The 2022 Banff Meeting Lung Report.

The Lung Session of the 2022 16th Banff Foundation for Allograft Pathology Conference-held in Banff, Alberta-focused on non-rejection lung allograft pathology and novel technologies for the detection of allograft injury. A multidisciplinary panel reviewed the state-of-the-art of current histopathologic entities, serologic studies, and molecular practices, as well as novel applications of digital pathology with artificial intelligence, gene expression analysis, and quantitative image analysis of chest computerized tomography. Current states of need as well as prospective integration of the aforementioned tools and technologies for complete assessment of allograft injury and its impact on lung transplant outcomes were discussed. Key conclusions from the discussion were: (1) recognition of limitations in current standard of care assessment of lung allograft dysfunction; (2) agreement on the need for a consensus regarding the standardized approach to the collection and assessment of pathologic data, inclusive of all lesions associated with graft outcome (eg, non-rejection pathology); and (3) optimism regarding promising novel diagnostic modalities, especially minimally invasive, which should be integrated into large, prospective multicenter studies to further evaluate their utility in clinical practice for directing personalized therapies to improve graft outcomes.

Transcriptional and Immune Landscape of Cardiac Sarcoidosis.

BACKGROUND: Cardiac involvement is an important determinant of mortality among sarcoidosis patients. Although granulomatous inflammation is a hallmark finding in cardiac sarcoidosis, the precise immune cell populations that comprise the granuloma remain unresolved. Furthermore, it is unclear how the cellular and transcriptomic landscape of cardiac sarcoidosis differs from other inflammatory heart diseases. METHODS: We leveraged spatial transcriptomics (GeoMx digital spatial profiler) and single-nucleus RNA sequencing to elucidate the cellular and transcriptional landscape of cardiac sarcoidosis. Using GeoMX digital spatial profiler technology, we compared the transcriptomal profile of CD68+ rich immune cell infiltrates in human cardiac sarcoidosis, giant cell myocarditis, and lymphocytic myocarditis. We performed single-nucleus RNA sequencing of human cardiac sarcoidosis to identify immune cell types and examined their transcriptomic landscape and regulation. Using multichannel immunofluorescence staining, we validated immune cell populations identified by single-nucleus RNA sequencing, determined their spatial relationship, and devised an immunostaining approach to distinguish cardiac sarcoidosis from other inflammatory heart diseases. RESULTS: Despite overlapping histological features, spatial transcriptomics identified transcriptional signatures and associated pathways that robustly differentiated cardiac sarcoidosis from giant cell myocarditis and lymphocytic myocarditis. Single-nucleus RNA sequencing revealed the presence of diverse populations of myeloid cells in cardiac sarcoidosis with distinct molecular features. We identified GPNMB (transmembrane glycoprotein NMB) as a novel marker of multinucleated giant cells and predicted that the MITF (microphthalmia-associated transcription factor) family of transcription factors regulated this cell type. We also detected additional macrophage populations in cardiac sarcoidosis including HLA-DR (human leukocyte antigen-DR)+ macrophages, SYTL3 (synaptotagmin-like protein 3)+ macrophages and CD163+ resident macrophages. HLA-DR+ macrophages were found immediately adjacent to GPMMB+ giant cells, a distinct feature compared with other inflammatory cardiac diseases. SYTL3+ macrophages were located scattered throughout the granuloma and CD163+ macrophages, CD1c+ dendritic cells, nonclassical monocytes, and T cells were located at the periphery and outside of the granuloma. Finally, we demonstrate mTOR (mammalian target of rapamycin) pathway activation is associated with proliferation and is selectively found in HLA-DR+ and SYLT3+ macrophages. CONCLUSIONS: In this study, we identified diverse populations of immune cells with distinct molecular signatures that comprise the sarcoid granuloma. These findings provide new insights into the pathology of cardiac sarcoidosis and highlight opportunities to improve diagnostic testing.

The over diagnosis of diffuse mesothelioma: An analysis of 311 cases with recommendations for the avoidance of pitfalls.

BACKGROUND: The diagnosis of mesothelioma may be challenging. We investigated a large database of cases in order to determine the frequency with which a diagnosis of mesothelioma was made incorrectly and the most frequent causes of error. DESIGN: A database including more than 4000 consultation cases of histologically confirmed mesothelioma was examined to identify cases in which mesothelioma was diagnosed by at least one pathologist when the available information pointed towards a different diagnosis. RESULTS: There were 311 cases misdiagnosed as mesothelioma. The most common category was metastatic carcinoma to the pleura or peritoneum (129 cases: 73 lung carcinomas, 15 renal cell carcinomas). The next most common category was primary lung cancer (111 cases: 55 sarcomatoid carcinoma, 56 pseudomesotheliomatous carcinoma). The third most common category was primary malignancies arising from or near the serosal membranes (33 cases). The fourth most common category was fibrous pleurisy (38 cases). The most common errors were failure to consider important radiographic information regarding the gross distribution of tumor, lack of awareness or consideration of another malignancy, overreliance on certain immunohistochemical results, and failure to perform certain diagnostic histochemical, immunohistochemical, or ultrastructural studies. CONCLUSIONS: There are a number of diagnostic pitfalls that can lead to the over diagnosis of mesothelioma. Careful attention to clinical and radiographic information as well as performance of appropriate ancillary tests can help to prevent such misdiagnoses. Detailed examples will be presented to assist in the avoidance of these pitfalls with emphasis on the most commonly observed errors.

Chronological trends in the causation of malignant mesothelioma: Fiber burden analysis of 619 cases over four decades.

Malignant mesothelioma is a relatively rare malignancy with a strong association with prior asbestos exposure. A percentage of cases is not related to asbestos, and fiber analysis of lung tissue is a useful methodology for identifying idiopathic or spontaneous cases. We have performed fiber analyses in more than 600 cases of mesothelioma over the past four decades and were interested in looking for trends in terms of fiber types and concentrations as well as percentages of cases not related to asbestos. Demographic information was also considered including patient age, gender, and tumor location (pleural vs. peritoneal). The histologic pattern of the tumor and the presence or absence of pleural plaques or asbestosis were noted. Fiber analysis was performed in 619 cases, using the sodium hypochlorite technique for digestion of lung tissue samples. Asbestos bodies were counted by light microscopy (LM) and coated and uncoated fibers by scanning electron microscopy (EM). The results were stratified over four decades. Trends that were observed included increasing patient age, increasing percentage of women, increasing percentage of peritoneal cases, and increasing percentage of epithelial histological type. There was a decreasing trend in the percentage of patients with concomitant asbestosis (p < 0.001). The percentage of cases with an elevated lung asbestos content decreased from 90.5% in the 1980s to 54.1% in the 2010s (p < 0.001). This trend also held when the analysis was limited to 490 cases of pleural mesothelioma in men (91.8% in the 1980s vs. 65.1% in the 2010s). There was a decrease in the median asbestos body count by LM from 1390 asbestos bodies per gram of wet lung in the 1980s to 38 AB/gm in the 2010s. Similar trends were observed for each of the asbestos fiber types as detected by EM. We conclude that there has been a progressive decrease in lung fiber content of mesothelioma patients during the past four decades, with an increasing percentage of cases not related to asbestos and an increase in median patient age.

Genetic variants in DDO and PEX5L in peroxisome-related pathways predict non-small cell lung cancer survival.

Peroxisomes play a role in lipid metabolism and regulation of reactive oxygen species, but its role in development and progression of non-small cell lung cancer (NSCLC) is not well understood. Here, we investigated the associations between 9708 single-nucleotide polymorphisms (SNPs) in 113 genes in the peroxisome-related pathways and survival of NSCLC patients from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) and the Harvard Lung Cancer Susceptibility (HLCS) study. In 1185 NSCLC patients from the PLCO trial, we found that 213 SNPs were significantly associated with NSCLC overall survival (OS) (p ≤ 0.05, Bayesian false discovery probability [BFDP] ≤ 0.80), of which eight SNPs were validated in the HLCS data set. In a multivariate Cox proportional hazards regression model, two independent SNPs (rs9384742 DDO and rs9825224 PEX5L) were significantly associated with NSCLC survival (hazards ratios [HR] of 1.17 with 95% CI [confidence interval] of 1.06-1.28 and 0.86 with 95% CI of 0.77-0.96, respectively). Patients with one or two protective genotypes had a significantly higher OS (HR: 0.787 [95% CI: 0.620-0.998] and 0.691 [95% CI: 0.543-0.879], respectively). Further expression quantitative trait loci analysis using whole blood and lung tissue showed that the minor allele of rs9384742 DDO was significantly associated with decreased messenger RNA (mRNA) expression levels and that DDO expression was also decreased in NSCLC tumor tissue. Additionally, high PEX5L expression levels were significantly associated with lower survival of NSCLC. Our data suggest that variants in these peroxisome-related genes may influence gene regulation and are potential predictors of NSCLC OS, once validated by additional studies.

Genetic variants of CHEK1, PRIM2 and CDK6 in the mitotic phase-related pathway are associated with nonsmall cell lung cancer survival.

The mitotic phase is a vital step in cell division and may be involved in cancer progression, but it remains unclear whether genetic variants in mitotic phase-related pathways genes impact the survival of these patients. Here, we investigated associations between 31 032 single nucleotide polymorphisms (SNPs) in 368 mitotic phase-related pathway genes and overall survival (OS) of patients with nonsmall cell lung cancer (NSCLC). We assessed the associations in a discovery data set of 1185 NSCLC patients from the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial and validated the findings in another data set of 984 patients from the Harvard Lung Cancer Susceptibility Study. As a result, we identified three independent SNPs (ie, CHEK1 rs76744140 T>C, PRIM2 rs6939623 G>T and CDK6 rs113181986 G>C) to be significantly associated with NSCLC OS with an adjusted hazard ratio of 1.29 (95% confidence interval = 1.11-1.49, P = 8.26 × 10-4 ), 1.26 (1.12-1.42, 1.10 × 10-4 ) and 0.73 (0.63-0.86, 1.63 × 10-4 ), respectively. Moreover, the number of combined unfavorable genotypes of these three SNPs was significantly associated with NSCLC OS and disease-specific survival in the PLCO data set (Ptrend < .0001 and .0003, respectively). Further expression quantitative trait loci analysis showed that the rs76744140C allele predicted CHEK1 mRNA expression levels in normal lung tissues and that rs113181986C allele predicted CDK6 mRNA expression levels in whole blood tissues. Additional analyses indicated CHEK1, PRIM2 and CDK6 may impact NSCLC survival. Taken together, these findings suggested that these genetic variants may be prognostic biomarkers of patients with NSCLC.

Potentially functional variants of ERAP1, PSMF1 and NCF2 in the MHC-I-related pathway predict non-small cell lung cancer survival.

BACKGROUND: Cellular immunity against tumor cells is highly dependent on antigen presentation by major histocompatibility complex class I (MHC-I) molecules. However, few published studies have investigated associations between functional variants of MHC-I-related genes and clinical outcomes of lung cancer patients. METHODS: We performed a two-phase Cox proportional hazards regression analysis by using two previously published genome-wide association studies to evaluate associations between genetic variants in the MHC-I-related gene set and the survival of non-small cell lung cancer (NSCLC) patients, followed by expression quantitative trait loci analysis. RESULTS: Of the 7811 single-nucleotide polymorphisms (SNPs) in 89 genes of 1185 NSCLC patients in the discovery dataset of the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial, 24 SNPs remained statistically significant after validation in additional 984 NSCLC patients from the Harvard Lung Cancer Susceptibility Study. In a multivariate stepwise Cox model, three independent functional SNPs (ERAP1 rs469783 T > C, PSMF1 rs13040574 C > A and NCF2 rs36071574 G > A) remained significant with an adjusted hazards ratio (HR) of 0.83 [95% confidence interval (CI) = 0.77-0.89, P = 8.0 × 10-7], 0.86 (0.80-0.93, P = 9.4 × 10-5) and 1.31 (1.11-1.54, P = 0.001) for overall survival (OS), respectively. Further combined genotypes revealed a poor survival in a dose-response manner in association with the number of unfavorable genotypes (Ptrend < 0.0001 and 0.0002 for OS and disease-specific survival, respectively). Also, ERAP1 rs469783C and PSMF1 rs13040574A alleles were associated with higher mRNA expression levels of their genes. CONCLUSION: These potentially functional SNPs of the MHC-I-related genes may be biomarkers for NSCLC survival, possibly through modulating the expression of corresponding genes.

Potentially functional genetic variants in PLIN2, SULT2A1 and UGT1A9 genes of the ketone pathway and survival of nonsmall cell lung cancer.

The ketone metabolism pathway is a principle procedure in physiological homeostasis and induces cancer cells to switch between glycolysis and oxidative phosphorylation for energy production. We conducted a two-phase analysis for associations between genetic variants in the ketone metabolism pathway genes and survival of nonsmall cell lung cancer (NSCLC) by analyzing genotyping data from two published genome-wide association studies (GWASs). In the discovery, we used a genotyping dataset from the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial in the multivariable Cox proportional hazards regression analysis. We used Bayesian false discovery probability (≤0.80) for multiple testing correction to evaluate associations between 25,819 (2,176 genotyped and 23,643 imputed) single-nucleotide polymorphisms (SNPs) in 162 genes and survival of 1,185 NSCLC patients. Subsequently, we validated the identified significant SNPs with an additional 984 NSCLC patients from the Harvard Lung Cancer Susceptibility GWAS study. Finally, we found that three independent and potentially functional SNPs in three different genes (i.e., PLIN2 rs7867814 G>A, SULT2A1 rs2547235 C>T and UGT1A9 rs2011404 C>T) were independently associated with risk of death from NSCLC, with a combined hazards ratio of 1.22 [95% confidence interval = 1.09-1.36 and p = 0.0003], 0.82 (0.74-0.91 and p = 0.0002) and 1.21 (1.10-1.33 and p = 0.0001), respectively. Additional expression quantitative trait loci analysis found that the survival-associated PLIN2 rs7867814 GA + AA genotypes, but not the genotypes of other two SNPs, were significantly associated with increased mRNA expression levels (p = 0.005). These results indicated that PLIN2 variants may be potential predictors of NSCLC survival through regulating the PLIN2 expression.

Novel genetic variants in KIF16B and NEDD4L in the endosome-related genes are associated with nonsmall cell lung cancer survival.

The endosome is a membrane-bound organ inside most eukaryotic cells, playing an important role in adaptive immunity by delivering endocytosed antigens to both MHC class I and II pathways. Here, by analyzing genotyping data from two published genome-wide association studies (GWASs), we evaluated associations between genetic variants in the endosome-related gene-set and survival of patients with nonsmall cell lung cancer (NSCLC). The discovery included 44,112 (3,478 genotyped and 40,634 imputed) single-nucleotide polymorphisms (SNPs) in 220 genes in a singlelocus analysis for their associations with survival of 1,185 NSCLC patients from the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial. After validation of the 821 survival-associated significant SNPs in additional 984 NSCLC patients from the Harvard Lung Cancer Susceptibility Study, 14 SNPs remained significant. The final multivariate stepwise Cox proportional hazards regression modeling of the PLCO dataset identified three potentially functional and independent SNPs (i.e., KIF16B rs1555195 C>T, NEDD4L rs11660748 A>G and rs73440898 A>G) with an adjusted hazards ratio (HR) of 0.86 (95% confidence interval [CI] = 0.79-0.94, p = 0.0007), 1.31 (1.16-1.47, p = 6.0 × 10-5 ) and 1.27 (1.12-1.44, p = 0.0001) for overall survival (OS), respectively. Combined analysis of the adverse genotypes of these three SNPs revealed a trend in the genotype-survival association (ptrend < 0.0001 for OS and ptrend < 0.0001 for disease-specific survival). Furthermore, the survival-associated KIF16B rs1555195T allele was significantly associated with decreased mRNA expression levels of KIF16B in both lung tissues and blood cells. Therefore, genetic variants of the endosome-related genes may be biomarker for NSCLC survival, possibly through modulating the expression of corresponding genes.

Novel genetic variants in HDAC2 and PPARGC1A of the CREB-binding protein pathway predict survival of non-small-cell lung cancer.

The CREB-binding protein (CBP) pathway plays an important role in transcription and activity of acetyltransferase that acetylates lysine residues of histones and nonhistone proteins. In the present study, we hypothesized that genetic variants in the CBP pathway genes played a role in survival of non-small-cell lung cancer (NSCLC). We tested this hypothesis using the genotyping data from the genome-wide association study (GWAS) dataset from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. In the single-locus analysis, we evaluated associations between 13 176 (1107 genotyped and 12 069 imputed) single-nucleotide polymorphisms (SNPs) in 72 genes and survival of 1185 patients with NSCLC. The identified 106 significant SNPs in the discovery were further validated in additional genotyping data from another GWAS dataset of 984 patients with NSCLC in the Harvard Lung Cancer Susceptibility Study. The combined results of two datasets showed that two independent, potentially functional SNPs (i.e., HDAC2 rs13213007G>A and PPARGC1A rs60571065T>A) were significantly associated with NSCLC overall survival, with a combined hazards ratio (HR) of 1.26 (95% confidence interval (CI), 1.09-1.45; P = .002) and 1.23 (1.04-1.47; P = .017), respectively. Furthermore, we performed an expression quantitative trait loci analysis and found that the survival-associated HDAC2 rs13213007A allele (GA+AA), but not PPARGC1A rs60571065A allele (TA+AA), was significantly associated with increased messenger RNA expression levels of HDAC2 in 373 lymphoblastoid cell lines. These results indicate that the HDAC2 rs13213007A allele is a potential predictor of NSCLC survival, likely by altering the HDAC2 expression.

CD68/CD31 immunohistochemistry double stain demonstrates increased accuracy in diagnosing pathologic antibody-mediated rejection in cardiac transplant patients.

Pathologic antibody-mediated rejection (pAMR) occurs in 10% of cardiac transplant patients and is associated with increased mortality. The endomyocardial biopsy remains the primary diagnostic tool to detect and define pAMR. However, certain challenges arise for the pathologist. Accurate identification of >10% of intravascular macrophages along with endothelial swelling, which remains a critical component of diagnosing pAMR, is one such challenge. We used double labeling with an endothelial and histiocytic marker to improve diagnostic accuracy. Twenty-two cardiac transplant endomyocardial biopsies were screened using a CD68/CD31 immunohistochemical (IHC) double stain. To determine whether pAMR diagnosis would change using the double stain, intravascular macrophage staining was compared to using CD68 alone. Twenty-two cardiac pAMR cases from patients were included. Fifty-nine percent of cases previously called >10% intravascular macrophage positive by CD68 alone were called <10% positive using the CD68/CD31 double stain. Not using the double stain was associated with a significant overcall. In C4d-negative cases, using the CD68/CD31 double stain downgraded the diagnosis of pAMR2 to pAMR1 in 32% of cases. It was concluded that more than one third of patients were overdiagnosed with pAMR using CD68 by IHC alone. We demonstrate the value of using a CD68/CD31 double stain to increase accuracy.

Unique Intradural Inflammatory Mass Containing Precipitated Morphine: Confirmatory Analysis by LESA-MS and MALDI-MS.

Opioids are often used for analgesia via continuous intrathecal delivery by implantable devices. A higher concentration and daily dose of opioid have been postulated as risk factors for intrathecal granuloma formation. We present a 42-year-old female patient with chronic abdominal pain from refractory pancreatitis, with an intrathecal drug delivery device implanted 21 years prior, delivering continuous intrathecal morphine. After many years without concerning physical signs or complaints, with gradual increases in daily morphine dose, she presented with rapidly progressive neurologic deficits, including lower extremity, bladder, and bowel symptoms. These symptoms were determined to be secondary to mass effect and local inflammation related to an intrathecal catheter tip granuloma, detected on magnetic resonance imaging of the spine. The mass was urgently resected. On histopathologic examination, this granuloma was found to be unique, in that in addition to the expected inflammatory components, it appeared to contain precipitated nonpolarizable crystals. These were identified as precipitated morphine using liquid extraction surface analysis-tandem mass spectrometry (LESA-MS/MS) and matrix-assisted laser desorption ionization-Fourier transform ion cyclotron resonance-mass spectrometry imaging (MALDI-FTICR-MSI). In addition to the unique finding of precipitated morphine crystals, the long-term follow-up of both morphine concentration and daily dose increases provides insight into the formation of intrathecal granulomas.

Essential role of PR-domain protein MDS1-EVI1 in MLL-AF9 leukemia.

A subgroup of leukemogenic mixed-lineage leukemia (MLL) fusion proteins (MFPs) including MLL-AF9 activates the Mecom locus and exhibits extremely poor clinical prognosis. Mecom encodes EVI1 and MDS1-EVI1 (ME) proteins via alternative transcription start sites; these differ by the presence of a PRDI-BF1-RIZ1 (PR) domain with histone methyltransferase activity in the ME isoform. Using an ME-deficient mouse, we show that ME is required for MLL-AF9-induced transformation both in vitro and in vivo. And, although Nup98-HOXA9, MEIS1-HOXA9, and E2A-Hlf could transform ME-deficient cells, both MLL-AF9 and MLL-ENL were ineffective, indicating that the ME requirement is specific to MLL fusion leukemia. Further, we show that the PR domain is essential for MFP-induced transformation. These studies clearly indicate an essential role of PR-domain protein ME in MFP leukemia, suggesting that ME may be a novel target for therapeutic intervention for this group of leukemias.

The role of EVI1 in myeloid malignancies.

The EVI1 oncogene at human chr 3q26 is rearranged and/or overexpressed in a subset of acute myeloid leukemias and myelodysplasias. The EVI1 protein is a 135 kDa transcriptional regulator with DNA-binding zinc finger domains. Here we provide a critical review of the current state of research into the molecular mechanisms by which this gene plays a role in myeloid malignancies. The major pertinent cellular effects are blocking myeloid differentiation and preventing cellular apoptosis, and several potential mechanisms for these phenomena have been identified. Evidence supports a role for EVI1 in inducing cellular quiescence, and this may contribute to the resistance to chemotherapy seen in patients with neoplasms that overexpress EVI1. Another isoform, MDS1-EVI1 (or PRDM3), encoded by the same locus as EVI1, harbors an N-terminal histone methyltransferase(HMT) domain; experimental findings indicate that this protein and its HMT activity are critical for the progression of a subset of AMLs, and this provides a potential target for therapeutic intervention.

Cardiac magnetic resonance imaging characterization of acute rejection in a porcine heterotopic heart transplantation model.

Preclinical disease models are important for the advancement of therapeutics towards human clinical trials. One of the difficult tasks of developing a well-characterized model is having a reliable modality with which to trend the progression of disease. Acute rejection is one of the most devastating complications that can occur following organ transplantation. Specifically in cardiac transplantation, approximately 12% of patients will experience at least one episode of moderate or severe acute rejection in the first year. Currently, the gold standard for monitoring rejection in the clinical setting is to perform serial endomyocardial biopsies for direct histological assessment. However, this is difficult to reproduce in a porcine model of acute rejection in cardiac transplantation where the heart is heterotopically transplanted in an abdominal position. Cardiac magnetic resonance imaging is arising as an alternative for serial screening for acute rejection in cardiac transplantation. This is an exploratory study to create and define a standardized cardiac magnetic resonance screening protocol for characterizing changes associated with the presence of acute rejection in this preclinical model of disease. Results demonstrate that increases in T1 mapping, T2 mapping, left ventricular mass, and in late gadolinium enhancement are significantly correlated with presence of acute rejection.

Ultrastructural Cardiac Pathology: The Wide (yet so very small) World of Cardiac Electron Microscopy.

Electron microscopy (EM) was a popular diagnostic tool in the 1970s and early 80s. With the adoption of newer, less expensive techniques, such as immunohistochemistry, the role of EM in diagnostic surgical pathology has dwindled substantially. Nowadays, even in academic centers, EM interpretation is relegated to renal pathologists and the handful of (aging) pathologists with experience using the technique. As such, EM interpretation is truly arcane-understood by few and mysterious to many. Nevertheless, there remain situations in which EM is the best or only ancillary test to ascertain a specific diagnosis. Thus, there remains a critical need for the younger generation of surgical pathologists to learn EM interpretation. Recognizing this need, cardiac EM was made the theme of the Cardiovascular Evening Specialty Conference at the 2023 United States and Canadian Academy of Pathology (USCAP) annual meeting in New Orleans, Louisiana. Each of the speakers contributed their part to this article, the purpose of which is to review EM as it pertains to myocardial tissue and provide illustrative examples of the spectrum of ultrastructural cardiac pathology seen in storage/metabolic diseases, cardiomyopathies, infiltrative disorders, and cardiotoxicities.

A machine learning algorithm improves the diagnostic accuracy of the histologic component of antibody mediated rejection (AMR-H) in cardiac transplant endomyocardial biopsies.

BACKGROUND: Pathologic antibody mediated rejection (pAMR) remains a major driver of graft failure in cardiac transplant patients. The endomyocardial biopsy remains the primary diagnostic tool but presents with challenges, particularly in distinguishing the histologic component (pAMR-H) defined by 1) intravascular macrophage accumulation in capillaries and 2) activated endothelial cells that expand the cytoplasm to narrow or occlude the vascular lumen. Frequently, pAMR-H is difficult to distinguish from acute cellular rejection (ACR) and healing injury. With the advent of digital slide scanning and advances in machine deep learning, artificial intelligence technology is widely under investigation in the areas of oncologic pathology, but in its infancy in transplant pathology. For the first time, we determined if a machine learning algorithm could distinguish pAMR-H from normal myocardium, healing injury and ACR. MATERIALS AND METHODS: A total of 4,212 annotations (1,053 regions of normal, 1,053 pAMR-H, 1,053 healing injury and 1,053 ACR) were completed from 300 hematoxylin and eosin slides scanned using a Leica Aperio GT450 digital whole slide scanner at 40X magnification. All regions of pAMR-H were annotated from patients confirmed with a previous diagnosis of pAMR2 (>50% positive C4d immunofluorescence and/or >10% CD68 positive intravascular macrophages). Annotations were imported into a Python 3.7 development environment using the OpenSlide™ package and a convolutional neural network approach utilizing transfer learning was performed. RESULTS: The machine learning algorithm showed 98% overall validation accuracy and pAMR-H was correctly distinguished from specific categories with the following accuracies: normal myocardium (99.2%), healing injury (99.5%) and ACR (99.5%). CONCLUSION: Our novel deep learning algorithm can reach acceptable, and possibly surpass, performance of current diagnostic standards of identifying pAMR-H. Such a tool may serve as an adjunct diagnostic aid for improving the pathologist's accuracy and reproducibility, especially in difficult cases with high inter-observer variability. This is one of the first studies that provides evidence that an artificial intelligence machine learning algorithm can be trained and validated to diagnose pAMR-H in cardiac transplant patients. Ongoing studies include multi-institutional verification testing to ensure generalizability.