Stain-Independent Deep Learning-Based Analysis of Digital Kidney Histopathology

Challenges in standardizing preimplantation kidney biopsy assessments and the potential of AI-Driven solutions

PURPOSE OF REVIEW

This review explores the variability in preimplantation kidney biopsy processing methods, emphasizing their impact on histological interpretation and allocation decisions driven by biopsy findings. With the increasing use of artificial intelligence (AI) in digital pathology, it is timely to evaluate whether these advancements can overcome current challenges and improve organ allocation amidst a growing organ shortage.

RECENT FINDINGS

Significant inconsistencies exist in biopsy methodologies, including core versus wedge sampling, frozen versus paraffin-embedded processing, and variability in pathologist expertise. These differences complicate study comparisons and limit the reproducibility of histological assessments. Emerging AI-driven tools and digital pathology show potential for standardizing assessments, enhancing reproducibility, and reducing dependence on expert pathologists. However, few studies have validated their clinical utility or demonstrated their predictive performance for long-term outcomes.

SUMMARY

Novel AI-driven tools hold promise for improving the standardization and accuracy of preimplantation kidney biopsy assessments. However, their clinical application remains limited due to a lack of proven associations with posttransplant outcomes and insufficient evaluation of predictive performance metrics. Future research should prioritize longitudinal studies using large-scale datasets, rigorous validation, and comprehensive assessments of predictive performance for both short- and long-term outcomes to fully establish their clinical utility.

Deep learning-based histopathological assessment of tubulo-interstitial injury in chronic kidney diseases

BACKGROUND

Chronic kidney disease (CKD) causes progressive and irreversible damage to the kidneys. Renal biopsies are essential for diagnosing the etiology and prognosis of CKD, while accurate quantification of tubulo-interstitial injuries from whole slide images (WSIs) of renal biopsy specimens is challenging with visual inspection alone.

METHODS

We develop a deep learning-based method named DLRS to quantify interstitial fibrosis and inflammatory cell infiltration as tubulo-interstitial injury scores, from WSIs of renal biopsy specimens. DLRS segments WSIs into non-tissue areas, glomeruli, tubules, interstitium, and arteries, and detects interstitial nuclei. It then quantifies these tubulo-interstitial injury scores using the segmented tissues and detected nuclei.

RESULTS

Applied to WSIs from 71 Japanese CKD patients with diabetic nephropathy or benign nephrosclerosis, DLRS-derived scores show concordance with nephrologists' evaluations. Notably, the DLRS-derived fibrosis score has a higher correlation with the estimated glomerular filtration rate (eGFR) at biopsy than scores from nephrologists' evaluations. Validated on WSIs from 28 Japanese tubulointerstitial nephritis patients and 49 European-ancestry patients with nephrosclerosis, DLRS-derived scores show a significant correlation with eGFR. In an expanded analysis of 238 Japanese CKD patients, including 167 from another hospital, deviations in eGFR from expected values based on DLRS-derived scores correlate with annual eGFR decline after biopsy. Inclusion of these deviations and DLRS-derived fibrosis scores improve predictions of the annual eGFR decline.

CONCLUSIONS

DLRS-derived tubulo-interstitial injury scores are concordant with nephrologists' evaluations and correlated with eGFR across different populations and institutions. The effectiveness of DLRS-derived scores for predicting annual eGFR decline highlights the potential of DLRS as a predictor of renal prognosis.

PDGF-D Is Dispensable for the Development and Progression of Murine Alport Syndrome

Alport syndrome is an inherited kidney disease, which can lead to glomerulosclerosis and fibrosis, as well as end-stage kidney disease in children and adults. Platelet-derived growth factor-D (PDGF-D) mediates glomerulosclerosis and interstitial fibrosis in various models of kidney disease, prompting investigation of its role in a murine model of Alport syndrome. In vitro, PDGF-D induced proliferation and profibrotic activation of conditionally immortalized human parietal epithelial cells. In Col4a3-/- mice, a model of Alport syndrome, PDGF-D mRNA and protein were significantly up-regulated compared with non-diseased wild-type mice. To analyze the therapeutic potential of PDGF-D inhibition, Col4a3-/- mice were treated with a PDGF-D neutralizing antibody. Surprisingly, PDGF-D antibody treatment had no effect on renal function, glomerulosclerosis, fibrosis, or other indices of kidney injury compared with control treatment with unspecific IgG. To characterize the role of PDGF-D in disease development, Col4a3-/- mice with a constitutive genetic deletion of Pdgfd were generated and analyzed. No difference in pathologic features or kidney function was observed in Col4a3-/-Pdgfd-/- mice compared with Col4a3-/-Pdgfd+/+ littermates, confirming the antibody treatment data. Mechanistically, lack of proteolytic PDGF-D activation in Col4a3-/- mice might explain the lack of effects in vivo. In conclusion, despite its established role in kidney fibrosis, PDGF-D, without further activation, does not mediate the development and progression of Alport syndrome in mice.

Reproducibility and prognostic ability of chronicity parameters in kidney biopsy - Comprehensive evaluation comparing microscopy and artificial intelligence in digital pathology

INTRODUCTION

Semi-quantitative scoring of various parameters in renal biopsy is accepted as an important tool to assess disease activity and prognostication. There are concerns on the impact of interobserver variability in its prognostic utility, generating a need for computerized quantification.

METHODS

We studied 94 patients with renal biopsies, 45 with native diseases and 49 transplant patients with index biopsies for Polyomavirus nephropathy. Chronicity scores were evaluated using two methods. A standard definition diagram was agreed after international consultation and four renal pathologists scored each parameter in a double-blinded manner. Interstitial fibrosis (IF) score was assessed with five different computerized and AI-based algorithms on trichrome and PAS stains.

RESULTS

There was strong prognostic correlation with renal function and graft outcome at a median follow-up ranging from 24 to 42 months respectively, independent of moderate concordance for pathologists scores. IF scores with two of the computerized algorithms showed significant correlation with estimated glomerular filtration rate (eGFR) at biopsy but not at the end of follow-up. There was poor concordance for AI based platforms.

CONCLUSION

Chronicity scores are robust prognostic tools despite interobserver reproducibility. AI-algorithms have absolute precision but are limited by significant variation when different hardware and software algorithms are used for quantification.

AI's Threat to the Medical Profession

This Viewpoint discusses the potential drawbacks of the use of artificial intelligence (AI) in medicine, for example, the loss of certain skills due to the reliance on AI, and how physicians should consider how to take advantage of the potential benefits of AI without losing control over their profession.

Banff Digital Pathology Working Group: Image Bank, Artificial Intelligence Algorithm, and Challenge Trial Developments

The Banff Digital Pathology Working Group (DPWG) was established with the goal to establish a digital pathology repository; develop, validate, and share models for image analysis; and foster collaborations using regular videoconferencing. During the calls, a variety of artificial intelligence (AI)-based support systems for transplantation pathology were presented. Potential collaborations in a competition/trial on AI applied to kidney transplant specimens, including the DIAGGRAFT challenge (staining of biopsies at multiple institutions, pathologists' visual assessment, and development and validation of new and pre-existing Banff scoring algorithms), were also discussed. To determine the next steps, a survey was conducted, primarily focusing on the feasibility of establishing a digital pathology repository and identifying potential hosts. Sixteen of the 35 respondents (46%) had access to a server hosting a digital pathology repository, with 2 respondents that could serve as a potential host at no cost to the DPWG. The 16 digital pathology repositories collected specimens from various organs, with the largest constituent being kidney (n = 12,870 specimens). A DPWG pilot digital pathology repository was established, and there are plans for a competition/trial with the DIAGGRAFT project. Utilizing existing resources and previously established models, the Banff DPWG is establishing new resources for the Banff community.

A Deep Learning-Based System Trained for Gastrointestinal Stromal Tumor Screening Can Identify Multiple Types of Soft Tissue Tumors

The accuracy and timeliness of the pathologic diagnosis of soft tissue tumors (STTs) critically affect treatment decision and patient prognosis. Thus, it is crucial to make a preliminary judgement on whether the tumor is benign or malignant with hematoxylin and eosin-stained images. A deep learning-based system, Soft Tissue Tumor Box (STT-BOX), is presented herein, with only hematoxylin and eosin images for malignant STT identification from benign STTs with histopathologic similarity. STT-BOX assumed gastrointestinal stromal tumor as a baseline for malignant STT evaluation, and distinguished gastrointestinal stromal tumor from leiomyoma and schwannoma with 100% area under the curve in patients from three hospitals, which achieved higher accuracy than the interpretation of experienced pathologists. Particularly, this system performed well on six common types of malignant STTs from The Cancer Genome Atlas data set, accurately highlighting the malignant mass lesion. STT-BOX was able to distinguish ovarian malignant sex-cord stromal tumors without any fine-tuning. This study included mesenchymal tumors that originated from the digestive system, bone and soft tissues, and reproductive system, where the high accuracy of migration verification may reveal the morphologic similarity of the nine types of malignant tumors. Further evaluation in a pan-STT setting would be potential and prospective, obviating the overuse of immunohistochemistry and molecular tests, and providing a practical basis for clinical treatment selection in a timely manner.