W e read with great interest the article by Sumiyoshi et al., 1 in which the authors report that the detection of chromogranin A (Cg A) mRNA using a new in situ hybridization method is far more sensitive than immunohistochemistry. In their series, 100% of 20 small cell carcinoma (SCLC) cases were positive for Cg A mRNA (100%), as opposed to 4 cases of SCLC (20%) detected by immunohistochemistry. Four reasons for the lack of reactivity to immunohistochemistry are given: 1) partial destruction of the epitope due to the formalin fixation and paraffin embedding processes, 2) rapid secretion of the synthesized Cg A protein, 3) defects in the translation of Cg A mRNA, and 4) defects in posttranslation processing of the Cg A protein. We think that a fifth reason has been overseen, namely, the masking of the epitope by formalin fixation. In recent years, several studies have demonstrated the efficiency of heat-mediated antigen retrieval methods in demasking antigens for immunohistochemistry. The benefits of these methods for the detection of Cg A have been noticed in the literature. [2][3][4] Using heat-mediated antigen retrieval in a pressure-cooker, we were able to detect Cg A in 81.5% of a series of 108 SCCL cases (unpublished results). Using a microwave oven, Guiter et al. had similar results with 20 cases of SCCL: 16% of their cases were positive for Cg A by conventional immunohistochemistry, and 80% were positive using microwave antigen retrieval. 5 The ImmunoMax method combines heat-mediated pretreatment antigen retrieval and biotinylated tyramine enhancement. 6 It was originally devised for immunohistochemistry. Curiously, Sumiyoshi et al. adapted the ImmunoMax method for their in situ hybridization method but did not use it in their immunohistochemical protocol. Had the authors also adapted the ImmunoMax for immunohistochemistry, their results concerning immunohistochemistry would have been quite different and would certainly have demonstrated better immunohistochemical sensitivity. We remain convinced that an efficient immunohistochemical protocol (i.e., involving pretreatment with heat) is better adapted to routine situations and more cost-effective than in situ hybridization for the detection of Cg A in SCLC.