A morphometrically-based classification rule for the diagnosis of primary mesothelial lesions

Our plain-English summary of A morphometrically-based classification rule for the diagnosis of primary mesothelial lesions, by F. Robutti, P.G. Betta, A. Donna, et al (The Journal of Pathology, 1990).

The study

• Authors: F. Robutti, P.G. Betta, A. Donna, M. Pavesi
• Journal: The Journal of Pathology
• Published: 1990;162(1):57-60
• Cited by: 5 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

A computer-aided morphometrical study was performed on histological specimens of reactive hyperplastic (n = 10) and malignant (n = 17) mesothelium. For each cell, seven nuclear features were measured and 13 parameters computed. Using stepwise variable selection, discriminant analysis chose the nuclear contour index, the standard deviation of the nuclear area, and the mean of the nuclear perimeter as discriminating features between hyperplastic and malignant mesothelium. The coefficients of these variables were included in a discriminant function which gave perfect discrimination between the two groups of lesions. When the function was assessed on a test set of hyperplastic (n = 10) and malignant (n = 17) mesothelial lesions treated as 'unknown', complete separation between these two diagnostic categories was achieved. This classification rule may help to increase the level of confidence with which a histological diagnosis of mesothelioma can be established.

In plain English

The authors, pathologists at two Italian hospitals, tested whether computer measurement of cell nuclei could tell reactive (benign) mesothelial hyperplasia apart from mesothelioma in tissue samples. They took 20 cases of reactive hyperplasia and 34 epithelial-type mesotheliomas (22 pleural, 12 peritoneal) from their files. They randomly split these into two groups, each with 10 benign cases and 17 mesotheliomas. An image analysis system measured seven nuclear features in 100 nuclei per case. In the first group, a statistical method picked three features (a nuclear shape measure, the variation in nuclear area, and the average nuclear perimeter) and combined them into a score. In that group, every mesothelioma scored below zero and every benign case scored above zero. The score was then applied to the second group, treated as unknown, and it again separated all benign cases from all mesotheliomas. The authors concluded that the method appeared objective and reproducible and should improve the accuracy of mesothelioma diagnosis. They noted that it requires the lesion first to be confirmed as mesothelial in origin, and that future studies would check whether it works on fluid (effusion) samples. The article did not examine asbestos exposure.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Diagnostic morphometry study (computer-aided image analysis of archived tissue sections), with a classification rule built on a training set and checked on a separate test set
• Who was studied: 20 cases of reactive mesothelial hyperplasia (12 of the pleura, 6 of the peritoneum, 2 of the tunica vaginalis testis) and 34 mesotheliomas of epithelial type (22 of the pleura, 12 of the peritoneum), taken from the authors' files at hospitals in Alessandria and Casale Monferrato, Italy. They were randomly split into a training set and a test set, each with 10 benign lesions and 17 mesotheliomas. Collection dates are not given.
• Main results:
  • In the training set, stepwise discriminant analysis selected three features as significant (P less than 0.05) for telling mesothelial hyperplasia from malignant mesothelioma: the nuclear contour index, the standard deviation of the nuclear area, and the mean of the nuclear perimeter.
  • The score was calculated as 16.7 (NCI) minus 0.005 (area SD) minus 0.03 (mean of perimeter) minus 67.
  • In the training set, the score completely separated the two groups: all mesotheliomas scored below zero and all benign cases scored above zero. The mesothelioma score had a mean of -3.23 (SD 0.39) and the mesothelial hyperplasia score had a mean of 7.47 (SD 1.58).
  • In the test set of 27 cases treated as unknown, using a cut-off score of zero, the rule classified 100 per cent of the benign and malignant mesothelial lesions correctly.
• The authors' conclusion: The authors concluded that the procedure appeared objective and reproducible and should therefore enhance the accuracy of the pathological diagnosis of mesothelioma.
• Limitations the authors state:
  • The procedure requires unequivocal prior recognition that the lesion is mesothelial in origin.
  • The method was applied to tissue sections only. The authors said future studies would determine whether it works equally well on cytology of serous effusions.

Topics

The National Library of Medicine indexes this study under subjects our site covers: Mesothelioma.

Medical terms in this study, explained in our Lexicon: Mesothelium, Mesothelioma.

Related studies

• Value of morphometry, texture analysis, densitometry, and histometry in the differential diagnosis and prognosis of malignant mesothelioma — B. Weyn, G. Van De Wouwer, et al. (1999)
• Deep-learning based classification distinguishes sarcomatoid malignant mesotheliomas from benign spindle cell mesothelial proliferations — J.R. Naso, A.B. Levine, et al. (2021)
• Malignant mesothelioma eight years after a diagnosis of atypical mesothelial hyperplasia — J. Scurry, M.A. Duggan (1999)
• Value of Glut-1 and Koc markers in the differential diagnosis of reactive mesothelial hyperplasia, malignant mesothelioma and pulmonary adenocarcinoma — Ö. Üçer, A.F. Dağli, et al. (2013)
• Computer-assisted differential diagnosis of malignant mesothelioma based on syntactic structure analysis — B. Weyn, G. van de Wouwer, et al. (1999)
• L1CAM expression in cystic mesothelial lesions: a comparison with adenomatoid tumours, well-differentiated papillary mesothelial tumours and malignant mesotheliomas — G. Karpathiou, F. Casteillo, et al. (2021)

Browse more research in The Journal of Pathology, or return to medical research.