ABSTRACT
Immune-related adverse events (irAEs) have become increasingly prevalent with immune checkpoint inhibitor (ICI) cancer treatment. We present a 79-year-old man with metastatic renal cell carcinoma who developed shortness of breath and hypercapnic respiratory insufficiency after his first cycle of nivolumab and ipilimumab. Laboratory data showed elevated creatinine kinase, troponins, and transaminases. Computed tomography of the chest demonstrated bilateral lower lobe atelectasis. Heart catheterization and endomyocardial biopsy were unremarkable. Electromyogram (EMG) and nerve conduction studies (NCS) of the limb muscles revealed mild diffuse myopathy, normal sensory nerve conductions, and low-amplitude motor responses. Subsequent diaphragmatic EMG and NCS demonstrated severe myopathy. ICI-mediated myopathy predominantly affecting diaphragmatic muscles was diagnosed. Treatment included intravenous methylprednisolone, infliximab, abatacept, rituximab, and plasmapheresis. He underwent tracheostomy placement on hospital day 11 due to minimal improvement. He was discharged to a long-term acute care hospital, but unfortunately, he died less than 1 month later due to recurrent infections. irAEs can affect any organ system, but diaphragmatic dysfunction is uncommon. Use of diaphragmatic EMG, NCS, ultrasound study, or biopsy can support the diagnosis. Treatment includes systemic steroids, plasmapheresis, immunosuppressive medications, respiratory support, and cessation of causative medications. ICI-related diaphragmatic dysfunction should be suspected in those patients at risk with hypoxia, hypercapnia, or prolonged invasive or noninvasive ventilation without a distinct etiology. This case report exemplifies the importance of multidisciplinary workup and management of respiratory symptoms and insufficiency to identify and ameliorate irAEs. Diaphragmatic involvement can be associated with significant morbidity and mortality despite early aggressive multimodal therapy.
INTRODUCTION
With increased use of immune checkpoint inhibitors (ICIs) for cancer, a heightened suspicion for immune-related adverse events (irAEs) is paramount because these may affect multiple organ systems. Involvement of the diaphragm is rare. We report a case of shortness of breath and hypercapnic respiratory insufficiency after one cycle of dual ICI therapy in a patient with renal cell carcinoma (RCC), and his presentation included elevation of muscle, cardiac, and liver enzymes. Through a series of testing and multidisciplinary approach, he was diagnosed with an irAE manifesting as severe diaphragmatic myopathy.
CASE REPORT
A 79-year-old man with metastatic clear cell RCC, coronary artery disease with history of cardiac stents, hypertension, and hyperlipidemia presented with dyspnea. Metastatic RCC was diagnosed by right renal and subcarinal lymph node biopsies. Formal staging elucidated stage IV RCC with mediastinal lymphadenopathy, thoracic spine pathologic fracture, right acetabular lytic lesions, and without intracranial metastases on magnetic resonance imaging of the brain. Three weeks before presentation, he received his first cycle of nivolumab and ipilimumab, and his peripheral oxygen saturation (SpO2) on room air was 95%. A few days after this infusion he developed an erythematous pruritic rash on the back that improved with topical triamcinolone cream.
Four days before presentation, he developed mild dyspnea and was started on prednisone 20 mg by his outpatient oncologist for suspicion of ICI-related pneumonitis. His respiratory symptoms persisted, and he proceeded to an outside emergency department where his SpO2 on room air was recorded as 56%. He had mild confusion, but his Glasgow Coma Scale (GCS) was 14 (Eyes 3, Verbal 5, Motor 6). A nonrebreather mask at 15 L/min increased the SpO2 to 98%, but this was immediately followed with an abrupt decrease in GCS to 7 (Eyes 2, Verbal 1, Motor 4). An arterial blood gas showed pH 7.11, pCO2 118.5 mm Hg, pO2 139 mm Hg. He was transferred to our institution on bilevel positive airway pressure (BPAP) 15/10 cmH2O with an improved pH of 7.3 and pCO2 of 68 mm Hg. Initial chest radiography revealed low lung volumes (Fig. 1A).
Radiographic imaging. Portable chest radiograph (A) on admission reveals low lung volumes (arrows) and blunting of the left costophrenic angle. Computed tomography of the chest without contrast demonstrates small bilateral pleural effusions (B, arrows) and opacities in the lung bases that may be atelectasis versus consolidations (C, arrows).
Radiographic imaging. Portable chest radiograph (A) on admission reveals low lung volumes (arrows) and blunting of the left costophrenic angle. Computed tomography of the chest without contrast demonstrates small bilateral pleural effusions (B, arrows) and opacities in the lung bases that may be atelectasis versus consolidations (C, arrows).
Physical exam revealed an oriented patient, intact cranial nerves II-XII, normal deep tendon reflexes, and 5/5 strength bilaterally in all extremity musculature except 4/5 was found in bilateral pelvic girdle musculature. No facial or neck muscle weakness was noted. Labs included a creatine kinase (CK) of 3178 units/L, CK-myocardial band 107 ng/mL, troponin T 685 ng/L, white blood cell 9.4 × 103/μL, hemoglobin 10.6 g/dL, N-terminal-pro-B-type natriuretic peptide 918 pg/mL, alanine aminotransferase 183 U/L, aspartate aminotransferase 188 U/L, total bilirubin 0.4 mg/dL, and aldolase 53.8 U/L. Autoimmune myositis panel, anti-3-hydroxy-3-methylglutarylr-CoA reductase, anti-muscle-specific kinase, anti-striational, anti-low density lipoprotein receptor-related protein 4 antibodies and anti-titin antibodies were negative. Computed tomography of the chest without contrast (Figs. 1B, C) demonstrated bilateral lower lobe atelectasis versus consolidation. Electrocardiogram showed normal sinus rhythm without significant ST changes, and transthoracic echocardiogram revealed normal left and right ventricular size and function without elevation of right ventricular systolic pressure. In addition to diuresis, early multimodal treatment was given with intravenous methylprednisolone, and plasmapheresis, followed by infliximab, rituximab, and abatacept (Fig. 2). CK trended downward within 48 hours after admission, but troponin remained elevated.
Biomarker trend. Trend of both creatinine kinase and troponin T demonstrated with early multi-modality therapy. Five treatments for plasma exchange were administered. Dosages of other medications were as follows: infliximab (5 mg/kg), rituximab (1000 mg), abatacept (500 mg). Solumedrol was administered initially at 1 g daily, then decreased to 1 mg/kg every 12 hours, then gradually tapered.
Biomarker trend. Trend of both creatinine kinase and troponin T demonstrated with early multi-modality therapy. Five treatments for plasma exchange were administered. Dosages of other medications were as follows: infliximab (5 mg/kg), rituximab (1000 mg), abatacept (500 mg). Solumedrol was administered initially at 1 g daily, then decreased to 1 mg/kg every 12 hours, then gradually tapered.
Six days after treatment, right and left heart catheterization showed a pulmonary capillary wedge pressure 18 mm Hg, mean pulmonary artery pressure 31 mm Hg, cardiac output 7.69 L/min, and cardiac index 3.49 L/min/m2 by Fick’s principle, and no significant coronary artery atherosclerosis. Endomyocardial biopsy was negative for inflammatory infiltration. Despite BPAP, respiratory acidosis persisted. Electromyogram (EMG) and nerve conduction studies (NCSs) of the limb muscles revealed mild diffuse myopathy, normal sensory nerve conductions, and low-amplitude motor responses. Subsequent diaphragmatic needle EMG and NCS showed severe myopathy. Phrenic stimulation motor responses were absent and could suggest unexcitable diaphragm muscles. Ultrasound thickness of the diaphragm remained at 1 mm without respiratory inspiratory thickening. ICI-mediated myopathy predominantly affecting diaphragmatic muscles and myocarditis were diagnosed. Despite biomarker trend and aggressive therapies, he had minimal improvement, so tracheostomy placement was performed for persistent hypercapnia and respiratory insufficiency. He was discharged to a long-term acute care hospital with 2 to 4 hours of tracheostomy collar daily and nocturnal pressure support ventilation. He had repeat hospitalizations due to infection (respiratory, urinary tract infection), and less than 1 month after discharge, he died.
DISCUSSION
Our case highlights a challenging presentation of irAE with concern for cardiac, muscular, rheumatologic, and neurologic abnormalities. ICIs may adversely affect multiple organ systems simultaneously or sequentially in the form of irAEs. Immune checkpoint signaling is a natural downregulatory process that the immune system uses to prevent autoimmunity and promote self-tolerance. Inhibiting programmed death 1, programmed death ligand 1, and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) immune checkpoints triggers an abundant immune response that is effective in cancer therapy, but also occasionally results in irAEs. The pathophysiology of these events is theorized as a cross interaction of T-cells with similar antigens of both malignant and normal cells.[1]
Neuromuscular involvement can occur with various mono/polyneuropathies, Guillain-Barré syndrome, neuromuscular junction disorders, cranial neuropathies, myositis, and polymyopathies.[2] Occasionally irAEs may present with bulbar symptoms and autoimmune myasthenia gravis and other overlap syndromes must be ruled out. Both phrenic neuropathies and isolated diaphragmatic myositis have only been described in a few case reports/series and reviews.[1–8] In a systematic review of 319 cases of ICI-associated neuromuscular disorders, three (0.9%) were diagnosed with phrenic neuropathy.[2] Once sepsis is excluded, suspicion of diaphragmatic involvement of irAEs should be considered, especially when patients present with dyspnea, imaging consistent with atelectasis, low lung volumes, or hemi-diaphragm elevation, and laboratory evidence of hypercapnia (Fig. 3). Labs vary but may include elevated CK in diaphragmatic myositis. Cross-reactivity with skeletal muscle injury elevates troponin T, whereas troponin I is more reflective of myocardial injury. Further evaluation of diaphragmatic dysfunction may include pulmonary function testing, and findings will be suggestive of neuromuscular weakness with decreased maximal inspiratory and expiratory pressures and restrictive ventilatory defect without diminished diffusing capacity of the lung for carbon monoxide.[3] Reduction of vital capacity when supine compared with when upright suggests neuromuscular diaphragmatic weakness.[9] Diaphragmatic ultrasound may reveal an absence of thickening at inspiration when compared with end-expiration.[9] Phrenic neuropathy may be diagnosed via NCS with low amplitudes of compound muscle action potentials. Diaphragmatic EMG results with fibrillation potentials and positive sharp waves are consistent with myopathy.[5] Although clinically similar, diaphragmatic weakness from critical illness or invasive (or noninvasive) ventilation would show normal resting muscle thickness >1.4 mm on diaphragmatic ultrasound with normal phrenic NCS.[7,10] Pathologic immunohistochemistry in autopsy reports of the diaphragm have shown mixed CD8+ and CD4+ lymphocytic infiltration of the diaphragm with type II fiber-specific atrophy indicating myositis.[6,7]
Symptoms and findings for diaphragmatic dysfunction. Unilateral and bilateral diaphragmatic paralysis may present differently in terms of symptoms and degree of change in testing.[3,9]
Symptoms and findings for diaphragmatic dysfunction. Unilateral and bilateral diaphragmatic paralysis may present differently in terms of symptoms and degree of change in testing.[3,9]
Treatment for these conditions is not standardized, and withdrawal of the offending ICI may be considered for moderate to severe adverse events. First-line treatment for most irAEs includes systemic steroids, followed by intravenous immunoglobin, plasmapheresis, and a variety of biologic agents. These biologics include anti–tumor necrosis factor alpha inhibitors, anti–B-cell CD20 monoclonal antibodies, anti-interleukin (IL)-6 receptor therapies, anti–IL-4 alpha therapies, anti–IL-17A therapies, anti–IL-23 alpha antibodies, anti–IL-12 and -23 therapies, CTLA-4 agonists, anti-CD52, anti-thymocyte globulin, mycophenolate mofetil, calcineurin inhibitors, cyclophosphamide, methotrexate, azathioprine, sulfasalazine, hydroxychloroquine, and Janus kinase inhibitors.[11] There is no current literature for diaphragmatic pacing in patients with phrenic neuropathy from irAEs. Despite these treatment options, survival is significantly decreased, with severe respiratory failure necessitating mechanical ventilation.[8] Irreversible myopathic or neuropathic inflammatory destruction before immunosuppressant therapy may be a plausible explanation for poor prognosis in severe ICI-related diaphragmatic dysfunction.
Our management of this case highlights the importance of early recognition, evaluation, and treatment. Second, the decision for early tracheostomy in this patient allowed for a quicker and safer disposition to a long-term acute care facility for prolonged ventilator weaning. Unfortunately, despite several lines of therapies the patient eventually developed recurrent infections and died.
CONCLUSION
irAEs may affect any organ system, but diaphragmatic dysfunction is a rare manifestation. Meticulous evaluation for concomitant and/or sequential irAEs is also paramount. Early recognition is crucial to prevent morbidity and mortality. Despite current treatment options, mortality remains high for diaphragmatic dysfunction in this patient population. Ultimately, multidisciplinary evaluation and management is vital for suspected irAE-related diaphragmatic dysfunction, and future research in facilitating prompt diagnosis, identifying preventive interventions, and intervening promptly is imperative.
References
Author notes
Walder JR, Faiz SA, Tummala S, et al. Diaphragmatic myopathy associated with dual immune checkpoint inhibitors in renal cell carcinoma.
Source of Support: This research is supported in part by the National Institutes of Health through MD Anderson’s Cancer Center Support Grant (CA016672).