Interstitial pneumonia with autoimmune features: the new consensus-based definition for this cohort of patients should be broadened

To the Editor:

We appreciate the research statement recently published by Fischer et al. [1] proposing new terminology: “interstitial pneumonia with autoimmune features” (IPAF) to characterise the heterogeneous group of patients with idiopathic interstitial pneumonia (IIP) who have a clinical flavour of underlying connective tissue disease (CTD) but do not meet the current American College of Rheumatology criteria for CTD. Indeed, a variety of terms and criteria have been used in studies to describe such patients, including “autoimmune featured interstitial lung disease”, “undifferentiated connective tissue disease associated interstitial lung disease”, and “lung-dominant connective tissue disease” [2–5]. Thus, the need for a consensus regarding the terminology and better characterisation of the clinical characteristics of this group of patients is evident. The proposed terminology, however, is not yet validated and does not include a provision for patients with clinical characteristics of CTD who do not meet American College of Rheumatology criteria for CTD or IPAF criteria, but are distinct from patients with IIP alone. To meet the proposed criteria for IPAF, a patient must have features of IIP on high-resolution computed tomography (HRCT) images and/or surgical lung biopsy. Alternative aetiologies of IIPs must be excluded by thorough clinical evaluation and the patient must fail to meet criteria for a specific CTD. The patient must then have at least one characteristic from at least two out of three domains: clinical, serological and morphological. We wish to share our concerns regarding the new IPAF classification.

Our primary concern with the proposed IPAF criteria is that usual interstitial pneumonia (UIP) is excluded from the morphological domain. This is problematic as UIP is a common pattern of IIP among patients with positive serology for CTD who do not meet criteria for specific CTD and do not meet clinical criteria for IPAF and, thus, are currently labelled as IPF. While this is in keeping with the 2011 criteria for the definition of IPF and may be the reason for excluding UIP in the proposed definition for IPAF, recent studies in patients with lung dominant CTD (defined as patients with positive serologies for CTD in the absence of extrathoracic features of a definite CTD) documented that the UIP pattern was the most common IIP pattern of disease, occurring in 57% of the 44 patients in this cohort [4, 6, 7]. Vij et al. [2] documented that 62% of patients with “autoimmune featured ILD” characterised by IIP, positive serologies and one or more symptoms suggestive of CTD but not meeting American College of Rheumatology criteria had a UIP pattern on HRCT. Studies of patients with IIP and autoimmune characteristics have produced different estimates of survival when compared to patients with chronic fibrosing lung disease, which may be related to differences in the prevalence of the UIP pattern between these cohorts [2, 8, 9]. UIP has been associated with poor survival among patients with CTD in some studies, but survival may still be better than that among patients with IPF [7, 10, 11]. These studies highlight the importance of the UIP pattern among these patients, and we strongly believe that UIP should be included in the morphological domain of the IPAF criteria.

We also have concerns regarding the IPAF clinical domain, which is limited to seven criteria. Based on the current IPAF definition, a patient with an abnormal CTD serology, UIP and oesophageal dysmotility and/or myalgias would not fulfil the proposed criteria for patients with IPAF. We feel that such patients differ from those with IIP that is limited to the lungs and may indeed have subclinical autoimmune disease and/or later manifest into a specific CTD. Perhaps these patients would be better acknowledged as having “autoimmune ILD”; autoimmune, simply based on positive CTD serology. Indeed, we appreciate that the proposed IPAF classification is intended for research purposes and would presumably lead to the design of clinical trials enrolling patients with IPAF as proposed by Fischer et al. [1]. We are concerned that by excluding UIP from the proposed morphological domain and excluding patients with other clinical features suggestive of subclinical autoimmune diseases, we would be prematurely disregarding such patients, labelling them as IIP (different from IPAF and by default including them into the category of IIP) and IPF if they have UIP. Such patients are arguably in greatest need of research efforts for effective treatments, yet may be excluded and left in the open, meeting neither criteria for IPAF nor IPF if the proposed IPAF criteria are widely applied to future studies/clinical trials.

While we applaud the efforts of the authors in proposing criteria for IPAF and acknowledge that this is a step in the right direction, we strongly believe the criteria should be reconsidered to encompass patients with autoimmune ILD who do not meet the proposed criteria for IPAF, particularly those with UIP, as well as those with positive serologies and subtle/less specific symptoms of CTD.