Carbapenems to Treat Multidrug and Extensively Drug-Resistant Tuberculosis: A Systematic Review

Background: Carbapenems (ertapenem, imipenem, meropenem) are used to treat multidrug-resistant (MDR-) and extensively drug-resistant tuberculosis (XDR-TB), even if the published evidence is limited, particularly when it is otherwise difficult to identify the recommended four active drugs to be included in the regimen. No systematic review to date has ever evaluated the efficacy, safety, and tolerability of carbapenems. Methods: A search of peer-reviewed, scientific evidence was carried out, aimed at evaluating the efficacy/effectiveness, safety, and tolerability of carbapenem-containing regimens in individuals with pulmonary/extra-pulmonary disease which was bacteriologically confirmed as M/XDR-TB. We used PubMed to identify relevant full-text, English manuscripts up to the 20 December 2015, excluding editorials and reviews. Results: Seven out of 160 studies satisfied the inclusion criteria: two on ertapenem, one on imipenem, and four on meropenem, all published between 2005 and 2016. Of seven studies, six were retrospective, four were performed in a single center, two enrolled children, two had a control group, and six reported a proportion of XDR-TB cases higher than 20%. Treatment success was higher than 57% in five studies with culture conversion rates between 60% and 94.8%. Conclusions: The safety and tolerability is very good, with the proportion of adverse events attributable to carbapenems below 15%.


Introduction
With over 480,000 new multidrug-resistant tuberculosis (MDR-TB, i.e., tuberculosis caused by Mycobacterium tuberculosis strains resistant, in vitro, to at least isoniazid and rifampicin) cases A search of peer-reviewed, scientific evidence was carried out, aimed at evaluating the efficacy/ effectiveness, safety, and tolerability of carbapenem-containing regimens in individuals with a pulmonary/extra-pulmonary disease, which was culture-and DST-confirmed as M/XDR-TB.
We used the database PubMed to identify any relevant full-text English manuscript without any time constraints until 20 December 2015. We decided to exclude conferences' abstracts because, on the basis of the word count, the information provided is too limited when assessing the above described objectives. A search for ongoing or recently completed trials on clinicaltrials.gov was also performed.
Several key-words (i.e., MDR-TB, XDR-TB, efficacy, effectiveness, safety, tolerability, carbapenem, meropenem, imipenem, ertapenem) were identified by the authors and combined to create ad hoc strings to retrieve the most relevant scientific articles. References' lists of the selected papers were analyzed in order to identify articles missed by the adopted search algorithms.
The search exclusion criteria were the following: (1) Case-reports describing fewer than 3 M/XDR-TB cases; (2) Experimental studies on animals with TB; (3) Reviews and editorials on carbapenems and M/XDR-TB; and (4) Unclear/unconfirmed M/XDR-TB diagnosis of treated patients.
The carbapenems considered for the present review were the following: imipenem, meropenem, and ertapenem (there is no published data on the in vivo use of doripenem, biapenem, panipenem, and faropenem in TB patients). Observational and experimental studies were considered suitable for the estimation of the efficacy/effectiveness, safety, and tolerability of the carbapenems.
Several demographic, epidemiological, clinical variables were collected: epidemiological design of the study, age and sex of the enrolled patients, duration of the study, location/s of the study (mono/multi-center, university/reference/peripheral hospital, country/ies), number and characteristics of treated M/XDR-TB patients, number and characteristics of control cases if any, dosage and duration of the administered drugs, adverse events and their severity if any, treatment outcomes recorded according to the WHO classification [44], sputum smear, and culture conversion.
Two authors independently performed the search and evaluated the titles and abstracts of the records according to the selection criteria. Potentially interesting articles were downloaded and critically assessed; when they satisfied the inclusion criteria, the planned information was retrieved and collected using a pre-designed electronic template. The entire process was carried out following the guidelines of the 2009 Preferred Reporting Items for Systematic Reviews and Meta-Analysis [45].

Results
A total of 159 records were obtained from the search. Nine studies satisfied the inclusion criteria and were deemed appropriate for a qualitative analysis. The entire selection process is summarized in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart ( Figure 1). The carbapenems considered for the present review were the following: imipenem, meropenem, and ertapenem (there is no published data on the in vivo use of doripenem, biapenem, panipenem, and faropenem in TB patients). Observational and experimental studies were considered suitable for the estimation of the efficacy/effectiveness, safety, and tolerability of the carbapenems.
Several demographic, epidemiological, clinical variables were collected: epidemiological design of the study, age and sex of the enrolled patients, duration of the study, location/s of the study (mono/multi-center, university/reference/peripheral hospital, country/ies), number and characteristics of treated M/XDR-TB patients, number and characteristics of control cases if any, dosage and duration of the administered drugs, adverse events and their severity if any, treatment outcomes recorded according to the WHO classification [44], sputum smear, and culture conversion. Two authors independently performed the search and evaluated the titles and abstracts of the records according to the selection criteria. Potentially interesting articles were downloaded and critically assessed; when they satisfied the inclusion criteria, the planned information was retrieved and collected using a pre-designed electronic template. The entire process was carried out following the guidelines of the 2009 Preferred Reporting Items for Systematic Reviews and Meta-Analysis [45].
Patients were recruited and exposed to carbapenems from 2001 [29] to 2015 [31,[41][42][43]. The mean number of patients exposed to single carbapenems were 44.6 per study, ranging from five [31] to 96 [41,43]. Only two studies enrolled children [30,32], and three studies included a control group [29,41,42]. The proportion of XDR-TB cases was higher than 20% in all but one study [32] ( Table 2). Daily dosages of meropenem varied in the selected studies; however, all the individuals with MDR-TB exposed to ertapenem were administered a single daily dose of 1 g [31,32].

Discussion
Aim of this systematic review is to describe the effectiveness, safety, and tolerability profile of carbapenems added to a background regimen, in the treatment of MDR-and XDR-TB cases.
Overall, the results of the systematic review show that carbapenems are safe and likely to be effective in treating M/XDR-TB.
Although not all the studies reported sputum smear and culture conversions rates, the available information confirms that the treatment outcomes of the cohorts treated with carbapenems was in general better than those reported in the literature [3,4,46].
In the largest cohort of MDR-TB cases available in the literature (almost 10,000 cases) the treatment success in the whole cohort was slightly higher than 60%, being much lower in XDR-TB cases: 43% in those meeting strictly the definition (e.g., being resistant to at least one fluoroquinolone and one second-line injectable drug), 30% among those with resistance to two second-line injectables and as low as 19% among those with additional resistance to ethambutol and pyrazinamide. In the cohorts in which carbapenems were used the treatment success ranged from 57.3% to 80.3% (although the definition was not standardized for the selected studies), and the culture conversion rates (which are considered predictors of success) ranged from 60% to almost 95%.
As recently shown [46], high sputum smear and culture conversion rates are likely to indicate low probability of acquiring drug resistance.
This systematic review highlights the observational nature of the studies performed to date. In the majority of the cases the authors of the selected studies carried out a description of case-series, excluding a control group from the epidemiological design. Furthermore, sample sizes were frequently small, on the basis of being mono-center studies and being performed in areas of low M/XDR-TB prevalence. All the studies were focused on difficult-to-treat cases. Evidence on children is limited and deserves further investigation, as well as a more accurate detailed profile on the safety and tolerability profile of the prescribed drugs. In particular, it would be important to assess the causal relationship between the occurrence of an adverse event and a specific drug. The majority of the papers we retrieved do not follow an international, standardized approach for the classification of adverse events, being more focused on effectiveness.
Another important topic which should be better analyzed by future studies is the patients' adherence to the prescribed drugs. No relevant information was provided by the studies included in this systematic review.
One of the most interesting findings is represented by the high effectiveness in cohorts of patients where the number of XDR-TB cases is high. It is clear that these preliminary results should be confirmed in order to better understand how and in which patient categories the carbapenem-containing regimens may make a difference.
Consecutive enrolment, missing blindness, and treatment randomization represent relevant methodological biases which may hamper the inference of the findings of the available studies. However, this review summarizes the current scientific evidence, highlighting the shortcomings which could be improved in the near future. In particular, more emphasis should be given to regimens including meropenem and ertapenem. Unfortunately, it was not possible to perform a meta-analysis. Data collected and summarized from the authors of the selected papers were heterogeneous and not sufficient to carry out a quantitative analysis. Moreover, only a few papers described the clinical activity of the three single molecules, including five on meropenem, two on ertapenem, and three focused on imipenem. Likely, the authors did not compute the necessary sample size to reduce the β error and to increase the statistical power of the study. All the methodological limitations have undoubtedly reduced the reliability of the clinical findings.
Randomized, controlled clinical trials should be implemented in order to elucidate the real therapeutic contribution of carbapenems in TB treatment success; different anti-TB regimens, as well as the different duration of the carbapenem's exposure, can create background noise which can hinder the interpretation of the carbapenem's contribution. The experimental studies should be based on the comparisons between the different carbapenems, as well as between the carbapenem and the current therapeutic alternatives. The investigators should consider all the demographic, epidemiological, microbiologic, and clinical characteristics, including the number of resistances to anti-TB drugs, which could influence the real impact of the pharmacodynamic action of the carbapenems. Furthermore, drug-drug interactions of carbapenems with the new anti-TB drugs bedaquiline and delamanid in terms of reduced/increased efficacy and safety should be investigated.
Of interest, there are two recent early bactericidal activity trials (phase-two trials) testing carbapenems; one trial evaluates the early bactericidal activity, safety, and tolerability of meropenem plus amoxycillin/CA and faropenem plus amoxycillin/CA in adult patients with newly-diagnosed pulmonary tuberculosis (NCT02349841) has recently been completed (results awaiting).
The second trial evaluating the early bactericidal activity study of faropenem (with amoxicillin/ clavulanic acid) in patients with pulmonary tuberculosis is currently recruiting (NCT02381470). These two phase-two trials will be of interest as they will be able to evaluate the individual bactericidal contribution of the penem/carbapenem drugs [47].
Economic analysis studies will also provide a proper perspective of the future role of carbapenems in the treatment armamentarium. The identification of new and more effective therapeutic strategies is crucial to reduce the increasing clinical, economic, and financial burden associated to the emergence of the MDR-TB epidemic, particularly in low-and middle-income countries (e.g., Southeast Asia and former Soviet Union countries) [48].

Conclusions
The systematic review is a contribution toward the best possible use of carbapenems when the number of active drugs necessary to design an effective regimen is lacking [3,4,9,49].
Although new drugs are becoming available to support the move towards TB elimination [50][51][52][53][54], repurposed drugs, like the carbapenems, exemplified by their tolerability, safety, and activity show a potential role in the TB treatment arena, and could be included in new TB drug combination trials, as better evidence is needed.