This study aims to develop a machine learning-based prediction model for patients with vulvar cancer. This model will utilize patient characteristics and disease features to determine the disease's prognosis. The scoring system will also include management information to facilitate prediction of clinical outcomes of different management strategies and potential management that would yield the best prognosis.

Vulvar cancer (VC) is a relatively rare gynecological cancer accounting for 5-8% of all cases [1].

It comes the fourth among the commonest gynecological cancers and tends to affect women after menopause with a median age of 68 years [2,3].

Risk factors include cervical intraepithelial neoplasia, prior history of cervical cancer, smoking, lichen sclerosus, and immunodeficiency syndromes [4-5]. As squamous cell carcinoma is considered the most common type of VC, there are two potential pathogenic pathways for squamous cell carcinoma of the vulva include chronic inflammatory processes and human papillomavirus (HPV) infection [6-7].

While VC may be asymptomatic, most cases are present with bleeding, discharge, vulvar mass, ulcer and/or pruritis. Furthermore, it can be presented by a groin mass which reflects inguinal lymph node involvement. VC may be confined to the primary site in 59% of cases while 30% and 6% of cases spread to regional lymph nodes and distant areas, respectively [8].

FIGO staging is considered the standard classification system that determines prognosis and management of newly diagnosed VC. However, there are numerous gaps in the current staging system that would limit full interpretation of prognosis and management guidance [9]. Although staging system primarily determines disease prognosis, the staging system does not consider all prognostic factors, such as disease stage and histopathology. In fact, factors other than lymph node metastasis may have a stronger predictive influence such as the severity of the disease, age, histologic type and adjuvant radiotherapy and chemotherapy [10].

Development of a prognostic and decision-making system, based on comprehensive inclusion of individual patient and disease characteristics, would facilitate accurate prediction of disease prognosis and determination of individualized management strategy

A retrospective multicenter cohort study will be conducted among at least 6 European gynecologic oncology centers.

Inclusion Criteria:

1. Women diagnosed with Vulvar cancer and treated at collaborating centers between January 1st, 2008, and December 31st, 2017.
2. Women aged 18 years old or older, complete follow-up on for at least 3 years, unless censored by mortality.

Exclusion criteria:

1. Women will be excluded from the study if there were lost to follow-up before 3 years post-treatment.
2. If the patient did not not receive their treatment in the receptive centers, and if they were diagnosed with synchronous cancers.

Inclusion Criteria:

• Women diagnosed with Vulvar cancer and treated at collaborating centers between January 1st, 2008, and December 31st, 2017
• women aged 18 years old or older, complete follow-up on for at least 3 years, unless censored by mortality.

Exclusion Criteria:

• Women will be excluded from the study if there were lost to follow-up before 3 years post-treatment
• If the patient did not receive their treatment in the receptive centers
• If the patient were diagnosed with synchronous cancers

All women who will be diagnosed with primary vulvar cancer at any stage, of all histological types and grades eligible for the study