To our knowledge, we have reported on the second case of sarcomatoid mesothelioma arising from MIS, following the first report (which was also presented by our research group) [4, 5]. Interestingly, both cases share the same genomic abnormalities (BAP1 retain, MTAP loss, and CDKN2A homozygous deletion). The presence of MTAP loss and CDKN2A homozygous deletion are suspected to be involved in the early progression to invasive lesions and/or sarcomatoid changes.
MIS was first proposed in 1992 [9]. However, until recently, it had been unclear if there was truly an in situ state. The concept of MIS has recently become accepted with the development of genome-based diagnostic methods [10]. BAP1 IHC, MTAP IHC, and CDKN2A FISH are gold-standard methods for distinguishing malignant mesothelioma from benign mesothelial lesions, with high sensitivity (80–90% in combination) and 100% specificity [11]. The sensitivity of BAP1 IHC is greater for epithelioid mesothelioma (60–70%) than for sarcomatoid mesothelioma (< 40%), while MTAP loss and CDKN2A homozygous deletions are detected more in sarcomatoid mesothelioma (~ 90%) than in epithelioid mesothelioma (60–70%) [11, 12]. It should be noted that high homozygous deletion of CDKN2A in pleural mesothelioma correlates with poor prognosis [13]. This may be associated with a high proportion of sarcomatoid features in MTAP-loss mesothelioma.
Although attention towards MIS has increased, indications for treatment and the timing of treatment are under discussion. Klebe et al. reported that the majority of MIS patients had only received follow-up treatment, while the remaining patients received active treatment [14]. These researchers also reported variation in the time of progression from in situ lesions; 15% of the MIS cases progressed to invasive mesothelioma within 6–12 months, while 35% showed progression ≥ 4 years later [14]. We speculate that one reason for this variation may be due to genetic abnormalities. In our opinion, careful follow-up or early intervention might be important for MIS patients, at least those with MTAP loss and/or CDKN2A homozygous deletions. However, this evidence is still preliminary.
The nine criteria suggesting mesothelioma on effusion cytology are as follows: 1) high cellularity, 2) large size, 3) papillary clusters, 4) acidophilic matrix, 5) macronucleoli, 6) protrusion from the cell membrane, 7) a prominent degree of cell-within-cell arrangement, 8) multinucleated cells, and 9) vacuoles overlapping the nuclei on Giemsa staining [15]. There is evidence that cell-in-cell engulfment is an important finding in the diagnosis of mesothelioma in body cavity fluids. For example, PM cells with a 9p21 homozygous deletion occasionally exhibit cell-in-cell engulfment, multinucleation, and/or multicellular clusters; this occurs more frequently than for reactive mesothelial cells [7, 8]. However, it is very difficult to suspect malignant mesothelioma in cases when only one out of nine criteria are met, such as in our case.
Cell-in-cell engulfment, named entosis by Overholtzer, is one form of the nonapoptotic cell death process in matrix-detached cells [15]. The stimulated mesothelial cells, whether reactive or malignant, easily adhere to each other and form a cell-to-cell apposition under the matrix detachment. In contrast, PM cells tend to show cell-in-cell engulfment more frequently [6–8]. Additionally, cell-in-cell engulfment is infrequently seen in histological samples. Therefore, this pattern may occur during or after the exfoliation of tumor cells into the body cavity fluid [2]. The rate of cell-in-cell engulfment in our case was 7.3%, which was much lower than in PM (17.4–20.6%) but was almost twice the average value in reactive mesothelial cells (3.5–4.1%) evidenced in previous reports [7, 8]. In our case, cell-in-cell engulfment was prominent (7.3%) compared to the low frequency of multinucleated cells (5.2%). Taken together, our findings suggest the necessity of paying attention to not only overt cytomorphological findings but also cell-in-cell engulfment, since this might be seen from the very early stages of mesothelioma.
In conclusion, we would like to emphasize that, MTAP loss and/or CDKN2A homozygous deletion in MIS may indicate aggressive features, based on our two MIS cases with MTAP loss and CDKN2A homozygous deletions that progressed to sarcomatoid mesothelioma. Re-biopsy of invasive lesions after diagnosing MIS is essential, especially for MIS presenting with these types of genetic abnormalities. When the number of mesothelial cells with cell-in-cell engulfment is high, even if there are no overt atypical cytological findings on effusion cytology, the possibility of mesothelioma should be considered.