Skip to main content
Download PDF
  • Systematic Review
  • Open access
  • Published:

CXCL9/CXCL10 as biomarkers the monitoring of treatment responses in Pulmonary TB patients: a systematic review and meta-analysis

BMC Infectious Diseases volume 24, Article number: 1037 (2024) Cite this article

Summary

Background

Tuberculosis (TB) remains a persistent threat to global public health and traditional treatment monitoring approaches are limited by their potential for contamination and need for timely evaluation. Therefore, new biomarkers are urgently required for monitoring the treatment efficacy of TB.

Methods

This study aimed to elucidate the levels of CXCL10 and CXCL9 in pulmonary TB patients who underwent anti-TB treatment. The data was acquired from five databases, including PubMed, Ovid, Web of Science, Embase, and the Cochrane Library. A meta-analysis of CXCL10 data from all time points was conducted. Furthermore, a trend meta-analysis of temporal data of CXCL10 and CXCL9 from multiple time points was also performed.

Results

It was revealed that patients who responded poorly to anti-TB treatment had higher serum levels relative to those who responded well (SMD: 1.23, 95% CI: -0.37–2.84) at the end of intensive treatment (2 months). Furthermore, heterogeneity was observed in these results, which might be because patients with a prior history of TB and different treatment monitoring methods than those selected in this study were also included. The analysis of alterations in CXCL10 and CXCL9 levels since the last collection time points indicated that their levels reduced with time.

Conclusion

In summary, the study revealed that reductions in CXCL10 levels during the first two months of anti-TB treatment are correlated with treatment responses. Furthermore, decreasing levels of CXCL9 during the treatment suggest that it may also serve as a biomarker with a similar value to CXCL10. Future in-depth studies are thus warranted to further probe the relevance of CXCL10 and CXCL9 in monitoring the treatment efficacy of TB.

Peer Review reports

Introduction

Tuberculosis (TB) is a mycobacterial infection caused by Mycobacterium tuberculosis. It is the second most prevalent cause of chronic respiratory disease in the world. The 2023 TB report published by the World Health Organization (WHO) [1] indicated that the global effects of the COVID-19 pandemic had adversely impacted TB patient diagnosis and treatment efforts in 2022, with treatment success rates of 88% and 63% for patients with drug-susceptible TB and multidrug-resistant/rifampicin-resistant TB, respectively. Techniques that can efficiently and early diagnose TB as well as monitor the treatment effect in patients are essential for ensuring efficacious treatment outcomes and preventing the emergence of resistant disease. However, currently, the monitoring of patients undergoing anti-TB treatment is a clinical challenge.

The WHO endorses sputum microscopy and culture as the primary approaches for monitoring TB patient treatment responses [2]. During the treatment, smears and cultures are routinely evaluated every month to assess patient responses to current treatment. Patients who exhibit persistent microbiological positivity at 2 months post-treatment evaluation are classified as slow responders, indicating a poor response to anti-TB therapy. Conversely, patients who achieve microbiological conversion within 2 months of treatment are classified as fast responders. However, these microbiology-based methods have some significant limitations, such as they markedly rely on the sputum sample collection, which is associated with a substantial risk of contamination from the oral cavity and upper respiratory tract. Furthermore, in some patients, the sputum collection process can be challenging such as in children. Moreover, the time-related restrictions and increased analytical timelines of sputum microscopy and culture methods also render them suboptimal for the rapid evaluation of patient treatment responses, thereby slowing the clinical decision-making for appropriate treatment regimens. Recently, several new promising biomarkers have been identified with high specificity and sensitivity for the monitoring of TB patient treatment outcomes, including RISK6 (mRNA signature 6-Marker), TB22 (mRNA signature 22-Marker), Lipoarabinomannan (LAM), etc [3,4,5,6,7].

There are certain inherent limitations to the use of monitoring methods that are reliant on particular omics approaches. While genomic markers can be extremely powerful, they are also very costly such that their widespread adoption in primary diagnostic and monitoring of a disease is infeasible, particularly in developing countries. The monitoring of the treatment outcomes in TB patients using transcriptomic data derived from different sequencing platforms or technologies can yield inconsistent data; therefore, larger sample sizes are necessary to establish universal standards. Thus, cost-effective and practical biomarkers for the time-sensitive monitoring of TB patient’s treatment efficacy are urgently required.

CXCL10 and CXCL9 are the members of the CXC chemokine family that are crucial regulators of cellular migration and inflammatory responses [8]. CXCL10 and CXCL9 exert their biological effects via CXCR3 signaling and trigger downstream immune activity [9, 10]. In pulmonary TB, CXCL9 and CXCL10 play crucial roles in recruiting chemotactic activated/effector cells to the site of TB infection, facilitating the formation of TB granulomas [11]. Their secretion is regulated by IFN-γ, which plays a key role in the immune response to TB. During effective anti-TB treatment, the pathogen load reduces, the inflammatory response in the body gradually weakens, and the level of IFN-γ decreases accordingly, which in turn reduces the levels of CXCL9 and CXCL10 [12, 13]. During TB progression, the patient’s serum CXCL9 and CXCL10 levels increase more than the healthy control individuals or those diagnosed with other pulmonary diseases [14,15,16]. Both CXCL10 and CXCL9 have been explored as auxiliary biomarkers to distinguish between patients with active and latent TB infection [12, 17, 18]. Although studies have indicated that the levels of CXCL9 and CXCL10 may guide clinical decision-making when monitoring patient treatment responses [19,20,21], there are only a few quantitative analyses on the relationship between CXCL9 and CXCL10 levels and anti-TB treatment response monitoring.

This systematic review and meta-analysis comprehensively analyzed CXCL9 and CXCL10 as a monitoring tool for anti-TB treatment outcomes. Furthermore, this study examines the relationship between CXCL9 and CXCL10 levels and treatment response in patients undergoing anti-TB therapy. The primary endpoint of the study was the microbiological and clinical outcome at 2 months post-treatment, while the secondary endpoint included longitudinal trends of CXCL9 and CXCL10 levels at various time points during treatment. Overall, this study aimed to address: (1) The correlation between CXCL9 and CXCL10 levels and the treatment response and (2) The longitudinal changes in CXCL9 and CXCL10 levels in in pulmonary TB patients throughout anti-TB treatment.

Methods

This study conducted a systematic review and meta-analysis of serum CXCL9 (MIG) and CXCL10 (IP-10) levels in pulmonary TB patients undergoing standard anti-TB treatment. The study selection criteria were based on the PRISMA checklist [22] (Table S1), and the PROSPERO (CRD42023480875) protocol was employed.

Search strategy

Studies were retrieved from the PubMed, Embase, Web of Science, Ovid, and Cochrane Library databases using the following search terms: (Tuberculosis or TB or Multidrug-Resistant Tuberculosis or tuberculosis Infection or XDR-TB OR LTBI) AND (Treatment Outcome or Efficacy or Treatment Effectiveness or Treatment response or monitor) AND (Chemokine CXCL9 or ‘Monokine Induced by gamma Interferon Chemokine’ or MIG or Small Inducible Cytokine B9 SCYB9 Chemokine) AND (Chemokine CXCL10 or CXCL10 or Cytokine IP-10 Protein or IP-10 OR ‘interferon-gamma-Inducible Protein of 10 kDa’).

Eligibility criteria

All studies published in English from January 1, 2005, to May 31, 2023, were eligible for inclusion in this meta-analysis. Inclusion criteria included (i) cross-sectional, cohort, and case-control as well as randomized controlled trials (RCTs), (ii) studies focused on changes in the levels of particular biomarkers during anti-TB treatment, (iii) studies on diagnosed TB patients, confirmed through bacteriological analysis, and (iv) studies with no age restrictions for patients but with a minimum of two-time points follow-up data during treatment. Exclusion criteria included: (i) studies on latent M. tuberculosis infections and non-tuberculous mycobacteria associated with other respiratory diseases, (ii) epidemiological studies, (iii) studies with patients who were only tested after initiating treatment, (iv) studies only focused on non-serum samples (such as pulmonary biopsy samples), (v) studies which analyzed stimulated CXCL10 levels, (vi) narrative or systematic reviews, meta-analyses, comments, conference abstracts, case reports, animal studies, editorials, or full-text articles that were not published in English, and vi) studies which employed non-standard therapeutic test references or acceptable reference standards including the culture of M. tuberculosis, Xpert MTB/RIF, smear microscopy, and clinical manifestations.

The patient’s recruitment was not restricted to any particular geographic region or type of healthcare system.

Study selection and risk of bias measurement

EndNote (version X9) was used to manage and evaluate all the selected studies, which were independently reviewed by two investigators (Z.Y. and J.Y.) who read the full text of potentially relevant articles and extracted the data. Of 267 studies, 251 were removed as duplicates after abstract/title and full-text review, while 16 were included for data extraction from full text. Two investigators (Z.Y. and J.Y.) independently scored all 16 articles using the QUADAS-2 tool [23] to assess their methodological quality. In case of any discrepancy, a third investigator (M.F.) resolved the issue after discussion and mutual consensus.

Data extraction

Before data extraction, two investigators (Z.W. and Y.C.) performed a thorough feasibility analysis of the established data extraction forms by randomly selecting the included studies. A third investigator (F.M.) was consulted in cases of disagreements between these authors. The data extraction form included: (i) basic information such as the origin country of the study, number of participants, baseline patient information, follow-up time points, and (ii) cytokine levels (average/median), as well as difference measurements (standard deviation, interval range) during each follow-up period. These data were independently extracted from the full text of the 16 included studies by two authors. In case no quantitative data were provided and could not be obtained after contacting the original authors, a third author (ZW) extracted graphical data from these studies. This approach was accepted only for extracting data from corresponding graphs [24, 25]. Extraction feasibility was assessed based on both article design reliability and graph extraction feasibility. When data were extracted from these figures, they were independently extracted by two investigators (Z.W. and Y.C.), and the average values were retained for analysis.

Data analysis

Following the data extraction and the assessments of central tendency and spread of these data at different follow-up time points, the extracted data were combined to produce standardized sample mean and standard deviation values and to assess corresponding trends. This study estimated normalized sample means and standard deviations using a data processing tool and employing the formulas provided by Wan et al. [26]. This approach integrates sample size with median, minimum, maximum, and/or interquartile range values to yield more accurate estimates. Since the follow-up times were varied in the studies, fold-change values for the analyzed biomarkers were measured relative to the previous collection time point, consistent with the report of Zimmer et al. [27].

To comprehensively analyze the changes in CXCL9/CXCL10 levels during treatment, this study employed two different analytical approaches. The first comprised a meta-analysis of all time points for CXCL9/CXCL10 levels, which was conducted using a random-effects model with a standardized mean difference (SMD) as the effect size. Patients who remained bacteriologically positive two months after anti-TB treatment (slow responders) were selected as the experimental groups. The control group comprised patients who were bacteriologically negative two months after initiating treatment (fast responders). The second approach of the study included a trend meta-analysis of CXCL9/CXCL10 levels during the treatment, which was conducted using a random intercept model. The estimated fold-change in CXCL9/CXCL10 levels was calculated with corresponding 95% confidence intervals (CIs). Covariance matrices and log-fold changes were calculated for each study through multivariate normal distribution simulations. Data were integrated using block diagonal matrix methods, and random intercept models were employed to analyze potential random effects. Furthermore, treatment responses for TB patients newly diagnosed using positive sputum smear test at the start of treatment were determined based on smear results before entrance into the treatment consolidation phase. Whereas, treatment responses for TB patients newly diagnosed using negative microscopic sputum smear test at the start of treatment were primarily based on risk factor scores or symptom scores.

The I2 statistic was employed to assess heterogeneity, and the outlier studies were identified using Galbraith plot analyses when assessing bias in research publications using the trim-and-fill method [28, 29]. A sensitivity analysis was also performed via an impact analysis by excluding or not excluding the study published by Annalisa et al. [30]. The heterogeneity in the literature was addressed through sensitivity analysis exclusion. For the meta-analysis of all time points, covariates such as patient age, history of tuberculosis, cytokine detection methods, and HIV history were introduced. Furthermore, a Meta-regression approach was employed to analyze sources of heterogeneity [31]. Significant differences were identified based on a two-sided p < 0.05. STATA 14.0 and R v4.2.2 were used for all statistical analyses [32].

Results

Study selection

Of 601 studies identified using the initial search strategy, 334 were duplicates and excluded from further examination. After a full-text review of the remaining 228 studies, 212 were excluded, while 16 were retained for analysis (Fig. 1) [30, 33,34,35,36,37,38,39,40,41,42,43,44,45,46,47]. Of the 212 excluded studies, 93 were excluded as they were not focused on treatment monitoring or other cytokines analysis, 42 were focused on non-TB mycobacteria or other respiratory diseases, and 29 were diagnostic tests or in vitro assays. Epidemiological analyses were also excluded from this study (n = 7). The remaining 16 studies were included in the quantitative synthesis and meta-analysis. Only 2 of these studies provided quantitative data on CXCL9 levels.

Fig. 1
figure 1

PRISMA flow chart for the study selection process

Full size image

Study characteristics

Of the 16 studies included in this meta-analysis, 14 were prospective analyses [38, 46]. The demographic characteristics of the patients from these 16 studies are presented in Table S2. These 14 studies were conducted in 8 different countries/regions, of which 14 were performed in a single country, while 2 were multicenter (bicentric) studies [44, 46]. Furthermore, 10 studies were performed in countries with moderate to high TB prevalence. Only one study included children [47]. Age ranges varied substantially among studies. Moreover, 5 studies did not report data related to HIV status, while ~ 50% of patients in one study were HIV-positive [44]. In 68% of the studies, the patient’s TB history was not documented. ELISAs were used to measure CXCL9/CXCL10 levels in 10 studies, while the remaining studies employed commercial Luminex kits or cell counting bead arrays (CBA). In addition, 7 studies classified patients as slow or fast responders when evaluating their reactions to conventional anti-TB treatment.

Quality and risk of bias assessment

The QUADAS-2 tool was employed to assess the potential risk of bias and quality of the included studies, with sources of potential bias classified into four distinct categories: “patient selection”, “index testing”, “reference standard” and “flow and timing” [23]. The majority of the included studies were prospective cohort studies, with 4 exhibiting a low risk of bias (Figure S2). Those studies in which healthy controls were included as a control group indicated a higher risk of patient selection bias and a generally higher overall risk of bias (N = 2) or a risk of bias classified as “unclear” (Figure S1). The generally higher overall risk of the two studies was primarily due to unclear referencing standards, procedural ambiguities, and timing issues. Most studies failed to directly observe or document patient’s compliance-related support measures. Just one study employed a double-blind approach when referencing criteria to interpret index results. Furthermore, 2 studies indicated a high degree of “flow and timing” risk, while the other studies revealed a generally low risk of bias (Fig. 2). As these were prospective cohort studies, all patients were subjected to the same criteria, and sample collection and processing were performed promptly.

Fig. 2
figure 2

QUADAS-2 risk of bias assessment results

Full size image

Comparisons between slow responders and fast responders

Since the design of most included studies varied, quantitative data on CXCL10 levels in slow and fast responders were only reported by 5 studies. Furthermore, CXCL9 data of slow and fast responders was only reported by 2 studies and therefore, was insufficient for any meta-analysis (Table S2). Comprehensive details of these 5 studies’ results are presented in Fig. 3. Elevated baseline (0 M) levels of CXCL10 were associated with a higher chance of a poor TB treatment response, with a pooled Standardized Mean Difference value of 0.45 (95% CI: 0.17–0.72). No significant differences were observed between the 2- and 6-month (2 M and 6 M) patient subgroups, which might be because of the high heterogeneity levels at these time points (2M: \(\:{I}^{2}\:\)= 96.3%, p-value < 0.001; 6M: \(\:{I}^{2}\:\)= 99.3%, p-value < 0.001).

A sensitivity analysis was conducted to explain the observed heterogeneity, which excluded the study by Annalisa (2005) as it had a significantly skewed effect size relative to other studies (Figure S2). Meta-regression results indicated that TB history was a source of heterogeneity (p = 0.016), whereas age, testing approach, and HIV were not the sources of significant heterogeneity (Figure S3). After excluding the study of Annalisa (2005), the overall heterogeneity decreased markedly (Figure S4). In the 2 M subgroup, high CXCL10 levels were linked to a greater chance of an unfavorable TB treatment response, with a pooled Standardized Mean Difference of 0.55 (95% CI: 0.02–1.08; \(\:{\text{I}}^{2}\) = 49.0%, p > 0.05). However, at 6 months, no significant differences were observed in CXCL10 levels between the slow and fast responders, with a pooled Standardized Mean Difference of 0 (95% CI: -0.52–0.51; \(\:{\text{I}}^{2}\) = 0.0%, p-value = 0.605).

Fig. 3
figure 3

Forest plots for SMD values comparing the levels of CXCL10 between slow and fast responders at 0, 2, and 6 months of treatment

CXCL10: IFN-gamma-Inducible Protein, 10 kDa; TB: tuberculosis; SMD: standardized mean difference

Full size image

CXCL9 and CXCL10 trend meta-analysis

The levels of CXCL9 and CXCL10 during the treatment were also evaluated through a trend meta-analysis, which compared these levels to previous time points. A meta-analysis of these results indicated that the fold-change values of CXCL9 and CXCL10 levels declined relative to previously recorded values (Table 1). Overall, this analysis included 11 and 4 studies focused on CXCL10 and CXCL9, with the fold change of -20.2 (95% CI: -56.4 to -16.6) and − 28.3 (95% CI: -40.1 to -16.7), respectively. The maximal fold-change values for CXCL10 and CXCL9 were − 69.5% and − 64.9%, respectively (Figure S5-6) and both the biomarkers indicated narrow confidence intervals.

Table 1 Combined sensitivity analysis (ρ) of weekly fold-change meta-regression and pooled fold-change data pertaining to the serum levels of CXCL9 and CXCL10 during anti-TB treatment
Full size table

Discussion

The objective of this study was to evaluate the levels of CXCL9 and CXCL10 as a standardized monitoring tool for anti-TB treatment responses. CXCL10 is a serological marker and has been investigated in studies focused on the diagnosis, clinical evaluation, and therapeutic monitoring of TB patients [15, 17, 48,49,50,51,52]. There have been several reports demonstrating the diagnostic performance of CXCL9 and CXCL10. For example, Sampath et al. [53] analyzed latent TB, drug-resistant TB (DR-TB), and drug-sensitive TB (DS-TB) patients and compared the differences in CXCL10 and CXCL9 levels between drug-resistant and drug-sensitive patients. DR-TB patients (CXCL9, median: 205.99; CXCL10, median: 1205.75) exhibited significantly increased levels of CXCL9 and CXCL10 compared to DS-TB patients (CXCL9, median: 134.48; CXCL10, median: 650.50). CXCL9 (AUC = 0.82, p < 0.0001) and CXCL10 (AUC = 0.84, p < 0.0001) effectively discriminate DR-TB from DS-TB. The results indicated that these chemokines may differentiate between disease stages. CXCL9 and CXCL10 are both CXC family chemokines that synergistically regulate host immune responses; however, there are only a few studies that have quantitatively assessed their role in therapeutic monitoring [54]. Therefore, this study analyzed these chemokines as biomarkers, and although the results are promising, further comprehensive biomarker-focused studies are required to improve timely and effective treatment monitoring for TB patients.

Quantitative meta-analysis results indicated that the levels of CXCL9 and CXCL10, as measured based on the mean fold-change in chemokine levels, decreased over the course of anti-TB treatment relative to baseline levels. Furthermore, CXCL10 levels were significantly higher in slow responders as compared to fast responders at 2 months of treatment (SMD: 0.55, 95% CI: 0.02–1.08). These results suggest that the level of CXCL10 could serve as an indicator to detect whether microbiological reversal occurs in patients. Analyses of mean fold-change relative to baseline confirmed that both CXCL9 and CXCL10 serum concentrations reduced during intensive treatment phases, with mean fold-changes of -20.2 (95% CI: -56.4 to -16.6) and − 28.3 (95% CI: -40.1 to -16.7), respectively, compared to prior analytical time points. Our analysis indicates that CXCL10 levels were not significantly different between slow responders and fast responders at 6 months of treatment. Additionally, there were no discernible differences in microbiological or clinical evaluations between fast and slow responders at 6 months of treatment. This observation is consistent across the included studies and supports the notion that CXCL10 levels, as reported in our study, are concordant with the microbiological examination results. Changes in the levels of CXCL10 and CXCL9 over the course of treatment in individual studies are presented in Figure S5 and Figure S6. CXCL10 demonstrating downward trends with time as compared to baseline, whereas CXCL9 levels tend to increase over time. However, these results are controversial. For example, Chung et al. [37] observed inconsistent increases or decreases in serum CXCL9 levels in slow and fast responders after the treatment, complicating the interpretation of CXCL9’s clinical relevance in TB treatment monitoring.

Here, the patient’s serum, blood, and/or plasma levels of CXCL9 and CXCL10 were analyzed, which indicates that easy analysis with minimal attendant risk of biohazard exposure or contamination from sample processing. This is a clear advantage over more traditional TB therapeutic monitoring strategies, which often entail sputum culture evaluation [55]. However, the magnitude of change in CXCL9 levels over the course of treatment was limited in this analysis, and substantial heterogeneity was detected among studies, hampering adequate analysis of CXCL9’s role in the detection and monitoring of TB.

To date, several studies have indicated novel approaches to monitor TB treatment, including serum-based transcriptomic analyses, metabolomics strategies, and the use of new imaging technologies to evaluate clinical signs and symptoms [56,57,58,59,60,61,62,63,64,65]. These approaches hold great promise as do not require sputum sample processing, specifically in a research field [66]. Sigal et al., for example, screened 70 infection, metabolism, and inflammation-related markers in serum samples collected from 319 pulmonary TB patients [67]. They revealed that the levels of SAA1, PCT, IL-1β, IL-6, CRP, PTX-3, and MMP-8 were strongly correlated with disease severity and early treatment response. However, the data obtained from patients is highly heterogeneous, which interferes with the implementation of similar testing in centralized laboratories and limits effective disease treatment via these approaches [68]. The sputum samples may be not required when CXCL9 and CXCL10 are used as therapeutic biomarkers in anti-TB treatment. However, the evidence is currently insufficient to confirm their ability to detect microbiological reversals in patients. The existing literature exhibits substantial variability in research methodologies and designs, complicating result comparisons and integration, and presenting challenges in explaining heterogeneity. Establishing standardized study designs is crucial for future investigations to comprehensively analyze the biomarker properties of CXCL9 and CXCL10. Currently, only one metabolomics analysis on 48 TB patients, 20 TB-DM patients, and 48 non-HIV-infected healthy controls has proposed that analyzing the shifts in metabolic activity during anti-TB treatment has been proposed as a viable treatment monitoring approach [69]. Based on the observed decrease in metabolite levels during the treatment, the authors developed a model (AUC = 0.91–0.97) capable of readily differentiating between these three treatment groups of patients. In this study, slow and fast responders were used for the indirect assessment of treatment response, since most clinical efficacy endpoints in the analyzed studies did not include disease recurrence. Currently, a significant concern is the standardization of clinical efficacy.

There are several limitations in this analysis. First, effective meta-analyses of longitudinal data markedly depend on the study’s experimental design, potentially contributing to varying levels of bias and possible gaps in the data. To minimize time-related variability, meta-analyses of all time points and trends were included in this study, focusing on fold-change values for CXCL9 and CXCL10 levels in treated TB patients. However, as observed by Zimmer et al. [27], these analytical outcomes depend on the data derived from the included studies, introducing a high risk of bias. Secondly, in this quantitative meta-analysis, data extraction was limited by the lack of data in the form of charts or numerical values in the analyzed studies. The extraction of data from graphs can introduce bias, although prior studies suggest that the overall magnitude of such bias is minor. Although efforts were made to minimize this source of bias, some subjectivity in result interpretation may persist. Third, relatively few relevant articles on CXCL9 were included in this study; therefore, CXCL10 was included in both meta-analyses, while CXCL9 was only included in the trend meta-analysis. Lastly, although limited efforts were made to assess the potential sources of heterogeneity, a wide range of complex factors can contribute to the incidence of heterogeneity. Furthermore, the overall quality of patient screening data in the included studies was not uniform, which might be another major source of potential bias.

Prompt assessment of the patient’s treatment response facilitates the formulation of a tailored treatment regimen based on individual circumstances. Fast responders may benefit from regimen adjustments such as shortened treatment duration and reduced drug dosage, which can mitigate the risk of drug-induced hepatotoxicity and serious adverse effects. Whereas, slow responders may indicate inadequate response to anti-TB therapy, indicating an increased risk of drug-resistant TB. Clinicians can rapidly evaluate patient drug resistance, enabling them to review and adjust treatment plans effectively. This includes optimizing drug combinations and dosages in current treatments. Clinicians should consider incorporating second-line drugs such as bedaquiline and fluoroquinolones. If drug resistance is confirmed, they should transition to a regimen recommended for MDR-TB or XDR-TB. Although this study offers new insight that may help guide clinical treatment planning for TB patients, only CXCL9 and CXCL10 were assessed, and the potential performance of other biomarkers warrants further research.

In conclusion, based on the current research in this field, further studies are required to assess the efficacy of utilizing specific chemokines as biomarkers to monitor TB patient treatment responses as there are no established clinical guidelines on their use. This systematic review and meta-analysis explored the potential of CXCL10 and CXCL9 levels as biomarkers for monitoring the treatment response in TB patients and provided a foundation to guide further research efforts in this field.

Data availability

No datasets were generated or analysed during the current study.

References

  1. World Health Organization. Global tuberculosis report 2023. Geneva, Switzerland: WHO; 2023.

    Google Scholar 

  2. Treatment of Tuberculosis. Guidelines. 4th edition. Geneva: World Health Organization; 2010.

  3. Drain PK, Gounder L, Grobler A, Sahid F, Bassett IV, Moosa M-YS. Urine lipoarabinomannan to monitor antituberculosis therapy response and predict mortality in an HIV-endemic region: a prospective cohort study. BMJ Open. 2015;5:e006833.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Heyckendorf J, Marwitz S, Reimann M, Avsar K, DiNardo AR, Günther G, et al. Prediction of anti-tuberculosis treatment duration based on a 22-gene transcriptomic model. Eur Respir J. 2021;58:2003492.

    Article  CAS  PubMed  Google Scholar 

  5. Kawasaki M, Echiverri C, Raymond L, Cadena E, Reside E, Gler MT, et al. Lipoarabinomannan in sputum to detect bacterial load and treatment response in patients with pulmonary tuberculosis: Analytic validation and evaluation in two cohorts. PLoS Med. 2019;16:e1002780.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Penn-Nicholson A, Mbandi SK, Thompson E, Mendelsohn SC, Suliman S, Chegou NN, et al. RISK6, a 6-gene transcriptomic signature of TB disease risk, diagnosis and treatment response. Sci Rep. 2020;10:8629.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Zak DE, Penn-Nicholson A, Scriba TJ, Thompson E, Suliman S, Amon LM, et al. A blood RNA signature for tuberculosis disease risk: a prospective cohort study. Lancet. 2016;387:2312–22.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Loetscher M, Gerber B, Loetscher P, Jones SA, Piali L, Clark-Lewis I, et al. Chemokine receptor specific for IP10 and mig: structure, function, and expression in activated T-lymphocytes. J Exp Med. 1996;184:963–9.

    Article  CAS  PubMed  Google Scholar 

  9. Loetscher P, Pellegrino A, Gong JH, Mattioli I, Loetscher M, Bardi G, et al. The ligands of CXC chemokine receptor 3, I-TAC, Mig, and IP10, are natural antagonists for CCR3. J Biol Chem. 2001;276:2986–91.

    Article  CAS  PubMed  Google Scholar 

  10. Piali L, Weber C, LaRosa G, Mackay CR, Springer TA, Clark-Lewis I, et al. The chemokine receptor CXCR3 mediates rapid and shear-resistant adhesion-induction of effector T lymphocytes by the chemokines IP10 and Mig. Eur J Immunol. 1998;28:961–72.

    Article  CAS  PubMed  Google Scholar 

  11. Lande R, Giacomini E, Grassi T, Remoli ME, Iona E, Miettinen M, et al. IFN-alpha beta released by Mycobacterium tuberculosis-infected human dendritic cells induces the expression of CXCL10: selective recruitment of NK and activated T cells. J Immunol. 2003;170:1174–82.

    Article  CAS  PubMed  Google Scholar 

  12. Chegou NN, Heyckendorf J, Walzl G, Lange C, Ruhwald M. Beyond the IFN-γ horizon: biomarkers for immunodiagnosis of infection with Mycobacterium tuberculosis. Eur Respir J. 2014;43:1472–86.

    Article  CAS  PubMed  Google Scholar 

  13. Azzurri A, Sow OY, Amedei A, Bah B, Diallo S, Peri G, et al. IFN-gamma-inducible protein 10 and pentraxin 3 plasma levels are tools for monitoring inflammation and disease activity in Mycobacterium tuberculosis infection. Microbes Infect. 2005;7:1–8.

    Article  CAS  PubMed  Google Scholar 

  14. Djoba Siawaya JF, Beyers N, van Helden P, Walzl G. Differential cytokine secretion and early treatment response in patients with pulmonary tuberculosis. Clin Exp Immunol. 2009;156:69–77.

    Article  CAS  PubMed  Google Scholar 

  15. Yu Y, Zhang Y, Hu S, Jin D, Chen X, Jin Q, et al. Different patterns of cytokines and chemokines combined with IFN-γ production reflect Mycobacterium tuberculosis infection and disease. PLoS ONE. 2012;7:e44944.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Wang X, Jiang J, Cao Z, Yang B, Zhang J, Cheng X. Diagnostic performance of multiplex cytokine and chemokine assay for tuberculosis. Tuberculosis. 2012;92:513–20.

    Article  CAS  PubMed  Google Scholar 

  17. Tebruegge M, Dutta B, Donath S, Ritz N, Forbes B, Camacho-Badilla K, et al. Mycobacteria-specific cytokine responses detect tuberculosis infection and distinguish latent from active tuberculosis. Am J Respir Crit Care Med. 2015;192:485–99.

    Article  CAS  PubMed  Google Scholar 

  18. Carrère-Kremer S, Kolia-Diafouka P, Pisoni A, Bolloré K, Peries M, Godreuil S, et al. QuantiFERON-TB gold plus assay in patients with latent vs. active tuberculosis in a low incidence setting: level of IFN-γ, CD4/CD8 responses, and release of IL-2, IP-10, and MIG. Front Microbiol. 2022;13:825021.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Chavez K, Ravindran R, Dehnad A, Khan IH. Gender biased immune-biomarkers in active tuberculosis and correlation of their profiles to efficacy of therapy. Tuberculosis (Edinb). 2016;99:17–24.

    Article  CAS  PubMed  Google Scholar 

  20. Yang Q, Cai Y, Zhao W, Wu F, Zhang M, Luo K, et al. IP-10 and MIG are compartmentalized at the site of disease during pleural and meningeal tuberculosis and are decreased after antituberculosis treatment. Clin Vaccine Immunol. 2014;21:1635–44.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Kumar NP, Moideen K, Nancy A, Viswanathan V, Shruthi BS, Sivakumar S, et al. Plasma chemokines are biomarkers of disease severity, higher bacterial burden and delayed sputum culture conversion in pulmonary tuberculosis. Sci Rep. 2019;9:18217.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Whiting PF, Rutjes AWS, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155:529–36.

    Article  PubMed  Google Scholar 

  24. Turner SL, Korevaar E, Cumpston MS, Kanukula R, Forbes AB, McKenzie JE. Effect estimates can be accurately calculated with data digitally extracted from interrupted time series graphs. Res Synth Methods. 2023;14:622–38.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Van der Mierden S, Spineli LM, Talbot SR, Yiannakou C, Zentrich E, Weegh N, et al. Extracting data from graphs: a case-study on animal research with implications for meta-analyses. Res Synth Methods. 2021;12:701–10.

    Article  PubMed  Google Scholar 

  26. Wan X, Wang W, Liu J, Tong T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14:135.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Zimmer AJ, Lainati F, Aguilera Vasquez N, Chedid C, McGrath S, Benedetti A, et al. Biomarkers that correlate with active pulmonary tuberculosis treatment response: a systematic review and Meta-analysis. J Clin Microbiol. 2022;60:e0185921.

    Article  PubMed  Google Scholar 

  28. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–60.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Lin L, Chu H. Quantifying publication bias in meta-analysis. Biometrics. 2018;74:785–94.

    Article  PubMed  Google Scholar 

  30. Azzurri A, Sow OY, Amedei A, Bah B, Diallo S, Peri G, et al. IFN-gamma-inducible protein 10 and pentraxin 3 plasma levels are tools for monitoring inflammation and disease activity in Mycobacterium tuberculosis infection. Microbes Infect. 2005;7(1):1–8.

    Article  CAS  PubMed  Google Scholar 

  31. Jackson D, Riley RD. A refined method for multivariate meta-analysis and meta-regression. Stat Med. 2014;33:541–54.

    Article  PubMed  Google Scholar 

  32. Viechtbauer W. Conducting Meta-analyses in R with the metafor Package. J Stat Soft. 2010;36:1–48.

    Article  Google Scholar 

  33. Alessandri AL, Souza AL, Oliveira SC, Macedo GC, Teixeira MM, Teixeira AL. Concentrations of CXCL8, CXCL9 and sTNFR1 in plasma of patients with pulmonary tuberculosis undergoing treatment. Inflamm Res. 2006;55:528–33.

    Article  CAS  PubMed  Google Scholar 

  34. Bai XJ, Li HM, Yang YR, Zhang JX, Liang Y, Wu XQ. Cytokine and soluble adhesion molecule profiles and biomarkers for treatment monitoring in re-treated smear-positive patients with pulmonary tuberculosis. Cytokine. 2018;108:9–16.

    Article  CAS  PubMed  Google Scholar 

  35. Chen Y-C, Chin C-H, Liu S-F, Wu C-C, Tsen C-C, Wang Y-H, et al. Prognostic values of serum IP-10 and IL-17 in patients with pulmonary tuberculosis. Dis Markers. 2011;31:101–10.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Chung W, Lee K, Jung Y, Kim Y, Park J, Sheen S, et al. Serum CXCR3 ligands as biomarkers for the diagnosis and treatment monitoring of tuberculosis. Int J Tuberculosis Lung Disease. 2015;19(12):1476–84.

    Article  CAS  Google Scholar 

  37. Chung WY, Yoon D, Lee KS, Jung YJ, Kim YS, Sheen SS et al. The usefulness of serum CXCR3 ligands for evaluating the early treatment response in tuberculosis. Med (United States). 2016;95(17).

  38. Hong JY, Lee HJ, Kim SY, Chung KS, Kim EY, Jung JY, et al. Efficacy of IP-10 as a biomarker for monitoring tuberculosis treatment. J Infect. 2014;68:252–p258.

    Article  PubMed  Google Scholar 

  39. Jayakumar A, Vittinghoff E, Segal MR, MacKenzie WR, Johnson JL, Gitta P, et al. Serum biomarkers of treatment response within a randomized clinical trial for pulmonary tuberculosis. Tuberculosis. 2015;95:415–20.

    Article  CAS  PubMed  Google Scholar 

  40. Kim SY, Kim J, Kim DR, Kang YA, Bong S, Lee J et al. Urine IP-10 as a biomarker of therapeutic response in patients with active pulmonary tuberculosis. BMC Infect Dis. 2018;18(1).

  41. Lee MR, Tsai CJ, Wang WJ, Chuang TY, Yang CM, Chang LY, et al. Plasma biomarkers can predict treatment response in tuberculosis patients: a prospective observational study. Med (United States). 2015;94(39):e1628.

    CAS  Google Scholar 

  42. Namuganga AR, Ssentalo Bagaya B, Chegou NN, Mayanja-Kizza H. Serum biomarkers for monitoring response to tuberculosis treatment: an assessment of the effect of different covariates among slow and fast treatment responders. Biomarkers. 2023.

  43. Pedersen JL, Barry SE, Bokil NJ, Ellis M, Yang Y, Guan G et al. High sensitivity and specificity of a 5-analyte protein and microRNA biosignature for identification of active tuberculosis. Clin Translational Immunol. 2021;10(6).

  44. Riou C, Peixoto BP, Roberts L, Ronacher K, Walzl G, Manca C et al. Effect of Standard Tuberculosis Treatment on Plasma Cytokine Levels in patients with active pulmonary tuberculosis. PLoS ONE. 2012;7.

  45. Stefanescu S, Cocos R, Turcu-Stiolica A, Shelby ES, Matei M, Subtirelu MS et al. Prediction of treatment outcome with inflammatory biomarkers after 2 months of therapy in pulmonary tuberculosis patients: preliminary results. Pathogens (Basel Switzerland). 2021;10(7).

  46. Tonby K, Ruhwald M, Kvale D, Dyrhol-Riise AM. IP-10 measured by dry plasma spots as biomarker for therapy responses in Mycobacterium Tuberculosis infection. Sci Rep. 2015;5.

  47. Zhao Y, Yang X, Zhang X, Yu Q, Zhao P, Wang J, et al. IP-10 and RANTES as biomarkers for pulmonary tuberculosis diagnosis and monitoring. Tuberculosis. 2018;111:45–53.

    Article  CAS  PubMed  Google Scholar 

  48. Peruhype-Magalhães V, de Araújo FF, de Morais Papini TF, Wendling APB, Campi-Azevedo AC, Coelho-Dos-Reis JG, et al. Serum biomarkers in patients with unilateral or bilateral active pulmonary tuberculosis: immunological networks and promising diagnostic applications. Cytokine. 2023;162:156076.

    Article  PubMed  Google Scholar 

  49. Fisher KL, Moodley D, Rajkumar-Bhugeloo K, Baiyegunhi OO, Karim F, Ndlovu H, et al. Elevated IP-10 at the protein and Gene Level Associates with Pulmonary TB. Front Cell Infect Microbiol. 2022;12:908144.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Mann TN, Warwick J, Chegou NN, Davis JH, Beltran CGG, Griffith-Richards S, et al. Biomarkers to predict FDG PET/CT activity after the standard duration of treatment for spinal tuberculosis: an exploratory study. Tuberculosis (Edinb). 2021;129:102107.

    Article  CAS  PubMed  Google Scholar 

  51. Kumar NP, Moideen K, Nancy A, Viswanathan V, Thiruvengadam K, Nair D, et al. Plasma chemokines are baseline predictors of unfavorable treatment outcomes in Pulmonary Tuberculosis. Clin Infect Dis. 2021;73:e3419–27.

    Article  PubMed  Google Scholar 

  52. Hur Y-G, Kang YA, Jang S-H, Hong JY, Kim A, Lee SA, et al. Adjunctive biomarkers for improving diagnosis of tuberculosis and monitoring therapeutic effects. J Infect. 2015;70:346–55.

    Article  PubMed  Google Scholar 

  53. Sampath P, Rajamanickam A, Thiruvengadam K, Natarajan AP, Hissar S, Dhanapal M, et al. Plasma chemokines CXCL10 and CXCL9 as potential diagnostic markers of drug-sensitive and drug-resistant tuberculosis. Sci Rep. 2023;13:7404.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Farber JM. Mig and IP-10: CXC chemokines that target lymphocytes. J Leukoc Biol. 1997;61:246–57.

    Article  CAS  PubMed  Google Scholar 

  55. Heyckendorf J, Georghiou SB, Frahm N, Heinrich N, Kontsevaya I, Reimann M, et al. Tuberculosis treatment monitoring and outcome measures: New Interest and New Strategies. Clin Microbiol Rev. 2022;35:e0022721.

    Article  PubMed  Google Scholar 

  56. Scriba TJ, Fiore-Gartland A, Penn-Nicholson A, Mulenga H, Mbandi SK, Borate B, et al. Biomarker-guided tuberculosis preventive therapy (CORTIS): a randomised controlled trial. Lancet Infect Dis. 2021;21:354–65.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Hoang LT, Jain P, Pillay TD, Tolosa-Wright M, Niazi U, Takwoingi Y, et al. Transcriptomic signatures for diagnosing tuberculosis in clinical practice: a prospective, multicentre cohort study. Lancet Infect Dis. 2021;21:366–75.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Berry MPR, Graham CM, McNab FW, Xu Z, Bloch SAA, Oni T, et al. An interferon-inducible neutrophil-driven blood transcriptional signature in human tuberculosis. Nature. 2010;466:973–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Blankley S, Graham CM, Levin J, Turner J, Berry MPR, Bloom CI, et al. A 380-gene meta-signature of active tuberculosis compared with healthy controls. Eur Respir J. 2016;47:1873–6.

    Article  PubMed  PubMed Central  Google Scholar 

  60. Odia T, Malherbe ST, Meier S, Maasdorp E, Kleynhans L, du Plessis N et al. The peripheral blood transcriptome is correlated with PET measures of lung inflammation during successful tuberculosis treatment. Front Immunol. 2021;11.

  61. Jaeger S, Karargyris A, Candemir S, Siegelman J, Folio L, Antani S, et al. Automatic screening for tuberculosis in chest radiographs: a survey. Quant Imaging Med Surg. 2013;3:89–99.

    PubMed  PubMed Central  Google Scholar 

  62. Murthy SE, Chatterjee F, Crook A, Dawson R, Mendel C, Murphy ME, et al. Pretreatment chest x-ray severity and its relation to bacterial burden in smear positive pulmonary tuberculosis. BMC Med. 2018;16:73.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. te Riele JB, Buser V, Calligaro G, Esmail A, Theron G, Lesosky M, et al. Relationship between chest radiographic characteristics, sputum bacterial load, and treatment outcomes in patients with extensively drug-resistant tuberculosis. Int J Infect Dis. 2019;79:65–71.

    Article  Google Scholar 

  64. Imperial MZ, Phillips PPJ, Nahid P, Savic RM. Precision-Enhancing Risk Stratification Tools for selecting Optimal Treatment durations in Tuberculosis clinical trials. Am J Respir Crit Care Med. 2021;204:1086–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Rockwood N, du Bruyn E, Morris T, Wilkinson RJ. Assessment of treatment response in tuberculosis. Expert Rev Respir Med. 2016;10:643–54.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  66. Sweeney TE, Braviak L, Tato CM, Khatri P. Genome-wide expression for diagnosis of pulmonary tuberculosis: a multicohort analysis. Lancet Respir Med. 2016;4:213–24.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  67. Sigal GB, Segal MR, Mathew A, Jarlsberg L, Wang M, Barbero S, et al. Biomarkers of tuberculosis severity and treatment effect: a Directed screen of 70 Host Markers in a Randomized Clinical Trial. EBioMedicine. 2017;25:112–21.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  68. Gupta-Wright A, den Boon S, MacLean EL, Cirillo D, Cobelens F, Gillespie SH, et al. Target product profiles: tests for tuberculosis treatment monitoring and optimization. Bull World Health Organ. 2023;101:730–7.

    Article  PubMed  PubMed Central  Google Scholar 

  69. Vrieling F, Alisjahbana B, Sahiratmadja E, van Crevel R, Harms AC, Hankemeier T, et al. Plasma metabolomics in tuberculosis patients with and without concurrent type 2 diabetes at diagnosis and during antibiotic treatment. Sci Rep. 2019;9:18669.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Funding

This work is supported by the National Key Research and Development Program of China (2022YFC2304800), Guangzhou Science and Technology Planning Project (202002030152, 202102020910, 202206010134, 202201010697, 2023A03J0539, 2023A03J0992), Guangdong Natural Science Foundation Project (2018A030313550, 2023A1515010461), Guangzhou Health Science and Technology Project (20231A011051, 20241A011049), Guangdong Bureau of Traditional Chinese Medicine Scientific Research Project (20231251), Guangdong Medical Science and Technology Research Fund Project (A2023284).

Author information

Author notes

  1. Zeyou Wei and Yuanjin Chen contributed equally to this work and share first authorship.

Authors and Affiliations

  1. State Key Laboratory of Respiratory Disease, Guangzhou Key Laboratory of Tuberculosis Research, Institute of Pulmonary Diseases, Guangzhou Chest Hospital, Institute of Tuberculosis, Guangzhou Medical University, 62 Hengzhigang Rd, Yuexiu District, Guangzhou, 510095, People’s Republic of China

    Zeyou Wei, Yuanjin Chen, Pengyan Dong, Zhihui Liu, Xiaomin Lai, Nan Wang, Hua Li, Qi Wang, Yu Yang & Fanrong Meng

  2. School of Public Health, Sun Yat-sen University, Shen Zhen, China

    Xiaomin Lai

  3. State Key Laboratory of Respiratory Disease, Guangzhou Key Laboratory of Tuberculosis Research, Department of Tuberculosis, Guangzhou Chest Hospital, Institute of Tuberculosis, Guangzhou Medical University, Guangzhou, P.R. China

    Lan Tao

  4. State Key Laboratory of Respiratory Disease, Guangzhou Key Laboratory of Tuberculosis Research, Department of Oncology, Guangzhou Chest Hospital, Institute of Tuberculosis, Guangzhou Medical University, Guangzhou, P.R. China

    Ning Su

Authors
  1. Zeyou Wei
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Yuanjin Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Pengyan Dong
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Zhihui Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Xiaomin Lai
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Nan Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Hua Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Qi Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Lan Tao
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Ning Su
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Yu Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Fanrong Meng
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Author contributions: ZW, YC, FM, and YY conceived and designed the study, undertook the review and data abstraction, and drafted the article. PD and YC designed the search strategy. ZW performed statistical analysis. All authors commented on and revised the article.

Corresponding authors

Correspondence to Yu Yang or Fanrong Meng.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wei, Z., Chen, Y., Dong, P. et al. CXCL9/CXCL10 as biomarkers the monitoring of treatment responses in Pulmonary TB patients: a systematic review and meta-analysis. BMC Infect Dis 24, 1037 (2024). https://doi.org/10.1186/s12879-024-09939-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12879-024-09939-0

Keywords

BMC Infectious Diseases

ISSN: 1471-2334

Contact us