Automated Evaluation of p16/Ki-67 Dual-Stain Cytology as a Biomarker for Detection of Anal Precancer in Men Who Have Sex With Men and Are Living With Human Immunodeficiency Virus.

BACKGROUND: Human papillomavirus-related biomarkers such as p16/Ki-67 "dual-stain" (DS) cytology have shown promising clinical performance for anal cancer screening. Here, we assessed the performance of automated evaluation of DS cytology (automated DS) to detect anal precancer in men who have sex with men (MSM) and are living with human immunodeficiency virus (HIV). METHODS: We conducted a cross-sectional analysis of 320 MSM with HIV undergoing anal cancer screening and high-resolution anoscopy (HRA) in 2009-2010. We evaluated the performance of automated DS based on a deep-learning classifier compared to manual evaluation of DS cytology (manual DS) to detect anal intraepithelial neoplasia grade 2 or 3 (AIN2+) and grade 3 (AIN3). We evaluated different DS-positive cell thresholds quantified by the automated approach and modeled performance compared with other screening strategies in a hypothetical population of MSM with HIV. RESULTS: Compared with manual DS, automated DS had significantly higher specificity (50.9% vs 42.2%; P < .001) and similar sensitivity (93.2% vs 92.1%) for detection of AIN2+. Human papillomavirus testing with automated DS triage was significantly more specific than automated DS alone (56.5% vs 50.9%; P < .001), with the same sensitivity (93.2%). In a modeled analysis assuming a 20% AIN2+ prevalence, automated DS detected more precancers than manual DS and anal cytology (186, 184, and 162, respectively) and had the lowest HRA referral rate per AIN2+ case detected (3.1, 3.5, and 3.3, respectively). CONCLUSIONS: Compared with manual DS, automated DS detects the same number of precancers, with a lower HRA referral rate.

Astrocytic glutamine synthetase is expressed in the neuronal somatic layers and down-regulated proportionally to neuronal loss in the human epileptic hippocampus.

Human mesial temporal lobe epilepsy (MTLE) features subregion-specific hippocampal neurodegeneration and reactive astrogliosis, including up-regulation of the glial fibrillary acidic protein (GFAP) and down-regulation of glutamine synthetase (GS). However, the regional astrocytic expression pattern of GFAP and GS upon MTLE-associated neurodegeneration still remains elusive. We assessed GFAP and GS expression in strict correlation with the local neuronal number in cortical and hippocampal surgical specimens from 16 MTLE patients using immunohistochemistry, stereology and high-resolution image analysis for digital pathology and whole-slide imaging. In the cortex, GS-positive (GS+) astrocytes are dominant in all neuronal layers, with a neuron to GS+ cell ratio of 2:1. GFAP-positive (GFAP+) cells are widely spaced, with a GS+ to GFAP+ cell ratio of 3:1-5:1. White matter astrocytes, on the contrary, express mainly GFAP and, to a lesser extent, GS. In the hippocampus, the neuron to GS+ cell ratio is approximately 1:1. Hippocampal degeneration is associated with a reduction of GS+ astrocytes, which is proportional to the degree of neuronal loss and primarily present in the hilus. Up-regulation of GFAP as a classical hallmark of reactive astrogliosis does not follow the GS-pattern and is prominent in the CA1. Reactive alterations were proportional to the neuronal loss in the neuronal somatic layers (stratum pyramidale and hilus), while observed to a lesser extent in the axonal/dendritic layers (stratum radiatum, molecular layer). We conclude that astrocytic GS is expressed in the neuronal somatic layers and, upon neurodegeneration, is down-regulated proportionally to the degree of neuronal loss.

LOC283731 promoter hypermethylation prognosticates survival after radiochemotherapy in IDH1 wild-type glioblastoma patients.

MGMT promoter methylation status is currently the only established molecular prognosticator in IDH wild-type glioblastoma multiforme (GBM). Therefore, we aimed to discover novel therapy-associated epigenetic biomarkers. After enrichment for hypermethylated fractions using methyl-CpG-immunoprecipitation (MCIp), we performed global DNA methylation profiling for 14 long-term (LTS; >36 months) and 15 short-term (STS; 6-10 months) surviving GBM patients. Even after exclusion of the G-CIMP phenotype, we observed marked differences between the LTS and STS methylome. A total of 1,247 probes in 706 genes were hypermethylated in LTS and 463 probes in 305 genes were found to be hypermethylated in STS patients (p values < 0.05, log2 fold change ± 0.5). We identified 13 differentially methylated regions (DMRs) with a minimum of four differentially methylated probes per gene. Indeed, we were able to validate a subset of these DMRs through a second, independent method (MassARRAY) in our LTS/STS training set (ADCY1, GPC3, LOC283731/ISLR2). These DMRs were further assessed for their prognostic capability in an independent validation cohort (n = 62) of non-G-CIMP GBMs from the TCGA. Hypermethylation of multiple CpGs mapping to the promoter region of LOC283731 correlated with improved patient outcome (p = 0.03). The prognostic performance of LOC283731 promoter hypermethylation was confirmed in a third independent study cohort (n = 89), and was independent of gender, performance (KPS) and MGMT status (p = 0.0485, HR = 0.63). Intriguingly, the prediction was most pronounced in younger GBM patients (<60 years). In conclusion, we provide compelling evidence that promoter methylation status of this novel gene is a prognostic biomarker in IDH1 wild-type/non-G-CIMP GBMs.

Spatial transcriptome analysis reveals Notch pathway-associated prognostic markers in IDH1 wild-type glioblastoma involving the subventricular zone.

BACKGROUND: The spatial relationship of glioblastoma (GBM) to the subventricular zone (SVZ) is associated with inferior patient survival. However, the underlying molecular phenotype is largely unknown. We interrogated an SVZ-dependent transcriptome and potential location-specific prognostic markers. METHODS: mRNA microarray data of a discovery set (n = 36 GBMs) were analyzed for SVZ-dependent gene expression and process networks using the MetaCore™ workflow. Differential gene expression was confirmed by qPCR in a validation set of 142 IDH1 wild-type GBMs that was also used for survival analysis. RESULTS: Microarray analysis revealed a transcriptome distinctive of SVZ+ GBM that was enriched for genes associated with Notch signaling. No overlap was found to The Cancer Genome Atlas's molecular subtypes. Independent validation of SVZ-dependent expression confirmed four genes with simultaneous prognostic impact: overexpression of HES4 (p = 0.034; HR 1.55) and DLL3 (p = 0.017; HR 1.61) predicted inferior, and overexpression of NTRK2 (p = 0.049; HR 0.66) and PIR (p = 0.025; HR 0.62) superior overall survival (OS). Additionally, overexpression of DLL3 was predictive of shorter progression-free survival (PFS) (p = 0.043; HR 1.64). Multivariate analysis revealed overexpression of HES4 to be independently associated with inferior OS (p = 0.033; HR 2.03), and overexpression of DLL3 with inferior PFS (p = 0.046; HR 1.65). CONCLUSIONS: We identified four genes with SVZ-dependent expression and prognostic significance, among those HES4 and DLL3 as part of Notch signaling, suggesting further evaluation of location-tailored targeted therapies.

Oral treatment with a zinc complex of acetylsalicylic acid prevents diabetic cardiomyopathy in a rat model of type-2 diabetes: activation of the Akt pathway.

BACKGROUND: Type-2 diabetics have an increased risk of cardiomyopathy, and heart failure is a major cause of death among these patients. Growing evidence indicates that proinflammatory cytokines may induce the development of insulin resistance, and that anti-inflammatory medications may reverse this process. We investigated the effects of the oral administration of zinc and acetylsalicylic acid, in the form of bis(aspirinato)zinc(II)-complex Zn(ASA)2, on different aspects of cardiac damage in Zucker diabetic fatty (ZDF) rats, an experimental model of type-2 diabetic cardiomyopathy. METHODS: Nondiabetic control (ZL) and ZDF rats were treated orally with vehicle or Zn(ASA)2 for 24 days. At the age of 29-30 weeks, the electrical activities, left-ventricular functional parameters and left-ventricular wall thicknesses were assessed. Nitrotyrosine immunohistochemistry, TUNEL-assay, and hematoxylin-eosin staining were performed. The protein expression of the insulin-receptor and PI3K/AKT pathway were quantified by Western blot. RESULTS: Zn(ASA)2-treatment significantly decreased plasma glucose concentration in ZDF rats (39.0 ± 3.6 vs 49.4 ± 2.8 mM, P < 0.05) while serum insulin-levels were similar among the groups. Data from cardiac catheterization showed that Zn(ASA)2 normalized the increased left-ventricular diastolic stiffness (end-diastolic pressure-volume relationship: 0.064 ± 0.008 vs 0.084 ± 0.014 mmHg/µl; end-diastolic pressure: 6.5 ± 0.6 vs 7.9 ± 0.7 mmHg, P < 0.05). Furthermore, ECG-recordings revealed a restoration of prolonged QT-intervals (63 ± 3 vs 83 ± 4 ms, P < 0.05) with Zn(ASA)2. Left-ventricular wall thickness, assessed by echocardiography, did not differ among the groups. However histological examination revealed an increase in the cardiomyocytes' transverse cross-section area in ZDF compared to the ZL rats, which was significantly decreased after Zn(ASA)2-treatment. Additionally, a significant fibrotic remodeling was observed in the diabetic rats compared to ZL rats, and Zn(ASA)2-administered ZDF rats showed a similar collagen content as ZL animals. In diabetic hearts Zn(ASA)2 significantly decreased DNA-fragmentation, and nitro-oxidative stress, and up-regulated myocardial phosphorylated-AKT/AKT protein expression. Zn(ASA)2 reduced cardiomyocyte death in a cellular model of oxidative stress. Zn(ASA)2 had no effects on altered myocardial CD36, GLUT-4, and PI3K protein expression. CONCLUSIONS: We demonstrated that treatment of type-2 diabetic rats with Zn(ASA)2 reduced plasma glucose-levels and prevented diabetic cardiomyopathy. The increased myocardial AKT activation could, in part, help to explain the cardioprotective effects of Zn(ASA)2. The oral administration of Zn(ASA)2 may have therapeutic potential, aiming to prevent/treat cardiac complications in type-2 diabetic patients.

Molecular profiling of long-term survivors identifies a subgroup of glioblastoma characterized by chromosome 19/20 co-gain.

Glioblastoma (GBM) is a devastating tumor and few patients survive beyond 3 years. Defining the molecular determinants underlying long-term survival is essential for insights into tumor biology and biomarker identification. We therefore investigated homogeneously treated, IDH (wt) long-term (LTS, n = 10) and short-term survivors (STS, n = 6) by microarray transcription profiling. While there was no association of clinical parameters and molecular subtypes with long-term survival, STS tumors were characterized by differential polarization of infiltrating microglia with predominance of the M2 phenotype detectable both on the mRNA and protein level. Furthermore, transcriptional signatures of LTS and STS predicted patient outcome in a large, IDH (wt) cohort (n = 468). Interrogation of overlapping genomic alterations identified concurrent gain of chromosomes 19 and 20 as a favorable prognostic marker. The strong association of this co-gain with survival was validated by aCGH in a second, independent cohort (n = 124). Finally, FISH and gene expression data revealed gains to constitute low-amplitude, clonal events with a strong impact on transcription. In conclusion, these findings provide important insights into the manipulation of the innate immune system by particularly aggressive GBM tumors. Furthermore, we genomically characterize a previously unknown, clinically relevant subgroup of glioblastoma, which can easily be identified through modern neuropathological workup.

Pathogenic role of the eight probably/possibly carcinogenic HPV types 26, 53, 66, 67, 68, 70, 73 and 82 in cervical cancer.

Eight HPV types (HPV26, 53, 66, 67, 68, 70, 73 and 82) that are phylogenetically closely related to 12 WHO-defined high-risk (HR) HPV have been rarely but consistently identified as single HPV infections in about 3% of cervical cancer (CxCa) tissues. Due to lack of biological data, these types are referred to as probable/possible (p) HR-HPV. To analyse their biological activity in direct comparison to HR-HPV types, we selected 55 formalin-fixed, paraffin-embedded (FFPE) CxCa tissues harbouring single pHR-HPV infections (2-13 cases per type) and 266 tissues harbouring single HR-HPV (7-40 cases per type) from a worldwide, retrospective, cross-sectional study. Single HPV infection was verified by two genotyping methods. Presence of type-specific spliced E6*I mRNA transcripts and expression of cellular proteins indicative of HPV transformation were assessed in all cases. In 55 CxCa tissues with pHR-HPV, E6*I mRNA expression was 100%; high p16(INK4a) , 98%; low pRb, 96%; low CyD1, 93%; and low p53, 84%. Compared to HPV16 tissues as a reference, individual frequencies of these five markers did not differ significantly, either for any of the eight pHR-HPV and the 11 other HR types individually or for the groups of pHR and HR types without HPV16. We conclude that the eight pHR-HPV types, when present as a single infection in CxCa, are biologically active and affect the same cellular pathways as any of the fully recognized carcinogenic HR-HPV types. Therefore we have provided molecular evidence of carcinogenicity for types currently classified as probably/possibly carcinogenic. Although this evidence is crucial for HPV-type carcinogenicity classification, per se it is not sufficient for inclusion of these HPV types into population-wide primary and secondary prevention programmes. Such decisions have to include careful estimation of effectiveness and cost-benefit analyses.

Interobserver reproducibility of cervical histology interpretation with and without p16 immunohistochemistry.

OBJECTIVES: Histopathological diagnosis of colposcopically identified cervical lesions is a critical step for the recognition of cervical cancer precursors requiring treatment. Although there have been efforts to standardize the histologic diagnosis of cervical biopsy specimens, in terms of terminology and use of biomarkers, there is no uniform approach in the pathology community. Adjunctive p16 immunohistochemistry (IHC) can highlight precancer diagnoses, with use recommendations outlined by the Lower Anogenital Squamous Terminology project. METHODS: We assessed the diagnostic reproducibility of cervical histopathological biopsy specimens with and without p16 staining among 2 expert pathologists. RESULTS: Interpretation of p16 IHC as positive vs negative was highly reproducible (92.5% agreement, κ = 0.85); greater variation was seen in the choice of which biopsy specimens required adjunctive p16 staining (78.0% agreement, κ = 0.43). Adjunctive p16 IHC did not significantly increase diagnostic agreement under multitiered grading systems (benign vs cervical intraepithelial neoplasia [CIN] 1/low-grade squamous intraepithelial lesion vs atypical squamous metaplasia vs CIN2/high-grade squamous intraepithelial lesion [HSIL] vs CIN3/HSIL-CIN3 vs cancer) (65.5% agreement, κ = 0.56 without p16; 70.0% agreement, κ = 0.58 with p16). However, when dichotomizing diagnoses based on clinical management (less than HSIL vs HSIL+), diagnostic agreement increased with p16 IHC (90.5% agreement, κ = 0.79 without p16; 92.0% agreement, κ = 0.84 with p16). For biopsy specimens taken from women positive for human papillomavirus (HPV) type 16, agreement was similar with or without adjunctive p16 (κ = 0.80 without p16; κ = 0.78-0.80 with p16). In contrast, p16 IHC substantially improved diagnostic agreement for cervical biopsy specimens taken from women positive for other high-risk HPV strains, producing improvements in κ from 0.03 to 0.24. CONCLUSIONS: Adjunctive p16 immunostaining provides useful information in the evaluation of cervical biopsies for precancer. In our study, we have demonstrated that it is highly reproducible between 2 pathologists, although the decision of which biopsies warrant its use is less so. Furthermore, although p16 IHC showed a limited increase in diagnostic reproducibility for all biopsies included in our study, it did demonstrate a more sizable gain in biopsies negative for HPV 16 but positive for other high-risk genotypes. Further studies are needed to clarify the role of p16 IHC and how it can be optimized for the detection of cervical precancer, particularly in HPV-vaccinated populations where types other than HPV 16 are relatively more important.

Identification of oropharyngeal squamous cell carcinomas with active HPV16 involvement by immunohistochemical analysis of the retinoblastoma protein pathway.

Viral oncogene RNA expression is regarded as reliable biomarker to identify oropharyngeal squamous cell carcinomas (OPSCC) with active HPV16 involvement. This study addressed whether the expression profile of the cellular proteins p16(INK4a), pRb, Cyclin D1 and p53 provide surrogate marker combinations that identify OPSCC with active HPV16 in situations where only formalin-fixed biopsies are available. Protein expression was analyzed by immunohistochemistry on tissue microarrays created from 188 OPSCC of which the HPV16 DNA and RNA status had been established previously from snap-frozen biopsies. Associations of single markers and of marker combinations with HPV16 DNA, viral RNA expression patterns and overall survival as primary end point were evaluated by statistical analysis. Most tumors with active HPV16 involvement (RNA(+) group; n = 40) showed a specific protein pattern, that is, high p16(INK4a) (80%), low pRb (85%), low Cyclin D1 (95%) and normal p53 (73%). This pattern was significantly different from the pattern observed in HPV DNA-negative tumors (HPV(-) group) and in HPV16 DNA-positive tumors lacking viral RNA expression patterns (RNA(-) group). The combination of high p16(INK4a) and low pRb levels was distinctly superior to p16(INK4a) alone; it was strongly associated with RNA(+) tumors (OR 41.4, 95%CI 10.7-162.5), with improved survival (HR 0.37, 95%CI 0.2-0.8) and had best predictive values (78% sensitivity, 93% specificity, 78% PPV, 93% NPV). In conclusion, if only formalin-fixed biopsy material is available, the marker combination high p16(INK4a) /low pRb is well suited to identify OPSCC with biologically active HPV16 which represent a distinct OPSCC entity with improved prognosis.

Submaxillary gland androgen-regulated protein 3A expression is an unfavorable risk factor for the survival of oropharyngeal squamous cell carcinoma patients after surgery.

Recently, increased expression of the submaxillary gland androgen-regulated protein 3A (SMR3A) was found in recurrent tumors of an orthotopic floor-of-mouth mouse tumor model after surgery. However, SMR3A expression in the pathogenesis of human malignancy and its correlation with the clinical outcome have not been addressed so far. We analyzed tissue microarrays with specimens from oropharyngeal squamous cell carcinoma (OPSCC) patients (n = 157) by immunohistochemistry and compared SMR3A expression with clinical and pathological features by statistical analysis. Strong SMR3A expression was found in almost 36 % of all primary OPSCCs. Although, SMR3A protein levels were not associated with any clinical or histopathological feature tested, univariate Kaplan-Meier analysis revealed a significant correlation between high SMR3A protein expression and poor progression-free (p = 0.02) and overall survival (p = 0.03). Furthermore, high SMR3A expression was an independent marker for poor clinical outcome [HR (SMR3A(high) vs. SMR3(low)) = 2.32; 95 % CI = 1.03-5.23] concerning overall survival in a multivariate analysis of OPSCC patients with surgery as primary therapy (n = 100). Our data demonstrate for the first time increased SMR3A protein expression in the pathogenesis of OPSCC, which serves as an unfavorable risk factor for patient survival.

CNN stability training improves robustness to scanner and IHC-based image variability for epithelium segmentation in cervical histology.

Background: In digital pathology, image properties such as color, brightness, contrast and blurriness may vary based on the scanner and sample preparation. Convolutional Neural Networks (CNNs) are sensitive to these variations and may underperform on images from a different domain than the one used for training. Robustness to these image property variations is required to enable the use of deep learning in clinical practice and large scale clinical research. Aims: CNN Stability Training (CST) is proposed and evaluated as a method to increase CNN robustness to scanner and Immunohistochemistry (IHC)-based image variability. Methods: CST was applied to segment epithelium in immunohistological cervical Whole Slide Images (WSIs). CST randomly distorts input tiles and factors the difference between the CNN prediction for the original and distorted inputs within the loss function. CNNs were trained using 114 p16-stained WSIs from the same scanner, and evaluated on 6 WSI test sets, each with 23 to 24 WSIs of the same tissue but different scanner/IHC combinations. Relative robustness (rAUC) was measured as the difference between the AUC on the training domain test set (i.e., baseline test set) and the remaining test sets. Results: Across all test sets, The AUC of CST models outperformed "No CST" models (AUC: 0.940-0.989 vs. 0.905-0.986, p < 1e - 8), and obtained an improved robustness (rAUC: [-0.038, -0.003] vs. [-0.081, -0.002]). At a WSI level, CST models showed an increase in performance in 124 of the 142 WSIs. CST models also outperformed models trained with random on-the-fly data augmentation (DA) in all test sets ([0.002, 0.021], p < 1e-6). Conclusion: CST offers a path to improve CNN performance without the need for more data and allows customizing distortions to specific use cases. A python implementation of CST is publicly available at https://github.com/TIGACenter/CST_v1.

Expression analysis of aldehyde dehydrogenase 1A1 (ALDH1A1) in colon and rectal cancer in association with prognosis and response to chemotherapy.

BACKGROUND: Aldehyde dehydrogenase 1A1 (ALDH1A1) has been described as a cancer stem cell marker and as a regulator of cellular chemoresistance. Therefore, ALDH1A1 has been suggested as potential biomarker to stratify patients into different risk categories for a "personalized" therapy approach. We have investigated the prognostic role of ALDH1A1 in primary colorectal cancer and its value in predicting response to chemotherapy in metastatic colorectal cancer. METHODS: Immunostaining against ALDH1A1 was performed on a paraffin-embedded tissue microarray including 659 primary colon cancer samples and 338 rectal cancer samples. Likewise, tissue of 44 palliatively resected colorectal liver metastases on whole-mount tissue slides was immunostained against ALDH1A1. Cytoplasmic, nuclear, and stromal expression of ALDH1A1 was assessed and merged with histopathological and clinical data. RESULTS: Univariate and multivariate analysis revealed that cytoplasmic and stromal expression of ALDH1A1 is not significantly associated with prognosis either in colon or in rectal cancer. Furthermore, cytoplasmic expression of ALDH1A1 does not predict response to palliative chemotherapy in patients with metastatic diseases. Intriguingly, as a novel finding, nuclear expression of ALDH1A1 was observed in a small subgroup of patients with colon cancer and rectal cancer. In colon cancer, nuclear expression was significantly associated with shortened overall survival by univariate and multivariate analysis. CONCLUSIONS: Immunohistochemical expression analysis of ALDH1A1 in colon cancer is useful for the detection of nuclear expression in a small subpopulation of patients and is associated with shorter survival. Cytoplasmic expression fails to be of clinical relevance as prognostic or predictive marker in colorectal cancer.

A comparison of dacron versus Flocked nylon swabs for anal cytology specimen collection.

OBJECTIVES: We compared the performance of commonly used Dacron versus flocked nylon swabs for anal cytology. STUDY DESIGN: From 23 HIV-positive men screened at Kaiser Permanente San Francisco (San Francisco, Calif., USA), 2 anal specimens were collected, 1 with each swab in random order, and placed into liquid cytology medium. Specimens were tested for cellularity by quantifying a genomic DNA (erv-3). The number of cells was assessed from prepared slides by automated image analysis. Performance was compared between swabs using 2-sample t tests and standard crossover trial analysis methods accounting for period effect. RESULTS: Flocked swabs collected slightly more erv-3 cells than Dacron for the first sample although not significantly (p = 0.18) and a similar number of erv-3 cells for the second sample (p = 0.85). Flocked swabs collected slightly more cells per slide than the Dacron swabs at both time periods although this was only significant in the second time period (p = 0.42 and 0.03 for first and second periods, respectively). In crossover trial analysis, flocked swabs outperformed Dacron for cell count per slide based on slide imaging (p = 0.03), but Dacron and flocked swabs performed similarly based on erv-3 quantification (p = 0.14). CONCLUSIONS: Further studies should determine whether flocked swabs increase the representation of diagnostically important cells compared to Dacron.

Therapeutic Effects of Inhibition of Sphingosine-1-Phosphate Signaling in HIF-2α Inhibitor-Resistant Clear Cell Renal Cell Carcinoma.

Specific inhibitors of HIF-2α have recently been approved for the treatment of ccRCC in VHL disease patients and have shown encouraging results in clinical trials for metastatic sporadic ccRCC. However, not all patients respond to therapy and pre-clinical and clinical studies indicate that intrinsic as well as acquired resistance mechanisms to HIF-2α inhibitors are likely to represent upcoming clinical challenges. It would be desirable to have additional therapeutic options for the treatment of HIF-2α inhibitor resistant ccRCCs. Here we investigated the effects on tumor growth and on the tumor microenvironment of three different direct and indirect HIF-α inhibitors, namely the HIF-2α-specific inhibitor PT2399, the dual HIF-1α/HIF-2α inhibitor Acriflavine, and the S1P signaling pathway inhibitor FTY720, in the autochthonous Vhl/Trp53/Rb1 mutant ccRCC mouse model and validated these findings in human ccRCC cell culture models. We show that FTY720 and Acriflavine exhibit therapeutic activity in several different settings of HIF-2α inhibitor resistance. We also identify that HIF-2α inhibition strongly suppresses T cell activation in ccRCC. These findings suggest prioritization of sphingosine pathway inhibitors for clinical testing in ccRCC patients and also suggest that HIF-2α inhibitors may inhibit anti-tumor immunity and might therefore be contraindicated for combination therapies with immune checkpoint inhibitors.

Accuracy and Efficiency of Deep-Learning-Based Automation of Dual Stain Cytology in Cervical Cancer Screening.

BACKGROUND: With the advent of primary human papillomavirus testing followed by cytology for cervical cancer screening, visual interpretation of cytology slides remains the last subjective analysis step and suffers from low sensitivity and reproducibility. METHODS: We developed a cloud-based whole-slide imaging platform with a deep-learning classifier for p16/Ki-67 dual-stained (DS) slides trained on biopsy-based gold standards. We compared it with conventional Pap and manual DS in 3 epidemiological studies of cervical and anal precancers from Kaiser Permanente Northern California and the University of Oklahoma comprising 4253 patients. All statistical tests were 2-sided. RESULTS: In independent validation at Kaiser Permanente Northern California, artificial intelligence (AI)-based DS had lower positivity than cytology (P < .001) and manual DS (P < .001) with equal sensitivity and substantially higher specificity compared with both Pap (P < .001) and manual DS (P < .001), respectively. Compared with Pap, AI-based DS reduced referral to colposcopy by one-third (41.9% vs 60.1%, P < .001). At a higher cutoff, AI-based DS had similar performance to high-grade squamous intraepithelial lesions cytology, indicating a risk high enough to allow for immediate treatment. The classifier was robust, showing comparable performance in 2 cytology systems and in anal cytology. CONCLUSIONS: Automated DS evaluation removes the remaining subjective component from cervical cancer screening and delivers consistent quality for providers and patients. Moving from Pap to automated DS substantially reduces the number of colposcopies and also achieves excellent performance in a simulated fully vaccinated population. Through cloud-based implementation, this approach is globally accessible. Our results demonstrate that AI not only provides automation and objectivity but also delivers a substantial benefit for women by reduction of unnecessary colposcopies.

Robust gridding of TMAs after whole-slide imaging using template matching.

Tissue microarrays (TMAs) represent an important approach for the high-throughput cellular analysis of large numbers of tissue samples on one single slide in research related to diagnostics and oncology. Whole-slide imaging now enables full scanning and subsequent image analysis of such TMAs. In contrast to automatically spotted RNA microarrays, TMAs are fabricated manually and mechanically by arranging hundreds of tissue cores in a single paraffin block. This procedure frequently results in quality problems severely hampering the later automatic image analysis of TMAs after whole-slide imaging. We therefore set out to (a) determine the extent of these quality issues in exemplary TMAs and (b) to develop a robust gridding method to identify the logical position coordinates of each TMA core on a virtual TMA slide. We present the first robust method identifying these coordinates by shifting a template grid over all cores of the TMA (template matching) and thereby measuring in how far the grid matches a predefined list of cores on the virtual TMA Slide. Analysis of 20 TMAs from Breast Cancer as well as 40 Head-and-Neck Cancer showed that frequent TMA layout issues comprise low staining, debris, core displacement, nonuniform background, missing cores, and rotated subarrays. On this highly demanding test comprising chromogen as well as fluorescence stained TMAs, our template matching method achieved an excellent position analysis. Of 8900 cores, 8864 (99.59%) were assigned properly. In all 60 slides of the test set, no localization error occurred. The automatic grid analysis of TMAs after whole-slide imaging is highly demanding and requires dedicated algorithms. We demonstrate such a method and evaluate its performance. © 2010 International Society for Advancement of Cytometry.

HIF-1α and HIF-2α differently regulate tumour development and inflammation of clear cell renal cell carcinoma in mice.

Mutational inactivation of VHL is the earliest genetic event in the majority of clear cell renal cell carcinomas (ccRCC), leading to accumulation of the HIF-1α and HIF-2α transcription factors. While correlative studies of human ccRCC and functional studies using human ccRCC cell lines have implicated HIF-1α as an inhibitor and HIF-2α as a promoter of aggressive tumour behaviours, their roles in tumour onset have not been functionally addressed. Herein we show using an autochthonous ccRCC model that Hif1a is essential for tumour formation whereas Hif2a deletion has only minor effects on tumour initiation and growth. Both HIF-1α and HIF-2α are required for the clear cell phenotype. Transcriptomic and proteomic analyses reveal that HIF-1α regulates glycolysis while HIF-2α regulates genes associated with lipoprotein metabolism, ribosome biogenesis and E2F and MYC transcriptional activities. HIF-2α-deficient tumours are characterised by increased antigen presentation, interferon signalling and CD8+ T cell infiltration and activation. Single copy loss of HIF1A or high levels of HIF2A mRNA expression correlate with altered immune microenvironments in human ccRCC. These studies reveal an oncogenic role of HIF-1α in ccRCC initiation and suggest that alterations in the balance of HIF-1α and HIF-2α activities can affect different aspects of ccRCC biology and disease aggressiveness.

Tumor Cell Load and Heterogeneity Estimation From Diffusion-Weighted MRI Calibrated With Histological Data: an Example From Lung Cancer.

Diffusion-weighted magnetic resonance imaging (DWI) is a key non-invasive imaging technique for cancer diagnosis and tumor treatment assessment, reflecting Brownian movement of water molecules in tissues. Since densely packed cells restrict molecule mobility, tumor tissues produce usually higher signal (a.k.a. less attenuated signal) on isotropic maps compared with normal tissues. However, no general quantitative relation between DWI data and the cell density has been established. In order to link low-resolution clinical cross-sectional data with high-resolution histological information, we developed an image processing and analysis chain, which was used to study the correlation between the diffusion coefficient (D value) estimated from DWI and tumor cellularity from serial histological slides of a resected non-small cell lung cancer tumor. Color deconvolution followed by cell nuclei segmentation was performed on digitized histological images to determine local and cell-type specific 2d (two-dimensional) densities. From these, the 3d cell density was inferred by a model-based sampling technique, which is necessary for the calculation of local and global 3d tumor cell count. Next, DWI sequence information was overlaid with high-resolution CT data and the resected histology using prominent anatomical hallmarks for co-registration of histology tissue blocks and non-invasive imaging modalities' data. The integration of cell numbers information and DWI data derived from different tumor areas revealed a clear negative correlation between cell density and D value. Importantly, spatial tumor cell density can be calculated based on DWI data. In summary, our results demonstrate that tumor cell count and heterogeneity can be predicted from DWI data, which may open new opportunities for personalized diagnosis and therapy optimization.

High numbers of PDCD1 (PD-1)-positive T cells and B2M mutations in microsatellite-unstable colorectal cancer.

DNA mismatch repair (MMR)-deficient cancers accumulate high numbers of coding microsatellite mutations, which lead to the generation of highly immunogenic frameshift peptide (FSP) neoantigens. MMR-deficient cells can grow out to clinically manifest cancers either if they evade immune cell attack or if local T-cells get exhausted. Therefore, a subset of MSI cancer patients responds particularly well to treatment with immune checkpoint inhibitors. We analyzed whether immune evasion in MMR-deficient cancer mediated by loss of HLA class I or II antigens is related to local immune cell activation status. Microsatellites located in Beta2-microglobulin (B2M) and the HLA class II-regulatory genes RFX5 and CIITA were analyzed for mutations in MMR-deficient colorectal cancers (n = 53). The results were related to CD3-positive and PDCD1 (PD-1)-positive T-cell infiltration. PDCD1 (PD-1)-positive T-cell counts were significantly higher in B2M-mutant compared to B2M-wild type tumors (median: 22.2 cells per 0.25 mm2 vs. 2.0 cells per 0.25 mm2, Wilcoxon test p = 0.002). Increasing PDCD1 (PD-1)-positive T-cell infiltration was significantly related to an increased likelihood of B2M mutations (OR = 1.81). HLA class II antigen expression status was significantly associated with enhanced overall T-cell infiltration, but not related to PDCD1 (PD-1)-positive T-cells. These results suggest that immune evasion mediated by B2M mutation-induced loss of HLA class I antigen expression predominantly occurs in an environment of activated PDCD1 (PD-1)-positive T cell infiltration. If B2M mutations interfere with anti-PDCD1 (PD-1)/CD274 (PD-L1) therapy success, we predict that resistance towards anti-PDCD1 (PD-1) therapy may - counterintuitively - be particularly common in patients with MMR-deficient cancers that show high PDCD1 (PD-1)-positive T cell infiltration.

Spatial intratumoral heterogeneity of proliferation in immunohistochemical images of solid tumors.

PURPOSE: The interactions of neoplastic cells with each other and the microenvironment are complex. To understand intratumoral heterogeneity, subtle differences should be quantified. Main factors contributing to heterogeneity include the gradient ischemic level within neoplasms, action of microenvironment, mechanisms of intercellular transfer of genetic information, and differential mechanisms of modifications of genetic material/proteins. This may reflect on the expression of biomarkers in the context of prognosis/stratification. Hence, a rigorous approach for assessing the spatial intratumoral heterogeneity of histological biomarker expression with accuracy and reproducibility is required, since patterns in immunohistochemical images can be challenging to identify and describe. METHODS: A quantitative method that is useful for characterizing complex irregular structures is lacunarity; it is a multiscale technique that exhaustively samples the image, while the decay of its index as a function of window size follows characteristic patterns for different spatial arrangements. In histological images, lacunarity provides a useful measure for the spatial organization of a biomarker when a sampling scheme is employed and relevant features are computed. The proposed approach quantifies the segmented proliferative cells and not the textural content of the histological slide, thus providing a more realistic measure of heterogeneity within the sample space of the tumor region. The aim is to investigate in whole sections of primary pancreatic neuroendocrine neoplasms (pNENs), using whole-slide imaging and image analysis, the spatial intratumoral heterogeneity of Ki-67 immunostains. Unsupervised learning is employed to verify that the approach can partition the tissue sections according to distributional heterogeneity. RESULTS: The architectural complexity of histological images has shown that single measurements are often insufficient. Inhomogeneity of distribution depends not only on percentage content of proliferation phase but also on how the phase fills the space. Lacunarity curves demonstrate variations in the sampled image sections. Since the spatial distribution of proliferation in each case is different, the width of the curves changes too. Image sections that have smaller numerical variations in the computed features correspond to neoplasms with spatially homogeneous proliferation, while larger variations correspond to cases where proliferation shows various degrees of clumping. Grade 1 (uniform/nonuniform: 74%/26%) and grade 3 (uniform: 100%) pNENs demonstrate a more homogeneous proliferation with grade 1 neoplasms being more variant, while grade 2 tumor regions render a more diverse landscape (50%/50%). Hence, some cases show an increased degree of spatial heterogeneity comparing to others with similar grade. Whether this is a sign of different tumor biology and an association with a more benign/malignant clinical course needs to be investigated further. The extent and range of spatial heterogeneity has the potential to be evaluated as a prognostic marker. CONCLUSIONS: The association with tumor grade as well as the rationale that the methodology reflects true tumor architecture supports the technical soundness of the method. This reflects a general approach which is relevant to other solid tumors and biomarkers. Drawing upon the merits of computational biomedicine, the approach uncovers salient features for use in future studies of clinical relevance.