Efficacy, safety, and tolerability of antimicrobial agents for complicated intra- abdominal infection: a systematic review and network meta-analysis

doi:10.21203/rs.3.rs-2431222/v1

Download PDF

Research Article

Efficacy, safety, and tolerability of antimicrobial agents for complicated intra- abdominal infection: a systematic review and network meta-analysis

https://doi.org/10.21203/rs.3.rs-2431222/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 21 Apr, 2023

Read the published version in BMC Infectious Diseases →

You are reading this latest preprint version

Background

Which antimicrobial agents provide the optimal efficacy, safety, and tolerability for the empirical treatment of complicated intra-abdominal infection (cIAI) remains unclear but is paramount in the context of evolving antimicrobial resistance. Therefore, updated meta-analyses on this issue are warranted.

Methods

We systematically searched 4 major electronic databases from their inception through October 2022. Randomized controlled trials examining antimicrobial agents for cIAIs treatment were included. Two reviewers independently assessed the quality of included studies utilizing the Cochrane Collaboration’s risk of bias tool as described in Cochrane Collaboration Handbook and extracted data from all manuscripts according to a predetermined list of interests. All meta-analyses were conducted using R software. The primary outcome was clinical success rate in patients with cIAI.

Results

Forty-five active-controlled trials with low to medium methodological quality and involving 14,267 adults with cIAI were included in network meta-analyses. The vast majority of patients with an APACHE II score < 10 had low risk of treatment failure or death. Twenty-one regimens were investigated. In the network meta-analyses, cephalosporin-based regimens were more effective and better tolerated than tigecycline with regard to clinical success rate (odds ratio [OR] = 1.3, 95% confidence interval [CI] 1.00 ~ 1.70) and discontinuation due to adverse events (OR = 0.69, 95% CI 0.44 ~ 1.00). No statistically significant differences were found among antimicrobial agents regarding microbiological success rates and all-cause mortality. A statistically significant trend was observed favoring carbapenems, which exhibited fewer discontinuations because of adverse events when compared with tigecycline (OR = 0.66, 95% CI 0.42 ~ 1.00). In pairwise meta-analyses, compared with tigecycline, carbapenems (2,315 participants, OR = 1.32, 95% CI 1.00 ~ 1.75, I² = 24.0%) were associated with superior clinical success rate. These results were of borderline statistical significance. Compared with tigecycline, carbapenems (2,304 participants, OR = 0.67, 95% CI 0.45 ~ 0.91; I² = 48.5%) were associated with significantly reduced risk of serious adverse events. According to the surface under the cumulative ranking curve analysis, cephalosporin-based regimens were more likely to be optimal among all treatments in terms of efficacy, and tigecycline was more likely to be worst regimen in terms of safety and tolerability.

Conclusion

This study suggests that cephalosporin-based regimens are optimal for the empiric treatment of patients with cIAIs and that tigecycline should be prescribed cautiously considering safety and tolerability concerns. However, it should be noted that data currently available for antimicrobial agent’ effectiveness, safety, and tolerability pertain mostly to lower-risk patients with cIAIs.

complicated intra-abdominal infection

antimicrobial agents

systematic review

network meta-analysis

efficacy

safety

tolerability

Intra-abdominal infections (IAIs) are a common problem in clinical practice and pose a major challenge for clinicians due to high morbidity and mortality rates. Traditionally, IAIs are classified as uncomplicated intra-abdominal infections and complicated intra-abdominal infections (cIAIs) (1). cIAIs are defined as infections that extend into a normally sterile area of the abdomen, such as gastroduodenal perforations, and commonly required a source control procedure (1–3). Successful management of these infections relies on timely and appropriate source procedures and appropriate empirical antimicrobial therapy. Source control procedures (e.g., percutaneous drainage) play a fundamental role by removing infected intra-abdominal fluid and tissue (1). Empiric antimicrobial therapy, albeit as a supplement to source control, is important in the overall treatment plan. Compared with proper selection of empiric antibiotics, inappropriate empiric antibiotic therapy is associated with higher risk of clinical failure, increased hospital length of stay, and higher medical costs (4, 5), thus highlighting the importance of selecting the optimal empirical therapy regimen.

cIAIs tend to be polymicrobial infections, which are generally caused by Gram-negative Enterobacteriaceae, Gram-positive cocci, and obligate anaerobes (1). Of these micro-organisms, Escherichia coli or Klebsiella pneumoniae, Streptococci and Bacteroides fragilis are predominantly isolated from intra-abdominal fluid and tissue cultures. The wide variety of microorganisms causing cIAIs has resulted in the use of broad-spectrum antimicrobial agents for empirical treatment (1, 3). Many antimicrobial agents are potential choices for treating cIAIs. In addition to older antimicrobials, several novel agents, such as eravacycline, ceftazidime/avibactam, and ceftolozane/tazobactam, has been approved for the treatment of patients with cIAIs (6, 7). Pairwise meta-analyses (8–10) have provided limited evidence suggesting optimal regimens for empirical therapy due to limitations associated with comparing only two interventions at a time. Given changes in patient’ characteristics and resistance over time in conjunction with the large number of antimicrobials available, proper selection of an empiric antimicrobial agent for cIAIs is challenging. In addition, inappropriate or overuse of antimicrobial therapy is associated with an increased risk of resistant pathogen emergence, higher medical costs, and even increased mortality risk. Therefore, it is important to determine the optimal empiric therapy for each cIAI and generatee hierarchies of the efficacy, safety, and tolerability of available antimicrobial agents. Network meta-analyses (NMAs) that simultaneously compare all interventions of interest by statistically combining direct and indirect evidence, is probably the best approach for identifying agents exhibiting the best efficacy, safety, and tolerability for the empirical treatment of cIAIs (11).

Thus, we conducted this network meta-analysis to compare all available drugs in the treatment of patients with cIAIs according to the specifications of the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) extension statement for reporting of systematic reviews incorporating network meta-analyses of healthcare interventions (12). The protocol was prospectively registered at PROSPERO International Prospective Register of Systematic Reviews (https://www.crd.york.ac.uk/PROSPERO/) under the registration number CRD42022313771.

2.1 Search Strategy and Study Selection

Three electronic databases, PubMed, EMBASE, and the Cochrane Central Register of Controlled Trials (CENTRAL), were systematically searched from their inception to October 2022 to identify eligible randomized controlled trials (RCTs). We also searched ClinicalTrials.gov to identify completed but unpublished RCTs. The references list of relevant meta-analyses, reviews, pooled analyses, and included trials were manually checked to identify additional studies. We used combinations of MeSH terms and text words around “intra-abdominal infection”, “antimicrobials”, and “randomized controlled trials”.

Two reviewers (WQK, YYW) independently selected the studies in accordance with pre-specified inclusion criteria, and any disagreements were settled via discussion. After removing duplicate studies, the reviewers screened the titles and abstracts of the remaining records, read the full text of the remaining reports, and identified eligible studies.

2.2 Inclusion Criteria

The inclusion criteria for studies were as follow: (1) RCTs; (2) included adults 18 years of age or older with a cIAI; (3) evaluated antimicrobial agents, including fluoroquinolone monotherapy or fluoroquinolone plus metronidazole, β-lactam/β-lactamase inhibitor monotherapy or β-lactam/β-lactamase inhibitor plus metronidazole, cephalosporin monotherapy or cephalosporin plus metronidazole, or carbapenems or other antimicrobial regimens that ensured coverage of common pathogens involved in IAIs, compared with control groups receiving active agents; (4) had to report at least one clinical-related outcome (e.g., clinical success rate or microbiological success rate);(5) constituted patients with cIAIs as a subgroup of the study population, and provided separate data for the group of patients with cIAIs;(6) published in English language or Chinese language.

Studies that evaluated aminoglycosides were excluded due to recommendations against their use in the 2017 Surgical Infection Society revised guidelines on the management of IAI (1) and aminoglycosides-associated nephrotoxicity. Studies focusing on patients with peritoneal dialysis associated peritonitis by reasons of differences in management were also excluded, as were studies evaluating agents in the same antibiotic class (e.g., meropenem versus imipenem/cilastatin). Non-RCTs, conference abstracts, and unpublished studies were also excluded. When duplicate studies were identified, only those that presented the most comprehensive and informative data were included.

2.3 Data Extraction

Two reviewers (WQK, YYW) independently extracted data from studies that met the inclusion criteria, and the data were managed using Microsoft Excel spreadsheets. One reviewer (WQK) then performed a thorough additional check to ensure that the data were accurate. The data included study characteristics (first author, study design, ClinicalTrials.gov Identifier [NCT], publication year, sample size, therapy duration), participant characteristics (mean age, sex ratio, mean acute physiology and chronic health evaluation II [APACHE II] score); interventions (drugs and dose); concomitant antibiotics; and outcomes of interest. The primary outcome was clinical success rate, which was defined as complete resolution or significant improvement in all signs and symptoms of the infection such that no additional antibiotic or procedure was necessary at the test-of-cure visit or the end-of-therapy visit. Secondary outcomes included microbiological success defined as eradication or presumed eradication of the baseline pathogen (if no postbaseline specimen was available for culture, microbiological outcome was based on clinical assessment), all-cause mortality, serious adverse events (SAEs), and discontinuation due to adverse events. As a result of the lack of a standard definition of SAE, we used the criteria as defined in each study. We preferentially extracted a modified intention-to-treat (ITT) population for clinical success, if possible. Otherwise, we used ITT population or available data (i.e., clinically evaluable population). For microbiological success, we extracted a microbiologically modified ITT population, or used the ITT population or available data (i.e., microbiologically evaluable population). The safety outcome was evaluated in the safety population. The total number of events and sample size were extracted for all outcomes. Discrepancies were resolved via discussion.

2.4 Risk of Bias Assessment

The quality of the included trials was assessed in accordance with the Cochrane Collaboration’s risk of bias tool, as described in the Cochrane Collaboration Handbook (13). Two investigators (WQK, YYW) independently determined the risk of bias to be low, unclear, or high based on the presence or absence of random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessors (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), or “other source of bias” (other bias). Discrepancies were resolved via discussion.

2.5 Data Analysis

Data were analyzed using a stepwise approach. Pairwise meta-analyses were performed first. The outcomes were dichotomous data, and results are presented as pooled odds ratios (ORs) with 95% confidence intervals (CIs). A random-effects model was used to derive pooled estimates across studies such models account for between-study differences. Between-study heterogeneity was quantitatively assessed using the I² statistic, with I² > 50% indicating high heterogeneity and I² < 50% indicating low heterogeneity(14). A Bayesian network meta-analysis (NMA) was then conducted using a random-effects model to combine direct and indirect comparisons via Markov chain Monte Carlo methods with the GeMTC package (version 0.8-2) in R (version 4.1.3) to calculate ORs and 95% credible intervals (15, 16). Four Markov chains were run simultaneously with 50 000 simulated draws after a burn-in of 20 000 iterations. Convergence of the model was assessed using trace plots and the Brooks-Gelman-Rubin statistic (17). Consistency was evaluated using node-splitting analysis comparing the differences between direct and indirect estimates for each comparison (18, 19). Model fit was assessed by comparing the deviance information criteria (DIC) between the consistency model and inconsistency model, which indicated a good fit if the number of DIC was approximated (20). Interventions were categorized in the network as follows: fluoroquinolone (FQ)-based regimens (e.g., moxifloxacin, ciprofloxacin plus metronidazole); β-lactam/β-lactamase inhibitor (BL/BLI)-based regimens (e.g., piperacillin-tazobactam, ceftolozane-tazobactam plus metronidazole); cephalosporin (CEP)-based regimens (e.g., ceftriaxone plus metronidazole); carbapenem (CAB) regimens (e.g., doripenem, imipenem-cilastatin). If data were available, ORs and 95% CIs for the primary outcome were calculated separately on the basis of APACHE II scores ≥ 10. In order to validate the robustness of findings for all outcomes, a series of four sensitivity analyses were performed to: (1) exclude trovafloxacin, which has been withdrawn due to serious hepatotoxicity; (2) remove studies at high risk of blinding bias; (3) remove studies with an enrolled sample size of ≤ 100 patients; (4) exclude studies with missing information regarding the APACHE II score. Furthermore, a network meta-regression analysis was conducted to explore the impact of study publication year on all outcomes. The surface under the cumulative ranking (SUCRA) curve and mean ranks were calculated to rank the treatments for each outcome (21). Comparison-adjusted funnel plots were generated to investigate the presence of small-study effects in the intervention network (22). The network command of Stata software version 14.0 was utilized to draw network plots and comparison-adjusted funnel plots (23).

3.1 Seleced Studies and Study Characteristics

The literature screening process is shown in Fig. 1. Electronic searches yielded 2,904 citations, and the full-text versions of 59 publications were subsequently reviewed. Of these, 16 publications were rejected based on the inclusion criteria. Forty-three studies comprising 44 trials met the inclusion criteria and were included in the network meta-analysis. One trial from ClinicalTrials.gov (24) that had completed results but had not been published met the criteria and was included. Forty-five English-language RCTs published between 1992 and 2021 were ultimately included in the NMA. Basic information derived from each of the 45 RCTs is shown in (see supplementary table 1). All studies were active-controlled trials. Twenty-one anti-infection regimens were investigated in the NMA (see supplementary box 1). A total of 38.5% (5,499/14,267) of the included subjects were female. The sample size varied from 56 to 1,043. Twenty-eight RCTs were multicenter studies with a double-blind design. The vast majority of patients had an APACHE II score ≤ 10, and thus considered at low risk for treatment failure or death (1). Nine trials did not report any data on disease severity. The mean age of included patients ranged from 35 to 66 years. The study duration ranged from 4 days to 14 days. Vancomycin, teicoplanin, linezolid, or daptomycin were concomitantly used against resistant gram-positive pathogens in 10 studies. Thirty-seven RCTs disclosed financial support from the pharmaceutical industry.

3.2 Quality of the Included Studies

Detailed random sequence generation was not reported in 20 of the included studies (44.4%), and 23 studies (51.1%) had inadequate allocation concealment. Fourteen trials (31.1%) were open-label design, 20 (44.4%) studies did not describe how participants and personnel were blind, and 23 (51.1%) studies were rated as having an unclear risk of outcome assessment blinding. Thirty-one (68.9%) trials were rated as having low-risk bias in terms of incomplete outcome data. Seventeen (37.8%) studies were considered to have high-risk bias on the basis of selective reporting. Thirty-five (77.8%) trials were deemed to have low-risk of other bias. (See supplementary table 2)

3.3 Assessment of Convergence, Mode Fit and Consistency

Trace plots and Brooks-Gelman-Rubin plots indicated good model convergence in the NMAs (see supplementary Fig. 1a ~ 1d). The consistency and inconsistency models exhibited good fit according to approximated DIC (See supplementary table 3). No significant inconsistencies were noted between direct and indirect evidence, as assessed using the node-splitting method (see supplementary table 4)

3.4 Clinical Success Rates

Node-splitting analyses revealed significant inconsistency in the comparison of CEP_based against BL_BLIs_based (P = 0.0389) because one study (Walker et al., 1993) had used a unique definition of clinical cure (ie., no evidence of infection at completion of treatment and no adverse drug reactions). As transitivity is a core assumption of NMAs, we excluded this study. Therefore, results for clinical success rates were obtained from 44 RCTs involving 16,390 participants covering 7 treatment regimens. The network plot for the outcome appeared in Fig. 2A. The clinical success rate was 82.9% (13,587/16,390). In the NMA, CEP_based regimens were more effective than tigecycline (TGC) with regard to clinical success rate, though the difference was of borderline statistical significance (OR = 1.3, 95% CI 1.00 ~ 1.70), as shown in Fig. 3A. Sensitivity meta-analyses confirmed the robustness of the ORs (see supplementary league table 1a ~ 1d). Pairwise meta-analyses revealed a borderline significant trend towards higher clinical success rates when CAB was compared with TGC (2,315 participants, OR = 1.32, 95% CI 1.00 ~ 1.75, I² = 24.0%) (see supplementary table 5). According to SUCRA curve analysis, CEP_based regimens had the highest probability (83.6%) of being the best regimens among all treatments in terms of clinical success rates (see supplementary table7), and tigecycline had the lowest probability (22.7%). Network meta-regression analysis suggested that publication year had no potential moderating association with the treatments (see supplementary table 6). Twelve of the included articles provided data relative to APACHE II scores ≥ 10 for clinical success rates, totaling 9 treatments arms, including ceftazidime/avibactam plus metronidazole, meropenem, ceftolozane/tazobactam plus metronidazole, ertapenem, piperacillin/tazobactam, cefotaxime plus metronidazole, moxifloxacin, ceftriaxone plus metronidazole, and tigecycline. The NMA results showed that none of the antimicrobial regimens were significantly superior, however, confidence intervals around ORs were very wide (see supplementary league table 5).

3.5 Microbiological Success Rates

Thirty-three RCTs including a total of 8,569 patients reported the microbiological success rates for 6 antimicrobial treatments. The network plot for the outcome appeared in Fig. 2B. Microbiological success was reported in approximately 86% of patients (7,371/8,569). The results of pairwise meta-analyses were generally consistent with those of NMAs (see supplementary table 5). No statistically significant differences were observed in terms of comparative efficacy of any of the treatment regimens in achieving microbiological success (Fig. 3A). Several sensitivity analyses confirmed the robustness of NMA results (see supplementary league table 2a ~ 2c). According to SUCRA curve analysis, cephalosporin-based regimens (76.1%) had the best efficacy in terms of microbiological success rate, and FQ_based regimens (32.4%) were the least efficacious among the treatments examined (see supplementary table 7). The result of network meta-regression analysis demonstrated that study publication year did not affect the outcome (see supplementary table 6).

3.6 All-cause Mortality

Thirty-eight eligible RCTs involving 15,302 patients provided all-cause mortality data for the 6 regimens. The network plot for the outcome appeared in Fig. 2C. Mortality occurred in 3.0% of patients (466/15,302). The results of pairwise and sensitivity meta-analyses were generally consistent with those of NMAs (see supplementary table 5 and league table 3a ~ 3c). No significant differences in all-cause mortality were observed among treatment regimens in the NMA (Fig. 3B). According to the SUCRA curve analysis, eravacycline (78.4%) had lowest all-cause mortality, and tigecycline (12.7%) was ranked the last (see supplementary table 7). Network meta-regression analysis revealed that publication year was not an effect modifier across trials in terms of all-cause mortality (see supplementary table 6).

3.7 Serious Adverse Events

Twenty-five RCTs involving 12,338 participants provided serious adverse event data for the 6 interventions. Serious adverse events were seen in 11.1% of patients (1,368/12,338). Node-split analysis revealed inconsistencies in three comparisons, including CEP_based versus CAB (P = 0.04), TGC versus CAB (P = 0.013) and TGC versus CEP_based (P = 0.007). As NMAs rely on an assumption of transitivity when incorporating all evidence (Chaimani et al., 2017), we only conducted pairwise meta-analyses. In the pairwise meta-analyses, carbapenems were associated with significantly fewer cases of serious adverse events than tigecycline (2,304 participants, OR = 0.67, 95% CI 0.45 ~ 0.91; I² = 48.5%) (see supplementary table 5).

3.8 Discontinuation Due to Adverse Events

Thirty trials covering 7,374 adults provided data for discontinuation due to adverse events. The network plot for the outcome appeared in Fig. 2D. A total of 7.7% of patients (568/7,374) discontinued treatment due to adverse events. In the NMA, the comparison of CAB and CEP_based treatment with TGC exhibited borderline statistical significance for fewer patients discontinuing because of adverse events (OR = 0.69, 95% CI 0.44 ~ 1.00; OR = 0.66, 95% CI 0.42 ~ 1.00, respectively) (Fig. 3B). Sensitivity and pairwise meta-analyses confirmed the robustness of the ORs (see supplementary table 5 and league table 4a ~ 4b). According to SUCRA curve analysis, eravacycline (90.9%) exhibited the best tolerability, and tigecycline (12.9%) exhibited the worst tolerability (see supplementary table 7). Network meta-regression analysis revealed that publication year was not an effect modifier across trials in terms of discontinuation due to adverse events (see supplementary table 6).

3.9 Comparison-adjusted Funnel Plot Analysis

We drew comparison-adjusted funnel plots to identify small- effects for all endpoints. There were no significant asymmetries in the comparison-adjusted funnel plots, indicating an absence of small- effects in terms of clinical success rate, microbiological success rate, all-cause mortality, and discontinuation due to adverse events (Fig. 4A ~ 4D)

To the best of our knowledge, the current analysis involving pooling of direct and indirect comparisons has provided the most comprehensive evidence with respect to comparisons of the efficacy, safety, and tolerability of various antimicrobial regimens for treating cIAIs. Our protocol was prospectively registered at PROSPERO in order to minimize the chance of duplication and reduce opportunities for reporting bias. Our NMA incorporated data for 7 antimicrobial regimens across 45 RCTs to compare differences and determine the relative ranking of antibiotics of empirical therapy for cIAIs in terms of efficacy, safety, and tolerability. Our study is timely considering the novel antimicrobials (e.g., ceftolozane/tazobactam, eravacycline) that clinicians can prescribe as a potential treatment plan for cIAIs in adults. Compared with several previous pairwise meta-analyses (8–10, 25–28), our study provided comprehensive evidence regarding the relative efficacy, safety, and tolerability of different antimicrobial regimens and clear evidence with respect to which specific antimicrobial regimens are the best candidates for the empirical treatment of cIAIs.

A major finding of our study was that cephalosporin-based regimens are slightly more efficacious than tigecycline with regard to clinical success rate and appear to provide the highest benefit among the examined therapeutic regimens in terms of clinical and bacteriological outcomes. These results did not change based on sensitivity analyses. Therefore, a cephalosporin plus metronidazole is considered as a superior therapeutic choice in patients with cIAIs. Few studies have explored the differences in clinical efficacy between cephalosporin-based regimens and tigecycline. The present results further support the recommendations of the guideline published by the Surgical Infection Society in 2017 (1), which suggested the use of a cephalosporin plus metronidazole for initial empiric therapy in low-risk patients with cIAIs. Clinical data from 3,233 cIAI patients published in 2016 by Ouyang et al. indicated that a cephalosporin plus metronidazole should be first-line option for empirical therapy cIAI patients (5), which strengthens our results. The comparison of carbapenems and tigecycline in our pairwise meta-analysis was of borderline significance, but no significant difference was observed in the NMA. A Bayesian meta-analysis of 15 studies comparing the effectiveness of individual carbapenems versus tigecycline showed consistent results (29). Thus, the borderline significance result in our study should be interpreted conservatively (30). Perhaps there is a small true difference between carbapenems and tigecycline in cIAI patients, but additional high-quality data are urgently needed to clarify this issue. Our NMAs indicated that other antimicrobial regimens (monotherapy or in combination with metronidazole) provide comparable effects in terms of clinical and microbiological outcomes, in line with previous meta-analyses (8–10, 25–28), which suggests these regimens may be suitable for selection as initial empirical therapy in cIAI patients.

The severity of illness is an important factor in guiding the subsequent selection of empirical antimicrobial therapy and optimization of the selection of source control. A high APACHE II score is associated with an increased risk of adverse outcomes in cIAI patients. The current guideline recommends that cIAI patients with an APACH II score ≥ 10 be considered at increased risk of adverse outcome in future management (1). The present study is the first NMA to examine differences between antimicrobial agents in terms of clinical efficacy in higher-risk cIAI patients, and none of the 9 treatments examined, including 2 novel β-lactam/β-lactamase combination antibiotics (ceftazidime/avibactam, ceftolozane/tazobactam), were superior. Regrettably, an insufficient number of events reported only in 12 studies increased the uncertainty of the results. A future study should examine more data from separate studies of high-risk patients to verify our findings.

The second main finding of this study was that no one antibiotic agent or regimen, including tigecycline, was associated with an increased risk of all-cause mortality, although tigecycline exhibited poor safety and tolerability, as defined by withdrawal due to adverse events, when compared with carbapenems and cephalosporins plus metronidazole. Excess mortality in patients who received tigecycline with some infections (i.e., hospital-acquired pneumonia) has aroused considerable concern (31, 32). Our results indicate that the use of tigecycline at the standard dose of an initial 100 mg followed by 50 mg every 12 h did not contribute to imbalanced mortality in cIAI patients. However, our findings suggest that tigecycline may not be a good first choice for empirical therapy in patients with cIAI because of its poor safety and tolerability profiles, which again confirms the current guideline(1). These findings diverged from a previous guideline issued in 2010 indicating the use of tigecycline for the initial empiric treatment of mild-to-moderate community-acquired infections (3). Our findings reflect the real-time and supplementary evidence regarding the efficacy and safety of tigecycline compared with other regimens for cIAI treatment. Indeed, other studies have shown higher rates of adverse events in patients treated with tigecycline relative to those treated with comparators (29, 33). Nevertheless, tigecycline remains an important treatment option for patients suspected of have an infection with an antibiotic-resistant pathogen (1, 34, 35).

Finally, considering the efficacy, safety, and tolerability profiles for cIAIs of novel antimicrobials, including eravacycline, ceftazidime/avibactam, ceftolozane/tazobactam, and imipenem/cilastatin/relebactam, our present NMA results indicated the noninferiority of these regimens to carbapenems or tigecycline in terms of clinical and microbiological outcomes and had favorable safety and their tolerability profiles, consistent with previous meta-analyses (25–27). Furthermore, our NMA provided evidence that two novel beta-lactam/beta-lactamase inhibitors plus metronidazole were similar to eravacycline for empirical therapy of cIAIs with respect to efficacy, safety, and tolerability. Given that the increasing prevalence of antibiotic-resistant gram-negative pathogens, especially carbapenem-resistant K. pneumoniae (36), our findings provide useful information to inform clinical physicians prescribing empiric antimicrobials when infection with difficult-to-treat gram-negative pathogens are suspected. These findings also provide evidence regarding effective carbapenem-sparing options to preserve the activity of these agents.

Several limitations of our NMA should be considered in the interpretation of our results. First, we failed to conduct an NMA for serious adverse events due to inconsistencies in some comparisons. Seven studies (37–43) used defined criteria for SAEs, whereas the remaining studies classified SAEs based on investigator’ judgement. Discrepancies between studies in terms of judging SAEs could lead to statistical inconsistencies. Second, as with other meta-analyses, heterogeneity was an inherent limitation. Although we used strict eligibility criteria to ensure that the studies included were as homogeneous as possible, differences in participant characteristics, study duration, and study design might have increased the heterogeneity. These confounding factors might have weakened the internal validity of the evidence. However, the publications did not provide sufficient patient-level data and study characteristics, which impeded the examination of sources of heterogeneity. Third, the methodological quality of some studies was low; for example, ≥ 50% of trials did not provide adequate information about allocation concealment, and 31.1% of studies lacked blinding of participants and outcome assessors. The high risk of bias could diminish the reliability and robustness of the findings, although sensitivity analyses confirmed the consistency of results for most comparisons. Fourth, we did not include placebo-controlled studies owing to ethical standards, which prevented assessment of the efficacy and safety of antibiotic treatment versus no antibiotic treatment. Therefore, we only synthesized outcome data for the relative efficacy and safety of agents used for the treatment of cIAIs. Only including active-controlled studies could have contributed to smaller relative effects compared with included placebo-controlled studies. Fifth, although we conducted a separate NMA of comparative efficacy of antibiotics in high- risk patients, only 12 studies with low event numbers were included. Therefore, the results were inconclusive, and more data will be needed to verify the findings. Sixth, most studies included few individuals more vulnerable to treatment failure or death. For example, some studies excluded patients with an APACHE II score > 30; thus, our findings should be extrapolated cautiously in this population. Seventh, as local bacterial epidemiology, pharmacokinetics and pharmacodynamics data for the antibacterial agents, cost-effectiveness and availability of antibiotics, and physiopathologic factors collectively played a role in the selection of empirical antimicrobial therapies, it is impossible to take into account all factors in this study. For example, there is high prevalence of ESBL- producing Escherichia coli in China. Approximately 53.8% and 39.3% of E. coli were resistant to levofloxacin and ampicillin-sulbactam (http://www.chinets.com/Document/Index?pageIndex=0) and they may be prescribed cautiously for cIAIs patients in China. However, about 90% of E. coli are susceptible to cefotaxime or ceftriaxone in North America, thus cephalosporin-based regimens were recommended for the treatment of lower-risk patients with cIAIs (1).

In summary, as the most recent and thorough meta-analysis on this subject to date, this work has important clinical implications. All comparisons between current drugs should be considered within the context of the limitations of this NMA. We hope that our results will provide helpful perspectives to facilitate decisions-making by patients and clinicians.

CEP_based, cephalosporin-based regimens; CAB, carbapenems; FQ_based, fluoroquinolone-based regimens; BL_BLIs_based, β-lactam/β-lactamase inhibitor-based regimens; TGC, tigecycline; cIAI, complicated intra-abdominal infection; RCT, randomized controlled trial; SUCRA, surface under the cumulative ranking curve; NMA, network meta-analysis; APACHE II, acute physiology and chronic health evaluation II; DIC, deviance information criterion; ORs, odds ratios; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-analyses; CENTRAL, Cochrane Central Register of Controlled Trials

Acknowledgments

None.

Funding

None.

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

AUTHOR CONTRIBUTIONS

Wen Qiang Kong and Ting Deng designed this study. Wen Qiang Kong and Yan Yan Wu selected the studies, extracted data and analyzed data. Wen Qiang Kong drafted this manuscript. Yun Feng Shu and Shi Qin Li checked this manuscript.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Mazuski JE, Tessier JM, May AK, Sawyer RG, Nadler EP, Rosengart MR, et al. The Surgical Infection Society Revised Guidelines on the Management of Intra-Abdominal Infection. Surg Infect (Larchmt). 2017;18(1):1–76.
Skrupky LP, Tellor BR, Mazuski JE. Current strategies for the treatment of complicated intraabdominal infections. Expert Opin Pharmacother. 2013;14(14):1933–47.
Solomkin JS, Mazuski JE, Bradley JS, Rodvold KA, Goldstein EJ, Baron EJ, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis. 2010;50(2):133–64.
Krobot K, Yin D, Zhang Q, Sen S, Altendorf-Hofmann A, Scheele J, et al. Effect of inappropriate initial empiric antibiotic therapy on outcome of patients with community-acquired intra-abdominal infections requiring surgery. Eur J Clin Microbiol Infect Dis. 2004;23(9):682–7.
Ouyang W, Xue H, Chen Y, Gao W, Li X, Wei J, et al. Clinical characteristics and antimicrobial patterns in complicated intra-abdominal infections: a 6-year epidemiological study in southern China. Int J Antimicrob Agents. 2016;47(3):210–6.
Scott LJ, Eravacycline. A Review in Complicated Intra-Abdominal Infections. Drugs. 2019;79(3):315–24.
van Duin D, Bonomo RA. Ceftazidime/Avibactam and Ceftolozane/Tazobactam: Second-generation beta-Lactam/beta-Lactamase Inhibitor Combinations. Clin Infect Dis. 2016;63(2):234–41.
Falagas ME, Peppas G, Makris GC, Karageorgopoulos DE, Matthaiou DK. Meta-analysis: ertapenem for complicated intra-abdominal infections. Aliment Pharmacol Ther. 2008;27(10):919–31.
Li Y, Chen L, Jiang J, Li X, Huang T, Liang X. Carbapenems vs beta-Lactam Monotherapy or Combination Therapy for the Treatment of Complicated Intra-abdominal Infections: Systematic Review and Meta-analysis of Randomized Controlled Trials. Open Forum Infect Dis. 2019;6(10):ofz394.
Mavros MN, Theochari NA, Kyriakidou M, Economopoulos KP, Sava JA, Falagas ME. Fluoroquinolone-based versus beta-lactam-based regimens for complicated intra-abdominal infections: a meta-analysis of randomised controlled trials. Int J Antimicrob Agents. 2019;53(6):746–54.
Salanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Res Synth Methods. 2012;3(2):80–97.
Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med. 2015;162(11):777–84.
Higgins JP, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928.
Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539–58.
van Valkenhoef G, Kuiper J. gemtc: network meta-analysis using Bayesian methods: R package version 0.8. 2018 [Available from: http://cran.r-project.org/web/packages /gemtc/index.html.
van Valkenhoef G, Lu G, de Brock B, Hillege H, Ades AE, Welton NJ. Automating network meta-analysis. Res Synth Methods. 2012;3(4):285–99.
Brooks S, Gelman A. General methods for monitoring convergence of iterative simulations. J Comput Graph Stat. 1998;7:434–55.
Higgins JP, Jackson D, Barrett JK, Lu G, Ades AE, White IR. Consistency and inconsistency in network meta-analysis: concepts and models for multi-arm studies. Res Synth Methods. 2012;3(2):98–110.
van Valkenhoef G, Dias S, Ades AE, Welton NJ. Automated generation of node-splitting models for assessment of inconsistency in network meta-analysis. Res Synth Methods. 2016;7(1):80–93.
Dias S, Welton NJ, Sutton AJ, Caldwell DM, Lu G, Ades AE. Evidence synthesis for decision making 4: inconsistency in networks of evidence based on randomized controlled trials. Med Decis Making. 2013;33(5):641–56.
Salanti G, Ades AE, Ioannidis JP. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol. 2011;64(2):163–71.
Chaimani A, Salanti G. Using network meta-analysis to evaluate the existence of small-study effects in a network of interventions. Res Synth Methods. 2012;3(2):161–76.
Chaimani A, Higgins JP, Mavridis D, Spyridonos P, Salanti G. Graphical tools for network meta-analysis in STATA. PLoS ONE. 2013;8(10):e76654.
Merck Sharp D, Ceftolozane/Tazobactam. (MK-7625A) Plus Metronidazole Versus Meropenem for Participants with Complicated Intra-abdominal Infection. 2021 [Available from: https://clinicaltrials.gov/ct2/show/NCT03830333?term=NCT03830333&draw=2&rank=1.
Cheng IL, Chen YH, Lai CC, Tang HJ. The use of ceftolozane-tazobactam in the treatment of complicated intra-abdominal infections and complicated urinary tract infections-A meta-analysis of randomized controlled trials. Int J Antimicrob Agents. 2020;55(2):105858.
Lan SH, Chang SP, Lai CC, Lu LC, Chao CM. The Efficacy and Safety of Eravacycline in the Treatment of Complicated Intra-Abdominal Infections: A Systemic Review and Meta-Analysis of Randomized Controlled Trials.J Clin Med. 2019;8(6).
Mikamo H, Yuasa A, Wada K, Crawford B, Sugimoto N. Optimal Treatment for Complicated Intra-abdominal Infections in the Era of Antibiotic Resistance: A Systematic Review and Meta-Analysis of the Efficacy and Safety of Combined Therapy With Metronidazole. Open Forum Infect Dis. 2016;3(3):ofw143.
Mu YP, Liu RL, Wang LQ, Deng X, Zhu N, Wei MD, et al. Moxifloxacin monotherapy for treatment of complicated intra-abdominal infections: a meta-analysis of randomised controlled trials. Int J Clin Pract. 2012;66(2):210–7.
Chen L, Liang X, Jiang J, Li X, Li Y. Carbapenems vs tigecycline for the treatment of complicated intra-abdominal infections: A Bayesian network meta-analysis of randomized clinical trials. Med (Baltim). 2019;98(40):e17436.
Hackshaw A, Kirkwood A. Interpreting and reporting clinical trials with results of borderline significance. BMJ. 2011;343:d3340.
Dixit D, Madduri RP, Sharma R. The role of tigecycline in the treatment of infections in light of the new black box warning. Expert Rev Anti Infect Ther. 2014;12(4):397–400.
Scaglione F. Comment on: Efficacy and safety of tigecycline: a systematic review and meta-analysis. J Antimicrob Chemother. 2011;66(12):2892–3. author reply 5–6.
Lauf L, Ozsvar Z, Mitha I, Regoly-Merei J, Embil JM, Cooper A, et al. Phase 3 study comparing tigecycline and ertapenem in patients with diabetic foot infections with and without osteomyelitis. Diagn Microbiol Infect Dis. 2014;78(4):469–80.
Sartelli M, Coccolini F, Kluger Y, Agastra E, Abu-Zidan FM, Abbas AES, WSES/GAIS/SIS-, et al. E/WSIS/AAST global clinical pathways for patients with intra-abdominal infections. World J Emerg Surg. 2021;16(1):49.
Sartelli M, Weber DG, Ruppe E, Bassetti M, Wright BJ, Ansaloni L, et al. Antimicrobials: a global alliance for optimizing their rational use in intra-abdominal infections (AGORA). World J Emerg Surg. 2016;11:33.
Hu F, Zhu D, Wang F, Wang M. Current Status and Trends of Antibacterial Resistance in China. Clin Infect Dis. 2018;67(suppl2):128–S34.
Chen Y, Zhu D, Zhang Y, Zhao Y, Chen G, Li P, et al. A multicenter, double-blind, randomized, comparison study of the efficacy and safety of tigecycline to imipenem/cilastatin to treat complicated intra-abdominal infections in hospitalized subjects in China. Ther Clin Risk Manag. 2018;14:2327–39.
Chen Z, Wu J, Zhang Y, Wei J, Leng X, Bi J, et al. Efficacy and safety of tigecycline monotherapy vs. imipenem/cilastatin in Chinese patients with complicated intra-abdominal infections: a randomized controlled trial. BMC Infect Dis. 2010;10:217.
Fomin P, Beuran M, Gradauskas A, Barauskas G, Datsenko A, Dartois N, et al. Tigecycline is efficacious in the treatment of complicated intra-abdominal infections. Int J Surg. 2005;3(1):35–47.
Garbino J, Villiger P, Caviezel A, Matulionyte R, Uckay I, Morel P, et al. A randomized prospective study of cefepime plus metronidazole with imipenem-cilastatin in the treatment of intra-abdominal infections. Infection. 2007;35(3):161–6.
Lucasti C, Popescu I, Ramesh MK, Lipka J, Sable C. Comparative study of the efficacy and safety of ceftazidime/avibactam plus metronidazole versus meropenem in the treatment of complicated intra-abdominal infections in hospitalized adults: results of a randomized, double-blind, Phase II trial. J Antimicrob Chemother. 2013;68(5):1183–92.
Oliva ME, Rekha A, Yellin A, Pasternak J, Campos M, Rose GM, et al. A multicenter trial of the efficacy and safety of tigecycline versus imipenem/cilastatin in patients with complicated intra-abdominal infections [Study ID Numbers: 3074A1-301-WW; ClinicalTrials.gov Identifier: NCT00081744]. BMC Infect Dis. 2005;5:88.
Weiss G, Reimnitz P, Hampel B, Muehlhofer E, Lippert H, Group AS. Moxifloxacin for the treatment of patients with complicated intra-abdominal infections (the AIDA Study). J Chemother. 2009;21(2):170–80.

No competing interests reported.

SupplementaryMaterial.pdf

Download PDF

Journal Publication

published 21 Apr, 2023

Read the published version in BMC Infectious Diseases →

Editorial decision: Major revision
01 Mar, 2023
Reviews received at journal
28 Feb, 2023
Reviews received at journal
04 Feb, 2023
Reviewers agreed at journal
27 Jan, 2023
Reviewers agreed at journal
24 Jan, 2023
Reviewers invited by journal
24 Jan, 2023
Editor assigned by journal
24 Jan, 2023
Editor invited by journal
09 Jan, 2023
Submission checks completed at journal
09 Jan, 2023
First submitted to journal
31 Dec, 2022

You are reading this latest preprint version

Efficacy, safety, and tolerability of antimicrobial agents for complicated intra- abdominal infection: a systematic review and network meta-analysis

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

1. Background

2. Methods

2.1 Search Strategy and Study Selection

2.2 Inclusion Criteria

2.3 Data Extraction

2.4 Risk of Bias Assessment

2.5 Data Analysis

3. Results

3.1 Seleced Studies and Study Characteristics

3.2 Quality of the Included Studies

3.3 Assessment of Convergence, Mode Fit and Consistency

3.4 Clinical Success Rates

3.5 Microbiological Success Rates

3.6 All-cause Mortality

3.7 Serious Adverse Events

3.8 Discontinuation Due to Adverse Events

3.9 Comparison-adjusted Funnel Plot Analysis

4. Discussion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1