Characteristics and Prognosis of Patients with Interstitial Pneumonia with Autoimmune Features (IPAF) Compared to Idiopathic Interstitial Pneumonia (IIP) and Connective Tissue Diseases Associated Interstitial Lung Diseases (CTD-ILD): A Systematic Review and Meta-Analysis

Abstract

Background: Interstitial pneumonia with autoimmune features (IPAF) is a term used to classify patients presenting with idiopathic interstitial pneumonia (IIP) alongside features suggestive of an underlying connective tissue disease (CTD), yet insufficient to establish a definitive CTD diagnosis. Since its introduction, the clinical significance of IPAF has remained uncertain, particularly whether it constitutes a distinct clinical entity with distinct clinical course and prognosis from those of IIP or CTD-associated interstitial lung disease (CTD-ILD). Prior to IPAF classification criteria, such patients were labelled with various nomenclatures such as undifferentiated CTD-ILD (UCTD-ILD), lung dominant CTD, or autoimmune featured ILD using overlapping classification criteria.

Aims: This systematic review aims to evaluate whether the baseline characteristics and clinical outcomes of patients with IPAF differ from those with IIP, including idiopathic pulmonary fibrosis (IPF), and CTD-ILD. Additionally, it seeks to identify prognostic factors associated with the outcomes of IPAF.

Methods: We conducted a comprehensive literature search using four electronic databases—Medline, Embase, Scopus, and Web of Science— in addition to the citations of the 2015 IPAF statement using Scopus. We included cross-sectional, case-control, and cohort studies that compared the baseline characteristics and outcomes of IPAF patients with those of IIP, IPF, or CTD-ILD patients, or that reported prognostic factors associated with IPAF. We also included studies that compared pre-IPAF entities (e.g., UCTD-ILD and related conditions) to other ILD groups. Primary outcomes of interest were all-cause mortality, disease progression, acute ILD exacerbation, and development of definitive CTD. Two independent reviewers conducted abstract screening, full-text screening, and data extraction in duplicate using predefined forms for each step. We assessed the risk of bias in each study using the Quality in Prognostic Studies (QUIPS) tool and rated the certainty of evidence using the GRADE approach, adapted for prognosis and prognostic factor studies. When possible, we conducted meta-analyses using a random-effects model to compare outcomes between IPAF and other ILD subtypes and to evaluate the effect of specific prognostic factors on IPAF outcomes.

Results: We included 43 studies derived from 41 distinct cohorts. Of these, 36 were retrospective, six were prospective, and one employed a mixed (retrospective and prospective) design. Thirty-two studies focused on IPAF, while 11 examined UCTD-ILD and related entities. According to the QUIPS tool, the majority of included studies were judged to carry a high risk of bias in one or more domains. Compared to CTD-ILD, IPAF patients were generally older, with a pooled median difference of 4.9 years. Conversely, they were younger than patients with IIP and IPF, with median differences of 4.5 and 5.6 years, respectively. Unlike IIP and IPF, most IPAF patients were female, reflecting a demographic profile more similar to that of CTD-ILD. IPAF patients were more likely to be current or former smokers than CTD-ILD patients, but less likely than those with IIP/IPF. However, baseline lung function did not consistently differ between IPAF and other ILD groups. IPAF was associated with significantly higher mortality compared to CTD-ILD (adjusted HR 1.74; 95% CI: 1.33–2.27; I² = 0%; low certainty), and lower mortality compared to IIP (unadjusted HR 0.53; 95% CI: 0.32–0.88; I² = 41%; very low certainty). There was no significant difference in mortality between IPAF and IPF (adjusted HR 0.85; 95% CI: 0.42–1.71; I² = 75%; very low certainty). Rates of disease progression were comparable between IPAF and CTD-ILD (RR 0.79; 95% CI: 0.44–1.43; I² = 44%; very low certainty), with a trend toward lower progression compared to IIP (RR 0.45; 95% CI: 0.20–1.00; I² = 34%; very low certainty). Similarly, rates of acute exacerbation did not differ significantly between IPAF and CTD-ILD (RR 0.92; 95% CI: 0.54–1.57; I² = 66%; very low certainty), or between IPAF and IIP (RR 0.83; 95% CI: 0.41–1.67; I² = 79%; very low certainty). Ten cohorts reported the incidence of definitive CTD among IPAF patients. The pooled incidence of de-novo CTD was 14% (95% CI: 11–18%; I² = 55%). Compared to IIP, IPAF was associated with a six-fold increased risk of developing CTD during follow-up (RR 6.00; 95% CI: 3.72–9.68; I² = 0%; low certainty). In contrast, UCTD-ILD and related entities did not consistently demonstrate differences in baseline characteristics or outcomes when compared to other ILD subtypes. In the multivariate analysis, independent predictors of all-cause mortality in IPAF were age (HR 1.05, 95% CI: 1.03-1.05, I2=27%, low certainty), male sex (HR 2.75, 95% CI: 1.34-5.63, I2=48%, low certainty), baseline ppDLCO (HR 0.68, 95% CI: 0.57-0.80, I² = 28%; very low certainty), and pathological or radiological UIP (HR 2.78, 95% CI: 1.59-4.85, I2=45%, very low certainty). Ever-smoking and baseline ppFVC were significant predictors of mortality only in the univariate analysis with unadjusted HR of 1.32 (95% CI: 1.04-1.67, I2=0%) and 0.91 (95% CI: 0.87-0.95, I2=0%) respectively. Similarly, longer six-minute walking distance and fulfillment of the clinical IPAF domain were also associated with reduced mortality in univariate analyses. Steroid therapy showed no significant association with mortality in either multivariate or univariate analyses.

Conclusion: Our systematic review supports the classification of IPAF as an intermediate phenotype within the ILD spectrum, exhibiting clinical features and outcomes that lie between those of CTD-ILD and IIP/IPF. It also underscores the need for well-designed prospective studies to refine IPAF classification criteria, delineate its clinical trajectories and phenotypes, and validate prognostic predictors that can more effectively guide future research and inform clinical decision-making.