We read with interest the article by Gong et al. in a recent issue of Cancer Cytopathology.1 The authors performed immunohistochemistry (IHC) on 100 cell-transferred pieces from 20 tumors and 2 reactive effusions. They demonstrated 97% concordant staining with IHC performed on previous cell blocks or smears obtained from the same patients. However, there was a 3% rate of false-negative staining and a 5% loss of specimen during the staining procedure (the article did not state at which step of the staining procedure the loss occurred).1 Destaining using 1% hydrogen chloride in 70% alcohol was performed on all cell-transferred pieces before IHC. Antigen retrieval was not employed in the immunostaining of cell-transferred pieces, with the exception of estrogen receptor determination. We agree with Gong et al. that the cell transfer technique (CTT) is extremely helpful for cytologic specimens lacking adequate cell blocks. We would further like to share our experience with CTT. We have utilized CTT at our institution over the last 1.5 years with great success. We have used it on fine-needle aspiration specimens and exfoliative cytology when there was limited material and an adequate cell block was not available for further testing. We also have applied CTT to small biopsy specimens when no significant residual tissue remained in the paraffin block. Our methodology is very similar to that described by Gong et al.,1 which is also recommended by the manufacturer of Mount Quick medium (New Comer Supply™, Middleton, WI.). In a pilot study, before full implementation at our institution, air-dried and fixed touch imprints were prepared from 12 fresh tumor and lymph node tissue samples. CTT was applied to Diff-Quik-stained (American Scientific Products, McGraw Park, IL) and Papanicolaou-stained touch imprints. IHC was simultaneously performed on the formalin-fixed tissues and their counterpart cell-transferred touch imprints. Approximately four to six cell-transferred pieces were prepared from each touch imprint. IHC of the touch imprints were performed with and without destaining, and with and without antigen retrieval. The following IHC stains were performed initially: pan cytokeratin(AE1/3), cytokeratin 7, cytokeratin 20, TTF1, CD45, CD20, and estrogen and progesterone receptors. The staining pattern of the formalin-fixed tissue was regarded as the gold standard. Our pilot study demonstrated 100% concordant IHC results between the formalin-fixed tissue and Papanicolaou-stained cell-transferred pieces treated with antigen retrieval. We found an approximately 15% false-negative rate when Papanicolaou-stained cell-transferred pieces were not subjected to antigen retrieval. Approximately 30% of air-dried cell-transferred pieces demonstrated false-negative findings and high background, nonspecific staining. Approximately 5–10% of cell-transferred samples that underwent destaining were lost or damaged during the process. In summary, the best IHC results were achieved on cell-transferred pieces obtained from fixed cytologic samples, treated with antigen retrieval, and not subjected to destaining. Since the inception of the pilot study, we have performed additional IHC stains on a variety of fine-needle aspiration and biopsy specimens with great success, including: cytokeratin 5/6, P63, S-100, HMB-45, melan-A, epithelial membrane antigen, CD3, CD15, CD30, CD117, polyclonal carcinoembryonic antigen, chromogranin, synaptophysin, prostate-specific antigen, calretinin, P16, P53, Ki-67, cyclin D1, WT-1, and CDX-2. In our experience, IHC performed on cell-transferred material is very accurate and comparable to results obtained on formalin-fixed tissue.
Read full abstract
Apr 1, 2006
Jun-Yi Lei + 3
Open Access
