Utility and limitations of Merlin immunohistochemistry for mesothelioma diagnosis in tissue sections and cell blocks

Distinguishing pleural mesothelioma (PM) from reactive mesothelial proliferations (RMP) can be challenging. In such cases, immunohistochemistry (IHC)-detected BAP1 or MTAP loss and FISH-detected CDKN2A homozygous deletion are effective. Merlin is the protein product of the NF2 gene, which is frequently altered in mesotheliomas. Recently, IHC-detected loss of Merlin was also shown to be useful in differentiating PM from RMP. To validate these findings, we examined Merlin IHC in PM cases, including for the first time cytologic material. Merlin IHC was performed on 67 PM cases, including 47 samples from tissues and 20 from cell blocks (CBs), and 29 RMP cases. In RMP, Merlin was expressed in cell membranes and cytoplasm, with no loss. Merlin expression was lost in 49 % of PM tissues. In discriminating PM from RMP, addition of Merlin IHC to the combination of BAP1 and MTAP IHC increased sensitivity from 77 % to 95 %. However, Merlin expression could not be assessed in 8.5 % of tissues and 25 % of CBs. In CBs, Merlin loss could be assessed only in sheeted or clustered tumor cells, because PM cells could not be identified precisely in few scattered tumor cells. In cases where both tissue biopsy and CBs were available, results matched in only 50 % of cases, suggesting uneven occurrence of Merlin loss in PM tissues. Our observations support the effectiveness of Merlin IHC in differentiating PM from RMP. However, investigators should be familiar with potential challenges in interpreting Merlin IHC results, especially in CBs.