In Situ Assessment of Porcine Osteochondral Repair Tissue in the Visible–Near Infrared Spectral Region

Abstract

Standard assessment of cartilage repair progression by visual arthroscopy can be subjective and may result in suboptimal evaluation. Visible–near infrared (Vis-NIR) fiber optic spectroscopy of joint tissues, including articular cartilage and subchondral bone, provides an objective approach for quantitative assessment of tissue composition. Here, we applied this technique in the 350–2,500 nm spectral region to identify spectral markers of osteochondral tissue during repair with the overarching goal of developing a new approach to monitor repair of cartilage defects in vivo. Full thickness chondral defects were created in Yucatan minipigs using a 5-mm biopsy punch, and microfracture (MFx) was performed as a standard technique to facilitate repair. Tissues were evaluated at 1 month (in adult pigs) and 3 months (in juvenile pigs) post-surgery by spectroscopy and histology. After euthanasia, Vis-NIR spectra were collected in situ from the defect region. Additional spectroscopy experiments were carried out in vitro to aid in spectral interpretation. Osteochondral tissues were dissected from the joint and evaluated using the conventional International Cartilage Repair Society (ICRS) II histological scoring system, which showed lower scores for the 1-month than the 3-month repair tissues. In the visible spectral region, hemoglobin absorbances at 540 and 570 nm were significantly higher in spectra from 1-month repair tissue than 3-month repair tissue, indicating a reduction of blood in the more mature repair tissue. In the NIR region, we observed qualitative differences between the two groups in spectra taken from the defect, but differences did not reach significance. Furthermore, spectral data also indicated that the hydrated environment of the joint tissue may interfere with evaluation of tissue water absorbances in the NIR region. Together, these data provide support for further investigation of the visible spectral region for assessment of longitudinal repair of cartilage defects, which would enable assessment during routine arthroscopy, particularly in a hydrated environment.

Visible and Near-Infrared Spectroscopy Enables Differentiation of Normal and Early Osteoarthritic Human Knee Joint Articular Cartilage

“…While there is currently no comparative study between ICRS and OARSI grades of human tissue, Vis-NIRS offers the advantage of objective and repeatable assessment of early-stage cartilage changes during arthroscopic repair surgery. In clinical settings, absolute values in the peak-specific analysis are likely to yield sub-optimal predictions due to the overlapping and non-specific nature of NIR peaks of cartilage constituents in the Vis-NIR range [ 21 , 34 ]. Models (e.g., SVM) trained on preprocessed (broadband) spectra of healthy and deteriorated tissue are more likely to evaluate tissue integrity effectively.…”

“…For preprocessing, Savitzky–Golay filter was used for data smoothing and spectral normalization was performed using localized standard normal variate (LSNV). Due to spectral saturation (noise) around the OH peak (1400–1520 nm), which is indicative of cartilage water content [ 21 ], each spectrum was trimmed to the spectral regions of 400–1400 nm and 1520–1850 nm. While the region (1000–1399 nm) has been linked to cartilage collagen content [ 1 , 31 ], the region (1600–1800 nm) is associated with the matrix proteoglycan content [ 1 , 31 ].…”