Distinguishing subungual melanoma from non-nail apparatus acral melanoma is clinically important due to its distinct clinical courses including high recurrence rate and short progression-free survival [14, 15]. Also, it has been reported that subungual melanoma harbors more distinct genomic alterations including CARD11, ARID2, ARID1A, ARID1B, PTPRB, and PTPRK genes, compared with acral melanoma .
Cyclin D1 in coordination with their catalytic partners CDK4 and CDK6, contributes to promoting cell cycle progression through transition from G1 to S phase . Cyclin D1 upregulation with amplification of CCND1 has been documented in various malignancies including breast, lung, colon, and oral cancers . Meanwhile, the expression rate of cyclin D1 in cutaneous melanoma compared with benign melanocytic nevus has been discordantly reported [9, 20, 21]. When it comes to acral melanoma, cyclin D1 has been reported to be overexpressed, resulting in constitutively activated MAPK signaling pathway without NRAS or BRAF mutations [22,23,24]. In this regard, we had expected cyclin D1 IHC staining would be an effective discriminator of benign SMP from subungual MIS.
However, cyclin D1 IHC staining seemed not to be reliable in distinguishing benign SMP from subungual MIS, in this study. Positive nuclear immunostaining for cyclin D1 was found in 4 out of 14 (28.6%) patients with benign SMP and 3 out of 13 (23.1%) patients with subungual MIS using cutoff of 92.5% which was determined regarding sensitivities and specificities using Youden index, with poor sensitivity of 23.1%. The lack of diagnostic value of cyclin D1 overexpression in these lesions may be due to the differences in the type of antibody used, the positive cell count system, and cutoff point for positivity . Also, it is important to note that cyclin D1 overexpression does not always represent amplification of CCND1; loss of cyclin D1 IHC staining does not always represent loss of function of CCND1, vice versa. Epigenetic regulation including DNA methylation at cytosine and histone acetylation can cause alteration in mRNA and protein expression . Therefore, further genetic tests including fluorescent in situ hybridization (FISH) for CCND1 may be needed for determining true CCND1 amplification to discriminate between benign SMP and subungual MIS .
PRAME gene is a member of cancer testis antigen (CTA) gene family encoding a membrane-bound protein recognized by T lymphocytes, causing autologous cytotoxic T cell-mediated immune response . Except for some distinct tissues including testis, ovary, placenta, adrenals, and endometrium, PRAME is not detected in healthy human tissues . Overexpression of PRAME was found to inhibit retinoic acid (RA) mediated cell differentiation, cell growth arrest, and apoptosis, contributing to tumorigenesis via inhibiting RA receptor signaling . PRAME has been reported to be overexpressed in a variety of malignancies including malignant melanoma, showing an utility in distinguishing between benign and malignant lesions . Although little has been studied about the diagnostic utility of PRAME in subungual melanocytic lesions, a recently published paper has validated the usefulness of PRAME expression in differentiating between melanoma and other nail unit melanocytic lesions . In the paper, subungual melanomas were all positive in PRAME IHC, and benign melanocytic lesions of nail unit were all negative. However, out of the 25 melanoma cases, 20 cases were invasive melanoma, while only 5 cases were MIS. Our paper provides additional insight from current literature focusing on MIS, which is more difficult to distinguish from benign lesions.
In this study, we demonstrated that PRAME IHC staining was a relia4ble discriminator of benign SMP from subungual MIS. Positive nuclear immunostaining for PRAME was found in 1 out of 14 (7.1%) patients with benign SMP and 10 out of 13 (76.9%) patients with subungual MIS using cutoff of 10%, showing modest sensitivity of 76.9% and good specificity of 92.9%. Cutoff (10%) for positivity of PRAME in this study is quite different from previously reported cutoff values ranging from 50 to 75% [7, 10, 28]. Differences in cutoff values among studies may be due to divergences in IHC staining methodology and inter-observer variability in IHC assessment . Also, the type of melanocytic lesions included in the studies might affect the differences in cutoff value for positivity of PRAME. Unlike previous studies, we assessed the expression of PRAME using only subungual melanocytic lesions, known to have different genomic alterations with lower mutation burden compared with other cutaneous melanomas [23, 30]. Also, small sample size of this study might have influenced the difference in cutoff value. Therefore, additional studies of PRAME IHC in a large cohort of melanocytic lesions with different subtypes are needed to determine whether cutoff value for PRAME positivity should differ according to the subtype of melanocytic lesions, or not.
Compared with cytogenetic studies including FISH, IHC has several advantages including more rapid turnaround time, lower cost, and higher accessibility. However, PRAME IHC can exhibit false positive and false negative results, confusing the correct diagnosis. Lezcano et al.  reported that diffuse nuclear immunoreactivity for PRAME was found in 13.6% of cutaneous melanocytic nevus. Also, Shyam et al.  demonstrated focal immunopositivity of PRAME from 5 to 10% in atypical but benign non-spitzoid melanocytic proliferations. In this study, 1 out of 14 (7.1%) patients with benign SMP exhibited PRAME nuclear immunostaining ranging from 10 to 20%. The significance of focal PRAME positivity in benign melanocytic lesions is still unclear, which necessitate further studies to determine the potential risk of malignant transformation related to focal PRAME positivity. Additional cytogenetic and molecular tests can help avoiding overdiagnosis in the cases with focal PRAME positivity .
Also, it has been reported that primary cutaneous melanoma could be completely negative for PRAME immunostaining . In this study, 3 out of 13 (23.1%) patients showed completely negative PRAME IHC. Histopathological analysis revealed there was no melanophage adjacent to cutaneous melanoma in these patients. Melanophage found in cutaneous melanoma is indicative of immune responses, predicting a relatively good prognosis for patients, possibly through tumor regression due to phagocytosis of melanoma cells . As PRAME gene encodes a membrane-bound protein recognized by T lymphocytes causing autologous cytotoxic T cell-mediated immune response , it is possible that lack of PRAME expression may lead to decreased immune response and reduced melanophages. While PRAME was found to be related with metastasis in uveal melanoma , little has been reported about the prognostic value of PRAME expression in cutaneous melanoma. Further studies are needed to determine the prognostic value of PRAME expression in cutaneous melanoma, regarding the role of PRAME in the relationship between immune response and oncogenesis.
It is important to keep in mind that IHC should be used in conjunction with clinical findings and histological analysis to make a final diagnosis. Clinically, we found that the patients with subungual MIS tended to be older and have wider lesions than the patients with benign SMP, with statistical significance. We also demonstrated that the patients with subungual MIS showed significantly more confluency, pagetoid melanocytosis, and severe atypia than the patients with benign SMP. Combining these clinical and histopathological features with PRAME IHC may significantly increase diagnostic power to distinguish benign SMP from subungual MIS.
This study has several limitations. First, this study was a retrospective study with relatively small sample size. Small sample size limited the usage of multivariate logistic regression analysis to develop a scoring system for the diagnosis of subungual melanoma in situ using PRAME expression. Second, this study was a single-center study confined to Korean patients. As the difference in the prevalence rate of subungual melanoma between the Asian group and other races including Western group is prominent enough to suppose difference in the biology of subungual melanoma in each group. Therefore, cyclin D1 and PRAME expression in melanocytic lesions of nail unit should be studied in other races including Western group. Also, this study evaluated no cytogenetic or molecular studies to compare with PRAME IHC in the diagnosis of subungual MIS.