Rapid development of severe acute respiratory distress syndrome after abatacept treatment in a patient with rheumatoid arthritis

Abstract

Abatacept is a biological agent that modulates T-cell costimulation by blocking CD28 signalling. This cytotoxic T-lymphocyte-associated antigen-4-Ig fusion protein was approved for treatment of rheumatoid arthritis (RA). However, a few case reports have revealed respiratory failure after abatacept treatment. In this report, we present a patient with RA who developed severe acute respiratory distress syndrome (ARDS) and who passed away 2 months after starting abatacept. A comprehensive analysis including radiology, blood examinations, infectious panel and flow cytometry lymphocyte analysis was done to determine the cause of respiratory failure. Since no infection was detected in this patient, an association between ARDS and abatacept is a strong possibility due to significant adverse reactions to the biological agent. Considering the rapid progression of respiratory failure after abatacept treatment in this report, we suggest that pulmonary function testing and lung structure evaluation be regarded throughout the early stage of treatment of patients with RA.

Background

Rheumatoid arthritis (RA) is a chronic inflammatory joint disease characterised by cartilage and bone damage.¹ However, this syndrome also results in extra-articular manifestations, such as rheumatoid nodules, atherosclerosis and lung disease.² Cardiovascular disease and interstitial lung disease (ILD) are the primary contributors to premature death due to RA, with an overall mortality between 10% and 20%.^{3 4} Subclinical interstitial lung abnormalities may be detected in 30%–50% of patients with RA, but the individual risk of progression to clinically significant ILD is not known.⁴ This failure could be associated with extra-articular manifestations of RA, comorbid conditions or drug therapy. Several studies have shown that biological and non-biological disease-modifying antirheumatic drugs (DMARDs) can result in lung toxicity and exacerbation of the underlying ILD.² Abatacept is a biological DMARD that appears to be an effective treatment for patients with RA-associated ILD.⁵ However, a few case reports have shown that abatacept can decrease lung function.^{6 7} In this report, we present a patient with RA who showed rapid exacerbation of severe acute respiratory distress syndrome (ARDS) after abatacept treatment.

Case presentation

The patient was a 74-year-old female non-smoker with medical records indicating double-seropositive RA of 30 years’ duration, Sjogren’s syndrome, arterial hypertension, dyslipidaemia, secondary pigmentation due to hydroxychloroquine, osteoporosis and calcaneal pseudoarthrosis.

The patient used methotrexate (15 mg/week), leflunomide (20 mg/day), prednisone (5 mg/day), folic acid (5 mg/day), celecoxib (200 mg/day), omeprazole (20 mg/day) and losartan (50 mg/12 hours), However, her clinical activity according to the Disease Activity Score Calculator for Rheumatoid Arthritis (DAS-28) score was 5.57 (high activity), with an anticyclic citrullinated peptide >3200 and a rheumatoid factor of 133. She had never received biological therapy until starting her treatment with abatacept (125 mg weekly), which reached five doses before discontinuation. One week after the last dose, she experienced dyspnoea and a productive cough without fever or haemoptysis. She was admitted to the emergency department 1 week later, and after the initial evaluation she was diagnosed with community-acquired pneumonia and was discharged with oral clarithromycin. Two weeks later, the patient came back to the emergency department in worse condition, with mild respiratory distress without hypoxaemia. Physical examination indicated the presence of crackles in both lower lobes of the lungs. She refused hospitalisation and was discharged with oral levofloxacin for 5 days.

Four weeks later, the patient was admitted to the emergency room for severe dyspnoea. The initial evaluation indicated normal mental status, tachycardia and poor distal perfusion without hypotension. Respiratory evaluation showed severe distress, tachypnoea, hypoxia (calculated oxygen saturation=94%) when using 100% oxygen inspired fraction (FiO₂) and diffuse bilateral crackles. She was transferred to the intensive care unit (ICU) and was started on empirical treatment with vancomycin and piperacillin/tazobactam.

Investigations

The first chest CT after discontinuing abatacept (figure 1A, B) found ground-glass opacities with centrilobular and paraseptal emphysematous bullae that compromise both the lower lobes, with signs of ILD probably due to long-term use of methotrexate.

Figure 1.

CT shows pulmonary distress progression in time after the last dose of abatacept. (A) Coronal and (B) axial slices of chest CT show distress progression with ground-glass opacities and centrilobular and paraseptal emphysematous bullae with early signs of diffuse interstitial disease of fibrotic appearance; changes can be secondary to long-term use of methotrexate. (C) Coronal and (D) axial slices of chest CT show distress progression with similar images 5 weeks after abatacept suspension. (E, F) Different axial slices of chest CT with distress progression 9 weeks after abatacept suspension.

The second chest CT (figure 1C, D) and X-ray showed progression of distress, without lung consolidation. Laboratory examinations revealed normal white cell count and mild anaemia. Arterial blood gases showed a pH of 7.494, carbon dioxygen partial pressure of 29.0 mm Hg, bicarbonate ion of 21.8 mmol/L and PaO₂:FiO₂ of 61. C reactive protein was 147 mg/L (normal value <3 mg/L) and lactate dehydrogenase was in the normal range (table 1).

Table 1.

Laboratory analysis

Variable	Reference range	Results before abatacept therapy was started	Results during the last ER admission	Results during ICU admission	Outcome results
WCC, ×109/L	4.5–11.50	4.01	8.54	13.34	21.21
Neutrophils, %	50–70	47	76	69	76
Lymphocytes, %	18–42	46	9	13	12
Monocytes, %	2–11	4	15	12	4
Band cells, %	0–5	0	0	0	6
Metamyelocytes, %	0	0	0	0	2
Eosinophils, %	1–3	2	0	0	0
Basophils, %	0–2	2	1	0	0
pH	7.370–7.440		7.494	7.390	6.960
PaCO2*, mm Hg	31.0–45.0		29.00	40.3	75.4
PaO2*, mm Hg	80.0–95.0		61.10	97.8	93.6
HCO3−, mmol/L	21.0–30.0		21.8	23.9	16.6
BE, mmol/L	−2.0 to +2.0		−0.20	−1.0	−16.6
sO2, %	95.0–100.0		93.6	97.4	91.3
FiO2, %	21		100	40	35
PaO2:FiO2	>250		61	244	208
CRP, mg/L	<5	3	147.0	151.6	112.4
Procalcitonin, ng/mL	0.5–2.0			1.01	0.53
Lactate dehydrogenase, UI/L	0–480			365	531
RF, UI/mL	0–20	133
Anti-CCP, UI/mL	25–50 (uncertain) >50 (positive)	>3200

*Arterial blood gases.

anti-CCP, anticyclic citrullinated peptide; BE, base excess; CRP, C reactive protein; ER, emergency room; FiO₂, oxygen inspired fraction; HCO₃⁻, bicarbonate ion; ICU, intensive care unit; PaCO₂, partial pressure of carbon dioxygen; PaO₂, partial pressure of oxygen; RF, rheumatoid factor; sO₂, calculated oxygen saturation; WCC, white cell count.

Differential diagnosis

Although the ICU team suspected an infectious aetiology during differential diagnosis, a comprehensive study including blood cultures, bronchial culture, urine culture and a respiratory panel via PCR (which included assessment for influenza A viruses, influenza B, human metapneumovirus, Mycoplasma pneumoniae, rhinovirus/human enterovirus, human parainfluenza viruses, respiratory syncytial virus, adenovirus, Bordetella pertussis, Chlamydia pneumoniae, and coronavirus HKU-1, NL-63, OC-43 and 229E) was performed, the results of which were negative. Bacilloscopies and tuberculosis cultures were also negative. A rapid HIV test was negative, as well as Pneumocystis jirovecii staining.

Complementary studies included flexible bronchoscopy with bronchoalveolar lavage (BALF) and microbiology studies, which showed negative results for infectious aetiology or diffuse alveolar haemorrhage. Differential cytology of BALF showed a discrete increase in polymorphonuclear cells (macrophages 87%, neutrophils 8%, lymphocytes 3%, eosinophils and basophils <1%) and the absence of leucocytes, and no bacteria were observed in BALF using Gram staining.

In the ICU, the third chest CT scan (figure 1E, F) ruled out pulmonary embolism and showed ground-glass opacities with diffuse distribution associated with interlobular septal thickening and multiple bullae and foci of alveolar filling in both lower lobes.

Finally, a flow cytometry analysis revealed reduced percentages of all subpopulations of CD3+ cells (figure 2), so pulmonary involvement due to rheumatological disease was less probable.

Figure 2.

Peripheral blood mononuclear cells isolated from the patient by flow cytometry. (A) CD56 and CD3 identified four populations: CD3+CD56− and total T-cells, CD3+CD56− as Natural Killer T cells, CD56+CD3− as natural killer cells, and CD56hi as CD56hi natural killer cells. (B) CD3 and CD20 identified B cells as CD3−CD20+. (C) From CD3+CD56− gate, CD8 and CD4 were used to identify cytotoxic T-cells and helper T-cells, respectively. (D) High expression of CD25 within the CD4+ population was used to identify regulatory T- cells.

Treatment

After ICU admission, vancomycin (2 g intravenously two times per day) and piperacillin/tazobactam (4.5 g intravenously three times a day) were empirically prescribed and administered, and then the patient was connected to mechanical ventilation (MV) to support respiratory function in addition to systemic corticoids (hydrocortisone 100 mg three times a day). After 1 week of MV, the patient’s arterial blood gas levels improved and her FiO₂ decreased to 35% (FiO₂ 0.35, PaO₂:FiO₂ 251).

Outcome and follow-up

The ARDS gradually worsened for 2 months until the patient died. In the last 2 days in the ICU, the patient showed respiratory acidosis and multisystemic failure. A second respiratory panel via PCR was positive for C. pneumoniae, and a chest CT showed severe fibrotic pattern in both lungs. After a meeting with the family, the patient was determined not to be a candidate for MV support. The patient passed away due to respiratory failure.

Discussion

Numerous studies over the past few years have supported the critical role of T-cell effector responses in RA; thus, therapeutic options have been further expanded to include strategies aimed at inhibiting cytokine production and costimulatory T-cell activation.⁸ However, biological therapy has been associated with an increased risk of ILD, such as hypersensitivity pneumonitis, and an increased risk of pulmonary exacerbation in patients with chronic comorbidities, such as asthma or ILD,⁹ and this association was reported for both abatacept and other biological therapies, such as antitumour necrosis factor-α (anti-TNF-α).^{9 10}

Abatacept is a fully humanised soluble fusion protein that comprised cytotoxic T-lymphocyte-associated antigen-4 (CTLA-4) and the Fc region of human IgG1. CTLA-4 binds to CD80/CD86 on antigen-presenting cells and inhibits T-cell costimulation and downstream inflammatory mediators, which are commonly increased in RA.¹¹ Current recommendations from the American College of Rheumatology suggest abatacept treatment for patients with previous pulmonary infection.¹² In our case, the medical history showed that the patient has suffered from RA for 30 years, in addition to radiological findings of chronic lung damage probably due to methotrexate and high clinical activity, which were relevant to the decision to initiate abatacept as the first biological therapy, despite the possibility of using alternatives such as anti-TNF-α therapies. However, acute lung damage has not been fully associated with abatacept in patients with RA.¹³ Wada et al⁶ reported for the first time in 2012 a case report showing the deterioration of interstitial pneumonia after abatacept administration in a patient with RA. They suggested that abatacept might be the cause of the deterioration of interstitial pneumonia, but they did not dismiss other possibilities, such as tacrolimus discontinuation, RA deterioration or viral infection. In 2016, Doğu et al⁷ published the second report of respiratory failure after abatacept administration in patients with RA, and the clinical findings were consistent with drug-induced acute respiratory failure.

The mechanisms involved in this process are unknown, as are the conditions or characteristics presented by patients with RA with abatacept-induced respiratory failure. In 2018, a national multicentre study assessed the efficacy of abatacept in patients with ILD associated with RA. They reported that 11 of 63 (17%) patients discontinued the drug due to adverse events. One of them died 2 months after abatacept withdrawal due to a flare-up of ILD. They concluded that abatacept appears to be effective for RA-associated ILD, as a good number of patients showed clinical improvement and remained stable after treatment.⁵ A study with 57 Chilean patients with RA did not report lung deterioration in any of these patients; therefore, our case was not influenced by race or environment.¹⁴

To elucidate the cause of death, it is possible to consider that respiratory failure may have been due to several causes, such as the activation of RA itself or bacterial-viral-fungus infections that were not detected by blood cultures, BALF cultures or respiratory panels via PCR. Another parameter to consider is C reactive protein, which is not a good infection marker in this case because in active RA, disease activity is usually elevated; however, the procalcitonin levels were normal according to many measures.

In conclusion, despite the safety and efficacy of abatacept compared with that of other therapeutic agents, as shown by several studies of patients with RA, it may be possible that a few patients may develop rapid acute respiratory failure after treatment. The main reason for such acute lung disease in patients with RA treated with this biological agent remains unknown. We think that the increased use of an increasing number of new immunomodulatory drugs in patients with RA makes it very important for clinicians to keep in mind such adverse reactions, which could be more frequently seen in the future.

Learning points.

• Acute respiratory failure may be due to abatacept use, and this association has been reported in a few case reports.

• A comprehensive evaluation for infections, rheumatoid arthritis flare and secondary causes of acute pneumonia is important to rule out differential diagnosis; however, previous chronic lung disease can be secondary to long-term use of methotrexate.

• Physicians should take into consideration these unusual but potentially severe adverse events when prescribing immunotherapy.