The median age was 72, and the most common site was head and neck (3/7, 42.9%), consistent with the literature (22). However, females (4/7, 57.1%) slightly outnumbered males (3/7, 42.9%), contrary to the literature (23), which may be due to the limited number of our cases.
The histopathological features in our series, such as the positive correlation between PNI and infiltrative growth pattern, the higher mitotic rate in cases with lymph node metastasis, and the occurrence of ulcerated tumors in older patients, may be associated with aggressive features in MCC, consistent with the literature (24, 25).
Our results showed that 57.1% (4/7) of cases were positive for MCPyV by IHC, matching the literature which reports rates of 46–90% (26, 27). In our series, the significant relationship we found between RB1 and MCPyV expression supports the notion that RB1 mutations predominantly play a role specifically in the pathogenesis of MCPyV (-) MCCs, where loss of RB1 expression is expected. Additionally, it indicates that in the absence of MCPyV IHC, RB1 can be utilized as an alternative to infer the virus status in MCC cases.
PD-L1 was evaluated as positive in 50% (2/4) of MCPyV (+) cases and 57.1% (4/7) of the total cases in our series. MCPyV (+) tumors tend to respond better to PD-L1 inhibitors (28) and more frequently express PD-L1 (29). While MCPyV or PD-L1 status of the tumor does not directly influence treatment decisions under existing protocols (6, 30), larger cohorts and further studies are needed in this area.
The role of MSI in MCC is less defined than in colorectal cancer, but the growing importance of immunotherapy has brought MMR proteins and PD-L1 forward as potential biomarkers for research. Gambichler et al. found that 16% (9/56) of the patients in their series had a loss of MMR proteins. Among the five cases that underwent MSI testing, four were found to be MSS (microsatellite stable), while one was classified as MSI-H (microsatellite instability-high). The MSI-H case was a patient with loss of MLH-1 and PMS-2 expression (31). Recently, Kestel et al. studied 12 patients with MCC in Turkey and found that expression was intact in all cases (32). Our one case (1/7, 14.3%) showed loss of nuclear expression with MLH-1 and was subsequently identified as MSS through NGS analysis (Case 7).
PRAME was positive in 42.9% of the cases (3/7), with two showing diffuse and strong positivity and one showing heterogenous weak staining. PRAME is not only a diagnostic marker for melanoma but has also recently emerged as an immunotherapy target in uveal melanoma (33). Elsensohn et al. detected PRAME positivity in 57% of 23 MCC cases, with 9% showing strong positivity, while Miller et al. observed strong positivity in 27% of 39 cases (18, 34). Our series presents the first results on PRAME expression in MCC from Turkey, and investigating this marker in larger cohorts may be important for identifying suitable candidates for immunotherapy in the future.
Given the rarity of MCC, the unknown dominance of either sun exposure or MCPyV in its pathogenesis locally, the limited number of studies, and the advances in immunotherapy, we decided to conduct a comprehensive literature review to provide a detailed overview of MCC studies in Turkey. The largest cohort was from a multicenter study (35), but despite contacting the corresponding author, we were unable to determine the exact number of patients contributed each center. After excluding confirmed duplications from this study, the total number of MCC cases reported from Turkey, including our own, was estimated at 227 ± 46. To prevent duplication and conflicting data, this study was excluded from further analysis, and all subsequent evaluations for MCC in Turkey were based on 227 cases. The mean age was 64 ± 18.3 years (range: 8–94) in the Turkish population. Among the 216 cases with reported gender, there was a slight male predominance (50.9%), with a male-to-female ratio of 1.04:1, consistent with the literature (36). In 198 cases, tumor location was known and distributed as follows: head and neck (29.8%) (19–21, 37–48), lower extremities (22.7%) (19–21, 37–39, 49–59), trunk (21.2%) (19–21, 37–39, 49, 50, 60–68), upper extremities (18.7%)(19–21, 37–39, 49, 69–72), and other sites (7.6%). In the other sites, 13 cases had primary nodal involvement (19, 73), 1 was in the oral cavity (74), 1 was paravertebral (75), and 1 was in the parotid gland (76). We noted that 40% of the cases located on the trunk (and 9% of the total), including one of our cases, were in the gluteal region (21, 37, 38, 49, 50, 60, 61, 63, 66, 67). Among the MCC cases, one (1/227, 0.44%) showed features of SCC (60), one (1/227, 0.44%) had squamous cell carcinoma in situ (CIS) (19), and one (1/227, 0.44%) was associated with actinic keratosis (47), displaying characteristics of a mixed tumor, with in situ case being MCPyV (-), consistent with findings in the literature that associate MCPyV negativity more frequently with mixed tumors (77). There were synchronous tumors in two cases (2/227, 0.88%), one being pulmonary small cell carcinoma(64) and the other both SCC and basal cell carcinoma (BCC) (47). As a risk factor for MCC, 7.5% of the cases (17/227) had a history of secondary malignancy (19, 21, 37, 64, 78), with the major being CLL (21, 37, 78). Other etiologies included renal transplantation in four cases (4/227, 1.76%) (56, 57, 63, 79), liver transplantation in one case (1/227, 0.44%) (80), chronic renal failure in one case (1/227, 0.44%) (58), chronic venous insufficiency in one case (1/227, 0.44%) (81), and rheumatoid arthritis in one case (1/227, 0.44%) (82). The tumor size was known for 128 cases, with a mean tumor diameter of 3.75 cm (range: 0.5–20 cm). Lymph node metastasis was present in 24.7% of the cases (56/227) (19–21, 37, 39, 45, 49–52, 54, 56–60, 63, 65, 69, 74, 76, 81, 83), and distant metastasis in also 24.7% (56/227) (19, 21, 32, 39, 40, 49, 52–54, 56–58, 63, 66–69, 71, 74, 81, 84, 85). Seventy-two (51.1%) of the 141 patients with known survival status were alive, and 69 (49.9%) were deceased. One hundred and thirty-one patients with available follow-up data had a mean follow-up period of 36.2 months. Wide excision was performed in 127 cases, and chemotherapy and/or radiotherapy was given in 89 cases. There was no information on whether the cases received immunotherapy, possibly because it is a more recent treatment option. In one of our cases (Case 7), immunotherapy was planned in addition to radiotherapy due to the presence of distant metastasis.
Histopathological examination revealed that infiltrative growth pattern (69.7%) was more common that nodular pattern (30.3%) among 76 cases with known TGP (19, 20, 55, 70). LVI was evident in 45.7% of 105 cases (19, 20, 38, 52, 63, 82). TILs were present in 65.8% of 76 cases (19–21). CK20 was positive in 95.4% (19, 20, 37, 38, 40, 44, 46, 47, 51, 55, 58, 63, 64, 66, 67, 69, 70, 72, 73, 75, 76, 82, 83) of 132 cases, and negative in only 6 patients (19, 37, 60, 74). CK20 staining pattern was known in 95 patients, with 48.4% showing positivity in the classic perinuclear dot-like pattern (19, 20, 40, 46, 50, 51, 53, 56, 58, 63, 72, 75, 76, 82, 83), while the remaining cases exhibited cytoplasmic, membranous, or mixed patterns (19, 20, 47, 64, 66, 69, 73). MMR proteins were investigated only in two studies on MCC, including ours, where no loss of expression in 21 cases (21/22, 95.4%), while MLH-1 loss was observed in only one of our cases 1/22 (4.6%), which was further classified as MSS through NGS analysis (32).
This study is the first to provide insights into the rate of MCPyV-positive cases among MCCs in Turkey to date. MCPyV positivity rates show significant geographical variation, with 25–30% in Australia (86), 60% in Japan (87), 80% in North America (15), and 85% in Germany (88). The MCPyV status was available in approximately 40.1% (91/227) of MCC cases reported from Turkey, with an MCPyV positivity rate of 70.3% (64/91). Although this appears lower than the rates reported in European and North American countries, the MCPyV positivity rate of 70.3% in Turkey is slightly higher than the countries in Asia, such as Japan. Despite the lack of statistical significance (p = 0.13), there was a female predominance in the MCPyV (+) group (36/64, 56.2%), while males predominated in the MCPyV (-) group (17/27, 63%), consistent with the literature (89, 90). Although age differences by virus status had been reported in various studies from Turkey (19–21), overall, the ages in both groups were similar, with no significant difference (p = 0.84). However, in the broader literature, virus-negative patients tend to be older than virus-positive patients, although the age difference appears to have a limited impact on MCC pathogenesis (89, 90). MCPyV (+) MCCs were more commonly localized in the upper extremities (19/50, 38%) and trunk (10/50, 20%), while head and neck localization (14/25, 56%) were more prominent in the MCPyV (-) group in Turkey (p = 0.007), in line with the literature (90, 91). It can be concluded that MCPyV plays a more significant role in tumor development in regions less exposed to sunlight, such as the trunk and upper extremities, while the oncogenic effect of sun exposure becomes more prominent in the head and neck region, for the MCC population in Turkey. The gluteal location of MCC is reported rarely in the English literature (52). It is noteworthy that 9% (17/227) of MCC cases in Turkey were localized to the gluteal region. The mean age of these patients was 63.1 ± 13.6, with a male-to-female ratio of 1.5:1. In our series, the case located on the gluteal region was a 68-year-old male. While UV radiation plays a significant role in the pathogenesis of MCC, in gluteal cases, factors such as MCPyV, environmental influences, genetic predisposition, and skin type are expected to be more prominent. Notably, only one case (1/17, 5.9%) was presented with immunosuppression due to kidney transplantation (63), while the remaining cases (16/17, 94.1%) had no identifiable etiologic risk factors, including our case. Moreover, lymph node metastasis was found in 24% (4/17) cases and distant metastasis in 35% (6/17), suggesting that gluteal MCCs are aggressive in nature and may be associated with poor prognosis. In conclusion, considering MCC in the differential diagnosis of patients presenting with a gluteal lesion, along with a multidisciplinary approach, may lead to earlier diagnosis and more effective treatment.
Immunohistochemistry revealed a mutant p53 pattern (10/21, 47.7%) and loss of Rb expression (12/18, 66.7%) more frequently in MCPyV (-) tumors, whereas MCPyV (+) group showed more wild-type p53 (51/57, 89.5%) and intact Rb expression (43/54, 79.6) in MCCs cases reported from Turkey (p = 0.0004 and 0.0008, respectively). MCPyV (-) MCCs commonly show TP53 mutations resulting in dysfunctional p53 protein, and the RB1 gene is frequently inactivated through mutations or deletions, causing uncontrolled cell cycle progression. Conversely, in MCPyV (+) tumors, wild type p53 and RB1 are typically retained, as viral proteins such as large T (LT) antigen inactivate Rb protein, driving tumorigenesis without the need for mutations in tumor suppressor genes (89, 90, 92). There was no difference between virus-positive and negative groups regarding TGP (n = 74), LVI (n = 78), TIL (n = 76), and metastasis (n = 73) in MCC cases with known virus status reported from Turkey (n = 91, p > 0.05). However, MCPyV (-) MCCs are more likely to exhibit frequent LVI, an infiltrative growth pattern, lower levels of TILs, and a higher likelihood of metastasis, whereas the opposite is expected for MCPyV (+) tumors according to the literature (93, 94). No significant difference in CK20 expression was observed between MCPyV-positive and MCPyV-negative MCC cases from Turkey (n = 91, p > 0.05). Harms et al. reported that CK20-negative MCCs may arise through virus-independent molecular pathways and could harbor distinct mutations (95). The small number of CK20-negative cases (2/91, 0.2%) in the Turkish cohort with known virus status may explain the lack of significant difference.
The primary limitations of our study were the small sample size, due to rarity of MCC, and its retrospective design. The presence of MCPyV was evaluated only by IHC, and molecular methods such as polymerase chain reaction (PCR) were not utilized. While comparing MCC studies from Turkey, the lack of assessment of standardized parameters in each study was also a limitation.