POS0879 LUNG ULTRASOUND FOR INTERSTITIAL LUNG DISEASE DETECTION IN A COHORT OF SYSTEMIC SCLEROSIS PATIENTS: ROLE OF B-LINE AND PLEURAL LINE MODIFICATIONS

Abstract

BackgroundLung ultrasound (LUS) is a technique that showed a high diagnostic accuracy for interstitial lung disease (ILD) detection in systemic sclerosis (SSc) patients and currently in progress of standardization. Traditionally, B-lines represented the finding of ILD, with the ≥10 total cut-off reported by Tardella et al. resulting to be closely related to moderate ILD detected on high resolution chest computed tomography (HRCT). Recently, Fairchild et al. proposed novel LUS criteria for the evaluation of pleural line, disclosing high accuracy and reproducibility.ObjectivesTo compare B lines cut-off with novel pleural line criteria and the respectively associated variables.MethodsWe enrolled 55 consecutive patients affected by SSc according to ACR/EULAR 2013 criteria who underwent respiratory functional tests (RFTs) during 2021, excluding smokers and those with arterial pulmonary hypertension. Twenty-four of them carried out a HRCT during a ± 6-months’ time. In the same day of RFTs, two certified blinded operators performed LUS for each patient applying the 14-areas technique proposed by Gutierrez et al., looking for the presence of total ≥10 cumulative B lines and the fulfilling of Fairchild’s criteria for pleural line. A 3-13 MHz operating linear probe was used. Clinical-demographic data and ongoing therapies were collected.ResultsAmong 55 total SSc patients, the agreement between the two operators for Fairchild’s criteria was almost perfect (Cohen’s kappa (k) =0.81) and substantial for ≥10 cumulative B-lines count (k=0.74). Fairchild’s criteria showed a higher diagnostic accuracy compared with ILD detected on HRCT, with an overall specificity (SP) and a positive predicted value (PPV) of 100% (Table 1). A negative correlation emerged between total lung capacity values (TLC%) and both B lines cut-off [first operator (IO): p 0.04, r -0.27; second operator (IIO): p 0.042, r-0.28] and pleural line criteria (IO: p 0.009, r -0.35; IIO: p 0.08, r – 0.36), but only the latter negatively correlated also with forced vital capacity values (FVC%) (IO: p 0.04, r – 0.27; IIO p 0.03, r -0.28). The ≥10 total B lines amount correlated positively with concurrent mycophenolate therapy (IO: p 0.09, r 0.28; IIO: p 0.005, r 0.37) and negatively with anti-centromere antibodies (IO: p 0.002, r -0.3; IIO p 0.009, r -0.34). The presence of digital ulcers showed a positive correlation with pleural line criteria (IO: p 0.03, r 0.29; IIO: p 0.005, r 0.37), with a significant association on multivariate analysis (IO: p 0.03, IIO: p 0.01).Table 1.Overall sensitivity, specificity, positive predictive value, negative predictive value of LUS compared to ILD detected on HRCT. C.I. 95% confidence interval.SE (C.I.)SP (C.I.)PPV (C.I.)NPV (C.I.)Fairchild’s criteria fulfilling0.91 (0.76 - 0,97)1 (0.78 – 1.000)1 (0.88 -1.00)0.82 (0,59 – 0.94)≥10 cumulative B-lines count0.73 (0.56 – 0.85)0.8 (0.55 – 0.93)0.89 (0.72 – 0.96)0.57 (0.36 – 0.75)ConclusionWe confirmed the feasibility and reliability of Fairchild’s recently proposed pleural line LUS criteria, that showed a higher diagnostic accuracy versus ≥10 cumulative B-lines count for ILD detected on HRCT, presenting SP and PPV values of 100% in SSc. Furthermore, these LUS features seem to differently associate with other aspects of the disease such as autoantibody specificity and vascular lesions, thus deserving future deeper evaluations. For the first time, we found that Fairchild’s criteria were associated with a clinical variable such as digital ulcers. Our results highlight the relevance of pleural line evaluation for ILD detection in SSc on LUS and its possible role towards a standardization of this diagnostic technique.