Introduction

Vulvar cancer is the fourth most common female genital cancer accounting for 1.4% of all new cases of cancer in women in Germany in 2016.1 Squamous cell carcinomas are the most common (>95%) followed by basal cell carcinoma, undifferentiated carcinoma and adenocarcinoma.2 Especially older women are majorly affected. Recent epidemiological studies have reported changes in the incidence rates of vulvar cancer in Germany, Europe and Asia, especially in younger women with an annual increase of 0.5–1.9%.1 3–9 The 5-year relative survival rate based on the extent of disease as reflected by the International Federation of Gynaecology and Obstetrics (FIGO) classification in the USA between 2012 and 2018 was 70.3%.10 The 5-year survival rate in the USA was 86.3% for FIGO I and 18.3% for FIGO IVB.11 Notably, most cases are diagnosed at an early stage; however, lymph nodes are already affected at the time of diagnosis in 20% of patients.1 Studies have shown that lymph node metastasis is a major prognostic factor.12 13 Various studies have reported varying results on whether age is a risk factor for recurrence.13 14 There are some studies on survival time depending on the FIGO stage whereas there are only a few on how age affects prognosis.

The present study aimed to determine the difference in the 5-year survival rate according to tumour stages in our large cohort of>17 000 women with vulvar cancer in Germany between 2004 and 2017. We also aimed to determine whether age at diagnosis influences the relative survival rate by comparing different age groups and their survival rates. Finally, we aimed to determine whether the distribution of tumour stages changes according to the age group. The impact of lymph node involvement and the local size of the tumour on the survival rate was also investigated. As vulvar cancer metastasis is primarily lymphogenous, positive inguinofemoral lymph nodes are expected when searching for distant metastasis.15 This leads to the hypothesis that staging for distant metastasis is not mandatory when regional lymph nodes are cancer-free.16 This hypothesis was tested in the present study.

MethodsData source

This retrospective study is based on the datasets of the German Centre for Cancer Registry Data (ZfKD) at the Robert Koch Institute and includes all reported cases of newly diagnosed malignant neoplasm of the vulva (International Classification of Diseases 10th Revision (ICD-10) C51) in Germany between 2004 and 2014 with follow-up until 2017.

ZfKD is a population-based cancer registry that uses the cause of death statistics to achieve the best coverage of cancer cases in Germany.17

Data on the patient’s dates of birth, diagnosis and death and the ICD-10 code, tumour–node–metastasis (TNM) classification as per the Union for International Cancer Control Classification of Malignant Tumours, grading and mortality status without further information on the cause of death were collected and analysed. The TNM classification was according to the fifth, sixth or seventh edition. During the study period, the classifications changed as well as the report in the registry.18 19 The end of follow-up was indicated by death notification in the registry.

Inclusion and exclusion criteria were defined. N=27.506 patients with a diagnosis of vulvar cancer (ICD-10 C51) were registered in the ZfKD between 2004 and 2014. Cases with incomplete staging, not assignable to FIGO classification or age<18 years were excluded. These criteria resulted in the exclusion of cases with missing T-stage (n=6855) unless M1 stage was present, cases of T0 with M1 (n=2) and cases with missing N stage (n=3620) unless T1a or M1 was present. In these cases, lymph node staging is not mandatory. Missing date of death (n=7), age<18 years (n=4) and use of TNM fourth edition (n=1) also led to exclusion. Finally, 17 017 patients were included in the descriptive analysis. Cases with missing T stage and M1 (n=154) and cases with death certificate only (n=50) were additionally excluded leaving 16 813 patients for the survival analyses.

There was no patient or public involvement in this study.

Staging and grouping

This study aimed to analyse the impact of different stages of vulvar cancer, age at diagnosis and lymph node status on the survival rates of patients with vulvar cancer. Therefore, the TNM stages in the data sets were converted into the 2009 FIGO classification according to the respective TNM classification.20 FIGO I includes tumours localised to the vulva whereas FIGO II includes tumours infiltrating neighbouring structures. FIGO III includes tumours with lymph node involvement and locally invasive grafted tumours. FIGO IVA includes tumours that invade the vaginal, bladder or rectal mucosa or are fixed to the pelvic bone and/or involve fixed or ulcerated inguinofemoral lymph nodes. FIGO IVB is any stage with metastasis involving pelvic lymph nodes.20 To investigate the influence of age on survival rates, the patients were divided into three groups according to the age at the time of diagnosis to represent young (premenopausal and perimenopausal women), middle-aged (postmenopausal women) and older women in their senior years. Holleczek et al conducted a cancer registry study regarding changes in the incidence of vulvar cancer in Saarland, a German federal state, defined the age groups they compared along the same lines: Group 1, ≤54 years; Group 2, 55–74 years; and Group 3, ≥75 years.4

Statistical methods

Categorical variables are presented as absolute and relative frequencies and continuous variables as means with SD and ranges and as additional medians with IQR. Survival time analyses were performed using Kaplan-Meier analyses and Cox regression. The Cox regression model included FIGO stage, age group and N and T stage as covariates with FIGO I, Group 1, N0 and T1 as references. Differences within the Kaplan-Meier analysis were analysed using log-rank tests. Relative survival was estimated using the function ‘period’ from the package ‘periodR’—for R (R Foundation for Statistical Computing, Vienna, Austria).21 The relative survival results were determined by comparing the overall survival of the study cohort with the survival of the German population of the corresponding sex and age group. The expected survival in the population was determined from mortality tables according to the ‘Ederer II’ method.22 For all comparisons, statistical significance was set at α=0.05. Statistical analysis was performed using SPSS Statistics for Windows (V.25.0; IBM, Armonk, New York, USA) and R (V.3.6.1).23 24 Python programming language (V.3.11) and Matplotlib (V.3.8.2) were used for data visualisation.25 26

ResultsDescription

Between 2004 and 2014, 27 506 patients were diagnosed with vulvar neoplasia in Germany. The number of vulvar cancer cases reported in the cancer registries has increased over the years, almost doubling between 2004 (n=1397) and 2009 (n=2680). After applying the inclusion and exclusion criteria, 17 017 cases were finally included in the study. The mean survival time was 109.8 months (95% CI: 108.7 to 110.9 months) and the median follow-up time was 57 months.

Notably, most women were diagnosed with localised stage I (n=9412; 55.3%) and II (n=2154; 12.7%) lesions followed by regional stage III (n=3734; 21.9%) and IVA (n=1096; 6.4%) lesions. Distant metastasis (stage IVB) was a rare occurrence (n=621; 3.6%) (table 1). Inguinofemoral lymph nodes were observed in 4832 cases. The median age at diagnosis was 70 years (range: 19–102 years) (table 1). Furthermore, most women were diagnosed in their eighth decade (70–79 years, n=5127; 30.1%). The IQR was between 57 and 78 years (21 years). Only 0.9% (n=243) of patients were aged≤30 years at the time of diagnosis. The median age at diagnosis increased with increasing stages (FIGO I: 68 years, FIGO IVB: 73 years) whereas the median age at death decreased with increasing stages (FIGO I: 80 years, FIGO IVB: 74 years) (table 1).

Table 1

Descriptive data of women diagnosed with vulvar cancer in Germany between 2004 and 2014

The distribution of patients according to their age group was as follows: Group 1 had 3555 patients (20.9%); Group 2, 7367 (43.3%); and Group 3, 6095 (35.8%) (table 1). Patients in Group 1 were significantly more often diagnosed with a low T-stage disease (79.5% (Group 1) vs 69.2% (Group 2) vs 62.9% (Group 3); p<0.001) and negative lymph nodes (65.1% (Group 1) vs 63.3% (Group 2) vs 59.4% (Group 3); p<0.001) than those in Groups 2 and 3. Therefore, more patients in Group 3 had significantly higher FIGO stages than those in Groups 1 and 2 (FIGO IVB: 4.7% vs 3.5% (Group 2) and 2.3% (Group 1); p<0.001) (figure 1).

Figure 1

Distribution of FIGO stages for women diagnosed with vulvar cancer in Germany between 2004 and 2014 according to age groups. The differences between and within the groups were significant (p<0.001). (Group 1: ≤54 years; Group 2: 55–74 years; Group 3: ≥75 years). FIGO, International Federation of Gynaecology and Obstetrics.

The risk of distant metastasis was 15.92 (95% CI: 12.17 to 20.94) times higher when lymph nodes were involved than when they were not. However, negative lymph nodes were observed in 13.7% (n=62/454) of patients with metastasis and documented N stage.

Overall and relative survival according to FIGO stage

The 5-year overall survival rate of all patients was 69.5% (95% CI: 68.7% to 70.3%) (figure 2). The overall 5-year survival rate for stage I was 84.3% (95% CI: 83.5% to 85.1%). It was significantly reduced for stages II and III with 65.1% (95% CI: 63.1% to 67.1%; p<0.001) and 51.8% (95% CI: 50.2% to 53.4%; p<0.001), respectively, and further reduced for stages IVA and IVB with 33.2% (95% CI: 30.3% to 36.1%; p<0.001) and 22.2% (95% CI: 18.3% to 26.1%; p<0.001), respectively (figure 2). Stage II patients had an increased risk of death within the first 5 years of disease which was more than twice that of stage I patients (HR: 2.4; 95% CI: 2.3 to 2.6) (figure 2).

Figure 2

Kaplan-Meier survival function and relative survival function with 95% CI for women diagnosed with vulvar cancer in Germany between 2004 and 2014 according to FIGO stage. FIGO, International Federation of Gynaecology and Obstetrics.

The 5-year relative survival rate decreased with increasing stage (figure 2). The differences between overall and relative survival rates were significant for FIGO I (84.3% vs 91.9%) and FIGO II (65.1% vs 75.6%) (figure 2). The 5-year relative survival rate for stage IVB was worse than the 5-year overall survival rate (21.3% vs 22.2%); however, the difference was not significant (figure 2).

Overall and relative survival according to age group

The 5-year overall survival rate differed significantly among the groups (Group 1 (89.4%) vs Group 2 (74.7%) vs Group 3 (51.3%); p≤0.001) (figure 3). Compared with Group 1, Groups 2 and 3 exhibited an increased mortality risk (HR: 2.62 and 6.05, respectively) (figure 3).

Figure 3

Kaplan-Meier survival function and relative survival function with 95% CI for women diagnosed with vulvar cancer in Germany between 2004 and 2014 according to age group. (Group 1: ≤54 years; Group 2: 55–74 years; Group 3: ≥75 years).

The 5-year relative survival rate showed the same trend as the 5-year overall survival rate; however, it is significantly higher than the 5-year overall survival rate in Groups 2 and 3 (Group 2: 78.1% vs 74.7%; Group 3: 68.0% vs 51.3%) (figure 3). The difference was not significant in Group 1 (89.2% vs 89.4%) (figure 3). The 5-year relative survival rates differed significantly among the three groups (p<0.001) (figure 3).

Mortality according to FIGO stage and age group

The analysis showed that mortality increased with increasing stage and age (figures 2 and 3). However, the increase in mortality risk across the stages was more pronounced in Group 1 (HR 25.3 (95% CI: 17.8 to 36.0) for FIGO IVB) than in Group 3 (HR 5.8 (95% CI: 5.0 to 6.9) for FIGO IVB) (table 2). Across all the FIGO stages and age groups, the mortality rate was highest within the first 3 years (figures 2 and 3).

Table 2

Cox regression of women with vulvar cancer in Germany between 2004 and 2014 for age subgroups by FIGO stage

The effects of N-stage and T-stage on survival

The N-stages and T-stages affected survival significantly (p<0.001). This significance remained even after adjusting for other comparisons. HR for N-stage was 1.79 (95% CI: 1.73 to 1.84) whereas that for T-stage was 1.88 (95% CI: 1.80 to 1.96). The higher the T-stages or N-stages, the higher the mortality risk.

Discussion

This study confirmed that the 5-year survival rate of patients with vulvar cancer depends strongly on the stage and age at diagnosis with the survival rate decreasing with increasing stage and age at diagnosis. Older women were significantly more diagnosed with higher stages than younger women. Furthermore, the survival chances, indicated by the HR, decreased more remarkably with advanced stages in younger than in older women. Lymph node status and tumour size, indicated by the T-stage, were further confirmed to be independent factors influencing variables for mortality (p<0.001). Adjusting for each factor did not change the overall results suggesting the robustness of the findings. Positive lymph node status was also a risk factor for metastasis (OR: 15). However, in approximately 14% of patients with metastasis, no positive lymph nodes were observed. Overall, vulvar carcinoma is a disease with good survival chances, partly because it is detected in early stages leading to a 5-year overall survival rate of two-thirds of patients. However, even if the disease is detected in an advanced stage with metastasis, one in five women survive for≥5 years. Mortality risk across all stages and ages was highest within the first 3 years after diagnosis.

The present study revealed that patients in Germany are more often diagnosed at an early stage (localised stage I or II) than at an advanced stage. According to studies from the USA, the Netherlands and Denmark, this distribution is similar in these countries7 8 10 27; however, patients in the USA and the Netherlands are less frequently diagnosed with stages I and II possibly due to differences in screening programmes and healthcare availability.10 27 There is compulsory health insurance for the entire population in Germany and an annual gynaecological screening programme for women whereas a lot of people are without health insurance in the USA.28 29 Early detection of carcinomas depends strongly on the insurance status. For example, a study on hepatocellular carcinoma reported that the frequency of early-stage was significantly higher among patients with insurance than among those without.30 The analysis showed that the survival rates across all stages in Germany are better than those in the USA and Japan.5 11 One possible explanation for this could be the treatment differences. A guideline for the treatment of patients with vulvar cancer was implemented during the study period in Germany; however, it was not established in Japan until 2015 and in the USA until 2016.16 31 32 The introduction of guidelines has led to better outcomes.33 In addition, Germany has better access to healthcare services than the USA as measured by the34 GBD 2019 Healthcare Access and Quality Collaborators by comparing the Healthcare Access and Quality Index of different Countries (87.0 (Germany) vs 80.6 (the USA)).34

In the present study, three age groups were defined and compared according to the 5-year survival rate: Group 1 (≤54 years), Group 2 (55–74 years) and Group 3 (≥75 years). The overall survival rate was expected to decrease with increasing age. However, it was hypothesised that the relative survival rate would not differ too much among the age groups because the age factor is virtually eliminated from relative survival rates. This hypothesis was wrong. The 5-year relative survival rate was lower than the 5-year overall survival rate in Group 1; however, the difference was not significant. In contrast, it was significantly higher than the 5-year overall survival rate in Groups 2 and 3. However, it was significantly worse among older patients than among younger patients (p<0.001) (figure 3). Meltzer-Gunnes et al from Norway compared three age groups according to the 5-year relative survival rate. They defined the groups differently: ≤60, 61–80 and≥81 years; however, they obtained the same result as age influenced the survival rate significantly (p<0.001).9 The data suggest that the survival rate is worse in Norway than in Germany (82.7%, 65.2 % and 44.1% vs 89.2%, 78.1% and 68.0%, respectively). However, due to the differently defined groups, this result remains uncertain. The reason for the differences in survival rates between young and older patients could be the significantly higher FIGO stages in older patients (figure 1). Higher FIGO stages are associated with lower survival rates. In addition, older women are more often untreated, according to the guidelines, due to comorbidity and their age and suffer more often from postoperative wound complications.35–37 There are potential explanations for the higher prevalence of human papillomavirus (HPV)-independent cancers with a worse prognosis than HPV-dependent cancers among older women.38 39 However, why older women are more frequently diagnosed with significantly higher stages than younger women remains unclear. Answering this question was not the present study’s aim; however, it has been mentioned that older women are more likely to develop HPV-independent vulvar cancer than younger women.38 Van de Nieuwenhof et al showed that HPV-independent differentiated vulvar intraepithelial neoplasia (VIN) progresses significantly faster into vulvar cancer than HPV-dependent usual VIN.40 The faster progression is a valid explanation for the higher stages observed in older patients. Furthermore, older women can be assumed to be less aware of changes in their vulva and cancer screening possibilities and have worse access to healthcare.34 41 The greater increase in mortality risk with increasing stage in younger women compared with older women remains unclear (table 2). An explanation may be the better physiological resilience of younger patients which induces better tolerance of the disease and therapy, especially in patients with low stages receiving simple therapy through surgery.42 Notably, more advanced carcinomas require more complex therapy and thereby burden the body in two ways: Through the cancer itself and the therapy. The effect of physiological resilience is minimised. Older patients suffer from frailty and have a higher mortality rate resulting in a flatter increase in mortality risk.43

Tanaka et al observed that the greatest decrease in survival rate in Japan was within the first 5 years after diagnosis.5 This is consistent with the findings of the present study in Germany with the greatest decrease observed within the first 3 years. The difference in the years could be explained by the size of the study cohort as Tanaka et al included 815 patients which was significantly fewer than those included in the present study. Furthermore, differences in lifestyle and predispositions might have led to the observed differences. The outdated and unrevised German guidelines on vulvar cancer recommend a 3-month follow-up interval within the first 3 years after diagnosis, a 6-month interval within the 4th and 5th year and an annual follow-up after that.42 A reasonable recommendation can be made based on the finding that mortality risk was highest within the first 3 years. Furthermore, the guideline recommended staging for metastasis only in patients with higher stages.42 This recommendation was based on the finding that metastasis occurs in 3.8% of patients with less than three affected lymph nodes and in 66% of patients with three or more affected lymph nodes.16 42 In the present study, 13.7% of all stage IVB patients had negative inguinofemoral lymph nodes. Therefore, staging for distant metastases should also be strongly considered in cases of negative inguinofemoral lymph nodes as recommended in the 2023 European Society of Gynaecological Oncology Guidelines for the Management of Patients with Vulvar Cancer.39

Notably, several studies have shown lymph node status to be a prognostic factor for recurrence and survival in women with vulvar cancer.13 44–48 Hami et al reported lymph node positivity as the most important risk factor for recurrence and Kumar et al reported lymph node status and tumour stage as the most important prognostic factors.13 48 These findings are consistent with those of the present study indicating that lymph node positivity is a significant risk factor for mortality (p=0.001) with tumour size (T-stage) also being an independent factor (p<0.001) as shown by the present data and the findings of Sznurkowski et al.46 Both factors are considered together in the FIGO classification making this staging the most reasonable for the most accurate prognosis.

Vulvar cancer is not a common disease. We obtained information on the survival of this disease by evaluating the dataset of the epidemiological cancer registry and studying approximately 17 000 patients with newly diagnosed invasive vulvar cancer over approximately 13 years. To our knowledge, the present study is the first presentation of structured data on the survival of a large cohort of women with vulvar cancer in Germany. Representative analyses can be derived from this. However, the present study had some limitations. First, the large study cohort could have led to statistical significance without clinical significance; however, this problem was addressed by analysing subgroups rather than the whole cohort. Second, as is typical for a registry study, only available information could be analysed; therefore, subsequent data collection was impossible. This resulted in the loss of approximately 10 000 diagnosed cases in this study owing to incomplete documentation. Nevertheless, the remaining cases were sufficient to obtain solid results. Third, HPV status as a risk factor, resection margin or recurrence rate are important and interesting factors that could not be investigated using the available data. Finally, the impact of applied therapy on the probability of survival could not be meaningfully assessed, although this would certainly be of great interest. These aspects remain to be investigated in future studies. As mentioned previously, every woman aged ≥20 years in Germany undergoes a gynecological check-up. They visit the gynaecologist once a year to have their vulva, vagina and cervix inspected and to undergo a Pap smear test on the cervix.28 This screening is assumed to play an important role in the increased rate of early detection of vulvar carcinomas in Germany and, therefore, could be a good recommendation for other countries.

In conclusion, this study revealed the significant impact of stage and age as prognostic factors on the survival of patients with vulvar cancer. Older women had a worse survival chance than younger women. In addition, older women were more often diagnosed at advanced stages. Tumour size and lymph node status were proven to be independent risk factors for mortality. Early detection, with concomitant small spread of the tumour and tumour-free lymph nodes, is the main controllable factor for the survival prognosis of the disease. Furthermore, we found that the greatest decrease in survival rate was within the first 3 years and staging for metastasis should always be considered, even if the lymph nodes are cancer-free.