543P - A multicenter retrospective study to investigate risk factors for immune checkpoint inhibitor-induced pneumonitis in non-small cell lung cancer patients with comorbid interstitial pneumonia

Background

Interstitial pneumonia (IP) is the common and poor prognostic comorbidities in patients with non-small cell lung cancer (NSCLC) and is a known risk factor for immune checkpoint inhibitor (ICI)-induced pneumonitis. The aim of this study is to assess the incidence, severity and risk factors for pneumonitis induced by immune checkpoint inhibitor (ICI) monotherapy in NSCLC patients with idiopathic IP.

Methods

This is a multicenter, retrospective study in patients with (1) advanced or recurrent NSCLC, (2) comorbid idiopathic IP, and (3) receiving ICI monotherapy as second-line or later therapy. All CTs at baseline and at the onset of pneumonitis were centrally adjudicated by a board-certified radiologist. We performed logistic regression analysis to identify risk factors for pneumonitis using clinical characteristics, laboratory data and radiological findings as variables.

Results

Sixty-five patients were analyzed, with a median age of 71 years, and 98.5% had a smoking history. Sixty percent received nivolumab, 21.5% received atezolizumab and 18.5% received pembrolizumab. The median cycle of ICI was 4 cycles. Radiological patterns of IP were UIP (11%), probable UIP (15%), indeterminate for UIP (51%), alternative (23%), respectively, with 23% having honeycomb lung and a median %FVC of 95%. The incidence of pneumonitis was 23.1% for all grade and 13.9% for ≥grade3. We could not identify any significant risk factors for ICI-induced pneumonitis in the multivariate logistic regression analysis. There was no difference in the incidence and severity of ICI-induced pneumonitis when comparing patients with and without honeycomb lung, or with FVC ≥80% and <80%.

Conclusions

There was no difference in the risk of ICI-induced pneumonitis according to specific findings, including the radiological pattern of comorbid IP, the presence of honeycomb lung, and pulmonary function tests. ICIs should be carefully administrated in NSCLC with any types of comorbid IP.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

S. Ikeda: Financial Interests, Personal, Advisory Board: AstraZeneca, Chugai, Ono, Daiichi Sankyo; Financial Interests, Personal, Invited Speaker: AstraZeneca, Chugai, Ono, Bristol Myers Squibb, Taiho, Eli Lilly, Pfizer, Boehringer Ingelheim, Takeda; Financial Interests, Personal, Research Grant: AstraZeneca, Chugai. J. Sakakibara-Konishi: Financial Interests, Personal, Research Grant: eli Lilly. K. Yokoo: Financial Interests, Personal, Invited Speaker: AstraZeneca, Chugai, Eli Lilly, Kyowa Kirin, MerckBiopharma, Novartis, Terumo, Takeda, Nippon Kayaku, Bristol Myers Squibb. T. Iwasawa: Financial Interests, Personal, Advisory Board: CANON Medical Systems, Ziosoft I; Financial Interests, Personal, Invited Speaker: Boehringer Ingelheim. T. Misumi: Financial Interests, Personal, Invited Speaker: Miyarisan. T. Ogura: Financial Interests, Personal, Invited Speaker: Japanese Boehringer Ingelheim, Shionogi; Financial Interests, Personal, Advisory Board: BMS, Japanese Boehringer Ingelheim, Taiho, Shionogi. All other authors have declared no conflicts of interest.