Full-field optical coherence tomography for the analysis of fresh unstained human lobectomy specimens

Full-field optical coherence tomography (FFOCT) is a real-time imaging technique that generates high-resolution three-dimensional tomographic images from unprocessed and unstained tissues. Lack of tissue processing and associated artifacts, along with the ability to generate large-field images quickly, warrants its exploration as an alternative diagnostic tool.

One section each from the tumor and from adjacent non-neoplastic tissue was collected from 13 human lobectomy specimens. They were imaged fresh with FFOCT and then submitted for routine histopathology. Two blinded pathologists independently rendered diagnoses based on FFOCT images.

Normal lung architecture (alveoli, bronchi, pleura and blood vessels) was readily identified with FFOCT. Using FFOCT images alone, the study pathologists were able to correctly identify all tumor specimens and in many cases, the histological subtype of tumor (e.g., adenocarcinomas with various patterns). However, benign diagnosis was provided with high confidence in roughly half the tumor-free specimens (others were diagnosed as equivocal or false positive). Further analysis of these images revealed two major confounding features: (a) Extensive lung collapse and (b) presence of smoker's macrophages. On a closer inspection, however, the smoker's macrophages could often be identified as distinct from tumor cells based on their relative location in the alveoli, size and presence of anthracosis. We posit that greater pathologist experience, complemented with morphometric analysis and color-coding of image components, may help minimize the contribution of these confounders in the future.

Our study provides evidence for the potential utility of FFOCT in identifying and differentiating lung tumors from non-neoplastic lung tissue. We foresee its potential as an adjunct to intra-surgical frozen section analysis for margin assessment, especially in limited lung resections.