In this study, reviewing 54 MCC cases, MCPyV and CK20 antibody clones used in the diagnosis were evaluated. MCPyV clone Ab3, although relatively understudied in the literature, demonstrated a higher positivity rate and was easier to evaluate compared to CM2B4. CK20 clone SP33, which has not been previously reported in MCC series in the literature, demonstrated aberrant staining in areas of necrosis. We observed no difference between clones SP33 and Ks20.8 in their ability to detect positive cases.

MCC, a rare and aggressive neuroendocrine skin carcinoma, has an etiology linked to MCPyV and sun damage, with notable geographical variations [33]. In our series, MCPyV positivity was observed in nearly 80% of cases, a rate comparable to those reported in Europe, where MCPyV is accounted as the major etiologic agent and may suggest that MCC in our region is more influenced by MCPyV than sun damage [24, 34, 35].

For MCPyV detection, a gold standard method has not been established. Since its discovery by Feng et al. in 2008, PCR-based techniques have been widely used due to their sensitivity in detecting MCPyV [22, 36]. It has been demonstrated that MCPyV contributes to tumorigenesis only after integrating into the host genome and undergoing specific mutations in MCC [37]. However, MCPyV has also been detected in non-MCC skin tumors and other malignancies where it does not play a role in pathogenesis, raising concerns about the specificity of PCR in these cases [22, 25, 38, 39].

The MCPyV antibody clone CM2B4, introduced in 2009, showed no reactivity in non-MCC tumors according to current studies [22, 25]. The lack of false positivity with CM2B4 (MCPyV PCR-negativity in CM2B4-positive tumors) supported the clinical utility of this antibody in the distinction of MCPyV-related MCC [19, 20, 29]. However, the sensitivity of this antibody (CM2B4) was slightly lower, as it identified approximately 70% of PCR-positive cases [19, 20, 40]. In 2012, the newer MCPyV antibody clone Ab3 was designed to enhance sensitivity [27].

In their initial study, Rodig et al. evaluated two MCPyV antibody clones in 57 cases and emphasized that Ab3 exhibited significantly greater sensitivity [27]. With the new clone, they detected 9 additional cases that CM2B4 had missed. However, CM2B4 remains commonly used in studies and routine practice, while reports on Ab3 are limited [21, 29, 30, 41]. The largest series comparing two antibody clones reported 90% positivity with Ab3 and 70% with CM2B4 [21]. In our series, approximately 80% of cases were positive with MCPyV clone Ab3 and 70% with clone CM2B4. Ab3 detected 5 additional cases that CM2B4 had missed, consistent with previous findings. In studies conducted to date, no cases have been reported that stained with CM2B4 but not with Ab3, as observed in our study.

In terms of specificity, studies indicate that both clones can occasionally exhibit immunoreactivity in non-neoplastic components [21, 36, 40]. With Ab3, a small subset of non-MCC skin cancers demonstrated focal and weak staining [21, 30]. In our study, although we did not test the antibodies on non-MCC tumors, we did not observe any aberrant staining in tumor surrounding tissues with either clone.

The extent and intensity of staining differ between the two clones. Ab3 has been reported to show higher extent and intensity, facilitating evaluation, while partial or weak staining has been reported in up to 25% of CM2B4-positive cases in literature [19, 26, 27, 41]. In our series, approximately 30% of CM2B4-positive cases showed low percentages or weak staining, complicating interpretation. Additionally, two of three cases with partial or weak Ab3 staining were negative with CM2B4. Similar findings of focal and weak staining with Ab3, yet negativity with CM2B4, have also been reported [30]. One study even described certain CM2B4 staining results as “uninterpretable” [20]. Based on these findings and our experience, we consider Ab3 to be easier to evaluate and more reliable than CM2B4, with higher sensitivity supported by current evidence.

Another finding that may vary geographically and etiologically in MCC is the expression status of CK20. CK20 plays a fundamental role in the diagnosis of MCC, and its perinuclear dot-like staining pattern is a distinctive and diagnostic feature for MCC [17]. However, it has been reported that approximately 10% of cases, particularly those negative for MCPyV, are CK20 negative [17, 18, 20, 22, 23, 42]. In our study, there were only two CK20 negative cases, which were also MCPyV negative. A study using next generation sequencing reported that CK20 and MCPyV-negative cases exhibit an ultraviolet (UV)-signature mutational profile [42]. Similarly, combined neuroendocrine carcinomas, which are mostly associated with squamous cell carcinoma, are MCPyV-negative and show a UV-signature mutational profile [1, 19]. Likewise, one of our MCPyV negative cases had Bowen disease overlying MCC indicating a UV radiation exposure.

Considering the key role of CK20 in MCC diagnosis, we investigated whether CK20 negativity might be influenced by factors beyond etiology. For this, we compared the commonly used Ks20.8 clone with the SP33 clone. To date, no studies have compared CK20 clones in MCC. In our series, two cases were negative with both clones, and no significant difference was observed in detecting positive cases. Diffuse staining was observed in a similar proportion of cases with both clones, although Ks20.8 showed stronger staining in about 40% of cases.

It is known that the typical DL staining pattern of CK20 in MCC can be accompanied by CS and M staining [19, 43]. In one-third of our cases, staining patterns differed between clones. Among these, Ks20.8 revealed additional CS staining in 40% and M staining in 60%. These variations highlight subtle enhancements between the clones. DL staining can be challenging to interpret, especially when weak or focal. In our series, we identified two cases where the DL staining was missed and incorrectly reported during routine practice (data not shown; author’s observation). This underscores the importance of selecting clones that produce sharper, more reliable staining to ensure accurate diagnosis.

We also noticed that SP33 stained necrotic areas in addition to viable tumor cells in 15 of our cases. In the same areas, no aberrant staining was observed with Ks20.8. Necrosis, observed in 70% of cases in our study, is a common feature of MCC due to its high grade. This aberrant SP33 staining could complicate evaluations, especially in resection specimens after neoadjuvant immunotherapy, where CK20 may play a role [44]. These findings highlight the need for careful consideration of clone selection.

This study has several limitations. The antibodies were only tested on skin and lymph nodes, and not on other tissues or non-MCC tumors. We also did not compare MCPyV antibody performance with PCR, which could have provided insights into their diagnostic accuracy. Lastly, our sample size may not fully represent the variability seen in routine practice. Future studies with larger cohorts and clone comparisons are needed to validate and expand these findings.

In conclusion, this study highlights the importance of selecting appropriate clones of MCPyV and CK20 antibodies in diagnosing MCC to enhance diagnostic accuracy. MCPyV antibody clone Ab3 demonstrated superior sensitivity and ease of interpretation compared to CM2B4, confirming its value in routine practice. While both CK20 clones Ks20.8 and SP33 showed comparable diagnostic performance in detecting positive cases, Ks20.8 exhibited stronger and more consistent staining patterns, which may aid in distinguishing challenging cases. The aberrant staining observed with SP33 in necrotic areas raises concerns, particularly in specimens from patients undergoing neoadjuvant therapy. These findings emphasize the need for further comparative studies to establish consensus on the optimal antibody clones for routine use in MCC diagnosis.