Primary Malignant Melanoma of the Cervix: An Integrated Analysis of Case Reports and Series

Melanoma, also known as malignant melanoma, is a type of malignant tumour that originates from melanocytes in the basal layer of the epidermis. Primary malignant melanomas of the female genital tract are rare. Similarly, primary malignant melanoma of cervix, which originates from cervical melanocytes, is an extremely rare disease and the second most common type of female melanoma in women aged between 15 to 44 years worldwide. To date, primary malignant melanoma of the cervix is characterized by poor patient prognosis and little consensus exists regarding the best treatment therapy. The situation is worsened by lack of clinical studies with large samples. Notably, surgery remains the preferred treatment option for patients with primary malignant melanomas of the cervix. Current treatments are based on Federation International of Gynecology and Obstetrics(2018) staging with reference to National Comprehensive Cancer Network guidelines. This study is in order to find a more suitable treatment modality for primary malignant melanoma of cervix. Therefore, we first conducted an integrated analysis of case reports and series to assess the impact of various factors on the prognosis of such patients. In summary, this is the first pooled analysis including 149 cases of primary cervical melanoma. We found that patients who underwent radical hysterectomy-based surgery, those with non-metastatic lymph nodes and those who underwent lymphadenectomy had significantly higher survival rates. In patients who had RH-based surgery, survival rates at the 24m time point of those who did not add other treatments was higher than those who did, but for those who had total hysterectomy-based surgery, the addition of other treatments to prolong median survival may be considered. In the overall analysis, age and lymphadenectomy were associated with increased and reduced risk of death in these patients, respectively. Although there is no statistical difference, stage III&IV, TAH, lymphatic metastases increase the risk of death; whereas radical hysterectomy was associated with reduced risk of death. In the subgroup analysis, for patients who have undergone radical hysterectomy-based surgery, lymphadenectomy reduces the risk of death, while lymphatic metastases and complementary other treatments increase the risk of death. For patients who have undergone total hysterectomy-based surgery, complementary treatment reduces the risk of death. In conclusion, via summarizing previous reports, the recommended treatment procedure for PMMC are radical hysterectomy and lymphadenectomy. The addition of other treatment options for patients who undergoing RH-based surgery need further study.


INTRODUCTION
Melanoma, also known as malignant melanoma (MM), is a type of malignant tumour that originates from melanocytes in the basal layer of the epidermis (1). It is considered one of the most aggressive cancer diseases due to its high malignancy and treatment resistance (2). Notably, MM patients have a survival period of less than 5 years (3), although they have very good prognosis if the disease is detected in its early stages (4). Globally, disease accounts for approximately 0.03% of all newly diagnosed cancers (5). Previous studies have shown that the high mortality rate is due to the aggressive metastatic potential of melanoma cells (6). Previous estimates have shown that approximately 132,000 and 48,000 new malignant melanoma cases and deaths, respectively, occur each year (7). In fact, MM is ranked the fifth most common cancer in the world, and its incidence is on the rise (8). Primary MMs of the female genital tract are rare, accounting for only 3-7% of all mucosal melanomas (9). On the other hand, primary malignant melanoma of the cervix (PMMC) is an extremely rare disease that originates from cervical melanocytes (10). It is the second most common type of female melanoma in women aged between 15 to 44 years worldwide (11).
Melanoma is a highly malignant tumor with a poor prognosis (12), while that of PMMC worse (13). To date, little consensus has been reached regarding the best treatment therapy for PMMC. Lack of clinical studies involving large sample sizes has also contributed to scarcity of optimal treatment options. Nevertheless, surgery remains the preferred treatment option (13,14). The current treatments are based on FIGO staging with reference to NCCN guidelines (15). This study is in order to find a more suitable treatment modality for PMMC. Therefore, we first conducted an integrated analysis of case reports and series to assess the impact of various factors on the prognosis of such patients.

Data Sources and Search Strategy
In this section, we describe a comprehensive analysis based on published case report data. Briefly, we searched three public databases, namely PubMed, EMBASE and the Cochrane Library, for case reports published from their inception until 31st January 2022. The search strategy involved the following terms: (((Cervixes) OR (Uterine Cervix)) OR (Cervix, Uterine)) OR (Cervix)) AND (Primary malignant melanoma). In addition, we performed a secondary search of the reference lists across relevant articles to identify additional eligible reports. Two independent reviewers (Aiping Min and Aizhen Fu) conducted the literature search, study selection and data extraction. Any discrepancy between them was resolved through consensus and arbitration by a third author (Meiyuan Huang).

Study Selection Criteria
All studies describing primary malignant melanoma of the cervix, with clinical information on the patients regardless of sample size, were included in this study. Conversely, articles that met the following criteria were excluded from the study: Primary melanoma of other genital tract origin, genital tract melanoma of undeterminable origin, reviews, books, as well as irrelevant. In cases where studies had overlapping data, we selected the study with the larger sample size.

Data Extraction
The following data were recorded: year of publication, patient's age at diagnosis, patient's symptoms, FIGO stage, whether surgery was performed, mode of surgery, presence or absence of lymphatic metastases, use of other treatment modalities and overall survival (OS) in months. If the patient's clinical information was mentioned in another article, and the corresponding full text was not available, we marked the survival status with UNCERTAIN.

Outcomes and Statistical Analysis
Outcomes assessed were age and OS. The definition for survival time was based on data from the individual participants reported. Missing and unidentifiable data were specified as NA, thus were not included in the statistical analysis. Categorical data were presented as frequencies and percentages. On the other hand, continuous data that conformed to normal distribution were presented as means and standard deviations (SD), while non-normally distributed data were presented as medians (range). In addition, Kaplan-Meier survival curves were generated to estimate OS of patients, and differences across subgroups compared using the Log-rank test. Statistical analysis was performed using the SURVIVAL package implemented in R language (16). A multivariate analysis of these predictors of survival for overall and subgroups was conducted using a Cox proportional risk model. Data followed by P<0.05 were considered statistically significant.

Search Results and Eligible Studies
The aforementioned search strategy resulted in a total of 233 articles across the screened databases. Finally, 113 articles, containing 149 cases, were found to be eligible and therefore included in the analysis. A summary of the main features of the included studies is outlined in Table 1, whereas the procedure for literature search and selection is shown in Figure 1.    IB1  RH+PLND  NA  no  87  0  50  VB  IB2  TAH+PLND  NA  C  16  1  58  VB  IB1  RH+PLND  NA  C  35  0  57  VB  IIA1  No  NA  C  4  1  42  VB  IIB  TAH  NA  R  9  1  63  VB  IIIB  No  NA  C+R  10  1  54  VB  IIB  RH+PLND  NA  C+R  33  1  78  VB  IIIB  No  NA  C+R  12  1  68  VB  IIA1  RH+PLND  NA  no  20  1

Patient Prognosis and Survival Rates
A total of 99 patients had information on survival time and status, of whom 39 survived (5-193 months) while 60 died. Corresponding information was not available for 32 patients, whereas the survival status of 18 patients could not be determined owing to a lack of full text. We generated Kaplan-Meier survival curves to evaluate OS of the patients, and obtained a median OS of 18 months in the cohort ( Figure 2). Next, we employed log-rank tests under stratified covariates (FIGO stage, Surgery or not, the extent of surgery, Lymphatic metastases or not, Lymphadenectomy or not, Supplementary treatment or not), to explore underlying factors that may affect patient prognosis. Results indicated that prognosis of patients significantly decreased with stage progression (P=0.00069; Figure 3). Notably, patients who did not have surgery had significantly worse prognosis compared to those who had RHbased and TAH-based surgery (P<0.0001), while those who had RH-based surgery had better prognosis than those who had TAH-based surgery ( Figure 4). Moreover, patients with nonlymphatic metastases had higher median survival times than those with lymphatic metastases, albeit with no statistical significance (P=0.056; Figure 5). Patients who underwent lymphadenectomy exhibited better prognosis than those who did not undergo the procedure (P<0.0001; Figure 6). Next, we compared surgical results (RH+TAH, RH, TAH) with and without other treatments, and found that in patients who underwent surgery(RH+TAH), there was no statistically significant differences in patient prognosis between groups with or without the addition of other treatment modalities (P=0.81; Figure 7). However, patients who added chemotherapy and others treatments(oncolytic virus Rigvir ® , argon laser, Argon heliumknife) had significantly longer median OS than those who  added the remaining treatment modalities. The exception to this was in patients who added radiotherapy or immunotherapy, where survival rates at the 60m time point were significantly higher than others ( Figure 8). Moreover, in patients who had RH-based surgery, we found survival rates at the 24m time point of those who did not add other treatments was higher than those who did (Figure 9; P=0.18). In order to exclude possible confounding effects due to inhomogeneous distribution of characteristics between patients with RH and patients with RH +T, we tested if the stage, LS, LM are the same ( Supplementary  Table). The results found no significant difference between the two groups for Stage, LS, but a significant difference for LM. Similarly, patients who had chemotherapy or other treatments (oncolytic virus Rigvir ® , argon laser, Argon heliumknife) exhibited significantly better median OS than those who had the remaining treatments. The exception was observed in patients who underwent radiotherapy, immunotherapy or with no other treatment added, whose survival rates at the 54m were significantly higher than others (Figure 10; P=0.14). In patients who had TAH-based surgery, we found no statistically significant differences in prognosis between the group with or without addition of other treatment modalities. However, the former group had higher median survival times than the latter (Figure 11; P=0.066). Furthermore, patients who added chemotherapy+radiotherapy exhibited significantly longer OS than those who added the remaining treatment modalities. The  Results of multivariate Cox hazard model analysis revealed that age, stage III&IV, TAH and lymph metastasis increased the risk of death, whereas RH and lymphadenectomy was associated with reduced risk of death ( Figure 13). In both the univariate and multifactorial cox regression risk models, only age and lymphatic resection showed consistency and could therefore be used as independent prognostic factors ( Table 2). For patients who have undergone RH-based surgery, lymphadenectomy reduces the risk of death, while lymphatic metastases and complementary other treatments increase the risk of death ( Figure 14). For patients who have undergone TAH-based surgery, lymphadenectomy seems to have little effect, while complementary treatment reduces the risk of death ( Figure 15).

DISCUSSION
Primary malignant melanoma of the cervix is an extremely rare disease. According to Norris and Taylor (129), cervical melanoma can be diagnosed based on four criteria, namely: presence of melanin in the normal cervical epithelium; absence  of melanoma elsewhere in the body; junction changes in the cervix; and metastasis following a pattern of cervical cancer. The disease is characterized by poor patient prognosis, especially if it is not detected in time or treated correctly (95). Previous studies have shown that the 5-year OS for patients with this cancer is approximately 10% and many patients die within three years of diagnosis (87.5%) (13). However, the present study revealed contrasting results. as evidenced by a 5-year OS of 27% and death rate of 65% within 3 years. We attribute this to the continuous improvement in surgery and other treatment modalities. The main prognostic factor for cervical melanoma is the FIGO stage at the time of diagnosis (13). This is consistent with the results the present study, which revealed that median survival time gradually decreased with increasing FIGO stage. Although no standard treatment modality has been developed for this condition, RH with pelvic lymph node dissection, partial vaginectomy remain the first-choice therapy for patients suitable for the procedure (130,131). Results of the present study indicated that RH-based surgery did improve patient survival times, which were significantly better than those of patients who had TAH-based surgery and those who did not have surgery. However, a total hysterectomy was performed in some cases.     Adjuvant pelvic radiotherapy may be considered for patients with positive surgical margins, parametrial involvement or histologically positive nodes. On the other hand, patients who are not suitable for radical surgery may be subjected to definitive external pelvic radiotherapy with/without brachytherapy, primarily for palliative purposes (131). Although our results were consistent with this conclusion, we believe that adjuvant other treatments are counterproductive in patients who have   undergone RH-based surgery. This conclusion is contrary to common sense. It may be related to the following reasons:1.The difference in LM numbers between the two groups was significant, while the median survival time was significantly lower for patients with lymph node metastases than for those without; 2.the sample size of patients supplemented with other treatments was not large enough; 3. each case came from a different study unit, so there was some variation in the quality of the procedure even for radical surgery. If the extent of surgery is inadequate, for example TAH, adjuvant other treatments may improve their median survival time. Immunotherapy, particularly immune checkpoint inhibitors, has shown great promise in cancer treatment (132), whereas immunotherapy based on immune checkpoint blockade is efficacious in treating melanoma (133). Although some studies have suggested that anti-PD-1 is associated with better OS compared to anti-CTLA4 in advanced/recurrent female genital tract melanoma (134), our results demonstrated that PD agents were not superior to the other adjuvant treatment modalities in patients with PMMC. This may be due to the small sample size of patients enrolled in this study.
In summary, this is the first pooled analysis including 149 cases of primary cervical melanoma. We found that patients who underwent RH-based surgery, those with non-lymph nodes metastatic and those who underwent lymphadenectomy had significantly higher survival rates. Based on the results of the analysis, the addition of other treatment options for patients who undergoing RH-based surgery is subject to further study, but for those who had TAH-based surgery, the addition of other treatments to prolong median survival may be considered. Notably, age and lymphadenectomy were associated with increased and reduced risk of death in these patients,

DATA AVAILABILITY STATEMENT
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

ETHICS STATEMENT
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

AUTHOR CONTRIBUTION
HC wrote the manuscript. AM and AF participated in the search strategy development. AM assisted in acquisition, analysis, or interpretation of data for the current work. HC prepared figures and tables. HW contributed to study concept and design. HC and MH double-checked the data and corrected the error in the tables. All authors contributed to critical manuscript revision.

FUNDING
This study was financially supported by Annual Science and Technology Steering Plan Project of Zhuzhou and Sichuan Science and Technology 259 Program (2021YFS0126).