Nontubercular mycobacterial cough

Summary

Habitual cough suppression leading to non-tuberculous mycobacteria infections and bronchiectasis has been reported. We present a case of a 55-year-old woman with a chronic history of cough with mild expectoration and frequent lower respiratory tract infections, remitting with antibiotic therapy and other supportive measures. She also reported habitual cough suppression for several years. She was eventually diagnosed with Mycobacterium avium complex (MAC) positive right middle lobe bronchiectasis—Lady Windermere syndrome and obstructive sleep apnoea (OSA), causing disabling symptoms limiting her daily activities. We aim to highlight two key issues—diagnosing MAC infections in a tuberculosis endemic country, and OSA and its long-term clinical implications.

Background

Bronchiectasis is a chronic and irreversible respiratory disease, causing cough, sputum production and frequent chest infections with devastating clinical complications, chronic limitation of quality of life and frequent usage of expensive healthcare.¹ The EMBARC (European Multicentre Bronchiectasis Audit and Research Collaboration) study was the first large scale, prospective observational study, conducted across 31 centres in India to study the epidemiology and disease characteristics of bronchiectasis in India. Tuberculosis was identified as the most frequent cause of bronchiectasis. The right lung was much more frequently involved than the left with predilection for middle and lower lobes alike.¹

The predilection of several pulmonary disorders, such as atelectasis, pneumonia, bronchiectasis and other chronic infections, to the right middle lobe (RML) was first suggested by Brock,^{2 3} when he hypothesised that the exceptionally long and narrow course of the RML, along with its poor mucociliary clearance and poor collateral ventilation makes it susceptible to these conditions. He further went on to state that post-tuberculous (TB) bronchiectasis was largely because of caseous necrosis of the lymph nodes surrounding the RML, which caused the deposition of lime salts in the RML bronchus, causing its atelectasis or even bronchiectasis, and termed it as ‘the RML syndrome’.

Although much has been made about tuberculosis and post-TB bronchiectasis, little progress has been made on studying the significance of non-tuberculous mycobacteria (NTM) causing bronchiectasis. Prince et al⁴ from Philadelphia, USA, reported a substantial increase in absolute numbers of Mycobacterium avium complex (MAC) positive bronchiectasis in immunocompetent patients without predisposing pulmonary disorders with a strong predilection to the RML. This was followed by Reich and Johnson⁵ publishing a case series of six elderly women with no predisposing pulmonary disorders, developing RML bronchiectasis due to NTM infections. The authors hypothesised that habitual cough suppression, considered fastidious, may have led to these cases and thus the term ‘Lady Windermere’ Syndrome was coined, from the Victorian-era play, Lady Windermere’s Fan, to convey the fastidious behaviour hypothesised.⁶

In the case presented here, knowledge of RML bronchiectasis in immunocompetent patients without predisposing lung disorders enabled us to take a guided history and consider NTM infection as the cause of bronchiectasis. With this came a heightened awareness of the barriers to diagnose and treat NTM infections, in our country.

We, therefore, aim to bring to light two key issues:

1. Diagnosis of MAC infections in a TB endemic country.

2. Obstructive sleep apnoea (OSA) as a key driver of cardiovascular morbidity and mortality.

Case presentation

A 55-year-old woman, presented to our institute with a 1-year history of intermittent low grade fevers, cough, which was worsening on lying down and sputum production, only a cupful in quantity, initially mucopurulent and later dry. The patient reported that she also had been having exertional dyspnoea for the past 6 months which progressed to grade 4 modified medical research council in the last 10 days. She also reported that she had been having bilateral swelling of her feet and legs for the past 4—5 months. She denied history of haemoptysis, sinusitis, or syncope. She never smoked and had no exposure to biomass fuel usage for heating for the past 30 years. She also denied alcohol use in her life. She denied a history of recurrent respiratory tract infections in her childhood. She had a history of Hypertension for 5 years and takes tablet atenolol 50 mg once daily. Her family history was unremarkable.

The patient’s daughter reported to us that her symptoms first started about 10 years ago with low-grade fever, cough and sputum production which prompted them to consult a medical practitioner close to their home. She was diagnosed with acute bronchitis and an antibiotic course was initiated along with other supportive treatment for symptomatic relief. Symptoms remitted with this treatment, but the patient apparently had several such episodes over the past 10 years. The attendant reported that her symptoms worsened over the past 1 year, limiting her functional status (as a homemaker). On asking for cough suppression, the daughter confided that chronic cough in women was considered a stigma in their community and she used cough suppression whenever she was in public places. The patient also told us that the taste of her sputum was repulsive, which prompted her to suppress cough.

The patient presented to our casualty in a conscious and oriented state and at presentation her pulse rate was 120 beats per minute, blood pressure was 120/70 mm Hg, temperature was 37.56°C and respiratory rate (RR) was 29 per minute. Her point-of-care blood sugar was 120 mg/dL (6.66 mmol/L). Her lips and tongue appeared cyanotic. She did not have clubbing and she had bilateral pitting type of pedal oedema, extending upto her shins. She also had abdominal obesity with a waist circumference of 118 cm. The rest of her general examination was unremarkable. An arterial sample for blood gas analysis was obtained and O2 inhalation via face mask was instituted.

Systemic examination was significant for a raised JVP (jugular venous pressure) with the upper level above the mandible. Measurement of the height of the column was not taken. Auscultation of the heart revealed a loud P2 with a pansystolic murmur, best heard along the left sternal border, accentuating with inspiration. Auscultation of the lungs was significant for coarse, leathery, pan-inspiratory crackles in the right anterior chest with bronchovesicular breathing. She also had fine rales in both lower lung fields posteriorly. She did not have a metabolic flap and the rest of the examination was unremarkable.

A provisional diagnosis of bronchiectasis and left heart failure were made. The ABG (arterial blood gases) revealed a hypercapnic respiratory failure with pCO2 (partial pressure of carbon dioxide) at 65.4 mm Hg (8.71 kPa) and pO2 (partial pressure of oxygen) at 39 mm Hg (5.2 kPa) and pH at 7.12 with HCO3-(bicarbonate) at 26.1 mmol/L. Intravenous furosemide along with non-invasive ventilation (NIV) with bilevel positive airway pressure was initiated with an Fio2 of 40%. The patient reported significant alleviation of symptoms and an ABG taken 6 hours later showed a pH of 7.27, pCO2 at 48.2 mm Hg (6.4 kPa) and a pO2 of 111 mm Hg (14.8 kPa). The HCO3- was at 22.4 mmol/L. Empirical antibiotic therapy was initiated with amoxicillin–clavulanic acid and azithromycin. Supportive therapy with bronchodilator nebulisation, antipyretics with temperature charting and diuretic therapy for heart failure was also advised.

A chest X-ray posteroanterior view taken at presentation is shown in figure 1.

Figure 1.

Chest X-ray PA view—enlargement of the cardiac silhouette with convexity of left atrial appendage. Mediastinal widening in the form of increased vascular pedicle width. Extensive bilateral airspace disease in the form of patchy ground glass opacification involving the basal zones. Upper lobe pulmonary venous diversion in the form of prominent bronchovascular markings (Antler Sign). PA, posteroanterior.

Apart from cor pulmonale, a key feature of left heart failure, evidenced by Stag’s Antler sign⁷ (cephalisation or diversion of the pulmonary veins of the upper lobes in pulmonary venous hypertension) is seen in this chest X-ray.

Laboratory investigations showed haemoglobin of 10.6 g% (106 g/L) with a normal leucocyte and platelet count. Her renal and liver parameters were unremarkable. Ultrasound of the abdomen showed a mild hiatus hernia. Congestive hepatopathy was not seen. PFT was attempted but the patient couldn’t perform the test properly. It was not pursued further. ECG showed normal sinus rhythm with sinus tachycardia. She tested negative for HIV, hepatitis B and hepatitis C. A Mantoux test was performed which showed an induration of 3 mm.

High-resolution CT chest was done the next day and the following findings were noted (figures 2 and 3).

Figure 2.

HRCT chest—axial section—cystic proximal airway bronchiectasis of the right middle lobe with cardiomegaly. HRCT, high-resolution CT.

Figure 3.

HRCT chest—sagittal section—cystic bronchiectasis of the proximal airways of the right middle lobe. Ground glass opacities are also noted. HRCT, high-resolution CT.

Over the course of her hospitalisation, the patient improved with nebulisations and diuretic therapy, relieving her respiratory distress and also improving her functional capacity. However, the patient continued to have persistent tachycardia and oxygen dependence. A transthoracic two-dimensional echo was done which showed an left ventricular ejection fraction of 57%, with concentric left ventricular hypertrophy with an enlarged left atrium chamber at 4.2 cm. She also had a mild-moderate mitral regurgitation. She had severe tricuspid regurgitation with an right ventricular systolic pressure of 91 mm Hg and an enlarged RA (right atrium) at 4.2 cm with a feebly collapsing and plethoric inferior vena cava at 1.8 cm. No pericardial effusion was noted.

Clinical suspicion for OSA warranted a review of history—the patient’s daughter reported that the patient had daytime somnolence and felt fatigued often. She also reported habitual snorning and unrefreshing sleep.

A diagnosis of heart failure with preserved ejection fraction (HFpEF) likely secondary to coronary artery disease and hypertension complicated by OSA was made. She also had superimposed chronic cor pulmonale due to persistent hypoxia secondary to RML bronchiectasis. She continued to have oxygen dependance with her SpO2 at 77%–80% without oxygen support.

Fibreoptic bronchoscopy was done which showed hyperamia and oedema of the RML and right lower lobe. A bronchial wash was done and the fluid was aspirated and sent for microscopy and culture. Microscopy with ZN stain showed non beaded mycobacteria. Cytology showed 23 leucocytes with a slight neutrophil predominance (60% neutrophils and 40% lymphocytes). The fluid was initially sent to Chennai for NTM culture, which was positive for MAC (figure 4).

Figure 4.

Non-beaded alcohol-fast mycobacteria (circled) on light microscopy.

A diagnosis of Lady Windermere syndrome was made and a decision to initiate therapy for MAC+RML bronchiectasis was taken. We prescribed rifampicin, ethambutol and clarithromycin (REC) with doses adjusted to body weight. Over the course of her hospitalisation for more than 3 weeks, the patient had significant symptomatic improvement with diuretic therapy and nebulisations, her oxygen saturation on room air improved to 92%, her oxygen dependence and NIV use reduced. pCO2 and pO2 were normal in subsequent ABGs. At the time of discharge, she had good exercise capacity, with a 6-minute walk test of 350m, her haemodynamics improved and her RR was less than 20. She was also counselled about OSA and was advised to lose weight.

The patient has been on regular follow-up since discharge. She finished a 12-month course of REC therapy. At her last visit, she did not report any symptoms of drug adverse effects or toxicities and her exercise capacity had remained the same. She did not have any fever or cough for the past 10 months. She also reported improvement in appetite and overall quality of life. Her abdominal circumference, however, remained at 118 cm.

Global health problem list

1. MAC versus Mycobacterium tuberculosis—enhancing diagnostic boundaries beyond endemicity.

2. OSA and HFpEF—an Indian perspective.

3. Serological diagnosis of MTB or MAC—the politico-medical landscape in India.

Global health problem analysis

MAC versus Mycobacterium tuberculosis: enhancing diagnostic boundaries beyond endemicity

Much has been made about M. tuberculosis and its natural history that NTM have largely remained overlooked. Brock’s report^{2 3} on ‘the RML syndrome’ in post-TB bronchiectasis provides a reasonable pathological basis for the predisposition of chronic lung disorders like bronchiectasis to the RML.

Over the next 30 years, a few case series reported significant incidence of atypical mycobacteria in immunocompromised patients, particularly in HIV infection. It was not until 1989 that Prince et al⁴ reported an absolute increase in case incidence of atypical mycobacteria positive bronchiectasis. The investigators noted a high prevalence of MAC disease in the RML.

Endemicity of tuberculosis allied to resource inequities (such as, the availability of bronchoscopy, and CT imaging, coupled with unregulated use of serological tests) has often led to initiation of antitubercular therapy (ATT) on clinical suspicion alone. This is further compounded by the low yield of smear positive sputum for TB (56.8%) and an even lower yield for MAC (33.3%).⁸ The same study also demonstrated a high sensitivity and specificity of BACTEC for both MTB and MAC, which often remains elusive for two reasons:

1. National Tuberculosis Elimination Programme (NTEP) has sanctioned only national reference laboratories (NRLs) and intermediate reference laboratories (IRLs) with culture and drug sensitivity testing (C DST) equipment. India has six NRLs and one IRL per state,⁹ which can be difficult to access for rural populations and rural health centres.

2. Although NTEP has sanctioned the use of these facilities in medical colleges,¹⁰ provision of equipment for culture and DST has not materialised.

Even in our institute, it is not standard practice to send specimen for culture, due to the same reasons stated above. However, DST, although for Rifampicin only, is possible due to line probe assays, available at the district NTEP centre.

In our case, a decision to initiate standard ATT was deferred because we considered the possibility of NTM infections as seen in Lady Windermere syndrome, that is, middle, and elderly aged women with no history of underlying lung disorders or immunocompromise. This prompted us to deliberate further, and once microscopy of bronchial wash fluid showed non beaded acid and alcohol fast mycobacteria, a culture had to be obtained for a definitive diagnosis. A possibility of the isolate being a contaminant was also considered, after inputs from the microbiology team. However, the clinical profile of the patient suggested that the isolate was most likely a pathogen.

We, therefore, initiated therapy for NTM in the interim and sent a bronchial wash sample for culture to Chennai, Tamil Nadu, (due to unavailability of BACTEC cultures for NTM in our state’s laboratory) and 4 weeks later, MAC was confirmed on culture.

Medical colleges, both public and private see 26% of suspected TB cases and treat 23% of smear positive TB cases.¹⁰ Therefore, the diagnostic and therapeutic responsibility of medical colleges and its faculty and students is incredibly significant and as such, the NTEP must ensure that all diagnostic modalities are made available to this sector. With this comes, the added benefit of detecting MAC cases definitively and thus directing appropriate treatment.

The unavailability of these diagnostic means (sometimes also including CT imaging and bronchoscopy), coupled with the high endemicity of MTB may lead to falsely low MAC cases.

A study from King George Medical University,¹¹ showed that out of 756 cases with suspected extrapulmonary TB, 227 had culture-positive mycobacteria, of which 165 (72.6%) had M. tuberculosis and a significant 62 (27.4%) had MAC positive on culture.

Several other studies have attempted to focus on the dangers of neglecting MAC¹² and a few have also found out that 88.6% of isolates of MAC are clinically significant.¹³ In our case, the microbiologist suggested that the NTM could be a contaminant. However, in this patient’s clinical context and with the bronchoscopy showing RML hyperaemia and oedema, the NTM isolate was considered pathogenic and therefore treatment for NTM was initiated. Long-term follow-up for 12 months also showed complete resolution of respiratory symptoms.

Therefore, we believe that the NTEP must urgently consider this issue and ramp up its strategies to enhance diagnosis and treatment of atypical mycobacterial infections too.

OSA and HFpEF: an Indian perspective

A literature-search analysis published in The Lancet, suggested that an estimated 1 billion people worldwide, in the age group of 30–69 years have mild to severe OSA hypopnoea syndrome (OSAHS).¹⁴ The study estimated that among 534 million people in India in the same age group, the prevalence of mild OSAHS was likely to be 52 million (defined by Apnoea–Hypopnoea Index ≥ 5 per hour) and the prevalence of severe OSAHS was likely to be 29 million (defined by Apnoea–Hypopnoea Index ≥15 per hour). This estimated prevalence was extrapolated from a two-stage, cross-sectional, community-based study in South Delhi, which screened 2860 patients of which 2505 patients were included in the final study.¹⁵ Characteristically, habitual snoring (452 reported it, of which 290 underwent polysomnography (PSG)—94/290 had diagnostic features of OSAHS on PSG) and abdominal obesity were the strongest risk factors. Other significant risk factors were—high waist-to-hip ratio, hypertension, male gender and a body mass index ≥25 kg/m².

A few other studies have attempted to establish the prevalence of OSAHS in the community (south Delhi and Bengaluru)^{16 17} and have demonstrated a consistently similar and significant 9%–10% prevalence in the 30–69 years age group, among both men and women in a community.

Therefore, it is imperative that OSAHS should be treated as a global health problem, both at the community level, by using screening questionnaires as demonstrated in the studies cited above, and at the individual patient level by enhancing the required diagnostics (PSG), which are woefully in short supply.

In our case, only a presumptive diagnosis of OSAHS could be made, based on a history of excessive daytime somnolence and fatigue, habitual snoring and clinical signs such as nocturnal Spo2 drops ≥4%, lasting for at least 10 seconds, and occurring at least 5 times in an hour. Our patient had 16 and 18 such drops in 1 hour measured on two consecutive nights, respectively.

OSA has been established with several cardiovascular disorders, including but not limited to—hypertension^{18 19} and coronary artery disease²⁰ and HFpEF. It has also been shown that OSA is strongly associated with frequent exacerbations of heart failure leading to hospitalisation.²⁰

A study of 125 patients with resistant hypertension, conducted in Brazil, showed that 64% had underlying OSA.²¹ In another study from Chicago, Illinois, USA,²² the average time elapsed between first recognition by the patient of a major feature of OSAS to sleep centre referral was 87.5 months. Only 4% of referrals were made as a result of the clinician eliciting a history of sleep-related complaints.

Thus, with an estimated 81 million people (52 million with mild OSAHS and 29 million with severe OSAHS) with OSA, Indian clinicians must be aware of the disease epidemiology and impact of OSAHS and must maintain a high degree of suspicion, to elicit appropriate history.

Serological diagnosis of MTB or MAC: the politico-medical landscape in India

In June 2012, the Ministry of Health and Family Welfare implemented a ban, on the import, manufacture and use of serological tests for the diagnosis of MTB. Yet, an estimated 1.5 million serological tests are performed every year in India, in the private sector. This amounts to an annual healthcare cost of ₹82.5 crore (US$15 million).²³

Several studies^24–26 have conclusively demonstrated the inefficacies and inaccuracies of serological tests for diagnosing TB. One study attempted to analyse why serological tests are so popular in the private sector in India²⁷—41 doctors from the private health sector were surveyed—ignorance of NTEP (formerly called Revised National Tuberculosis Control Programme (RNTCP)) guidelines was reported by more than 90%. A few lab technicians, also surveyed in the same study, reported that RNTCP does not give guidance on quality of sputum.²⁷ In this way, serological tests not only contribute to significant healthcare costs but are also poorly sensitive to MTB which further lends to underdiagnosis of NTM.

Further, a study conducted in Southern California, Los Angeles, USA, demonstrated a 24.1% cross-reactivity of IgG TB antibodies against culture confirmed NTM.²⁸ Thus, several potential NTM cases can be misdiagnosed and remain inadequately treated.

The message is abundantly clear—national guidelines have been established for the diagnosis and management of tuberculosis. Although the NTEP has significantly improved diagnostic and therapeutic accessibility to patients with TB, focus on NTM still remains primitive. Specimen culture remains the only way to differentiate MTB and NTM and therefore accessibility to culture, especially in medical colleges should be prioritised. We also would like to highlight the irrational use of serological tests for diagnosis of TB and suggest that the NTEP should take notice and implement stringent measures to curb their use.

Patient’s perspective.

‘My mother feels much better these days and I never thought this could have been possible, given that she had been sick for so long. Both of us were surprised by the fact that cough suppression could lead to this. I’m not sure I understand the message of your case report but if it helps the community, then I hope it gets the recognition.

I have also gotten a better understanding of abdominal fat and blood pressure and sleep apnoea. I think I have it too and I am trying to change my diet and exercise more often to lose weight. For my mother, it is a bit difficult, but her BP has been under good control. Thanks to both of you and the other madam for treating us well.’

- Communicated to us by the daughter of the patient

Learning points.

• Habitual cough suppression causing Mycobacterium avium complex+ right middle lobe bronchiectasis is known as ‘Lady Windermere’ syndrome. The natural history can be in years or often decades and requires a high index of suspicion.

• Non-tuberculous mycobacteria (NTM) infections in immunocompetent patients are common and are significantly prevalent.

• Culture remains the only time-tested way to definitively diagnose NTM infections. Therefore, efforts to ramp up accessibility to equipment and facilities for mycobacterial culture should be prioritised.

• Obstructive sleep apnoea is a severely under-recognised cause of cardiovascular morbidity and mortality. It is now estimated to be the most common cause of secondary hypertension.

• Stringent measures should be taken to curb the use of serological methods for diagnosis of tuberculous. These tests can further undermine the true incidence and prevalence of NTM infections.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s)