Micronodular Pattern of Organizing Pneumonia or Hypersensitivity Pneumonia Induced by an Immune Checkpoint Inhibitor?

To the Editor:

I read with great interest the report entitled “Durvalumab-induced Organizing Pneumonia with a Diffuse Micronodular Pattern in a Patient with Lung Cancer” (1). In that report by Yamasaki and colleagues, diffuse centrilobular, primarily ground-glass micronodules on chest computed tomographic (CT) imaging was diagnosed as organizing pneumonia (OP) by transbronchial biopsy (TBB) with bronchoscopy. The authors proposed the importance of considering OP as a differential diagnosis in addition to hypersensitivity pneumonitis and infectious bronchiolitis in patients presenting with diffuse centrilobular micronodules during immune checkpoint inhibitor (ICI) therapy. That report is important because immunotherapy has become a standard of care in oncology, and drug-induced pneumonia is more frequent and important. However, we have some concerns regarding their diagnosis and conclusion.

OP, cryptogenic or secondary, is a clinicopathological entity characterized by granulation tissue plugs in the lumen of small airways, alveolar ducts, and alveoli. An unusual radiological presentation corresponding to diffuse micronodular pattern mimicking miliary lung infiltration has been reported, and the prevalence of this micronodular pattern of OP (MNOP) in radiological series is estimated to range from 10% to 24% (2, 3). The typical MNOP on CT imaging is widespread lung micronodules without diffuse ground-glass opacities, airspace consolidation, or cavitation.

CT patterns of drug-induced pneumonia were classified based on the American Thoracic Society and European Respiratory Society international multidisciplinary classification of interstitial pneumonias as acute interstitial pneumonia/diffuse alveolar damage–like pattern, hypersensitivity pneumonia (HP)-like pattern, OP-like pattern, nonspecific interstitial pneumonia–like pattern, or others (4). It has been reported that ICIs induce the OP-like pattern in 19–65% and the HP-like pattern in 16–22% among drug-induced pneumonias (5). The HP-like pattern includes a centrilobular diffuse micronodular pattern and bronchiolitis-like appearance.

The radiographic HP-like pattern and MNOP are difficult to differentiate. In addition, it is occasionally hard to make a histologically precise diagnosis for OP or HP by TBB because of the small size of the specimen. In large series of OP, the histological diagnosis was obtained on TBB specimens from 31% to 67% of cases. Importantly, most cases of MNOP were diagnosed by surgical lung biopsy (2). Regarding HP, the meta-analysis of 11 studies on TBB revealed a diagnostic yield of 64.3% (6). My concern on the report by Yamasaki and colleagues (1) is that the diffuse centrilobular micronodules were of primarily ground-glass appearance and not well defined on chest CT imaging, suggesting rather an HP-like pattern. Another concern is that the diagnosis for this unusual entity of OP was made by TBB alone. If the case showed HP-type radiographic pattern, it is not rare in ICI therapy.

Finally, I agree with the authors’ message. During ICI therapy, considering drug-induced pneumonias, including OP or HP, as a differential diagnosis is important because they usually have a favorable response to steroid treatment. In addition to radiological examination, performing examination of BAL fluid and TBB is recommended whenever possible as a means of ruling out infections and neoplastic lesions.