High Prevalence of Active Tuberculosis in Adults and Children with Idiopathic Inflammatory Myositis as Compared with Systemic Lupus Erythematosus in a Tuberculosis Endemic Country: Retrospective Data Review from a Tertiary Care Centre in India

Aim: Infections are the leading cause of morbidity and mortality in idiopathic inflammatory myositis (IIM) with India being endemic for Tuberculosis (TB). We compared and contrasted the prevalence, clinical profile and outcomes of active TB in IIM with systemic lupus erythematosus (SLE). Methods: Medical records were reviewed for adults and children with IIM (Bohan and Peter criteria) and SLE (ACR criteria) at a tertiary care hospital in India from January 2015 to October 2017. Follow-up was recorded until February 2020 for all those who had developed active TB. Results: Of 167 (132 adults and 35 juvenile) IIM and 280 (131 adults and 149 juvenile) SLE, active TB occurred in 24 (14.4%) IIM (22 (16.7% of 132) adults; 2 (5.71% of 35) juvenile) and 18 (6.4%) SLE [(8 (6.1% of 131) adults; 10 (6.7% of 149) juvenile, p-value < 0.01]. Patients with IIM had higher odds of developing TB as compared with SLE [OR 2.24 (CI 1.5–5.5), p=0.007]. The risk of developing active TB was 68-fold and 30.4-fold higher in patients with IIM and SLE, respectively, as compared with the general population. Extrapulmonary forms were more common (14/24). Nearly half developed TB during active IIM, at a glucocorticoid dose of 0.25 (0–1.5) mg/kg/day. Over a follow-up duration of 27 months (8–184), all were cured of TB, though prolonged course of anti-tuberculous treatment was required in 25%, and five IIM relapsed during treatment. Conclusion: Patients with IIM have increased risk of active TB, with common extrapulmonary forms, slow response, and relapses during treatment.


INTRODUCTION
A diagnosis of Idiopathic Inflammatory Myopathies (IIM) entails significant morbidity, and, at times, even mortality. 1 In the developing world, infections are the leading contributor to such ORIGINAL morbidity. Tuberculosis (TB) is one such infection, that remains a particular challenge in this part of the world. The emergence of drug-resistant strains has only added to the problem. Polypharmacy, in the setting of chronic illnesses, further compounds the problem. Recent estimates suggest the prevalence of TB in India to be 3.2 cases per thousand population. 2 Patients with IIM often require high doses of glucocorticoids for long periods. Poor cough reflex from pharyngeal and respiratory muscle involvement and immobility results in poor handling of both aerosols and oral secretions. Moreover, some patients IIM also have underlying complement pathway defects, such as Mannose Binding Lectin deficiency which is reasonably common and confers a higher risk of TB. 3,4 Children are more likely to harbour such inherited deficiencies. 5 Over the last two decades, public health initiatives in India have attempted to address the issues of rising prevalence, treatment default, and the emergence of drug resistance in TB with partial success. 2 On the same lines, there have been efforts towards decreasing the usage of glucocorticoids in rheumatic disorders too. The changing dynamics of therapeutic practices could have a bearing on the prevalence of tuberculosis in these diseases, and also influence the ways this problem is addressed. Since a diagnosis of IIM entail long term treatment with high doses of glucocorticoids, we sought to determine the prevalence of active TB in adults and children with IIM and compare it with patients diagnosed with Systemic Lupus Erythematosus (SLE), another rheumatic disease that requires high dose glucocorticoids and are associated with higher incidence and prevalence of TB. 6,7 METHODS Medical records from paper charts and electronic medical records were reviewed for adults and juvenile patients with definite and probable IIM by Bohan and Peter criteria 1975) 8 following up at a tertiary care hospital in India from January 2015 to February 2020. The approach was all-inclusive irrespective of a past history of TB. Radiographs of the chest are usually obtained in all patients before beginning immunosuppression, ruling out the possibility of active pulmonary TB at initiation of immunosuppressant (IS) therapy. Patients with drug-induced myopathies, inherited or degenerative myopathies and endocrine, infectious, or metabolic muscle disorders were excluded. Institute ethical clearance for waiver of consent for retrospective data review was obtained [2017-41-IP-76] as per local guidelines. 9 Demographic variables and clinical profiles, as well as disease activity, were noted. Records of adults and children with SLE (ACR criteria 1997) collected for a previous study were drawn for comparison with IIM. 10 Assuming a confidence interval of 95% and a margin of error 2% with a 1.8% prevalence of TB in IIM from an-other study, 11 the sample size was calculated to be 170. An event of TB was defined by either the demonstration of acid-fast bacilli (AFB) in the tissue or fluid obtained or presumptive diagnosis 12 based on clinical and radiological features and improvement with anti-tuberculous therapy (ATT). Information on the organs involved, past or family history of TB, duration of ATT, drug interactions and adverse events, glucocorticoid and other immunosuppressive drugs received were obtained. A glucocorticoid (GC) dose >0.5 mg/kg was defined as high dose steroids. Details on outcomes such as cure of TB, resistance, relapses of TB and the primary disease, and death were recorded. Active disease and relapse of IIM were defined as per physician assessment-based on new/subacute muscle weakness alongside elevated muscle enzymes or weakness requiring an increase in immunosuppression as assessed by the treating physician. A worsening of skin rash was also recorded as flare of disease in those with dermatomyositis or overlap myositis. The presence of interstitial lung disease (ILD) was defined as the presence of lung fibrosis by high-resolution computed tomography and evidence of restrictive physiology on pulmonary function tests. Screening for ILD was done only in those with breathlessness on exertion or crackles on respiratory examination. Chronic liver disease defined as a progressive deterioration of liver functions for more than six months, which includes synthetic functions, detoxification and excretory functions, assessed using a combination of ultrasound liver, liver function tests, coagulation profile and elastography. 13 Overlap myositis defined by presence of myositis and fulfilling one CTD criteria. [14][15][16][17][18] Antisynthetase syndrome defined by Connor's criteria. 19 Complete response, partial response and relapse in SLE as defined by EULAR/ERA-EDTA criteria. 20

STATISTICAL ANALYSIS
Comparisons were drawn between SLE and IIM patients. Descriptive statistics and non-parametric tests were used for comparisons. Results are expressed as median ± 2SD. The analysis was done using GraphPad Prism for Mac (trial version 7).

RESULTS
The IIM cohort had 132 adults and 35 juvenile IIM patients with mean age at diagnosis 32 and eight years respectively ( Table 1 The prevalence of tuberculosis in adults and children with IIM Active TB was seen in 24 (14.4%) IIM patients ( Table  2) with 22 (16.7%) being in adults and two (5.71%) in juvenile IIM cases. Of those with TB, 18 had DM; three had PM, four anti-synthetase syndrome and seven had overlap myositis. Lung was most common focus of TB, 10 patients had pulmonary TB and 2 patients had isolated pleural effusion. Extra-pulmonary TB was seen in 12 patients: Lymph node and disseminated TB was found in four patients each. Four patients had unusual site of involvement. Two had bone TB; one of them had pyomyositis in addition. One patient each had renal and gastrointestinal involvement (Figure 1). TB occurred at median 2 (1-2.5) years after the onset of IIM in adults, and recurrence was seen in one of the 22 (4.5%) cases 5.5 years after complete treatment. On the other hand, TB was diagnosed at the disease onset in one of the two cases of jDM, with a recurrence of MDR sputum positive pulmonary TB 4.5 years later.

Prevalence of tuberculosis in adults and children with SLE
One-hundred thirty-one adults and 149 children with SLE were included for comparison. Of the 131 (8:1 female:male) adult SLE patients (median age 32.4 years, mean follow-up 3 years, 68 (51.9) had lupus nephritis and 4.6 had chronic kidney disease) eight developed Tuberculosis (6.10%). Half the patients developed TB within 2 years and a half were extra-pulmonary (1 lymph node, 1 peritoneal, 1 bone, 1 CNS). Of the 149 (female:male 9:1) paediatric SLE patients (median age 13.7 years, mean follow up 6 years, all had lupus nephritis) and 10 (6.71%) developed TB. Six of 10 had TB within six months of SLE diagnosis and half were extra-pulmonary (2 pericardial, 1 lymph node and 2 disseminated). One of them died of disseminated TB (Figure 1).

Comparison of SLE and IIM patients with TB
The overall incidence of TB in IIM was 14.37/100 patients (2/100 patient years) and in SLE was 6.42/100 patients (2.386/100 patient years). The incidence of TB in adult IIM patient was 16.6/100 patients while in juvenile IIM

Diagnosis of TB in IIM
In almost half the cases (n=13), the diagnosis was based on definite evidence of mycobacterial infection, i.e. smear evidence of acid-fast bacilli (n=7), histology (n=4), or growth in cultures (n=3) ( Table 3). An empiric diagnosis was made in 11 patients. One patient was suspected to have miliary TB, but died before the diagnosis could be confirmed. One patient of DM had culture-proven pre-extremely drug-resistant (Pre-XDR) infection. She improved on therapy but suffered kanamycin induced hearing loss. IIM patients who developed TB had a median disease duration of 2 (0.2-12) years.

Diagnosis of TB in SLE
The diagnosis of TB was empiric in most cases with adult SLE while in seven paediatric lupus it was based    Table 1). None of them was suspected to be drug resistant.

TUBERCULOSIS IN MYOSITIS
The course of TB in IIM Fourteen were followed up after the diagnosis of TB, of which one was lost after one month of treatment. The median GC dose was 0.25 (0.1-1.8) mg/kg/day at the time of diagnosis of TB. Nine of the 13 followed up (69.2%) had 10 adverse events related to ATT: 5 had drug-induced hepatitis, one had pancreatitis, 2 had a skin rash; one of them was related to para-amino salicylate sodium (PAS) (Two patients had more than one adverse effect). Median follow-up duration after the diagnosis of TB was 82 (69.3-126.3) months, while median TB free interval was 72.5 (57-107) months. The median duration of ATT was 9 months (5.3-11.2). A resolution of TB was documented in all who followed up. Two patients had a recurrence of TB, at 66 and 54 months respectively, despite taking a complete course of ATT. One of them developed military TB and the other sputum positive resistant TB. There was one death from TB in this cohort. A prolonged course of ATT was required in 25% of patients as compared with 2% in the general population. Confirmed drug resistance was 3%, similar to the general population.
The treatment with ATT resulted in a relapse of myositis in 5 cases, of which flare of muscle disease was seen in 3 patients, while 2 had cutaneous involvement. However, by the last follow up visit, all patients had attained remission.

The course of Tuberculosis in SLE
Out of 18 cases of TB in patients with SLE, 10 were in remission, five had relapse of SLE during or after completion of ATT and 3 had persistent active disease. Among these two were lost to follow up, 14 were cured of TB and two died. Two out of 290 SLE patients (714 per 100000 patients) suffered mortality, as compared with 32 per 100000 rates in the general population. 2 SLE patients with tuberculosis suffered 22 times higher mortality than that seen from TB in the general population.

DISCUSSION
Idiopathic inflammatory myositis is known to be a debilitating disease with a high risk of infections. We have previously described infections as the leading cause of in-hospital mortality in patients with IIM. 1 Of the various infections, Tuberculosis is one that remains a particular challenge in India. Apart from being a Tuberculosis endemic zone, the emergence of drug-resistant strains has further compounded the picture in this region. We found that 14% of patients with IIM developed TB, and this risk was 68 times higher than the general population, and 2.24 times higher than SLE patients. Extra-pulmonary, atypical, and disseminated forms of TB were more common. A sizeable number required prolonged courses of ATT, although all improved and prevalence of drug resistance was comparable with the general population. Although the literature is replete with case reports of TB at odd sites in IIM patients, data on the prevalence of TB in a cohort of IIM in the developing countries is scant. 21 In  21 Although rituximab use has been considered safer, Gazaix-Fontaine et al. have described TB even after RTX use, suggesting that background disease, as well as TB prevalence in the country, might be contributively. 27 This merits further exploration in larger multicentre studies. Intravenous Immunoglobulin could be a safe therapy for the management of active diseases in the setting of TB. 28 Apart from the activation of latent bacilli, de-novo infection could result from an immunosuppressed state. Both SLE and IIM can, at times, have an associated underlying immunodeficiency, more so in children. 29 Apart from inherited complement defects, most common being C4 deficiency, seen with both SLE and IIM, polymorphisms 141 TITLE and copy number variations in C2, C4, and C4A gene have associated with both SLE and polymyositis. 30,31,32 Of particular note, Mannose Binding Lectin (MBL) deficiency is reasonably common and confers a high risk of Tuberculosis. 33 The higher prevalence of TB in IIM could also be contributed by a failure to clear pooled secretions. Besides, IIM patients often require higher doses of steroids for more extended periods as compared to SLE. Moreover, LTBI is almost universal in the Indian population; hence screening practice for LTBI is not employed in India. 34 The only study on the use of prevention in 97 SLE patients in India reported a remarkable decline in Tuberculosis from 11.6% to 2% over a two-year observation period. 35 Another paper confirmed that in patients who seroconvert for TST while on biologics, Tuberculosis could be prevented with Isoniazid prophylaxis. 36 Such initiatives can go a long way in reducing both morbidities, and possibly even mortality. Thus, prevention can be considered in the TB endemic setting in patients receiving high dose corticosteroids like patients with IIM, keeping in mind the challenges from polypharmacy and consequent pill burden, drug interactions, and risk of emergence of resistant strains. Notably, we found various atypical sites of TB in the current study, paralleling patterns seen in other immunosuppressed states such as HIV. The extrapulmonary TB is common in other rheumatology disease settings such as TNF-inhibitors induced reactivation of TB. 37 Muscle involvement with TB is infrequent; however, one case of the 24 suggesting a possible preference of the TB bacilli to involve inflamed sites with greater vascularity. Interestingly Nagayama et al. have previously described a case of Tuberculous myo-fasciitis in a patient with DM. 38 Subsequently, upon review of the eight instances of such Tubercular involvement in literature, three were found to have IIM. 38 It seems plausible that an inflamed muscle may act as a nidus for infection.
We also noted that a sizeable number of cases developed drug-induced adverse effects. Polypharmacy could complicate the management of TB in IIM. Transaminitis from anti TB treatment can be confused with a relapse of muscle disease, mandating caution while interpreting laboratory tests. With better understanding of the risk, predictors and outcomes of TB in IIM, it may become possible to devise less toxic regimens for management. The strengths of our study include a modest sample size, a comparator group of SLE patients, and follow-up of a subgroup of cases. The limitations are a retrospective nature with the consequent possibility of ascertainment bias as well as under-reporting. There was lack of data on personal history of TB which merits exploration. Further, complete data on immunosuppression received is not available due to retrospective nature of the study. Since the prevalence of TB has gradually changed over the decades with changing treatment practices, disparate time periods of study in the SLE and IIM patients may account for some differences. However, this being the first such study from India assumes relevance as a pilot study. We hope that this exploratory study will pave the way for future prospective studies to understand more about TB in these diseases. The differences in prevalence of TB among various IIM also need to be studied in a prospective study.
Ours being a tertiary care centre receives a referral from more severe illness, and hence prevalence in milder forms of myositis such as amyopathic myositis cannot be ascertained. Besides, the lack of a stringent referral system could have led us to miss cases who reported new symptoms of TB to the general practitioner on follow-up.

CONCLUSION
Patients with IIM have increased risk of active TB, with common extrapulmonary forms, slower response, and relapses during treatment.

AUTHOR CONTRIBUTION STATEMENT
LG, NR and AZ were involved in data collection and collation.
LG, RA and AA were involved in planning and revising the manuscript critically. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.