Life-threatening respiratory failure requiring extra-corporeal membrane oxygenation secondary to the anti-synthetase syndrome

Abstract

Veno-venous extra-corporeal membrane oxygenation (VV ECMO) provides support in severe acute respiratory failure (SARF) refractory to maximal conventional ventilatory support. ECMO does not treat the lungs per se, but allows time for the underlying condition to reverse or resolve. Common indications include acute respiratory distress syndrome (ARDS) and life-threatening asthma. On occasion, rarer conditions causing respiratory failure are diagnosed during ECMO support. The anti-synthetase syndrome (ASS) comprises of a group of conditions characterised by the presence of anti-aminoacyl-tRNA-synthetase antibodies with one or more of interstitial lung disease, inflammatory myositis and/or arthritis. Mainstay of management is immunosuppression. Here, we present two patients requiring ECMO support for SARF, whose respiratory function failed to respond to usual treatment for their assumed pneumonia. Both showed a rapid improvement in respiratory function and oxygenation once immunosuppressive therapy was instigated. Further testing revealed anti-synthetase antibodies, therefore both went on to receive further immunosuppression and an ultimately good outcome. Despite life-threatening respiratory failure, VV ECMO support allowed time for stabilisation, diagnosis and treatment. Outcomes in acute inflammatory interstitial pneumonitis are improved if immunosuppressive treatment is initiated as soon as possible. Our experience with these two cases have led to an institutional change in practice to send an urgent auto-antibody screen (including extractable nuclear antigen panel) on admission for all our SARF patients.

Case report 1

Our first case is a 47-year-old previously fit and well male, who presented to his local hospital with a three-day history of shortness of breath, productive cough and fever. Clinical and radiological studies indicated multilobar community-acquired pneumonia of unknown organism and he was treated accordingly. He deteriorated into Type 2 respiratory failure, necessitating invasive ventilation. Despite maximal conventional treatment, hypoxia worsened and he was referred and accepted for extra-corporeal membrane oxygenation (ECMO) on day 5. He was commenced on Veno-venous (VV) ECMO support via a right internal jugular 31 F dual lumen Avalon Elite^© cannula, achieving flows of 4.5–5 L/min prior to transfer to his local ECMO centre. Initial blood tests on transfer demonstrated raised inflammatory markers – C-reactive protein 435 mg/L, white cell count (WCC) 26.3 × 10⁹/L with a predominant neutrophilia. Percutaneous dilatational tracheostomy was performed on day 9.

Despite an improvement in inflammatory markers with broad spectrum antimicrobial treatment, oxygenation and compliance issues persisted during the first two weeks of ECMO support. This necessitated 50–60% inspired oxygen concentration and the use of prone positioning alongside ECMO support to achieve an arterial partial pressure of oxygen greater than 6 kPa. Extensive microbiological investigations including sputum, blood, urine and broncho-alveolar lavage (BAL) culture for bacteria, mycobacteria and fungi, respiratory viral PCR from throat swab and BAL, legionella urinary antigen and atypical pneumonia screen (mycoplasma, Q fever and psittacosis) yielded no positive results.

Initial chest X-rays demonstrated severe bilateral consolidation (Figure 1) and a computed tomography (CT) chest on day 10, five days post ECMO cannulation showed extensive consolidation with ground glass opacities affecting 70% of the pulmonary parenchyma, with pulmonary hypertension and right heart strain (Figure 2), which was confirmed by transoesophageal echocardiography. A trial of methylprednisolone (500 mg/day for three days followed by a 14-day tapering steroid regime) to treat acute respiratory distress syndrome (ARDS) with failure to progress as per our institution’s policy was commenced on day 13. From day 16 onwards he demonstrated persistent improvement in oxygenation alongside slow radiological resolution allowing weaning from ECMO support after a successful 14-h trial off on day 28 with acceptable respiratory and gas exchange parameters (PaO₂ ≥ 8 kPa with pH ≥ 7.3 on FiO₂ ≤ 50%), spontaneously breathing with a pressure support of 10–15 cmH₂O over 10 cmH₂O positive end expiratory pressure (PEEP) and respiratory rate ≤30.

Figure 1.

Case 1: Chest X-ray on arrival to our ECMO centre.

Figure 2.

Case 1: CT thorax at day 10.

Unfortunately, 48 h post ECMO decannulation he deteriorated, requiring raised inspired oxygen to 80% and peak inspiration pressures of 30–34 cmH₂O to achieve adequate gas exchange. Immunological testing yielded negative results for antineutrophil cytoplasmic antibodies (ANCA), antinuclear antibodies (ANA), anti-glomerular basement membrane (anti-GBM) and anti double-stranded DNA (anti-dsDNA) with normal immunoglobulin and complement profiles, but on day 32, a standard extractable nuclear antigen (ENA) panel revealed the presence of anti-Jo antibodies. Steroid therapy was recommenced at 60 mg prednisone daily, and, following discussion with rheumatology and respiratory consultant colleagues, rituximab 1 g was administered on day 33. There was a rapid and dramatic improvement in oxygenation and compliance in following days. Although critical care acquired weakness slowed his respiratory weaning, he was successfully tracheostomy decannulated on day 48 and transferred to the ward for ongoing rehabilitation on day 49.

During follow-up at six months, imaging demonstrated bilateral peripheral and mainly lower zone interstitial shadowing consistent with a degree of pulmonary fibrosis (Figure 3), but he reported being back at work part-time with much improved exercise tolerance, and was cycling 2-3km several times a week. He remains on low-dose maintenance prednisone and six monthly rituximab infusions having experienced worsening of respiratory symptoms on both azathioprine and mycophenolate mofetil and is followed up 3–6 monthly by local rheumatology and respiratory teams.

Figure 3.

Case 1: Chest X-ray at follow-up appointment (six months post ECMO de-cannulation).

Case report 2

Our second case is a 39-year-old female with insulin-dependent diabetes, who presented to her local hospital with a seven-day history of productive cough and pyrexia. She was diagnosed with community-acquired pneumonia and treated accordingly. Three days after admission, her condition deteriorated, with increased oxygen requirements and she required intubation, invasive ventilation and renal replacement therapy for an acute kidney injury. As she was requiring a very high concentration of inspired oxygen to maintain an acceptable PaO₂, she was referred and accepted for ECMO support on day 4. VV ECMO was initiated at her local hospital via a 19 F multi-stage drainage cannula in her femoral vein and 17 F return cannula in her right internal jugular vein and she was transferred to our ECMO centre (Figure 4). Flows were maintained at 3.0–3.5 L/min and she was weaned to ‘rest ventilator settings’ (FiO₂ 0.3–0.4, pressure controlled ventilation of 20/10 cm H20 with respiratory rate of 10 breaths per minute). Her admission blood tests were consistent with an infectious aetiology with a CRP of 245 mg/L and WCC of 18.5 × 10⁹/L which was predominantly due to neutrophilia. She underwent percutaneous dilatational tracheostomy on day 9.

Figure 4.

Case 2: Chest X-ray on arrival to our ECMO centre.

She initially seemed to respond well to treatment and was successfully weaned off ECMO support by day 10. Unfortunately, oxygen requirements increased over subsequent days and despite numerous measures including prone positioning and inhaled nitric oxide, her respiratory function worsened and she was placed back on ECMO via a 27 F Avalon Elite^© dual lumen cannula in her right internal jugular vein on day 18 achieving similar ECMO and ventilator settings as before. CT scan post re-cannulation demonstrated frank bilateral lung consolidation with areas of necrosis (Figure 5).

Figure 5.

Case 2: CT thorax at day 18.

Microbiological testing revealed pneumococcus in the original sputum sample from the referring hospital, but all other microbiological tests were returned as negative, including blood, sputum, urine and BAL cultures, urinary antigen for legionella, PCR for respiratory viruses and atypical pneumonia screen. She failed to improve when treated with various antimicrobial regimens. Immunological tests yielded negative results for rheumatoid factor, ANCA, anti-dsDNA, anti-ENA and anti-GBM, with a normal immunoglobulin and complement profile, although a positive ANA was detected (1 in 400 with a positive homogeneous pattern). On closer questioning, her family reported that she had been suffering from muscle and joint pains, diagnosed as fibromyalgia, for a couple of years. Seven days post re-cannulation for ECMO, an extended myositis panel revealed the presence of anti-PL-12 antibodies. That same day she was started on three days of pulsed methyl-prednisolone 1 g once a day followed by maintenance steroids, and her immunosuppressive regime was consolidated with rituximab 1 g on day 24 with rapid and dramatic improvement in gas exchange and compliance. Twenty-four hours later, respiratory function had improved to the extent she could be successfully weaned off ECMO support with acceptable ventilatory and gas exchange parameters as in case 1.

Within a week, her FiO₂ was down to 35%, and she was spontaneously breathing via a tracheostomy mask. Her critical care recovery was complicated by on-going need for renal replacement therapy, a prolonged ileus and critical care acquired weakness. She was transferred back to her local hospital on day 50 to continue rehabilitation. During follow-up at six months, her main symptom was myalgia with no reported breathlessness. Chest X-ray demonstrated no evidence of extensive volume loss or pulmonary fibrosis (Figure 6), and pulmonary function tests were within normal limits. She had regained independent renal function.

Figure 6.

Chest X-ray at follow-up appointment (six months post ECMO de-cannulation).

Discussion

Anti-synthetase syndrome (ASS) is a term used to describe a group of rare chronic autoimmune inflammatory myopathies comprising a subset of polymyositis and dermatomyositis of unknown cause.¹ Clinical features include fever, rash, myositis and myopathy, polyarthritis, Raynaud’s phenomenon, ‘mechanic’s hands’, Gottron’s papules and interstitial lung disease (ILD).

It is 2–3 times more prevalent in women and presents throughout adulthood with a peak age of onset in the fifth and sixth decades of life.²

The hallmark of the syndrome is the presence of serum autoantibodies against aminoacyl-tRNA synthetases. At least 10 different antibodies have been described including anti-Jo-1 (anti-histadyl tRNA synthetase), anti-PL-7 and PL-12, anti-OJ, anti-Zo, anti-EJ, anti-KS, anti-YRS, anti-SRP and anti-Wa – with anti-Jo-1 being by far the most commonly described.¹ The diagnosis is confirmed by the presence of one or more of these antibodies alongside evidence of at least one of:

1. ILD

2. inflammatory myopathy

3. inflammatory polyarthritis affecting small joints.³

Acute respiratory failure as the first presentation in ASS is rare, but reports within the literature are increasing.^3,4 The use of ECMO in these patients is rarer still but there are published case series from Germany⁵ and England⁶ describing the use of ECMO in ILD with at least one patient from each series having a confirmed diagnosis of the ASS.^5,6

Once the diagnosis has been confirmed, the mainstay of treatment is immunosuppression. Corticosteroids are typically used first-line, with other immunosuppressants, including cyclophosphamide, azathioprine, cyclosporine and tacrolimus, used in those who fail to respond to steroids.⁷ It is policy at our institution to consider a trial of IV methylprednisolone in all patients with unresolving ARDS who are still reliant on ECMO support by day 10 (to be commenced before day 14), particularly in those with any suspicion of underlying autoimmune pathology and those with signs of fibrotic change on CT imaging. Whilst the evidence base behind steroid treatment in ARDS is controversial, we have observed that some patients demonstrate dramatic clinical improvement within a few days of starting treatment, whilst others are non-responders. We give a pulse of 500 mg IV methylprednisolone for three days in younger patients with single organ failure, followed by a tapering regime in responders and use a more conservative dose of 2 mg/kg infusion over 24 h in older patients or those with on-going need for multi-organ support with a slower taper. The initial improvement in case 1 was most likely secondary to the dose of intravenous methylprednisolone with subsequent deterioration related to weaning and discontinuation of steroids. It is perfectly possible that many or all of the ‘steroid responders’ we encounter have a previously undiagnosed underlying steroid-responsive condition causing or perpetuating their ARDS.

Rituximab is a chimeric monoclonal antibody which targets a cluster of 20 different antigens (CD20) found on the surface of B-cells. It is licensed for primary treatment of certain lymphomas, granulomatosis with polyangitis and microscopic polyangitis and in severe active rheumatoid arthritis resistant to other disease-modifying antirheumatic drugs, but is increasingly used off label for a wide range of auto-immune and inflammatory diseases.^8,9 There are published case series describing the use of rituximab to good effect in patients with treatment refractory ASS.^10–12 However, its use as second line following steroids in the absence of other immunosuppressants or in combination with cyclophosphamide is limited to case reports¹³ Rituximab has been associated with a number of adverse events in the critically ill, most relating to developing new infections, but also isolated cases of cytokine release syndrome, a potentially life-threatening systemic inflammatory reaction. Pulmonary toxicity (organising, desuamative, interstitial or granulomatous pneumonitis) has also been reported.¹⁴ As in the cases described above, patients receiving rituximab should always be pre-medicated with steroids, antihistamine and antipyretic agents to minimise the risk of infusion reactions.

In both of these cases, rituximab was advised by rheumatology and respiratory colleagues due to previous experience with the use of rituximab in ASS for both induction and maintenance of remission, the life-threatening nature of their respiratory failure and to minimise consequences on future fertility.

Prognosis in ASS is most affected by the degree of ILD and any irreversible fibrosis that may have developed underlining the vital importance of early diagnosis and treatment.¹ Both of our patients required immunosuppression to reverse their respiratory failure. Results from ENA testing, particularly the extended panels, may take some time, and the tests are not available universally. Immunosuppression in the face of potential infection, without serological confirmation of appropriate autoantibodies, risks causing significant deterioration. Within our institution we have changed practice to send an ENA panel on admission in all our severe acute respiratory failure (SARF) patients, with an extended myositis panel sent after discussion with rheumatology and respiratory colleagues in those with a suggestive history of ASS. Case 2 also highlights the vital importance of taking a thorough history from family members, as her joint pains, previously labelled as fibromyalgia, may have expedited discussion with rheumatology colleagues and led to an earlier diagnosis and treatment had their importance been appreciated on admission.

Conclusions

Both these patients presented with life-threatening respiratory failure requiring ECMO support for presumed community-acquired pneumonia but failed to improve with standard treatment. When this occurs, it is imperative not to get fixated on the initial diagnosis and to consider a wide differential of alternative diagnoses. In both of these patients, positive antibody results led us to diagnose ASS and treat with immunosuppressive agents, with dramatic improvements noted shortly thereafter. Outcomes are improved if immunosuppressive treatment in acute inflammatory pneumonitis is commenced rapidly, but this must be balanced with substantial risks of immunosuppression in a patient with a severe infective pneumonia. It is therefore vital to diagnose these patients promptly. As a direct result of these cases, we have changed our clinical practice to request an urgent auto-antibody screen including ENA panel in all of our SARF patients on admission.

ASS should be remembered as a rare form of acute ILD presenting with all the clinical and radiological hallmark features of a severe community-acquired pneumonia, and the inflammatory ILDs must be considered in the differential diagnosis of all presumed infective pneumoniae that fail to respond to appropriate antimicrobial treatment.