Immunohistochemical reactivity in mesothelioma and adenocarcinoma: a stepwise logistic regression analysis

Our plain-English summary of Immunohistochemical reactivity in mesothelioma and adenocarcinoma: a stepwise logistic regression analysis, by A. Dejmek, A. Hjerpe (APMIS, 1994).

The study

• Authors: A. Dejmek, A. Hjerpe
• Journal: APMIS
• Published: 1994;102(4):255-64
• Cited by: 35 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Histological sections from 103 malignant mesotheliomas and 43 adenocarcinoma metastases in pleural biopsies were investigated for reactivity against a panel of 11 different antibodies. The size of the material allowed the evaluation by stepwise logistic regression analysis, which selected five parameters of major importance: vimentin reactivity in epithelial cells, reactivity to low-molecular-weight keratins in fibrous cells, strong membrane accentuation of EMA reactivity, and lack of reactivity to LeuM1 and BerEp4. Three of these criteria were sufficient to identify a mesothelioma with high specificity and with a sensitivity of approximately 70%. Whilst the monoclonal anti-CEA tested was the most valuable single parameter, it did not add any diagnostic information to the combination of criteria selected by the stepwise logistic regression analysis. However, this antibody can be used to exclude most of the adenocarcinomas from further analysis with the more extensive panel.

In plain English

This was a laboratory study of tumor tissue. It tested whether antibody staining (immunohistochemistry) could distinguish malignant mesothelioma from adenocarcinoma that had spread to the pleura. It did not study asbestos exposure. The authors stained tissue from 103 malignant mesotheliomas from two Swedish hospitals, Malmo General and Huddinge University, and from 43 adenocarcinoma metastases of the pleura from Malmo General Hospital autopsy files. A monoclonal anti-CEA antibody stained 70% of the adenocarcinomas and none of the 103 mesotheliomas. Stepwise logistic regression selected five criteria: vimentin staining in epithelial cells, CAM5.2 staining in fibrous cells, EMA staining strongly concentrated at cell membranes, and negative results for LeuM1 and BerEp4 (weak focal staining was counted as negative). Requiring the first or second criterion plus one other identified 77% of mesotheliomas, with 100% specificity in the material studied. A stricter rule, requiring three criteria, identified 69% of mesotheliomas. The authors concluded that the anti-CEA antibody could be used as a first step to exclude most adenocarcinomas before applying the larger panel. They noted that adenocarcinomas were underrepresented in their material.

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 immunohistochemistry study of archived tumor tissue sections, analyzed with two-by-two contingency tables and stepwise logistic regression
• Who was studied: 103 malignant mesotheliomas (29 epithelial, 63 biphasic, 11 fibrous) from the autopsy and biopsy files of Malmo General Hospital and Huddinge University Hospital, Sweden, and 43 adenocarcinoma metastases of the pleura (from lung, colon, breast, ovary, pancreas, prostate and stomach) consecutively retrieved from the autopsy files of Malmo General Hospital
• Exposure: Not an exposure study. No asbestos or fiber exposure was assessed. The study tested tissue reactivity to a panel of 11 commercially available antibodies.
• Main results:
  • The monoclonal anti-CEA antibody stained 70% of the adenocarcinomas and none of the 103 mesotheliomas. The authors called it the most valuable single antibody.
  • Vimentin staining in epithelial cells was seen in about half of the mesotheliomas, but in only one adenocarcinoma (a breast cancer).
  • LeuM1, B72.3 and BerEp4 stained 88%, 75% and 66% of adenocarcinomas respectively, compared with 33%, 23% and 15% of mesotheliomas.
  • Stepwise logistic regression selected five criteria: vimentin reactivity in epithelial cells, CAM5.2 reactivity in fibrous cells, EMA reactivity with strong membrane accentuation, lack of reactivity to LeuM1, and lack of reactivity to BerEp4. In the scoring, cases with weak focal reactivity were assigned to the negative group.
  • When the first or second criterion plus one of the other four was met, sensitivity for mesothelioma was 77%, with 100% predictive value and specificity in the investigated material.
  • The more stringent rule, requiring three of the five criteria (p greater than 0.99), reduced sensitivity, the proportion of mesotheliomas identified, to 69%.
  • Among mesotheliomas with an epithelial component, the corresponding sensitivities were 76% and 72%.
  • Although CEA was the most valuable single parameter, it was excluded from the final model, mainly because it varied together with EMA and intermediate filaments.
• The authors' conclusion: The authors concluded that CEA did not improve the model. They concluded that the monoclonal anti-CEA antibody could still be used first, as a screening step, to exclude most adenocarcinomas, with the five-criterion panel then applied to the CEA-negative cases.
• Limitations the authors state:
  • Adenocarcinomas were underrepresented in the material, so the authors suggested using a more stringent p-value limit.
  • In practice it can be hard to tell a very proliferative stroma from a fibrous tumor component, especially in a small biopsy.
  • Keratin staining can also occur in stromal fibroblasts in adenocarcinomas.

Topics

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

Medical terms in this study, explained in our Lexicon: Mesothelioma, Adenocarcinoma, Malignant Mesothelioma, Monoclonal.

Related studies

• Immunohistochemical panels for differentiating epithelial malignant mesothelioma from lung adenocarcinoma: a study with logistic regression analysis — R. Carella, G. Deleonardi, et al. (2001)
• Immunohistochemical distinction of malignant mesothelioma from pulmonary adenocarcinoma with anti-surfactant apoprotein, anti-Lewisa, and anti-Tn antibodies — M. Noguchi, T. Nakajima, et al. (1989)
• Monoclonal antibody SM047 as an immunohistochemical marker of ovarian adenocarcinoma — W.G. McCluggage, P. Maxwell, et al. (2001)
• HBME-1, MOC-31, WT1 and calretinin: an assessment of recently described markers for mesothelioma and adenocarcinoma — J. Oates, C. Edwards (2000)
• Calretinin, thrombomodulin, CEA, and CD15: a useful combination of immunohistochemical markers for differentiating pleural epithelial mesothelioma from peripheral pulmonary adenocarcinoma — C.E. Comin, L. Novelli, et al. (2001)
• Adenocarcinoma simulating mesothelioma: a clinicopathologic and immunohistochemical study of 29 cases — M.N. Koss, M. Fleming, et al. (1998)

Browse more research in APMIS, or return to medical research.