You are here

  • Home
  • Archive
  • Volume 79, Issue 2
  • Association of indicators of extensive disease and rifampin-resistant tuberculosis treatment outcomes: an individual participant data meta-analysis

Article Text

Article menu
Tuberculosis
Original research
Association of indicators of extensive disease and rifampin-resistant tuberculosis treatment outcomes: an individual participant data meta-analysis
Free
  1. Jonathon R Campbell1,2,3,
  2. Sarah K Brode4,5,
  3. Pennan Barry6,
  4. Mayara Lisboa Bastos1,
  5. Maryline Bonnet7,
  6. Lorenzo Guglielmetti8,
  7. Russell Kempker9,
  8. Dzmitry Klimuk10,
  9. Rafael Laniado Laborín11,
  10. Vladimir Milanov12,
  11. Rupak Singla13,
  12. Alena Skrahina10,
  13. Anete Trajman3,14,
  14. Tjip S van der Werf15,
  15. Piret Viiklepp16,
  16. Dick Menzies1,2,3
  1. 1 Faculty of Medicine and Health Sciences, McGill University, Montreal, Quebec, Canada
  2. 2 Respiratory Epidemiology and Clinical Research Unit, Research Institute of the McGill University Health Centre, Montreal, Quebec, Canada
  3. 3 Montreal Chest Institute & McGill International TB Centre, McGill University, Montreal, Quebec, Canada
  4. 4 West Park Healthcare Centre, Toronto, Ontario, Canada
  5. 5 Department of Medicine, University of Toronto, Toronto, Ontario, Canada
  6. 6 Tuberculosis Control Branch, California Department of Public Health, Richmond, California, USA
  7. 7 University of Montpellier, Montpellier, France
  8. 8 Immunology and Infectious Diseasese, Sorbonne Universite, Paris, France
  9. 9 Department of Medicine, Emory University School of Medicine, Atlanta, Georgia, USA
  10. 10 Republican Scientific and Practical Centre for Pulmonology and Tuberculosis, Minsk, Belarus
  11. 11 Facultad de Medicina, Universidad Autónoma de Baja California, Mexicali, Mexico
  12. 12 Occupational Diseases, Medical University-Sofia, Sofia, Bulgaria
  13. 13 Tuberculosis and Respiratory Diseases, National Institute of Tuberculosis and Respiratory Diseases, New Delhi, India
  14. 14 Department of Internal Medicine, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
  15. 15 Departments of Internal Medicine, Infectious Diseases, Pulmonary Diseases, and Tuberculosis, UMC Groningen, Groningen, The Netherlands
  16. 16 Department of Registries, National Institute for Health Development, Tallinn, Estonia
  1. Correspondence to Dr Jonathon R Campbell, McGill University, Montreal, Canada; jonathon.campbell{at}mcgill.ca

Abstract

Background Indicators of extensive disease—acid fast bacilli (AFB) smear positivity and lung cavitation—have been inconsistently associated with clinical rifampin-resistant/multidrug-resistant tuberculosis (RR/MDR-TB) outcomes. We evaluated the association of these indicators with end-of-treatment outcomes.

Methods We did an individual participant data meta-analysis of people treated for RR/MDR-TB with longer regimens with documented AFB smear and chest radiography findings. We compared people AFB smear-negative without cavities to people: (1) smear-negative with lung cavities; (2) smear-positive without lung cavities and (3) AFB smear-positive with lung cavities. Using multivariable logistic regression accounting for demographic, treatment and clinical factors, we calculated adjusted ORs (aOR) for any unfavourable outcome (death, lost to follow-up, failure/recurrence), and mortality and treatment failure/recurrence alone.

Results We included 5596 participants; included participants significantly differed from excluded participants. Overall, 774 (13.8%) were AFB smear-negative without cavities, 647 (11.6%) only had cavities, 1424 (25.4%) were AFB smear-positive alone and 2751 (49.2%) were AFB smear-positive with cavities. The median age was 37 years (IQR: 28–47), 3580 (64%) were male and 686 (12.5%) had HIV. Compared with participants AFB smear-negative without cavities, aOR (95% CI) for any unfavourable outcome was 1.0 (0.8 to 1.4) for participants smear-negative with lung cavities, 1.2 (0.9 to 1.5) if smear-positive without cavities and 1.6 (1.3 to 2.0) if AFB smear-positive with lung cavities. Odds were only significantly increased for mortality (1.5, 95% CI 1.1 to 2.1) and failure/recurrence (2.2, 95% CI 1.5 to 3.3) among participants AFB smear-positive with lung cavities.

Conclusion Only the combination of AFB smear-positivity and lung cavitation was associated with unfavourable outcomes, suggesting they may benefit from stronger regimens.

  • Tuberculosis
  • Respiratory Infection

Data availability statement

The data are accessible through University College London who now hold the data as a publicly accessible repository. Access to this is governed by an oversight committee, and applications must be made to that committee. There are legal restrictions in terms of signed data sharing agreements with all data contributors. So, application must be made to access the data. But it is publicly accessible and inquiries can be made through this link: https://www.ucl.ac.uk/global-health/research/tb-ipdplatform.

https://doi.org/10.1136/thorax-2023-220249

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Acid fast bacilli sputum smear positivity and radiological findings of lung cavitation are independent predictors of unfavourable outcomes in drug-susceptible tuberculosis, but among people with drug-resistant tuberculosis, these two factors have been inconsistently associated with treatment outcomes.

    WHAT THIS STUDY ADDS

    • Using a large individual participant database of people being treated with longer rifampicin-resistant tuberculosis regimens, acid fast bacilli sputum smear positivity or radiological findings of lung cavitation alone were not associated with unfavourable end-of-treatment outcomes.

    • Only participants who were both sputum smear positive and had lung cavitation were at increased odds of unfavourable outcomes, particularly failure and recurrence.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • People who have both evidence of sputum smear positivity and lung cavitation may benefit from stronger and/or longer treatment regimens.

    • These findings may have implications on appropriateness of shorter rifampicin-resistant tuberculosis treatment regimens for certain patient phenotypes.

    Background

    Of the 10.6 million people estimated by the WHO to have developed tuberculosis (TB) in 2021, 450 000 had rifampin-resistant or multidrug-resistant TB (RR/MDR-TB).1 RR/MDR-TB is associated with substantially worse treatment outcomes when compared with drug-susceptible TB.1 Among people initiating treatment for RR/MDR-TB in 2019, 60% experienced treatment success, compared with 86% among people initiating treatment for drug-susceptible TB in the same year.2

    Poor rates of RR/MDR-TB treatment success have led to widespread efforts to identify risk factors associated with unfavourable treatment outcomes. Extensive disease is one such risk factor, often defined based on results of sputum smear microscopy for acid-fast bacilli (AFB) and chest radiological findings of lung cavitation.3 AFB smear-positivity correlates with extent of disease and bacterial burden,4 while lung cavities generally form after primary infection and are a marker of prolonged disease.5 Among people with drug-susceptible TB, both AFB smear-positivity and lung cavitation predict unfavourable treatment outcomes across several prediction models.6–8 Findings among people with RR/MDR-TB have varied. In the STREAM randomised controlled trials,9 10 lung cavitation, but not AFB smear, was associated with increased risk of unfavourable outcomes, while in the ZeNiX study11 nearly all unfavourable outcomes occurred among persons with lung cavitation at enrolment. Conversely, in TB-PRACTECAL,12 neither smear nor lung cavitation appeared to modulate risk of unfavourable outcomes. In studies evaluating longer RR/MDR-TB regimens,3 13 AFB smear-positivity at treatment initiation was associated with lack of treatment response by 6 months, but lung cavitation was inconsistently associated.

    Using a large individual participant dataset of people initiating longer RR/MDR-TB treatment regimens, we aimed to evaluate the association of two commonly available indicators of extensive disease—AFB smear and lung cavitation findings—with end-of-treatment outcomes.

    Methods

    Search strategy and study selection

    Using a previously described database of RR/MDR-TB patients,14 we conducted an individual participant data meta-analysis. The database contains studies identified from systematic reviews performed in 2010 and 2016,15 16 as well as from a WHO public call for data in 2018.17 Studies were eligible for inclusion if study investigators agreed to share their data and the data shared included at least 25 patients: (1) with evidence of rifampin resistance; (2) with complete treatment information; (3) with known end-of-treatment outcomes and (4) receiving treatment with longer RR/MDR-TB regimens (ie, intended duration of treatment ≥18 months).

    Deidentified data on each participant were provided by study investigators. Data requested included baseline demographic and clinical characteristics of participants, such as age, sex, body mass index, HIV infection and antiretroviral therapy use, TB treatment history, AFB smear results, radiologic findings, and genotypic and phenotypic drug-susceptibility results; treatment information; and end-of-treatment outcomes. Each study was assessed for risk of selection bias and quality.

    The characteristics of included studies and their quality are provided in online supplemental appendix tables 1–3 and have been described previously.18 Overall, 52 studies conducted in 37 countries/regions containing 12 938 participants initiating RR/MDR-TB treatment from 1993 to 2016 are included in the individual participant database. From this database, we excluded all studies where AFB smear and lung cavitation findings were missing for >50% of the study population, and from the remaining studies, all participants with incomplete AFB smear and lung cavitation information.

    Supplemental material

    Definitions and outcomes

    The database contains data on AFB sputum smear result and absence/presence of lung cavities on chest X-ray at treatment initiation, without information on AFB smear grade or the extent of cavitation. We used these data to create a four-level categorical variable describing these two indicators of extensive disease: (1) AFB smear-negative without lung cavitation; (2) AFB smear-negative with lung cavitation; (3) AFB smear-positive without lung cavitation and (4) AFB smear-positive with lung cavitation. World Bank definitions of country-level income in 2018 were used to categorise the country where each participant was treated as low-income and lower-middle income, upper-middle-income or high-income.19 Following previous analyses, we created categories for the year of treatment initiation as participants started treatment over several years: 1993–2003, 2004–2008, 2009–2012 and 2013–2016.

    We considered any drug received for at least 1 month as used during treatment, unless the total treatment duration was less than 1 month, in which case all drugs were considered as used during treatment. We only considered drugs contained within WHO group A, B or C RR/MDR-TB drugs. Group A drugs included levofloxacin, moxifloxacin, bedaquiline and linezolid, while group B and C drugs included clofazimine, cycloserine, terizidone, delamanid, ethambutol, pyrazinamide, amikacin, streptomycin, para-aminosalicylic acid (PAS), ethionamide, prothionamide, meropenem and imipenem-cilastatin.20

    End-of-treatment outcomes were defined by study investigators and largely followed Laserson criteria21 (see online supplemental appendix table 4 for individual study criteria). We classified outcomes as death of any cause, failure or recurrence, lost to follow-up (lost to follow-up, transfer, unknown outcome) and success (cure or treatment complete without evidence of recurrence). Timing of culture conversion among those culture positive at baseline was defined by study authors. We calculated time to event as the time difference from the date of treatment initiation to the date of the event.

    Data analysis

    For primary analyses, we had three outcomes: any unfavourable outcome (failure, lost to follow-up, recurrence and death) compared with treatment success; death during treatment vs survival; and treatment failure or recurrence compared with disease-free survival (ie, excluding those who died during treatment). In secondary analyses, we evaluated the time to several events (any unfavourable outcome, death, failure or recurrence and lost to follow-up) among those with precisely known treatment duration, as well as the outcome of culture conversion among those culture positive at baseline.

    Descriptive analyses were performed stratified by AFB smear result and lung cavitation findings. Categorical variables were described using frequencies (n, %) and compared across groups using chi-square tests, while continuous variables were described using medians and IQR and compared across groups using Kruskal-Wallis tests.

    As data for potential confounders were missing for some participants (online supplemental appendix table 5), we imputed these data. Missing data for age, sex, body mass index, HIV and antiretroviral treatment, bilateral disease, previous TB treatment, and drug susceptibility results for fluoroquinolones, second-line injectables, streptomycin, ethionamide and prothionamide, ethambutol, pyrazinamide and PAS were imputed using multivariate imputation by chained equations (mice,22 V.3.13.0 in R). We performed 25 Gibb’s sampling iterations to generate each of the 20 imputed datasets22; original data were preserved within each imputed dataset and imputed data were only used where data were missing.

    We considered levofloxacin, moxifloxacin, amikacin, streptomycin, ethambutol, pyrazinamide, PAS, ethionamide and prothionamide as effective if they were used and drug susceptibility tests confirmed susceptibility, or if they were imputed as susceptible when data were missing. All other drugs were considered effective if they were used and drug resistance had not been proven, as for these drugs, drug susceptibility was largely unavailable, and when available was resistant in <10% of isolates.

    Using the imputed datasets, we assessed the association of AFB smear and lung cavitation findings on each outcome. We performed logistic regression with generalised linear mixed models, treating each study as the clustering variable, to assess this association (reference group: AFB smear-negative without lung cavitation) adjusted for a priori selected covariates informed by previous analyses.16 18 23 Covariates included age (continuous), sex, underweight (<18.5 kg/m2) body mass index, HIV coinfection and antiretroviral therapy use, country-level income, previous TB treatment history, resistance to fluoroquinolones and/or second-line injectables, presence of bilateral lung disease, number of effective WHO Group A drugs received, number of effective WHO group B and C drugs received, and year of treatment initiation. We analysed each imputed dataset to estimate adjusted ORs (aOR) and SEs, then pooled them according to Rubin’s Rules24 to estimate overall aOR and 95% CI.

    For each of our primary outcomes, we assessed potential effect modification. We did analyses stratified on the following variables: age (based on median age, <37 vs ≥37 years), sex, underweight body mass index, country-level income, bilateral disease, previous treatment history, HIV coinfection, resistance to fluoroquinolones and/or second-line injectables, use of bedaquiline and/or linezolid, number of effective drugs received, and year of treatment initiation. Likelihood ratio tests between models with versus without interaction terms for AFB smear and lung cavitation findings and the stratifying variable were performed to assess if there were any significant differences (p<0.05).

    Next, to further assess the robustness of our primary analysis, we performed two complete-case analyses. First, we only included participants who had completely known information on all covariates included in our multivariable analysis (‘strict complete case analysis’). Second, as the most commonly missing variable was drug susceptibility testing, we included participants with completely known information for all covariates except drug susceptibility testing, which we imputed (‘complete case except drug susceptibility testing analysis’).

    In secondary analysis, we included participants with a precisely known treatment duration and calculated the time to: any unfavourable outcome, death during treatment, failure or recurrence and declaration of lost to follow-up. These times were calculated for each of the four AFB smear and lung cavitation strata. We constructed cumulative incidence plots for each outcome in a competing risks framework,25 with success being treated as a censoring event.

    Finally, we evaluated the impact of AFB smear and/or lung cavitation on culture conversion. Among participants who were culture positive at treatment initiation with a known culture conversion outcome (did not convert, converted), we estimated the adjusted HR (aHR) using a Cox proportional hazards model. The model was adjusted for the same covariates as in our primary analyses and was performed on each imputed dataset, with resultant aHR and SEs pooled according to Rubin’s Rules.

    Data analysis was conducted in R (V.4.1.0) using packages mice (V.3.13.0) for imputation, lme4 (V.1.1-23) for logistic regression and survival (V.3.2-11) for culture conversion.

    Results

    Of 52 studies in the individual participant database, we excluded 17 studies (7192 participants) as AFB smear and lung cavitation information was missing for >50% of participants. In the remaining studies, 150 participants were excluded due to missing AFB and lung cavitation information, leaving 35 studies and 5596 participants (figure 1). Excluded participants significantly differed from those included, with the largest differences in country-level income, HIV coinfection and year of treatment initiation; these differences were driven by exclusion of large, recent studies from South Africa where chest X-ray findings were unavailable (online supplemental appendix table 6).

    Figure 1
    Figure 1

    PRISMA diagram for studies and populations included and excluded from the analysis. AFB, acid fast bacilli; CXR, chest X-ray; IPD, individual participant data.

    Characteristics of included participants are reported in table 1. Among the 5596 participants, 774 (13.8%) were AFB smear-negative without cavities, 647 (11.6%) were AFB smear-negative with lung cavities, 1424 (25.4%) were AFB smear-positive without lung cavities and 2751 (49.2%) were AFB smear-positive with cavities. Participants in each of these groups differed significantly on nearly all characteristics. The median (IQR) age was 37 years (28–47), 3580 (64%) were male, 686 (13% with known status) were living with HIV, and 998 (18%) and 746 (13%) received bedaquiline and linezolid, respectively. South Africa, South Korea, Russia and Philippines were the most common countries of treatment, where 2661 (48%) of participants were treated (online supplemental appendix table 7).

    View this table:
    Table 1

    Characteristics of included participants

    Primary analysis

    Among included participants, unfavourable outcomes occurred in 220/774 (28.4%) participants AFB smear-negative without cavities, 214/647 (33.1%) participants AFB smear-negative with cavities, 492/1424 (34.6%) participants AFB smear-positive without cavities and 1158/2751 (42.1%) participants AFB smear-positive with cavities. In adjusted analyses (table 2), when compared with participants with AFB smear-negative and without cavities, odds of unfavourable outcomes were 1.6 times (95% CI 1.3 to 2.0) higher among those AFB smear-positive with cavities, but not significantly different among those who were AFB smear-positive without cavities (aOR 1.2, 95% CI 0.96 to 1.5) or who were AFB smear-negative with cavities (aOR 1.0, 95% CI 0.8 to 1.4). Similar trends in adjusted analyses were seen for other primary outcomes, with participants AFB smear-positive with cavities having 1.5 times (95% CI 1.1 to 2.1) higher odds of mortality and 2.2 times (95% CI 1.5 to 3.3) higher odds of failure or recurrence, but no significant differences among participants who were only AFB smear-positive or only had cavities.

    View this table:
    Table 2

    Multivariable estimates of primary outcomes

    Findings from stratified analyses evaluating odds of any unfavourable outcome were largely consistent with those in the overall population (table 3). For any unfavourable outcome, we found evidence of effect modification by HIV-infection status (p=0.043) and resistance to fluoroquinolones and/or second-line injectables (p=0.023). Findings for the outcomes of mortality and failure or recurrence are reported in online supplemental appendix tables 8 and 9. We found evidence of effect modification for the outcome of mortality by receipt of bedaquiline and/or linezolid (p=0.012) and year of treatment initiation (p=0.015), while for failure or recurrence only by HIV-infection status (p=0.014).

    View this table:
    Table 3

    Multivariable estimates and stratified analyses assessing effect modification of AFB and CXR cavitation on all unfavourable outcomes

    We included 1806/5596 (32.3%) participants in our strict complete case analysis. Participants included in this analysis were more likely to be treated in a high-income country, while being less likely to be living with HIV or be resistant to fluoroquinolones and/or second-line injectables (online supplemental appendix table 10). We included 3343/5596 (59.7%) participants in our complete case except drug susceptibility testing analysis—participants in this analysis were more similar to those in the primary analysis but still were less likely to be resistant to fluoroquinolones and/or second-line injectables (online supplemental appendix table 11). In both complete case analyses (online supplemental appendix table 12), results were generally consistent with the primary analysis. For both analyses, participants who were AFB smear-positive with cavities still had significantly higher odds of any unfavourable outcome and the specific outcome of failure or recurrence; however, no group had increased odds of mortality. In the strict complete case analysis only, participants AFB smear-positive without cavities also had increased odds of any unfavourable outcome and failure or recurrence.

    Secondary analysis

    When evaluating the cumulative incidence of any unfavourable outcome, failure or recurrence, death and lost to follow-up (figure 2), we included 5279/5596 (94.3%) participants who had a precisely known duration of treatment. Participants who were AFB smear-negative and without lung cavitation had the lowest incidence of each event. In contrast, participants who were AFB smear-positive with lung cavitation typically had the highest incidence of each event, particularly for failure or recurrence. For this event, cumulative incidence across groups was similar in the first year of treatment before diverging later in treatment. By month 24, the cumulative incidence of failure or recurrence for those AFB smear-positive with cavities was 7.3% (95% CI 6.2% to 8.6%), while among those AFB smear-negative without cavities it was 3.4% (95% CI 2.0% to 6.1%), among those AFB smear-negative with cavities 3.3% (95% CI 2.0% to 5.7%) and among those AFB smear-positive without cavities 4.7% (95% CI 3.4% to 6.4%).

    Figure 2
    Figure 2

    Time to event for different treatment outcomes (note: different y-axis scales), with the 24-month cumulative incidence (95% CI) for each group. Time to any unfavourable outcome (A), failure or recurrence (B), death (C) and lost to follow-up (D). CXR, chest X-ray.

    When evaluating time to culture conversion, we included 4274/5596 (76.3%) participants who were culture positive at baseline with a known culture conversion outcome. We excluded 252 participants because they were culture negative at baseline and 1070 participants due to absence of culture and/or culture conversion information. Participants excluded from this analysis significantly differed from those included (online supplemental appendix table 13). Overall, 540/4274 (12.6%) participants culture positive at baseline did not culture convert; the vast majority (425/540; 78.7%) were resistant to fluoroquinolones and/or second-line injectables. We found when compared with participants AFB smear-negative without cavities, all other participants had slower times to culture conversion. Adjusted HRs (where values below one suggest slower times to culture conversion, relative to persons with AFB smear-negative and no cavities) were 0.86 (95% CI 0.74 to 0.98) for those AFB smear-negative with lung cavities, 0.74 (95% CI 0.66 to 0.84) for those AFB smear-positive without lung cavities, and 0.62 (95% CI 0.56 to 0.69) for those both AFB smear-positive and with lung cavities (table 4).

    View this table:
    Table 4

    Cox proportional hazards model for time to culture conversion

    Discussion

    In this study of 5596 people from diverse settings treated with longer RR/MDR-TB regimens, we found only the combination of AFB smear-positivity and lung cavitation was associated with any unfavourable outcome, mortality and failure or recurrence, when compared with those who were AFB smear-negative and without lung cavitation. These findings were strongest for the outcome of failure or recurrence and largely consistent in analyses stratified on important patient, treatment and clinical characteristics. These data suggest people being treated for RR/MDR-TB with these features may require stronger and/or longer treatment regimens.

    Time to culture conversion was significantly longer among persons with AFB smear-positivity and lung cavitation present alone or together, similar to findings from the endTB observational study which included participants with RR/MDR-TB receiving bedaquiline and/or delamanid.3 However, our findings for this interim treatment outcome were not replicated in the analysis of end-of-treatment outcomes, where only the combination of indicators were associated with unfavourable outcomes. This adds to the body of evidence suggesting time to culture conversion—generally used as an interim outcome for treatment monitoring20—may poorly predict end-of-treatment outcomes.26

    We found the occurrence of unfavourable outcomes between groups stratified on AFB smear result and presence of lung cavities was similar early in treatment before diverging later in treatment (figure 2). This trend differs from previous analyses stratified on HIV18 and low body mass index,23 and the phenomenon was most striking for the outcome of failure or recurrence. The formation of lung cavities occurs through liquefaction of caseous necrotic tissue.27 The resulting high-oxygen-area of the cavity surrounding the necrotic tissue is a site for rapid bacterial replication, while the necrotic area is poorly penetrated by host immune cells and anti-TB drugs.5 28 Meanwhile, AFB smear grading is based on mycobacterial load in the sputum.29 The presence of both AFB smear-positivity and lung cavitation, therefore, likely increases risk for treatment failure or recurrence due to incomplete eradication of bacteria due to its high burden and/or poor penetration of drugs into cavities,28 as has been seen in people with drug-susceptible TB.8 30

    Though this study focused on people receiving long (18–24 months) RR/MDR-TB regimens, its findings have implications for patients treated with newer and shorter RR/MDR-TB regimens. In studies of shorter RR/MDR-TB regimens, both lung cavitation and AFB smear-positivity have been inconsistently associated with unfavourable outcomes.9–12 31 This could be due to the relatively small sample sizes in such studies, making them underpowered for subgroup analyses. However, these could be critical factors in deciding appropriateness of shorter regimens, as has been demonstrated in drug-susceptible TB.32 For example, in the ZeNix trial, the rate of unfavourable outcomes among participants without lung cavities was 1.5% (1/67) while it was 16.2% (18/111) among those with lung cavities.11 The use of individual participant data meta-analysis of shorter RR/MDR-TB regimens would better clarify the role of AFB smear-positivity and lung cavitation in deciding the appropriate length of treatment, and would be facilitated by subscription to standardised collection and reporting of patient, clinical and treatment characteristics and outcomes in TB research.33

    Key strengths of our analyses include the diverse nature of studies and participants analysed, which should increase generalisability of our findings. The use of individual participant data and multiple imputation for missing covariate data allowed for adjustment of numerous covariates and reflection of the uncertainty in estimates. Additional analyses on effect modification and time to interim (culture conversion) and end-of-treatment outcomes further elucidated the impact of indicators of extent of disease and how they might vary by key characteristics. These data can be used by clinicians to potentially increase regimen duration and strengthen its composition, as is often done with drug-susceptible TB,34 or consider adjunctive therapies (such as host-directed therapies35).

    Our study has limitations. We lacked detailed information on AFB smear grade and extent of lung cavitation—which was based on chest X-ray and are known to have inter-reader and intrareader variability.36 These factors may further modify unfavourable treatment outcomes37 38 and should be evaluated in future studies. We did not have start and end dates of different TB drugs available in all datasets, and to align definitions across studies, we considered all drugs received. This may result in misclassification bias with respect to the total number of effective drugs received at any one time. Not all studies included in the individual participant database had both AFB smear and lung cavitation findings and were excluded. This may introduce selection bias as those excluded significantly differed from those included, though some findings aligned with other large studies in RR/MDR-TB.3 Adherence is an important determinant of treatment outcomes,39 however, we lacked these data and there could be residual confounding. Nearly all studies included in this meta-analysis used directly observed treatment, which should have encouraged a high level of treatment adherence.40 Finally, the number of participants in this analysis initiating treatment from 2013 to 2016 and receiving new/repurposed drugs of bedaquiline and linezolid were relatively few.

    In summary, we found only when AFB smear-positivity and lung cavitation were present together were they associated with unfavourable RR/MDR-TB outcomes. Future research should evaluate if longer and/or stronger treatment regimens improve outcomes among people with these indicators, how smear grade and number and size of lung cavities impact outcomes, as well as use individual participant data meta-analysis to determine if these findings are consistent among people treated with shorter RR/MDR-TB regimens.

    Data availability statement

    The data are accessible through University College London who now hold the data as a publicly accessible repository. Access to this is governed by an oversight committee, and applications must be made to that committee. There are legal restrictions in terms of signed data sharing agreements with all data contributors. So, application must be made to access the data. But it is publicly accessible and inquiries can be made through this link: https://www.ucl.ac.uk/global-health/research/tb-ipdplatform.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study used individual patient data provided by the investigators of the original studies, who obtained informed consent from all participants as appropriate for their original study designs. All data received were anonymised. This analysis was approved by the Research Institute of the McGill University Health Centre (Montreal, QC, Canada; BMB-07-021t), as well as by participating sites, as necessary.

    Acknowledgments

    The authors would like to acknowledge all members of the Collaborative Group for the Meta-Analysis of Individual Patient Data in MDR-TB treatment–2019 who contributed data making this analysis possible.

    References

    1. ↵
      1. WHO
      . Global tuberculosis report: 2022. Geneva: World Health Organization; 2022. Available: https://www.who.int/teams/global-tuberculosis-programme/tb-reports/global-tuberculosis-report-2022
    2. ↵
      1. WHO
      . Tuberculosis data repository. Geneva: World Health Organization 2021; Available: https://www.who.int/teams/global-tuberculosis-programme/data [Accessed 12 Sep 2022].
    3. ↵
      2. Franke MF ,
      3. Khan P ,
      4. Hewison C , et al
      . Culture conversion in patients treated with bedaquiline and/or delamanid. A prospective multicountry study. Am J Respir Crit Care Med 2021;203:111–9. doi:10.1164/rccm.202001-0135OC
      OpenUrl
    4. ↵
      2. Olaru ID ,
      3. Heyckendorf J ,
      4. Grossmann S , et al
      . Time to culture positivity and Sputum smear microscopy during tuberculosis therapy. PLoS One 2014;9:e106075. doi:10.1371/journal.pone.0106075
    5. ↵
      2. Ong CWM ,
      3. Elkington PT ,
      4. Friedland JS
      . Pulmonary cavitation, and matrix metalloproteinases. Am J Respir Crit Care Med 2014;190:9–18. doi:10.1164/rccm.201311-2106PP
      OpenUrlCrossRefPubMed
    6. ↵
      2. Peetluk LS ,
      3. Ridolfi FM ,
      4. Rebeiro PF , et al
      . Systematic review of prediction models for pulmonary tuberculosis treatment outcomes in adults. BMJ Open 2021;11:e044687. doi:10.1136/bmjopen-2020-044687
    7. ↵
      2. Gopalan N ,
      3. Srinivasalu VA ,
      4. Chinnayan P , et al
      . Predictors of unfavorable responses to therapy in rifampicin-sensitive pulmonary tuberculosis using an integrated approach of radiological presentation and sputum mycobacterial burden. PLOS ONE 2021;16:e0257647. doi:10.1371/journal.pone.0257647
    8. ↵
      2. Romanowski K ,
      3. Balshaw RF ,
      4. Benedetti A , et al
      . Predicting tuberculosis relapse in patients treated with the standard 6-month regimen: an individual patient data meta-analysis. Thorax 2019;74:291–7. doi:10.1136/thoraxjnl-2017-211120
      OpenUrlAbstract/FREE Full Text
    9. ↵
      2. Nunn AJ ,
      3. Phillips PPJ
      . Trial of a shorter regimen for rifampin-resistant tuberculosis. N Engl J Med 2019;380:1201–13. doi:10.1056/NEJMoa1811867
      OpenUrlCrossRefPubMed
    10. ↵
      2. Goodall RL ,
      3. Meredith SK ,
      4. Nunn AJ , et al
      . Evaluation of two short standardised regimens for the treatment of rifampicin-resistant tuberculosis (STREAM stage 2): an open-label, multicentre, randomised, non-inferiority trial. Lancet 2022;400:1858–68. doi:10.1016/S0140-6736(22)02078-5
      OpenUrl
    11. ↵
      2. Conradie F ,
      3. Bagdasaryan TR ,
      4. Borisov S , et al
      . Bedaquiline–Pretomanid–Linezolid regimens for drug-resistant tuberculosis. N Engl J Med 2022;387:810–23. doi:10.1056/NEJMoa2119430
      OpenUrlCrossRef
    12. ↵
      2. Nyang’wa B-T ,
      3. Berry C ,
      4. Kazounis E , et al
      . A 24-week, all-oral regimen for rifampin-resistant tuberculosis. N Engl J Med 2022;387:2331–43. doi:10.1056/NEJMoa2117166
      OpenUrlCrossRef
    13. ↵
      2. Hewison C ,
      3. Ferlazzo G ,
      4. Avaliani Z , et al
      . Six-month response to delamanid treatment in MDR TB patients. Emerg Infect Dis 2017;23:1746–8. doi:10.3201/eid2310.170468
      OpenUrl
    14. ↵
      2. Ahmad N ,
      3. Ahuja SD ,
      4. Akkerman OW , et al
      . Treatment correlates of successful outcomes in pulmonary multidrug-resistant tuberculosis: an individual patient data meta-analysis. Lancet 2018;392:821–34. doi:10.1016/S0140-6736(18)31644-1
      OpenUrlCrossRefPubMed
    15. ↵
      2. Bastos ML ,
      3. Lan Z ,
      4. Menzies D
      . An updated systematic review and meta-analysis for treatment of multidrug-resistant tuberculosis. Eur Respir J 2017;49:1600803. doi:10.1183/13993003.00803-2016
    16. ↵
      2. Ahuja SD ,
      3. Ashkin D ,
      4. Avendano M , et al
      . Multidrug resistant pulmonary tuberculosis treatment regimens and patient outcomes: an individual patient data meta-analysis of 9,153 patients. PLoS Med 2012;9:e1001300. doi:10.1371/journal.pmed.1001300
    17. ↵
      1. WHO
      . Public call for individual patient data on treatment of rifampicin and multidrug-resistant (MDR/RR-TB) tuberculosis. 2018. Available: https://www.who.int/news-room/articles-detail/public-call-for-individual-patient-data-on-treatment-of-rifampicin-and-multidrug-resistant-(mdr-rr-tb)-tuberculosis [Accessed 29 Jun 2021].
    18. ↵
      2. Bisson GP ,
      3. Bastos M ,
      4. Campbell JR , et al
      . Mortality in adults with multidrug-resistant tuberculosis and HIV by antiretroviral therapy and tuberculosis drug use: an individual patient data meta-analysis. Lancet 2020;396:402–11. doi:10.1016/S0140-6736(20)31316-7
      OpenUrl
    19. ↵
      1. World Bank
      . World bank country and lending groups. 2021. Available: https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups [Accessed 29 Jun 2021].
    20. ↵
      1. WHO
      . WHO Consolidated guidelines on tuberculosis. Module 4, treatment: drug-resistant tuberculosis treatment. Geneva World Health Organization; 2020. Available: https://www.who.int/publications/i/item/9789240007048 [Accessed Jul 2020].
    21. ↵
      2. Laserson KF ,
      3. Thorpe LE ,
      4. Leimane V , et al
      . Speaking the same language: treatment outcome definitions for multidrug-resistant tuberculosis. Int J Tuberc Lung Dis 2005;9:640–5.
      OpenUrlPubMedWeb of Science
    22. ↵
      2. van Buuren S ,
      3. Groothuis-Oudshoorn K
      . Mice: multivariate imputation by chained equations in R. J Stat Softw 2011;45:1–67. doi:10.18637/jss.v045.i03
      OpenUrlCrossRef
    23. ↵
      2. Campbell JR ,
      3. Chan ED ,
      4. Falzon D , et al
      . Low body mass index at treatment initiation and rifampicin-resistant tuberculosis treatment outcomes: an individual participant data meta-analysis. Clin Infect Dis 2022;75:2201–10. doi:10.1093/cid/ciac322
      OpenUrl
    24. ↵
      2. Rubin DB
      . Multiple imputation for nonresponse in surveys. New York: John Wiley & Sons, 1987. doi:10.1002/9780470316696
    25. ↵
      2. Aalen OO ,
      3. Johansen S
      . An empirical transition matrix for non-homogeneous markov chains based on censored observations. Scand J Stat 1978;5:141–50.
      OpenUrlWeb of Science
    26. ↵
      2. Phillips PPJ ,
      3. Mendel CM ,
      4. Burger DA , et al
      . Limited role of culture conversion for decision-making in individual patient care and for advancing novel regimens to confirmatory clinical trials. BMC Med 2016;14:19. doi:10.1186/s12916-016-0565-y
    27. ↵
      2. Dannenberg AM
      . Pathogenesis of pulmonary tuberculosis. Am Rev Respir Dis 1982;125(3 Pt 2):25–9. doi:10.1164/arrd.1982.125.3P2.25
      OpenUrlPubMedWeb of Science
    28. ↵
      2. Urbanowski ME ,
      3. Ordonez AA ,
      4. Ruiz-Bedoya CA , et al
      . Cavitary tuberculosis: the gateway of disease transmission. Lancet Infect Dis 2020;20:e117–28. doi:10.1016/S1473-3099(20)30148-1
      OpenUrl
    29. ↵
      2. Lohmann EM ,
      3. Koster BFPJ ,
      4. le Cessie S , et al
      . Grading of a positive sputum smear and the risk of mycobacterium tuberculosis transmission. Int J Tuberc Lung Dis 2012;16:1477–84. doi:10.5588/ijtld.12.0129
      OpenUrlCrossRefPubMed
    30. ↵
      2. Colangeli R ,
      3. Jedrey H ,
      4. Kim S , et al
      . Bacterial factors that predict relapse after tuberculosis therapy. N Engl J Med 2018;379:823–33. doi:10.1056/NEJMoa1715849
      OpenUrlCrossRefPubMed
    31. ↵
      2. Ndjeka N ,
      3. Campbell JR ,
      4. Meintjes G , et al
      . Treatment outcomes 24 months after initiating short, all-oral bedaquiline-containing or Injectable-containing rifampicin-resistant tuberculosis treatment regimens in South Africa: a retrospective cohort study. Lancet Infect Dis 2022;22:1042–51. doi:10.1016/S1473-3099(21)00811-2
      OpenUrl
    32. ↵
      2. Imperial MZ ,
      3. Nahid P ,
      4. Phillips PPJ , et al
      . A patient-level pooled analysis of treatment-shortening regimens for drug-susceptible pulmonary tuberculosis. Nat Med 2018;24:1708–15. doi:10.1038/s41591-018-0224-2
      OpenUrlCrossRefPubMed
    33. ↵
      2. Campbell JR ,
      3. Falzon D ,
      4. Mirzayev F , et al
      . Improving quality of patient data for treatment of multidrug- or rifampin-resistant tuberculosis. Emerg Infect Dis 2020;26. doi:10.3201/eid2603.190997
    34. ↵
      2. Johnston JC ,
      3. Cooper R ,
      4. Menzies D
      . Chapter 5: treatment of tuberculosis disease. Can J Respir Crit Care Sleep Med 2022;6:66–76. doi:10.1080/24745332.2022.2036504
      OpenUrl
    35. ↵
      2. Wallis RS ,
      3. Ginindza S ,
      4. Beattie T , et al
      . Adjunctive host-directed therapies for pulmonary tuberculosis: a prospective, open-label, phase 2, randomised controlled trial. Lancet Respir Med 2021;9:897–908. doi:10.1016/S2213-2600(20)30448-3
      OpenUrl
    36. ↵
      2. Nel M ,
      3. Franckling-Smith Z ,
      4. Pillay T , et al
      . Chest imaging for pulmonary TB-an update. Pathogens 2022;11:161. doi:10.3390/pathogens11020161
    37. ↵
      2. Kokebu DM ,
      3. Ahmed S ,
      4. Moodliar R , et al
      . Failure or relapse predictors for the STREAM stage 1 short regimen for RR-TB. Int J Tuberc Lung Dis 2022;26:753–9. doi:10.5588/ijtld.22.0073
      OpenUrl
    38. ↵
      2. Hernandez-Romieu AC ,
      3. Little BP ,
      4. Bernheim A , et al
      . Increasing number and volume of cavitary lesions on chest computed tomography are associated with prolonged time to culture conversion in pulmonary tuberculosis. Open Forum Infect Dis 2019;6:ofz443. doi:10.1093/ofid/ofz443
    39. ↵
      2. Chimeh RA ,
      3. Gafar F ,
      4. Pradipta IS , et al
      . Clinical and economic impact of medication non-adherence in drug-susceptible tuberculosis: a systematic review. Int J Tuberc Lung Dis 2020;24:811–9. doi:10.5588/ijtld.19.0754
      OpenUrlPubMed
    40. ↵
      2. Alipanah N ,
      3. Jarlsberg L ,
      4. Miller C , et al
      . Adherence interventions and outcomes of tuberculosis treatment: a systematic review and meta-analysis of trials and observational studies. PLoS Med 2018;15:e1002595. doi:10.1371/journal.pmed.1002595

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Twitter @JCampbellMcGill

    • Contributors JRC, SKB, PB, MLB, MB, LG and RM were responsible for the design and conception of the study. JRC, SKB, PB, MLB, MB, LG, RK, DK, RLL, VM, RS, AS, AT, TSvdF, PV and RM contributed to data collection and curation. JRC was responsible for the data analysis and drafting the first version of the manuscript. JRC, SKB, PB, MLB, MB, LG, RK, DK, RLL, VM, RS, AS, AT, TSvdF, PV and RM revised the manuscript and made significant intellectual contributions to its content. JRC is the overall guarantor of the manuscript and its content.

    • Funding JRC receives salary support from the McGill University Health Centre Foundation and the McGill University Department of Medicine.

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.