A population-based study of tuberculosis incidence among rheumatic disease patients under anti-TNF treatment.

INTRODUCTION: Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis. The advent of immunobiologic therapy with TNF inhibitors agents, has been associated with a significant increase in incident cases of tuberculosis in this population.
OBJECTIVE: To estimate the incidence of tuberculosis in patients receiving TNF inhibitors therapy for rheumatic diseases. As secondary objectives, we sought to evaluate mortality and the clinical impact of screening for latent tuberculosis infection.
METHODS: This retrospective study included patients with rheumatic diseases of Public Health System from the Brazilian state, a high TB incidence area, who received prescriptions of TNF inhibitors agents between 2006 and 2016.
RESULTS: A total of 5853 rheumatic disease patients were included. Patients were predominantly women (68.7%) aged 49.5 (± 14.7) years old. Forty-three cases of TB were found (2.86 cases per 1000 person-years; 18 times higher than in the general population). Adalimumab and certolizumab users presented a higher risk for TB development compared to etanercept users (RR: 3.11, 95%CI 1.16-8.35; 7.47, 95%CI 1.39-40.0, respectively). In a subgroup of patients, screening for latent tuberculosis infection was performed in 86% of patients, and 30.2% had a positive tuberculin skin test. Despite latent TB treatment, TB was diagnosed in 2 out of 74 (2.7%) patients. Overall, TB diagnosis did not increase mortality.
CONCLUSION: In this population-based study of rheumatic disease patients from a high incident area, TNF inhibitor exposure was associated with an 18-time increased TB incidence. Adalimumab and certolizumab were associated with greater and earlier TB diagnosis compared to etanercept.

Introduction

Tumor necrosis factor-alpha (TNF-α) is a cytokine involved in the pathogenesis of several systemic rheumatic diseases.[1,2] With the advent of TNF inhibitors therapy, treatment of these diseases has advanced markedly and clinical outcomes have improved, especially in patients refractory to conventional therapy.[3-7] The benefits of TNF inhibitors therapy have been well established in several studies that have demonstrated efficacy in controlling disease activity in rheumatoid arthritis (RA), ankylosing spondylitis (AS), psoriatic arthritis (PsA), and juvenile idiopathic arthritis (JIA).[8-10] Five TNF inhibitors agents are currently available for use in Brazil: infliximab (IFX), etanercept (ETN), adalimumab (ADA), golimumab (GOL), and certolizumab pegol (CZP).

TNF-α is known to play a role in the control of infectious diseases, particularly those caused by intracellular microorganisms such as Mycobacterium tuberculosis.[11] Its role is particularly important in organizing the activation and maintenance of granuloma.[12,13] Accordingly, despite its efficacy, TNF inhibitors therapy has been shown to increase the incidence of infections in general and of serious infections (such as tuberculosis) in particular.[14,15]

In 2016, an estimated 10.4 million incident cases of tuberculosis (TB) will have occurred worldwide.[16] Brazil ranks 20th among the 30 countries with the highest TB burden in the world, and accounts for approximately one-third of all incident cases of TB in the Americas.[17] In 2017, TB incidence rate of 33.5 cases per 100,000 population were reported in Brazil. Rates in the state of Rio Grande do Sul (RS) appear to be higher than the national average, with an incidence rate of 39.5 cases per 100,000 population; this makes RS one of the four states with the highest incidence of new-onset TB in Brazil.[18]

The risk of developing TB is higher in individuals with RA when compared to the general population.[19] This risk is fourfold higher in RA patients on TNF inhibitors therapy when compared to that of anti-TNF-naive RA patients. [15,20-22] In areas with a higher incidence of TB, such as in Asian countries, a nearly 26-fold greater risk of TB was found in those exposed to TNF inhibitors agents.[23] In patients given infliximab, risk could be up to 30 times greater than in the general population.[24,25]

Considering this increased risk, screening and treatment of latent tuberculosis infection (LTBI) has been recommended prior to initiation of TNF inhibitors therapy.[2,26-29] Screening for LTBI has been shown to reduce the risk of TB reactivation. According to the Spanish Society of Rheumatology registry of patients on immunobiologicals, BIOBADASER, this practice has managed to reduce the number of incident cases of TB in patients starting TNF inhibitors TNF therapy by 78% between 2002 and 2006, when it was implemented.[29,30]

A previous study carried out in Brazil, based on records from the Brazilian Registry of Biologic Therapy Monitoring (BiobadaBrasil) maintained by the Brazilian Society of Rheumatology, found a TB incidence in RA patients with TNF inhibitors exposure of 2.8 cases per 1,000 exposed.[31] Given the higher-than-average incidence of TB in southern Brazil, it is believed that the rate of new TB cases in patients with rheumatic diseases exposed to TNF inhibitors therapy in Rio Grande do Sul may also be significantly higher than that of the general population.[18] Within this context, the primary objective of this study is to estimate the incidence of tuberculosis in patients receiving TNF inhibitors therapy for rheumatic diseases. As secondary objectives, we sought to evaluate mortality and the influence of screening for latent tuberculosis infection on clinical outcomes in this population.

Methods

Study design and patients

This population-based retrospective cohort study included all Public Health System patients from the state of Rio Grande do Sul who were prescribed and dispensed TNF inhibitors therapy for rheumatic diseases from 2006 to 2016. Inclusion was based on the records of the statewide Exceptional Circumstance Drug Dispensing Program, considering those ICD-10 codes covered by the Brazilian Clinical Protocols and Therapeutic Guidelines for rheumatic diseases. Because TB is a notifiable disease in Brazil, definition of TB cases was based on TB reporting data, obtained from the Notifiable Diseases Information System. This system includes all patients with a confirmed diagnosis of tuberculosis, as well as the clinical presentation, treatment regimen and final follow-up outcome. Positive cases being defined by patients with clinical diagnosis associated with bacteriological diagnosis.

Mortality data were obtained from the Mortality Information System, provided by the state centralizing agency for all death certificates in this state. To unify information from these different databases, the linkage technique was used, whereby standardized information is used to find the same individual across several data sources. The variables used for linkage were patient’s name, mother’s name, and date of birth.

Screening for latent tuberculosis prior to the initiation of TNF inhibitors is recommended for all patients who are candidates for therapy and drug release is required. In subgroup of patients being followed at the outpatient rheumatology clinic of Hospital de Clínicas de Porto Alegre (HCPA) underwent a tuberculin skin test prior to initiation of TNF inhibitors therapy; if LTBI was detected, positive tuberculin skin test or suggestive chest X-ray findings without history of previous tuberculosis treatment, isoniazid treatment was indicated. This information was obtained from a review of patients’ medical records. In addition, patients with and without a history of treatment for LTBI were evaluated for potential TB infection.

Statistical analyses

Data were analyzed in SPSS version 18.03.

No sample calculation was performed since all samples were scheduled to be included in the study.

Descriptive analyses of the incidence rate of tuberculosis, calculated per 1000 exposed patient-years and per 100,000 exposed patients, were carried out. Person- years were calculated from the first day of TNF inhibitor therapy to the date of tuberculosis infection, or death or discontinuation of TNF inhibitor use.

For incidence density analysis, we used the time of use related to the first exposure for calculation of incidence related to the first exposure, also the total time of use of any TNF inhibitor during follow-up to define cumulative incidence density over the course of the study.

In addition, descriptive analyses of all data related to age, sex, underlying disease, duration of anti-TNF exposure, and number of anti-TNF agents used were performed. Among TB cases, we evaluated demographic characteristics, underlying rheumatic disease, TNF inhibitors agent used, and time elapsed from initiation of TNF inhibitors therapy to onset of TB.

Fig 1 illustrates the model used to validate a TB case as related to TNF inhibitors TNF agent exposure. A case of TB was considered related to anti-TNF therapy if it occurred up to 90 days after ETN, ADA, GOL, or CZP were last dispensed to the patient, and up to 180 days after IFX was last dispensed. The 90-day period was defined based on a previous study to evaluate the incidence of tuberculosis, and time on drug influence was still considered.[32] Specifically, considering infliximab this time was extended to 180 days in view of some evidence pointing to the presence of drug circulating up to 28 weeks post exposure.[33]

Fig 1

Model for drug-related tuberculosis cases.

Cases of TB during TNF inhibitors use. Cases of TB associated with recente use of TNF inhibitors. Cases of TB not related to TNF inhibitors use.

To compare the incidence of TB among different drugs, incidence was evaluated as percentage, with Fisher’s exact test for statistical significance. P-values <0.05 were considered statistically significant.

Disease-free survival time was analyzed in all patients receiving TNF inhibitors therapy by the Kaplan–Meier method, with head-to-head comparison of drugs using the Mantel-Cox log-rank test for the event of interest.

Patient survival was also evaluated using Kaplan–Meier analysis, which included all patients in the cohort. The last evaluation of mortality records was defined as the time point for censoring. The log-rank test was used to compare all-cause mortality versus deaths due to TB.

In the subgroup of patients who underwent screening for LTBI, categorical comparison was performed using Pearson’s chi-square test or Fisher’s exact test. Again, p-values <0.05 were considered significant for all analyses.

This study was approved by the Research Ethics Committees of Hospital de Clínicas de Porto Alegre and the School of Public Health/Rio Grande do Sul State Department of Health with opinion number 66184417.8.3001.5312.

Results

A total of 5853 patients and 6902 courses of TNF inhibitors therapy were included, from 2006 to 2016. The sample was predominantly female (68.7%), with a mean age of 49.1 years (Table 1). All five TNF inhibitors agents approved for use in Brazil by the Public Health System were covered: adalimumab accounted for 2980 courses of treatment (43.8%), etanercept for 2543 (37.4%), infliximab for 623 (9.1%), golimumab for 492 (7.2%), and certolizumab pegol for 164 (2.4%).

Table 1

Demographic and clinical features in patients in patients exposed to TNF inhibitors.

Patients’ features	RA(n = 3653)	AS(n = 1150)	PsA(n = 872)	JIA(n = 124)	Other spondyloarthritis(n = 54)	Total(n = 5853)	p*
Female	3004 (82.2)*	467 (40.6)	446 (51.1)	76 (61.2)	29 (53.7)	4022 (68.7)	<0,001
Age (years)—mean (SD)	52.7 (12.8)d	42.8 (11.9)b	47.7 (12.5)c	13.1 (7.51)a	44.0 (13.5)bc	49.1 (14.1)	<0,001
≤65 years	3039 (83.2)	1112 (96.7)*	798 (91.5)*	124 (100)*	51 (94.4)	5124 (87.6)
>65 years	614 (16.8)*	38 (3.3)	74 (8.5)	0	3 (5.6)	729 (12.4)
TNF inhibitors agent	4212	1345	1030	151	60	6802
IFX	430 (10.2)*	113 (8.4)	69 (6.6)	7 (4.6)	4 (6.6)	623 (9.1)	0,001
ETN	1441 (34.1)	533 (39.2)*	427 (41.4)*	118 (78.1)*	24 (40)	2543 (37.3)	<0,001
ADA	1695 (40.2)	696 (51.74)*	529 (51.3)*	28 (18.5)	32 (53.3)	2980 (43.8)	<0,001
GOL	483 (11.4)*	3 (0.2)	4 (0.3)	2 (1.3)	0	492 (7.2)	<0,001
CZP	163 (3.8)*	0	1 (0.1)	0	0	164 (2.4)	<0,001
Number of patients per TNF inhibitors							0,006
1 TNF inhibitors	3090 (84.6)*	953 (83)	713 (81.7)	92 (74.2)	48 (89)	4896 (83.6)
2 or more TNF inhibitors	563 (15.4)	197 (17)	159 (18.3)	32 (25.8)*	6 (11)	957 (16.4)
Duration of TNF inhibitors therapy—mean years (SD)	2.55 (2.03)b	2.67 (1.71)b	2.59 (1.85)b	2.87 (2.27)b	1.56(1.55)a	2.58 (1.95)	<0,001

Abbreviations: RA, rheumatoid arthritis; AS, ankylosing spondylitis; PsA, psoriatic arthritis; JIA, juvenile idiopathic arthritis; TNF, tumor necrosis factor; IFX, infliximab; ETN, etanercept; ADA, adalimumab; GOL, golimumab; CZP, certolizumab pegol; SD, standard deviation;

* p significant

variables with letters not repeated = p significant

Of the patients included, 3653 (62.4%) had been diagnosed with rheumatoid arthritis, 1150 (19.7%) with ankylosing spondylitis, 872 (14.9%) with psoriatic arthritis, 124 (2.1%) with juvenile idiopathic arthritis, and 54 (0.9%) with other inflammatory spondyloarthropathies.

In this sample of TNF inhibitors users, 4896 (83.6%) used only one TNF inhibitor throughout the study period, while 957 (16.4%) were exposed to two or more drugs of this class. The average duration of follow-up, ie exposure to TNF inhibitors therapy was 2.58 ± 1.95 years.

Of the 5853 patients included, 43 received a diagnosis of TB during follow-up; 28 of these occurred during the first exposure. The characteristics of these patients are described in Table 2. The mean age of these patients was 49.5 ± 14.7 years. No significant relationship was found with age >65 years (p = 0.298). Twenty-four (55.8%) were women, 28 (65.1%) had RA, 8 (18.6%) had AS, 6 (13.9%) had PsA, and 1 (2.3%) had JIA. Regarding anti-TNF agents, 27 (62.8%) cases were associated with ADA, 10 (23.2%) with ETN, 3 (7%) with IFX, 2 (4.6) with CZP and 1 (2.3%) with GOL.

Table 2

Demographic and clinical features in patients with tuberculosis.

Patients with tuberculosis	RA(n = 28)	AS(n = 8)	PsA(n = 6)	JIA(n = 1)	Total(n = 43)	p
Female	19 (67.8)	4 (50)	1 (16.7)	0)	24 (55.8)	0,080
Age (years)—mean (SD)	52 (14.1)	43.5 (11.1)	51.6 (14.3)	13.4	49.5 (14.7)	0,298
≤65 years	21 (75)	7 (100)	5 (83.3)	1 (100)	35 (81.4)
>65 years	7 (25)	0	1 (16.7)	0	8 (18.6)
TNF inhibitors agent
IFX	3 (10.7)	0	0	0	3 (7)	0,501
ETN	6 (21.4)	2 (25)	1 (16.7)	1 (100)	10 (23.2)	0,344
ADA	16 (57.1)	6 (75)	5 (83.3)	0	27 (62.8)	0,344
GOL	1 (3.6)	0	0	0	1 (2.3)	0,771
CZP	2 (7.1)	0	0	0	2 (4.6)	0,771
Patients with prior TNF inhibitors	11 (37.4)	0	4 (66.7)*	0	15 (34.9)	0,028
Time to active TB§ (years)—mean (SD)	0.5 (0.1)	0.9 (0.7)	1.3 (1.1)	1.7	0.9 (0.5)	0,981
Site of TB						0,678
Pulmonary	17 (60.7)	5 (62.5)	3 (50)	1 (100)	26 (60.5)
Extrapulmonary	7 (25)	3 (37.5)	3 (50)	0	13 (30.2)
Pulmonary and extrapulmonary	4 (14.3)	0	0	0	4 (9.3)

Abbreviations: RA, rheumatoid arthritis; AS, ankylosing spondylitis; PsA, psoriatic arthritis; JIA, juvenile idiopathic arthritis; TNF, tumor necrosis factor; IFX, infliximab; ETN, etanercept; ADA, adalimumab; GOL, golimumab; CZP, certolizumab pegol; SD, standard deviation.

§Time from onset of TNF inhibitors to development of tuberculosis

* p significant

The overall incidence rate of TB in the study population was 734.7 cases per 100,000 exposed. In patient-years, this incidence corresponds to 2.73 per 1000 patient-years exposed (considering first exposure to an TNF inhibitor agent), while the cumulative incidence of TB in the overall study population was 2.86 per 1000 patient-years. Disaggregated by rheumatic disease, the cumulative incidence was 3 per 1000 patient-years for RA (9305 patient-years exposed), 2.61 per 1000 patient-years for AS (3054 patient-years exposed), 2.66 per 1000 patient-years for PsA (2249 patient-years exposed), and 2.8 per 1000 patient-years for JIA (353 patient-years exposed), with no significant differences across groups.

Fig 2 shows the number of TB cases among the different TNF inhibitors agents in relation to the total number of cases, with 0.7% of cases occurring with ADA, 1.5% with CZP, 0.2% with ETN, 0.3% with GOL, and 0.4% with IFX. The proportion of cases of TB was significantly higher among ADA users than among ETN users (p = 0.043, Fisher’s exact test). Fifteen patients had prior exposure to another TNF inhibitor agent. The mean time elapsed from initiation of TNF inhibitors therapy to onset of TB was 0.9 ± 0.5 years overall (i.e., for all anti-TNF agents).

Fig 2

Tuberculosis—Proportion related to first exposure to TNF inhibitors in percentage.

*Positive association with tuberculosis. **Negative association with tuberculosis. p = 0.043, Fisher’s exact test.

The time tuberculosis-free drug use (Fig 3) differed significantly across different TNF inhibitors agents, with a higher likelihood of developing TB at first exposure with ADA versus ETN (p = 0.01) and a shorter time to onset of TB with CZP therapy than with any other TNF inhibitors agent (p < 0.05).

Fig 3

A: Survival curves for patients on tumor necrosis factor antagonists as a function of whether they developed tuberculosis during follow-up (log-rank p < 0.05). B: Probability of tuberculosis considering disease-free drug use in years (* p< 0.02; § p<0.01 for all TNF by log rank test).

The variables gender, age, duration of exposure, number of exposures, rheumatic disease and exposure medication were evaluated in the regression model. When performing multivariate Poisson regression analysis, the number of exposures remained associated with the development of TB (RR = 1.17, 95% CI 1.01 to 1.37, p = 0.043) indicating that for each exposure the incidence increased of TB increased by 17%. The first exposure medication also remained after Poisson regression significantly associated with the risk of developing tuberculosis, among medications, ADA presented a 3.11-fold higher risk of TB when compared to ETN (RR = 3.11, 95% CI 1.16 to 8.35, p = 0.024), in addition to this medication, the use of CZP increased the risk of tuberculosis in 7.47-fold compared to ETN (RR = 7.47, 95% CI 1.39–40.0, p = 0.019). The other medications were not significantly different from the risk of ETN (p> 0.5).

The subsample designed to evaluate LTBI treatment enrolled patients from the assisted therapy center of the outpatient rheumatology clinic of Hospital de Clínicas de Porto Alegre. A total of 268 patients in this subgroup were on TNF inhibitors therapy, among these patients, 232 (86%) underwent a tuberculin (PPD) skin test before initiation of TNF inhibitors therapy; 70 (30.2%) tested positive (wheal >5mm). Ninety-three patients underwent repeat tuberculin skin test, of whom 3 ultimately converted to a reactive result. Of the 70 patients with a reactive PPD test, 43 (23.3%) had RA, 20 (50%) had AS, and 7 (24.1%) had PsA; patients with AS were significantly more likely to test positive than patients with RA or PsA (p = 0.004). Seventy-four patients completed treatment for latent TB. There were 5 cases of TB diagnosed during TNF inhibitors therapy. Of these patients, 2 had completed treatment for LTBI, 2 had been nonreactive on PPD skin test, and 1 had not undergone PPD testing (a PPD test had been performed more than 2 years before initiation of TNF inhibitors therapy). The number of cases of tuberculosis was compared between groups with treatment for latent tuberculosis, non-reactive PPD and PPD not known without significant difference between the groups (p = 0.379, p = 0.700 respectively by Fischer’s exact test). There was no significant difference in cases of tuberculosis across these different rheumatic diseases.

Regarding all-cause mortality, there were 250 deaths in the overall sample (4.3%). Among patients with TB, 2 (both with RA) had a TB-related death. As shown in Fig 4, the mean 10-year overall survival rate of these patients by the Kaplan–Meier method was 95.7%, with no statistically significant difference on the log rank test (p = 0.2) among all-cause deaths and deaths from tuberculosis.

Fig 4

Survival in relation patients with TNF inhibitors use with deaths due to general causes and deaths due to tuberculosis.

(p = 0,20 by log rank test). 0 = deaths from all causes. 1 = deaths due to tuberculosis.

Discussion

Brazil is a country with high TB incidence, particularly so in Rio Grande do Sul.[18] Our study found a TB incidence at first anti-TNF exposure of 2.73 per 1000 patient-years exposed. This rate is similar to that found in a previous Brazilian study of RA patients (2.86 per 1000 patient-years)[31] and in studies conducted in the first years of TNF inhibitors prescribing in countries such as Italy and Canada, which reported incidence rates of 2.46 per 1000 and 2.57 per 1000 patient-years exposed, respectively. [14,34] These countries had a much lower incidence of TB in the general population than Brazil. However, as the studies were conducted at the very dawn of TNF inhibitors therapy for the treatment of rheumatic diseases, protocols for LTBI screening and treatment were not yet well established, and the risk of reactivation of granulomatous disease was consequently higher, with a greater number of cases in their populations. The implementation of LTBI therapy revealed a 7-fold higher likelihood of TB reactivation in those exposed to TNF inhibitors therapy who did not follow guidelines for latent TB treatment, and a 78% reduction in the rate of active TB cases among those who did [29,35].

In this sense, our subgroup analysis of patients from the HCPA outpatient rheumatology clinic evaluated LTBI screening and treatment prior to the start of TNF inhibitors therapy. The tuberculin test is recommended in our country for latent tuberculosis evaluation, although it is known that Interferon Gamma Release Assay has better reproducibility and specificity than the tuberculin test, but not yet defined protocol of use in areas with higher incidence of tuberculosis in the general population. [36,37]. A tuberculin test was performed in 86% of patients prior to initiation of immunobiologics, and approximately 30% tested positive, a rate similar to those described elsewhere in the literature.[29,38,39] Comparison across different rheumatic diseases revealed a significantly higher rate of PPD positivity among individuals with AS compared to those with RA or PsA; this finding is consistent with the well-known lower responsiveness of RA patients to tuberculin.[40] However, despite this greater anergy to the Mantoux reaction and consequent risk of false-negative results, this group of patients with RA did not have a higher risk of active TB. Ultimately, there was no significant difference in the rate of incident TB infection among patients with different rheumatic diseases.

As in previous investigations, our study showed a high incidence of TB in patients with rheumatic diseases exposed to TNF inhibitors therapy: 735.29 per 100,000 exposed, approximately 18 times higher than that of the general population of the state of Rio Grande do Sul (39.5 per 100,000).[24]

Considering the different TNF inhibitors agents in use, a higher incidence of TB at first exposure was associated with ADA and CZP therapy, while the lowest incidence was found among patients on ETN. It bears stressing that the only statistically significant difference found on head-to-head comparison was between ADA and ETN, with a positive association of TB cases with ADA therapy and a negative association between ETN therapy and onset of TB, with a risk about 3 times greater of TB development when compared to ETN. This differential risk of TB development in relation to ETN therapy has been reported in previous studies.[7] The lower risk of TB in ETN users has been associated with its mechanism of action. ETN tends to bind to TNF in a less stable manner, which could be associated with less structural damage to granuloma, especially when compared to ADA and IFX.[41-43] We found a trend toward higher incidence of TB among patients on CZP. However, the difference was not statistically significant, probably due to the small number of patients on this drug; however, an interesting—and statistically significant—finding regarding CZP therapy was that it was associated with the highest likelihood of TB development in the first year of exposure compared to the other TNF inhibitors agents. In addition to presenting about 7 times greater risk of developing TB when compared to ETN. Some studies have noted a trend toward frequent TB cases with CZP therapy, but real-life studies of this drug are scarce due to its more recent introduction in relation to other TNF inhibitors agents. Nevertheless, data from clinical trials also support an association with a high number of TB cases.[44] To date, studies with BiobadaBrasil registry data have not shown any incident cases of TB in users of GOL and CZP, unlike in our cohort.[31,45]

In addition to the risk associated with the type of TNF inhibitor, another feature that can be assessed due to the design of our study was the increased risk of developing tuberculosis related to the number of exposures to TNF inhibitors, with a 17% increase in risk each time. although exposure time was not at higher risk as in previous studies [46].

Considering the different rheumatologic diseases evaluated, in our study we found no statistically significant difference between the incidence of tuberculosis in the different rheumatologic pathologies that were under study. Despite the increased risk of developing tuberculosis in rheumatoid arthritis patients compared to the general population, when comparing the risk of tuberculosis with the use of TNF inhibitors among inflammatory arthropathies, our study and previous studies found no significant difference. [24,47]

Regarding the mean time to TB development, active granulomatous disease was detected already in the first year of exposure to TNF inhibitors therapy (mean time elapsed, 0.9 ± 0.5 years). This is corroborated by several studies which have suggested a higher risk of TB development in the first months of TNF inhibitors therapy, a phenomenon attributable to reactivation of latent disease.[46,48-50] In addition, considering the disease-free survival time, there was a significantly higher likelihood of developing TB over the years in patients exposed to ADA than in those exposed to ETN, as well as a significantly higher likelihood of TB onset in the first year of exposure among CZP users in relation to all other TNF inhibitors agents.

Analysis of TB sites revealed a predominance of pulmonary tuberculosis (n = 26, 60.4%), as in previous studies.[34,51] Extrapulmonary TB was found in 13 cases (30.2%), a rate lower than that reported in previous studies of national registry data.[31,50]

In this cohort of TNF inhibitors users, the overall survival rate was 95.7%. All-cause mortality was thus 4.3% over the study period. Among cases of incident TB, there were 2 deaths attributable to TB infection (4.6%). These findings are similar to previous reports of mortality data among TB patients receiving TNF inhibitors therapy. TB-attributable mortality rates are consistently lower than in the general population, probably due to the presence of HIV-coinfected patients in the general population.[52,53]

The present study is the first large investigation to evaluate the incidence of TB in patients receiving TNF inhibitors therapy for rheumatic diseases in the state of Rio Grande do Sul, and one of the first in Brazil to evaluate the incidence of TB in such a large population (n = 5853) with different rheumatic conditions over such a long follow-up period, using real-life, population-based data.

Some limitations of this study are inherent to retrospective designs. As our data were retrieved from existing records, no information could be obtained on LTBI screening and treatment prior to initiation of TNF inhibitors therapy, except in our subgroup analysis. In addition, exposure to TNF inhibitors agents was assessed solely by the proxy indicators of drug dispensing and discontinuation. We did not analyze data on concomitant use of other disease-modifying antirheumatic drugs, for instance, nor did we assess whether TNF inhibitors therapy was used in combination with glucocorticoids, which could constitute an additional risk factor for development of TB in this population. In addition to these limitations in our study we were unable to evaluate possible host-related risk factors for TB such as drug or alcohol use, associated comorbidities, and socioeconomic conditions. We also had fewer prescriptions of infliximab and golimumab, and their negative association with tuberculosis should be interpreted with caution. Nevertheless, this was the largest retrospective cohort study of patients receiving TNF inhibitors therapy for rheumatic diseases conducted to date in Brazil. Its retrospective design notwithstanding, the fact that TB and mortality records were obtained from SINAN and from the SUS mortality information system respectively ensures that these data were reliable, because TB is a notifiable disease in Brazil and the associated mortality data were fed into the mortality information system from official death certificates. Besides that, our study included only patients using the public health system, not including patients with TNF inhibitors prescription through health insurance or private acquisition, and may be a selection bias, however, a Brazilian study showed in the sample that the portion of patients with health insurance represented only about 15% of immunobiological prescriptions.[45]

Conclusions

The use of TNF inhibitors therapy by patients with systemic rheumatic diseases significantly increases the number of incident cases of tuberculosis in this population, with an 18-time increased TB incidence in these patients when compared to the general population. Considering the different TNF inhibitors agents used by our sample, our study demonstrated that a greater number of cases of TB were associated with ADA than with ETN, Finally, we found that newer drugs such as CZP, which have had a much shorter postmarketing surveillance period for assessment of their potential for infectious adverse events, appear to be associated with a higher frequency of incident TB. This highlights the need for proper screening and treatment of latent tuberculosis, as well as the surveillance of new contacts with individuals with tuberculosis, annual screening among people with initial negative screening as recent recommendations, and discussion of the use of interferon gamma release quantification tests more broadly in this population.

2 Sep 2019

PONE-D-19-21256

A POPULATION-BASED STUDY OF TUBERCULOSIS INCIDENCE AMONG RHEUMATIC DISEASE PATIENTS UNDER ANTI-TNF TREATMENT

PLOS ONE

Dear Mrs Sartori,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please address all comments of the Reviewers and all of the statistical analyses need to  be reviewed by a statistician

We would appreciate receiving your revised manuscript by Oct 17 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Mahmoud Abu-Shakra, MD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: General comment.

The paper confirm the increased risk of TB reactivation associated with anti-TNFs in and endemic TB Country. Apart from some English language imprecisions that should be corrected, the report is exhaustive and provides useful data for clinicians.

Specific comments.

Page 3. Line 2. References 1 and 2 should be written in superscript.

Page 4. First sentence. The Authors should correct this phrase by adding that RA itself increases the risk of TB (add the referencei.e. Yeh JJ, et al. PLoS One. 2014 Oct 22;9(10):e110922.

Page 5. Line 3 and followings, The AA. should better address the notification system in Brazil. In some Countries also the suspected cases of TB should be notified. Is the same in Brazil, or are only definite TB diagnoses included? This issue is not negligible in terms of TB incidence calculation. Does the system include information from the patients' clinical chart?

Page 5. Last paragraph. Apart from the cohort observed at the Hospital of Porto Alegre, did the patients undergo LTBI screening? The AA. should indicate this.

Page 5. Last paragrafh, The AA. should indicate if LTBI diagnosis was done by TST only, or, at least in a percentage of patients, by Quantiferon, combined or not with TST.

Page 10. Line 10. The AA. should insert a sentence on the better accuracy of Quantiferon TB Plus test for LTBI screening. (see. Baddley JW, et al. Clin Microbiol Infect. 2018 Jun;24 Suppl 2:S10-S20.

Page 12. Discussion section. Last paragraph. Regarding the study limits, the AA, should add the absence of information on the host-related additional risk factors for TB, including drug or alcohol abuse, comorbidities, socio-economic conditons, etc.

Page 13. Conclusions. The text is redundant and with some repetitions. I suggest to shorten this section by underlining the main results and emphasizing the need of accurate LTBI screening (adding Quantiferon-TB Plus test).

Reviewer #2: This was a cohort study of tuberculosis in rheumatic disease patients who have been treated with anti-TNF. It is useful to get estimates from other parts of the world, especially those with a higher incidence in the underlying population, to better understand what the risk is of this adverse outcome. The linkage of data across registers in Brazil was an excellent way to answer this study question. There are some basic methodological and epidemiological pieces of the study that were not described that need clarification.

Major comments

- The TB cases were based on reporting data. Did these reports depend on TB being culture-verified? Could any TB be reported to the system or did it have to be reported through a lab that found a positive test?

- The study design and patients does not describe person-time (start of follow-up, end of follow-up). Given that there are incidence rates and survival estimates in the paper, how person-time was calculated needs to be included.

- Is this a new-user design? Are the exposures to TNFi all instances of the 1st use of TNFi? Or 1st use of the particular TNFi? This should be made more clear in the methods. In the results it states that the number of exposures was used as a variable in the model, so this indicates that people had previous exposures – of the same TNFi or another TNFi?

- It is odd that incidence is “evaluated as a percentage” when incidence is a rate – the number of cases, per person per year. There is an aspect of time that is missing. It is hard to compare these results to those from other studies when percentages are given.

- A summary of the follow-up time should be included in the results (see STROBE guidelines, if needed)

- Are the survival curves adjusted for anything?

- A table 1 in a cohort study usually has the exposure as the header of the table so that the reader can see how characteristics of the study differ by exposure. This is necessary for the reader to understand other factors, such as age and sex, which are potentially related to starting on a TNFi.

- To have p values for every line of every table and every line in every graph is a lot of tests and does not seem necessary (for example, we can see clearly that the age is different across rheumatic diseases in Table 1, it’s not necessary to do a statistical test for it).

- In Figure 2, it says there was a positive association, compared to what? These are unadjusted for age and sex, which is a major limitation. The reader is interested in rates (unadjusted are ok, sex- and age-standardized are even better) and then the adjusted rate ratios are helpful for comparing rates. Statistically comparing unadjusted rates is not very useful for inference.

- There wasn’t much discussion about potential differences in TB risk in the different rheumatic diseases. This could be a limitation if all diseases are put together in this study.

Minor comments

- In the abstract results there is a typo, it should be “Forty-three cases”

- In the discussion, one usually starts with a first sentence that summarizes the results. It would be a better start to the discussion if the first sentence (describing a recent study) is moved elsewhere and the results of the present study are summarized.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

17 Oct 2019

Dear Publishers,

We respectfully welcome and appreciate the valuable contributions of the reviewers, with their suggestions and correction requests that certainly qualify and give greater consistency to our study.

The following describes the treatment given to each of the changes proposed by the reviewers for resubmission of the manuscript PONE-D-19-21256 (A POPULATION-BASED STUDY OF TUBERCULOSIS INCIDENCE AMONG RHEUMATIC DISEASE PATIENTS UNDER ANTI-TNF TREATMENT).

Reviewer 1:

Introduction:

#Page 3. Line 2. References 1 and 2 should be written in superscript.

Thank you for your observation, the correction in the body of the text (pages 3, lines 1 and 2).

#Page 4. First sentence. The Authors should correct this phrase by adding that RA itself increases the risk of TB (add the referencei.e. Yeh JJ, et al. PLoS One. 2014 Oct

We thank and adjust information about the risks of tuberculosis in this patient population, adding reference suggested by relevance. (page 4, line 1 and 2).

Methods:

#Page 5. Line 3 and followings, The AA. should better address the notification system in Brazil. In some Countries also the suspected cases of TB should be notified. Is the same in Brazil, or are only definite TB diagnoses included? This issue is not negligible in terms of TB incidence calculation. Does the system include information from the patients' clinical chart?

This information has been added to the methods section to clarify the definition of tuberculosis cases in the Brazilian reporting system.

“This system includes all patients with a confirmed diagnosis of tuberculosis, as well as the clinical presentation, treatment regimen and final follow-up outcome.” (Page 5, line 8)

#Page 5. Last paragraph. Apart from the cohort observed at the Hospital of Porto Alegre, did the patients undergo LTBI screening? The AA. should indicate this.

The request has been fulfilled and included in an attempt to provide clarity.

“Screening for latent tuberculosis prior to the initiation of TNF inhibitors is recommended for all patients who are candidates for therapy and drug release is required.” (page 5 – last paragraph)

#Page 5. Last paragrafh, The AA. should indicate if LTBI diagnosis was done by TST only, or, at least in a percentage of patients, by Quantiferon, combined or not with TST.

Thank you for your observation, the correction in the body of the text.

“positive tuberculin skin test or suggestive chest X-ray findings without history of previous tuberculosis treatment” (page 5 – last paragraph).

Discussion:

#Page 10. Line 10. The AA. should insert a sentence on the better accuracy of Quantiferon TB Plus test for LTBI screening. (see. Baddley JW, et al. Clin Microbiol Infect. 2018 Jun;24 Suppl 2:S10-S20.

Relevant information about the use of another method for identifying latent tuberculosis was added to the discussion.

“The tuberculin test is recommended in our country for latent tuberculosis evaluation, although it is known that Interferon Gamma Release Assay has better reproducibility and specificity than the tuberculin test, but not yet defined protocol of use in areas with higher incidence of tuberculosis in the general population” (page 10 – last paragraph).

#Page 12. Discussion section. Last paragraph. Regarding the study limits, the AA, should add the absence of information on the host-related additional risk factors for TB, including drug or alcohol abuse, comorbidities, socio-economic conditons, etc.

Information was mentioned in the discussion as limitations of our study.

“In addition to these limitations in our study we were unable to evaluate possible host-related risk factors for TB such as drug or alcohol use, associated comorbidities, and socioeconomic conditions” (page 12 – last paragraph).

Conclusion:

#Page 13. Conclusions. The text is redundant and with some repetitions. I suggest to shorten this section by underlining the main results and emphasizing the need of accurate LTBI screening (adding Quantiferon-TB Plus test).

The findings have been rewritten to meet the proposed objectives. (page 14 - last paragraph).

Reviewer 2:

Major comments

Methods:

# The TB cases were based on reporting data. Did these reports depend on TB being culture-verified? Could any TB be reported to the system or did it have to be reported through a lab that found a positive test?

This information has been added to the methods section to clarify the definition of tuberculosis cases in the Brazilian reporting system.

“Positive cases being defined by patients with clinical diagnosis associated with bacteriological diagnosis.” (page5, line 10)

# The study design and patients does not describe person-time (start of follow-up, end of follow-up). Given that there are incidence rates and survival estimates in the paper, how person-time was calculated needs to be included.

This information was added to the methods in order to provide more clarity in the definitions used.

“Person- years were calculated from the first day of TNF inhibitor therapy to the date of tuberculosis infection, or death or discontinuation of TNF inhibitor use” (page 6, line 3)

# Is this a new-user design? Are the exposures to TNFi all instances of the 1st use of TNFi? Or 1st use of the particular TNFi? This should be made more clear in the methods. In the results it states that the number of exposures was used as a variable in the model, so this indicates that people had previous exposures – of the same TNFi or another TNFi?

This information was added to the methods in order to provide more clarity in the definitions used. For TNF inhibitor exposure, this is a new-user design.

“For incidence density analysis, we used the time of use related to the first exposure for calculation of incidence related to the first exposure, also the total time of use of any TNF inhibitor during follow-up to define cumulative incidence density over the course of the study.” (page 6, second paragraph)

Results:

# It is odd that incidence is “evaluated as a percentage” when incidence is a rate – the number of cases, per person per year. There is an aspect of time that is missing. It is hard to compare these results to those from other studies when percentages are given.

The incidence density is presented in person time for better comparison with studies as mentioned (page 8 - second paragraph).

As noted, we corrected information referring to figure 2 in the body of the text and in the figure itself, in which we had no intention of presenting incidence rates and density, but only presentation in the number of cases in proportion.

“Figure 2 shows the number of TB cases among the different TNF inhibitors agents in relation to the total number of cases, with 0.7% of cases occurring with ADA, 1.5% with CZP, 0.2% with ETN, 0.3% with GOL, and 0.4% with IFX. The proportion of cases of TB was significantly higher among ADA users than among ETN users (p = 0.043, Fisher’s exact test)” (page 8 – third paragraph).

#A summary of the follow-up time should be included in the results (see STROBE guidelines, if needed)

Adjustment was performed in the body of the text as suggested and oriented guidelines STROBE.

“The average duration of follow-up, ie exposure to TNF inhibitors therapy was 2.58 ± 1.95 years”. (page 7 – sixth paragraph).

# Are the survival curves adjusted for anything?

As described in the body text adjustment was performed.

“The variables gender, age, duration of exposure, number of exposures, rheumatic disease and exposure medication were evaluated in the regression model”. (page 8 – last paragraph).

# A table 1 in a cohort study usually has the exposure as the header of the table so that the reader can see how characteristics of the study differ by exposure. This is necessary for the reader to understand other factors, such as age and sex, which are potentially related to starting on a TNFi.

"Thank you for your relevant comment on table 1. In the submitted version we tried to follow STROBE´s recommendation to give characteristics of study participants (eg demographic, clinical, social) and information on exposures and potential confounders, but we are open to modify it, if necessary. In this population-based study, our primary goal was to estimate the incidence of tuberculosis in patients receiving TNF inhibitors (TNFi) therapy for rheumatic diseases. Considering that all participants were exposed to TNFi, we thought it would be more informative to report on how each disease subgroup could differ, as treatment protocols are unique and TNFi exposure could vary significantly among rheumatic diseases. As the number of TNFi exposures is expected to be a confounding factor, we included this variable in regression model and, ultimately, found it as an independent predictor of TB development, as reported in the results section. Nonetheless, if this is not the best way to present our participants, we would be pleased to change it accordingly."

# To have p values for every line of every table and every line in every graph is a lot of tests and does not seem necessary (for example, we can see clearly that the age is different across rheumatic diseases in Table 1, it’s not necessary to do a statistical test for it).

Thanks for the comment, we are grateful for the observation, however, we understand that keeping the value of p shows the differences more clearly. Despite our understanding, if not necessary, we agree to their withdrawal.

# In Figure 2, it says there was a positive association, compared to what? These are unadjusted for age and sex, which is a major limitation. The reader is interested in rates (unadjusted are ok, sex- and age-standardized are even better) and then the adjusted rate ratios are helpful for comparing rates. Statistically comparing unadjusted rates is not very useful for inference.

A) The comparison was made between users of adalimumab and etanercept.

“The proportion of cases of TB was significantly higher among ADA users than among ETN users (p = 0.043, Fisher’s exact test)” (page 8 – third paragraph).

B) We did not perform adjustment in this case because it is not a rate but only the demonstration of the number of tuberculosis cases by each TNF inhibitor. number of exposures, exposure time, type of medication and rheumatologic disease taken to the regression model.

#There wasn’t much discussion about potential differences in TB risk in the different rheumatic diseases. This could be a limitation if all diseases are put together in this study.

As suggested considering the design of our study we included assessment of the incidence density of tuberculosis due to rheumatologic disease separately.

“Disaggregated by rheumatic disease, the cumulative incidence was 3 per 1000 patient-years for RA (9305 patient-years exposed), 2.61 per 1000 patient-years for AS (3054 patient-years exposed), 2.66 per 1000 patient-years for PsA (2249 patient-years exposed), and 2.8 per 1000 patient-years for JIA (353 patient-years exposed), with no significant differences across groups.” (page 8 – second paragraph).

Minor comments

# In the abstract results there is a typo, it should be “Forty-three cases”

Bug pointed out has been fixed. (abstract - page 2)

# In the discussion, one usually starts with a first sentence that summarizes the results. It would be a better start to the discussion if the first sentence (describing a recent study) is moved elsewhere and the results of the present study are summarized.

The initial discussion session was rewritten and adjusted as suggested for a better presentation of our study findings.

“Brazil is a country with high TB incidence, particularly so in Rio Grande do Sul.18 Our study found a TB incidence at first anti-TNF exposure of 2.73 per 1000 patient-years exposed. This rate is similar to that found in a previous Brazilian study of RA patients (2.86 per 1000 patient-years)31 and in studies conducted in the first years of TNF inhibitors prescribing in countries such as Italy and Canada, which reported incidence rates of 2.46 per 1000 and 2.57 per 1000 patient-years exposed, respectively.” (page 10 – second paragraph)

We are at your disposal for any necessary clarifications.

Sincerely,

The authors

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

28 Oct 2019

A POPULATION-BASED STUDY OF TUBERCULOSIS INCIDENCE AMONG RHEUMATIC DISEASE PATIENTS UNDER ANTI-TNF TREATMENT

PONE-D-19-21256R1

Dear Dr. Sartori,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Mahmoud Abu-Shakra, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Elizabeth Arkema

12 Nov 2019

PONE-D-19-21256R1

A population-based study of tuberculosis incidence among rheumatic disease patients under anti-TNF treatment

Dear Dr. Sartori:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Mahmoud Abu-Shakra

Academic Editor

PLOS ONE