Hypothesis: HEG1 and claudin‐4 staining will allow a diagnosis of epithelioid and biphasic mesothelioma versus non‐small‐cell lung carcinoma with only two stains in most cases

Abstract

Hypothesis: HEG1 and claudin-4 staining will allow a diagnosis of epithelioid and biphasic mesothelioma versus non-small-cell lung carcinoma with only two stains in most cases Separation of mesothelioma from metastatic carcinoma requires immunohistochemical support, with small batteries of stains recommended as a startingpoint, but these numbers commonly expand to 10, 12 or more stains, a process that is not only expensive but frequently generates anomalous or confounding results, leading to even more stains. Here we review data on HEG1 clone SKM9-2, a new (now commercially available) mesothelioma marker and claudin-4, a broad-spectrum carcinoma marker, to ask whether these two stains are sufficient, by themselves, to separate mesotheliomas from non-small-cell lung (NSCLC) as well as other carcinomas. Data for HEG1, derived from four laboratories, showed membrane staining in 393 of 434 (91%) epithelioid/biphasic mesotheliomas and one of 360 (0.3%) NSCLC (sensitivity 91%, specificity 99.7%). Reports from seven laboratories evaluating claudin-4 in NSCLC showed positivity in 469 of 502 (93%) carcinomas and weak positivity in five of 463 (1.0%) epithelioid/ biphasic mesotheliomas (sensitivity 93%, specificity 98.9%). Comparable results were found with carcinomas from other sites, except for serous and thyroid carcinomas, some of which react with HEG1 but are also positive for claudin-4. For sarcomatoid mesotheliomas, HEG1 sensitivity is modest and staining sometimes difficult to interpret. We hypothesise that the combination of HEG1 and claudin-4 immunostaining will potentially allow the separation of epithelioid/ biphasic mesotheliomas from NSCLC carcinomas with high accuracy using only two immunostains in most cases. This combination will probably also work for carcinomas from most other sites, but more reports on HEG1 SKM9-2 staining of carcinomas other than NSCLC are needed. This approach would greatly simplify the diagnosis of mesothelioma.

New developments in mesothelial pathology

“…It is often suggested that one should start with two mesothelioma and two carcinoma markers, but in practice this is not what happens: I reviewed 100 consecutive consultation cases and found that the median number of markers employed was nine; in some cases as many as 30 stains were used, and 93% of the cases had more than four stains applied 42 . What often happens is that some of the staining results are contradictory, and then even more stains are applied!…”

“…The proposition to be discussed here in summary form is that a combination of HEG1 (clone SKM9‐2) and claudin‐4 fulfils these requirements. The interested reader is referred to Churg & Naso 42 for full details.…”

“…Table 3 shows a suggested interpretation scheme. It should also be remembered that this idea really applies to tumours with an epithelial component because, both HEG1 and claudin‐4 have, at best, modest sensitivity for saromatoid mesotheliomas and sarcomatoid carcinomas 42,45 …”

“…It is often suggested that one should start with two mesothelioma and two carcinoma markers, but in practice this is not what happens: I reviewed 100 consecutive consultation cases and found that the median number of markers employed was nine; in some cases as many as 30 stains were used, and 93% of the cases had more than four stains applied. 42 What often happens is that some of the staining results are contradictory, and then even more stains are applied! A further complication is that the number of stains that in theory might be useful is simply too large to be practical, except in a very specialised facility; most laboratories are not going to maintain stocks of 10 or 20 antibodies for tumours that are fairly uncommon in everyday practice.…”

“…A summary of published information on HEG1 staining is shown in Table 2 and the detailed data from which Table 2 is drawn are available in Churg & Naso. 42 In six reports, HEG1 was expressed on 393 of 434 (91%) of epithelioid and biphasic mesotheliomas; 47 of 75 (63%) sarcomatoid mesotheliomas and only one of 360 (0.3%) non-small cell lung carcinomas (NSCLC), for a specificity of 99.7%. For epithelioid tumours, only membrane staining (Figure 10) should be counted as positive, because a small percentage of carcinomas show weak cytoplasmic staining.…”