Checkpoint Inhibitor Pneumonitis: A Pulmonologist’s Perspective

Get Permission

Oncologists have become accustomed to seeing pneumonitis associated with immune checkpoint inhibitor treatment in non–small cell lung cancer (NSCLC), requiring intervention by pulmonologists in the management of severe cases. At CHEST 2019, the annual meeting of the American College of Chest Physicians, specialists shared their insights in the treatment of this population, noting that diagnosis and treatment can be difficult.

Karthik S. Suresh, MD

Karthik S. Suresh, MD

Catherine R. Sears, MD

Catherine R. Sears, MD

Karthik S. Suresh, MD, Assistant Professor of Medicine at Johns Hopkins University School of Medicine, Baltimore, discussed the differential diagnosis and workup of patients suspected of having pneumonitis related to immune checkpoint inhibitor therapy, and Catherine R. Sears, MD, Assistant Professor of Medicine at the Indiana University School of Medicine, Indianapolis, described its management. Dr. Sears was lead author of the 2019 official American Thoracic Society research statement that identified knowledge gaps and research priorities in this area.1

Nuances of Diagnosis

The reported incidence of checkpoint inhibitor–related pneumonitis is 3% to 5% in randomized controlled trials, but it is higher—10% to 20%—in retrospective studies that have involved multidisciplinary adjudication. At Johns Hopkins, 19% of 205 patients in a data set for NSCLC treated with immune checkpoint inhibitors were diagnosed with pneumonitis after adjudication.

“Hopkins is a referral center, so there may be referral bias, but regardless, I think the consensus is that as these drugs have moved out of clinical trials, the incidence of pneumonitis is likely to be higher than previously thought—probably somewhere between 10% and 15%,” said Dr. Suresh.

Clinically, pneumonitis presents with dyspnea, cough, and exertional desaturation and is primarily a diagnosis of exclusion. Briefly, chest imaging reveals new infiltrates, and new infection must be ruled out. Time to onset varies greatly—from within a few hours of first infusion, to a day or two later, to more than a year after checkpoint inhibitor initiation. Some studies have suggested that the higher grades usually present early, but this observation has not been validated.

Presentation and Differential Diagnosis

“The differential diagnosis of [immune checkpoint inhibitor] pneumonitis includes other immune-related adverse events that mimic it, and common conditions that are comorbid in the NSCLC population, irrespective of [checkpoint inhibitor] therapy,” Dr. Suresh said. In his presentation, he divided the differential diagnosis into these categories and outlined scenarios for which specific tests, such as bronchoscopy, are indicated.

For the patient with nonspecific symptoms, such as cough, dyspnea, fatigue, and hypoxia, the following factors can raise flags for certain conditions:

  • Hypercarbia: immune checkpoint inhibitor–associated myasthenia gravis
  • Superventricular tachycardia, shock, volume overload: immune checkpoint inhibitor–associated myocarditis
  • Risk factors for tuberculosis: immune checkpoint inhibitor–induced reactivation of tuberculosis (especially in patients from endemic areas)
  • Recent cytotoxic chemotherapy: diffuse alveolar hemorrhage, opportunistic infections
  • Recent high-dose steroid taper:Pneumocystis carinii pneumonia, Nocardia
  • Increase in size of tumor: pseudoprogression.

“These are some of the [immune checkpoint inhibitor]-related considerations that we keep in mind when seeing these patients,” he said. “We also don’t want to lose track of the fact that a lot of these patients have underlying obstructive lung disease and other diseases that complicate their presentation…. There was a recent report of a fatal case of asthma after initiation of an [immune checkpoint inhibitor]. We need to ask patients about their underlying obstructive lung disease and whether that has worsened [after checkpoint inhibitor initiation]. This can be unrelated to the pneumonitis.”

Other conditions are also in the differential diagnosis: heart failure, aspiration pneumonia, and thromboembolism. “Among the patients referred to us who we work up for pneumonitis, 30% to 40% will have one of these other entities,” he said.

Considerations by Indication

“Another way to approach this issue is to understand why these patients were put on [immune checkpoint inhibitors] in the first place,” Dr. Suresh continued. “That can sometimes frame our aggressiveness toward bronchoscopy and how we arrange things on our differential diagnosis. For instance, the patient receiving an [immune checkpoint inhibitor] as first-line therapy, for all intents and purposes, was otherwise immunocompetent before treatment. You can contrast this, for example, with the bone marrow transplant patient who received [checkpoint inhibitors] after multiple rounds of chemotherapy. These patient populations are very different.”

Dr. Suresh outlined the specific immune checkpoint inhibitor indications and considerations this way:

  • Immune checkpoint inhibitor as first-line regimen: Assume the patient is otherwise immunocompetent, with a lower suspicion for opportunistic infection, diffuse alveolar hemorrhage, or other drug toxicity.
  • Concurrent checkpoint inhibitor plus chemotherapy (paclitaxel/platinum): Consider drug-induced pneumonitis that is not caused by an immune checkpoint inhibitor in addition to checkpoint inhibitor pneumonitis.
  • Heavy pretreatment (ie, transplant patients, salvage therapy): A higher suspicion for opportunistic infection or diffuse alveolar hemorrhage if cytopenic; with bone marrow transplant, consider graft-vs-host disease and idiopathic pneumonia syndrome.
  • Treatment immediately after curative surgery and radiotherapy (ie, stage III NSCLC): Consider restriction from video-assisted thoracoscopic surgery and radiation pneumonitis.

Workup Recommendations

Workup of patients with suspected pneumonitis includes sputum cultures and nasopharyngeal swabs. It is especially important to look for evidence of viral infection in the winter months. Other tests include exertional oximetry, computed tomography of the chest, and pulmonary function tests.

“Possibly useful” clinical factors include lymphocytosis in bronchoalveolar lavage fluid; negative cultures and bronchoalveolar lavage neutrophils < 50%; lack of eosinophils and recent exposure to paclitaxel; and lack of evidence of heart failure or history of uncontrolled occupational lung disease. There are also factors for which utility is unclear: time of symptom onset, radiographic appearance, history of other immune-related adverse events, and tumor response to checkpoint inhibitor therapy.

“There is no one biomarker or serum test that has any reasonable sensitivity or specificity for pneumonitis. The most useful thing, actually, is multidisciplinary adjudication of this disease,” he said.

Management of High-Grade Pneumonitis

“Pneumonitis often looks like infection, and this is where we [pulmonologists] can be helpful to oncologists,” Dr. Sears said.

Patients with grade 3 pneumonitis definitely need a workup for an infectious cause, including blood and sputum cultures and swabs and often bronchoscopy with bronchoalveolar lavage. Transbronchial biopsy is not recommended unless steroids fail to induce a prompt response in the patient, radiologic findings are atypical, or there is a concern about tumor progression. If infection is likely, empiric antibiotics are usually started. In the absence of infection or improvement within 24 to 48 hours, steroid-sparing agents should be initiated.


  • In patients undergoing immunotherapy for NSCLC who have nonspecific lung symptoms, clinicians should suspect checkpoint inhibitor–related pneumonitis, which is more common than clinical trials suggest.
  • The differential diagnosis includes other immune-related adverse events that mimic checkpoint inhibitor pneumonitis, and common conditions that are comorbid in the NSCLC population, irrespective of checkpoint inhibitor therapy.
  • For grade 2 and higher, withdraw the checkpoint inhibitor and quickly start immunosuppression, escalating treatment as necessary.

Steroid-sparing agents include infliximab, intravenous immunoglobulin, mycophenolate mofetil, tocilizumab, and cyclophosphamide—the use of which has largely been extrapolated from their proven utility in other inflammatory lung diseases and immune-related toxicities. “There are often reasons to use one vs another in an individual patient,” she said.

Rechallenge with immune checkpoint inhibitors is being done “more and more,” especially in patients who respond well to the first such drug and whose grade 2 or 3 pneumonitis resolved promptly, she added. Approximately one-third of rechallenged patients will develop pneumonitis again, but they are likely to respond to steroids if they responded initially. It is important to wean steroids (10 mg prednisone equivalent per day or less) before checkpoint inhibitor retreatment, as they may negatively impact treatment outcomes. If managed promptly and appropriately, approximately 85% of all comers with checkpoint inhibitor pneumonitis will recover.

Infections and Pneumonitis

“One should have a high suspicion for infection in these patients, especially opportunistic infections,” Dr. Sears said. Infections (bacterial, fungal, viral, parasitic) account for 1% to 2% of deaths in patients on immune checkpoint inhibitors. Infections described with immune checkpoint inhibitors include tuberculosis, listeriosis, invasive aspergillosis, cytomegalovirus, and pneumocystis. However, typical and atypical infections are more common in patients receiving immunosuppressive treatment for an immune-related adverse event.

“When patients are not responding the way you would expect, or if there are radiologic signs or clinical symptoms that make you think this is more than pneumonitis, you really should look for infections,” Dr. Sears emphasized.

As with steroids, patients should remain on antibiotics only as long as necessary, as the use of antibiotics has been negatively associated with the clinical activity of immune checkpoint inhibitors. 

DISCLOSURE: Dr. Suresh reported no conflicts of interest. Dr. Sears has served as a scientific consultant for Biodesix and bioAffinity.


1. Sears CR, Peikert T, Possick JD, et al: Knowledge gaps and research priorities in immune checkpoint inhibitor-related pneumonitis: An official American Thoracic Society research statement. Am J Respir Crit Care Med 200:e31-e43, 2019.