ENKTL originates from mature T and NK-cell lymphoma, which is rare, and regional and ethnic differences in the disease have been noted. It has been rarely reported in North American and European countries but is common in East Asian and South American countries, especially China , accounting for 15% to 30% of all lymphomas in China. However, regardless of whether the involvement of the skin is the primary or secondary manifestation of ENKTL, cutaneous ENKTL is extremely rare and has a poor prognosis .
Cutaneous ENKTL is more frequent in male individuals than in female individuals, and occurs frequently in middle-aged adults. In this study, 83.3% of the patients were men, and the male-to-female patient ratio was 5:1, which may be attributed to the limited size of the sample. The median age of the patients was 64.5 years, and 66.7% of the patients were older than 60 years, indicating a slightly higher proportion of elderly individuals in comparison with previous studies. In recent years, the published literature on cutaneous ENKTL is mainly focused on cases of primary cutaneous ENKTL, mostly in the form of case reports or reported in small series, while reports of nasal ENKTL with cutaneous involvement are even rarer.
In this study, among all subsets of cutaneous ENKTL, primary cutaneous ENKTL was the most common. All forms of nasal ENKTL with cutaneous involvement present within 2 years after the initial diagnosis of nasal ENKTL , while the mean time in our group was 15 months (range : 2 to 30 months). Nasal ENKTL with secondary spread to the skin was more likely than primary cutaneous ENKTL to present with generalized skin lesions . The distribution of cutaneous ENKTL in the extremities was as high as 50.0%, followed by the trunk or head and neck. In this study, the clinical macroscopic findings of cutaneous ENKTL were characterized by erythema, papules, subcutaneous nodules, and ulceration, which was similar with the previous reports by Liang et al . In all such patients, lesions were more common in the lower extremities, which may be related to T-cell homing . Furthermore, the cutaneous lesions of the lower extremities were more likely to progress to ulceration. The causes of ulceration formation are consistent with the characteristics of tumor cells destroying blood vessels and secondary ischemic necrosis in histopathology. These groups of cutaneous ENKTL showed no remarkable differences in terms of the clinical features of the cutaneous lesions, except in terms of stage.
More than half of the patients with primary cutaneous ENKTL developed extracutaneous lesions several months after the appearance of skin lesions. However, in this study, involvement of the lymph nodes, nasal cavity, or sinus and lungs was observed by PET/CT in patients showing skin lesions, while bone marrow involvement was rare. Therefore, a full diagnostic workup is required when skin lesions are found, and attention should be paid to lesions outside the skin, which have a certain influence on disease evaluation and prognosis judgment. Moreover, the results of this study showed B symptoms in approximately 66.6% of the patients, which can occur at any time in the disease course. Many studies have shown that B symptoms are a poor prognostic factor for ENKTL [9-10].
In this group, histopathological examinations typically showed polymorphous diffuse and angiocentric lymphoid infiltrate involving dermis and subcutaneous region, in association with angiodestruction and coagulative necrosis. The tumor cells showed irregular folded nuclei. In addition, cells size ranges from small to large, but most patients mainly showed medium-sized cells or a mixture with many types of cells. Histopathological assessments are of great significance to the diagnosis of ENKTL, but they are difficult to evaluate. Besides, the clinical manifestations of cutaneous ENKTL at an early stage are not typical, almost all patients were misdiagnosed with inflflammatory lesions at onset, which can easily lead to missed diagnosis and misdiagnosis. Therefore, detection based on immunohistochemistry and EBER in situ hybridization detection should be performed at an early stage. If necessary, multiple biopsies are recommended to improve accurate diagnosis of cutaneous ENKTL.
Immunohistochemical findings for this disease often show the neoplastic cells are positive for CD2, CD3ε, cytotoxic protein (TIA-1, perforin, and GranB) and CD56, frequently negative for other T-lineage markers and B-cell antigen (CD20). CD56 is a marker for NK cells, expressed in a subset of CD4 and CD8 cells, and is positive in 74% to 76% of cases of ENKTL . In accordance with published literature, there are no significant differences in the clinicopathological features between CD56 positive and CD56 negative cases . Therefore, for rare CD56 negative cases, the detection of EBV and expression of cytotoxic proteins are required for diagnosis. The high expression levels of Ki-67 are related to the volume of the primary lesion. Large tumors (volume >10 cm3) may have a higher ability to show tumor cell proliferation, and their prognosis is significantly worse than that in the low-expression group, which can be used as one of the prognostic indicators of ENKTL.
CD20 is a specific marker of B-cell lymphoma that serves as a target for therapeutic monoclonal antibodies for the treatment of B-cell lymphomas and leukemias. CD20 expression in T-cell lymphoma has been gradually reported in recent years. However, abnormal expression of CD20 in ENKTL is very rare, mainly occurs in extranasal ENKTL, and is associated with an advanced and poor prognosis. Abnormal expression of CD20 in ENKTL suggests a high proliferation rate, similar to the high expression level of Ki-67 in tumor cells . But now, the cutaneous lesions of three patients in this group were CD20-positive, as well as high expression levels of Ki-67 (60%-90%), and they were all in advanced stage, indicating a worse prognosis. It is difficult to distinguish abnormal expression of CD20 in ENKTL from B-cell lymphoma with exceptional expression of T cell markers. 3 cases of CD20 abnormal expression of ENKTCL patients expressed multiple NK/T cell markers, and B cell markers only expressed CD20. On the contrary, in addition to weakly expressing PAX-5, B-cell lymphoma cells with exceptional expression of T-cell markers showed diffuse expression of CD20, CD79a, CD19, BOB-1 and other B-cell markers. In short, CD20 increases the difficulty of diagnosis, so diagnosis requires a combination of multiple tests pathology, immunohistochemistry, EBER, and gene rearrangement.
Interestingly, One patient in this group showed CD20-positive primary nasal lesions. With the progression and spread of the disease, lesions of the penis and scrotum showed weak positivity for CD20. Finally, cutaneous lesions showed a CD20-negative status. In contrast, one case of primary nasal ENKTCL showed acquisition of CD20 expression in cutaneous relapse. Immunohistochemistry of the primary and relapsed lesions in these two cases of ENKTL with secondary spread to the skin showed only discordant CD20 expression, which may be due to tumor transformation of progenitor cell subsets that co-express CD20 and NK-cell markers, or the neoplastic process after neoplastic transformation, with the latter seeming more plausible . Nevertheless, the underlying mechanism remains to be further studied.
EBER are EBV-encoded small mRNA expressed at all stages of EBV infection. All nine patients were EBER-positive, further indicating that EBV infection was closely related to the incidence of ENKTL. Moreover, there was no difference in histopathology and immunophenotypic changes in nasal and extranasal ENKTL, but the detection rate of the former EBV was higher than that of the latter . The pathogenesis of EBV tumors is different between elderly and young patients may be related to the immune degradation caused by aging . However, EBV-negative results do not exclude the possibility of an ENKTL diagnosis. EBV-negative ENKTL is a rare subtype with limited available data. Previous studies have reported EBV-negative patients showed an inert clinical outcome and had a better prognosis than those with positive patients. But Nicolae et al  showed that EBV-negative patients was indistinguishable clinically and pathologically from positive patients, with a similar fulminant clinical course. By the way, measuring circulating EBV DNA can be used to determine a treatment algorithm and serve as an objective criteria for follow-up of their apeutic efficiency.
Currently, there is no consensus regarding the proper treatment strategy, and the main treatments for ENKTL patients are chemotherapy, radiotherapy, and chemotherapy combined with radiotherapy. For localized disease (I/II), radiotherapy alone has a good short-term effect, but the rate of late recurrence and metastasis is higher in patients receiving this treatment . Combined therapy can be administered as simultaneous chemoradiotherapy or sequential chemoradiotherapy, both of which have good effects on the treatment of early ENKTL patients. However, patients with advanced stage (III/IV), relapse, and refractory disease mainly receive systemic chemotherapy alone, since the target lesions are extensive and may not be suitable for local radiotherapy. Avilés et al.  compared the prognosis of ENKTL patients treated with radiotherapy alone, chemotherapy alone and combined chemoradiotherapy, and found that complete response(CR), 5-year progression-free survival (PFS) and 5-year OS of combined chemoradiotherapy were all higher than those treated with radiotherapy alone or chemotherapy alone.
To date, there is no standard chemotherapeutic protocol for ENKTL. Conventional CHOP regimens (anthracycline-containing) was confirmed ineffective to ENKTL . The present study found that chemotherapy regimens based on pegaspargase (PEG-ASP) or L-asparaginase (L-ASP) yielded promising results in the treatment of ENKTL. However, L-ASP has many adverse reactions, limiting its clinical application, while PEG-ASP can effectively preserve the original activity of L-ASP with few adverse reactions. PEG-ASP can be combined with gemcitabine and chidamide in chemotherapy, and in recent years, the latter has achieved significant efficacy in the treatment of ENKTL. The use of radiotherapy and chemotherapy combined with autologous hematopoietic stem cell transplantation (ASCT) has been demonstrated to improve treatment efficacy [20-21]. In this group, one patient with diffuse skin lesions received ASCT after complete remission of chemotherapy, and survived for 31 months without tumor recurrence, showing significant efficacy. However, the number of cases of HSCT for ENKTL are relatively rare, and its clinical efficacy remains to be further studied. In addition, some new therapeutic methods, such as biological targeted therapy and cellular immunotherapy, are still being explored to provide new therapeutic directions for ENKTL.
Primary cutaneous ENKTL has been reported to be less aggressive and to show a better prognosis than nasal ENKTL with cutaneous involvement, and patients with single lesions had lower mortality and better prognosis than those with multiple lesions, while the prognosis of primary cutaneous ENKTL with secondary nasal lesions was not significantly different from that of nasal ENKTL [7,22]. However, the mortality and prognosis of patients with localized and multiple cutaneous lesions in this group were slightly different from those in previous studies, which may be attributed to the limited number of samples collected or the improvements associated with the current advanced treatment regimens.
Previous studies have shown that the median survival time ranges from 2 to 15 months for patients with cutaneous ENKTL in most series, and the estimated 5-year survival of 0% . Jiang L et al  reported that the 3-year OS rates of cutaneous ENKTL were 73.9 % for patients who achieved complete response (CR) compared with 10.3 % for patients who did not. Takata K et al  found that the 5-year overall survival rate of primary cutaneous ENKTL was 25%. However, the prognosis of patients in this group was poor and the 2-year OS was only 27.3%, and we would continue to follow up the patients. On the other hand, untreated patients had a 75% mortality rate overall, while one 92-year-old untreated patient with local recurrence is still alive and has lived for 35 months. The exact cause is unclear, but it could be the microenvironment of senescent tissue is less capable of supporting rapid tumor growth .
In summary, cutaneous ENKTL is rare in clinical practice, and is characterized by high invasiveness and a poor response to chemotherapy. For cases in which clinical symptoms do not match the signs, patients show progressive aggravation, or long-term treatment is not curative, especially with B symptoms and lymph node enlargement, we should be highly alert to the possibility of this disease. Timely histopathological and immunohistochemical examinations, early diagnosis, and treatment can improve the prognosis of patients.