Objective: To describe the clinical,radiological and pathological features of non-specific interstitial pneumonia (NSIP), and to evaluate the multidisciplinary approach to the diagnosis of NSIP.
Methods: The clinical data, lung CT scans and pathologic slides of patients with a diagnosis or a probable diagnosis of NSIP on discharge were re-evaluated by a panel of respiratory physicians, radiologists and pathologists. A pathological diagnosis and a clinical diagnosis were made by the panel according to the 2002 classification by American Thoracic Society/European Respiratory Society.
Results: In the 7 cases diagnosed by surgical lung biopsy, diffuse bronchiectasis was found to be the predominant feature in 1 case, and organizing pneumonia in another case. The remaining 5 cases met the criteria of NSIP. CT guided percutaneous lung biopsy was performed in 2 probable cases, for which the specimens were inadequate for a definite diagnosis, but because of a lack of specific findings, and with the consistent clinical and radiological features, the diagnosis of probable NSIP was maintained. In the patients with NSIP or probable NSIP, ground glass opacities were the predominant CT features, without the typical appearance of idiopathic pulmonary fibrosis. Of the 2 probable cases, 1 died from disease deterioration 20 days after lung biopsy, and 1 was found to have multiple myositis/dermatomyositis 2 years after the initial diagnosis. Among the 5 definite cases, one was confirmed to have multiple myositis/dermatomyositis, but no underlying causes were found for the other 4 cases. One patient died from progressive fibrosis and respiratory failure 3 years after the initial diagnosis.
Conclusions: The radiological manifestations of NSIP were not diagnostic, and therefore surgical lung biopsy was the procedure of choice in making a definite diagnosis. The clinical and pathological diagnosis of NSIP needs a multidisciplinary approach by respiratory physicians, radiologists and pathologists. The search for a possible underlying cause is an important part of the dynamic diagnostic process.