Search

Deep learning-based computer vision methods have recently made remarkable breakthroughs in the analysis and classification of cancer pathology images. However, there has been relatively little investigation of the utility of deep neural networks to synthesize medical images. In this study, we evaluated the efficacy of generative adversarial networks to synthesize high-resolution pathology images of 10 histological types of cancer, including five cancer types from The Cancer Genome Atlas and the five major histological subtypes of ovarian carcinoma. The quality of these images was assessed using a comprehensive survey of board-certified pathologists (n = 9) and pathology trainees (n = 6). Our results show that the real and synthetic images are classified by histotype with comparable accuracies and the synthetic images are visually indistinguishable from real images. Furthermore, we trained deep convolutional neural networks to diagnose the different cancer types and determined that the synthetic images perform as well as additional real images when used to supplement a small training set. These findings have important applications in proficiency testing of medical practitioners and quality assurance in clinical laboratories. Furthermore, training of computer-aided diagnostic systems can benefit from synthetic images where labeled datasets are limited (e.g. rare cancers). We have created a publicly available website where clinicians and researchers can attempt questions from the image survey (http://gan.aimlab.ca/). © 2020 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Rationale: Podocyte infolding glomerulopathy (PIG) is a newly described condition with only 37 cases reported worldwide. Due to its rarity, the pathogenesis and evolution of this disease is unclear. This case report contributes to our collective knowledge about the clinical and histological progression of this disease. Presenting concerns of the patient: Over the course of a year, a 52-year-old Malaysian woman with no known prior medical history developed progressively worsening edema and other findings consistent with nephrotic syndrome. Diagnosis: Unlike most patients with PIG, this patient did not have any autoimmune disease. She was Hepatitis B core antibody positive with a Hepatitis B surface antibody >1000, suggesting prior Hepatitis B infection with immunity. A renal biopsy was performed which was consistent with PIG. A second renal biopsy was done 2 years later which again showed characteristic findings of PIG with worsened podocyte effacement but no interval change in chronicity. Interventions: The patient was treated with blood pressure control and renin-angiotensin-aldosterone system (RAAS) blockade with irbesartan and spironolactone. She was also treated with prednisone at 1 mg/kg for 2 months followed by a taper for a total of 7 months of prednisone treatment. Outcomes: The patient had a partial response to a course of prednisone. However, since stopping steroids, her proteinuria and renal function has been gradually worsening. Teaching points: PIG is mostly found in patients of East Asian descent. It presents as proteinuria and is often associated with autoimmune disease but can be idiopathic. It is characterized on renal biopsy by infolding or protrusion of podocyte cytoplasm into glomerular basement membrane, as well as intramembranous cytoplasmic microspherules or microtubules. Atypical membranous nephropathy should be ruled out prior to diagnosis. Unlike membranous nephropathy, PIG usually responds at least partially to steroid monotherapy. To our knowledge, this is the first reported case of PIG from North America. Furthermore, it is the first case of PIG with repeat biopsy showing interval worsening of PIG rather than either resolution of PIG or transformation of PIG to a different diagnosis.

This chapter proposes a summary of the current understanding of kidney allograft rejection, evolving from a traditional conceptualization of discrete episodes to a model that recognizes phases of immune activation directed toward the allograft. This starts with detection of well-understood forms of early T cell-mediated rejection (TCMR) or antibody-mediated rejection (AMR) that typify initiation of allograft directed immune injury, and maturation to more complex and chronic forms of rejection that are characterized most often by integrated T and B responses and memory. The risk of progression to chronic forms of rejection may depend on the degree with which alloimmune responsiveness is consistently subdued by immunosuppressant medication, but also the degree with which either the innate and cognate effector responses are secondarily activated or by which adaptive, regulatory responses restrain inflammation.

BACKGROUND Due to a lack of consensus on SB for pediatric kidney transplant recipients, we evaluated the yield and clinical utility of SB findings at various time points post-transplant. METHODS Patients transplanted at a single institution between 2014 and 2020 with at least one SB at 1.5, 3, 6, 12, and 24 months post-transplant were included. Additional biopsies were done for indication (IB). TCMR was classified by Banff criteria (score ≥i1t1). RESULTS Forty-seven patients had 142 biopsies (SB = 113, IB = 29); 19 (40.4%) of whom experienced at least one TCMR episode in the first-year post-transplant. The greatest SB yield of any pathologic abnormality was at 6 months (57.1%; P

Background Early dysfunction of renal allografts may be associated with vascular injury, which raises the specter of active rejection processes that require medical intervention. In our practice, we have encountered patients who present with delayed graft function and demonstrate a unique pattern of endothelial cell injury that raises concern for rejection in their biopsy. Therefore, we sought to systematically determine the biopsy characteristics and outcome of these patients. Methods During a 17-year period at the University of Washington in Seattle, United States, we identified 24 cases of a distinct arterial vasculopathy presenting in the first year posttransplantation. This early transplant arteriopathy (ETA) is characterized by endothelial cell swelling and intimal edema but without the intimal arteritis that defines vascular rejection. Results Approximately 1% of transplant biopsies during the study period showed ETA, almost all of which were in deceased donor organs (96%), and most presented with delayed graft function (54%) or increased serum creatinine (38%) soon after transplantation (median 13 days; range, 5-139). In this study, 77% of patients were managed expectantly, with only 2 patients (7.6%) subsequently developing acute vascular rejection. Except for 1 patient who died, all patients had functioning allografts at 1 year follow-up. Conclusion Recognizing ETA and distinguishing it from vascular rejection is important to prevent over-treatment because most patients appear to recover allograft function rapidly with expectant management.

Deep learning-based computer vision methods have recently made remarkable breakthroughs in the analysis and classification of cancer pathology images. However, there has been relatively little investigation of the utility of deep neural networks to synthesize medical images. In this study, we evaluated the efficacy of generative adversarial networks (GANs) to synthesize high resolution pathology images of ten histological types of cancer, including five cancer types from The Cancer Genome Atlas (TCGA) and the five major histological subtypes of ovarian carcinoma. The quality of these images was assessed using a comprehensive survey of board-certified pathologists (n = 9) and pathology trainees (n = 6). Our results show that the real and synthetic images are classified by histotype with comparable accuracies, and the synthetic images are visually indistinguishable from real images. Furthermore, we trained deep convolutional neural networks (CNNs) to diagnose the different cancer types and determined that the synthetic images perform as well as additional real images when used to supplement a small training set. These findings have important applications in proficiency testing of medical practitioners and quality assurance in clinical laboratories. Furthermore, training of computer-aided diagnostic systems can benefit from synthetic images where labeled datasets are limited (e.g., rare cancers). We have created a publicly available website where clinicians and researchers can attempt questions from the image survey at http://gan.aimlab.ca/.

The bioactive sphingolipid ceramide 1-phosphate (C1P) regulates cell division in a variety of cell types including macrophages. However, the mechanisms involved in this action are not completely understood. In the present work we show that C1P stimulates the release of vascular endothelial growth factor (VEGF) in RAW264.7 macrophages, and that this growth factor is essential for stimulation of cell proliferation by C1P. The stimulation of VEGF release was dependent upon activation of the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (PKB-1 also known as Akt-1), and mitogen-activated protein kinase-kinase (MEK)/extracellularly regulated kinase-2 (ERK-2) pathways, as inhibition of these kinases with selective pharmacological inhibitors or with specific gene silencing siRNA, abrogated VEGF release. A key observation was that sequestration of VEGF with a neutralizing antibody, or treatment with VEGF siRNA abolished C1P-stimulated macrophage growth. Also, inhibition of the pathways involved in C1P-stimulated VEGF release inhibited the stimulation of macrophage growth by C1P. Moreover, blockade of VEGF receptor-2 (VEGFR-2), which is the primary receptor for VEGF, with the pharmacological inhibitor DMH4, or with specific VEGFR-2 siRNA, substantially inhibited C1P-stimulated cell growth. It can be concluded that stimulation of VEGF release is a key factor in the promotion of macrophage proliferation by C1P.

Expression of human equilibrative nucleoside transporter 1 (hENT1) in pancreatic ductal adenocarcinoma (PDAC) has been postulated to be a marker of sensitivity to gemcitabine. However, heterogeneity in the studies attempting to quantify hENT1 expression in patients with PDAC treated with gemcitabine has yielded inconclusive results that impede the adoption of hENT1 expression as a predictive biomarker. Tissue microarrays consisting of PDAC specimens from 227 patients acquired between 1987 and 2013 annotated with treatment and outcome information were subjected to staining with two antibodies for hENT1 (10D7G2 and SP120) on a single automated platform and scored by two independent pathologists blinded to treatment and outcome. The resultant scores were subjected to individual predictive disease-specific survival analysis and to unsupervised hierarchical clustering to generate a multi-marker classification. Tumour cell staining prevalence using either SP120 or 10D7G2 was predictive of gemcitabine sensitivity (p = 0.02; p = 0.01). When combined, three groups emerged, classified as SP120Low_10D7G2Low, SP120Low_10D7G2High, and SP120High_10D7G2High, in which adjuvant gemcitabine conferred median survival differences of 0.2, 0.8, and 1.5 (p = 0.76, p = 0.06, p = 0.01) years, respectively. These results were largely replicated in multivariable analysis with the P value for the SP120Low_10D7G2High cluster achieving statistical significance (p = 0.03). These data suggest that either antibody for hENT1 can be used to predict gemcitabine sensitivity in resected PDAC. However, using both antibodies adds valuable information that enables the stratification of patients who can expect to have a good, intermediate, and poor response to adjuvant gemcitabine.

BACKGROUND AND OBJECTIVES: Fibrillary GN has been defined as an immune complex-mediated GN with amyloid-like fibrils larger than amyloid which are IgG positive and Congo red negative. With discovery of DNAJB9 as a highly sensitive and specific marker for fibrillary GN, the specificity of the morphologic criteria for establishing the diagnosis of fibrillary GN has come into question. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: We sought to (1) determine anatomic characteristics that best define fibrillary GN and (2) identify clinical and pathologic features that predict outcomes. RESULTS: We retrospectively reviewed kidney biopsies from patients diagnosed with fibrillary GN or suspected fibrillary GN between 1997 and 2017 (n=266, 65% female, median age 61). Approximately 11% of kidney biopsies had one or more unusual feature including monotypic deposits, Congo red positivity, or unusual fibril diameter. Fibrillary GN as a possible monoclonal gammopathy of renal significance represented

Deficiencies in DNA mismatch repair have been associated with inferior response to 5-FU in colorectal cancer. Pancreatic ductal adenocarcinoma is similarly treated with pyrimidine analogs, yet the predictive value of mismatch repair status for response to these agents has not been examined in this malignancy. A tissue microarray with associated clinical outcome, comprising 254 resected pancreatic ductal adenocarcinoma patients was stained for four mismatch repair proteins (MLH1, MSH2, MSH6 and PMS2). Mismatch repair deficiency and proficiency was determined by the absence or presence of uniform nuclear staining in tumor cells, respectively. Cases identified as mismatch repair deficient on the tissue microarray were confirmed by immunohistochemistry on whole slide sections. Of the 265 cases, 78 (29%) received adjuvant treatment with a pyrimidine analog and 41 (15%) showed a mismatch repair-deficient immunoprofile. Multivariable disease-specific survival in the mismatch repair-proficient cohort demonstrated that adjuvant chemotherapy, regional lymph-node status, gender, and the presence of tumor budding were significant independent prognostic variables (P≤0.04); however, none of the eight clinico-pathologic covariates examined in the mismatch repair-deficient cohort were of independent prognostic significance. Univariable assessment of disease-specific survival revealed an almost identical survival profile for both treated and untreated patients with a mismatch repair-deficient profile, while treatment in the mismatch repair-proficient cohort conferred a greater than 10-month median disease-specific survival advantage over their untreated counterparts (P=0.0018). In this cohort, adjuvant chemotherapy with a pyrimidine analog conferred no survival advantage to mismatch repair-deficient pancreatic ductal adenocarcinoma patients. Mismatch repair immunoprofiling is a feasible predictive marker in pancreatic ductal adenocarcinoma patients, and further prospective evaluation of this finding is warranted.

We tested whether loss of eukaryotic elongation factor 2 kinase (eEF2K) activity in macrophages suppresses development of atherosclerosis by transplanting bone marrow from mice with mutant eEF2K into ldlr−/− mice. Sixteen weeks after high-fat diet feeding, mutant eEF2K hematopoietic chimeras had a dramatically reduced level of atherosclerotic plaque formation. M1-skewed macrophages from eEF2K knock-in mice have less tumour necrosis factor-α release and a lesser ability to induce expression of endothelial cell markers, providing a potential explanation for the role of eEF2K. Because eEF2K activity in cells of the hematopoietic compartment contributes to atherosclerosis development, drugs inhibiting eEF2K might have a beneficial effect in treatment of atherosclerosis.

Familial hypobetalipoproteinemia (FHBL) is a rare genetic disorder of lipid metabolism that is associated with abnormally low serum levels of low-density lipoprotein (LDL) cholesterol and apolipoprotein B. It is an autosomal co-dominant disorder, and depending on zygosity, the clinical manifestations may vary from none to neurological, endocrine, hematological or liver dysfunction. Nonalcoholic fatty liver disease is common in persons with FHBL, however progression to nonalcoholic steatohepatitis is unusual. We describe here a patient with a novel APOB mutation, V703I, which appears to contribute to the severity of the FHBL phenotype. He had liver enzyme abnormalities, increased echogenicity of the liver consistent with steatosis, very low LDL cholesterol at 0.24 mmol/l (normal 1.8–3.5 mmol/l) and an extremely low apolipoprotein B level of 0.16 g/l (normal 0.6–1.2 g/l). APOB gene sequencing revealed him to be a compound heterozygote with two mutations (R463W and V703I). APOB R463W has previously been reported to cause FHBL. Genetic sequencing of his first-degree relatives identified the APOB V703I mutation in his normolipidemic brother and father and the APOB R463W mutation in his mother and sister, both of whom have very low LDL cholesterol levels. These results suggest that the APOB V703I mutation alone does not cause the FHBL phenotype. However, it is possible that it has a contributory role to a more aggressive phenotype in the presence of APOB R463W.

We recently reported that oxidized LDL (oxLDL) induces an oscillatory increase in intracellular calcium ([Ca(2+)](i)) levels in macrophages. Furthermore, we have shown that these [Ca(2+)](i) oscillations mediate oxLDL's ability to inhibit macrophage apoptosis in response to growth factor deprivation. However, the signal transduction pathways by which oxLDL induces [Ca(2+)](i) oscillations have not been elucidated. In this study, we show that these oscillations are mediated in part by intracellular mechanisms, as depleting extracellular Ca(2+) did not completely abolish the effect. Inhibiting sarco-endoplasmic reticulum ATPase (SERCA) completely blocked [Ca(2+)](i) oscillations, suggesting a role for Ca(2+) reuptake by the ER. The addition of oxLDL resulted in an almost immediate activation of sphingosine kinase (SK), which can increase sphingosine-1-phosphate (S1P) levels by phosphorylating sphingosine. Moreover, S1P was shown to be as effective as oxLDL in blocking macrophage apoptosis and producing [Ca(2+)](i) oscillations. This suggests that the mechanism in which oxLDL generates [Ca(2+)](i) oscillations may be 1) activation of SK, 2) SK-mediated increase in S1P levels, 3) S1P-mediated Ca(2+) release from intracellular stores, and 4) SERCA-mediated Ca(2+) reuptake back into the ER.

Oxidized LDL (OxLDL) is thought to play a role in the pathogenesis of early as well as advanced stages of atherosclerosis. One possible mechanism involves local upregulation of pro-inflammatory cytokines such as vascular endothelial growth factor (VEGF). This study was done to define the mechanism by which OxLDL increases secretion of VEGF in macrophages. The murine leukemia-derived RAW 264.7 macrophage cell line as well as mouse peritoneal macrophages and human monocyte-derived macrophages were used in these studies. Cells were exposed to native low-density lipoprotein (LDL), acetylated LDL, and LDL that had been modified by oxidation with copper or ferrous ions or by exposure to auto-oxidation products of arachidonic acid for 16h, and VEGF was then assayed in medium. Pharmacological inhibitors of phosphatidylinositol 3-kinase (PI3K) or PKCzeta blocked VEGF secretion by OxLDL. Inhibitors of other protein kinase C (PKC) subtypes had no effect, and neither did inhibitors of mitogen activated protein kinase kinase (MAPKK). We found that LDL with oxidative modification of either its lipid or protein component can induce VEGF expression. Higher degrees of oxidation of LDL conferred higher potency to induce VEGF. Macrophages from mice lacking both scavenger receptors A (SR-A) and CD36 were fully responsive to OxLDL with regard to VEGF secretion. These macrophages show an 85% reduction in OxLDL uptake compared to macrophages from wild-type mice. Macrophages from mice lacking LOX-1 were also fully responsive to oxLDL with regard to VEGF secretion. We conclude that VEGF upregulation is mediated through PI3K and PKCzeta, and does not involve the above three scavenger receptors or require uptake of oxidized LDL.

Macrophages are prominent components of human atherosclerotic lesions and they are believed to accelerate the progression and/or complications of both early and advanced atherosclerotic lesions. We and others have shown that oxidized low-density lipoprotein (oxLDL) induces growth and inhibits apoptosis in murine bone marrow-derived macrophages. In this study, we sought to characterize the oxidative modification of LDL that is responsible for this prosurvival effect. We found that both the modified lipid and the modified protein components of oxLDL can increase the viability of macrophages. The key modification appeared to involve derivatization of amino groups in apoB or in phosphatidylethanolamine by lipid peroxidation products. These reactive oxidation products were primarily unfragmented hydroperoxide- or endoperoxide-containing oxidation products of linoleic acid or arachidonic acid. LC-MS/MS studies showed that some of the arachidonic acid-derived lysine adducts were isolevuglandins that contain lactam and hydroxylactam rings. MS/MS analysis of linoleic acid autoxidation adducts was consistent with 5- or 6-membered nitrogen-containing heterocycles derived from unfragmented oxidation products. The amine modification by oxidation products generated a fluorescence pattern with an excitation maximum at 350nm and emission maximum at 430nm. This is very similar to the fluorescence spectrum of copper-oxidized LDL.

Oxidized LDL (oxLDL) promotes lipid accumulation as well as growth and survival signaling in macrophages. OxLDL uptake is mainly due to scavenger receptors SR-AI/II and CD36. However, other scavenger receptors such as lectin-like oxLDL receptor-1 (LOX-1) may also play a role. We used mice with targeted inactivation of the LOX-1 gene to define the role of this receptor in the uptake of oxLDL and in activation of survival pathways. There was no difference in uptake or degradation of 125I-oxLDL in unstimulated macrophages from wild-type and LOX-1 knockout mice and no difference in the rate of clearance of oxLDL from plasma in vivo. However, when expression of LOX-1 was induced with lysophosphatidylcholine, oxLDL uptake and degradation increased 2-fold in wild-type macrophages but did not change in LOX-1 knockout macrophages. Macrophages lacking LOX-1 showed the same stimulation of PKB phosphorylation and enhancement of survival by oxLDL as wild-type cells. These data show that LOX-1 does not alter the uptake of oxLDL in unstimulated macrophages and is not essential for the pro-survival effect of oxLDL in these cells. However, LOX-1 expression is highly inducible by lysophosphatidylcholine and pro-inflammatory cytokines, and if that occurred in macrophages within atheromas, LOX-1 could substantially increase oxLDL uptake by lesion macrophages.

Objective— Macrophage survival and proliferation is believed to be a contributing factor in the development of early atherosclerotic lesions. Oxidized low density lipoprotein (oxLDL), a key mediator in the pathogenesis of this disease, has been shown to block apoptosis in macrophages deprived of growth factor. In this report, we investigate the mechanism of oxLDL-mediated macrophage survival. Methods and Results— OxLDL, but not native LDL (nLDL), induces an immediate and oscillatory increase in intracellular calcium ([Ca 2+ ] i ). We also show that the calcium/calmodulin dependent kinase, eukaryotic elongation factor-2 kinase (eEF2 kinase), is activated in response to oxLDL, an effect that can be blocked by inhibiting calcium mobilization. Furthermore, selective inhibition of eEF2 kinase reverses the prosurvival effect of oxLDL and results in cellular apoptosis. p38 MAP kinase, a negative regulator of eEF2 kinase, is activated on growth factor withdrawal, a response that can be inhibited by oxLDL. Finally, we show that oxLDL, by activating eEF2 kinase, phosphorylates and therefore inhibits eEF2, resulting in an overall decrease in protein synthesis. Conclusion— These results indicate a novel signaling pathway in which oxLDL can block macrophage apoptosis by mobilizing calcium and activating eEF2 kinase.