Evidence-based pathology

Novel prognostic histopathological grading system in oral squamous cell carcinoma based on tumour budding and cell nest size shows high interobserver and intraobserver concordance

AIMS

Squamous cell carcinoma of the oral cavity (OSCC) is a common tumour entity with a variable, partially highly aggressive clinical course. Recently, we proposed a novel (three-tiered) clinically useful grading scheme strongly associated with patient outcome in OSCC, consisting of a sum score of the histomorphological patterns tumour budding and cell nest size which outperforms WHO based grading algorithms currently in use. The aim of our study was to probe for interobserver and intraobserver reliability of this novel grading system.

METHODS

108 OSCC were retrospectively scored according to the proposed grading scheme by three independent pathologists-two experienced head and neck pathologists and one pathologist in training-blinded to each other's scoring results.

RESULTS

The Cohen's Kappa (κ) values for concordance rates between experienced pathologists were κ=0.97 for the overall grade, κ=0.97 for budding activity and κ=0.91 for cell nest size, indicating a strong interobserver reliability of our proposed grading system. Initial interobserver agreement was markedly lower with the pathologist in training (κ=0.55 for overall grade) but improved significantly after a training session (κ=0.87 for overall grade). Intraobserver concordance was high (κ=0.95 for overall grade), indicating a high reproducibility of the algorithm.

CONCLUSIONS

In conclusion, our study indicates that OSCC grading based on our proposed novel scheme yields an excellent inter-reader and intrareader agreement, further supporting the suitability of this grading system for routine pathological practice.

Evidence-based pathology: systematic literature reviews as the basis for guidelines and best practices

CONTEXT

Evidence-based medicine has been proposed as a new paradigm for the identification and evaluation of medical information. Best available evidence or data are identified and used as the basis for the diagnosis and treatment of individual patients. Evidence-based pathology has adapted basic evidence-based medicine concepts to the specific needs of pathology and laboratory medicine.

OBJECTIVES

To briefly review the history and basic concepts of evidence-based medicine and evidence-based pathology, describe how to perform and interpret systematic reviews, and discuss how to integrate best evidence into guidelines.

DATA SOURCES

PubMed (National Library of Medicine, Washington, DC) and Web of Science (Thompson Reuters, New York, New York) were used.

CONCLUSIONS

Evidence-based pathology provides methodology to evaluate the quality of information published in pathology journals and apply it to the diagnosis of tissue samples and other tests from individual patients. Information is gathered through the use of systematic reviews, using a method that is less biased and more comprehensive than ad hoc literature searches. Published data are classified into evidence levels to provide readers with a quick impression about the quality and probable clinical validity of available information. Best available evidence is combined with personal experience for the formulation of evidence-based, rather than opinion-based, guidelines that address specific practice needs.

Improving Renal Cell Carcinoma Classification by Automatic Region of Interest Selection

In this paper, we present an improved automated system for classification of pathological image data of renal cell carcinoma. The task of analyzing tissue biopsies, generally performed manually by expert pathologists, is extremely challenging due to the variability in the tissue morphology, the preparation of tissue specimen, and the image acquisition process. Due to the complexity of this task and heterogeneity of patient tissue, this process suffers from inter-observer and intra-observer variability. In continuation of our previous work, which proposed a knowledge-based automated system, we observe that real life clinical biopsy images which contain necrotic regions and glands significantly degrade the classification process. Following the pathologist's technique of focusing on selected region of interest (ROI), we propose a simple ROI selection process which automatically rejects the glands and necrotic regions thereby improving the classification accuracy. We were able to improve the classification accuracy from 90% to 95% on a significantly heterogeneous image data set using our technique.

Evaluation of the invasion front pattern of squamous cell cervical carcinoma by measuring classical and discrete compactness

The invasion front pattern of squamous cell carcinoma (SCC) is a conspicuous histological phenomenon, which is assessed without precise criteria. The current study was performed to introduce the classical (C(C)) and discrete compactness (C(D)) as new morphometric parameters for quantification of this pattern. A retrospective analysis of 76 surgically treated patients with cervical carcinoma was conducted and the pattern of invasion was qualitatively classified as closed, finger-like or diffuse, respectively, by two pathologists. After digitization of the histological slides with a field of view of 10.4 mm x 8.3mm, tumor areas were labeled and C(C) and C(D) were computed based on the drawings (binary images). Additionally, intraindividual variation of compactness was evaluated for 12 selected tumors. The qualitative pattern assessment by the pathologists was moderately reproducible with an interobserver agreement of 72% and a kappa coefficient of 0.44. The values of C(C) and C(D) referring to the invasion front patterns assigned by both pathologists were significantly different between the three classified groups (p< or =0.01 and p< or =0.0001), so that, both theoretically and in practice, compactness regards the same morphological feature. In due consideration of the analysis of the area under the ROC (receiver operating characteristic) curves and the variation coefficient of different tumor regions, C(D) is more suitable for practical use than C(C). Tumors with a microscopic invasion into the parametria and with lymph-vascular space invasion were found to have a lower value of C(D), which indicates a more diffuse pattern of invasion (p=0.028 and p=0.033). We conclude that the discrete compactness C(D) is a new and reproducible parameter for a computer assisted quantification of the invasion front pattern and, thus, defines a further phenotypic feature of SCC of the uterine cervix.

Computation of semantic similarity within an ontology of breast pathology to assist inter-observer consensus

Computer-assisted consensus in medical imaging involves automatic comparison of morphological abnormalities observed by physicians in images. We built an ontology of morphological abnormalities in breast pathology to assist inter-observer consensus. Concepts of morphological abnormalities extracted from existing terminologies, published grading systems and medical reports were organized in an taxonomic hierarchy and furthermore linked by the relation "is a diagnostic criterion of" according to diagnostic meaning. We implemented position-based, content-based and mixed semantic similarity measures between concepts in this ontology and compared the results with experts' judgment. The position-based similarity measure using both taxonomic and non-taxonomic relations performed as well as the other measures and was used for automatic comparison of morphological abnormalities within the IDEM computer-assisted consensus platform.

Architectural analysis of oral cancer, dysplastic, and normal epithelia

BACKGROUND

We present a novel, automated, and quantitative approach to evaluate local epithelial tissue architecture based on mathematical graph theory.

METHODS

Four hundred forty-one images of three diagnostic classes of oral epithelium (normal, dysplastic, and neoplastic) were analysed. The epithelial compartment was partitioned into exclusive areas associated with each nucleus to approach the theoretical cell extents. The spatial arrangement of cells in neighbourhoods of two sizes was characterised by constructing graph networks based on the cell centroids and recording 29 statistical properties. We analysed 104,627 and 67,590 neighbourhoods of diameters 37.5 and 75 microm, respectively.

RESULTS

The discrimination power of the architectural descriptors was evaluated by using discriminant analysis. The best neighbourhood discrimination rate was 75% for normal versus carcinoma. For the pooled data, discrimination into three classes based on largest number of neighbourhoods associated with each class was 100% correct. Case-wise, discrimination rates were 67%, 100%, and 80% correct for normal, premalignant, and malignant. When considering two classes, discrimination rates was 89% (normal) and 100% (malignant) correct, with 71% of premalignant cases assigned to the malignant class.

CONCLUSIONS

The results indicate that unbiased and reproducible quantification of tissue architectural features is possible and may provide valuable morphological information for diagnostic purposes.

Invasive front grading: reliability and usefulness in the management of oral squamous cell carcinoma

BACKGROUND

The value of histological grading was examined with emphasis on reliability of assessment in 102 cases of intraoral squamous cell carcinoma from Northern Ireland with known outcome.

METHODS

Two pathologists independently graded the invasive tumour front blinded to the stage and outcome.

RESULTS

Intraobserver agreement was acceptable but interobserver agreement was not satisfactory. The degree of keratinisation was assessed most consistently while nuclear polymorphism was the least reliable feature. Multivariate survival analysis showed that the total grading score was associated with overall survival while the pattern of tumour invasion was the most valuable feature in estimating regional lymph node involvement. The number of positive lymph nodes was strongly associated with regional relapse, while the treatment modality and status of the surgical margins correlated with local relapse.

CONCLUSIONS

Grading of selected features in OSCC is reliable and can facilitate treatment planning.

Approach to ground-glass opacification of the lung

There area number of diseases that present with ground-glass opacification of the lung as a primary manifestation on chest radiography and thin-section computed tomography (CT). These diseases cannot be clearly categorized into the classic classification scheme of airspace and interstitial disease because there are features of both categories seen in the imaging and histologic findings. Ground-glass opacification has, therefore, been categorized as nonspecific by many radiologists. The fact that both the airspaces and interstitial tissues are often involved should have little importance when evaluating radiographs or high-resolution CT (HRCT) images. The role of the radiologist is evolving and is becoming more significant in the clinical evaluation of a patient presenting with so-called interstitial lung disease. In this article, an approach is described that hopefully will lead to an effective and narrow differential diagnosis when a radiologist is confronted with ground-glass opacification. The critical features in evaluating such cases include: (1) the duration of clinical symptoms; (2) the presence or absence of lung fibrosis, especially honeycombing; (3) the patient's history of smoking; and (4) the distribution of the disease. Each of these features is considered in this article in conjunction with the diseases that lead to ground-glass opacification. The recent thinking regarding nonspecific interstitial pneumonia and usual interstitial pneumonia also is discussed.

The boundaries of cellular pathology: how pathologists see their clinical role

There has been much recent public attention on the effects of the practices of cellular pathology on patients and their relatives. It is important to establish clearly pathologists' views about their professional relationships to clinicians and patients. A national postal questionnaire survey was performed to investigate how cellular pathologists perceive their role in clinical practice and whether there are important differences between different groups of pathologists. Responses were received from 773 pathologists, of whom 167 were trainees. On the basis of responses to 25 attitude statements, it appears that the profession shares a core belief that pathologists are clinicians rather than scientists and sees diagnosis as its main clinical responsibility. However, the role that each pathologist feels clear about differs significantly between pathologists, with a stronger consensus over function, responsibility, and professional image than on the clinical practice of pathology, job satisfaction, and the impact of technological change. Academic activities are a minority interest, although academic pathologists express better job satisfaction and more radical views of professional practice. Young trainees are more likely to work in teaching hospitals, to be female, to have trained outside the UK, and to have had experience in other specialties. They have, however, a more restricted view of the clinical role of pathology than accredited pathologists. Most cellular pathologists see themselves as clinicians with a special role in diagnosis, but how this is interpreted and the extent to which pathologists involve themselves in clinical decision-making is inconsistent. This professional dilemma should be addressed by research to obtain more systematic knowledge of the clinical impact of cellular pathology and especially the ill-defined links between diagnosis and clinical decision-making.

Biliary dysplasia as a marker of cholangiocarcinoma in primary sclerosing cholangitis

BACKGROUND/AIMS

Indentification of biliary dysplasia in a primary sclerosing cholangitis (PSC) liver biopsy may indicate developing cholangiocarcinoma. The objectives were to determine whether biliary dysplasia can be recognised reproducibly in PSC and to compare the frequency in cases with and without cholangiocarcinoma.

METHODS

Liver biopsies from 26 PSC cases with concurrent or subsequent cholangiocarcinoma (within 2 years) were assessed for biliary dysplasia independently by three liver pathologists. This was done in two stages: initially, without agreement on criteria, and subsequently after such agreement. Liver biopsies from 60 PSC cases without cholangio-carcinoma were also assessed.

RESULTS

Reproducibility for biliary dysplasia without prior agreement on criteria was only marginally better than random (kappa=0.129). In contrast, after prior agreement on criteria, reproducibility was moderate (kappa=0.44). Biliary dysplasia was agreed to be present by all three pathologists in 23% and 19% of biopsies in the first and second round, respectively, from patients with cholangiocarcinoma, but in none of the patients without cholangiocarcinoma.

CONCLUSION

Criteria for biliary dysplasia can be agreed and the entity recognised in liver biopsies. The strong association of biliary dysplasia with cholangiocarcinoma in PSC suggests use of dysplasia as a marker for current or developing malignancy.

Benchmarking general practice use of pathology services: a model for monitoring change

AIMS

To identify a model to assess general practitioner use of pathology services that could be applied to assess specific interventions designed to promote best practice.

METHODS

A database containing standardised requesting data for 22 general practices was constructed. The database contained 28 tests covering 95% of general practitioner activity, distributed across pathology, and it was evaluated during two sequential six month periods. A comparison of ranks of requesting activity between different time periods was undertaken by calculating Pearson rank correlation coefficients. Requesting numbers were also adjusted for patients' age and sex distributions within the 22 practices for a sample of three high volume tests. The effects of distributing requesting guidelines and details of requesting activity were assessed during two sequential three month periods.

RESULTS

Requesting activity was extremely stable during the two baseline periods for most test (r > 0.80 for 20 of the 28 tests). Several less discriminatory tests were identified. Age and sex adjustment had minimal impact on the ranks of requesting activity. Requesting activity during the two three month periods after distributing guidelines and comparative details of individual requesting activity showed little change (overall correlation coefficient, 0.844 between baseline and intervention periods).

CONCLUSIONS

Ranking general practitioners requesting activity adjusted for practice list size provides a reproducible means of measuring requesting activity for most pathology tests performed in general practice. Activity was not influenced by age or sex of patients on the practice list. Distributing requesting guidelines and individual requesting activity on their own do not have any measurable impact on requesting activity. More innovative (possibly multiple) interventions might be required to influence general practitioner requesting practice.

Mapping the evidence base of pathology

Evidence-based medicine (EBM) is a paradigm that is beginning to exert an influence in related fields such as surgery, general practice, psychiatry, and now pathology. For example, a survey has shown that 70 per cent of therapeutic interventions in clinical haematology delivered in a district general hospital were based on research-derived evidence. A prerequisite to evidence-based pathology is availability of, and access to, the evidence base. As a discipline, pathology has certain characteristics that makes information retrieval more challenging. Nevertheless, a number of evidence-seeking techniques can be utilized to maximize the chances of success: focusing the question, use of an evidence-seeking protocol, and application of methodological filters. A number of key information sources are reviewed for their usefulness and a comparison is made with the yield from the World Wide Web. Conclusions are drawn from an example of the evidence-seeking process based on a clinical scenario involving immune thrombocytopenic purpura.

Internal quality assurance activities of a surgical pathology department in an Australian teaching hospital

AIM

To assess the role of a quality assurance programme in improving the service provided by a surgical pathology department.

METHODS

A continuous internal quality assurance study of the activities of an anatomical pathology department in an Australian teaching hospital was undertaken over a five year period. This addressed all steps involved in the production of a surgical pathology report. These were addressed in an open forum which included technical, scientific, clerical, and medical staff. Minor errors not needing immediate action were discussed and incorporated into laboratory practice. For major discrepancies with potential implications for patient management supplementary reports were issued and the relevant clinician informed of the outcome.

RESULTS

Comprehensive peer review of 8.9% of the total workload of the department (3530 cases) and all the frozen sections (916 cases) over a period of five years, beginning in 1991, led to comments on some aspects of the original report by the reviewer in 19.6% of the cases. The great majority of the comments were minor, concerning issues related to the microscopic findings (4%), macroscopic description (3.1%), clerical aspects (3%), typographical errors (3%), coding errors (2.7%), technical errors including poor sections and incorrect labelling (1.7%), inadequate clinical history (1.2%), and incomplete or incomprehensible diagrams (0.9%). In two cases (0.05%) the original report did not state proximity of the tumour to surgical margins and in three of the frozen sections (0.3%) the original diagnosis was incorrect. However, in these cases the frozen section assessment did not alter the overall management of the cases.

CONCLUSIONS

This study highlights the importance of a review system in detecting errors in surgical pathology reporting. Recognition of the fact that surgical pathology is not infallible has improved the end product. It has also minimised interobserver variability in the department, resulting in a uniform approach among the pathologists to macroscopic description, specimen sampling, special stains, and histological reporting.

IDEM: a Web application of case-based reasoning in histopathology

Different software engineering and artificial intelligence methods can be used to design Internet retrieval of prototypical medical images. We used the case-based reasoning (CBR) approach to provide an 'intelligent' access to a collection of illustrated medical cases through the Internet. This paper presents a Web interface for the CBR system IDEM (image and diagnosis from examples in medicine) in the domain of breast pathology. Thanks to the definition of a similarity measure between the descriptions of cases we propose a flexible querying of the case-base and a quantitative browsing among cases through similarity links. The resemblance rates provided by the system argue for the quality and the relevancy of the retrieved data. The flexibility of the querying process is robust to missing information and could be adapted to a daily practice. The CBR approach is a promising method for a clinical relevant and an efficient retrieval of reference images and diagnosis clues through Internet.

Is the invasive front of an oral carcinoma the most important area for prognostication?

The prognosis for patients with oral squamous cell carcinomas is difficult to predict. There is therefore a great need for more reliable prognostic markers which will be of help in the treatment decisions. In this review I present a hypothesis which suggests that molecular and morphological characteristics at the invasive front area of various carcinomas may reflect tumour prognosis better than other parts of the tumour. It is now known that several molecular events of importance for tumour spread like gains and losses of adhesion molecules, secretion of proteolytic enzymes, increased cell proliferation and initiation of angiogenesis occur at the tumour-host interface (invasive front). Consequently, our group has recently developed a simple morphological malignancy grading system that restricts the evaluation to the deep invasive front area of the tumour. Several studies have shown that this system is a significantly better predictor of prognosis than traditionally used morphological systems. All studies performed so far show that invasive front grading is a valuable supplement to clinical staging, suggesting that it should be introduced into the clinic.