sensitivity analyses were conducted. Life-years gained were increased for sorafenib compared to best supportive care (mean ± standard deviation: 1.58 ± 0.17 vs 1.05 ± 0.10 life-years gained/sorafenib patient and best supportive care, respectively). Lifetime total costs were $US40,639 ± $US3052 for sorafenib and $US7, 804 ± $US1349 for best supportive care. The incremental cost-effectiveness ratio was $US62,473/life-year gained.

CONCLUSIONS

  The economic evaluation indicates that sorafenib is cost-effective compared to best supportive care, with a cost-effectiveness ratio within the established threshold that US society is willing to pay (i.e. $US50,000-$US100,000) and significantly lower than alternative thresholds suggested in recent years ($US183,000-$US264,000/life-year gained, or $US300,000/quality-adjusted life-year) in oncology.

Comparative efficacy and acceptability of pharmacotherapeutic agents for anxiety disorders in children and adolescents: a mixed treatment comparison meta-analysis

OBJECTIVE

to compare efficacy and acceptability of different pharmacotherapeutic agents for treating anxiety disorders in children and adolescents.

METHODS

A recently conducted Cochrane Review on pharmacotherapy for anxiety disorders in children and adolescents was updated. A mixed treatment comparison meta-analysis using Bayesian Markov Chain Monte Carlo simulation was used to perform the indirect comparison. We calculated relative risk ratios (RR) with 95% credible interval (CrI) using placebo as the common comparator.

RESULTS

Data were combined from 16 clinical trials that randomized children to six different treatment strategies, including placebo. Fluoxetine, fluvoxamine, paroxetine, sertraline, and venlafaxine were more efficacious than placebo. Venlafaxine was significantly less efficacious than fluvoxamine (RR = 0.60; 95% CrI 0.35-0.95) and paroxetine (RR = 0.65; 95% CrI 0.44-0.93). Fluoxetine, fluvoxamine, paroxetine, and sertraline had higher acceptability profile than placebo. Venlafaxine was less tolerated than fluvoxamine (RR = 0.16; 95% CrI 0.01-0.64), paroxetine (RR = 0.21; 95% CrI 0.05-0.59), and sertraline (RR = 0.31; 95% CrI 0.08-0.83). Fluvoxamine had a higher rate of clinical response and acceptability compared to other treatments in the network, with probability of 47.5% and 50.6% of being the most efficacious and well-tolerated treatment, respectively.

CONCLUSION

Clinically important differences exist between commonly prescribed pharmacotherapeutic agents for treating anxiety among children in terms of both efficacy and acceptability in favor of fluvoxamine. Fluvoxamine might be the best choice when starting treatment for anxiety disorders among children and adolescents because it has the most favorable balance between benefits and acceptability.

Multiparameter evidence synthesis in epidemiology and medical decision-making

Meta-analysis has been well-established for many years, but has been largely confined to pooling evidence on pair-wise contrasts. Broader forms of synthesis have also been described, apparently re-invented in disparate fields, each time taking different computational approaches. The potential value of Bayesian estimation of a joint posterior parameter distribution and simultaneously sampling from it for decision analysis has also been appreciated. However, applications have been relatively few in number, sometimes stylized, and presented mainly to a statistical methods audience. As a result, the potential for multiparameter evidence synthesis in both epidemiology and health technology assessment has remained largely unrecognized. The advent of flexible software for Bayesian Markov chain Monte Carlo in the shape of WinBUGS has the made these earlier strands of work more widely available. Researchers can now carry out synthesis at a realistic level of complexity. The Bristol programme has not only contributed to a growing body of literature on how to synthesize different evidence structures, but also on how to check the consistency of multiple information sources and how to use the resulting models to prioritize future research.

Economic evaluation of sorafenib in the treatment of hepatocellular carcinoma in Canada

BACKGROUND

A randomized phase III trial of sorafenib vs. placebo in hepatocellular carcinoma (HCC) demonstrated that sorafenib significantly prolonged overall survival (OS) compared to placebo.

RESEARCH DESIGN AND METHODS

A Markov model was developed to evaluate the cost-effectiveness of sorafenib vs. best supportive care (BSC) in HCC from the perspective of the Canadian provincial Ministry of Health. The model followed survival and time to progression (TTP) in monthly cycles based on the extrapolation of patient level trial data. Health effects were expressed as life-years gained (LYG). Resource use included drugs, physician visits, laboratory tests, scans, and hospitalizations. Unit costs were gathered from public sources and were expressed in 2007 Canadian Dollars. Costs and effects were evaluated over a lifetime and discounted at 5%. Results were presented as mean +/- standard deviation. Deterministic and probabilistic sensitivity analyses were conducted.

RESULTS

LYG was longer for sorafenib (1.52 +/- 0.16 vs. 1.03 +/- 0.09 LYG/patient for sorafenib and BSC, respectively). The lifetime total costs were $47,511 +/- 3 656 for sorafenib and $10,376 +/- 1 649 for BSC, resulting in an incremental cost-effectiveness ratio (ICER) of $75,821/LYG, and deterministic ICER of $75,759/LYG. The results were most sensitive to OS, TTP and BSC costs after progression. Sensitivity analyses results showed that the model was robust.

CONCLUSIONS

The economic evaluation indicates that sorafenib is cost-effective as compared to BSC in HCC. Limitations include multiple data sources, use of expert opinion for resource use, and the lack of utility data.

Evaluation of networks of randomized trials

Randomized trials may be designed and interpreted as single experiments or they may be seen in the context of other similar or relevant evidence. The amount and complexity of available randomized evidence vary for different topics. Systematic reviews may be useful in identifying gaps in the existing randomized evidence, pointing to discrepancies between trials, and planning future trials. A new, promising, but also very much debated extension of systematic reviews, mixed treatment comparison (MTC) meta-analysis, has become increasingly popular recently. MTC meta-analysis may have value in interpreting the available randomized evidence from networks of trials and can rank many different treatments, going beyond focusing on simple pairwise-comparisons. Nevertheless, the evaluation of networks also presents special challenges and caveats. In this article, we review the statistical methodology for MTC meta-analysis. We discuss the concept of inconsistency and methods that have been proposed to evaluate it as well as the methodological gaps that remain. We introduce the concepts of network geometry and asymmetry, and propose metrics for the evaluation of the asymmetry. Finally, we discuss the implications of inconsistency, network geometry and asymmetry in informing the planning of future trials.