Introduction and objectives The peripheral lung contains a range of structural elements (small airways, blood vessels and lymphatics) together with infiltrating inflammatory cells. These components exist together in complicated spatial arrangements. Lung disease is frequently accompanied by changes in both lung architecture and the number and distribution of inflammatory cells. Light microscopy has been the conventional technique of choice in understanding these changes and relationships but provides only 2-D representations of a complex 3-D network.

We selected to use micro computed tomography (µ-CT) to image structural elements in the peripheral lung. We aimed to reconstruct the 3-D architecture by combining the µ-CT data with immunohistochemistry (IHC) to positively identify the principal structural elements and inflammatory cells.

Methods Human lung tissue was fixed in formalin, embedded in paraffin wax and subjected to µ-CT scanning. The tissue was then sectioned and immunostained for pancytokeratin (airways), collagen IV (blood vessels), D2–40 (lymphatic vessels) and CD68 (macrophages). The resulting images were used to guide the segmentation of the 3-D µ-CT image stack. IHC, using neurofilament antibodies, was also used on multiple lung samples to attempt to identify nerve fibres in the parenchymal tissue.

Results The main structural elements of the lung periphery could be identified, segmented out and their 3-D architecture examined. Macrophages were found throughout the tissue in large quantities and were most concentrated around the blood vessels and lymphatics. Lymphatic vessels were especially dense in the pleural region and elsewhere were intertwined with blood vessels. Despite being readily identifiable in bronchial samples, nerve fibres were not identified using IHC in the parenchyma.

Conclusions Combining µ-CT and IHC provides a robust method to positively identify important structural elements of the peripheral lung and to localise inflammatory cells in 3-D, thus allowing a detailed review of their spatial relationships. Alternative methodologies may however be advantageous regarding identifying parenchymal nerve fibres for reconstruction. µ-CT and IHC together create a highly accurate 3-D reconstruction but this method remains time consuming; advances in automation and improved tools are required to fully exploit the research potential.