Histochemistry--Epithelial (or Mixed) Tumors

Histochemical tests, less expensive than immunostains, can sometimes be diagnostic. Alcian blue (AB) stains acid mucin that is produced by epithelial mesotheliomas. AB is made specific for acid mucin by staining 2 slides, one of which has been pretreated with hyaluronidase to remove the hyaluronic acid of acid mucin (AB ± H). A positive stain shows blue cytoplasmic droplets in the untreated slide and absence (or decrease) of staining of cytoplasmic droplets in the slide treated with hyaluronidase. The removal of stromal stain indicates that the digestion worked, but has no diagnostic significance. Positive in about 50% of epithelial mesotheliomas, this stain is definitive. Its sensitivity is low because of the solubility of hyaluronic acid in aqueous fixatives, and it is seldom used nowadays [1].


An alcian blue stain of an epithelial mesothelioma shows staining of vacuoles (arrows) (left panel). In addition there is staining of tumor stroma. The section in the right panel was treated with hyaluronidase before staining with alcian blue. There is no staining of either vacuoles or stroma. Only the vacuolar staining reaction is specific for mesothelioma: stromal staining of any tumor will be destroyed after treatment with hyaluronidase.


Production of neutral mucin signifies adenocarcinoma. The PAS stain after diastase digestion (PAS-D) stains cytoplasmic neutral mucin droplets red. A positive result is a cytoplasmic vacuole with a target-like appearance--a red rim separated from a central bull's eye by a clear space sometimes with radiating spokes (arrow points to one example, figure B). This stain works well on both cytologic and histologic preparations. These droplets are sometimes also stained by alcian blue, but hyaluronidase does not decrease the intensity of the stain. Another stain for mucin, the mucicarmine stain, can stain both acid and neutral mucin, and thus has no use in distinguishing carcinoma from mesothelioma. Caution: Rare PAS-D positive mesotheliomas have been reported [1,2]. In such cases, clinical features, ultrastructure, and immunohistochemical studies will signify mesothelioma.


A neoplastic cell has cytoplasmic droplets of neutral mucin that indicate adenocarcinoma (PAS-D stain).

The PAS stain (without diastase) colors red a number of different structures that contain carbohydrates: glycogen, mucin, basement membranes, lysosomes, leukocyte granules, and most fungi. Some tumors, including epithelial malignant mesotheliomas, have cytoplasmic glycogen, which is removed by pretreatment of the slide by diastase.


This PAS stain of a mesothelioma shows fine cytoplasmic droplets of glycogen (left). Pretreatment of the slide with diastase removes the glycogen, as shown on the right. The fine cytoplasmic stain that remains is in leukocyte granules that do not contain glycogen. Cytoplasmic droplets of neutral mucin, as seen in adenocarcinomas, are larger.

Immunohistochemistry: Several antibodies (B72.3, CD-15 (Leu-M1), and CEA) are often used to distinguish adenocarcinomas (which usually stain positively) from mesotheliomas (which usually do not stain) [3]. Several antibodies are used because none is completely sensitive or specific for adenocarcinoma. Positive and negative control slides are needed to evaluate the staining. More than 10% of viable tumor cells should stain in a carcinoma. A recent caveat concerns false positive staining with Leu-M1 and CEA in mesotheliomas. This staining usually occurs in tumors that are rich in acid mucin. It can be avoided by pretreatment of the sections with hyaluronidase [4].

Sarcomatous tumors show diffuse staining with antibodies to cytokeratin. This feature is helpful in differentiating the tumor from other sarcomas, which are usually negative, or only focally positive, for cytokeratin [1]. Carcinomas, mesotheliomas, and reactive pleural fibrosis can all be expected to stain with antibodies to the usual pancytokeratin.

Addendum: Two new antibodies that stain most epithelial mesotheliomas, but only rare adenocarcinomas, are now available for use on paraffin sections. Both antibodies stain normal, reactive, and neoplastic mesothelium. The first antibody reacts with calretinin, a calcium-binding protein that is expressed in neural tissues and non-neural cell types that include mesothelioma. When nuclear staining alone, rather than cytoplasmic staining, is evaluated, the sensitivity and specificity can be very high, >90% and 98%, respectively. The second antibody--cytokeratin 5/6 (CK 5/6)--reacts with types 5 and 6 of the many types of keratin intermediate filaments. These filaments are more frequent in mesothelial cells than in cells of adenocarcinoma. The antibody gives cytoplasmic staining, and has a sensitivity and specificity for epithelial mesothelioma of 92% and 86%, respectively. CK 5/6 reactivity may also be found in squamous carcinoma, transitional carcinoma, and ovarian carcinoma. Therefore, the results must be interpreted in the clinical setting. Experience in performance and evaluation is necessary for all of these stains, and concordant results from several tests should be used for diagnosis [5].



This immunoperoxidase stain for B72.3 shows positive membrane and cytoplasmic staining signifying that the tumor is an adenocarcinoma.

Electron microscopy is sometimes useful in distinguishing epithelial mesotheliomas from carcinomas. While most ultrastructural features are similar, mesotheliomas are characterized by having long, narrow, branching microvilli with a length to width ratio of around 10 to 15:1. In general, spindled cells of mesothelioma lack microvilli, but sometimes cells intermediate between epithelial and spindled with long microvilli can be found in predominantly spindled tumors. By contrast, adenocarcinomas have short, stubby microvilli [1].


1. McCaughey W, Colby T, Battifora H, Churg A, Corson J, Greenberg S, Grimes M, et al. Diagnosis of diffuse malignant mesothelioma: experience of a US/Canadian mesothelioma panel. Mod Pathol 1991; 4:342-353.

2. MacDougall D, Wang S, Zidar B. Mucin-positive epithelial mesothelioma. Arch Pathol Lab Med 1992; 116:874-880.

3. Brown R, Clark G, Tandon A, Allred D. Multiple-marker immunohistochemical phenotypes distinguishing malignant pleural mesothelioma from pulmonary adenocarcinoma. Hum Pathol 1993; 24:347-354.

4. Robb J. Mesothelioma versus adenocarcinoma: false-positive CEA and Leu-M1 staining due to hyaluronic acid. Hum Pathol 1989; 20:400.

5. Cury P, Butcher D, Fisher C, Corrin B, Nicholson A. Value of the mesothelium-associated antibodies thrombomodulin, cytokeratin 5/6, calretinin, and CD44H in distinguishing epithelioid pleural mesothelioma from adenocarcinoma metastatic to the pleura. Mod Pathol 2000; 13:107-112.


Table of Contents