Search

Objective: To seek the optimal melanin-removal method for hematoxylin and eosin (HE) staining, immunohistochemistry and molecular detection. Methods: Thirty-eight paraffin tissue samples of malignant melanoma diagnosed at the Fujian Cancer Hospital, Fuzhou, China between January 2018 and March 2022 were collected and used to make a tissue microarray. Melanin in these cases was removed using warm hydrogen peroxide, double oxidation depigmentation, modified potassium permanganate-oxalic acid or trichloroisocyanuric acid, followed by HE staining. The cases were divided into two cohorts: one was subject to the one of the above four methods to remove melanin first, followed by immunohistochemistry (SOX-10, Ki-67, HMB45 and Melan A), while the other was subject to immunohistochemical staining first and then a melanin removal. Following that, seventeen melanin-rich paraffin tissue samples were collected and depigmented using the methods described above. DNA extraction was then done, followed by assessments of DNA content and quality. Moreover, the completeness of melanin removal, the effect on HE and immunohistochemical staining, and the quality of DNA were compared between the depigmented methods. Results: Regarding the effectiveness of melanin removal, the modified potassium permanganate-oxalic acid and the warm hydrogen peroxide methods were the most effective, and both showed residual melanin in only 5.26% (2/38) of the cases. The trichloroisocyanuric acid method showed residual melanin in 10.53% (4/38) of the cases. The worst was the double oxidation depigmentation method, which showed pigment residue in 15.79% (6/38) of the cases. For HE staining, the percentage of good staining with the warm hydrogen peroxide method was 92.11%, higher than the other three methods. For immunohistochemical staining, the mean staining scores of immunohistochemistry first followed by melanin removal with modified potassium permanganate-oxalic acid, double oxidation and trichloroisocyanuric acid were 20.84, 26.63 and 35.02, respectively. These immunohistochemical staining scores were higher than those of melanin removal first followed by immunohistochemistry (8.70, 15.41 and 21.22, respectively). The mean staining score of melanin removal by warm hydrogen peroxide method followed by immunohistochemistry was 33.57, superior to that of immunohistochemistry followed by the melanin removal (19.96). Moreover, the staining scores of HMB45, MelanA and Ki-67 with immunohistochemical staining followed by trichloroisocyanuric acid method were 36.45, 33.79, and 36.24, respectively, while the staining score of SOX10 with melanin removal by warm hydrogen peroxide followed by immunohistochemistry was 34.39. The DNA was significantly degraded by modified potassium permanganate-oxalic acid, double oxidation depigmentation and trichloroisocyanuric acid, whereas the mean concentration of DNA extracted after melanin removal by hydrogen peroxide method was 59.59 μg/L, substantially higher than that of DNA extracted without melanin removal (30.3 μg/L, P =0.001). The A 260 / A 280 of DNA extracted after melanin removal by hydrogen peroxide was between 1.8 and 2.0 in all cases, and the A 260 / A 230 was above 2.0 in sixteen cases, suggesting high purity of DNA. However, the DNA extracted without removing the melanin showed poor purity, with A 260 / A 280 below 1.8 in eight cases and A 260 / A 230 below 2.0 in sixteen cases. Conclusions: Warm hydrogen peroxide showed the least melanin residue, superior HE staining and a minimal effect on DNA purity/quality compared to the other three methods. It thus appears most suitable for PCR, NGS and other molecular detection. Melanin removal with trichloroisocyanuric acid after immunohistochemical staining has the least melanin residual, and thus could be the most convenient and efficient. However, it is noted that the efficacy of the same depigmentation method varies with different antibodies. Therefore, the optimal depigmentation method should be selected based on the specific markers of interest.