Diffuse alveolar haemorrhage associated with subsequent development of ANCA positivity and emphysema in three young adults

A 14 year-old young woman with lethargy, exertional dyspnea, microcytic hypochromic anemia, persistent cough, and multiple episodes of haemoptysis was referred to the Royal Brompton Hospital (RBH) respiratory paediatric service. Lung function was characterized by a mild restrictive pattern (forced vital capacity (FVC) 73%, forced expiratory volume in 1 s (FEV1) 77.5%, total lung capacity (TLC) 77%, carbon monoxide transfer factor (TLCO) 85% and transfer coefficient (KCO) 110%). HRCT revealed widespread ground glass opacification throughout both lungs (Fig. 1-a). A bronchoalveolar lavage (BAL) revealed increasingly haemorrhagic returns and abundant haemosiderin laden macrophages. A surgical lung biopsy showed findings consistent with DAH (Fig. 2). In the absence of identifiable associations, the patient was diagnosed with IPH and started on hydroxychloroquine. However, her disease remained inadequately controlled with frequent flares over the years, with poor compliance a possible contributor.

Aged 21, she was admitted to the RBH adult interstitial lung disease (ILD) unit with worsening breathlessness, frequent “infectious” exacerbations, and daily small volume haemoptysis. At the time, she denied any connective tissue disease (CTD) symptoms. She only had an intermittent smoking history over the previous 3 years, totaling one pack-year. Autoimmune screen was negative. Although atypical non-specific anti-neutrophil cytoplasmic antibody (ANCA) staining was found by indirect immunofluorescence [NOVA-lite, INOVA Diagnostics Inc., San Diego, USA]), both MPO and anti-proteinase 3 (PR3) were negative (fluorescence enzyme immunoassay on Immunocap 250 analyser [Thermo Scientific ImmunoDiagnostics, Milton Keynes, UK]). A chest HRCT again revealed diffuse ground glass opacification, in addition to interlobular septal thickening (Fig. 1 b). On lung function, she had developed mild obstruction and TLCO had significantly worsened (48%), without echocardiographic evidence of pulmonary hypertension. Over subsequent years, varying doses of oral and intravenous corticosteroids were used. Both Azathioprine and Methotrexate were not tolerated. Mycophenolate mofetil achieved control of the haemoptysis for a number of years, although with the need of ongoing prednisolone associated with significant weight gain. In her late twenties, the intercurrent haemoptysis recurred. Mild renal impairment was identified with microscopic haematuria but no red blood cell casts.

At the age of 32, the patient developed p-ANCA, with MPO positivity of 3.5 u/ml, confirmed on a second sample (27 u/ml). On HRCT, in addition to the known ground-glass opacities, new emphysema was evident (Fig. 1 c-d), distributed alongside pulmonary vessels, in keeping with interstitital emphysema. She was started on intravenous cyclophosphamide with partial response followed by Rituximab which was continued at six monthly intervals. At her last follow up visit, 10 months after initiation of Rituximab, she reported improved breathlessness and no further episodes of haemoptysis. Renal function was stable although slightly abnormal and lung function remained stable with mild obstruction and disproportionate reduction in DLCO and KCO, suggestive of the development of emphysema.

A 34-year-old man was referred to the RBH ILD unit with a two-year history of breathlessness on exertion and recurrent haemoptysis and anaemia requiring blood transfusions. Initial investigations at his local hospital included a chest HRCT scan showing diffuse ground glass opacity (Fig. 3 a), a negative autoimmune screen, normal lung function tests and a surgical lung biopsy histologically characteristic of DAH (Fig. 4). The patient was a farm tractor driver and his job involved spraying farming chemicals and working in grain stores, although he always wore a protective mask. He was an ex-smoker, having stopped 10 years prior to presentation (5 pack-year). He denied CTD symptoms. A repeat autoimmune screen including ANCA was negative apart from mild ANA positivity with a speckled pattern. Remaining blood tests revealed hypochromic microcytic anemia, and positive anti-thyroid antibodies. While positive anti-transglutaminase IgA antibodies were found, gastroenterology review including duodenal biopsies excluded coeliac disease. On BAL, Pearls stained pigmented macrophages were seen. Lung function tests were characterized by normal spirometry and mild gas transfer reduction (FEV1 127%, FVC 120.5%, TLC 106%, TLCO 84% and KCO 76%). He was started on low dose prednisone with resolution of the haemoptysis. He remained functionally and clinically stable for several years, slowly tapering the prednisolone dose. However, 6 years later he started complaining of episodic small volume haemoptysis and arthralgia, responsive to short courses of increased dose prednisol one. A repeat HRCT revealed more obvious ground glass infiltrates (Fig. 3 b). Eight years from first presentation p-ANCA with MPO positivity was identified (14 u/ml), confirmed on a second test (12 u/ml). In view of ongoing episodic small volume haemoptysis, methotrexate was added to background low dose prednisolone. Despite relatively good control of haemoptysis, with only occasional episodes consistently responsive to increased short course of prednisolone, eight years after his initial presentation, minor areas of emphysema began to appear on HRCT (Fig. 3 c) which progressively worsened becoming more evident in HRCT 2 years later. Perivascular emphysema was again observed (Fig. 3 d). In parallel with CT changes, a progressive worsening in his gas transfer was observed, with essentially stable spirometry but TLCO of 41% and KCO of 42% and normal echocardiographic evaluation.

A 16-year old young man was transitioned to the Royal Brompton ILD service from the paediatric respiratory department with a diagnosis of IPH. He had the first episode of haemoptysis aged 7 and since then experienced recurrent episodes of haemoptysis, shortness of breath and cough. He was given a diagnosis of IPH based initially on symptoms, recurrent shadowing on imaging (Fig. 5 a) and negative autoimmune screen. Interestingly, aged 6 he was diagnosed with Hashimoto thyroiditis. Exacerbations of his IPH were strongly associated with flare ups of his thyroiditis. Until the age of 10, his symptoms and imaging findings responded to intermittent courses of oral corticosteroids of various doses.

Aged 11 he was admitted to the local hospital with haemoptysis and severe respiratory failure. He was intubated and transferred to the paediatric intensive care unit at RBH. He underwent BAL and surgical lung biopsy. The BAL revealed large numbers of haemosiderin-laden macrophages on Perl’s stain. The biopsy showed DAH without supporting evidence of vasculitis (Fig. 6). His HRCT at the time of the biopsy demonstrated diffuse ground glass opacity, perivascular emphysema and a small cavity in the right lower lobe (Fig. 5 b) which raised suspicion for vasculitis but repeat ANCA testing was negative. He was treated with antibiotics and intravenous corticosteroids with good response and resolution of the cavity on subsequent imaging. His spirometry prior to this acute episode showed FEV1 41% and FVC 37%.

Upon discharge, he was commenced on maintenance treatment with hydroxychloroquine, low dose prednisolone and azithromycin 250 mg three times weekly. Short term courses of increased doses of prednisolone were used for flare ups. A repeat HRCT 4 years after the biopsy showed extensive emphysema with a perivascular distribution, and superimposed diffuse ground glass opacities (Fig. 5 c). The patient had never smoked nor had he ever been exposed to passive smoking.

When he transitioned to the adult ILD clinic and 8 years after the biopsy, aged 19, the radiological picture had further progressed and was characterized by extensive emphysema and pleuroparenchymal fibroelastosis (PPFE) of both upper lobes (Fig. 6 d). Lung function was severely impaired with FEV1 54%, FVC 46%, FEV1/FVC 99%, TLCO 33% and KCO 68% reflecting the combination of severe emphysema and PPFE. A repeat autoimmune screen revealed a p-ANCA with MPO positivity (4.6 u/mL). Following these results, he was started on low dose azathioprine, with ongoing hydroxycloroquine and azithromycin, with current symptomatic and functional stability.

The development of specific antibodies +/− other typical autoimmune extrapulmonary manifestations after the initial pulmonary presentation have been described in a number of scenarios [3‐6]. Specifically, to the development of ANCA MPO positivity, longitudinal assessments have revealed new onset of MPO+ or PR3+ and/or overt vasculitis on follow up in patients initially presenting as idiopathic interstitial pneumonias [3].

However, there are only a few reports detailing subsequent autoimmune serology (other than ANCA) conversions following an initial diagnosis of IPH [7, 8]. Le Clainche et al. [4] report on 15 children diagnosed with IPH, of which only one developed ANCA on follow up (specificity not reported). However, as the children were not screened for ANCA at presentation, it is impossible to conclude whether this was a real seroconversion.

ANCA MPO seroconversion occurring in adulthood in IPH after many years has been described, to our knowledge, only in another case [5]. However, the patient presented by Freitas et al. was a 21 year old girl diagnosed with IPH at the age of 4, who developed ANCA positivity 12 years after presentation, aged 16, and had a good response to immunosuppression. Furthermore, no emphysematous changes on follow up were mentioned. By contrast, our cases developed ANCA positivity in adulthood, and all continue to require ongoing immunosuppressive treatment to control the DAH, suggesting the possible role played by autoimmune phenomena in this rare disease, even when initially deemed “idiopathic”. In support of this, the second case had positive anti-transglutaminase IgA antibodies and autoimmune thyroiditis and the third case had autoimmune thyroiditis.

During follow up, our cases developed cystic opacities consistent with emphysema, not seen at presentation. In the first patient, limited emphysema was first seen 10 years after presentation. The cystic areas became more prominent over the following 6 years, with concurrent isolated reduction in DLCO. The second patient was first noted to have developed emphysematous changes on HRCT 8 years after the initial presentation, and 20 years after having completely stopped smoking, with further significant worsening of radiological emphysema after another 2 years. In parallel, there was marked worsening in gas transfer with persistently normal lung volumes. The third patient was diagnosed with IPH in early childhood and was first noted to have emphysema on HRCT 4 years after the diagnosis, aged 11, with subsequent progression despite the lack of active or passive smoking exposure.

The development of emphysema in the context of interstitial processes even outside of a significant history of smoking is being increasingly recognized. In a cohort of never-smoker patients with rheumatoid arthritis related ILD, 27% had associated emphysema with obstructive functional indices [6], while in systemic sclerosis related ILD patients with concomitant emphysema, the prevalence of never smokers was 33% [9]. In a cohort of 233 hypersensitivity pneumonia (HP) patients evaluated for the presence of emphysema and PPFE, 23% of patients with emphysema were never-smokers [7].

In AAV, development of emphysema has been described in a few retrospective case series and case reports [8, 10‐12]. In a cohort of AAV patients with 79 granulomatosis with polyangiitis (GPA) cases and 61 microscopic polyangiitis (MPA) cases, global emphysematous changes incidence was 13% with no significant differences between the two groups [12]. In the largest published series so far, Yagamata et al. reported emphysematous lesions in 37% of patients in a cohort of MPA patients, more frequently characterized by low attenuation areas rather than cystic lesions, although the Authors did not link the emphysematous lesions with alveolar haemorrhage per se [13].

The pathogenesis of emphysema on a background of DAH and ANCA positivity is unclear. One possibility is that the emphysema seen in our three cases could be secondary to the inflammatory process present within the capillaries/small vessels leading to the destruction of the alveolar walls though the release of noxious insults including proteolytic enzymes and free oxygen radicals. The observation of a striking peribronchovascular distribution to the emphysema supports this hypothesis. This has been suggested in the context of capillaritis [14] and in hypocomplementemic urticarial vasculitis, a rare immune complex related small vessel vasculitis, characterized by low C1q and C4 complement levels. Serial histopathological analysis of one case described by Hunt et al. [15] showed that the initial changes of capillaritis evolved into emphysema, and obstructive airways disease. Several other case reports have presented this association in the context of hypocomplementemic urticarial vasculitis [16‐22].

It is also possible that the ANCA antibodies themselves have pathogenic activity. Neutrophils can be activated by MPO-ANCA causing the release of MPO, a cationic protein expressed by several immune cells including neutrophils and macrophages, resulting in an inflammatory response [23]. In an animal model of smoke-induced emphysema, a myeloperoxidase inhibitor prevented progression of the emphysematous changes [24] Further support for a role played by autoantibodies and specifically ANCA, comes from the finding of significantly increased proportion of combined pulmonary fibrosis an emphysema (CPFE) among patients with ANA positivity and a higher MPO-ANCA positivity compared to idiopathic pulmonary fibrosis (IPF) patients without emphysema. Furthermore, CPFE patients with ANA and/or ANCA positivity showed an increased number of CD20+ cells forming lymphoid follicles within the fibrotic interstitium in areas adjacent to fibroblastic foci [25].

In the first two cases, we cannot completely exclude that the development of cystic changes and/or emphysema was related to the previous history of smoking, and certainly cigarette smoking may have contributed to its development. However, the first case had a truly trivial smoking history (one pack year), and stopped by age 21, when her HRCT showed no cystic/emphysematous lesions. The second patient had a 5-pack year history but he stopped and had been a non-smoker for 10 years by the time he had his initial HRCT, which did not reveal any emphysematous destruction. Furthermore, the third patient developed extensive emphysema at a very young age with no exposure to cigarette smoke, supporting a direct link between DAH, ANCA positivity and emphysema.

Further studies are required to better understand the relationship between DAH, ANCA positivity and development of emphysema. ANCA positivity seems to be retrospectively associated with a worse outcome in children with IPH [26]. Also, the development of emphysema in the contest of CTDs and HP has been associated to a worse prognosis [6, 7, 9].

IPH is a rare disease and little is known about the mechanisms involved in its development. However, its known association with coeliac disease [27], its response to immunosuppressive therapy and its possible association with autoimmune features [4, 28, 29] suggests that a contribution of immunological overactivity, even if a clear cut autoimmune disease is not identified.

Immunosuppressive treatment should always be considered in these patients even in the context of an idiopathic onset. Patients with IPH should have ANCA screening not only at presentation but also during follow up, especially when the response to therapy is poor or in presence of suspected renal involvement. However, it is worth noting that ANCA positivity is not always identified in AAV, such that the presence of ANCA is not essential to confirm the diagnosis or to start immunosuppressive therapy, if needed [30]. Full lung function tests including DLCO should be monitored frequently even in case of stable spirometric values, as subtle worsening due to development of emphysematous changes could develop even after many years. This could aid in identifying the cases which could benefit from a more aggressive therapy and perhaps prevent the development of complications such as emphysematous changes.