Background: It is unknown whether patients with interstitial lung disease (ILD) and only some features of autoimmunity have a different natural history from those with a defined connective tissue disease (CTD-ILD). The classification criteria for “ILD with autoimmune features” (IPAF) may not be able to characterize all these patients, especially those with a usual interstitial pneumonia (UIP) pattern [1].

Objectives: To determine clinical characteristics and predictive factors for progression in a cohort of ILD patients with features of autoimmunity, through the application of classification criteria for IPAF and specific CTD, whenever possible.

Methods: We retrospectively selected a cohort of consecutive patients with ILD as onset manifestation and features of autoimmunity (at least 1 autoantibody and/or 1 clinical sign/symptom), evaluated by our multidisciplinary unit from March 2009 to March 2020. All the final diagnoses were revised according to the latest CTD and IPAF criteria. Patients were followed up for 33 (16.5-69.5) months.

Results: Of the 101 patients enrolled (67.4±10.9 yrs, F/M ratio 65/36), 53 (52.5%) and 37 (36.6%) respectively satisfied the CTD and IPAF criteria. Eleven patients (10.9%) did not satisfy IPAF criteria because of only 1 item (clinical or serologic) within the IPAF domains and a UIP pattern; we defined this group as “autoimmune” UIP (AI-UIP). All the 8 patients initially classified as undifferentiated CTD had sufficient IPAF criteria. Among the IPAF patients (68.2±10.1 years, F/M ratio 20/17), the most common findings were: Nonspecific interstitial pneumonia pattern (56.8%), antinuclear antibodies positivity (43.2%) and arthritis (24.3%). The combination of a positive morphologic and serologic domain was the most common to reach the diagnosis (48.6%). Some IPAF patients had features not included in IPAF criteria, such as non-anti-synthetase myositis-specific antibodies (21.6%), objective sicca syndrome (13.5%) and anti-myeloperoxidase antibodies (2.7%). Over a median of 17 months, 2 IPAF patients (5.4%) developed a definite UIP pattern, while 4 (10.8%) a specific CTD. Comparing the IPAF, CTD-ILD and AI-UIP groups, no statistically significant differences were found in the mean age, sex distribution, smoking habits and mean duration of the disease. However, IPAF patients had a significantly higher prevalence of arterial hypertension and left-sided heart failure and a lower predominance of UIP pattern as expected (10.8% vs. 32.1% vs. 100%, p <0.01). Although no differences were found at the diagnosis, at 1 year the proportion of IPAF patients with radiological progression of the fibrosis and/or functional deterioration (defined by a decline in FVC of ≥ 10% and/or DLCO of ≥ 15% predicted) was lower to that of CTD-ILD and AI-UIP (17.1% vs. 31.4% vs. 63.6%, p 0.01). Fewer IPAF patients needed oxygen support (8.6% vs. 31.4% vs. 36.4, p 0.02). Considering the overall 101 patients, having an IPAF and a UIP pattern respectively predicted a slower (OR: 0.37, p 0.04) and a faster (OR: 3.56, p 0.01) ILD progression at the multivariate analysis.

Conclusion: In our cohort, IPAF criteria were useful to identify a subset of patients with a slower ILD progression and a possible evolution to CTD (10-15% of cases) [2]. These criteria do not characterize all the patients with a UIP pattern and limited features of autoimmunity, which seem to have a worse prognosis, independently from the final diagnosis. Further studies are needed to clarify if the prognosis of AI-UIP is different from that of idiopathic pulmonary fibrosis.

REFERENCES:

[1]Graney, et al. Ann Am Thorac Soc 2019;16(5):525-33.

[2]Sebastiani, et al. Biomedicines 2021,9,17.