Advances in Digital Pathology: From Artificial Intelligence to Label-Free Imaging

BACKGROUND

Digital pathology, in its primary meaning, describes the utilization of computer screens to view scanned histology slides. Digitized tissue sections can be easily shared for a second opinion. In addition, it allows tissue image analysis using specialized software to identify and measure events previously observed by a human observer. These tissue-based readouts were highly reproducible and precise. Digital pathology has developed over the years through new technologies. Currently, the most discussed development is the application of artificial intelligence to automatically analyze tissue images. However, even new label-free imaging technologies are being developed to allow imaging of tissues by means of their molecular composition.

SUMMARY

This review provides a summary of the current state-of-the-art and future digital pathologies. Developments in the last few years have been presented and discussed. In particular, the review provides an outlook on interesting new technologies (e.g., infrared imaging), which would allow for deeper understanding and analysis of tissue thin sections beyond conventional histopathology.

KEY MESSAGES

In digital pathology, mathematical methods are used to analyze images and draw conclusions about diseases and their progression. New innovative methods and techniques (e.g., label-free infrared imaging) will bring significant changes in the field in the coming years.

Prognostic Role of FGFR Alterations and FGFR mRNA Expression in Metastatic Urothelial Cancer Undergoing Checkpoint Inhibitor Therapy

OBJECTIVE

To examine the disease-specific survival(DSS) after checkpoint inhibitor(CPI) therapy based on FGFR alterations and FGFR mRNA expression levels in patients with metastatic urothelial cancer(mUCa) within a multi-center cohort.

METHODS

Within a cohort of 72 patients with mUCa from five academic centers in Germany FGFR alterations, as well as FGFR1-4 mRNA expression levels in tumor samples from the primary tumor or metastatic sites. Spearman rank correlations, logistic regression, as well as Kaplan-Meier survival analyses and univariate Cox proportional hazards regression models were employed to examine the impact of different FGFR patterns on the DSS after CPI treatment.

RESULTS

FGFR3 mutations or gene fusions (gene alterations) were detected in 16.9% of all samples. Patients with or without FGFR3 gene alterations did not show different oncological outcomes undergoing CPI treatment. Low expression of FGFR2 mRNA alone, as well as the combination of either low FGFR2mRNA expression and FGFR3 gene alteration or high FGFR3mRNA expression (P = 0.027), identified a subgroup of patients with unfavorable outcomes, comprising 40% of the total cohort. This trend was also observed in univariate Cox proportional hazards regression analysis(FGFR3 gene alteration: Hazard ratio(HR) 5.33, 95%Confidence interval(CI)1.76-15.0, P = 0.004; FGFR3mRNA expression:HR 3.04, 95%CI 1.40-7.13, P = 0.005).

CONCLUSION

Assessment of FGFR mRNA expression identified a high-risk subgroup of patients with mUCa. These patients showing overexpression of FGFR3 mRNA were found to have unfavorable DSS after CPI treatment. Using this approach may be suitable for identifying a patient population with poor response to CPI treatment, which may benefit from early FGFR inhibition.

Identification of patients at risk for biochemical recurrence after radical prostatectomy with intra-operative frozen section

OBJECTIVE

To identify patients at risk for biochemical recurrence (BCR) of prostate cancer (PCa) after radical prostatectomy (RP) with intra-operative whole-mount frozen section (FS) of the prostate.

MATERIAL AND METHODS

We examined differences in BCR between patients with initial negative surgical margins at FS, patients with final negative surgical margins with initial positive margins at FS without residual PCa after secondary tumour resection, and patients with final negative surgical margins with initially positive margins at FS with residual PCa in the secondary tumour resection specimen. Institutional data of 883 consecutive patients undergoing RP were collected. Intra-operative whole-mount FS was routinely used to check for margin status and, if necessary, to resect more periprostatic tissue in order to achieve negative margins. Patients with lymph node-positive disease or final positive surgical margins were excluded from the analysis. Kaplan-Meier curves and multivariable Cox proportional hazards regression analyses adjusting for clinical covariates were employed to examine differences in biochemical recurrence-free survival (BRFS) according to the resection status mentioned above.

RESULTS

The median follow-up was 22.4 months. The 1- and 2-year BRFS rates in patients with (81.0% and 72.9%, respectively; P = 0.001) and without residual PCa (90.3% and 82.3%, respectively; P = 0.033) after secondary tumour resection were significantly lower compared to patients with initial R0 status (93.4% and 90.9%, respectively). On multivariable Cox regression only residual PCa in the secondary tumour resection was associated with a higher risk of BCR compared to initial R0 status (hazard ratio 1.99, 95% confidence interval 1.01-3.92; P = 0.046).

CONCLUSION

Despite being classified as having a negative surgical margin, patients with residual PCa in the secondary tumour resection specimen face a high risk of BCR. These findings warrant closer post-RP surveillance of this particular subgroup. Further research of this high-risk subset of patients should focus on examining whether these patients benefit from early salvage therapy and how resection status impacts oncological outcomes in the changing landscape of PCa treatment.

Label-free, automated classification of microsatellite status in colorectal cancer by infrared imaging

Challenging histopathological diagnostics in cancer include microsatellite instability-high (MSI-H) colorectal cancer (CRC), which occurs in 15% of early-stage CRC and is caused by a deficiency in the mismatch repair system. The diagnosis of MSI-H cannot be reliably achieved by visual inspection of a hematoxylin and eosin stained thin section alone, but additionally requires subsequent molecular analysis. Time- and sample-intensive immunohistochemistry with subsequent fragment length analysis is used. The aim of the presented feasibility study is to test the ability of quantum cascade laser (QCL)-based infrared (IR) imaging as an alternative diagnostic tool for MSI-H in CRC. We analyzed samples from 100 patients with sporadic CRC UICC stage II and III. Forty samples were used to develop the random forest classifier and 60 samples to verify the results on an independent blinded dataset. Specifically, 100% sensitivity and 93% specificity were achieved based on the independent 30 MSI-H- and 30 microsatellite stable (MSS)-patient validation cohort. This showed that QCL-based IR imaging is able to distinguish between MSI-H and MSS for sporadic CRC - a question that goes beyond morphological features - based on the use of spatially resolved infrared spectra used as biomolecular fingerprints.

Integrated Fourier Transform Infrared Imaging and Proteomics for Identification of a Candidate Histochemical Biomarker in Bladder Cancer

Histopathological differentiation between severe urocystitis with reactive urothelial atypia and carcinoma in situ (CIS) can be difficult, particularly after a treatment that deliberately induces an inflammatory reaction, such as intravesical instillation of Bacillus Calmette-Guèrin. However, precise grading in bladder cancer is critical for therapeutic decision making and thus requires reliable immunohistochemical biomarkers. Herein, an exemplary potential biomarker in bladder cancer was identified by the novel approach of Fourier transform infrared imaging for label-free tissue annotation of tissue thin sections. Identified regions of interest are collected by laser microdissection to provide homogeneous samples for liquid chromatography-tandem mass spectrometry-based proteomic analysis. This approach afforded label-free spatial classification with a high accuracy and without interobserver variability, along with the molecular resolution of the proteomic analysis. Cystitis and invasive high-grade urothelial carcinoma samples were analyzed. Three candidate biomarkers were identified and verified by immunohistochemistry in a small cohort, including low-grade urothelial carcinoma samples. The best-performing candidate AHNAK2 was further evaluated in a much larger independent verification cohort that also included CIS samples. Reactive urothelial atypia and CIS were distinguishable on the basis of the expression of this newly identified and verified immunohistochemical biomarker, with a sensitivity of 97% and a specificity of 69%. AHNAK2 can differentiate between reactive urothelial atypia in the setting of an acute or chronic cystitis and nonmuscle invasive-type CIS.

[Tumor grading of the hepatobiliary system]

Tumors of the liver, intrahepatic and extrahepatic bile ducts as well as the gallbladder are very heterogeneous and show different biological behavior. The 4‑stage (i.e. well, moderately, poorly and undifferentiated) grading system for hepatocellular carcinoma proposed by the WHO takes tumor size and architecture as well as the extent of cell and nuclear pleomorphism into account. In addition, the WHO defines some special forms of hepatocellular carcinoma. For carcinomas of intrahepatic bile ducts the WHO provides a 3‑stage (well, moderately and poorly differentiated) grading system, which is based on architectural and cytological changes. At this localization there are also additional special histological forms that have to be dealt with outside the grading system described. The WHO proposes a 3‑stage (well, moderately and poorly differentiated) grading system for carcinomas of the extrahepatic bile ducts and the gallbladder, which considers the proportion of glands contained within the adenocarcinoma. Similar to cancers of the liver and intrahepatic bile ducts there are also numerous special histological forms, which are explained in this article.

Noninvasive Diagnosis of High-Grade Urothelial Carcinoma in Urine by Raman Spectral Imaging

The current gold standard for the diagnosis of bladder cancer is cystoscopy, which is invasive and painful for patients. Therefore, noninvasive urine cytology is usually used in the clinic as an adjunct to cystoscopy; however, it suffers from low sensitivity. Here, a novel noninvasive, label-free approach with high sensitivity for use with urine is presented. Coherent anti-Stokes Raman scattering imaging of urine sediments was used in the first step for fast preselection of urothelial cells, where high-grade urothelial cancer cells are characterized by a large nucleus-to-cytoplasm ratio. In the second step, Raman spectral imaging of urothelial cells was performed. A supervised classifier was implemented to automatically differentiate normal and cancerous urothelial cells with 100% accuracy. In addition, the Raman spectra not only indicated the morphological changes that are identified by cytology with hematoxylin and eosin staining but also provided molecular resolution through the use of specific marker bands. The respective Raman marker bands directly show a decrease in the level of glycogen and an increase in the levels of fatty acids in cancer cells as compared to controls. These results pave the way for "spectral" cytology of urine using Raman microspectroscopy.

Histological changes in endocrine and exocrine pancreatic tissue from patients exposed to incretin-based therapies

AIMS

Incretin-based therapies have been associated with an increased risk of pancreatitis. Recently, various histological abnormalities have been reported in human pancreatic tissue from brain-dead organ donors who had been exposed to incretin-based drugs. In the present study we examined pancreatic tissue collected at surgery.

METHODS

Human pancreatic tissue from 7 type 2-diabetic patients treated with incretin-based drugs (type 2-I), 6 diabetic patients without incretin treatment (type 2-NI), 11 patients without diabetes (no diabetes group) and 9 brain-dead organ donors (BDOD group) was examined.

RESULTS

Fractional beta-cell area was reduced in the type 2-NI group compared to the group without diabetes (P < .05), but there was no difference compared to the type 2-I patients. Alpha-cell area (P = .30), beta-cell replication (P = .17) and alpha-cell replication (P = .91) were not different. There were also no differences in acinar cell (P = .13) and duct cell replication (P = .099). Insulin-positive duct cells were more frequent in the type 2-I and the BDOD groups (P = .034). No co-expression of insulin and glucagon was detected. Pancreatic intraepithelial neoplasia (PanIN) lesions were very rare, all low-grade (PanIN 1a and 1b) and tended to occur more frequently in the type 2-I group (P = .084).

CONCLUSIONS

The present results did not reveal marked histological abnormalities in the pancreas of incretin-treated patients with type 2 diabetes. Low numbers of specimens available and a large inter-individual variability of the findings warrant caution regarding the interpretation of histological data concerning drug effects on the human pancreas.

Dynamic changes of tumor gene expression during repeated pressurized intraperitoneal aerosol chemotherapy (PIPAC) in women with peritoneal cancer

BACKGROUND

Intraperitoneal chemotherapy is used to treat peritoneal cancer. The pattern of gene expression changes of peritoneal cancer during intraperitoneal chemotherapy has not been studied before. Pressurized intraperitoneal aerosol chemotherapy is a new form of intraperitoneal chemotherapy using repeated applications and allowing repeated tumor sampling during chemotherapy. Here, we present the analysis of gene expression changes during pressurized intraperitoneal aerosol chemotherapy with doxorubicin and cisplatin using a 22-gene panel.

METHODS

Total RNA was extracted from 152 PC samples obtained from 63 patients in up to six cycles of intraperitoneal chemotherapy. Quantitative real-time PCR was used to determine the gene expression levels. For select genes, immunohistochemistry was used to verify gene expression changes observed on the transcript level on the protein level. Observed (changes in) expression levels were correlated with clinical outcomes.

RESULTS

Gene expression profiles differed significantly between peritoneal cancer and non- peritoneal cancer samples and between ascites-producing and non ascites-producing peritoneal cancers. Changes of gene expression patterns during repeated intraperitoneal chemotherapy cycles were prognostic of overall survival, suggesting a molecular tumor response of peritoneal cancer. Specifically, downregulation of the whole gene panel during intraperitoneal chemotherapy was associated with better treatment response and survival.

CONCLUSIONS

In summary, molecular changes of peritoneal cancer during pressurized intraperitoneal aerosol chemotherapy can be documented and may be used to refine individual treatment and prognostic estimations.

[Reference ranges of antioxidant parameters in whole blood (erythrocytes) in a Thüringen region]

BACKGROUND

The oxidant stress is characterized by measurement of the activities of glutathione peroxidase, superoxiddismutase and also by concentrations of glutathione and selenium in erythrocytes. A standardization of the methods of determination is very important.

MATERIAL AND METHODS

In erythrocytes of blood donors (n = 101) the parameters glutathione peroxidase, glutathione, superoxiddismutase and selenium were determined.

RESULTS

The following results of the antioxidant parameters in erythrocytes of blood donors were found: Selenium 67.1 +/- 20.1 nmol/mmol Hb, glutathione peroxidase 842 +/- 290 U/mmol Hb, glutathione 108 +/- 48 mumol/mmol Hb, superoxiddismutase 15.8 +/- 6.4 U/mumol Hb.

CONCLUSION

Selenium, glutathione peroxidase, glutathione and superoxiddismutase in erythrocytes of blood donors are normally distributed. There are no significant differences between men and women. The use of "own reference values" is necessary because no standardization of the methods of determination exists.

Macroevolution by transposition: drastic modification of DNA recognition by a type I restriction enzyme following Tn5 transposition

We have characterized a novel mutant of EcoDXXI, a type IC DNA restriction and modification (R-M) system, in which the specificity has been altered due to a Tn5 insertion into the middle of hsdS, the gene which encodes the polypeptide that confers DNA sequence specificity to both the restriction and the modification reactions. Like other type I enzymes, the wild type EcoDXXI recognizes a sequence composed of two asymmetrical half sites separated by a spacer region: TCA(N7)RTTC. Purification of the EcoDXXI mutant methylase and subsequent in vitro DNA methylation assays identified the mutant recognition sequence as an interrupted palindrome, TCA(N8)TGA, in which the 5' half site of the wild type site is repeated in inverse orientation. The additional base pair in the non-specific spacer of the mutant recognition sequence maintains the proper spacing between the two methylatable adenine groups. Sequencing of both the wild type and mutant EcoDXXI hsdS genes showed that the Tn5 insertion occurred at nucleotide 673 of the 1221 bp gene. This effectively deletes the entire carboxyl-terminal DNA binding domain which recognizes the 3' half of the EcoDXXI binding site. The truncated hsdS gene still encodes both the amino-terminal DNA binding domain and the conserved repeated sequence that defines the length of the recognition site spacer region. We propose that the EcoDXXI mutant methylase utilizes two truncated hsdS subunits to recognize its binding site. The implications of this finding in terms of subunit interactions and the malleability of the type I R-M systems will be discussed.

Transmembrane cation movements during infection of Lactobacillus lactis by bacteriophage LL-H

Phage LL-H-induced cation (K+, Na+, Mg2+, Ca2+, Cd2+) movements in Lactobacillus lactis bacteria have been studied. The effects of the m.o.i. and external cation concentration have been quantified. LL-H-induced effluxes showed cation specificity: K+ but practically no Mg2+ was lost during LL-H infection at low and moderate m.o.i. (up to about 100). Simultaneously to K+ efflux, divalent cation influxes were observed. These were dependent on the m.o.i. and on concentrations of external divalent cations and were concomitant with phage DNA transport, as concluded from the timing of the first phage-promoted biochemical changes in host cell metabolism and from electron microscopical observations. Host energy was not mobilized with phage-induced divalent cation influx. Several features of divalent cation influxes support the view that divalent cations have to be cotransported into the cell as counterions of LL-H DNA. Phage DNA associated with divalent cations may be the basic feature of the divalent cation dependence of LL-H infection.

Involvement of the bacterial groM gene product in bacteriophage T7 reproduction. II. A reduced level of ion concentrations causes the blockage of T7 maturation in K-12-M cells

Cellular leakage observed in Escherichia coli K-12-M shortly after T7 infection might be the cause of arrested phage morphogenesis. We observed in this strain, but not in the normal host, a drastic reduction of the intracellular concentration of potassium (60%), magnesium (40%), putrescine (90%), and spermidine (40%), whereas ATP was not significantly reduced. Leakage started about 1 min after the addition of phage and was arrested 3 to 5 min postinfection. Larger molecules such as o-nitrophenyl-beta-D-galactopyranoside could not enter the cells, showing that the permeability of the membrane was not generally affected. To prevent their leakage, we increased the outside concentrations of several small molecules and ions. The yield of progeny phage was substantially increased by the addition of 100 mM MgSO4.