Evaluation of Human Kidney for Transplantation Using Polarization-Sensitive Optical Coherence Tomography

Introduction: Kidney transplant is the preferred treatment for patients who have end-stage renal diseases. Currently, there is a worldwide massive shortage of donor kidneys due to the increasing demand for renal transplantation. Kidney donor profile index (KDPI) is a traditional evaluation criterion of deceased donor kidneys to characterize donor-related risks and predict graft outcomes. With the rapid increase of demand of deceased donor kidneys, the expanded criteria donor (ECD) with less limitations on the donor age (> 60 years old) and death causes (hypertension, cerebrovascular, or terminal serum creatinine) is introduced to extend the donor pool. Pre-transplant donor biopsy (PTDB) is a standardized assessment to estimate the nephron mass of pre-transplant donor kidneys to determine the single or dual kidney transplantation or being discarded. Banff histopathological consensus criteria (BHCC) is created to be the interpretation of pre-transplantation kidney biopsies, which can be utilized to correlate histopathological findings with graft outcomes and assess interobserver variability of histopathologic scores. At present, BHCC is based on the histological staining of one piece of kidney tissue sliced from a specific location of the deceased donor kidney. It is doubted whether one piece of sliced tissue for biopsy from a deceased donor kidney can be used to evaluate the entire kidney quality based on present histopathological criteria. Additionally, the distribution characterization of renal tissues over the entire donor kidney is unable to be demonstrated in present clinical biopsies due to multiple slicing of renal tissues at different locations.

Materials and

Methods: 20 deceased donor kidneys sourced from LifeShare of Oklahoma were utilized in our experiment for imaging and histopathological examinations. All donor kidneys were preserved using hypothermic machine perfusion (HMP) to maintain physiological integrity and were transported to the laboratory for imaging within 48 hours post-donation. A PS-OCT system incorporating detector units for concurrent acquisition of orthogonal polarizations and high-speed imaging capabilities was employed to conduct comprehensive scans of multiple locations within the entire donor kidney. The imaging protocol encompassed a field-of-view (FOV) measuring 4.3 × 1.5 × 2.0 mm³ (X, Y, Z) with a resolution of 1500 × 270 × 787 pixels in the X, Y, and Z axes, resulting in a sampling resolution of 2.8 × 5.5 × 2.6 μm³ during kidney imaging sessions. Following PS-OCT imaging, all labeled regions spanning the entirety of the kidney underwent slicing and subsequent histological processing to facilitate a comparative analysis with corresponding PS-OCT findings. We utilized a Pix2pix model featuring adaptive loss functions, specifically tailored for image-to-image translation tasks such as segmentation and denoising. total of 269 cross-sectional images extracted from each 3D PS-OCT dataset corresponding to distinct anatomical regions underwent preprocessing to facilitate automated recognition of tubule microstructures. Five thresholds of 0.5, 0.6, 0.7, 0.8, and 0.9 were used to extract the cumulative fibrosis signals (0.0 ~ 1.0) from each raw PS-OCT DOPU image. Five featured fibrosis extractions with the DOPU value of 0.5~1.0, 0.6~1.0, 0.7~1.0, 0.8~1.0, and 0.9~1.0 were collected for each DOPU image.

Results, Conclusions, and Discussions: Our analysis revealed heterogeneous histopathological distribution on each surface, and the distribution characterization varied across surfaces within the same donor kidney. We used the current clinical MAPI standard to evaluate donor kidney viability and observed that the heterogeneous distribution of kidney tissue quality affects the comprehensive assessment of viability. Viability assessment also varies across different regions of the kidney. Donor kidneys with lower average MAPI scores (M = 1) exhibited relatively lower heterogeneity distribution (MAPI scores ranging from 0 to 3). However, kidneys with higher average MAPI scores showed a more heterogeneous distribution in histopathological examinations of kidney tissues (MAPI scores ranging from 3 to 12 and 6 to 15). To quantify the density distribution of tubules across the entire kidney, we calculated the spatial density, representing the percentage of tubules over kidney tissues. We demonstrated the spatial density distribution for donor kidneys, and revealed variations among different kidneys. Specifically, Kidneys exhibited various spatial densities of tubules. Additionally, we mapped the tubular diameter distribution characterization over the entire kidneys, based on calculating the central diameter of all tubules within each 3D PS-OCT dataset. Notably, Kidneys displayed a heterogeneous distribution of tubular diameter. We utilized histopathological scores for fibrosis from the Banff Histopathological Consensus Criteria (BHCC) under the MAPI standard to correlate with PS-OCT DOPU fibrosis values. Linear fitting was applied to correlate the percentage of fibrosis in DOPU images with the histopathological scores. Optimal fitting occurred at a threshold of 0.7 for extracting fibrosis from DOPU images, as indicated by the highest R-Square value. We found a high agreement in fibrosis distribution across the entire kidney surface between histopathological scores and extracted fibrosis in DOPU images. Additionally, a dice coefficient of 0.8 between the overlap of fibrosis regions from histopathological scores and extracted fibrosis in DOPU images indicated a high correlation.

Acknowledgements (Optional): This work was supported by grants from the Stephenson Cancer Center (3P30CA225520), Institutional Research Grant number 134128‐IRG‐19‐142‐01 from the American Cancer Society, National Science Foundation (OIA-2132161, 2238648), National Institute of Health (R01DK133717, R01CA269897), Department of Defense (W81XWH-22-1-0415), Oklahoma Shared Clinical and Translational Resources (NIGMS U54GM104938), Oklahoma Center for the Advancement of Science and Technology (HR23-071, HR 23-069), the medical imaging COBRE (P20 GM135009), and the Midwest Biomedical Accelerator Consortium (MBArC), an NIH Research Evaluation and Commercialization Hub (REACH). Histology service provided by the Tissue Pathology Shared Resource was supported in part by the National Institute of General Medical Sciences COBRE Grant P20GM103639 and National Cancer Institute Grant P30CA225520 of the National Institutes of Health.