Rapidly progressive course of pauci-immune pulmonary capillaritis in a 70-year-old Asian male refractory to immunosuppression and plasma exchange

Abstract

A 70-year-old man presented with acute respiratory failure, alveolar infiltrates and haemoptysis requiring supplemental oxygen. Flexible bronchoscopy with bronchoalveolar lavage identifies diffuse alveolar haemorrhage. Clinical and serological evaluations do not identify a precise aetiology and histopathology establishes the diagnosis of isolated pauci-immune pulmonary capillaritis. The patient received induction therapy with high dose methylprednisolone at 1000 mg/day for 5 days and weekly rituximab at 375 mg/m² scheduled over 4 weeks. Although the patient demonstrated clinical improvement after the first week, he experienced a rapid relapse requiring mechanical ventilation. His induction rituximab regimen was continued and plasma exchange was initiated. Despite these therapies, the patient’s condition deteriorated and passed away. Our case adds insight to the management of this rare entity and describes the use of plasma exchange as salvage therapy.

Background

The diffuse alveolar haemorrhage (DAH) syndromes are a heterogeneous group of rare disorders characterised by bleeding into the alveoli as a result of inflammatory injury to the alveolar arterioles, venules or capillaries. Its aetiology is commonly autoimmune and can be characterised as a primary pulmonary vasculitis (eg, antineutrophil cytoplasmic antibody-associated vasculitis (AAV), anti-glomerular basement membrane (GBM) disease, polyarteritis nodosa) or secondary to medications, malignancy, infection or systemic vasculitis (ie, systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), systemic sclerosis and inflammatory myositis).¹ Clinical findings may include haemoptysis, radiographical alveolar infiltrates, iron deficiency anaemia and hypoxaemia. In the setting of a systemic vasculitis, glomerulonephritis (GN) may also be seen. When there is absence of clinical and serological evidence to support a specific aetiology and the lungs are the sole manifestation of disease, the DAH is classified as either a pauci-immune pulmonary capillaritis or idiopathic pulmonary haemosiderosis. The former is characterised by histological evidence of alveolar capillaritis (ie, neutrophilic infiltration of the alveolar septa, loss of capillary structural integrity and erythrocyte infiltration into the alveolar-interstitial space) and the latter is characterised by the absence of said features.²

General principles of management for DAH include supplemental oxygen therapy, antimicrobial therapy for superimposed infections, corticosteroid therapy for attenuating the vasculitis and targeted immunosuppression based on the specific aetiology of DAH. In contrast to AAV or anti-GBM disease, there is a paucity of data and no clinical guidelines for the management for isolated pauci-immune pulmonary capillaritis. In this report, we describe a rapidly progressive course of isolated pauci-immune pulmonary capillaritis and chronicle our management strategies to add to the body of literature regarding this rare entity.

Case presentation

A 70-year-old man of Burmese descent presented to the emergency room with a main problem of progressive lower extremity oedema, decreased urine output and worsening dyspnoea on exertion. These symptoms developed over the past 2 months despite careful diuretic management in the outpatient setting by his nephrologist. Associated symptoms include coughing with scant haemoptysis, declining exercise tolerance over the past 2 months, anorexia and 4.5 kilograms of unintentional weight loss.

He has a previous medical and surgical history of chronic kidney disease stage 4 (attributed to hypertensive nephrosclerosis and diabetic nephropathy), diabetes mellitus on insulin therapy, diabetic retinopathy, hypertension, coronary artery disease status postcoronary artery bypass grafting and hyperlipidaemia. Prescribed medications include furosemide 80 mg two times a day, nifedipine XL 90 mg once a day, aspirin 81 mg once a day, clopidogrel 75 mg once a day, atorvastatin 10 mg once a day and ranolazine 500 mg two times a day. He did not take any herbal or homeopathic remedies. Family history was significant for ‘pulmonary haemorrhage’ in his older sister who developed and passed away from the condition in her 50s. Social history is relevant for monogamy, never-smoker and no biomass fuel exposure, no industrial dust or metal exposure, no pets at home, no alcohol consumption and no recreational drug use. He initially immigrated from Burma 20 years prior to admission and denied any history of tuberculosis or other pulmonary infections.

Vital signs on presentation were temperature 36.1°C, pulse rate 51 beats/min, blood pressure 160/94 mm Hg, respiratory rate 24 breaths/min, oxygen saturation 93% on room air, body mass index 31.32 kg/m². Physical examination was relevant for the presence of jugular venous distension, bilateral rales with a basilar predominance, scattered upper airway wheezing and 4+ dependent oedema. Notably, there was an absence of abnormal heart sounds, purpuric or petechial skin rash, fingernail clubbing, synovitis, facial sinus tenderness or neuropathy.

Blood work results revealed mild anaemia with haemoglobin of 126 g/L, leucocyte count of 8.1×10⁹ / L, platelet count 198×10⁹ / L, normal coagulation profile, blood urea nitrogen of 32 mg/dL, serum creatinine (Scr) of 3.4 mg/dL (from previous value of 1.5 mg/dL 1 year prior), B-type natriuretic peptide (BNP) of 14 400 pg/mL and troponin-I of 0.029 ng/mL. Urinalysis (UA) revealed 2+ proteinuria, 2 red blood cells per high power field and four white blood cells per high power field with a specific gravity of 1.025 and pH 5.0. ECG demonstrated evidence of a prior inferior wall infarct and a chronic right bundle branch block but without acute findings.

Chest X-ray (CXR) 4 days prior to admission revealed mild cardiomegaly and no pulmonary infiltrates or effusions (figure 1A). In contrast the admission CXR shows bilateral alveolar infiltrates (figure 1B). CT of the chest revealed prominent central airspace opacities with peripheral sparing and bilateral small pleural effusions with associated compressive atelectasis (figure 2). Lung ultrasonography revealed heterogeneous distribution of B-lines throughout the lung fields with areas of pleural sparing (figure 3). Transthoracic echocardiography demonstrated left ventricular ejection fraction of 50% without regional wall motion abnormalities, significant valvular disease, elevation in pulmonary artery systolic pressure or any right ventricular dysfunction.

Figure 1.

Chest X-ray of the patient. (A) Taken 4 days prior to admission showing mild cardiomegaly but relatively clear lung parenchyma. (B) Taken on admission showing diffuse patchy alveolar infiltrates.

Figure 2.

CT of the lungs. These three axial images are taken at different anatomical locations. Findings include bilateral small pleural effusions greater on the right side and reaching the upper third of the right thorax and the middle of the left thorax. Mild compressive atelectasis is seen. Prominent central airspace opacities with peripheral sparing are noted.

Figure 3.

Lung ultrasound. This image demonstrates B lines emanating from the pleural line representing the alveolar-interstitial syndrome.

Initial treatment with intravenous diuresis with supplemental oxygen was performed to treat cardiorenal syndrome. Despite achieving a net negative volume balance, improvement in his dependent oedema and a return to his baseline SCr of 1.5 mg/dL, he had persistent pulmonary alveolar infiltrates and worsening hypoxaemia requiring intensive care unit (ICU) admission for bilevel positive pressure airway administration.

Investigations

After stabilisation of respiratory failure, the patient underwent a right heart catheterisation (table 1). Results were consistent with precapillary pulmonary hypertension (PH) but the degree of PH could not explain the patient’s degree of hypoxaemia. Because of an increasing quantity of haemoptysis to three spoonful per day, the patient underwent endotracheal intubation and a decision was made to perform flexible bronchoscopy (FB) with bronchoalveolar lavage (BAL).

Table 1.

Right heart catheterisation

Haemodynamic pressures	mm Hg
PA (S/D/M)	56/12/30
PCW (S/D/M)	17/28/17
RA(a/v/M)	9/9/6
RV (s/edp)	62/13
TPG	13
Haemodynamic oxygen saturation	Hgb (g/L)	% Saturation	O2 content (mL O2/dL)
Ao	83	76.1	8.7
PA	83	46.7	5.35
Haemodynamic calculations
Heart rate (beats/minute)	56
Body surface area (m2)	2.03
Height (cm)	170
Weight (kg)	91.2
CO by Fick (L/min)	6.67
CI by Fick (L/min/m2)	3.29
VO2 (mL/kg/min)	223.58
VO2 indexed (mL/kg/min/m2)	110.4
Haemodynamic resistances
PVR (Wood units)	3.95
PVR (dsc−5)	315.58
PVR indexed (Wood units)	1.95
PVR indexed (dsc−5)	155.84

Ao, aorta; CI, cardiac index; CO, cardiac output; PA, pulmonary artery; PCW, pulmonary capillary wedge; PVR, pulmonary vascular resistance; RA, right atrium; RV, right ventricle; S/D/M, systolic/diastolic/mean; TPG, transpulmonary gradient; VO2, oxygen consumption.

FB revealed normal airways without endobronchial lesions but copious fresh blood and clots were seen in the airways. BAL of the left lower lobe and right lower lobe was performed with progressively bloody alveolar fluid with sequential fluid aspiration (figure 4). Cytology demonstrated the presence of haemosiderin-laden macrophages and absence of malignant cells. BAL fluid cell count was significant for 17% lymphocytes (box 1). Culture data were unrevealing for bacteria, acid-fast bacilli or fungal organisms. The diagnosis of DAH was established.

Figure 4.

Bronchoalveolar lavage showing progressively bloody return indicating diffuse alveolar haemorrhage (starting from the syringe on the right of the image and progressing in a left-ward direction).

Box 1. Results of the bronchoalveolar lavage fluid analysis.

BAL fluid cell count

RBC count 0.0485×10¹²/L

WBC count 800/mm³

Differential

Monocytes/macrophages (57%)

PMN (26%)

Lymphocytes (17%)

Eosinophils (0%)

Basophils (0%)

BAL cytology findings

Haemosiderin laden macrophages

Lysed red blood cells

Positive Prussian blue staining

Negative for malignancy

At the time of diagnosis, the SCr remained at baseline and repeat UA and microscopic examination did not show a nephritic sediment. A spot urine protein-to-creatinine ratio was 0.6. Therefore, a percutaneous renal biopsy was deferred. Serological tests for primary and secondary pulmonary vasculitis and pulmonary-renal syndrome were performed (table 2) but were unrevealing.

Table 2.

Results of serological Testing

	Result	Reference range
Anti-nuclear antibody	Negative	Negative
C-ANCA/proteinase 3 antibody (AI)	Negative	Negative
CCP antibody, IgG (units)	<16	<16
Centromere antibody (AI)	<1.0	<1.0
C3 (mg/dL)	108	80–160
C4 (mg/dL)	66.2	16–48
DsDNA antibody (IU/mL)	2	<30.0
GBM antibody IgG (AI)	<1.0	<1.0
Histone antibody (U)	<1.0	<1.0
HIV 1/2 antigen/antibody	Negative	Negative
HBsAg	Negative	Negative
HBsAb	Positive	Negative
HCV Ab	Negative	Negative
Jo-1 antibody (AI)	<1.0	<1.0
P-ANCA/myeloperoxidase antibody (AI)	Negative	Negative
Rheumatoid factor	Negative	Negative
RNP antibody (AI)	<1.0	<1.0
Scl-70 antibody (AI)	<1.0	<1.0
SS-A antibody (AI)	<1.0	<1.0
SS-B antibody (AI)	<1.0	<1.0
Sm antibody (AI)	<1.0	<1.0
Transglutaminase IgG (U/mL)	1	<15

C-ANCA, cytoplasmic-anti-neutrophil cytoplasmic antibody; CCP, citrullinated cyclic peptide; dsDNA, double-stranded DNA; GBM, glomerular basement membrane; HBsAb, hepatitis B surface antibody; HBsAg, hepatitis B surface antigen; HCV, hepatitis C virus; HIV, human immunodeficiency virus; P-ANCA, perinuclear anti neutrophil cytoplasmic antibody; RNP, ribonuclear protein; Sm, Smith.

Differential diagnosis

The primary differential diagnoses for a patient with respiratory failure, alveolar infiltrates and haemoptysis include congestive heart failure due to cardiomyopathy or valvular disease, necrotising pneumonia, endobronchial malignancy with parenchymal metastases, pulmonary infarction and DAH.

The differential diagnosis for acute kidney injury (AKI) superimposed with the aforementioned clinical and respiratory findings include cardiorenal syndrome associated with worsening cardiac function, pulmonary-renal syndrome associated an acute GN, sepsis-related AKI as a consequence of bacterial pneumonia, interstitial nephritis associated with a primary viral pulmonary infection or zoonotic infection (eg, leptospirosis), systemic amyloidosis from an underlying paraproteinaemia and renal vein thrombosis associated with pulmonary embolism and infarction.

On initial evaluation of our patient, the cardinal features of weight gain, progressive oedema and dyspnoea in the setting of a rise in SCr all pointed to the cardiorenal syndrome. This was supported by UA findings of a concentrated urine, acellular urine sediment and appropriately acidic urine pH. The 2+ urine dipstick protein was likely a consequence of urinary concentration as well as a reflection of his underlying proteinuric chronic kidney disease. Further, there was no initial clinical or laboratory evidence of infectious, infiltrative, thrombotic or vasculitic processes.

On treatment with an adequate course of intravenous diuresis, many of the patient’s clinical metrics (ie, weight, BUN, SCr, jugular venous pressure and leg oedema) of cardiorenal syndrome improved. However, the persistence of radiographic infiltrates, respiratory failure and right heart catheterisation numbers shifted the differential diagnosis towards non-cardiogenic etiologies of respiratory failure such as DAH, bronchoalveolar malignancy, fungal lung infections such as Aspergillus and Pneumocystis and mycobacterial disease.

Once the BAL results confirmed the diagnosis of DAH and excluded infection, the differential diagnosis expanded to systemic vasculitis (eg, granulomatosis with polyangiitis, microscopic polyangiitis, eosinophilic granulomatosis with polyangiitis, antiphospholipid syndrome and Henoch-Schonlein purpura), antibody-mediated disorders (eg, anti-GBM disease, coeliac disease and connective tissue disease-related immune complex disease), drugs (eg, cocaine, trimellitic anhydride, isocyanates and D-penicillamine), AIDS, diffuse alveolar damage and pulmonary veno-occlusive disease.

Treatment

On the diagnosis of DAH, the patient started on pulse dose methylprednisolone 1 g for a 5-day course. Despite completion of this steroid regimen, the haemoptysis persisted, and the alveolar infiltrates progressed on CXR. Surgical lung biopsy (SLB) was debated as it was initially felt the risks of thoracoscopic lung biopsy were excessively high in a patient in respiratory failure. However, after the initial week of therapy and observation, the patient’s oxygenation improved to only requiring 2 L of nasal cannula oxygenation. Ultimately, SLB was performed with a right video-assisted thoracoscopic surgery with a wedge biopsy of the middle and upper lobes at areas of diffuse parenchymal inflammation. The wedge biopsy pathology was sent to a referral centre for interpretation and revealed diffuse acute and chronic alveolar haemorrhage with areas of organising connective tissue suggestive of a healing capillaritis (figure 5A). Haemosiderin laden macrophages are seen accumulating within the alveolar spaces and interstitial spaces, signifying the presence of prior alveolar haemorrhage (figure 5B, C). Immunofluorescence microscopy did not demonstrate any staining.

Figure 5.

Surgical lung biopsy. (A) Low power view (H&E stain) showing diffuse acute and chronic haemorrhage. There are also patchy areas of organising connective tissue centred on alveolar walls and other areas with polypoid collections of alveolar fibrin, macrophages and haemosiderin. No granulomas, active neutrophilic capillaritis or vasculitis are seen. (B) High power view (H&E stain) highlighting the organising connective tissue, alveolar fibrin, macrophages and neutrophils. Vascular congestion is present. (C) High power view (Prussian Blue stain) demonstrating numerous haemosiderin laden macrophages.

In the absence of clinical or serological evidence of a precise autoimmune aetiology, the diagnosis of idiopathic isolated pauci-immune pulmonary capillaritis was established. Methylprednisolone was continued at 125 mg two times a day intravenously, azathioprine (AZA) was initiated at 50 mg two times a day with plans for a gradual uptitration to 2.5 mg/kg/day, and rituximab (RTX) induction was started at 375 mg/m² once weekly for 4 weeks. Thiopurine methyltransferase genetic analysis was sent prior to initiation of AZA and was negative for poor metaboliser alleles. Pneumocystis pneumonia (PcP) prophylaxis was also started using daily double-strength trimethoprim-sulfamethoxazole.

Over the course of the second hospitalisation week, the patient’s haemoptysis, oxygen requirements and radiographic infiltrates gradually improved to the point the patient was liberated to room air and was discharged to a subacute rehabilitation facility on oral prednisone 60 mg once a day with plans for outpatient continuation of induction rituximab therapy and gradual down-titration of prednisone.

Outcome and follow-up

One day after his discharge to a rehabilitation facility, the patient developed haemoptysis again and hypoxaemia to 76% on room air. He was admitted to the ICU for endotracheal intubation and mechanical ventilation for worsening hypoxaemic respiratory failure and rapidly redeveloping radiographic infiltrates on CXR. Further investigations showed oliguria with a rise in SCr and drop in haematocrit, increase in weight of 2.5 kg from discharge, development of jugular venous distension and increased central venous pressure (as estimated by inferior vena cava ultrasonography) compared with prior examinations. The prime differential diagnoses for his acute decompensation included worsening of his cardiorenal syndrome, progressive DAH as a result of de-escalation of corticosteroid dosage and opportunistic infections such as PcP.

Pulse dose methylprednisolone was reinitiated at 500 mg intravenous every 12 hours, AZA was discontinued due to leucopenia, the RTX induction regimen was continued and intravenous diuresis dosage was stepped up to attain a net negative volume balance. In light of ongoing haemoptysis, bronchoscopy was repeated and showed bloody mucoid plugs in the dependent bronchi with persistent BAL evidence of DAH. Repeat BAL culture data demonstrated absence of bacterial, mycobacterial or fungal organisms.

Over the course of the following 5 days, the patient’s oxygenation requirements worsened despite attaining a net negative volume balance and escalating SCr. With the diagnosis of refractory DAH despite an optimal immunosuppressive regimen, the decision to initiate plasma exchange (PLEX) was made. The rationale was to use PLEX as a last-line salvage therapy despite not having sufficient clinical and pathological evidence of an immunoglobulin or immune-complex-mediated disease process. The initial prescription was at an exchange rate of 50 mL/kg with fresh frozen plasma and albumin replacement.

Despite these extensive medical efforts, the patient developed refractory hypoxaemia with progressive pulmonary infiltrates after the second plasma exchange that was refractory to diuresis and escalating titration of positive end-expiratory pressure. The patient underwent cardiac arrest and passed away subsequently.

Discussion

Role of surgical lung biopsy

When clinical data are unrevealing in a patient with DAH and rapidly progressive respiratory failure, SLB is indicated. In this setting, the objectives are to exclude infection and malignancy and to diagnose a potentially treatable aetiology by ascertaining the pathogenic mechanism of disease. For instance, isolated anti-GBM disease limited to the lungs (occurs in 5%–10% of cases) can be diagnosed by immunofluorescence showing a linear IgG staining pattern of pulmonary capillary basement membranes.^{3 4} A granular immune complex deposition of pulmonary capillary basement membranes is seen in SLE.⁵ In other cases, DAH due to AAV can show neutrophilic infiltration of the capillary walls without involvement of larger vessels.⁶

A review of the literature has revealed many recommendations for the management of the common autoimmune aetiologies of DAH such as AAV and anti-GBM disease.³ However, the literature on the management of isolated pauci-immune pulmonary capillaritis is sparse and there are no consensus treatment guidelines or large clinical trials to guide the specific management of this disease. Therefore, management choices are based on expert opinion and are often derived from the data obtained from clinical trials on the more common aetiologies of DAH. Data from an SLB should be used when extrapolating treatment algorithms from clinical trials of other established aetiologies of vasculitis.

Isolated pauci-immune pulmonary capillaritis

Pauci-immune pulmonary capillaritis is an idiopathic form of isolated DAH that presents without other clinical features of systemic vasculitis or circulating auto-antibodies such as antineutrophil cytoplasmic antibodies (ANCA), anti-GBM antibodies or antiphospholipid antibodies. In a series of 36 patients with biopsy proven pulmonary capillaritis, only 4 patients were considered to be idiopathic.⁷ Median follow-up of these patients was 116.5 months and there were no subsequent manifestations of systemic disease.

Pathologically, pauci-immune pulmonary capillaritis is diagnosed by the identification of fibrinoid necrosis of alveolar and vessel walls and the presence of apoptotic cellular debris and haemosiderin-laden macrophages along with the absence of immunofluorescent staining for immune complexes or immunoglobulins.^{2 8} AAV can also have a similar pathological presentation, though usually would have clinical evidence of other small-vessel involvement (eg, palpable purpura, GN, destructive airway lesions, pulmonary nodules or cavities, mononeuritis multiplex and/or sinus disease) and presence of circulating ANCA. ANCA testing has good sensitivity for AAV (80%–90%)^{9 10} and the specificity and positive predictive values will depend on the clinical presentation and range can approach 98% in a patient presenting with multisystem manifestations of disease.^{11 12} Because DAH was the sole manifestation of small-vessel vasculitis in our patient, this reduces the overall diagnostic accuracy of the ANCA test.¹³

In our patient, the SLB demonstrated a healing capillaritis without clear evidence for a specific pathogenical mechanism. Because there was no extra-pulmonary evidence of a specific small-vessel vasculitis, it was difficult to ascertain whether this represented a true case of idiopathic pulmonary capillaritis or a late presentation of AAV that no longer demonstrated circulating ANCA. This made the initial diagnosis and selection of a tailored immunosuppressive regimen and prognostication challenging.

Immunosuppression strategies

In the specific management of DAH, use of a high dose ‘pulse’ methylprednisolone should be given once a diagnosis is established even before the precise aetiology is elucidated.¹⁴ Additional immunosuppressive agents, such as cyclophosphamide (CYC) or RTX, are indicated when there is the concomitant presence of GN, vasculitis or progression of DAH despite pulse corticosteroid therapy.¹⁵ It is not known if randomised-controlled trial proven strategies for vasculitides such as AAV or anti-GBM disease may be fully generalisable to pauci-immune pulmonary capillaritis.

Because the pathogenesis of isolated pauci-immune pulmonary capillaritis is unknown, best practices are anecdotal and based on centre experience and case series. Given the phenotypic similarities between isolated pauci-immune pulmonary capillaritis and idiopathic pulmonary haemosiderosis, it is reasonable to assume some degree of generalisability given the documented response of the latter to corticosteroids (CS),¹⁶ AZA¹⁷ and CYC.¹⁸

In the largest series of eight patients with isolated pauci-immune pulmonary capillaritis, the majority presented with respiratory failure requiring mechanical ventilation. Following treatment with high-dose CS and CYC, seven out of the eight patients survived the index hospitalisation.^{6 19} However, two out of these seven patients experienced relapse. Notably, none of these survivors in the 4-year follow-up period developed any clinical or serological evidence of systemic vasculitis. More recently, this regimen was successfully used in a paediatric patient with isolated pauci-immune pulmonary capillaritis.²⁰

Role for rituximab

In the current era of vasculitis therapies, the role of the anti-CD20 monoclonal antibody RTX has been expanding and many centres now prefer its use over CYC. It is known that suppression of B-lymphocytes production alone does not necessarily correlate with clinical response. This is shown by data from the RA literature has shown that two doses of RTX at 1000 mg can lead to nearly complete suppression of B-lymphocyte which nadir at week 16 and only slowly rebound over the course of 1 year but without effect on clinical response.²¹ Analogously, pharmacological depletion of anti-double-stranded DNA in SLE patients did not lead to clinical responses in the EXPLORER trial.²²

Despite this, it is increasingly clear in large clinical trials that RTX does lead to demonstrably improved remission rates in AAV. The landmark RAVE trial demonstrated that induction therapy with RTX was non-inferior to CYC for the induction of remission with a comparable side effect profile.²³ The RITUXVAS trial showed RTX had a comparable rate of sustained remission, a similar median time to remission and no significant adverse events at 12 months compared with CYC.²⁴ Further, longer term studies in AAV have been able to show that an induction regimen of RTX followed by placebo led to superior remission outcomes compared with a regimen of CYC followed by AZA.²⁵ In one cohort of patients with AAV, maintenance RTX given at 6-month intervals was found to be superior to maintenance AZA with significantly lower relapse rates.²⁶ With this abundance of data, the paradigm of management in AAV has clearly shifted towards favouring initial use of RTX over CYC. In a case report, RTX was used successfully to induce remission of idiopathic pauci-immune pulmonary capillaritis after failure of high-dose CS and CYC.²⁷

Role for plasma exchange

PLEX has a central role in the management for anti-GBM disease. However, guidelines are absent regarding its use in other aetiologies of DAH.²⁸ Based on expert opinion, it may be reasonable to consider patients with severe or progressive DAH refractory to CS and other immunosuppressive agents.²⁹ For PLEX to be a rational choice as a blood purification technique for the treatment of vasculitis, there must be a pathogenic antibody or other protein to be removed extracorporeally. Further, at least five separate treatments over a 7-day to 10-day period is usually required to remove 90% of a patient’s initial total body burden of these substances.³⁰

Its efficacy in the management of vasculitis was first demonstrated in Goodpasture’s syndrome in a randomised-controlled trial of 17 patients (all with advanced renal insufficiency and pulmonary involvement) in which PLEX and immunosuppression use led to a significantly higher rate of antibody reduction.³¹ A large retrospective series of 71 patients with anti-GBM disease demonstrated that the use of PLEX (50 mL/kg, maximum 4 L) and use of prednisolone and CYC appear to confer better long-term survival over 20 years for those with an SCr under 500 μmol/L (5.65 mg/dL).³²

For the management of AAV, early non-standardised and uncontrolled clinical trials suggested a potential benefit for the addition of plasma exchange to immunosuppression alone. The MEPEX trial, a randomised controlled trial of 137 patients with AAV with advanced renal insufficiency randomised to PLEX or pulse dose methylprednisolone, appeared to show improvement in renal recovery and a suggestion of overall survival based on the renal recovery status at 6 weeks.³³ However, the recent PEXIVAS trial has tempered the enthusiasm for PLEX in AAV as it did not reduce the risk of end-stage renal disease or death in patients with AAV.³⁴

Conclusion

Despite the increasing armamentarium of immunosuppressive medications, there has been no large-scale clinical trials or any clear data regarding the clinical course of isolated pauci-immune pulmonary capillaritis. Given the abundance of data on the use of immunosuppressants, namely RTX, in the management of AAV and the use of PLEX in anti-GBM disease, expeditious use of these therapies for DAH due to pauci-immune pulmonary capillaritis may be reasonable given many similarities between these disease entities.

To summarise, we present a rare case of DAH due to isolated pauci-immune pulmonary capillaritis that is refractory to treatment with high-dose corticosteroids, intravenous rituximab and plasma exchange. Our case highlights the concept that data from other clinical trials of DAH may not be generalisable to the management of isolated pauci-immune pulmonary capillaritis despite phenotypic similarities between these conditions.

Learning points.

• Diffuse alveolar haemorrhage (DAH) syndrome is characterised by small vessel vasculitis at the alveolar and septal level and has numerous aetiologies.

• When DAH is not accompanied with clinical or serological evidence of primary or systemic vasculitis and is accompanied by histopathological evidence of capillaritis, the entity is termed isolated pauci-immune pulmonary capillaritis.

• The clinical course of isolated pauci-immune pulmonary capillaritis is not well described but appears to portend a more favourable prognosis than other primary pulmonary vasculitides (eg, anti-neutrophil cytoplasmic antibody-associated vasculitis).

• There is a paucity of rigorous evidence-based management strategies for the management of isolated pauci-immune pulmonary capillaritis. Strategies vary by treatment centre expertise and include corticosteroids, cyclophosphamide, rituximab and azathioprine.

• The role of plasma exchange in isolated pauci-immune pulmonary capillaritis is unknown but may be reasonable as salvage therapy for rapidly progressive DAH refractory to high-dose corticosteroids.