Rheumatology International

, Volume 29, Issue 9, pp 1047–1050

Mycobacterium tuberculosis infection precipitates SLE in patients from endemic areas

Authors

    • Department of Immunology and AutoimmunedisordersNational Institute of Immunohaematology (ICMR)
  • Manisha Patwardhan
    • Department of Immunology and AutoimmunedisordersNational Institute of Immunohaematology (ICMR)
  • Vandana Pradhan
    • Department of Immunology and AutoimmunedisordersNational Institute of Immunohaematology (ICMR)
Original Article

DOI: 10.1007/s00296-009-0903-x

Cite this article as:
Ghosh, K., Patwardhan, M. & Pradhan, V. Rheumatol Int (2009) 29: 1047. doi:10.1007/s00296-009-0903-x
  • 182 Views

Abstract

In a cohort of 70 systemic lupus erythematosus (SLE) patients diagnosed over 2 years period, 14 patients were found to have confirmed antecedent tuberculosis (20.0%) which was 40 times higher (p < 0.001, 95%CI 36.2–48.6) than the prevalence of tuberculosis in the local population. Demonstration of anti-histone antibodies in similar proportion of SLE patients with and without antecedent anti-TB treatment and similar proportion of renal involvement (36 vs. 40%) between anti-TB drug exposed and non-exposed patients ruled out the drugs to be causative factors in precipitating SLE in these patients with antecedent history of tuberculosis. A cohort of 30 confirmed pulmonary tuberculosis patients were also studied in parallel to demonstrate high incidence of autoantibodies in these patients but no SLE. This study suggests a role of prior tuberculosis in precipitating SLE in genetically predisposed patients.

Keywords

SLEM. tuberculosisAutoimmunityEndemic areaDrug induced SLE

Introduction

Systemic lupus erythematosus (SLE) is a major non organ specific autoimmune disorder affecting both the sexes, though women are affected more commonly than men. The hall mark of this autoimmune disorder is the presence of autoantibodies to single stranded and double stranded DNA. In addition, autoantibodies to different proteins associated with DNA and many other autoantibodies to different proteins and drugs have been detected in this disorder. The activity in this disease varies from time to time and is evaluated by clinical SLE disease activity index (SLEDAI) score and diagnosis of the disease is largely clinical on the basis of ARA (American Rheumatological Association) criteria supported by a few laboratory investigations like positive anti-nuclear antibody test. There are several papers both from non-endemic [1, 2] and endemic [3, 4] areas of the world showing an increased prevalence of tuberculosis in SLE patients. The causes were discussed and were mostly related to treatment related immunosuppression in this disease.

However, Mycobacterium tuberculosis is a strong immunogen and its product is used extensively as an adjuvant (Freund’complete Adjuvant) for laboratory induction of a strong immune response, or in the clinics for immunotherapy against various cancers [5, 6]. Hence, there is a distinct possibility that in a genetically predisposed individual infection with M. tuberculosis may precipitate SLE. In the present study, 70 patients of SLE were studied over 2 years for the evidence of past and concurrent infection TB and this report documents prevalence of tuberculosis concurrent or preceding the diagnosis of SLE in India.

Materials and methods

Patients attending Rheumatology clinics in the hospitals in Mumbai over last 2 years were included in this study. The diagnosis of SLE was confirmed on the basis of fulfilling ARA revised diagnostic Criteria for SLE with ANA positivity (>1:300). Detailed serological tests like anti-ds DNA, anti-Sm, anti-SS A, anti-SS B were also done using standard investigation kits (ANA-BioRad, USA), (ds DNA from IFA Euroimmune),(ANA-BLOT from Euroimmune), (anti-histone ELISA from Pharmacia).

Seventy patients fulfilled the criteria of SLE. Thirty patients of confirmed pulmonary tuberculosis as evidenced by clinical features, sputum positivity for acid fast bacilli and positive chest X-ray were also included in this study. All these 30 patients were on anti-tuberculosis treatment for more than 2 months consisting of a combination of Rifampicin, INH, ethambutol and pyrazinamide (First 2 months of therapy). Serum from these tuberculosis patients was collected for testing of ANA, ds-DNA, anti-SS A, anti-SS B antibodies. None of these 30 TB patients had any clinical evidence of SLE. All the patients of SLE and 30 patients of confirmed pulmonary TB were also tested for anti-histone antibodies. Standard tests of statistical significance was done using the online Graphpad programme.

Results

Fourteen patients out of 70 (20.0%) suffering from SLE were found to have suffered and treated for infection with M. tuberculosis. Except one patient (Table 1 RS) all other patients developed SLE within 6 months to 2 years of diagnosis of tuberculosis. They were also treated by a combination of Rifampicin, INH, Ethambutol and Pyrazinamide (for first 2 months in cases of pulmonary tuberculosis). Details of seroreactivity of the 14 cases are presented in Table 1. The serological reactivity of 30 PTB patients is presented in Table 2.
Table 1

Serological and other details of SLE with TB patients

 

Age

Sex

Type of TB

Involvement of kidney

Time to diagnosis of SLE after TB

ANA

ANCA

AHA

ANA BLOT

IFA (Titre + patterns)

IFA

ELISA

WB

VP

29

F

Pulm TB

 

2 years 2 months

>1:300 Speckled

   

PD

30

M

Pulm TB

LN (MPGN)

10 months

1:300 Speckled

  

Sm-RNP

SM

15

F

Pulm TB

 

1 year

>1:300 Homogen

P-ANCA

 

Sm-RNP

TG

26

F

Abd Koch

 

2 years

>1:300 Homogen

 

Positive

SS-B

RS

32

F

Pulm TB

LN (MPGN)

8 years

1:300 Speckled

  

SS-B

MJ

11

M

Pulm TB

 

13 months

1:300 Homogen

 

Positive

Sm

FP

36

F

Pulm TB

LN

12 months

1:300 Homogen

  

Sm RNP/SS-A

SG

36

M

Abd koch

 

6 months

>1:300 Speckled

  

Sm-RNP

AG

9

M

Pulm TB

 

6 months

1:300 Homogen

 

Positive

Sm

RM

17

F

Pulm TB

Type IV LN

12 months

1:300 Homogen

  

Sm-RNP

GA

25

F

Pulm TB

 

9 months

1:300 Speckled

  

SS-A

AS

21

F

Pulm TB

LN

2 years

>1:300 Speckled

 

Positive

SS-B

KP

26

F

Pulm TB

LN

1 year

>1:300 Homogen

P_ANCA

Positive

SS-B

SS

30

F

Pulm TB

 

6 months

1:300 Homogen

P-ANCA

 

SS-B

All samples were anti-ds DNA positive when tested by IFA technique. ANA, anti-nuclear antibody; ANCA, anti-neutrophil cytoplasmic antibody; AHA, anti-histone antibody; WB, ANA Western blot

Table 2

Comparison of autoantibodies in various groups

Characteristics

SLE without TB

SLE with TB

TB patients

No tested

56

14

30

Sex

 Males

5

4

10

 Females

50

10

20

Autoantibodies tested

 ANA(anti-nuclear ab)

56 (100%)

14 (100%)

3 (10%)

 Anti-ds DNA ab

56 (100%)

14 (100%)

0 (0%)

 ANCA(anti-neutrophil cytoplasmic ab)

16 (28.5%)

6 (42.8%)

4 (13.3%)

 AHA(anti-histone ab)

22 (390%)

5 (35.7%)

9(30%)

In short 5/14 patients with preceding tuberculosis infection were positive for anti-histone antibodies whereas 20/56 patients with SLE and no history of tuberculosis (hence no treatment with anti-tuberculosis drugs) positive for anti-histone antibodies. None of the 30 TB, patients had any renal problems (i.e. they have normal serum creatinine and no proteinuria).

Discussion

There are several reports of increased prevalence of TB in patients with SLE in both endemic areas and non-endemic areas [14]. Mumbai is an endemic area for tuberculosis with a prevalence of 500/100,000 population. In our patients we looked into the prevalence of preceding Tubercular infection in SLE patients which turned out to be 14/70 i.e. 40 times higher than the population prevalence (95%CI 36.2–48.6 p < 0.001).

Studies by Balakrishnan et al. from Mumbai showed prevalence of tuberculosis was extremely high in SLE patients. However, they did not look into the prevalence of antecedent TB in these patients. After the diagnosis of SLE, majority of the patients usually receive immunosuppressive treatment which predispose them to tubercular infection. Hence, finding higher prevalence of TB in SLE patients should not be surprising in an endemic area, but what is surprising that this finding applies equally to non-endemic areas of the world also. Moreover, some of the symptoms of SLE itself like chronic lymphadenopathy, pleural effusion, ascites, pericardial effusion and X-ray changes in the lungs may be mistaken for tubercular infection in an endemic area before SLE is diagnosed. Hence, there could be an element of overdiagnosis of TB in SLE patients before SLE is diagnosed if strict criteria for the diagnosis of TB is not followed.

We have used strict criteria where either acid fast bacilli had been demonstrated in the sputum or a biopsy specimen had shown classical caseating granuloma for the diagnosis of the tuberculosis.

Antecedent TB in a patient of SLE leads to following arguments:
  1. 1.

    Association of prior TB with SLE is purely coincidental.

     
  2. 2.

    Anti-tubercular drugs particularly INH is known to produce drug induced Lupus.

     
  3. 3.

    Prior tuberculosis in some way precipitates SLE in a genetically predisposed person.

     
Let us look into each of the argument one by one.
  1. (1)

    Several reports including our own have shown extremely high prevalence of tuberculosis in SLE patients, hence, this association is unlikely to be coincidental, 40 times higher prevalence in antecedent TB in SLE patient is statistically significant.

     
  2. (2)

    Drug induced lupus classically do not cause renal involvement 6/14 (41%) of our patients at presentation with prior tuberculosis and anti-TB drug treatment had renal involvement where as 20/56 (36%) patients of SLE without prior tuberculosis had renal involvement in the present series (Chi square test, p > 0.05). This data is similar to that reported for classical SLE from different non-endemic areas [7]. Proportion of our patients with antecedent TB showing anti-histone antibody positivity were not different from SLE patients who never received anti-TB drugs. More over in many patients in this SLE developed month after anti-TB drugs were stopped. This is not usual for drug induced lupus. Moreover, anti-histone antibodies developed due to anti-tubercular and other drugs have quite different characteristics [8] with reference to anti-histone antibodies.

     

Classical models of experimental autoimmune diseases such as mycobacteria induced arthritis [9, 10, 11] have shown many features of classical autoimmunity. Autoantibodies similar to that found in SLE patients have been regularly detected. [1214] in high proportion of patients with pulmonary tuberculosis similar to this study. A study by Schoenfield et al. showed that monoclonal antibodies raised against M. tuberculosis can cross react with DNA. It was hypothesised that it is possible in individuals with appropriate genetic background, mycobacterial infection could lead to anti-DNA antibodies without clinical manifestations of SLE[1315] this is clearly shown in Table 2 .

Present study partly solves the jigsaw puzzle of the relationship of SLE with tuberculosis by clinical data that a large number of SLE patients (21.4%) in endemic area has antecedent tuberculosis before developing SLE. It may not be out of place to mention here that a study conducted in Romania also found antecedent tuberculosis in a large subset of their SLE patients [16] in early eighties.

It may be argued that patients who are genetically predisposed to develop SLE are also predisposed to develop tuberculosis, but one study tends to show that autoimmunity probably developed as a natural selection to evolution against tuberculosis [17]. Recently one of the antigens Pst S1 from M. tuberculosis was found to be a potent biological response modifiers in addition to various heat-shock proteins that have been isolated from M. tuberculosis [18].

Hence, if we combine all the evidences from experimental autoimmune arthritis, cross reactivity of many monoclonal antibodies against M. tuberculosis to various components of DNA, demonstration of anti-nuclear and anti-histone antibodies in patients with pulmonary tuberculosis and finally the clinical data showing substantial number of patients (40 times than normal prevalence) developing antecedent TB before developing SLE, we are left with a persuasive reason to believe that antecedent tuberculosis may precipitate SLE in a fair number of patients in an endemic area through its immunopotentiating and immunomodulatory action.

Conflict of interest statement

None.

Copyright information

© Springer-Verlag 2009