Negative Dysplasia Research Articles

Deep learning for automatic diagnosis of gastric dysplasia using whole-slide histopathology images in endoscopic specimens.

Distinguishing gastric epithelial regeneration change from dysplasia and histopathological diagnosis of dysplasia is subject to interobserver disagreement in endoscopic specimens. In this study, we developed a method to distinguish gastric epithelial regeneration change from dysplasia and further subclassify dysplasia. Meanwhile, optimized the cross-hospital diagnosis using domain adaption (DA). 897 whole slide images (WSIs) of endoscopic specimens from two hospitals were divided into training, internal validation, and external validation cohorts. We developed a deep learning (DL) with DA (DLDA) model to classify gastric dysplasia and epithelial regeneration change into three categories: negative for dysplasia (NFD), low-grade dysplasia (LGD), and high-grade dysplasia (HGD)/intramucosal invasion neoplasia (IMN). The diagnosis based on the DLDA model was compared to 12 pathologists using 100 gastric biopsy cases. In the internal validation cohort, the diagnostic performance measured by the macro-averaged area under the receiver operating characteristic curve (AUC) was 0.97. In the independent external validation cohort, our DLDA models increased macro-averaged AUC from 0.67 to 0.82. In terms of the NFD and HGD cases, our model's diagnostic sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were significantly higher than junior and senior pathologists. Our model's diagnostic sensitivity, NPV, was higher than specialist pathologists. We demonstrated that our DLDA model could distinguish gastric epithelial regeneration change from dysplasia and further subclassify dysplasia in endoscopic specimens. Meanwhile, achieved significant improvement of diagnosis cross-hospital.

Role of DNA Flow Cytometry in the Diagnosis of Malignancy in Bile Duct Biopsies Using Paraffin-Embedded Tissue.

Histopathologic evaluation of bile biopsies for biliary strictures is frequently challenging and is affected by interobserver disagreement. Reliable ancillary tests that can help differentiate benign from malignant are not available. This study aimed to evaluate whether DNA content abnormalities detected by flow cytometry on formalin-fixed, paraffin-embedded (FFPE) tissue can help differentiate benign/reactive, dysplastic from malignant cell populations in bile duct biopsies. We performed DNA flow cytometry on 30 FFPE bile duct biopsies in 5 well-defined diagnostic categories: (1) negative for dysplasia (NED), (2) low-grade dysplasia (LGD), (3) high-grade dysplasia (HGD), (4) carcinoma (CA), and (5) indefinite for dysplasia (IND). Abnormal DNA content was detected in 0 NED, 5 LGD (62.5%), 2 HGD (33.3%), 3 CA (60%), and 4 IND (80%) samples. As a diagnostic marker, the estimated sensitivity, specificity, positive predictive value, and negative predictive value were 63%, 100%, 100%, and 50%, respectively, for diagnosing HGD or CA. DNA flow cytometry analysis is a useful ancillary test for the interpretation of bile duct biopsies. DNA content abnormalities, when correlated with histologic findings, will not only help confirm the morphologic impression but also identify patients who are at a higher risk of developing malignancy.

A Prospective Validation and Observer Performance Study of a Deep Learning Algorithm for Pathologic Diagnosis of Gastric Tumors in Endoscopic Biopsies.

Gastric cancer remains the leading cause of cancer-related deaths in Northeast Asia. Population-based endoscopic screenings in the region have yielded successful results in early detection of gastric tumors. Endoscopic screening rates are continuously increasing, and there is a need for an automatic computerized diagnostic system to reduce the diagnostic burden. In this study, we developed an algorithm to classify gastric epithelial tumors automatically and assessed its performance in a large series of gastric biopsies and its benefits as an assistance tool. Using 2,434 whole-slide images, we developed an algorithm based on convolutional neural networks to classify a gastric biopsy image into one of three categories: negative for dysplasia (NFD), tubular adenoma, or carcinoma. The performance of the algorithm was evaluated by using 7,440 biopsy specimens collected prospectively. The impact of algorithm-assisted diagnosis was assessed by six pathologists using 150 gastric biopsy cases. Diagnostic performance evaluated by the AUROC curve in the prospective study was 0.9790 for two-tier classification: negative (NFD) versus positive (all cases except NFD). When limited to epithelial tumors, the sensitivity and specificity were 1.000 and 0.9749. Algorithm-assisted digital image viewer (DV) resulted in 47% reduction in review time per image compared with DV only and 58% decrease to microscopy. Our algorithm has demonstrated high accuracy in classifying epithelial tumors and its benefits as an assistance tool, which can serve as a potential screening aid system in diagnosing gastric biopsy specimens.

Interobserver agreement in pathologic evaluation of bile duct biopsies

Intraductal biopsy is commonly used for preoperative evaluation of the etiology of biliary strictures. Interpretation of intraductal biopsies is frequently challenging. The diagnosis often suffers from interobserver disagreement, which has not been studied in the literature. We sought to assess interobserver concordance in the interpretation of intraductal biopsies. Eighty-five biopsies were retrieved, falling into five diagnostic categories: negative for dysplasia (NED), indefinite for dysplasia (IND), low-grade dysplasia (LGD), high-grade dysplasia (HGD), and carcinoma (CA). Eight gastrointestinal pathologists blindly reviewed all the slides. Agreement among pathologists was analyzed using Fleiss κ and weighted concordance coefficient S∗. A face-to-face consensus/training session was held to discuss the classification criteria, followed by a second round review. The overall interobserver agreement was fair in the first round review (κ=0.39; S∗=0.56) and improved to moderate in the second round review (κ=0.48; S∗=0.69). The agreement before and after consensus meeting was substantial to nearly perfect for CA (κ=0.65, S∗=0.83; and κ=0.80, S∗=0.91), fair for HGD (κ=0.28, S∗=0.69; and κ=0.40, S∗=0.63), and moderate for NED (κ=0.47, S∗=0.50; and κ=0.47, S∗=0.53). Agreement improved from fair to moderate for LGD (κ=0.36, S∗=0.61; and κ=0.49, S∗=0.71) and slight to fair for IND (κ=0.16, S∗=0.51; and κ=0.33, S∗=0.50). Compared with Hollande's fixed specimens, the agreement was higher in almost all diagnostic categories in formalin-fixed biopsies. Overall, interobserver concordance was improved after a consensus/training session. Interobserver reproducibility was high at the end of the diagnostic spectrum (CA) but fair to moderate for other diagnostic categories.

Diagnosis and risk stratification of Barrett’s dysplasia by flow cytometric DNA analysis of paraffin-embedded tissue

ObjectiveThe diagnosis of dysplasia in Barrett’s oesophagus (BO) can be challenging, and reliable ancillary techniques are not available. This study examines if DNA content abnormality detected by flow cytometry can...

Immunohistochemistry in the diagnosis of dysplasia in chronic inflammatory bowel disease colorectal polyps

Background and study aimsDevelopment of cancer is the most significant complication in inflammatory bowel disease (IBD). Distinguishing true dysplasia from reactive atypia in polyps is difficult, leading sometimes to the unsatisfactory diagnosis of “indefinite for dysplasia”. Therefore, there is a need for the development of markers that can help improve diagnosis.We evaluated the diagnostic value of the expression of AMACR, Ki67 and p53 by immunohistochemistry in the diagnosis of dysplasia in polyps developed on IBD. Patients and methodsForty colorectal polyps in IBD were studied. These had been diagnosed over a period of 11years. Dysplasia was classified according to the Vienna Classification (version 2000). Immunohistochemistry was performed using anti-AMACR, anti-Ki67 and anti-p53 antibodies. ResultsPolyps were classified as follows: 21 negative for dysplasia (ND), 10 indefinite for dysplasia (IFD), 6 low-grade dysplasia (LGD), 1 high-grade dysplasia (HGD) and 2 adenocarcinomas (ACA).AMACR positivity was observed in all polyps with HGD and ACA, 5 of the 6 LGD polyps and 3 of the 10 IFD (p=0.007).p53 immunostaining showed nuclear staining in the basal part of the crypts in 8 of the 10 IFD lesions. In ACA and HGD polyps, p53 positivity was typically observed in all epithelial cell layers (p=0.004).ACA and HGD showed diffuse and scattered staining of Ki67 along the full length of the crypts. Five lesions with LGD had extension of Ki-67 positive cells up to and into the surface epithelium. Ki67 staining in all IFD lesions was restricted to the basal third of the crypt (p<0.001).By combining the three markers, a relationship with dysplasia was statistically significant (p<0.001). Sensitivity ranged from 66.7% to 88.9% and specificity from 71.4% to 100%.The positive predictive value (PPV) for detecting dysplasia using these different antibodies ranged from 66.7% to 100% and the negative predictive value (NPV) for excluding dysplasia ranged from 85.7% to 93.3%. ConclusionsThe high degree of sensitivity and specificity of AMACR, p53 and Ki67 for dysplasia in IBD suggests that these antibodies, when combined, may be useful to detect neoplastic epithelium in this condition.

Whole slide image cytometry: a novel method to detect abnormal DNA content in Barrett's esophagus

Barrett's esophagus (BE) is a precursor of esophageal adenocarcinoma (EAC). Both low-grade dysplasia (LGD) and high-grade dysplasia (HGD) are associated with an increased risk of progression to EAC. However, histological interpretation and grading of dysplasia (particularly LGD) is subjective and poorly reproducible. This study has combined whole slide imaging with DNA image cytometry to provide a novel method for the detection of abnormal DNA content through image analysis of tissue sections. A total of 20 cases were evaluated, including 8 negative for dysplasia (NFD), 6 LGD, and 6 HGD. Feulgen-stained esophageal sections were scanned in their entirety. Barrett's mucosa was interactively chosen for automatic nuclei segmentation where irrelevant cell types were ignored. The combined DNA content histogram for all nuclei within selected image regions was then obtained. In addition, three histogram measurements were computed, including xER-5C, 2cDI, and DNA-MG. Visual evaluation suggested the shape of DNA content histograms from NFD, LGD, and HGD cases exhibiting identifiable differences. The histogram measurements, xER-5C, 2cDI, and DNA-MG, were shown to be effective in differentiating metaplastic from dysplastic cases with statistical significance. Moreover, they also successfully separated NFD, LGD, and HGD patients with statistical significance. Whole slide image cytometry is a novel and effective method for the detection of abnormal DNA content in BE. Compared with histological review, it is more objective. Compared with flow cytometry and cytology-preparation image cytometry, it is low cost, simple to use, only requires a single 1 μm section, and facilitates selection of tissue and topographical correlation. Whole slide image cytometry can detect differences in DNA content between NFD, LGD, and HGD patients in this cross-sectional study. Abnormal DNA content detection by whole slide image cytometry is a promising biomarker of progression that could affect future diagnostics in BE.

Mo1896 Recognition of the Null Staining Pattern Increases the Utility of p53 IHC in Barrett's Esophagus

Background: Mutation and/or deletion of p53, a cell cycle regulatory gene, is a promising biomarker for predicting the risk of neoplastic progression in Barrett's esophagus (BE). Although overexpression of p53 by immunohistochemistry (IHC) is a useful surrogate for point mutations, complete absence of p53 protein by IHC in neoplastic cells (null pattern) has recently been shown to be highly correlated with truncation and deletion mutations. This study was designed to evaluate p53 expression patterns in BE with and without dysplasia. Design: 2,817 biopsies with a diagnosis of BE with and without dysplasia that also had a p53 IHC stain performed between Jan 1, 2010 and Nov 12, 2012 were identified in the files of Miraca Life Sciences. This included 139 with high grade dysplasia (HGD), 274 with low grade dysplasia (LGD), 214 indefinite for dysplasia (IND) and 2190 negative for dysplasia (ND). IHC stains were classified as: wild type (WT, 1-15% nuclear staining throughout); point mutation pattern (PMP, foci with .50% nuclear staining); and null mutation pattern (NMP, foci with complete absence of nuclear staining). Results: Abnormal p53 IHC expression patterns were detected in 442/2,817 (15.7%) biopsies, including 372/442 (84.2%) with PMP, 57/442 (12.9%) with NMP, and 13/442 (2.9%) with both patterns in the same biopsy; 70 (15.4%) of the putative p53 mutations were NMP. The frequency of p53 mutation patterns increased with increasing grades of dysplasia (Figure 1). While both PMP and NMP were identified in cases with and (rarely) without dysplasia, NMP was disproportionately found in biopsies with HGD than LGD, IND, and ND (32/133, 24.1% vs. 38/309, 12.3%; p=0.0019). 133/139 (95.7%) of HGD had abnormal p53 IHC. Conclusions: p53 null staining pattern is readily identifiable, and accounts for 15.4% of all p53 mutations detectable by IHC in BE in our series. Recognition of the NMP increases the sensitivity of detection of p53 IHC mutation patterns in BE, and the vast majority of biopsies with HGD were found to have abnormal p53 IHC. These results suggest that, due to historical lack of recognition of the null pattern, the utility of p53 IHC as an adjunctive diagnostic and predictive marker in BE biopsies may be underestimated.

Should LSIL with ASC-H (LSIL-H) in cervical smears be an independent category? A study on SurePath™ specimens with review of literature

BackgroundCervical smears exhibiting unequivocal features of 'low grade squamous intraepithelial lesion' (LSIL) are occasionally also admixed with some cells suspicious for, but not diagnostic of, 'high grade squamous intraepithelial lesion' (HSIL). Only a few studies, mostly reported as abstracts, have evaluated this concurrence. In this study, we evaluate the current evidence that favors a distinct category for "LSIL, cannot exclude HSIL" (LSIL-H), and suggest a management algorithm based on combinations of current ASCCP guidelines for related interpretations.MethodsWe studied SurePath™ preparations of cervical specimens from various institutions during one year period. Cytohisto correlation was performed in cases with cervical biopsies submitted to our institution. The status of HPV DNA testing was also noted in some LSIL-H cases with biopsy results.ResultsOut of 77,979 cases 1,970 interpreted as LSIL (1,523), LSIL-H (146), 'atypical squamous cells, cannot exclude HSIL' (ASC-H) (109), and HSIL (192) were selected. Concurrent biopsy results were available in 40% (Total 792 cases: 557 LSIL, 88 LSIL-H, 38 ASCH, and 109 HSIL). Biopsy results were grouped into A. negative for dysplasia (ND), B. low grade (HPV, CIN1, CIN1 with HPV), and C. high grade (CIN 2 and above).The positive predictive values for various biopsy results in relation to initial cytopathologic interpretation were: a. LSIL: (557 cases): ND 32% (179), low grade- 58% (323), high grade- 10% (55); b. LSIL-H: (88 cases): ND 24% (21), low grade- 43% (38), high grade- 33% (29); c. ASCH: (38 cases): ND 32% (12), low grade- 37% (14), high grade- 31% (12); d. HSIL (109 cases): ND 5% (6), low grade 26% (28), high grade 69% (75). The patterns of cervical biopsy results in cases reported as LSIL-H were compared with that observed in cases with LSIL, ASC-H, and HSIL.94% (32 of 34) of LSIL-H were positive for high risk (HR) HPV, 1 was negative for HR HPV but positive for low risk (LR), and 1 LSIL-H was negative for HR and LR both.ConclusionLSIL-H overlapped with LSIL and ASC-H, but was distinct from HSIL. A management algorithm comparable to ASC-H and HSIL appears to be appropriate in LSIL-H cases.

Computerized morphometry as an aid in determining the grade of dysplasia and progression to adenocarcinoma in Barrett's esophagus

The aims of this study were to use computerized morphometry in order to differentiate between the degree of dysplasia and to predict progression to invasive adenocarcinoma in Barrett's esophagus (BE). Biopsies from 97 patients with BE graded by a consensus forum of expert gastrointestinal pathologists were available for morphometrical analysis. The study group included 36 biopsies negative for dysplasia (ND), none of which progressed to carcinoma; 16 indefinite for dysplasia (IND) and 21 low-grade dysplasia (LGD), of which three progressed in each group and 24 high-grade dysplasia (HGD), of which 15 progressed to invasive carcinoma. Computerized morphometry was used for measuring indices of size, shape, texture, symmetry and architectural distribution of the epithelial nuclei. Low-grade dysplasia was best differentiated from the ND group by nuclear pseudostratification (P=0.036), pleomorphism (P<0.01), and chromatin texture (margination, P<0.01) and from the HGD group by nuclear area (P<0.01), pleomorphism (P<0.01), chromatin texture (margination, P<0.01), symmetry (P<0.01), and orientation (P=0.027). These results were validated on a new set of cases (n=55) using a neural network model, resulting in an accuracy of 89% for differentiating between the ND and LGD groups and 86% for differentiating between the LGD and HGD groups. Within the HGD group, univariate significant predictors of the progression interval to carcinoma were: indices of nuclear texture (heterogeneity: P=0.0019, s.d.-OD: P=0.005) and orientation: P=0.022. Nuclear texture (heterogeneity) was the only independent predictor of progression (P=0.004, hazard=11.54) by Cox's multivariate test. This study proposes that computerized morphometry is a valid tool for determining the grade of dysplasia in BE. Moreover, histomorphometric quantification of nuclear texture is a powerful tool for predicting progression to invasive adenocarcinoma in patients with HGD.

Expression of the high mobility group proteins HMGI(Y) correlates with malignant progression in Barrett's metaplasia.

Expression of the high mobility group proteins HMGI(Y) has been shown to be a marker of malignancy in thyroid and pancreatic lesions and to correlate significantly with malignant progression in the colon. The aim of this study was to determine whether HMGI(Y) expression is associated with malignant progression in Barrett's metaplasia (BM). Immunoperoxidase staining for HMGI(Y) was performed on sections of formalin-fixed paraffin-embedded endoscopic esophageal biopsies from 42 patients with BM. These consisted of 19 biopsies negative for dysplasia (ND), 16 with low-grade dysplasia (LGD)/indeterminate for dysplasia (IND), and 7 with high-grade dysplasia (HGD)/adenocarcinoma (CA). The percentage of positive cells was recorded, and nuclear HMGI(Y) immunoreactivity in >10% of the cells was considered positive. Statistical analysis was performed using Fisher's exact test. Positive HMGI(Y) staining was detected in 2 of 19 (11%) cases ND, 5 of 16 (30%) LGD/IND cases, and 7 of 7 (100%) HGD/CA cases. Biopsies with HGD/CA were significantly more likely to be positive for HMGI(Y) than biopsies ND (P < 0.0001) or with LGD/IND (P = 0.0046). We conclude that HMGI(Y) expression is significantly associated with malignant progression in BM. Additional studies are needed to determine whether BM biopsies that are ND or LGD/IND and positive for HMGI(Y) are more likely to progress to adenocarcinoma.

Receptor activator of nuclear factor kappaB (RANK) is expressed as a late event during malignant progression in Barrett's metaplasia

Receptor activator of NF-kappaB (RANK) activation induces several survival signals including nuclear factor kappaB (NF-kappaB) and activation and upregulation of Bcl-xL. The aim of this work was to determine whether RANK is expressed in oesophageal adenocarcinoma (EA) and its precursor, Barrett's metaplasia (BM). Sections of formalin-fixed and paraffin-embedded tissue from 23 oesophagectomy specimens showing EA and variable degrees of dysplasia in BM were immunohistochemically stained for RANK. All 23 carcinomas (100%) showed strong RANK immunoreactivity in most cancer cells. In 8 cases with high-grade dysplasia (HGD), RANK immunoreactivity was weaker and detected in a smaller percentage of cells. RANK was not detected in any area of BM that was negative for dysplasia (ND) (17 cases), or showed low-grade dysplasia (LGD) (9 cases). These results indicate that RANK is expressed as a late event during malignant progression in BM. Studies are needed to determine whether interfering with RANK signalling affects the survival of EA cells.

Early progression stage of malignancy of uterine cervical dysplasia as revealed by immunohistochemical demonstration of increased DNA-instability.

The degree of DNA-instability as revealed by the immunohistochemical staining with anti-single-stranded DNA antibody after acid hydrolysis (DNA-instability test) was used as a marker of malignancy. This was applied to mild dysplasia (42 cases), moderate dysplasia (43 cases), severe dysplasia (27 cases), squamous cell carcinoma in situ (CIS) (21 cases), invasive squamous cell carcinoma (SCC) (31 cases) and normal (7 cases) human uterine cervix. The expression of tumour suppressor gene p53 and oncogene bcl-2 was detected immunohistochemically. Proliferative activity was evaluated by PCNA immumohistochemistry and the quantitative analysis of the number, mean area, the largest area and maximum shape irregularities of AgNOR in a nucleus were performed for all these cases. The distribution of numeric chromosomal aberrations of chromosome 17 was also investigated in some of these cases. The results showed that 31 SCC (100%), 21 CIS (100%), 21 severe dysplasia (77.77%), 28 moderate dysplasia (65.11%), and 14 mild dysplasia (33.33%) were positively stained by the DNA-instability test diffusely or sporadically, indicating their malignancy. Reflecting the malignant character, these cases showed a remarkable increase in the PCNA-index with the loss of polarity of PCNA positive cell distribution and also an increase in number, mean and largest sizes and maximum shape irregularity of AgNOR dots. The mean chromosome index for chromosome 17, p53 and bcl-2 immunostaining positivity were also found to be significantly increased in moderate and severe dysplasia and in cancerous cases in comparison to normal and mild dysplasia cases. Moreover, the DNA-instability-test positive dysplasia cases showed statistically significant increased values of PCNA-index, AgNOR parameters, mean chromosome index, p53 and bcl-2 expression in comparison to those of DNA-instability-test negative dysplasia cases. In conclusion, some mild dysplasia (33.33%) and most of the moderate (65.11%) and severe dysplasia (77.77%) were regarded as malignant in nature, existing at an early stage of progression of malignancy.