Safety and efficacy of hydroxychloroquine as prophylactic against COVID-19 in healthcare workers: a meta-analysis of randomised clinical trials

OBJECTIVE

We studied the safety and efficacy of hydroxychloroquine (HCQ) as pre-exposure prophylaxis for COVID-19 in healthcare workers (HCWs), using a meta-analysis of randomised controlled trials (RCTs).

DATA SOURCES

PubMed and EMBASE databases were searched to identify randomised trials studying HCQ.

STUDY SELECTION

Ten RCTs were identified (n=5079 participants).

DATA EXTRACTION AND SYNTHESIS

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were used in this systematic review and meta-analysis between HCQ and placebo using a Bayesian random-effects model. A pre-hoc statistical analysis plan was written.

MAIN OUTCOMES

The primary efficacy outcome was PCR-confirmed SARS-CoV-2 infection and the primary safety outcome was incidence of adverse events. The secondary outcome included clinically suspected SARS-CoV-2 infection.

RESULTS

Compared with placebo, HCWs randomised to HCQ had no significant difference in PCR-confirmed SARS-CoV-2 infection (OR 0.92, 95% credible interval (CI): 0.58, 1.37) or clinically suspected SARS-CoV-2 infection (OR 0.78, 95% CI: 0.57, 1.10), but significant difference in adverse events (OR 1.35, 95% CI: 1.03, 1.73).

CONCLUSIONS AND RELEVANCE

Our meta-analysis of 10 RCTs investigating the safety and efficacy of HCQ as pre-exposure prophylaxis in HCWs found that compared with placebo, HCQ does not significantly reduce the risk of confirmed or clinically suspected SARS-CoV-2 infection, while HCQ significantly increases adverse events.

PROSPERO REGISTRATION NUMBER

CRD42021285093.

Nonpharmacologic and Pharmacologic Treatments of Adult Patients With Major Depressive Disorder: A Systematic Review and Network Meta-analysis for a Clinical Guideline by the American College of Physicians

BACKGROUND

Primary care patients and clinicians may prefer alternative options to second-generation antidepressants for major depressive disorder (MDD).

PURPOSE

To compare the benefits and harms of nonpharmacologic treatments with second-generation antidepressants as first-step interventions for acute MDD, and to compare second-step treatment strategies for patients who did not achieve remission after an initial attempt with antidepressants.

DATA SOURCES

English-language studies from several electronic databases from 1 January 1990 to 8 August 2022, trial registries, gray literature databases, and reference lists to identify unpublished research.

STUDY SELECTION

2 investigators independently selected randomized trials of at least 6 weeks' duration.

DATA EXTRACTION

Reviewers abstracted data about study design and conduct, participants, interventions, and outcomes. They dually rated the risk of bias of studies and the certainty of evidence for outcomes of interest.

DATA SYNTHESIS

65 randomized trials met the inclusion criteria; eligible data from nonrandomized studies were not found. Meta-analyses and network meta-analyses indicated similar benefits of most nonpharmacologic treatments and antidepressants as first-step treatments. Antidepressants had higher risks for discontinuation because of adverse events than most other treatments. For second-step therapies, different switching and augmentation strategies provided similar symptomatic relief. The certainty of evidence for most comparisons is low; findings should be interpreted cautiously.

LIMITATIONS

Many studies had methodological limitations or dosing inequalities; publication bias might have affected some comparisons. In some cases, conclusions could not be drawn because of insufficient evidence.

CONCLUSION

Although benefits seem to be similar among first- and second-step MDD treatments, the certainty of evidence is low for most comparisons. Clinicians and patients should focus on options with the most reliable evidence and take adverse event profiles and patient preferences into consideration.

PRIMARY FUNDING SOURCE

American College of Physicians. (PROSPERO: CRD42020204703).

Effect of dietary patterns on cardiovascular risk factors in people with type 2 diabetes. A systematic review and network meta-analysis

AIMS

To identify the most effective dietary pattern for improving cardiovascular risk factors in people with type 2 diabetes.

METHODS

PubMed, Embase, the Cochrane library, SCOPUS and Web of Science were systematically searched for randomized controlled trials comparing the effects of dietary patterns on body weight, blood pressure, HbA1c and lipids after 6 and 12 months. Treatment effects were synthesized using Bayesian network meta-analysis. Six-month changes in HbA1c, SBP and LDL-C were used to estimate relative risk reductions (RRR) for cardiovascular events.

RESULTS

Seventy-three RCTs on eight different dietary patterns were included. All reduced body weight and HbA1c after 6 months, with the largest effects from the low carbohydrate (body weight -4.8 kg, 95 %credibility interval (95 %CrI) -6.5;-3.2 kg) and Mediterranean diet (HbA1c -1.0 %, 95 %CrI -15;-0.4 % vs usual diet). There were no significant 6-month blood pressure or lipid effects. Dietary patterns had non-statistically significant 12-months effects. The Mediterranean diet resulted in the largest expected RRR for cardiovascular events: -16 % (95 %CI -31;3.0) vs usual diet.

CONCLUSIONS

In patients with type 2 diabetes, all dietary patterns outperformed usual diet in improving body weight and HbA1c after 6 months and clinically relevant cardiovascular risk reduction could be achieved. There was insufficient evidence to select one optimal dietary pattern.

Theory and practice of Bayesian and frequentist frameworks for network meta-analysis

Network meta-analysis (NMA) is an increasingly popular statistical method of synthesising evidence to assess the comparative benefits and harms of multiple treatments in a single analysis. Several automated software packages facilitate conducting NMA using either of two alternative approaches, Bayesian or frequentist frameworks. Researchers must choose a framework for conducting NMA (Bayesian or frequentist) and select appropriate model(s), and those conducting NMA need to understand the assumptions and limitations of different approaches. Bayesian models are more frequently used and can be more flexible but require checking additional assumptions and greater statistical expertise that are often ignored. The present paper describes the important theoretical aspects of Bayesian and frequentist models for NMA and the applications and considerations of contrast-synthesis and arm-synthesis NMAs. In addition, we present evidence from a limited number of simulation and empirical studies that compared different frequentist and Bayesian models and provide an overview of available automated software packages to perform NMA. We will conclude that when analysts choose appropriate models, there are seldom important differences in the results of Bayesian and frequentist approaches and that network meta-analysts should therefore focus on model features rather than the statistical framework.

Morbidity of the Donor Site and Complication Rates of Breast Reconstruction with Autologous Abdominal Flaps: A Systematic Review and Meta-Analysis

Background. Numerous studies have evaluated the use of autologous abdominal tissue for breast reconstruction; nevertheless, complications and donor site morbidity rates vary significantly. The study aims to compare the literature regarding morbidity of the donor site and complication rates of breast reconstruction with autologous abdominal flaps. Methods. The databases of MEDLINE, EBSCO, Scopus, Wiley Library, and Web of Sciences were searched for studies that compared different flaps in terms of complications and donor site morbidity. The procedures studied included pedicled transverse rectus abdominis myocutaneous flap (pTRAM), free TRAM (fTRAM), deep inferior epigastric perforator (DIEP), and superficial inferior epigastric artery perforator (SIEA) flaps. A total of 34 studies were included. Of these, 28 were retrospective studies and 9 were prospective cohort studies. Results. When compared to DIEP, fTRAM flaps were found to have a decreased incidence of flap fat necrosis, hematoma, and total thrombotic events, yet a higher risk of donor site hernia/bulging. pTRAM flaps were also associated with an increased risk of hernia/bulging at the donor site, as well as wound infection, yet flap hematoma was less common. On the other hand, SIEA flaps showed the lowest risk of donor site hernia/bulging while still having a high risk of wound infection. Conclusion. fTRAM procedures comparatively had the least complications. However, regarding flap choice, patients would benefit most from a case-by-case analysis, taking into consideration individual risk factors and preferences.

Interpreting and assessing confidence in network meta-analysis results: an introduction for clinicians

Purpose

We aimed to provide clinicians with introductory guidance for interpreting and assessing confidence in on Network meta-analysis (NMA) results.

Methods

We reviewed current literature on NMA and summarized key points.

Results

Network meta-analysis (NMA) is a statistical method for comparing the efficacy of three or more interventions simultaneously in a single analysis by synthesizing both direct and indirect evidence across a network of randomized clinical trials. It has become increasingly popular in healthcare, since direct evidence (head-to-head randomized clinical trials) are not always available. NMA methods are categorized as either Bayesian or frequentist, and while the two mostly provide similar results, the two approaches are theoretically different and require different interpretations of the results.

Conclusions

We recommend a careful approach to interpreting NMA results and the validity of an NMA depends on its underlying statistical assumptions and the quality of the evidence used in the NMA.

Comparison of Various Vagal Maneuvers for Supraventricular Tachycardia by Network Meta-Analysis

BACKGROUND

Vagal maneuvers (VagMs) are recommended as the first-line treatment of supraventricular tachycardia (SVT). However, the optimal type of VagMs remains unproven.

AIM

This study aims to compare the effectiveness and adverse events amongst VagMs on SVT via network meta-analyses (NMAs).

METHODS

We systematically searched randomized controlled trials (RCTs) that involved adults with SVT and compared VagMs without language restrictions. We determined the initial and final responses of conversion rate to sinus rhythm and adverse events. Risk of bias (RoB) was appraised by Cochrane revised tool, and contribution matrix was calculated. NMAs were synthesized using frequentist random-effects model and presented as relative risk (RR) with 95% CI. The order of probability was presented as surface under the cumulative ranking curve analysis (SUCRA). Sensitivity analysis was performed using both Bayesian and frequentist approach with fixed- or random-effects models. Certainty of evidence (CoE) was rated by using the Grading of Recommendations, Assessment, Development, and Evaluations methodology.

RESULTS

Fourteen RCTs with 2,180 patients were enrolled. Small portion of mixed estimates was contributed from high overall RoB studies. Compared with carotid sinus massage (CSM), the modified Valsalva maneuver (MVM) was the most effective VagM after initial performance [SUCRA: 0.9992, RR: 5.47 (1.77-16.93)] and at the end of study [SUCRA: 1.0000, RR: 3.62 (2.04-6.39), CoE: high]. The standard VM did not elicit better conversion rate to the sinus rhythm than CSM at the initial response [SUCRA: 0.4395, RR: 1.97 (0.63-6.15)] and at the end of the study [SUCRA: 0.4795, RR: 1.64 (0.94-2.87), CoE: moderate]. The SUCRA value of CSM at the initial and final responses was the least one amongst three VagMs (0.0613 and 0.0205, respectively). Adverse events amongst three VagMs were similar (CoE: low). Sensitivity analyses yielded consistent results.

CONCLUSION

We recommended MVM as the first choice of VagM for rhythm conversion before the pharmacological management of SVT.

Antioxidant micronutrient supplements for adult critically ill patients: A bayesian multiple treatment comparisons meta-analysis

BACKGROUND & AIMS

Antioxidant micronutrients (AxMs) have been administered to critically ill adults attempting to counteract the oxidative stress imposed during critical illness. However, results are conflicting and relative effectiveness of AxMs regimens is unknown. We conducted a Bayesian multi-treatment comparison (MTC) meta-analysis to identify the best AxM treatment regimen that will improve clinical outcomes.

METHODS

PubMed, EMBASE, Web of Science and Cochrane databases were searched from the inception of databases through August 2020. Randomized controlled trials (RCT) comparing AxMs supplementations with placebo among critically ill adults were included. Two authors assessed trial quality using Cochrane risk of bias tool and assessed certainty of evidence (CoE). A random effect model, non-informative priors Bayesian MTC meta-analysis using gemtc package in R version 3.6.2 was performed. AxMs treatment effect on clinical outcomes (mortality, infection rates, intensive care unit (ICU) and hospital stays and ventilator days) were represented by absolute risk differences (ARD) for dichotomous outcomes and mean differences (MD) for continuous outcomes. Prior to final analysis, we repeated the search through January 2021.

RESULTS

37 RCT (4905 patients) were included with 16 direct comparisons. With respect to mortality, the ARD for "vitamin E" compared with placebo was centred at -0.19 [95%CrI: -0.54,0.16; very low CoE] and was ranked the best treatment for mortality reduction as per surface under the cumulative ranking curve (SUCRA 0.71, 95%CrI: 0.07,1.00). A combination of "selenium, zinc and copper" was ranked the best for lowest ICU stay [-9.40, 95% CrI: -20.0,1.50; low CoE]. A combination of "selenium, zinc, copper and vitamin E" was ranked the best treatment for infection risk reduction [-0.22, 95% CrI: -0.61,0.17; very low CoE]. Ventilator days were least with a combination of "selenium, zinc and manganese" [2.80, 95% CrI: -6.30,0.89; low CoE]. Hospital stay was the lowest using a combination of "selenium, zinc and copper" [-13.00, 95% CrI: -38.00,13.00; very low CoE]. There is substantial uncertainty present in the rankings due to wide and overlapping 95% CrIs of SUCRA scores for the treatments.

CONCLUSIONS

Studies on critically ill adult patients have suggested a possible beneficial effects of certain AxM supplementations over and above the recommended dietary allowance. However, evidence does not support their use in clinical practice due to the low confidence in the estimates. Current studies evaluating specific AxMs or their combinations are limited with small sample sizes.

REGISTRATION

PROSPERO, CRD42020210199.

TAKE-HOME MESSAGE

Evidence suggesting a potential benefit of AxMs use more than recommended doses in critically ill adults is weak, indicating that there is no justification for this practice.

Predictive P-score for treatment ranking in Bayesian network meta-analysis

Background

Network meta-analysis (NMA) is a widely used tool to compare multiple treatments by synthesizing different sources of evidence. Measures such as the surface under the cumulative ranking curve (SUCRA) and the P-score are increasingly used to quantify treatment ranking. They provide summary scores of treatments among the existing studies in an NMA. Clinicians are frequently interested in applying such evidence from the NMA to decision-making in the future. This prediction process needs to account for the heterogeneity between the existing studies in the NMA and a future study.

Methods

This article introduces the predictive P-score for informing treatment ranking in a future study via Bayesian models. Two NMAs were used to illustrate the proposed measure; the first assessed 4 treatment strategies for smoking cessation, and the second assessed treatments for all-grade treatment-related adverse events. For all treatments in both NMAs, we obtained their conventional frequentist P-scores, Bayesian P-scores, and predictive P-scores.

Results

In the two examples, the Bayesian P-scores were nearly identical to the corresponding frequentist P-scores for most treatments, while noticeable differences existed for some treatments, likely owing to the different assumptions made by the frequentist and Bayesian NMA models. Compared with the P-scores, the predictive P-scores generally had a trend to converge toward a common value of 0.5 due to the heterogeneity. The predictive P-scores’ numerical estimates and the associated plots of posterior distributions provided an intuitive way for clinicians to appraise treatments for new patients in a future study.

Conclusions

The proposed approach adapts the existing frequentist P-score to the Bayesian framework. The predictive P-score can help inform medical decision-making in future studies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-021-01397-5.

Risk of recurrence among patients with HR-positive, HER2-negative, early breast cancer receiving adjuvant endocrine therapy: A systematic review and meta-analysis

BACKGROUND

A systematic review and meta-analysis was conducted to assess breast cancer (BC) outcomes among patients with early-stage hormone receptor positive (HR+), human epidermal growth factor receptor 2-negative (HER2-) BC, receiving adjuvant endocrine therapy.

METHODS

Randomized controlled trials (RCTs) and real-world evidence (RWE) studies were identified using Ovid MEDLINE®, Embase, and Evidence-Based Medicine Reviews. Clinical and methodological similarities including alignment of outcome definitions with standardized definitions for efficacy endpoints criteria were assessed to evaluate feasibility of conducting a meta-analysis. Where feasible, 5-year probabilities of BC recurrence or death were estimated using a Bayesian hierarchical arm-based model.

RESULTS

Of 21 included studies, 8 RCTs and 4 RWE studies reported outcome data of interest. There was heterogeneity in outcome reporting, as well as variation in recurrence risk amongst studies with aligned reporting. Of the 12 studies, 10 were considered for inclusion in a meta-analysis of BC recurrence or death. Only a subgroup analysis of node-positive patients (3 studies; n = 7307) was deemed feasible. The 5-year probability of BC recurrence or death was 17.2% (95% credible interval: 14.6%-20.3%).

CONCLUSION

Although studies reporting recurrence outcomes were limited, there remains a high risk of BC recurrence, especially among node-positive patients. Approximately 1 in 6 women with node-positive HR+/HER2- early-stage BC receiving endocrine therapy experience recurrence or death within 5-years of initiating treatment, suggesting a need for novel treatments for this population.

Comparisons of Efficacy and Safety between Triple (Inhaled Corticosteroid/Long-Acting Muscarinic Antagonist/Long-Acting Beta-Agonist) Therapies in Chronic Obstructive Pulmonary Disease: Systematic Review and Bayesian Network Meta-Analysis

BACKGROUND

Various combinations of inhaled corticosteroid (ICS), long-acting muscarinic antagonist (LAMA), and long-acting beta-agonist (LABA) have been used as triple therapy for stable chronic obstructive pulmonary disease (COPD).

OBJECTIVE

Our study was conducted to answer whether there were significant differences among various combinations in efficacy, for reducing exacerbation or mortality, and in safety, for increasing cardiovascular events or pneumonia.

METHOD

We searched parallel-group randomized controlled trials (RCTs) comparing ICS/LAMA/LABA with other inhaled drugs in patients with stable COPD for at least 12 weeks in PubMed, EMBASE, the Cochrane Library, and clinical trial registries from inception to December 31, 2019. We conducted a network meta-analysis with Bayesian statistics using a random-effects model with heterogeneous variance structure (PROSPERO, CRD42019126757).

RESULTS

Nine different combinations of ICS/LAMA/LABA were identified in 21 RCTs containing 29,892 patients with moderate to very severe COPD. We could not find any significant evidence suggesting a better treatment for reducing total exacerbations or all-cause mortality among ICS/LAMA/LABA combinations. There were also no significant differences in moderate to severe exacerbation, COPD-related mortality, or cardiovascular disease-related mortality among ICS/LAMA/LABA combinations, and the risk of major adverse cardiovascular events was not different. A significantly lower risk of pneumonia was found in fluticasone propionate (FP)/glycopyrrolate/salmeterol (SAL) than FP/tiotropium/SAL {median odds ratio [OR] (95% credible interval [CrI]) = 0 [0-0.72]} and FP/umeclidinium/SAL {median OR (95% Crl) = 0 [0-0.97]}.

CONCLUSION

There were no significant differences in clinical outcomes, including acute exacerbation and all-cause mortality among various ICS/LAMA/LABA combinations in patients with moderate to very severe COPD.

A Bayesian dose-response meta-analysis model: A simulations study and application

Dose-response models express the effect of different dose or exposure levels on a specific outcome. In meta-analysis, where aggregated-level data is available, dose-response evidence is synthesized using either one-stage or two-stage models in a frequentist setting. We propose a hierarchical dose-response model implemented in a Bayesian framework. We develop our model assuming normal or binomial likelihood and accounting for exposures grouped in clusters. To allow maximum flexibility, the dose-response association is modelled using restricted cubic splines. We implement these models in R using JAGS and we compare our approach to the one-stage dose-response meta-analysis model in a simulation study. We found that the Bayesian dose-response model with binomial likelihood has lower bias than the Bayesian model with normal likelihood and the frequentist one-stage model when studies have small sample size. When the true underlying shape is log-log or half-sigmoid, the performance of all models depends on choosing an appropriate location for the knots. In all other examined situations, all models perform very well and give practically identical results. We also re-analyze the data from 60 randomized controlled trials (15,984 participants) examining the efficacy (response) of various doses of serotonin-specific reuptake inhibitor (SSRI) antidepressant drugs. All models suggest that the dose-response curve increases between zero dose and 30-40 mg of fluoxetine-equivalent dose, and thereafter shows small decline. We draw the same conclusion when we take into account the fact that five different antidepressants have been studied in the included trials. We show that implementation of the hierarchical model in Bayesian framework has similar performance to, but overcomes some of the limitations of the frequentist approach and offers maximum flexibility to accommodate features of the data.

Bayesian Methods for Meta-Analyses of Binary Outcomes: Implementations, Examples, and Impact of Priors

Bayesian methods are an important set of tools for performing meta-analyses. They avoid some potentially unrealistic assumptions that are required by conventional frequentist methods. More importantly, meta-analysts can incorporate prior information from many sources, including experts’ opinions and prior meta-analyses. Nevertheless, Bayesian methods are used less frequently than conventional frequentist methods, primarily because of the need for nontrivial statistical coding, while frequentist approaches can be implemented via many user-friendly software packages. This article aims at providing a practical review of implementations for Bayesian meta-analyses with various prior distributions. We present Bayesian methods for meta-analyses with the focus on odds ratio for binary outcomes. We summarize various commonly used prior distribution choices for the between-studies heterogeneity variance, a critical parameter in meta-analyses. They include the inverse-gamma, uniform, and half-normal distributions, as well as evidence-based informative log-normal priors. Five real-world examples are presented to illustrate their performance. We provide all of the statistical code for future use by practitioners. Under certain circumstances, Bayesian methods can produce markedly different results from those by frequentist methods, including a change in decision on statistical significance. When data information is limited, the choice of priors may have a large impact on meta-analytic results, in which case sensitivity analyses are recommended. Moreover, the algorithm for implementing Bayesian analyses may not converge for extremely sparse data; caution is needed in interpreting respective results. As such, convergence should be routinely examined. When select statistical assumptions that are made by conventional frequentist methods are violated, Bayesian methods provide a reliable alternative to perform a meta-analysis.

A network meta-analysis for neoadjuvant and adjuvant treatments for resectable squamous cell carcinoma of esophagus

The optimal treatment for resectable esophageal squamous cell carcinoma (ESCC) is still a debatable point; however, randomized trials for strategies including neoadjuvant or adjuvant chemotherapy (CT), radiotherapy, or chemoradiotherapy (CRT) are not always available. This network meta-analysis aimed to identify an effective approach through indirect comparisons. An extensive literature search comparing multimodality treatment and surgery was performed, and a network meta-analysis was conducted with the frequentist method. Twenty-three trials including a total of 3636 ESCC patients were included. Neoadjuvant CRT and neoadjuvant CT, which were recommended by most guidelines for esophageal cancer, were associated with an overall survival advantage compared with surgery alone (HR = 0.43, 95% CI 0.26-0.73; HR = 0.71, 95% CI 0.32-1.59). A statistically significant survival benefit from neoadjuvant CRT compared with neoadjuvant CT could not be demonstrated in our study (HR = 0.61, 95% CI 0.32-1.17, P = 0.08). Our network meta-analysis showed that both neoadjuvant CRT and neoadjuvant CT were effective in improving the survival of patients with ESCC. Individual clinical decisions need further study in the future.

Comparative efficacy of different ultrasound-guided ablation for the treatment of benign thyroid nodules: Systematic review and network meta-analysis of randomized controlled trials

BACKGROUND

Percutaneous ablation is currently deemed an additionally treatment option for benign thyroid nodules in the world, but possibly different effect among the ablation modalities is not clear. So we aim to evaluate the efficacy and complications of thermal/chemical ablation by network meta-analysis.

MATERIALS AND METHODS

In the network meta-analysis, PubMed, EMBASE and the Cochrane Library databases were searched from 1980 to 2020. Studies of adults with thyroid benign nodules under percutaneous ablation therapy were included. Percentage mean volume change, symptom score change, cosmetic score change and complications were evaluated by network meta-analysis.

RESULTS

In the network meta-analysis, Radiofrequency Ablation(RFA) with 2 treatment sessions group was associated with the highest reduction for the mean volume change during 6-month follow-up (MD = 79.09 and 95% CrI:48.23-89.94). There is no significant difference in the incidence of complications. Subgroup analysis showed that 2 sessions of Radiofrequency Ablation (RFA) ranks the highest probability (surface under the cumulative ranking curve (SUCRA) values 77.9) of being the most efficacious treatment for solid or predominantly solid benign nodules. Ethanol ablation (EA) ranked first (SUCRA value 81.1) in the treatment for cyst or predominantly cyst benign nodules.

CONCLUSION

RFA appears to be superior to other US-guided percutaneous ablation in reducing benign thyroid nodule volume during short- and long-term follow-up. In the subgroup analysis, RFA with 2 treatment sessions showed the most significant effectiveness for solid benign thyroid nodules and EA showed more effectiveness to decrease the volume of cyst benign thyroid nodules.

Comparative Efficacy and Safety of PD-1/PD-L1 Inhibitors for Patients with Solid Tumors: A Systematic Review and Bayesian Network Meta-analysis

Purpose: The clinical use of immunotherapies targeting programmed death 1 (PD-1)/programmed death ligand 1 (PD-L1) is rapid expanding, but the equivalency of these inhibitors remains unclear. We aimed to comprehensively compare the efficacy and safety of PD-1/PD-L1 inhibitors with a systematic review and Bayesian network meta-analysis

Methods: We searched PubMed, Web of Knowledge, related reviews and abstracts for randomized controlled trials of five PD-1/PD-L1 inhibitors for patients with solid tumors before November 30^th, 2018. We estimated summary hazard ratios (HRs) for overall survival (OS) and progression-free survival (PFS), and odds ratios (ORs) for grade 3-5 treatment-related adverse events (TrAEs) using pairwise and network meta-analysis with random-effects. This study was registered with PROSPERO (#CRD42018116624).

Results: Totally, 43 reports of 35 trials comprising 21261 patients were eligible for the analysis. Nivolumab, pembrolizumab, atezolizumab and durvalumab were more effective than control treatment, and no significant differences were identified in OS and PFS between any two inhibitors. Avelumab was associated with significantly inferior OS to nivolumab (HR 1.37, 95%CrI 1.05-1.78) and pembrolizumab (HR 1.33, 95%CrI 1.02-1.73), and with inferior PFS to nivolumab (HR 1.60, 95%CrI 1.03-2.51). Compared with placebo, nivolumab had increased risk of grade 3-5 TrAEs (OR 2.35, 95%CrI 1.35-4.17). Compared with standard-of-care, nivolumab (OR 0.39, 95%CrI 0.28-0.54), pembrolizumab (OR 0.43, 95%CrI 0.30-0.60), atezolizumab (OR 0.37, 95%CrI 0.21-0.64) and avelumab (OR 0.24, 95%CrI 0.12-0.48) significantly reduced grade 3-5 TrAEs. There were not significant differences in grade 3-5 TrAEs between any two inhibitors.

Conclusion: This Bayesian network meta-analysis revealed that nivolumab, pembrolizumab, atezolizumab and durvalumab yielded equivalent survival, while avelumab was associated with unfavorable survival. PD-1/PD-L1 inhibitors were comparable in the risk of TrAEs, and safer than conventional therapies.

Meta-analysis of rare adverse events in randomized clinical trials: Bayesian and frequentist methods

BACKGROUND/AIMS

Regulatory approval of a drug or device involves an assessment of not only the benefits but also the risks of adverse events associated with the therapeutic agent. Although randomized controlled trials (RCTs) are the gold standard for evaluating effectiveness, the number of treated patients in a single RCT may not be enough to detect a rare but serious side effect of the treatment. Meta-analysis plays an important role in the evaluation of the safety of medical products and has advantage over analyzing a single RCT when estimating the rate of adverse events.

METHODS

In this article, we compare 15 widely used meta-analysis models under both Bayesian and frequentist frameworks when outcomes are extremely infrequent or rare. We present extensive simulation study results and then apply these methods to a real meta-analysis that considers RCTs investigating the effect of rosiglitazone on the risks of myocardial infarction and of death from cardiovascular causes.

RESULTS

Our simulation studies suggest that the beta hyperprior method modeling treatment group-specific parameters and accounting for heterogeneity performs the best. Most models ignoring between-study heterogeneity give poor coverage probability when such heterogeneity exists. In the data analysis, different methods provide a wide range of log odds ratio estimates between rosiglitazone and control treatments with a mixed conclusion on their statistical significance based on 95% confidence (or credible) intervals.

CONCLUSION

In the rare event setting, treatment effect estimates obtained from traditional meta-analytic methods may be biased and provide poor coverage probability. This trend worsens when the data have large between-study heterogeneity. In general, we recommend methods that first estimate the summaries of treatment-specific risks across studies and then relative treatment effects based on the summaries when appropriate. Furthermore, we recommend fitting various methods, comparing the results and model performance, and investigating any significant discrepancies among them.

Exploiting relationships between outcomes in Bayesian multivariate network meta-analysis with an application to relapsing-remitting multiple sclerosis

In multivariate network meta-analysis (NMA), the piecemeal nature of the evidence base means that there may be treatment-outcome combinations for which no data is available. Most existing multivariate evidence synthesis models are either unable to estimate the missing treatment-outcome combinations, or can only do so under particularly strong assumptions, such as perfect between-study correlations between outcomes or constant effect size across outcomes. Many existing implementations are also limited to two treatments or two outcomes, or rely on model specification that is heavily tailored to the dimensions of the dataset. We present a Bayesian multivariate NMA model that estimates the missing treatment-outcome combinations via mappings between the population mean effects, while allowing the study-specific effects to be imperfectly correlated. The method is designed for aggregate-level data (rather than individual patient data) and is likely to be useful when modeling multiple sparsely reported outcomes, or when varying definitions of the same underlying outcome are adopted by different studies. We implement the model via a novel decomposition of the treatment effect variance, which can be specified efficiently for an arbitrary dataset given some basic assumptions regarding the correlation structure. The method is illustrated using data concerning the efficacy and liver-related safety of eight active treatments for relapsing-remitting multiple sclerosis. The results indicate that fingolimod and interferon beta-1b are the most efficacious treatments but also have some of the worst effects on liver safety. Dimethyl fumarate and glatiramer acetate perform reasonably on all of the efficacy and safety outcomes in the model.

Comparison of Poly (ADP-ribose) Polymerase Inhibitors (PARPis) as Maintenance Therapy for Platinum-Sensitive Ovarian Cancer: Systematic Review and Network Meta-Analysis

Simple Summary

Poly (ADP-ribose) polymerase inhibitors (PARPis; inhibitors of a family of enzymes that are primary involved in DNA repair) are considered to be the drug of choice in maintenance therapy for platinum-sensitive ovarian cancer. However, despite the FDA approval of three such agents and their availability in clinical practice, thus far, no clinical trial investigated them in a head-to-head direct comparison. In this study, we used a statistical approach that allows comparing direct and indirect evidence (network meta-analysis) in order to compare the three FDA-approved PARPis (olaparib, niraparib and rucaparib). To this end, we used data from six randomized control trials involving a total of 2270 ovarian cancer patients. Interestingly, we found no significant differences in clinical outcomes (overall survival and progression-free survival) between the three agents. However, niraparib was found to be associated with higher risk of certain adverse events (thrombocytopenia, neutropenia, constipation, and headaches) compared to the other two PARPis.

Abstract

Background: Three PARPis (olaparib, niraparib and rucaparib) are currently FDA-approved as maintenance therapy in newly diagnosed and recurrent ovarian cancer. However, thus far, no trial has compared the three approved PARPis in the overall population, in patients with BRCA mutations, or in those with wild-type BRCA. Methods: A frequentist network meta-analysis was used for indirect comparisons between the different PARPis with respect to progression free survival (PFS), overall survival (OS), and adverse events. Results: Overall, six randomized clinical trials involving 2,770 patients, were included in the analysis. Results from the indirect comparisons revealed no statistically significant differences between the three PARPis with respect to PFS or OS in the entire population and in patients with mutated and wild-type BRCA, separately. Niraparib showed a statistically significant increased risk for grade 3 and 4 thrombocytopenia (risk-difference [RD] from placebo: 0.3; 95% confidence interval [CI], 0.27‒0.34) and any grade neutropenia (RD from placebo: 0.22; 95% CI, 0.18‒0.25) as compared with the other PARPis. Conclusion: No statistically significant difference was found between the three PARPis with respect to PFS or OS (overall and in subpopulations by BRCA status). There is, however, a statistical difference in toxicity as niraparib is associated with a greater risk for thrombocytopenia and neutropenia.

Antioxidant micronutrient supplementation in critically ill adults: A systematic review with meta-analysis and trial sequential analysis

BACKGROUND

The role of antioxidant micronutrient (AxM) supplementation in the critically ill patients has been controversial, and recent trials have suggested a tendency to harm. Therefore, we performed a systematic review with meta-analysis and trial sequential analysis (TSA) of randomized controlled trials (RCT) to examine the effect of AxM supplementation on clinical outcomes among critically ill adults.

METHODS

PubMed, EMBASE, Cochrane, CINAHL, LILACS, DARE, SCOPUS, and Web of sciences databases were searched from inception to March 2019. RCTs that compared AxM supplements with placebo in adult critically ill patients and reporting mortality as an outcomes were included. Trial quality was assessed using updated cochrane risk of bias (RoB-II) tool. Primary outcome was all-cause mortality. Secondary outcomes were 28-day mortality, intensive care unit (ICU) and hospital length of stay (LOS), ventilator days and infection between the two groups. Outcomes were summarised using random-effects estimators. Quality of evidence (QOE) was rated using Grading of Recommendations, Assessment, Development and Evaluation. Prior to final analysis, we repeated the search through September 2019. R version 3.6.2 and STATA version 13 were used for all statistical analyses.

RESULTS

Pooled analysis of 34 trials with 4678 patients revealed that AxM supplementation was associated with possible reduction in all-cause mortality (relative risk [RR], 0.89 [95%CI 0.79 to 0.99], TSA adjusted CI 0.77 to 1.03; Low QOE). Fragility index and number needed to treat were 1 and 41, respectively. Eight studies with low RoB (RR, 1.08; 95%CI 0.95 to 1.23; TSA CI, 0.64 to 1.82; moderate QOE) did not show mortality reduction with AxM supplementation.

SECONDARY OUTCOMES

ICU LOS (weighted mean difference [WMD], -0.84; 95%CI -1.50 to -0.18; moderate QOE), hospitalization days (WMD, -2.83; 95%CI -3.91to -1.75; low QOE) and ventilator days (WMD, -1.87; 95%CI -3.60 to -0.14; very low QOE) showed a statistically significant benefit with AxM supplementation. In meta-regression analysis, neither the duration of AxM therapy nor the dosage of selenium, which was the most widely studied AxM, reported an association with mortality.

CONCLUSION

Although AxM supplementation was associated with possible reduction in all-cause mortality, results from the TSA and studies with low RoB showing null effect suggest that the evidence of benefit is questionable. Secondary outcomes attained statistically significant benefit with AxM supplements, but the certainity of evidence was low. To summarize, current evidence does not justify administration of AxM in critically ill patients.

REGISTRATION

PROSPERO, CRD42019125898.

Clinical evidence for association of neoadjuvant chemotherapy or chemoradiotherapy with efficacy and safety in patients with resectable esophageal carcinoma (NewEC study)

BACKGROUND

The efficacy and safety of neoadjuvant treatment over surgery alone and that of neoadjuvant chemoradiotherapy (NCRT) over neoadjuvant chemotherapy (NCT) in resectable esophageal carcinoma remains inconclusive. This study (NewEC) used global data to comprehensively evaluate these comparisons and to provide a preferable strategy for patient subsets.

METHODS

This study included a meta-analysis of randomized controlled trials (RCTs) identified from inception to May 2019 from PubMed, EMBASE, the Cochrane Central Register of Controlled Trials, ClinicalTrials.gov, and congresses and a registry-based cohort study with patients from Massachusetts General Hospital (Massachusetts, USA) and Guangdong Provincial People's Hospital (Guangzhou, China) recruited from November 2000 and June 2017, to cross-validate the comparisons among NCRT versus NCT versus surgery. The GRADE approach was used to assessed quality of evidence in meta-analysis. Neural network machine learning propensity score-matched analysis was used to account for confounding by patient-level characteristics in the cohort study. The primary endpoint was overall survival (OS). The study was registered with PROSPERO CRD42017072242 and ClinicalTrials.gov NCT04027543.

FINDINGS

Of 22,070 studies assessed, there were 38 (n = 6,993 patients) eligible RCTs. Additionally, 423 out of 467 screened patients were included in the cohort study. The results from trials showed that NCT had a better OS than surgery alone (hazard ratio [HR] 0·88, 95% confidence interval [CI] 0·79-0·98; high quality) and was only favorable for adenocarcinoma (HR 0·83, 95% CI 0·72-0·96; moderate quality). High-quality evidence showed a significantly better OS for NCRT than surgery alone (HR 0·74, 95% CI 0·66-0·82) for both adenocarcinoma (HR 0·73, 95% CI 0·62-0·86) and squamous cell carcinoma (SCC) (HR 0·73, 95% CI 0·65-0·83). The OS benefit of NCRT over NCT was seen in the pairwise (HR 0·78, 95% CI 0·62-0·99; high quality) and network (HR 0·82, 95% CI 0·72-0·93; high quality) meta-analyses, with similar results before (HR 0·60, 95% CI 0·40-0·91) and after (HR 0·44, 95% CI 0·25-0·77) matching in the cohort study, leading to a significantly increased 5-year OS rate in both adenocarcinoma and SCC before and after matching. The increased benefits from NCT or NCRT were not associated with the risk of 30-day or in-hospital mortality.

INTERPRETATION

NewEC Study provided high-quality evidence supporting the survival benefits of NCRT or NCT over surgery alone, with NCRT presenting the greatest benefit for resectable esophageal carcinoma.

FUNDING

National Science and Technology Major Project, the National Natural Science Foundation of China, the Natural Science Foundation of Guangdong Province, the Guangzhou Science and Technology Major Program, the Medical artificial intelligence project of Sun Yat-Sen Memorial Hospital, the Guangdong Science and Technology Department, the Guangdong Province Medical Scientific Research Foundation, and Guangdong Provincial People's Hospital Intermural Program.

Bayesian inference for network meta-regression using multivariate random effects with applications to cholesterol lowering drugs

Low-density lipoprotein cholesterol (LDL-C) has been identified as a causative factor for atherosclerosis and related coronary heart disease, and as the main target for cholesterol- and lipid-lowering therapy. Statin drugs inhibit cholesterol synthesis in the liver and are typically the first line of therapy to lower elevated levels of LDL-C. On the other hand, a different drug, Ezetimibe, inhibits the absorption of cholesterol by the small intestine and provides a different mechanism of action. Many clinical trials have been carried out on safety and efficacy evaluation of cholesterol lowering drugs. To synthesize the results from different clinical trials, we examine treatment level (aggregate) network meta-data from 29 double-blind, randomized, active, or placebo-controlled statins +/$-$ Ezetimibe clinical trials on adult treatment-naïve patients with primary hypercholesterolemia. In this article, we propose a new approach to carry out Bayesian inference for arm-based network meta-regression. Specifically, we develop a new strategy of grouping the variances of random effects, in which we first formulate possible sets of the groups of the treatments based on their clinical mechanisms of action and then use Bayesian model comparison criteria to select the best set of groups. The proposed approach is especially useful when some treatment arms are involved in only a single trial. In addition, a Markov chain Monte Carlo sampling algorithm is developed to carry out the posterior computations. In particular, the correlation matrix is generated from its full conditional distribution via partial correlations. The proposed methodology is further applied to analyze the network meta-data from 29 trials with 11 treatment arms.

GRADE approach to rate the certainty from a network meta-analysis: avoiding spurious judgments of imprecision in sparse networks

When direct and indirect estimates of treatment effects are coherent, network meta-analysis (NMA) estimates should have increased precision (narrower confidence or credible intervals compared with relying on direct estimates alone), a benefit of NMA. We have, however, observed cases of sparse networks in which combining direct and indirect estimates results in marked widening of the confidence intervals. In many cases, the assumption of common between-study heterogeneity across the network seems to be responsible for this counterintuitive result. Although the assumption of common between-study heterogeneity across paired comparisons may, in many cases, not be appropriate, it is required to ensure the feasibility of estimating NMA treatment effects. This is especially the case in sparse networks, in which data are insufficient to reliably estimate different variances across the network. The result, however, may be spuriously wide confidence intervals for some of the comparisons in the network (and, in the Grading of Recommendations Assessment, Development, and Evaluation approach, inappropriately low ratings of the certainty of the evidence through rating down for serious imprecision). Systematic reviewers should be aware of the problem and plan sensitivity analyses that produce intuitively sensible confidence intervals. These sensitivity analyses may include using informative priors for the between-study heterogeneity parameter in the Bayesian framework and the use of fixed effects models.

Comparative effectiveness of pharmacologic interventions to prevent shivering after surgery: a network meta-analysis

INTRODUCTION

Although many drugs have been studied to prevent postoperative shivering, their comparative effectiveness is unknown. We attempted to assess the comparative effectiveness of the pharmacologic strategies to prevent shivering after surgery including intravenous (IV) and intrathecal (IT) meperidine, IV and IT dexmedetomidine, IV and IT clonidine, nefopam, tramadol, ketamine, and serotonin receptor antagonists.

EVIDENCE ACQUISITION

We implemented a frequentist network meta-analysis of randomized trials. Random effect model was selected according to deviance information criteria. The incidence of moderate or severe shivering and the overall incidence of postoperative shivering were compared in all patients and subgroups of general and spinal anesthesia.

EVIDENCE SYNTHESIS

Seventy trials with 6062 participants were included. All interventions except clonidine IT and dexmedetomidine IT significantly decreased the incidence of moderate or severe shivering. All interventions except clonidine IT significantly decreased the overall incidence of shivering. Nefopam IV was ranked best regarding both the incidence of moderate or severe shivering (odds ratio [OR] 0.04, 95% confidence interval [CI] 0.04-0.25, SUCRA 86.2, moderate quality-evidence), and the overall incidence of shivering (OR 0.07, 95% CI 0.03 to 0.15, SUCRA 85.7, moderate quality-evidence), which was consistent in the subgroups of anesthesia.

CONCLUSIONS

Nefopam was ranked best regarding both severity and incidence of postoperative shivering. In addition to nefopam, tramadol, meperidine IV and IT, and dexmedetomidine IV were ranked high. However, there was significant heterogeneity regarding the individual drug regimen and surgery type, precluding firm conclusion. Further randomized trials are required to compare the efficacy of the drugs with high rank.

Motion analysis of non-model organisms using a hierarchical model: Influence of setup enclosure dimensions on gait parameters of Swinhoe's striped squirrels as a test case

In in-vivo motion analyses, data from a limited number of subjects and trials is used as proxy for locomotion properties of entire populations, yet the inherent hierarchy of the individual and population level is usually not accounted for. Despite the increasing availability of hierarchical model frameworks for statistical analyses, they have not been applied extensively to comparative motion analysis. As a case study for the use of hierarchical models, we analyzed locomotor parameters of four Swinhoe's striped squirrels. The small-bodied arboreal mammals exhibit brief bouts of rapid asymmetric gaits. Spatio-temporal parameters on runways with experimentally varied dimensions of the setup enclosure were compared to test for their potentially confounding effects. We applied principal component analysis to evaluate changes to the overall locomotor pattern. A common, non-hierarchical, pooled statistical analysis of the data revealed significant differences in some of the parameters depending on enclosure dimensions. In contrast, we used a hierarchical Bayesian generalized linear model (GLM) that considers subject specific differences and population effects to compare the effect of enclosure dimensions on the measured parameters and the principal components. None of the population effects were confirmed by the hierarchical GLM. The confounding effect of a single subject that deviates in its locomotor behavior is potentially bigger than the influence of the experimental variation in enclosure dimensions. Our findings justify the common practice of researchers to intuitively select an enclosure with dimensions assumed as "non-constraining". Hierarchical models can easily be designed to cope with limited sample size and bias introduced by deviating behavior of individuals. When limited data is available-a typical restriction of in-vivo motion analyses of non-model organisms-density distributions of the Bayesian GLM used here remain reliable and the hierarchical structure of the model optimally exploits all available information. We provide code to be adjusted to other research questions.

Systematic reviews and cancer research: a suggested stepwise approach

Systematic reviews, with or without meta-analysis, play an important role today in synthesizing cancer research and are frequently used to guide decision-making. However, there is now an increase in the number of systematic reviews on the same topic, thereby necessitating a systematic review of previous systematic reviews. With a focus on cancer, the purpose of this article is to provide a practical, stepwise approach for systematically reviewing the literature and publishing the results. This starts with the registration of a protocol for a systematic review of previous systematic reviews and ends with the publication of an original or updated systematic review, with or without meta-analysis, in a peer-reviewed journal. Future directions as well as potential limitations of the approach are also discussed. It is hoped that the stepwise approach presented in this article will be helpful to both producers and consumers of cancer-related systematic reviews and will contribute to the ultimate goal of preventing and treating cancer.

Systemic treatments for metastatic cutaneous melanoma

BACKGROUND

The prognosis of people with metastatic cutaneous melanoma, a skin cancer, is generally poor. Recently, new classes of drugs (e.g. immune checkpoint inhibitors and small-molecule targeted drugs) have significantly improved patient prognosis, which has drastically changed the landscape of melanoma therapeutic management. This is an update of a Cochrane Review published in 2000.

OBJECTIVES

To assess the beneficial and harmful effects of systemic treatments for metastatic cutaneous melanoma.

SEARCH METHODS

We searched the following databases up to October 2017: the Cochrane Skin Group Specialised Register, CENTRAL, MEDLINE, Embase and LILACS. We also searched five trials registers and the ASCO database in February 2017, and checked the reference lists of included studies for further references to relevant randomised controlled trials (RCTs).

SELECTION CRITERIA

We considered RCTs of systemic therapies for people with unresectable lymph node metastasis and distant metastatic cutaneous melanoma compared to any other treatment. We checked the reference lists of selected articles to identify further references to relevant trials.

DATA COLLECTION AND ANALYSIS

Two review authors extracted data, and a third review author independently verified extracted data. We implemented a network meta-analysis approach to make indirect comparisons and rank treatments according to their effectiveness (as measured by the impact on survival) and harm (as measured by occurrence of high-grade toxicity). The same two review authors independently assessed the risk of bias of eligible studies according to Cochrane standards and assessed evidence quality based on the GRADE criteria.

MAIN RESULTS

We included 122 RCTs (28,561 participants). Of these, 83 RCTs, encompassing 21 different comparisons, were included in meta-analyses. Included participants were men and women with a mean age of 57.5 years who were recruited from hospital settings. Twenty-nine studies included people whose cancer had spread to their brains. Interventions were categorised into five groups: conventional chemotherapy (including single agent and polychemotherapy), biochemotherapy (combining chemotherapy with cytokines such as interleukin-2 and interferon-alpha), immune checkpoint inhibitors (such as anti-CTLA4 and anti-PD1 monoclonal antibodies), small-molecule targeted drugs used for melanomas with specific gene changes (such as BRAF inhibitors and MEK inhibitors), and other agents (such as anti-angiogenic drugs). Most interventions were compared with chemotherapy. In many cases, trials were sponsored by pharmaceutical companies producing the tested drug: this was especially true for new classes of drugs, such as immune checkpoint inhibitors and small-molecule targeted drugs.When compared to single agent chemotherapy, the combination of multiple chemotherapeutic agents (polychemotherapy) did not translate into significantly better survival (overall survival: HR 0.99, 95% CI 0.85 to 1.16, 6 studies, 594 participants; high-quality evidence; progression-free survival: HR 1.07, 95% CI 0.91 to 1.25, 5 studies, 398 participants; high-quality evidence. Those who received combined treatment are probably burdened by higher toxicity rates (RR 1.97, 95% CI 1.44 to 2.71, 3 studies, 390 participants; moderate-quality evidence). (We defined toxicity as the occurrence of grade 3 (G3) or higher adverse events according to the World Health Organization scale.)Compared to chemotherapy, biochemotherapy (chemotherapy combined with both interferon-alpha and interleukin-2) improved progression-free survival (HR 0.90, 95% CI 0.83 to 0.99, 6 studies, 964 participants; high-quality evidence), but did not significantly improve overall survival (HR 0.94, 95% CI 0.84 to 1.06, 7 studies, 1317 participants; high-quality evidence). Biochemotherapy had higher toxicity rates (RR 1.35, 95% CI 1.14 to 1.61, 2 studies, 631 participants; high-quality evidence).With regard to immune checkpoint inhibitors, anti-CTLA4 monoclonal antibodies plus chemotherapy probably increased the chance of progression-free survival compared to chemotherapy alone (HR 0.76, 95% CI 0.63 to 0.92, 1 study, 502 participants; moderate-quality evidence), but may not significantly improve overall survival (HR 0.81, 95% CI 0.65 to 1.01, 2 studies, 1157 participants; low-quality evidence). Compared to chemotherapy alone, anti-CTLA4 monoclonal antibodies is likely to be associated with higher toxicity rates (RR 1.69, 95% CI 1.19 to 2.42, 2 studies, 1142 participants; moderate-quality evidence).Compared to chemotherapy, anti-PD1 monoclonal antibodies (immune checkpoint inhibitors) improved overall survival (HR 0.42, 95% CI 0.37 to 0.48, 1 study, 418 participants; high-quality evidence) and probably improved progression-free survival (HR 0.49, 95% CI 0.39 to 0.61, 2 studies, 957 participants; moderate-quality evidence). Anti-PD1 monoclonal antibodies may also result in less toxicity than chemotherapy (RR 0.55, 95% CI 0.31 to 0.97, 3 studies, 1360 participants; low-quality evidence).Anti-PD1 monoclonal antibodies performed better than anti-CTLA4 monoclonal antibodies in terms of overall survival (HR 0.63, 95% CI 0.60 to 0.66, 1 study, 764 participants; high-quality evidence) and progression-free survival (HR 0.54, 95% CI 0.50 to 0.60, 2 studies, 1465 participants; high-quality evidence). Anti-PD1 monoclonal antibodies may result in better toxicity outcomes than anti-CTLA4 monoclonal antibodies (RR 0.70, 95% CI 0.54 to 0.91, 2 studies, 1465 participants; low-quality evidence).Compared to anti-CTLA4 monoclonal antibodies alone, the combination of anti-CTLA4 plus anti-PD1 monoclonal antibodies was associated with better progression-free survival (HR 0.40, 95% CI 0.35 to 0.46, 2 studies, 738 participants; high-quality evidence). There may be no significant difference in toxicity outcomes (RR 1.57, 95% CI 0.85 to 2.92, 2 studies, 764 participants; low-quality evidence) (no data for overall survival were available).The class of small-molecule targeted drugs, BRAF inhibitors (which are active exclusively against BRAF-mutated melanoma), performed better than chemotherapy in terms of overall survival (HR 0.40, 95% CI 0.28 to 0.57, 2 studies, 925 participants; high-quality evidence) and progression-free survival (HR 0.27, 95% CI 0.21 to 0.34, 2 studies, 925 participants; high-quality evidence), and there may be no significant difference in toxicity (RR 1.27, 95% CI 0.48 to 3.33, 2 studies, 408 participants; low-quality evidence).Compared to chemotherapy, MEK inhibitors (which are active exclusively against BRAF-mutated melanoma) may not significantly improve overall survival (HR 0.85, 95% CI 0.58 to 1.25, 3 studies, 496 participants; low-quality evidence), but they probably lead to better progression-free survival (HR 0.58, 95% CI 0.42 to 0.80, 3 studies, 496 participants; moderate-quality evidence). However, MEK inhibitors probably have higher toxicity rates (RR 1.61, 95% CI 1.08 to 2.41, 1 study, 91 participants; moderate-quality evidence).Compared to BRAF inhibitors, the combination of BRAF plus MEK inhibitors was associated with better overall survival (HR 0.70, 95% CI 0.59 to 0.82, 4 studies, 1784 participants; high-quality evidence). BRAF plus MEK inhibitors was also probably better in terms of progression-free survival (HR 0.56, 95% CI 0.44 to 0.71, 4 studies, 1784 participants; moderate-quality evidence), and there appears likely to be no significant difference in toxicity (RR 1.01, 95% CI 0.85 to 1.20, 4 studies, 1774 participants; moderate-quality evidence).Compared to chemotherapy, the combination of chemotherapy plus anti-angiogenic drugs was probably associated with better overall survival (HR 0.60, 95% CI 0.45 to 0.81; moderate-quality evidence) and progression-free survival (HR 0.69, 95% CI 0.52 to 0.92; moderate-quality evidence). There may be no difference in terms of toxicity (RR 0.68, 95% CI 0.09 to 5.32; low-quality evidence). All results for this comparison were based on 324 participants from 2 studies.Network meta-analysis focused on chemotherapy as the common comparator and currently approved treatments for which high- to moderate-quality evidence of efficacy (as represented by treatment effect on progression-free survival) was available (based on the above results) for: biochemotherapy (with both interferon-alpha and interleukin-2); anti-CTLA4 monoclonal antibodies; anti-PD1 monoclonal antibodies; anti-CTLA4 plus anti-PD1 monoclonal antibodies; BRAF inhibitors; MEK inhibitors, and BRAF plus MEK inhibitors. Analysis (which included 19 RCTs and 7632 participants) generated 21 indirect comparisons.The best evidence (moderate-quality evidence) for progression-free survival was found for the following indirect comparisons:• both combinations of immune checkpoint inhibitors (HR 0.30, 95% CI 0.17 to 0.51) and small-molecule targeted drugs (HR 0.17, 95% CI 0.11 to 0.26) probably improved progression-free survival compared to chemotherapy;• both BRAF inhibitors (HR 0.40, 95% CI 0.23 to 0.68) and combinations of small-molecule targeted drugs (HR 0.22, 95% CI 0.12 to 0.39) were probably associated with better progression-free survival compared to anti-CTLA4 monoclonal antibodies;• biochemotherapy (HR 2.81, 95% CI 1.76 to 4.51) probably lead to worse progression-free survival compared to BRAF inhibitors;• the combination of small-molecule targeted drugs probably improved progression-free survival (HR 0.38, 95% CI 0.21 to 0.68) compared to anti-PD1 monoclonal antibodies;• both biochemotherapy (HR 5.05, 95% CI 3.01 to 8.45) and MEK inhibitors (HR 3.16, 95% CI 1.77 to 5.65) were probably associated with worse progression-free survival compared to the combination of small-molecule targeted drugs; and• biochemotherapy was probably associated with worse progression-free survival (HR 2.81, 95% CI 1.54 to 5.11) compared to the combination of immune checkpoint inhibitors.The best evidence (moderate-quality evidence) for toxicity was found for the following indirect comparisons:• combination of immune checkpoint inhibitors (RR 3.49, 95% CI 2.12 to 5.77) probably increased toxicity compared to chemotherapy;• combination of immune checkpoint inhibitors probably increased toxicity (RR 2.50, 95% CI 1.20 to 5.20) compared to BRAF inhibitors;• the combination of immune checkpoint inhibitors probably increased toxicity (RR 3.83, 95% CI 2.59 to 5.68) compared to anti-PD1 monoclonal antibodies; and• biochemotherapy was probably associated with lower toxicity (RR 0.41, 95% CI 0.24 to 0.71) compared to the combination of immune checkpoint inhibitors.Network meta-analysis-based ranking suggested that the combination of BRAF plus MEK inhibitors is the most effective strategy in terms of progression-free survival, whereas anti-PD1 monoclonal antibodies are associated with the lowest toxicity.Overall, the risk of bias of the included trials can be considered as limited. When considering the 122 trials included in this review and the seven types of bias we assessed, we performed 854 evaluations only seven of which (< 1%) assigned high risk to six trials.

AUTHORS' CONCLUSIONS

We found high-quality evidence that many treatments offer better efficacy than chemotherapy, especially recently implemented treatments, such as small-molecule targeted drugs, which are used to treat melanoma with specific gene mutations. Compared with chemotherapy, biochemotherapy (in this case, chemotherapy combined with both interferon-alpha and interleukin-2) and BRAF inhibitors improved progression-free survival; BRAF inhibitors (for BRAF-mutated melanoma) and anti-PD1 monoclonal antibodies improved overall survival. However, there was no difference between polychemotherapy and monochemotherapy in terms of achieving progression-free survival and overall survival. Biochemotherapy did not significantly improve overall survival and has higher toxicity rates compared with chemotherapy.There was some evidence that combined treatments worked better than single treatments: anti-PD1 monoclonal antibodies, alone or with anti-CTLA4, improved progression-free survival compared with anti-CTLA4 monoclonal antibodies alone. Anti-PD1 monoclonal antibodies performed better than anti-CTLA4 monoclonal antibodies in terms of overall survival, and a combination of BRAF plus MEK inhibitors was associated with better overall survival for BRAF-mutated melanoma, compared to BRAF inhibitors alone.The combination of BRAF plus MEK inhibitors (which can only be administered to people with BRAF-mutated melanoma) appeared to be the most effective treatment (based on results for progression-free survival), whereas anti-PD1 monoclonal antibodies appeared to be the least toxic, and most acceptable, treatment.Evidence quality was reduced due to imprecision, between-study heterogeneity, and substandard reporting of trials. Future research should ensure that those diminishing influences are addressed. Clinical areas of future investigation should include the longer-term effect of new therapeutic agents (i.e. immune checkpoint inhibitors and targeted therapies) on overall survival, as well as the combination of drugs used in melanoma treatment; research should also investigate the potential influence of biomarkers.

Multivariate and network meta-analysis of multiple outcomes and multiple treatments: rationale, concepts, and examples

Organisations such as the National Institute for Health and Care Excellence require the synthesis of evidence from existing studies to inform their decisions—for example, about the best available treatments with respect to multiple efficacy and safety outcomes. However, relevant studies may not provide direct evidence about all the treatments or outcomes of interest. Multivariate and network meta-analysis methods provide a framework to address this, using correlated or indirect evidence from such studies alongside any direct evidence. In this article, the authors describe the key concepts and assumptions of these methods, outline how correlated and indirect evidence arises, and illustrate the contribution of such evidence in real clinical examples involving multiple outcomes and multiple treatments

Comparison of anti-malarial drugs efficacy in the treatment of uncomplicated malaria in African children and adults using network meta-analysis

Background

Artemisinin-based combination therapy (ACT) and novel drug combinations are available and used in African countries to treat uncomplicated malaria. Network meta-analysis methods are rarely and poorly applied for the comparison of their efficacies. This method was applied on a set of randomized controlled trials to illustrate its usefulness.

Methods

A literature review available in Pubmed was conducted in July 2016. Eligible studies, conducted in sub-Saharan Africa, published between 2002 and 2016, focused on randomized controlled trials of at least two artemisinin-based combinations to treat uncomplicated malaria in children and adults. Agglomerate data were: the number of PCR-corrected adequate clinical and parasitological response (ACPR) on day 28, used as the primary endpoint in all interventions, the number of participants and the list of treatments. A Bayesian random effect meta-analysis using a binary outcome was the method to compare the efficacy. Ranking measure was used to obtain a hierarchy of the competing interventions.

Results

In total, 76 articles were included; 13 treatment regimens were involved and tested in 36,001 patients. Using artemether–lumefantrine (AL) as the common comparator for the entire network, 12 relative treatment effects were estimated and indirect comparisons were obtained. Dihydroartemisinin–piperaquine (DHAP) was shown to be more effective than AL (odds ratio [OR] = 1.92; 95% CI 1.30–2.82; 19,163 patients), ASAQ (OR = 1.70; 95% CI 1.10–2.64; 14,433 patients), and amodiaquine–sulfadoxine–pyrimethamine (AQSP): OR = 2.20; 95% CI 1.21–3.96; 8863 patients. Artesunate–amodiaquine (ASAQ) was comparable to AL (OR = 1.11; 95% CI 0.84–1.45; 21,235 patients). No significant difference was found between artesunate and mefloquine (ASMQ) and AL (OR = 1.20; 95% CI = 0.52-2.8; 13,824 participants). According to treatment ranking, among the WHO-recommended ACT medicines, DHAP was shown to be the most efficacious.

Conclusions

Based on the available evidence, this study demonstrated the superiority of DHAP among currently recommended artemisinin-based combinations. The application of the methods described here may be helpful to gain better understanding of treatment efficacy and improve future decisions. However, more data are needed to allow robust conclusions about the results in comparison with novel drugs. Further surveillance of the efficacy of anti-malarial drugs and clinical trials are needed to closely follow the evolution of the epidemiology of drug-resistant malaria in Africa.

Electronic supplementary material

The online version of this article (doi:10.1186/s12936-017-1963-0) contains supplementary material, which is available to authorized users.

Using structural equation modeling for network meta-analysis

BACKGROUND

Network meta-analysis overcomes the limitations of traditional pair-wise meta-analysis by incorporating all available evidence into a general statistical framework for simultaneous comparisons of several treatments. Currently, network meta-analyses are undertaken either within the Bayesian hierarchical linear models or frequentist generalized linear mixed models. Structural equation modeling (SEM) is a statistical method originally developed for modeling causal relations among observed and latent variables. As random effect is explicitly modeled as a latent variable in SEM, it is very flexible for analysts to specify complex random effect structure and to make linear and nonlinear constraints on parameters. The aim of this article is to show how to undertake a network meta-analysis within the statistical framework of SEM.

METHODS

We used an example dataset to demonstrate the standard fixed and random effect network meta-analysis models can be easily implemented in SEM. It contains results of 26 studies that directly compared three treatment groups A, B and C for prevention of first bleeding in patients with liver cirrhosis. We also showed that a new approach to network meta-analysis based on the technique of unrestricted weighted least squares (UWLS) method can also be undertaken using SEM.

RESULTS

For both the fixed and random effect network meta-analysis, SEM yielded similar coefficients and confidence intervals to those reported in the previous literature. The point estimates of two UWLS models were identical to those in the fixed effect model but the confidence intervals were greater. This is consistent with results from the traditional pairwise meta-analyses. Comparing to UWLS model with common variance adjusted factor, UWLS model with unique variance adjusted factor has greater confidence intervals when the heterogeneity was larger in the pairwise comparison. The UWLS model with unique variance adjusted factor reflects the difference in heterogeneity within each comparison.

CONCLUSION

SEM provides a very flexible framework for univariate and multivariate meta-analysis, and its potential as a powerful tool for advanced meta-analysis is still to be explored.

An investigation of the impact of using different methods for network meta-analysis: a protocol for an empirical evaluation

BACKGROUND

Network meta-analysis, a method to synthesise evidence from multiple treatments, has increased in popularity in the past decade. Two broad approaches are available to synthesise data across networks, namely, arm- and contrast-synthesis models, with a range of models that can be fitted within each. There has been recent debate about the validity of the arm-synthesis models, but to date, there has been limited empirical evaluation comparing results using the methods applied to a large number of networks. We aim to address this gap through the re-analysis of a large cohort of published networks of interventions using a range of network meta-analysis methods.

METHODS

We will include a subset of networks from a database of network meta-analyses of randomised trials that have been identified and curated from the published literature. The subset of networks will include those where the primary outcome is binary, the number of events and participants are reported for each direct comparison, and there is no evidence of inconsistency in the network. We will re-analyse the networks using three contrast-synthesis methods and two arm-synthesis methods. We will compare the estimated treatment effects, their standard errors, treatment hierarchy based on the surface under the cumulative ranking (SUCRA) curve, the SUCRA value, and the between-trial heterogeneity variance across the network meta-analysis methods. We will investigate whether differences in the results are affected by network characteristics and baseline risk.

DISCUSSION

The results of this study will inform whether, in practice, the choice of network meta-analysis method matters, and if it does, in what situations differences in the results between methods might arise. The results from this research might also inform future simulation studies.

Survival After Neoadjuvant and Adjuvant Treatments Compared to Surgery Alone for Resectable Esophageal Carcinoma: A Network Meta-analysis

OBJECTIVE

This network meta-analysis compared overall survival after neoadjuvant or adjuvant chemotherapy (CT), radiotherapy (RT), or combinations of both (chemoradiotherapy, CRT) or surgery alone to identify the most effective approach.

SUMMARY BACKGROUND DATA

The optimal treatment for resectable esophageal cancer is unknown.

METHODS

A search for randomized controlled trials reporting on neoadjuvant and adjuvant therapies was conducted. Using a network meta-analysis, treatments were ranked based on their effectiveness for improving survival.

RESULTS

In 33 eligible randomized controlled trials, 6072 patients were randomized to receive either surgery alone (N = 2459) or neoadjuvant CT (N = 1332), RT (N = 58), and CRT (N = 1196) followed by surgery or surgery followed by adjuvant CT (N = 542), RT (N = 383), and CRT (N = 102). Twenty-one comparisons were generated. Neoadjuvant CRT followed by surgery compared with surgery alone was the only treatment to significantly improve survival [hazard ratio (HR) = 0.77, 95% confidence interval (CI): 0.68-0.87]. When trials were grouped considering neoadjuvant and adjuvant therapies and surgery alone, neoadjuvant therapies combined with surgery compared with surgery alone showed a survival advantage (HR = 0.83, 95% CI 0.76-0.90), whereas surgery along with adjuvant therapies showed no significant survival advantage (HR = 0.87, 95% CI 0.67-1.14). A subgroup analysis of neoadjuvant therapies showed a superior effectiveness of neoadjuvant CRT and surgery compared with surgery alone (HR = 0.77, 95% CI 0.68-0.87).

CONCLUSIONS

This network meta-analysis showed neoadjuvant CRT followed by surgery to be the most effective strategy in improving survival of resectable esophageal cancer. Resources should be focused on developing the most effective neoadjuvant CRT regimens for both adenocarcinomas and squamous cell carcinomas of the esophagus.

Diverse criteria and methods are used to compare treatment effect estimates: a scoping review

OBJECTIVES

To determine what criteria researchers use to assess whether the estimates of effect of an intervention on a dichotomous outcome are different when obtained using different study designs.

STUDY DESIGN AND SETTING

Scoping review of the literature. We included studies of dichotomous outcomes in which authors compared the estimates of effects from different study designs. We performed searches in electronic databases and in the list of references of relevant studies. Two reviewers independently selected studies and abstracted data. We created a list of the criteria used to compare estimates of effects between study designs, described their main features, and classified them using a clinical perspective.

RESULTS

We included 26 studies, from which we identified 24 criteria. Most of the studies focused on comparing estimates from observational studies and randomized controlled trials (n = 19). The most common criteria aimed to determine whether there was a difference or not (n = 18), provided guidance for such a judgment (n = 16), and were based on the point estimates (n = 11). We judged 14 criteria to be appropriate and classified them as either statistically related or clinically related.

CONCLUSION

We found that diverse criteria are used to compare effect estimates between study designs. Familiarity with these would aid in the interpretation of results from different studies regarding the same question.

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.

Breast Cancer Chemoprevention: A Network Meta-Analysis of Randomized Controlled Trials

BACKGROUND

Several agents have been advocated for breast cancer primary prevention. However, few of them appear effective, the associated severe adverse effects limiting their uptake.

METHODS

We performed a comprehensive search for randomized controlled trials (RCTs) reporting on the ability of chemoprevention agents (CPAs) to reduce the incidence of primary breast carcinoma. Using network meta-analysis, we ranked CPAs based simultaneously on efficacy and acceptability (an inverse measure of toxicity). All statistical tests were two-sided.

RESULTS

We found 48 eligible RCTs, enrolling 271 161 women randomly assigned to receive either placebo or one of 21 CPAs. Aromatase inhibitors (anastrozole and exemestane, considered a single CPA class because of the lack of between-study heterogeneity; relative risk [RR] = 0.468, 95% confidence interval [CI] = 0.346 to 0.634), arzoxifene (RR = 0.415, 95% CI = 0.253 to 0.682), lasofoxifene (RR = 0.208, 95% CI = 0.079 to 0.544), raloxifene (RR = 0.572, 95% CI = 0.372 to 0.881), tamoxifen (RR = 0.708, 95% CI = 0.595 to 0.842), and tibolone (RR = 0.317, 95% CI = 0.127 to 0.792) were statistically significantly associated with a therapeutic effect, which was restricted to estrogen receptor-positive tumors of postmenopausal women (except for tamoxifen, which is active also during premenopause). Network meta-analysis ranking showed that the new selective estrogen receptor modulators (SERMs) arzoxifene, lasofoxifene, and raloxifene have the best benefit-risk ratio. Aromatase inhibitors and tamoxifen ranked second and third, respectively.

CONCLUSIONS

These results provide physicians and health care regulatory agencies with RCT-based evidence on efficacy and acceptability of currently available breast cancer CPAs; at the same time, we pinpoint how much work still remains to be done before pharmacological primary prevention becomes a routine option to reduce the burden of this disease.

A Bayesian missing data framework for generalized multiple outcome mixed treatment comparisons

Bayesian statistical approaches to mixed treatment comparisons (MTCs) are becoming more popular because of their flexibility and interpretability. Many randomized clinical trials report multiple outcomes with possible inherent correlations. Moreover, MTC data are typically sparse (although richer than standard meta-analysis, comparing only two treatments), and researchers often choose study arms based upon which treatments emerge as superior in previous trials. In this paper, we summarize existing hierarchical Bayesian methods for MTCs with a single outcome and introduce novel Bayesian approaches for multiple outcomes simultaneously, rather than in separate MTC analyses. We do this by incorporating partially observed data and its correlation structure between outcomes through contrast-based and arm-based parameterizations that consider any unobserved treatment arms as missing data to be imputed. We also extend the model to apply to all types of generalized linear model outcomes, such as count or continuous responses. We offer a simulation study under various missingness mechanisms (e.g., missing completely at random, missing at random, and missing not at random) providing evidence that our models outperform existing models in terms of bias, mean squared error, and coverage probability then illustrate our methods with a real MTC dataset. We close with a discussion of our results, several contentious issues in MTC analysis, and a few avenues for future methodological development.

A multilevel approach to network meta-analysis within a frequentist framework

Meta-analysis is a powerful tool to summarize knowledge. Pairwise or network meta-analysis may be carried out with multivariate models that account for the dependence between treatment estimates and quantify the correlation across studies. From a different perspective, meta-analysis may be viewed as a special case of multilevel analysis having a hierarchical data structure. Hence, we introduce an alternative frequentist approach, called multilevel network meta-analysis, which also allows to account for publication bias and the presence of inconsistency. We propose our approach for a three-level data structure set-up: arms within studies at the first level, studies within study designs at the second level and design configuration at the third level. This strategy differs from the traditional frequentist modeling because it works directly on an arm-based data structure. An advantage of using multilevel analysis is its flexibility, since it naturally allows to add further levels to the model and to accommodate for multiple outcome variables. Moreover, multilevel modeling may be carried out with widely available statistical programs. Finally, we compare the results from our approach with those from a Bayesian network meta-analysis on a binary endpoint which examines the effect on mortality of some anesthetics at the longest follow-up available. In addition, we compare results from the Bayesian and multilevel network meta-analysis approaches on a publicly available "Thrombolytic drugs" database. We also provide the reader with a blueprint of SAS codes for fitting the proposed models, although our approach does not rely on any specific software.