Efficacy and Safety of Adjunctive Recombinant Human Interleukin-2 for Patients with Pulmonary Tuberculosis: A Meta-Analysis

Background The results of previous clinical trials evaluating the efficacy and safety of recombinant human interleukin-2 (rhuIL-2) for adult patients with pulmonary tuberculosis showed inconsistent results. Accordingly, a comprehensive systematic review and meta-analysis was performed. Methods Relevant randomized controlled trials (RCTs) were retrieved by searching the PubMed, Embase, Cochrane's Library, Web of Science, Wanfang, and CNKI databases. A random-effects model was used to combine the results. Results 18 RCTs with 2630 patients were included in this meta-analysis. Pooled results showed that adjunctive rhuIL-2 significantly increased the odds of sputum culture conversion to negative (risk ratio [RR]: 1.27, 95% CI: 1.09 to 1.47, p=0.002, I2 = 80%), sputum smear conversion to negative (RR: 1.35, 95% CI: 1.17 to 1.57, p < 0.001, I2 = 83%), radiographic focus absorption (RR: 1.17, 95% CI: 1.06 to 1.30, p=0.002, I2 = 72), and cavity closure (RR: 1.24, 95% CI: 1.09 to 1.40, p < 0.001, I2 = 23). The use of rhuIL-2 was not related to any severe adverse events which led to discontinuation of the treatment. Results showed that rhuIL-2 was related to an increased risk of fever (RR: 2.46, 95% CI: 1.29 to 4.70, p=0.006, I2 = 0%). The incidence of other adverse events, such as musculoskeletal pain, hepatic injury, and renal toxicity, was not significantly different between groups (p all >0.05). Conclusions rhuIL-2 is an effective adjunctive immunotherapy for patients with pulmonary tuberculosis.


Introduction
Tuberculosis (TB) is a major infectious disease and a serious public health problem worldwide, which is caused by the infection of Mycobacterium tuberculosis (Mtb) [1,2]. As of 2020, the World Health Organization (WHO) has proximately estimated 9.9 million new cases and 1.28 million deaths of TB in the global population [3]. Currently, the cornerstone for the treatment of TB is the standard chemotherapy [4,5]. However, some inevitable problems during chemotherapy still exist, including a long treatment course, severe adverse efects, poor compliance, and resistance to multiple drugs [6,7]. Accordingly, eforts to develop novel adjunctive therapy for patients with TB are still of great signifcance in clinical medicine and public health.
Accumulating evidence suggests that a host's ability to recognize, respond to, and regulate MTB determines the occurrence, development, and outcome of TB [8][9][10]. Because activated macrophages and specifc T cells work in concert to protect the host against TB through the production and interaction of innate immune cells, treatments that enhance protective immunity or regulate adaptive immunity against TB are potential adjuvants for patients with advanced disease [11,12]. It has been confrmed in preclinical studies that immune activation and regulation are both mediated by interleukin-2, a cytokine associated with T1-type immune responses [13,14]. Besides, pilot studies also showed that IL-2 could cause diferential gene expression in peripheral blood mononuclear cells (PBMCs) stimulated by Mtb [15] and enhance the proliferation and transformation of CD4 + T cells and NK cells [16], which might collectively enhance the anti-TB efcacy of the standard chemotherapy. However, previous clinical trials evaluating the infuence of recombinant human interleukin-2 (rhuIL-2) as adjuvant to chemotherapy in adult patients with pulmonary TB showed inconsistent results [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33]. Besides, it remains largely unknown whether the potential efcacy and safety of adjunctive rhuIL-2 treatment are similar in patients with drug-susceptible and multidrugresistant TB (MDR-TB) and in patients with newly diagnosed and recurrent TB. Terefore, we performed a metaanalysis to comprehensively summarize the efcacy and safety of adjunctive rhuIL-2 treatment on the basis of standard chemotherapy in adults with pulmonary TB.

Methods
Te PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement [34,35] and the Cochrane Handbook guidelines [36] were followed during the design and implementation of the study.

Study Selection.
Studies that fulflled the following criteria were included: [1] full-length articles published in English or Chinese; [2] designed as parallel-group RCTs; [3] adult patients who were diagnosed with HIV-seronegative pulmonary TB and randomly allocated to a treatment group with adjunctive rhuIL-2 and a control group without rhuIL-2 on the basis of standard chemotherapy for TB; and [4] reported at least one of the following efcacy outcomes, including the proportion of patients with sputum culture conversion to negative, the proportion of patients with sputum smear conversion to negative, the proportion of patients with radiographic focus absorption, and the proportion of patients with radiographic cavity closure. Radiographic changes of the pulmonary TB focuses were rated to the following four grades as previously described: marked absorption (meaning signifcant improvement of more than half of initial abnormalities), moderate absorption (meaning defnite improvement better than initial abnormalities but less than a half ), no changes (no certain diference in flms compared with original lesion), and deterioration (being worse than initial abnormalities or spreading to another area) [37]. Te combined proportions of patients with marked and moderate absorption were considered as those with focus absorption. No restriction was applied to the dosage, route, and duration of rhuIL-2 treatment [37].
Nonrandomized studies, studies including patients without pulmonary TB, or studies that did not report the outcomes of interest were excluded. For studies with overlapped patient population, the one with the largest sample size was included for the meta-analysis.

Data Extraction and Quality Assessment.
Database searches, data extraction, and quality evaluation were conducted by two independent authors. If disagreement occurred, it was resolved by discussion with the corresponding author. We extracted data regarding study information (frst author, publication year, and study country), study design (blind or open-label), and patient information (number of patients, mean age, sex, MDR or drug-susceptible TB, and newly diagnosed or recurrent pulmonary TB), background treatments, dosages, routes, and the duration of rhuIL-2 treatment, regimens of controls, and follow-up duration). Quality evaluation was achieved using the Cochrane's Risk of Bias Tool [36] according to the following aspects: [1] random sequence generation, [2] allocation concealment, [3] blinding of participants and personnel, [4] blinding of outcome assessors, [5] incomplete outcome data, [6] selective outcome reporting, and [7] other potential bias.

Statistical Analysis.
Te methodology of statistics is generally considered with the previous published metaanalysis involving RCTs [38]. Te infuence of adjunctive rhuIL-2 on the proportion of patients who achieved the efcacy outcomes were presented as risk ratios (RRs) and the corresponding 95% confdence intervals (CIs). Besides, the infuence of rhuIL-2 on the risks of common adverse events, including fever, musculoskeletal pain, hepatic injury, and renal toxicity, were also summarized as RRs and 95% CIs. We used the Cochrane's Q-test to detect the heterogeneity [39]. Te I 2 statistic was also calculated, and an I 2 >50% refected signifcant heterogeneity [40]. Pooled analyses were calculated using a random-efects model because this method incorporates the infuence of potential heterogeneity and provides a more generalized result [36]. Sensitivity analysis by exclusion of one study at a time was used to evaluate the infuence of each study on the pooled results of the meta-analysis [36]. Additionally, subgroup analyses were performed to evaluate whether the results were similar in patients with MDR-TB and drug-susceptible TB and in patients with newly diagnosed and recurrent TB. Meta-regression analyses were performed to evaluate the possible infuence of patient and treatment characteristics on the efcacy outcomes, including the number of patients, the mean age, the mean daily dose, routes, and the duration of rhuIL-2 treatment. Publication bias was evaluated by visual inspection of funnel plots and Egger's regression asymmetry test [41]. Diferences with p < 0.05 were considered statistically signifcant. Te RevMan (Version 5.1; Cochrane, Oxford, UK) and Stata software (version 12.0; Stata Corporation, College Station, TX) were used for the statistical analyses.

Safety
Outcomes. Te use of rhuIL-2 was not related to any severe adverse events which led to discontinuation of the treatment in any of the included studies. Results showed that rhuIL-2 was related to an increased risk of fever (RR: 2.46, 95% CI: 1.29 to 4.70, p � 0.006, I 2 � 0%; Figure 6(a)). Te incidence of other adverse events, such as musculoskeletal pain, hepatic injury, and renal toxicity, was not signifcantly diferent between groups (Figures 6(b)-6(d), p all >0.05).

Publication Bias.
Te funnel plots for the meta-analyses of the efect of rhuIL-2 on the outcomes of sputum culture conversion, sputum smear conversion, radiographic focus absorption, and cavity closure were symmetrical, suggesting low risk of publication biases (Figures 7(a)-7(d)). Egger's regression tests also suggested low risks of publication biases (p � 0.17, 0.35, 0.32, and 0.19, respectively). Te publication biases underlying the meta-analyses of adverse events were difcult to estimate because only 3 ∼ 8 studies were included for each outcome.

Discussion
In this systematic review and meta-analysis, we pooled the results of 18 available RCTs, and the results showed that adjunctive treatment with rhuIL-2 on the basis of standard chemotherapy for adult patients with pulmonary TB was associated with improved sputum bacterial elimination and improved radiographic changes. Subgroup analyses showed that the benefts of adjunctive rhuIL-2 remained only among patients with MDR-TB or recurrent TB. Besides, no serious adverse events related to the use of rhuIL-2 were reported. Te use of rhuIL-2 may increase the risk of fever, which is generally mild and could be adequately controlled after symptomatic treatment. Taken together, these fndings suggest that rhuIL-2 is an efective and safe adjunctive immunotherapy for patients with pulmonary TB who are treated with standard chemotherapy, which is associated with the improved microbiologic and radiographic outcomes.
To the best of our knowledge, only one previous metaanalysis evaluated the potential role of rhuIL-2 as adjunctive immunotherapy in patients with TB [42]. Although the results of the meta-analysis also suggested that rhuIL-2 may improve the sputum TB elimination in patients with TB, only four RCTs were included in the meta-analysis, and only two studies are available for the individual outcomes of sputum culture or smear conversion, which made the results  Journal of Tropical Medicine 5                      of the meta-analysis less convincing [42]. Besides, the previous meta-analysis failed to show that rhuIL-2 was associated with improved radiographic changes, and the safety of rhuIL-2 was also unable to be determined because of the limited studies available [42]. Our study has a few strengths in methodology as compared to the previous one. First, an extensive literature search was performed in fve electronic databases, which retrieved 18 RCTs for the subsequent meta-analysis. Te number of the overall included patients was much larger for the current meta-analysis as compared to that of the previous one (2630 versus 656). Second, in addition to the confrmed beneft of rhuIL-2 on sputum TB elimination by pooling 10 and 14 RCTs, respectively, the results of the meta-analysis also indicated that adjunctive rhuIL-2 on the basis of standard chemotherapy may also improve the radiographic changes of patients with TB, including the absorption of the pulmonary focus and the closure of cavities. Moreover, since patients with MDR-TB and recurrent TB have confrmed to be associated with worse prognosis than those with drug-susceptible TB and newly diagnosed TB [43,44], we performed a subgroup analysis to evaluate if the potential therapeutic efcacy of rhuIL-2 remained for these patients, and the fndings confrmed the consistent beneft of rhuIL-2 on microbiologic and radiographic outcomes even in patients with MDR-TB and recurrent TB. Finally, safety outcomes were also evaluated in this meta-analysis, and we found that no severe adverse events related to the use of rhuIL-2 were reported in any of the included studies, and additional treatment with rhuIL-2 only increased the incidence of fever without afecting the hepatic or renal adverse events in patients with pulmonary TB. Taken together, the fndings of the current meta-analysis suggested that adjunctive rhuIL-2 is efective and safe in patients with pulmonary TB. Te mechanisms underlying the potential therapeutic efcacy of adjunctive rhuIL-2 for patients with pulmonary TB are likely to be mainly dependent on the role of IL-2 for the restoration and stimulation of the innate immunity of the host against Mtb. In order for macrophages to kill mycobacteria, Mtb-specifc T lymphocytes are essential [45]. It is possible that a dysfunctional cell-mediated immune response to infection with Mtb can lead to the progression of the primary infection or to reactivation of TB [45]. Previous studies have shown that IL-2 produced by T1 cells is essential for the cellular immunity, which however was shown to decrease in patients with TB [46]. Correspondingly, a  subsequent study showed that a restored IL-2 level and a signifcantly elevated IL-2/IFN-c ratio may be a marker for the successful elimination of Mtb infection [47], suggesting the possible therapeutic implication of exogenous IL-2 for patients with TB. Consistently, a recent preclinical study in a mouse model of T cell dysfunction by persistent Mtb antigen stimulation found a signifcant decrease in IL-2 production, and the exogenous IL-2 administration restored antigenspecifc T cell responses and protective efcacy [48]. Moreover, a recent study suggested that defciency of IL-2 inducible T cell kinase may impair the early pulmonary protection against Mtb infection in mice probably due to the reduced endogenous IL-2 production [49]. Te molecular mechanisms underlying the benefts of rhuIL-2 for TB are to be investigated.
Te results of the subgroup analyses suggested that in the patients who achieved sputum smear conversion to negative, the rhuIL-2 signifcantly improved the treatment outcomes in patients with drug-susceptible TB. However, among the proportion of patients who achieved sputum culture conversion to negative, such kind of efects became nonsignifcant. Similarly, for the patients with newly diagnosed TB, the usage of rhuIL-2 signifcantly increased the odds of sputum smear conversion to negative, but the efect became nonsignifcant for sputum culture conversion to negative. Tis may be explained by the low sensitivity of sputum smear for the detection of TB. Indeed, a smear-positive result requires more acid-fast bacilli than sputum culture, and its sensitivity is limited to over 10,000 biological/ml in sputum [50]. Smears for acid-fast bacilli are afected by the specimen material, the patient's intermittent discharge of bacteria, the number of bacteria in the specimen, and many other factors, resulting in low sensitivity [51]. In addition, the results of the subgroup analyses showed that the benefts of rhuIL-2 on sputum TB elimination was mainly driven by studies of patients with MDR-TB and recurrent TB, and the favorable infuence of rhuIL-2 on some radiographic change may also be more remarkable in patients with MDR-TB and recurrent TB, such as the closure of pulmonary cavities. Tese fndings highlight an important role of adjunctive rhuIL-2 for patients with MDR-TB and recurrent TB, which is clinically important because anti-TB treatment in these patients is more challenging [52,53]. Te mechanisms are not fully determined. However, it could be hypothesized that the innate immunity of the host against Mtb may be impaired more severely in patients with MDR-TB and recurrent TB as compared to those with drug-susceptible and newly diagnosed TB. In fact, a previous study suggested a worse cellular immune function and a lower level of IL-2 in patients with recurrent TB as compared to those of newly diagnosed TB [54]. Future studies are warranted for further investigation.
Our study also has limitations. Firstly, although 18 RCTs were included in the meta-analysis, high-quality large-scale RCTs which evaluate the possible infuence of adjunctive rhuIL-2 on clinical outcomes in patients with pulmonary TB remain lacking. Moreover, for the most of the included studies, the follow-up durations are relatively short. Largescale RCTs with adequate follow-up durations are needed to determine the potential infuence of adjunctive rhuIL-2 on the risk of TB recurrence and mortality in these patients, as well as the long-term safety. Besides, the dose, route, and duration of rhuIL-2 administration varied among the included studies. Although results of meta-regression analyses failed to show that diference in these factors have signifcant infuences on the efcacies of rhuIL-2, these results should be interpreted with caution because of the limited available datasets for the analyses. Future studies are needed to determine the optimal regimens for adjunctive rhuIL-2 in patients with pulmonary TB. Finally, only patients with pulmonary TB were included in this meta-analysis. Future studies are needed to determine the possible therapeutic role of adjunctive rhuIL-2 for patients with extra-pulmonary TB.

Conclusions
In conclusion, the results of the meta-analysis indicate that rhuIL-2 is an efective adjunctive immunotherapy for patients with pulmonary TB, particularly for those with MDR-TB and recurrent TB. Large-scale clinical studies are needed to evaluate the infuence of adjunctive rhuIL-2 on long-term clinical prognosis and to determine the optimal regimen of rhuIL-2 for the treatment of patients with pulmonary TB.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.