Prostate cancer treatment is shifting from radical to focal therapy. Instant tumor visualization on a microscopic level is crucial for clinical application of focal therapy. Optical coherence tomography (OCT) produces instant tissue visualization on a µm scale and the attenuation of OCT signal as a measure of tissue organization. The objective is to correlate qualitative and quantitative OCT analysis with histopathology.
Twenty prostates were analyzed by needle based OCT after radical prostatectomy. For precise correlation, whole mount histology slides were cut through the OCT trajectory. OCT images were classified in eight histological categories. Two reviewers independently performed assessment of the OCT images into these categories. Quantitative attenuation coefficient was used to discriminate stroma and malignant tissue. Sensitivity and specificity for detection of malignancy on OCT was calculated.
Visual analyses showed that OCT can reliably differentiate between fat, cystic and regular atrophy and benign glands. Differentiation of benign stroma and inflammation and also malignancy Gleason 3 and 4 is more difficult. Sensitivity and specificity for detection of malignancy on OCT were calculated at 77% and 75%. Quantitative analysis by means of the attenuation coefficient for differentiation between stroma and malignancy showed no significant difference (4.39 mm-1 vs. 5.31 mm-1).
Precise correlation of histology and OCT is possible and helps us to understand what we see and measure on OCT. Visual malignancy detection shows reasonable sensitivity and specificity. Our future studies focus on improving discrimination of malignancy with OCT for example by combining an extra imaging modality.
The objective of this study is to demonstrate the feasibility of needle-based optical coherence tomography (OCT) and functional analysis of OCT data along the full pullback trajectory of the OCT measurement in the prostate, correlated with pathology. OCT images were recorded using a commercially available C7-XR™ OCT Intravascular Imaging System interfaced to a C7 Dragonfly™ intravascular 0.9-mm-diameter imaging probe. A computer program was constructed for automated image attenuation analysis. First, calibration of the OCT system for both the point spread function and the system roll-off was achieved by measurement of the OCT signal attenuation from an extremely weakly scattering medium (Intralipid® 0.0005 volume%). Second, the data were arranged in 31 radial wedges (pie slices) per circular segments consisting of 16 A-scans per wedge and 5 axial B-scans, resulting in an average A-scan per wedge. Third, the decay of the OCT signal is analyzed over 50 pixels (500 μm) in depth, starting from the first found maximum data point. Fourth, for visualization, the data were grouped with a corresponding color representing a specific μoct range according to their attenuation coefficient. Finally, the analyses were compared to histopathology. To ensure that each single use sterile imaging probe is comparable to the measurements of the other imaging probes, the probe-to-probe variations were analyzed by measuring attenuation coefficients of 0.03, 6.5, 11.4, 17, and 22.7 volume% Intralipid®. Experiments were repeated five times per probe for four probes. Inter- and intraprobe variation in the measured attenuation of Intralipid samples with scattering properties similar to that of the prostate was <8% of the mean values. Mean attenuation coefficients in the prostate were 3.8 mm−1 for parts of the tissue that were classified as benign (SD: 0.8 mm−1, minimum: 2.2 mm−1, maximum: 8.9 mm−1) and 4.1 mm−1 for parts of tissue that were classified as malignant (SD: 1.2 mm−1, minimum: 2.5 mm−1, maximum: 9.0 mm−1). In benign areas, the tissue looked homogeneous, whereas in malignant areas, small glandular structures were seen. However, not all areas in which a high attenuation coefficient became apparent corresponded to areas of prostate cancer. This paper describes the first in-tissue needle-based OCT imaging and three-dimensional optical attenuation analysis of prostate tissue that indicates a correlation with pathology. Fully automated attenuation coefficient analysis was performed at 1300 nm over the full pullback. Correlation with pathology was achieved by coregistration of three-dimensional (3-D) OCT attenuation maps with 3-D pathology of the prostate. This may contribute to the current challenge of prostate imaging and the rising interest in focal therapy for reduction of side effects occurring with current therapies.