Meta-regression models to address heterogeneity and inconsistency in network meta-analysis of survival outcomes

Bayesian pairwise meta-analysis of time-to-event outcomes in the presence of non-proportional hazards: A simulation study of flexible parametric, piecewise exponential and fractional polynomial models

BACKGROUND

Traditionally, meta-analysis of time-to-event outcomes reports a single pooled hazard ratio assuming proportional hazards (PH). For health technology assessment evaluations, hazard ratios are frequently extrapolated across a lifetime horizon. However, when treatment effects vary over time, an assumption of PH is not always valid. The Royston-Parmar (RP), piecewise exponential (PE), and fractional polynomial (FP) models can accommodate non-PH and provide plausible extrapolations of survival curves beyond observed data.

METHODS

Simulation study to assess and compare the performance of RP, PE, and FP models in a Bayesian framework estimating restricted mean survival time difference (RMSTD) at 50 years from a pairwise meta-analysis with evidence of non-PH. Individual patient data were generated from a mixture Weibull distribution. Twelve scenarios were considered varying the amount of follow-up data, number of trials in a meta-analysis, non-PH interaction coefficient, and prior distributions. Performance was assessed through bias and mean squared error. Models were applied to a metastatic breast cancer example.

RESULTS

FP models performed best when the non-PH interaction coefficient was 0.2. RP models performed best in scenarios with complete follow-up data. PE models performed well on average across all scenarios. In the metastatic breast cancer example, RMSTD at 50-years ranged from -14.6 to 8.48 months.

CONCLUSIONS

Synthesis of time-to-event outcomes and estimation of RMSTD in the presence of non-PH can be challenging and computationally intensive. Different approaches make different assumptions regarding extrapolation and sensitivity analyses varying key assumptions are essential to check the robustness of conclusions to different assumptions for the underlying survival function.

Challenges and Opportunities in Interdisciplinary Research and Real-World Data for Treatment Sequences in Health Technology Assessments

With an ever-increasing number of treatment options, the assessment of treatment sequences has become crucial in health technology assessment (HTA). This review systematically explores the multifaceted challenges inherent in evaluating sequences, delving into their interplay and nuances that go beyond economic model structures. We synthesised a 'roadmap' of literature from key methodological studies, highlighting the evolution of recent advances and emerging research themes. These insights were compared against HTA guidelines to identify potential avenues for future research. Our findings reveal a spectrum of challenges in sequence evaluation, encompassing selecting appropriate decision-analytic modelling approaches and comparators, deriving appropriate clinical effectiveness evidence in the face of data scarcity, scrutinising effectiveness assumptions and statistical adjustments, considering treatment displacement, and optimising model computations. Integrating methodologies from diverse disciplines-statistics, epidemiology, causal inference, operational research and computer science-has demonstrated promise in addressing these challenges. An updated review of application studies is warranted to provide detailed insights into the extent and manner in which these methodologies have been implemented. Data scarcity on the effectiveness of treatment sequences emerged as a dominant concern, especially because treatment sequences are rarely compared in clinical trials. Real-world data (RWD) provide an alternative means for capturing evidence on effectiveness and future research should prioritise harnessing causal inference methods, particularly Target Trial Emulation, to evaluate treatment sequence effectiveness using RWD. This approach is also adaptable for analysing trials harbouring sequencing information and adjusting indirect comparisons when collating evidence from heterogeneous sources. Such investigative efforts could lend support to reviews of HTA recommendations and contribute to synthesising external control arms involving treatment sequences.

Implementing Multilevel Network Meta-Regression for Time-To-Event Outcomes: A Case Study in Relapsed Refractory Multiple Myeloma

OBJECTIVES

Multilevel network meta-regression (ML-NMR) leverages individual patient data (IPD) and aggregate data from a network of randomized controlled trials (RCTs) to assess the comparative efficacy of multiple treatments, while adjusting for between-study differences. We provide an overview of ML-NMR for time-to-event outcomes and apply it to an illustrative case study, including example R code.

METHODS

The case study evaluated the comparative efficacy of idecabtagene vicleucel (ide-cel), selinexor+dexamethasone (Sd), belantamab mafodotin (BM), and conventional care (CC) for patients with triple-class exposed relapsed/refractory multiple myeloma in terms of overall survival. Single-arm clinical trials and real-world data were naively combined to create an aggregate data artificial RCT (aRCT) (MAMMOTH-CC versus DREAMM-2-BM versus STORM-2-Sd) and an IPD aRCT (KarMMa-ide-cel versus KarMMa-RW-CC). With some assumptions, we incorporated continuous covariates with skewed distributions, reported as median and range. The ML-NMR models adjusted for number of prior lines, triple-class refractory status, and age and were compared using the leave-one-out information criterion. We summarized predicted hazard ratios and survival (95% credible intervals) in the IPD aRCT population.

RESULTS

The Weibull ML-NMR model had the lowest leave-one-out information criterion. Ide-cel was more efficacious than Sd, BM, and CC in terms of overall survival. Effect modifiers had minimal impact on the model, and only triple-class refractory was a prognostic factor.

CONCLUSIONS

We demonstrate an application of ML-NMR for time-to-event outcomes and introduce code that can be used to aid implementation. Given its benefits, we encourage practitioners to utilize ML-NMR when population adjustment is necessary for comparisons of multiple treatments.

Diagnostic accuracy of tests for tuberculous pericarditis: A network meta-analysis

Tuberculous pericarditis (TBP) is a relatively uncommon but potentially fatal extrapulmonary manifestation of tuberculosis. Despite its severity, there is no universally accepted gold standard diagnostic test for TBP currently. The objective of this study is to compare the diagnostic accuracy of the most commonly used tests in terms of specificity, sensitivity, negative predictive value (NPV), and positive predictive value (PPV), and provide a summary of their diagnostic accuracies. A comprehensive literature review was performed using Scopus, MEDLINE, and Cochrane central register of controlled trials, encompassing studies published from start to April 2022. Studies that compared Interferon Gamma Release Assay (IGRA), Xpert MTB/RIF, Adenosine Deaminase levels (ADA), and Smear Microscopy (SM) were included in the analysis. Bayesian random-effects model was used for statistical analysis and mean and standard deviation (SD) with 95% confidence intervals were calculated using the absolute risk (AR) and odds ratio (OR). Rank probability and heterogeneity were determined using risk difference and Cochran Q test, respectively. Sensitivity and specificity were evaluated using true negative, true positive, false positive, and false negative rates. Area under the receiver operating characteristic (AUROC) was calculated for mean and standard error. A total of seven studies comprising 16 arms and 618 patients were included in the analysis. IGRA exhibited the highest mean (SD) sensitivity of 0.934 (0.049), with a high rank probability of 87.5% for being the best diagnostic test, and the AUROC was found to be 94.8 (0.36). On the other hand, SM demonstrated the highest mean (SD) specificity of 0.999 (0.011), with a rank probability of 99.5%, but a leave-one-out analysis excluding SM studies revealed that Xpert MTB/RIF ranked highest for specificity, with a mean (SD) of 0.962 (0.064). The diagnostic tests compared in our study exhibited similar high NPV, while ADA was found to have the lowest PPV among the evaluated methods. Further research, including comparative studies, should be conducted using a standardized cutoff value for both ADA levels and IGRA to mitigate the risk of threshold effect and minimize bias and heterogeneity in data analysis.

Effectiveness of Interventions Aimed at Changing Movement Patterns in People With Patellofemoral Pain: A Systematic Review With Network Meta-analysis

OBJECTIVE: To summarize the effectiveness of interventions for changing movement during weight-bearing functional tasks in people with patellofemoral pain (PFP). DESIGN: Systematic review with network meta-analysis (NMA). LITERATURE SEARCH: Medline, Embase, CINAHL, SPORTDiscus, and Cochrane Library were searched from inception up to May 2023. STUDY SELECTION CRITERIA: Randomized controlled trials involving people with PFP and nonsurgical, nonpharmacological interventions on task kinematics were included. DATA SYNTHESIS: NMA was conducted for frontal knee movement data, and pairwise meta-analysis was used to pool data when NMA was not possible. Reduced movements were those changes that indicated movements occurring with less amplitude. The GRADE approach was used to grade the certainty of the evidence. RESULTS: Thirty-seven trials were included (n = 1235 participants). Combining knee/hip exercises with internal feedback had the strongest effect on reducing frontal knee movements (standardized mean difference [SMD] from NMA = -2.66; GRADE: moderate evidence). On pairwise comparisons, the same combination of interventions reduced frontal hip movements (SMD = -0.47; GRADE: moderate evidence) and increased sagittal knee movements (SMD = 1.03; GRADE: moderate evidence), with no effects on sagittal hip movements (GRADE: very low evidence), compared to knee/hip exercises alone. There was no effect for single applications of braces on the frontal knee movement (GRADE: very low evidence) and taping on movements of the knee, hip, and ankle (GRADE: very low to low evidence) compared to no intervention. CONCLUSION: Knee/hip exercises combined with internal feedback techniques may change knee and hip movements in people with PFP. The combination of these interventions can reduce frontal knee and hip movements, and can increase sagittal knee movements. J Orthop Sports Phys Ther 2023;53(12):1-13. Epub 14 September 2023. doi:10.2519/jospt.2023.11956.

Multi-state network meta-analysis of progression and survival data

Multiple randomized controlled trials, each comparing a subset of competing interventions, can be synthesized by means of a network meta-analysis to estimate relative treatment effects between all interventions in the evidence base. Here we focus on estimating relative treatment effects for time-to-event outcomes. Cancer treatment effectiveness is frequently quantified by analyzing overall survival (OS) and progression-free survival (PFS). We introduce a method for the joint network meta-analysis of PFS and OS that is based on a time-inhomogeneous tri-state (stable, progression, and death) Markov model where time-varying transition rates and relative treatment effects are modeled with parametric survival functions or fractional polynomials. The data needed to run these analyses can be extracted directly from published survival curves. We demonstrate use by applying the methodology to a network of trials for the treatment of non-small-cell lung cancer. The proposed approach allows the joint synthesis of OS and PFS, relaxes the proportional hazards assumption, extends to a network of more than two treatments, and simplifies the parameterization of decision and cost-effectiveness analyses.

Neoadjuvant immunotherapy and chemotherapy regimens for the treatment of high-risk, early-stage triple-negative breast cancer: a systematic review and network meta-analysis

BACKGROUND

Patients with triple-negative breast cancer (TNBC) are generally younger and more likely to experience disease recurrence and have the shortest survival among all breast cancer patients. Recently, neoadjuvant delivery of the programmed cell death protein-1 inhibitor pembrolizumab plus chemotherapy followed by adjuvant pembrolizumab was approved for patients with high-risk, early-stage TNBC, but this treatment regimen has not been evaluated in head-to-head trials with other neoadjuvant treatment regimens. Therefore, the objective of this study was to estimate the relative efficacy of neoadjuvant pembrolizumab + chemotherapy followed by adjuvant pembrolizumab versus other neoadjuvant treatments for early-stage TNBC through a systematic review and network meta-analysis (NMA).

METHODS

EMBASE, MEDLINE, Cochrane Central Register of Controlled Trials, conference abstracts, and clinical trial registries were searched for randomized controlled trials evaluating neoadjuvant treatments for early-stage TNBC. NMA was performed to estimate relative treatment effects among evaluated interventions.

RESULTS

Five trials met the inclusion criteria and were included in the NMA. The relative efficacy of neoadjuvant pembrolizumab + chemotherapy followed by adjuvant pembrolizumab was favorable to paclitaxel followed by anthracycline + cyclophosphamide in terms of pathologic complete response (pCR), event-free survival (EFS), and overall survival; paclitaxel + carboplatin followed by anthracycline + cyclophosphamide in terms of pCR and EFS; paclitaxel + bevacizumab followed by anthracycline + cyclophosphamide + bevacizumab in terms of pCR; and paclitaxel + carboplatin + veliparib followed by anthracycline + cyclophosphamide in terms of EFS.

CONCLUSIONS

Neoadjuvant pembrolizumab + chemotherapy followed by adjuvant pembrolizumab confers benefits in response and survival outcomes versus alternative neoadjuvant treatments for early-stage TNBC.

Clinical outcomes of different revascularization approaches for patients with multi-vessel coronary artery disease: A network meta-analysis

BACKGROUND

As surgical techniques continue to evolve, the optimal approach for revascularizing multi-vessel coronary artery disease (CAD) remains a matter of ongoing debate. Accordingly, our objective was to compare and contrast various surgical techniques utilized in the management of multi-vessel CAD.

METHODS

A systematic literature review was performed using PubMed, Embase, and Cochrane central register of controlled trials from inception to May 2022. Random-effects network meta-analysis was performed for the primary outcome; target vessel revascularization (TVR), and secondary outcomes; mortality, major adverse cardiac and cerebrovascular events, postoperative myocardial infarction, new-onset atrial fibrillation, stroke, new-onset dialysis, in patients undergoing percutaneous coronary intervention (PCI) with a stent, off-pump coronary bypass graft, on-pump coronary artery bypass graft (ONCABG), hybrid coronary revascularization, minimally-invasive coronary artery bypass, or robot-assisted coronary artery bypass (RCAB) surgeries.

RESULTS

A total of 8841 patients were included from 23 studies. The analysis showed that ONCABG had the highest freedom from TVR, with a mean (SD) absolute risk of 0.027 (0.029); although ONCABG was found to be superior to all other methods, it was only significantly better than first-generation stent PCI. While RCAB did not demonstrate significant superiority over other treatments, it showed a greater probability of preventing postoperative complications. Notably, no significant heterogeneity was calculated for any of the reported outcomes.

CONCLUSIONS

ONCABG shows a better rank probability compared to all other techniques for preventing TVR, while RCAB offers greater freedom from most postoperative complications. However, given the absence of randomized controlled trials, these results should be interpreted with caution.

Indirect treatment comparison of lurbinectedin versus other second-line treatments for small-cell lung cancer

Aim: Compare lurbinectedin versus other second-line (2L) small-cell lung cancer (SCLC) treatments. Methods: An unanchored matching-adjusted indirect comparison connected the platinum-sensitive SCLC cohort of a single-arm lurbinectedin trial to a network of three randomized controlled trials (oral and intravenous [IV] topotecan, and platinum re-challenge) identified by systematic literature review. Network meta-analysis methods estimated relative treatment effects. Results: In platinum-sensitive patients, lurbinectedin demonstrated a survival benefit and favorable safety profile versus oral and IV topotecan and platinum re-challenge (overall survival, hazard ratio [HR]: 0.43; 95% credible interval [CrI]: 0.27, 0.67; HR: 0.43; 95% CrI: 0.26, 0.70; HR: 0.42; 95% CrI: 0.30, 0.58 respectively). Conclusion: Lurbinectedin showed a robust survival benefit and favorable safety versus other SCLC treatments in 2L platinum-sensitive SCLC.

Non-opioid analgesics for the prevention of chronic postsurgical pain: a systematic review and network meta-analysis

BACKGROUND

Chronic postsurgical pain is common after surgery. Identification of non-opioid analgesics with potential for preventing chronic postsurgical pain is important, although trials are often underpowered. Network meta-analysis offers an opportunity to improve power and to identify the most promising therapy for clinical use and future studies.

METHODS

We conducted a PRISMA-NMA-compliant systematic review and network meta-analysis of randomised controlled trials of non-opioid analgesics for chronic postsurgical pain. Outcomes included incidence and severity of chronic postsurgical pain, serious adverse events, and chronic opioid use.

RESULTS

We included 132 randomised controlled trials with 23 902 participants. In order of efficacy, i.v. lidocaine (odds ratio [OR] 0.32; 95% credible interval [CrI] 0.17-0.58), ketamine (OR 0.64; 95% CrI 0.44-0.92), gabapentinoids (OR 0.67; 95% CrI 0.47-0.92), and possibly dexmedetomidine (OR 0.36; 95% CrI 0.12-1.00) reduced the incidence of chronic postsurgical pain at ≤6 months. There was little available evidence for chronic postsurgical pain at >6 months, combinations agents, chronic opioid use, and serious adverse events. Variable baseline risk was identified as a potential violation to the network meta-analysis transitivity assumption, so results are reported from a fixed value of this, with analgesics more effective at higher baseline risk. The confidence in these findings was low because of problems with risk of bias and imprecision.

CONCLUSIONS

Lidocaine (most effective), ketamine, and gabapentinoids could be effective in reducing chronic postsurgical pain ≤6 months although confidence is low. Moreover, variable baseline risk might violate transitivity in network meta-analysis of analgesics; this recommends use of our methods in future network meta-analyses.

SYSTEMATIC REVIEW PROTOCOL

PROSPERO CRD42021269642.

A Comparison of Alternative Network Meta-Analysis Methods in the Presence of Nonproportional Hazards: A Case Study in First-Line Advanced or Metastatic Renal Cell Carcinoma

OBJECTIVES

Network meta-analysis (NMA) of time-to-event outcomes based on constant hazard ratios can result in biased findings when the proportional hazards (PHs) assumption does not hold in a subset of trials. We aimed to summarize the published non-PH NMA methods for time-to-event outcomes, demonstrate their application, and compare their results.

METHODS

The following non-PH NMA methods were compared through an illustrative case study in oncology of 4 randomized controlled trials in terms of progression-free survival and overall survival: (1) 1-step or (2) 2-step multivariate NMAs based on traditional survival distributions or fractional polynomials, (3) NMAs with restricted cubic splines for baseline hazard, and (4) restricted mean survival NMA.

RESULTS

For progression-free survival, the PH assumption did not hold across trials and non-PH NMA methods better reflected the relative treatment effects over time. The most flexible models (fractional polynomials and restricted cubic splines) fit better to the data than the other approaches. Estimated hazard ratios obtained with different non-PH NMA methods were similar at 5 years of follow-up but differed thereafter in the extrapolations. Although there was no strong evidence of PH violation for overall survival, non-PH NMA methods captured this uncertainty in the relative treatment effects over time.

CONCLUSIONS

When the PH assumption is questionable in a subset of the randomized controlled trials, we recommend assessing alternative non-PH NMA methods to estimate relative treatment effects for time-to-event outcomes. We propose a transparent and explicit stepwise model selection process considering model fit, external constraints, and clinical validity. Given inherent uncertainty, sensitivity analyses are suggested.

Heterogeneity in meta-analysis: a comprehensive overview

In recent years, meta-analysis has evolved to a critically important field of Statistics, and has significant applications in Medicine and Health Sciences. In this work we briefly present existing methodologies to conduct meta-analysis along with any discussion and recent developments accompanying them. Undoubtedly, studies brought together in a systematic review will differ in one way or another. This yields a considerable amount of variability, any kind of which may be termed heterogeneity. To this end, reports of meta-analyses commonly present a statistical test of heterogeneity when attempting to establish whether the included studies are indeed similar in terms of the reported output or not. We intend to provide an overview of the topic, discuss the potential sources of heterogeneity commonly met in the literature and provide useful guidelines on how to address this issue and to detect heterogeneity. Moreover, we review the recent developments in the Bayesian approach along with the various graphical tools and statistical software that are currently available to the analyst. In addition, we discuss sensitivity analysis issues and other approaches of understanding the causes of heterogeneity. Finally, we explore heterogeneity in meta-analysis for time to event data in a nutshell, pointing out its unique characteristics.

Clinical outcomes in patients relapsed/refractory after ≥2 prior lines of therapy for follicular lymphoma: a systematic literature review and meta-analysis

BACKGROUND

Patients with follicular lymphoma (FL) can have high response rates to early lines of treatment. However, among FL patients relapsed/refractory (r/r) after ≥2 prior lines of therapy (LOT), remission tends to be shorter and there is limited treatment guidance. This study sought to evaluate the clinical outcomes for r/r FL after ≥2 prior LOT identified through systematic literature review.

METHODS

Eligible studies included comparative or non-comparative interventional or observational studies of systemic therapies among adults with FL r/r after ≥2 prior LOT published prior to 31st May 2021. Prior LOT must have included an anti-CD20 monoclonal antibody and an alkylating agent, in combination or separately. Overall response rate (ORR) and complete response (CR) were estimated using inverse-variance weighting with Freeman-Tukey double-arcsine transformations. Kaplan-Meier (KM) curves for progression-free survival (PFS) and overall survival (OS) estimated by reconstructing digitized curves using the Guyot algorithm, and survival analyses were conducted, stratified by ≥2 prior LOT and ≥ 3 prior LOT groups (as defined in the source material). Restricting the analyses to the observational cohorts was investigated as a sensitivity analysis.

RESULTS

The analysis-set included 20 studies published between 2014 and 2021. Studies were primarily US and/or European based, with the few exceptions using treatments approved in US/Europe. The estimated ORR was 58.47% (95% confidence interval [CI]: 51.13-65.62) and proportion of patients with CR was 19.63% (95% CI: 15.02-24.68). The median OS among those ≥2 prior LOT was 56.57 months (95% CI: 47.8-68.78) and median PFS was 9.78 months (95% CI: 9.01-10.63). The 24-month OS decreased from 66.50% in the ≥2 prior LOT group to 59.51% in the ≥3 prior LOT group, with a similar trend in PFS at 24-month (28.42% vs 24.13%).

CONCLUSIONS

This study found that few r/r FL patients with ≥2 prior LOT achieve CR, and despite some benefit, approximately 1/3 of treated patients die within 24 months. The shorter median PFS with increasing prior LOT suggest treatment durability is suboptimal in later LOT. These findings indicate that patients are underserved by treatments currently available in the US and Europe.

Termination of seizures in the paediatric age group, best benzodiazepine and route of administration: A network meta-analysis

This network meta-analysis aims to compare various benzodiazepines and their route of administration using the data published exclusively in randomized controlled trials (RCTs). Two thousand two hundred sixty-three children presenting with an episode of seizure to ER or to a paramedic where they were administered a benzodiazepine as the first-line treatment were included. All the outcomes were measured for their mean with 95% CI and rank probability. The primary outcome was the number of successful seizure cessation. Secondary outcomes were the time interval between drug administration and seizure cessation, the time interval between patient arrival and seizure cessation and the number of episodes of seizure recurrence after drug administration. For the number of successful cessations, intramuscular midazolam showed the highest mean and best rank probability with a value of .881 (.065) and 57.9%, respectively. For the time of cessation, both intravenous lorazepam (IVL) and intravenous diazepam showed a mean of 3.30 (1.30) with IVL having the highest rank probability of 32%. For total time for cessation, intranasal midazolam showed the best mean and rank probability with a value of 4.3 (1.1) and 55%, respectively. Buccal midazolam showed the lowest mean with a value of .106 (.084) for rate of recurrence. Although there was no significant difference between the treatments, but based on the rank probability, IVL shows more promising results for patients who already have an established intravenous line, and for patients presenting in the ER without an intravenous line, the first line of treatment should be INM as it shows the highest rank probability in total time with second-highest successful cessation rate.

Challenges of modelling approaches for network meta-analysis of time-to-event outcomes in the presence of non-proportional hazards to aid decision making: Application to a melanoma network

BACKGROUND

Synthesis of clinical effectiveness from multiple trials is a well-established component of decision-making. Time-to-event outcomes are often synthesised using the Cox proportional hazards model assuming a constant hazard ratio over time. However, with an increasing proportion of trials reporting treatment effects where hazard ratios vary over time and with differing lengths of follow-up across trials, alternative synthesis methods are needed.

OBJECTIVES

To compare and contrast five modelling approaches for synthesis of time-to-event outcomes and provide guidance on key considerations for choosing between the modelling approaches.

METHODS

The Cox proportional hazards model and five other methods of estimating treatment effects from time-to-event outcomes, which relax the proportional hazards assumption, were applied to a network of melanoma trials reporting overall survival: restricted mean survival time, generalised gamma, piecewise exponential, fractional polynomial and Royston-Parmar models.

RESULTS

All models fitted the melanoma network acceptably well. However, there were important differences in extrapolations of the survival curve and interpretability of the modelling constraints demonstrating the potential for different conclusions from different modelling approaches.

CONCLUSION

The restricted mean survival time, generalised gamma, piecewise exponential, fractional polynomial and Royston-Parmar models can accommodate non-proportional hazards and differing lengths of trial follow-up within a network meta-analysis of time-to-event outcomes. We recommend that model choice is informed using available and relevant prior knowledge, model transparency, graphically comparing survival curves alongside observed data to aid consideration of the reliability of the survival estimates, and consideration of how the treatment effect estimates can be incorporated within a decision model.

Systematic literature review and network meta-analysis of pembrolizumab versus other interventions for previously untreated, unresectable or metastatic, microsatellite instability-high or mismatch repair-deficient colorectal cancer

Aim: To compare pembrolizumab with competing interventions for previously untreated, unresectable or metastatic microsatellite instability-high or mismatch repair-deficient colorectal cancer. Method: Trials were identified via a systematic literature review and synthesized using a Bayesian network meta-analysis with time-varying hazard ratios (HRs). Results: Using intention-to-treat data, HRs for overall survival were generally in favor of pembrolizumab but not statistically significant; however, statistical significance was reached versus all comparators by month 16 when accounting for crossover. Estimated HRs for progression-free survival significantly favored pembrolizumab versus all comparators by month 12. Pembrolizumab was also superior to all comparators in terms of grade ≥3 adverse events. Conclusion: These analyses suggest that pembrolizumab is a highly efficacious and safe treatment in this population.

Effects of antidepressant medicines on preventing relapse of unipolar depression: a pooled analysis of parametric survival curves

BACKGROUND

Major depressive disorder is characterized by a high risk of relapse. We aimed to compare the prophylactic effects of different antidepressant medicines (ADMs).

METHODS

PubMed, Cochrane Central Register of Controlled Trials, Embase and the Web of Science were searched on 4 July 2019. A pooled analysis of parametric survival curves was performed using a Bayesian framework. The main outcomes were hazard ratios (HRs), relapse-free survival and mean relapse-free months.

RESULTS

Forty randomized controlled trials were included. The 1-year relapse-free survival for ADM (76%) was significantly better than that for placebo (56%). Most of the relapse difference (86.5%) occurred in the first 6 months. Most HRs were not constant over time. Proof of benefit after 6 months of follow-up was not established partially because of small differences between the drug and placebo after 6 months. Almost all studies used an 'enriched' randomized discontinuation design, which may explain the high relapse rates in the first 6 months after randomization.

CONCLUSIONS

The superiority of ADM v. placebo was mainly attributed to the difference in relapse rates that occurred in the first 6 months. Our analysis provided evidence that the prophylactic efficacy was not constant over time. A beneficial effect was observed, but the prevention of new episodes after 6 months was questionable. These findings may have implications for clinical practice.

Adjuvant Therapy for Resectable Biliary Tract Cancer: A Bayesian Network Analysis

Background: Selecting proper postoperative adjuvant therapy is of great importance for prolonging overall survival (OS) of patients with biliary tract cancer (BTC). OS is commonly affected by high rate of postoperative recurrence and metastasis. Purpose: The present study aimed to identify the optimal adjuvant therapy for BTC patients. Method: A comprehensive search was carried out on Pubmed, Web of science, and Embase databases to acquire articles regarding BTC therapy approaches. Subsequently, the hazard ratio (HR) and its 95% confidence intervals (CIs) were applied to evaluate the efficacy of different adjuvant therapy regimens. The GemTc (GemTc.0.8-2) and R (R.3.6.0) software were employed to perform statistical analyses. Result: Data from 22 articles, including 14,646 patients, were quantitatively analyzed. The results showed that in terms of 5-year OS, gemcitabine (GEM) was considered as the optimal adjuvant therapy for BTC compared with chemoradiotherapy (CRT; HR = 0.59; 95% CI = 0.34-0.97), observation (OB; HR = 0.49; 95% CI = 0.33-0.73), and radiotherapy (RT; HR = 0.40; 95% CI = 0.22-0.71). Additionally, 5-fluorouracil (5-FU) exhibited improved efficacy compared with RT (HR = 0.52; 95% CI = 0.29-0.91) and OB (HR = 0.63; 95% CI = 0.43-0.92). When the efficacy of 5-FU was compared with that of GEM, the results showed that 5-FU (HR = 1.29) was more effective than GEM. Furthermore, CRT and RT prolonged positive resection margin (R⁺)-OS (HR = 0.69; 95% CI = 0.49-1.00) and positive lymph node-(N⁺)-OS (HR = 0.22; 95% CI = 0.074-0.66) in BTC patients. In terms of median recurrence-free survival (RFS) and 1-year OS, the differences were not statistically significant among different therapeutic interventions. Conclusion: The present study suggested that GEM could be used as a first-line adjuvant therapy for resected BTC patients. Additionally, CRT could be the optimal treatment approach for R⁺ and N⁺ patients.

Comparative efficacy of various preventive methods for exposure keratopathy among critically ill patients: A network meta-analysis of randomized controlled trials

BACKGROUND

There is a wide variety of preventive methods currently available for the treatment of exposure keratopathy. Because of a lack of evidence from head-to-head randomized controlled trials (RCTs), the relative effects of these preventive methods in exposure keratopathy patients remain unclear. The purpose of our study is to carry out a network meta-analysis comparing the efficacy of different methods for the prevention of exposure keratopathy and rank these nursing methods for practical consideration.

METHODS

A literature search was performed of the MEDLINE (PubMed), EMBASE, Web of Science, China National Knowledge Infrastructure Library (CNKI), China Science and Technology Journal Database (Weipu), WanFang Database and China Biology Medicine disc. Two authors independently extracted data from each included RCTs according to a predesigned Excel spreadsheet and assessed the methodological quality of included RCTs using the Cochrane risk of bias tool. Data was analyzed using the R (V.3.6.2) and the Stata (V.15.0).

RESULTS

21 RCTs involving 2022 patients and evaluating 11 preventive methods were included. Rankings based on posterior probabilities revealed that artificial tear ointment might be the best way to prevent exposure keratopathy (35%), polyethylene covers might be the second-best (31%), swimming goggles might be the third-best (21%), foam dressing might be the fourth-best (18%).

CONCLUSIONS

This network meta-analysis indicated that artificial tear ointment, polyethylene covers, swimming goggles and foam dressing might be selected for the prevention of exposure keratopathy in intensive care unit patients, which is important in future research. Although evidence is scant, more attention should be paid to head-to-head comparisons of the most commonly used prevention methods in this field.

Assessing the Value of Cemiplimab for Adults With Advanced Cutaneous Squamous Cell Carcinoma: A Cost-Effectiveness Analysis

OBJECTIVES

To evaluate the cost-effectiveness of cemiplimab in patients with advanced cutaneous squamous cell carcinoma (CSCC) from a payer perspective in the United States.

METHODS

A partitioned survival model was developed to assess the cost-effectiveness of cemiplimab versus historical standard of care (SOC). All inputs were identified based on a systematic literature review, supplemented by expert opinion where necessary. Clinical inputs for cemiplimab were based on individual patient data from a cemiplimab phase 2 single-arm trial (NCT27060498). For SOC, analysis was based on a pooled analysis of single-arm clinical trials and retrospective studies evaluating chemotherapy and epidermal growth factor receptor inhibitors (cetuximab, erlotinib, and gefitinib) identified via a systematic literature review (6 of the 27 included studies). Overall survival and progression-free survival were extrapolated over a lifetime horizon. Costs were included for drug acquisition, drug administration, management of adverse events, subsequent therapy, disease management, and terminal care. Unit costs were based on published 2019 US list prices.

RESULTS

In the base case, cemiplimab versus SOC resulted in an incremental cost-effectiveness ratio of $99 447 per quality adjusted-life year (QALY), where incremental costs and QALYs were $372 108 and 3.74, respectively. At a willingness-to-pay threshold of $150 000/QALY, the probabilistic sensitivity analysis suggests a 90% probability that cemiplimab is cost-effective compared to SOC. Scenario analyses resulted in incremental cost-effectiveness ratios ranging from $90 590 to $148 738.

CONCLUSIONS

Compared with historical SOC, cemiplimab is a cost-effective use of US payer resources for the treatment of advanced CSCC and is expected to provide value for money.

Perineural and intravenous dexamethasone and dexmedetomidine: network meta-analysis of adjunctive effects on supraclavicular brachial plexus block

Both perineural and intravenous dexamethasone and dexmedetomidine are used as local anaesthetic adjuncts to enhance peripheral nerve block characteristics. However, the effects of dexamethasone and dexmedetomidine based on their administration routes have not been directly compared, and the relative extent to which each adjunct prolongs sensory blockade remains unclear. This network meta-analysis sought to compare and rank the effects of perineural and intravenous dexamethasone and dexmedetomidine as supraclavicular block adjuncts. We sought randomised trials investigating the effects of adding perineural and intravenous dexamethasone or dexmedetomidine to long-acting local anaesthetics on supraclavicular block characteristics, including time to block onset and durations of sensory, motor and analgesic blockade. Data were compared and ranked according to relative effectiveness for each outcome. Our primary outcome was sensory block duration, with a 2-h difference considered clinically important. We performed a frequentist analysis, using the GRADE framework to appraise evidence. One-hundred trials (5728 patients) were included. Expressed as mean (95%CI), the control group (local anaesthetic alone) had a duration of sensory block of 401 (366-435) min, motor block duration of 369 (330-408) min and analgesic duration of 435 (386-483) min. Compared with control, sensory block was prolonged most by intravenous dexamethasone [mean difference (95%CI) 477 (160-795) min], followed by perineural dexamethasone [411 (343-480) min] and perineural dexmedetomidine [284 (235-333) min]. Motor block was prolonged most by perineural dexamethasone [mean difference (95%CI) 294 (236-352) min], followed by intravenous dexamethasone [289 (129-448)min] and perineural dexmedetomidine [258 (212-304)min]. Analgesic duration was prolonged most by perineural dexamethasone [mean difference (95%CI) 518 (448-589) min], followed by intravenous dexamethasone [478 (277-679) min] and perineural dexmedetomidine [318 (266-371) min]. Intravenous dexmedetomidine did not prolong sensory, motor or analgesic block durations. No major network inconsistencies were found. The quality of evidence for intravenous dexamethasone, perineural dexamethasone and perineural dexmedetomidine for prolongation of supraclavicular sensory block duration was 'low', 'very low' and 'low', respectively. Regardless of route, dexamethasone as an adjunct prolonged the durations of sensory and analgesic blockade to a greater extent than dexmedetomidine. Differences in block characteristics between perineural and intravenous dexamethasone were not clinically important. Intravenous dexmedetomidine did not affect block characteristics.

Comparative effectiveness of pulmonary arterial hypertension drugs in treatment-naive patients: a network meta-analysis

Aim: No network meta-analysis has been conducted to study efficacy of drug therapies specific for treatment of pulmonary arterial hypertension in treatment-naive patients only. Methods: Randomized controlled trials on pulmonary arterial hypertension-specific drug therapies were searched and a Bayesian network meta-analysis was performed. The 6-min walking distance (6MWD) and all-cause mortality were efficacy outcomes, whereas discontinuation due to adverse events was a safety-related outcome. Results: Analysis included 3.713 patients from 21 trials. Combination of ambrisentan and tadalafil showed the greatest impact on 6MWD, followed by epoprostenol and intravenous treprostinil (high dose). The latter two demonstrated marked effect size on mortality, although not statistically significant. Conclusion: According to 6MWD, ambrisentan/tadalafil combination was considered as most effective among all comparisons. Prospero ID: CRD42019110832.

A Bayesian Network Meta-analysis of Add-on Drug Therapies Specific for Pulmonary Arterial Hypertension

Background: Recently published meta-analyses did not discriminate between drug agents used for initial and sequential combination therapy. Objective: To assess the comparative efficacy of drugs specific for the treatment of pulmonary arterial hypertension (PAH) as add-on therapies based on 6-minute walk distance (6MWD), all-cause mortality, and discontinuation due to adverse events (AEs). Methods: EMBASE, PubMed, Cochrane Library, and ClinicalTrials.gov were searched until December 9, 2018, for the randomized, placebo-controlled clinical trials (RCTs) conducted on primarily adult patients diagnosed with PAH. Data extracted from applicable RCTs were as follows: for 6MWD mean change from baseline, the total number of patients, and the number of patients with events, per treatment. Network meta-analysis (NMA) was conducted in a Bayesian framework. Results: A total of 16 RCTs were eligible for analysis, with 4112 patients. Add-on therapy with tadalafil or inhaled treprostinil performed better than endothelin receptor antagonists alone [27 m; 95% credible interval (CrI): (11, 43); and 19 m; 95% CrI: (10, 27); respectively]. Add-on therapy with macitentan or bosentan performed better than phosphodiesterase type 5 inhibitors alone [26 m; 95% CrI: (6.4, 45); and 22 m; 95% CrI: (5.1, 38); respectively]. Differences in all-cause mortality and discontinuation due to AEs were nonsignificant. Conclusion and Relevance: Our NMA evaluated efficacy and safety of add-on therapies in patients with PAH. None of the previous meta-analyses evaluated RCTs focusing solely on patients pretreated with another PAH-specific drug therapy. Our results support guideline recommendations on combination therapy in PAH patients and add the quantitative perspective on which sequential therapy demonstrated the greatest effect size.

Relative efficacy of treatment options in transplant-ineligible newly diagnosed multiple myeloma: results from a systematic literature review and network meta-analysis

Established treatments for transplant-ineligible (TNE) patients with newly diagnosed multiple myeloma (NDMM) include melphalan and prednisone (MP) combined with either bortezomib (VMP) or thalidomide (MPT), or lenalidomide plus low-dose dexamethasone (Rd). New treatments for TNE NDMM include Rd plus bortezomib (RVd) and daratumumab plus VMP (VMP + D), daratumumab plus lenalidomide and dexamethasone (D + Rd). Relative efficacy of these treatments was compared using a network meta-analysis. Eight trials identified by a systematic literature review were included in the primary analysis; hazard ratios (HRs) for overall survival (OS) and progression-free survival (PFS) were used. Rd was superior to other MP-based regimens for OS and PFS. There was strong evidence that, compared with Rd, both D + Rd and RVd improved PFS (HR 0.57; 95% credible interval (CrI) 0.43, 0.73 and HR 0.72; 95% CrI 0.56, 0.91, respectively). However, there was strong evidence only for RVd in respect to OS (HR 0.72; 95% CrI 0.52, 0.96).

Comparison of ventilatory modes to facilitate liberation from mechanical ventilation: protocol for a systematic review and network meta-analysis

INTRODUCTION

Timely liberation from invasive mechanical ventilation is important to reduce the risk of ventilator-associated complications. Once a patient is deemed ready to tolerate a mode of partial ventilator assist, clinicians can use one of multiple ventilatory modes. Despite multiple trials, controversy regarding the optimal ventilator mode to facilitate liberation remains. Herein, we report the protocol for a systematic review and network meta-analysis comparing modes of ventilation to facilitate the liberation of a patient from invasive mechanical ventilation.

METHODS AND ANALYSIS

We will search MEDLINE, EMBASE, PubMed, the Cochrane Library from inception to April 2019 for randomised trials that report on critically ill adults who have undergone invasive mechanical ventilation for at least 24 hours and have received any mode of assisted invasive mechanical ventilation compared with an alternative mode of assisted ventilation. Outcomes of interest will include: mortality, weaning success, weaning duration, duration of mechanical ventilation, duration of stay in the acute care setting and adverse events. Two reviewers will independently screen in two stages, first titles and abstracts, and then full texts, to identify eligible studies. Independently and in duplicate, two investigators will extract all data, and assess risk of bias in all eligible studies using the Modified Cochrane Risk of Bias tool. Reviewers will resolve disagreement by discussion and consultation with a third reviewer as necessary. Using a frequentist framework, we will perform random-effect network meta-analysis, including all ventilator modes in the same model. We will calculate direct and indirect estimates of treatment effect using a node-splitting procedure and report effect estimates using OR and 95% CI. We will assess certainty in effect estimates using Grading of Recommendations Assessment, Development and Evaluation methodology.

ETHICS AND DISSEMINATION

Research ethics board approval is not necessary. The results will be disseminated through publication in a peer-reviewed journals.

PROSPERO REGISTRATION NUMBER

CRD42019137786.

Integrating expert opinion with clinical trial data to extrapolate long-term survival: a case study of CAR-T therapy for children and young adults with relapsed or refractory acute lymphoblastic leukemia

BACKGROUND

Long-term clinical outcomes are necessary to assess the cost-effectiveness of new treatments over a lifetime horizon. Without long-term clinical trial data, current practice to extrapolate survival beyond the trial period involves fitting alternative parametric models to the observed survival. Choosing the most appropriate model is based on how well each model fits to the observed data. Supplementing trial data with feedback from experts may improve the plausibility of survival extrapolations. We demonstrate the feasibility of formally integrating long-term survival estimates from experts with empirical clinical trial data to provide more credible extrapolated survival curves.

METHODS

The case study involved relapsed or refractory B-cell pediatric and young adult acute lymphoblastic leukemia (r/r pALL) regarding long-term survival for tisagenlecleucel (chimeric antigen receptor T-cell [CAR-T]) with evidence from the phase II ELIANA trial. Seven pediatric oncologists and hematologists experienced with CAR-T therapies were recruited. Relevant evidence regarding r/r pALL and tisagenlecleucel provided a common basis for expert judgments. Survival rates and related uncertainty at 2, 3, 4, and 5 years were elicited from experts using a web-based application adapted from Sheffield Elicitation Framework. Estimates from each expert were combined with observed data using time-to-event parametric models that accounted for experts' uncertainty, producing an overall distribution of survival over time. These results were validated based on longer term follow-up (median duration 24.2 months) from ELIANA following the elicitation.

RESULTS

Extrapolated survival curves based on ELIANA trial without expert information were highly uncertain, differing substantially depending on the model choice. Survival estimates between 2 to 5 years from individual experts varied with a fair amount of uncertainty. However, incorporating expert estimates improved the precision in the extrapolated survival curves. Predictions from a Gompertz model, which experts believed was most appropriate, suggested that more than half of the ELIANA patients treated with tisagenlecleucel will survive up to 5 years. Expert estimates at 24 months were validated by longer follow-up.

CONCLUSIONS

This study provides an example of how expert opinion can be elicited and synthesized with observed survival data using a transparent and formal procedure, capturing expert uncertainty, and ensuring projected long-term survival is clinically plausible.

An indirect treatment comparison of the efficacy of pembrolizumab versus competing regimens for the adjuvant treatment of stage III melanoma

Objective: To determine the efficacy of pembrolizumab relative to other treatments used in stage III melanoma by conducting a systematic literature review (SLR) and network meta-analysis (NMA).

Methods: A SLR was conducted to identify randomized clinical trials (RCTs) evaluating approved adjuvant treatments including interferon-containing regimens, BRAF-inhibitors, and PD-L1 inhibitors in stage III melanoma patients. Relative treatment effects for recurrence-free survival (RFS) were synthesized with Bayesian NMA models that allowed for hazard ratios (HRs) to vary over time.

Results: Included studies formed a connected network of evidence composed of eight trials. In high-risk stage III patients, the HR for pembrolizumab vs observation decreased significantly over time with the superiority of pembrolizumab over observation becoming statistically meaningful before 3 months. By 9 months, the HR for pembrolizumab vs observation was statistically significantly lower than the HR for most other treatments vs observation, with the exception of ipilimumab and biochemotherapy due to overlapping 95% credible intervals. In BRAF + patients, pembrolizumab was statistically significantly better than observation after 3 months. The HR for both BRAF-inhibitors vs observation increased significantly over time and pembrolizumab was statistically superior to both BRAF-inhibitors after 15 months.

Conclusions: Pembrolizumab results in statistically significantly improved RFS compared to all competing regimens after 9 months, except ipilimumab and biochemotherapy, for the adjuvant treatment of stage III melanoma. However, point estimate HRs vs observation for pembrolizumab are much lower than those for ipilimumab. In BRAF + patients, the advantage of pembrolizumab versus competing interventions increases over time with respect to RFS.

Indirect comparison of nivolumab ± ipilimumab (CheckMate 032) versus other treatments for recurrent small-cell lung cancer

Aim: To estimate the comparative efficacy of nivolumab ± ipilimumab versus alternative treatments for small-cell lung cancer after at least one prior line of chemotherapy. Materials & m ethods: A systematic literature review identified six randomized controlled trials (RCTs) that could be connected in a network. The Kaplan-Meier survival curves from these RCTs were synthesized using network meta-analysis models. Aggregate-level matching was used to connect CheckMate 032 to the RCTs. Results: CheckMate 032 was connected to the network by Amrubicin Clinical Trial-1. Nivolumab ± ipilimumab had a more durable tumor response and more favorable long-term survival versus topotecan via intravenous and versus amrubicin. Conclusion: Compared with chemotherapies for recurrent small-cell lung cancer, nivolumab ± ipilimumab improves response duration, which may translate to long-term survival benefits.

What have we learned from network meta-analyses applied to critical care?

It is widely accepted in modern medicine that medical decisions must be supported by scientific evidence. Identifying the best intervention when several options are available constitute a great challenge for every clinician. Traditional meta-analysis (TMA) allows summarizing evidence from studies that compare the same two interventions for one event (head to head studies or direct comparisons). Network meta-analysis (NMA) is a relatively new procedure that allows to compare multiple interventions for one event, even when non-head to head studies have been conducted (indirect evidence). Other advantages of NMA include increasing the accuracy of the results and ranking all the interventions according to their effectiveness. These features are of paramount importance as: 1) they summarize information from events (e.g. diseases or outcomes) that has more than two possible interventions (e.g. treatments or procedures); 2) they strengthen the level of guideline recommendations; and 3) they identify new hypotheses based on indirect comparison. As this is a narrative review, all manuscripts have been selected from PubMed according to our best knowledge with the aim to illustrate different features, options or applications of NMA in critical care. First, we provide a description of the usefulness, interpretation, assumptions and main plots related to NMAs. Second, we analyzed some examples of NMAs related to critical care medicine. Third, we include a pragmatic approach about how results from NMAs can improve the clinical practice as well an R script with a database to conduct an NMAs and reproduce figures and tables that have been shown here. As a conclusion, NMA is an established, robust, objective and reproducible statistic technique that has been applied to several critical care areas. Clinical practice guidelines have started to include NMA evidence to support their recommendations. In future years, it seems highly probable that this technique will increase it applicability in almost all areas of critical care medicine.

Assessing the consistency assumptions underlying network meta-regression using aggregate data

When numerous treatments exist for a disease (Treatments 1, 2, 3, etc), network meta‐regression (NMR) examines whether each relative treatment effect (eg, mean difference for 2 vs 1, 3 vs 1, and 3 vs 2) differs according to a covariate (eg, disease severity). Two consistency assumptions underlie NMR: consistency of the treatment effects at the covariate value 0 and consistency of the regression coefficients for the treatment by covariate interaction. The NMR results may be unreliable when the assumptions do not hold. Furthermore, interactions may exist but are not found because inconsistency of the coefficients is masking them, for example, when the treatment effect increases as the covariate increases using direct evidence but the effect decreases with the increasing covariate using indirect evidence.

We outline existing NMR models that incorporate different types of treatment by covariate interaction. We then introduce models that can be used to assess the consistency assumptions underlying NMR for aggregate data. We extend existing node‐splitting models, the unrelated mean effects inconsistency model, and the design by treatment inconsistency model to incorporate covariate interactions. We propose models for assessing both consistency assumptions simultaneously and models for assessing each of the assumptions in turn to gain a more thorough understanding of consistency.

We apply the methods in a Bayesian framework to trial‐level data comparing antimalarial treatments using the covariate average age and to four fabricated data sets to demonstrate key scenarios.

We discuss the pros and cons of the methods and important considerations when applying models to aggregated data.

Using Evidence from Randomised Controlled Trials in Economic Models: What Information is Relevant and is There a Minimum Amount of Sample Data Required to Make Decisions?

Evidence from randomised controlled trials (RCTs) is used to support regulatory approval and reimbursement decisions. I discuss how these decisions are typically made and argue that the amount of sample data and regulatory authorities' concerns over multiplicity are irrelevant when making reimbursement decisions. Decision analytic models (DAMs) are usually necessary to meet the requirements of an economic evaluation. DAMs involve inputs relating to health benefits and resource use that represent unknown true population parameters. Evidence about parameters may come from a variety of sources, including RCTs, and uncertainty about parameters is represented by their joint posterior distribution. Any impact of multiplicity is mitigated through the prior distribution. I illustrate my perspective with three examples: the estimation of a treatment effect on a rare event; the number of RCTs available in a meta-analysis; and the estimation of population mean overall survival. I conclude by recommending that reimbursement decisions should be followed by an assessment of the value of sample information and the DAM revised structurally as necessary and to include any new sample data that may be generated.

Pharmacological Treatment of Arterial Hypertension in Children and Adolescents: A Network Meta-Analysis

Pharmacological treatment is indicated in children and adolescents with hypertension unresponsive to lifestyle modifications, but there is not enough evidence to recommend 1 class of antihypertensive drugs over others. We performed a network meta-analysis to compare the results of available randomized clinical trials on pharmacological treatment of pediatric hypertension. From a total of 554 potentially relevant studies, 13 randomized placebo-controlled clinical trials enrolling ≥50 patients and a follow-up ≥4 weeks were included. The reduction of systolic blood pressure (SBP) and diastolic BP (DBP) after treatment were the coprimary end points. A total of 2378 pediatric patients, with a median age of 12 years, were included in the analysis. After a median follow-up of 35 days, lisinopril and enalapril were found to be superior to placebo in reducing SBP and DBP, whereas only for DBP, losartan was found to be superior to placebo and lisinopril and enalapril were found to be superior to eplerenone. Angiotensin-converting enzyme inhibitors and angiotensin receptor blockers were associated with a greater SBP and DBP reduction compared with placebo, likewise the mineralocorticoid receptor antagonist was inferior to angiotensin-converting enzyme inhibitors in DBP reduction. The analysis was adjusted for study-level mean age, percentage of women, mean baseline blood pressure, and mean weight, only the latter significantly affected DBP reduction. According to the present analysis, angiotensin-converting enzyme inhibitors and angiotensin receptor blockers could represent the best choice as antihypertensive treatment for pediatric hypertension. However, because of the paucity of available data for the other classes of antihypertensive drugs, definitive conclusions are not allowed and further randomized controlled trials are warranted.

Evaluation of biomarkers for treatment selection using individual participant data from multiple clinical trials

Biomarkers that predict treatment effects may be used to guide treatment decisions, thus improving patient outcomes. A meta-analysis of individual participant data (IPD) is potentially more powerful than a single-study data analysis in evaluating markers for treatment selection. Our study was motivated by the IPD that were collected from 2 randomized controlled trials of hypertension and preeclampsia among pregnant women to evaluate the effect of labor induction over expectant management of the pregnancy in preventing progression to severe maternal disease. The existing literature on statistical methods for biomarker evaluation in IPD meta-analysis have evaluated a marker's performance in terms of its ability to predict risk of disease outcome, which do not directly apply to the treatment selection problem. In this study, we propose a statistical framework for evaluating a marker for treatment selection given IPD from a small number of individual clinical trials. We derive marker-based treatment rules by minimizing the average expected outcome across studies. The application of the proposed methods to the IPD from 2 studies in women with hypertension in pregnancy is presented.

Multiple Score Comparison: a network meta-analysis approach to comparison and external validation of prognostic scores

BACKGROUND

Prediction models and prognostic scores have been increasingly popular in both clinical practice and clinical research settings, for example to aid in risk-based decision making or control for confounding. In many medical fields, a large number of prognostic scores are available, but practitioners may find it difficult to choose between them due to lack of external validation as well as lack of comparisons between them.

METHODS

Borrowing methodology from network meta-analysis, we describe an approach to Multiple Score Comparison meta-analysis (MSC) which permits concurrent external validation and comparisons of prognostic scores using individual patient data (IPD) arising from a large-scale international collaboration. We describe the challenges in adapting network meta-analysis to the MSC setting, for instance the need to explicitly include correlations between the scores on a cohort level, and how to deal with many multi-score studies. We propose first using IPD to make cohort-level aggregate discrimination or calibration scores, comparing all to a common comparator. Then, standard network meta-analysis techniques can be applied, taking care to consider correlation structures in cohorts with multiple scores. Transitivity, consistency and heterogeneity are also examined.

RESULTS

We provide a clinical application, comparing prognostic scores for 3-year mortality in patients with chronic obstructive pulmonary disease using data from a large-scale collaborative initiative. We focus on the discriminative properties of the prognostic scores. Our results show clear differences in performance, with ADO and eBODE showing higher discrimination with respect to mortality than other considered scores. The assumptions of transitivity and local and global consistency were not violated. Heterogeneity was small.

CONCLUSIONS

We applied a network meta-analytic methodology to externally validate and concurrently compare the prognostic properties of clinical scores. Our large-scale external validation indicates that the scores with the best discriminative properties to predict 3 year mortality in patients with COPD are ADO and eBODE.

Network meta-analysis including treatment by covariate interactions: Consistency can vary across covariate values

BACKGROUND

Many reviews aim to compare numerous treatments and report results stratified by subgroups (eg, by disease severity). In such cases, a network meta-analysis model including treatment by covariate interactions can estimate the relative effects of all treatment pairings for each subgroup of patients. Two key assumptions underlie such models: consistency of treatment effects and consistency of the regression coefficients for the interactions. Consistency may differ depending on the covariate value at which consistency is assessed. For valid inference, we need to be confident of consistency for the relevant range of covariate values. In this paper, we demonstrate how to assess consistency of treatment effects from direct and indirect evidence at various covariate values.

METHODS

Consistency is assessed using visual inspection, inconsistency estimates, and probabilities. The method is applied to an individual patient dataset comparing artemisinin combination therapies for treating uncomplicated malaria in children using the covariate age.

RESULTS

The magnitude of the inconsistency appears to be decreasing with increasing age for each comparison. For one comparison, direct and indirect evidence differ for age 1 (P = .05), and this brings results for age 1 for all comparisons into question.

CONCLUSION

When fitting models including interactions, the consistency of direct and indirect evidence must be assessed across the range of covariates included in the trials. Clinical inferences are only valid for covariate values for which results are consistent.

Bayesian one-step IPD network meta-analysis of time-to-event data using Royston-Parmar models

Network meta‐analysis (NMA) combines direct and indirect evidence from trials to calculate and rank treatment estimates. While modelling approaches for continuous and binary outcomes are relatively well developed, less work has been done with time‐to‐event outcomes. Such outcomes are usually analysed using Cox proportional hazard (PH) models. However, in oncology with longer follow‐up time, and time‐dependent effects of targeted treatments, this may no longer be appropriate. Network meta‐analysis conducted in the Bayesian setting has been increasing in popularity. However, fitting the Cox model is computationally intensive, making it unsuitable for many datasets. Royston‐Parmar models are a flexible alternative that can accommodate time‐dependent effects. Motivated by individual participant data (IPD) from 37 cervical cancer trials (5922 women) comparing surgery, radiotherapy, and chemotherapy, this paper develops an IPD Royston‐Parmar Bayesian NMA model for overall survival. We give WinBUGS code for the model. We show how including a treatment‐ln(time) interaction can be used to conduct a global test for PH, illustrate how to test for consistency of direct and indirect evidence, and assess within‐design heterogeneity. Our approach provides a computationally practical, flexible Bayesian approach to NMA of IPD survival data, which readily extends to include additional complexities, such as non‐PH, increasingly found in oncology trials.

Network Meta-analysis of Progression-Free Survival and Overall Survival in First-Line Treatment of BRAF Mutation-Positive Metastatic Melanoma

INTRODUCTION

The present study aimed to inform an economic evaluation of dabrafenib and trametinib combination as first-line treatment of metastatic melanoma in a Canadian setting. A network meta-analysis was conducted to estimate hazard ratios (HRs) for progression-free survival (PFS)and overall survival (OS) of dabrafenib plus trametinib versus other first-line treatments of BRAF mutation-positive metastatic melanoma including dabrafenib, trametinib, vemurafenib, ipilimumab, and dacarbazine (DTIC).

METHODS

HRs for PFS and OS were from randomized controlled trials identified from systematic literature reviews. HRs for PFS and OS (adjusted for crossover as appropriate) were analyzed using multivariate and univariate Bayesian network meta-analysis.

RESULTS

In multivariate network-meta analyses (HRs for PFS and OS estimated simultaneously to account for the correlation of treatment effects on PFS and OS), HRs (95% credible interval) for PFS and OS favored dabrafenib plus trametinib [PFS: 0.23 (0.18-0.29) versus DTIC, 0.32 (0.24-0.42) versus ipilimumab plus DTIC, 0.52 (0.32-0.83) versus trametinib, 0.57 (0.48-0.69) versus vemurafenib, and 0.59 (0.50-0.71) versus dabrafenib]; OS [0.41 (0.29-0.56) versus DTIC, 0.52 (0.38-0.71) versus ipilimumab plus DTIC, 0.68 (0.47-0.95) versus trametinib, 0.69 (0.57-0.84) versus vemurafenib, and 0.72 (0.60-0.85) versus dabrafenib]. The beneficial effects on OS of dabrafenib plus trametinib versus ipilimumab plus DTIC and versus trametinib were attenuated when HRs were estimated using univariate network meta-analysis (HRs for PFS and OS estimated separately).

CONCLUSION

This analysis demonstrates improved PFS and OS with dabrafenib + trametinib versus dabrafenib, trametinib, vemurafenib, ipilimumab plus DTIC, and DTIC as first-line treatment for patients with BRAF mutation-positive metastatic melanoma.

FUNDING

Novartis Pharmaceuticals.

Assumptions of Mixed Treatment Comparisons in Health Technology Assessments - Challenges and Possible Steps for Practical Application

The validity of mixed treatment comparisons (MTCs), also called network meta-analysis, relies on whether it is reasonable to accept the underlying assumptions on similarity, homogeneity, and consistency. The aim of this paper is to propose a practicable approach to addressing the underlying assumptions of MTCs. Using data from clinical studies of antidepressants included in a health technology assessment (HTA), we present a stepwise approach to dealing with challenges related to checking the above assumptions and to judging the robustness of the results of an MTC. At each step, studies that were dissimilar or contributed to substantial heterogeneity or inconsistency were excluded from the primary analysis. In a comparison of the MTC estimates from the consistent network with the MTC estimates from the homogeneous network including inconsistencies, few were affected by notable changes; that is, a change in effect size (factor 2), direction of effect or statistical significance. Considering the small proportion of studies excluded from the network due to inconsistency, as well as the number of notable changes, the MTC results were deemed sufficiently robust. In the absence of standard methods, our approach to checking assumptions in MTCs may inform other researchers in need of practical options, particularly in HTA.

Treatment of idiopathic pulmonary fibrosis: a network meta-analysis

BACKGROUND

Idiopathic pulmonary fibrosis (IPF) is an interstitial lung disease associated with high morbidity and mortality. Effective treatments for IPF are limited. Several recent studies have investigated novel therapeutic agents for IPF, but very few have addressed their comparative benefits and harms.

METHODS

We performed a Bayesian network meta-analysis (NMA) to assess the effects of different treatments for IPF on mortality and serious adverse events (SAEs). We searched MEDLINE and EMBASE for randomized controlled trials (RCTs) up to August 2015. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach served to assess the certainty in the evidence of direct and indirect estimates. We calculated the surface under the cumulative ranking curve (SUCRA) for each treatment. We included parallel group RCTs, including factorial designs, but excluded quasi-randomized and cross-over trials. Studies were only included if they involved adult (≥18 years of age) patients with IPF as defined by the 2011 criteria and examined one of the 10 interventions of interest (ambrisentan, bosentan, imatinib, macitentan, N-acetylcysteine, nintedanib, pirfenidone, sildenafil, prednisone/azathioprine/N-acetylcysteine triple therapy, and vitamin K antagonist).

RESULTS

A total of 19 RCTs (5,694 patients) comparing 10 different interventions with placebo and an average follow-up period of 1 year fulfilled the inclusion criteria. SUCRA analysis suggests nintedanib, pirfenidone, and sildenafil are the three treatments with the highest probability of reducing mortality in IPF. Indirect comparison showed no significant difference in mortality between pirfenidone and nintedanib (NMA OR, 1.05; 95% CrI, 0.45-2.78, moderate certainty of evidence), pirenidone and sildenafil (NMA OR, 2.26; 95% CrI, 0.44-13.17, low certainty of evidence), or nintedanib and sildenafil (NMA OR 2.40; 95% CrI, 0.47-14.66, low certainty of evidence). Sildenafil, pirfenidone, and nintedanib were ranked second, fourth, and sixth out of 10 for SAEs.

CONCLUSION

In the absence of direct comparisons between treatment interventions, this NMA suggests that treatment with nintedanib, pirfenidone, and sildenafil extends survival in patients with IPF. The SAEs of these agents are similar to the other interventions and include mostly dermatologic and gastrointestinal manifestations. Head-to-head comparisons need to confirm these findings.

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.

Efficacy of limited fluid resuscitation in patients with hemorrhagic shock: a meta-analysis

BACKGROUNDS

The objective of this meta-analysis was to evaluate the efficacy of limited fluid resuscitation during active hemorrhage compared with regular fluid resuscitation and provide strong evidences for the improvement of fluid resuscitation strategies in uncontrolled hemorrhagic shock.

METHODS

Electronic searches were performed using PubMed, Medline, Embase and CNKI in accordance with pre-set guidelines. Clinical trials and observation studies were included or excluded according to the criteria. The endpoints examined were mortality, hemoglobin (Hb), platelets (PLT), hematocrit (Hct), prothrombin Time (PT), activated partial thromboplastin time (APTT), base excess (BE), blood lactic acid (BLA) and the main complications, such as multiple organ dysfunction syndrome (MODS) and acute Respiratory Distress Syndrome (ARDS). Risk ratios (RR), mean differences (MDs) and 95% confidence intervals (95% Cl) were calculated using fixed/random effect model.

RESULTS

The search indentified 11 studies including 1482 subjects. 725 hemorrhagic patients were treated with limited fluid resuscitation while 757 patients undertook regular fluid resuscitation during active hemorrhage. Limited fluid resuscitation had its advantage to reduce the mortality in hemorrhagic shock (RR = 0.67; 95% CI = 0.56-0.81; P < 0.0001) and easily controlled the blood routine index close to normal compared with regular fluid resuscitation (Hb: MD = 13.04; 95% CI = 2.69-23.38; P = 0.01. PLT: MD = 23.16; 95% CI = 6.41-39.91; P = 0.007. Hct: MD = 0.02; 95% CI = 0.02-0.03; P < 0.00001). LFR also had shorter PT and APTT compared with RFR (PT: MD = -2.81; 95% CI = -3.44--2.17; P < 0.00001 and APTT: MD = -5.14; 95% CI = -6.16--4.12; P < 0.00001). As for blood gas analysis, LFR reduced the decrease of BE (MD = 2.48; 95% CI = 1.11-3.85; P = 0.0004) and increase of BLA (MD = -0.65; 95% CI = -0.85--0.44; P < 0.00001). Besides, LFR may also reduce the occurrence of postoperative complications (MODS: RR= 0.37; 95% CI = 0.21-0.66; P = 0.0008. ARDS: RR = 0.35; 95% CI = 0.21-0.60; P < 0.0001).

CONCLUSION

The results provide convincing evidence that support the continued investigation and use of limited fluid resuscitation during active hemorrhage in the trauma setting.

Network Meta-analysis: Users' Guide for Surgeons: Part II - Certainty

In the previous article (Network Meta-analysis: Users' Guide for Surgeons-Part I, Credibility), we presented an approach to evaluating the credibility or methodologic rigor of network meta-analyses (NMA), an innovative approach to simultaneously addressing the relative effectiveness of three or more treatment options for a given medical condition or disease state. In the second part of the Users' Guide for Surgeons, we discuss and demonstrate the application of criteria for determining the certainty in effect sizes and directions associated with a given treatment option through an example pertinent to clinical orthopaedics.

Network Meta-analysis: Users' Guide for Surgeons: Part I - Credibility

Conventional meta-analyses quantify the relative effectiveness of two interventions based on direct (that is, head-to-head) evidence typically derived from randomized controlled trials (RCTs). For many medical conditions, however, multiple treatment options exist and not all have been compared directly. This issue limits the utility of traditional synthetic techniques such as meta-analyses, since these approaches can only pool and compare evidence across interventions that have been compared directly by source studies. Network meta-analyses (NMA) use direct and indirect comparisons to quantify the relative effectiveness of three or more treatment options. Interpreting the methodologic quality and results of NMAs may be challenging, as they use complex methods that may be unfamiliar to surgeons; yet for these surgeons to use these studies in their practices, they need to be able to determine whether they can trust the results of NMAs. The first judgment of trust requires an assessment of the credibility of the NMA methodology; the second judgment of trust requires a determination of certainty in effect sizes and directions. In this Users' Guide for Surgeons, Part I, we show the application of evaluation criteria for determining the credibility of a NMA through an example pertinent to clinical orthopaedics. In the subsequent article (Part II), we help readers evaluate the level of certainty NMAs can provide in terms of treatment effect sizes and directions.