The road less travelled: thoracic ultrasound, advanced imaging and artificial intelligence for early diagnosis of non-expandable lung in malignant pleural effusion.

Abstract

Malignant pleural effusion (MPE) affects up to 15% of cancer patients, with nearly 30% of symptomatic cases developing non-expandable lung (NEL), a condition characterised by the lung's failure to fully re-expand post-drainage, thereby impeding proper pleural apposition and leading to several adverse outcomes. Inadequate diagnostic certainty leads to prolonged hospitalisation, repeated invasive procedures, drainage complications, high pleurodesis failure rates, increased healthcare costs and diminished patient quality of life. Conventional diagnostic methods, predominantly based on post-procedural chest radiography and computed tomography, frequently delay accurate diagnosis, underscoring the need for noninvasive pre-procedural techniques. Emerging evidence supports thoracic ultrasound, particularly the application of M-mode during breath-hold, as a promising modality for early NEL detection by identifying the absent sinusoidal sign and reduced lung movement. Experimental approaches, including speckle tracking imaging, two-dimensional shear wave elastography and quantitative ultrasound assessments via the lung/liver echogenicity ratio, also show potential, albeit with limitations that warrant further validation. Integration of artificial intelligence into multimodal imaging workflows may enhance diagnostic precision and predictive modelling, ultimately facilitating personalised therapeutic strategies and transforming the management of NEL in MPE. These innovations promise to reduce invasive diagnostics and healthcare costs while improving patient outcomes and quality of life in MPE-associated NEL.