Vulvar cancers account for 0.6% of female cancers in the US.1 In New Zealand, the Ministry of Health registered 70 vulvar cancer cases in 2014 with rates reported to be between 0.2 to 0.3% between 2003 to 2013.2 Surveillance Epidemiology and End Results (SEER) databases between 1973–2004 documented a rise in the incidence of invasive vulvar tumours by 1% per year.3 Risk factors for developing vulvar cancers include human papilloma virus (HPV), accounting for 40% of vulvar cancer cases.4 Squamous cell histology has been reported in up to 95% of vulvar cancers.5 Most patients present with pruritus followed by vulvar bleeding, discharge, dysuria and pain.5 Presence of inguinal or femoral nodes is the most important prognostic factor for survival.6

Patients with early stage vulvar cancer (Stage 1 and 2) should undergo excision of the primary tumour. Surgical margins greater than 1cm have been suggested to reduce the risk of local recurrence.7 There is growing evidence for sentinel node evaluation in these patients.8 Radiation therapy (RT) is offered to patients with close or positive margins (Stage 2) or with inguinal lymph node metastases to reduce risk of locoregional recurrence and improve survival.9,10 Patients with locally advanced (Stage 3 and 4a) disease who are surgical candidates should be recommended inguinofemoral lymphadenectomy.11 Chemoradiation or radiation alone is recommended in patients who are not surgical candidates.5

In New Zealand, studies providing an update of factors affecting therapeutic management of patients with vulvar cancers are limited. The objectives of our study were to review patient characteristics, treatment choices and outcomes of patients with vulvar cancers in the Waikato Region, New Zealand.

Methods

Retrospective review of all newly diagnosed vulvar cancer cases registered in a regional cancer centre with a catchment population of 720,000 between 1 January 2000 and 31 August 2015. Cases were identified from the databases of the departments of clinical coding, radiation oncology and medical oncology of Waikato Hospital. Using the patient’s National Health Index (NHI) number, details of hospital admissions, electronic patient files and individual patient’s clinic notes were reviewed and evaluated for completeness of pathological reporting and treatment-related side effects. Data reviewed include ethnicity, age at diagnosis, comorbidities using Age-Adjusted Charlson Comorbidity Index Score (ACCIS), smoking status, body mass index, clinical presentation, primary therapy (surgery, radiation or combined chemoradiation), date of histological diagnosis, staging, complications of primary therapy, recurrences and mortality. For staging, the Federation International Gynaecologic Oncology (FIGO) and TNM classification system were used. ACCIS for each individual patient was calculated, which allowed comparison between cases with regards to treatment, toxicities and outcomes. Patients were reviewed weekly during chemoradiation and 2–3 monthly thereafter. Median follow-up was 33 months. Overall survival was calculated from the date of diagnosis to the date of death from any cause or date last known alive, using the Kaplan Meier method. Health and Disability Ethics Committees (HDEC) review was not required for this study as per Standard Operating Procedures HDEC Version 2.0 August 2014.

In the 15-year period, 60 cases with vulvar cancer were identified from the three different databases. Thirteen cases were excluded; 11 cases had another histological diagnosis, which included vulvar basal cell carcinoma (3), extramammary Paget’s disease of the vulva (2), vulvar adenocarcinoma (2), vulvar melanoma (1), leiomyosarcoma (1), metastatic adenoid cystic carcinoma (1) and high-grade neuroendocrine cancer (1). Two cases with vulvar squamous cancer were excluded as they were diagnosed prior to year 2000. This report covers 47 cases of vulvar squamous cell carcinoma. Between 2000–2004 eleven cases were identified, fourteen cases in 2005–2009 period and eighteen cases in 2010–2014 period.

Results

Forty-seven patients had vulvar squamous cell cancer graded as well differentiated (28%), moderately differentiated (60%) and poorly differentiated (11%). Depth of invasion 1mm or less was seen in four cases versus 33 cases reported having depths of invasion greater than 1mm. Tumour margins (<8mm) were initially identified in 13 cases with seven of the cases requiring re-excision, and among these only three cases had clear margins following re-excision. Depth of invasion, lymphovascular invasion and vulvar intraepithelial neoplasia appeared to be reported in 79%, 68% and 55% respectively. However, p16, p53 and HPV DNA and Ki67 were poorly reported; ie, 21%, 6% and 6% of cases respectively. Patient characteristics are shown in Table 1.

Table 1: Vulvar cancer patient demographics and treatment according to stage.

‍

ACCIS: Age Adjusted Charlson Comorbidity Index Score                       ND: Node Dissection

RT: Radiotherapy                                                                              Rx: Treatment

TE: Tumore Excised                                                                           Rec: Recurrence

Dx: Diagnosis

The median age of our patients was 69 years (range 39–94). Eighty-one percent (38 patients) were of European ethnicity and 19% (nine patients) Māori. Thirty-six percent (n=17) of patients were either current smokers or ex-smokers; 45% (n=21) were non-smokers. Smoking status was unknown in 19% (n=9). Median age at presentation for Māori was 57 years versus 76 years for European. Vulvar pruritus, a lump, pain and bleeding were the most common initial symptoms. Thirty-one percent (n=15) of cases had a background of lichen sclerosus. Hypertension, obesity, ischaemic heart disease, diabetes and second malignancies were the top five comorbidities observed among patients (Figure 1). Each patient’s ACCIS was evaluated (median score 4).

Figure 1: Top five comorbidities in squamous cell vulvar cancer by percentage (%).

Over the 15 years, treatment had changed. More formalised imaging, emerging sentinel node evaluation and discussion in multidisciplinary meetings were observed in the latter years. Overall, only 36% of our patients had an initial staging CT scan; 6% had a CT scan at suspected progression and 2% had PET CT scans. Nine percent had an initial MRI scan for evaluation. We observed 54% mortality rate among patients with lymph node involvement versus 39% without lymph node involvement at diagnosis with a median follow-up of 33 months. Chemotherapy was only observed to be used from 2007 onwards and not between 2000–2006. Sixty-six percent of cases were discussed in a multidisciplinary meeting (MDM) at initial diagnosis. This is explained by the change in recommendations over the course of time with cases diagnosed prior to 2005 less likely to be discussed in an MDM.

Treatment and complications (refer to Table 1)

Surgery

Patients with early stage disease (Stage 1 and 2) were more likely to undergo local excision of their primary tumour compared to patients with advanced disease (Stage 3 and 4) (94% versus 88%). Lymph node dissections were less likely to be performed in early stage disease versus advanced disease (48% versus 61%). The low rates of node dissection in Stage 1 are due to advanced age and comorbidities; (five patients), negative sentinel node and patient choice (one patient each respectively).

Surgical complications were more common in advanced stage versus early stage (42% versus 33%). The most common acute wound complication was cellulitis (21%). The most common chronic complication was lymphedema (17%). No treatment-related deaths were observed. Only three patients underwent sentinel node (two bilateral and one unilateral) based management. Among these, one case was positive and lead to inguinal node dissection with positive lymph node involvement. The remaining two negative sentinel node cases have since had no groin recurrences. Eleven percent were deemed inoperable due to extensive tumour infiltration (two patients), and extensive comorbidities (three patients).

Chemoradiotherapy

Eleven patients (23%) received concurrent chemoradiation either in the form of 5 fluorouracil (5FU)-mitomycin (63%, seven patients) or weekly cisplatin (36%, four patients). The median age of patients in 5FU-mitomycin versus cisplatin arms was 69 years and 50 years, respectively. From our small number of patients who received chemotherapy (n=11), we observed that weekly cisplatin was a much better tolerated regime compared to 5FU-mitomycin (Table 2). Comparing between the two chemotherapy arms, ACCIS median for 5FU-mitomycin versus cisplatin were 4 and 2.5, respectively.

Table 2: Complications observed in the 5FU-mitomycin and cisplatin chemoradiation group by stage and age.

‍

Thirty-eight patients (80%) received radiation treatment with either curative (mean 50 Gray, range 45–61 Gray) or palliative intent (which included primaries: vulva and groin; and secondaries: lumbar spine and lung). Radiotherapy infield recurrence was seen in four patients (Table 1). Among the remaining patients who did not receive radiation treatment, four (8.5%) were unfit for treatment due to comorbidities and five (10.6%) had their treatment completed with surgery.

A total of eleven patients with Stage 1 disease received radiotherapy. Seven were administered radiotherapy due to high-risk features such as positive margins (three cases), inoperable (two cases), upstaged (two cases). In the remaining four cases, radiotherapy was administered at cancer recurrence. In three cases that had positive tumour margins, ie, (two had persistent positive tumour margins despite surgical re-excision, and one patient was surgically unfit for re-excision). Two patients’ tumours were located close to the perineum or anal verge and thus deemed inoperable. Two cases were upstaged during follow-up postoperatively with inguinal nodes and thus managed with radiotherapy.

Five-year overall survival is shown in Figure 2. Whereas five-year disease free specific survival for Stages 1, 2, 3 and 4 were 94%, 60%, 59% and 29% respectively. The discrepancy between overall survival and disease-specific survival is reflected by this comorbid cohort of patients. For example, two patients with Stage 1 had died of a second malignancy, and another patient deemed unfit for surgical excision of the primary tumour due to advanced age and comorbidities. A review of Stage 2 patients revealed four of the five patients with advanced age (median age of 78 years). One Stage 2 patient was inoperable and eventually died of ischaemic heart disease. Another Stage 2 patient developed delirium postoperatively and was hence unfit for further adjuvant treatment.

Figure 2: Kaplan Meier survival curve by stage of vulvar cancer.

Independent of stage, overall survival outcomes were better for (ACCIS) <5 (60%) versus 41% for scores ≥5 (Figure 3). At the time of data analysis, 22 of the 47 patients had already deceased; with a median time to death from date of diagnosis among deceased patients reported at 23 months. The median follow-up of 33 months (range 3–161 months) was uncensored for death with a mean of 44 months.

Figure 3: Bar graph showing Age-Adjusted Charlson Comorbid Index Score (ACCIS) versus number of patients alive and deceased.  

Discussion

We report one of the largest single-centre reviews of vulvar cancers in Australasia. Multiple factors including patient choice, tumour location, advanced age, comorbidities and treatment complications have influenced and individualised treatment. Variation in treatment over the course of time, such as evolving sentinel node evaluation, advancements in imaging modalities and greater proportion of cases discussed in MDM in the latter years were observed. Although surgical and chemoradiotherapy were associated with high morbidity, we observed that cisplatin radiotherapy was much better tolerated compared to 5FU-mitomycin, however this requires validation in larger prospective studies. Independent of stage, patients with fewer comorbidities had a better overall survival.

Our study showed that Māori tend to present with vulvar tumours younger, with less comorbidities compared to Europeans. Evidence shows that the median life expectancy at birth in New Zealand for non-Māori and Māori is 82 years and 75 years respectively.12 Wu et al13 studied vaginal cancers among different races in the US and showed that African, Asian, Pacific Island and older women were more likely to be diagnosed with advanced disease, and these groups had lower five-year relative survival rates than their Caucasian, non-Hispanic, and younger counterparts. Babarinsa et al14 add that most patients present late and default during treatment. Our study was too small to evaluate overall survival differences between ethnicities.

The rate of HPV pathological reporting in our centre was poor. Improving HPV reporting will be useful in the future as the presence of “HPV-16 antibodies confer a 5.3-fold risk of vulvar neoplasia. High antibody levels are associated with 20-fold risk, more commonly seen in smokers” probably related to decreased clearance of HPV infection among smokers.4,15 We identified 31% cases had lichen sclerosus prior to diagnosis of vulvar cancer, which suggests the well-known premalignant feature seen more commonly in the elderly. Lichen sclerosus and VIN rates in both ethnicities were similar in our study.

Another predisposing factor of vulvar cancer includes low socioeconomic status and smoking. Thirty-six percent of our patients had prior smoking history. Socioeconomic status was not evaluated in our study. We identified metabolic syndrome to be highly prevalent. At least 20% of our patients had hypertension, diabetes, ischaemic heart disease and obesity. The Metabolic Syndrome and Cancer project identified metabolic syndrome to increase vulvar cancer risk (hazard ratio 1.78).16

Factors tailoring towards different surgical approaches in the 15-year period include comorbidities, age, depth of invasion and lateralisation of the tumour. Twelve local recurrences (eight vulvar and four groins) were identified. Although the numbers of patients receiving chemoradiation were small, no local recurrence was observed in our patients who received chemoradiation. A vulvar squamous cancers study in Italy in 502 patients showed that 37% developed recurrences. “Five-year survival rate was 60% for perineal recurrences, 27% for inguinal and pelvic recurrences, 15% for distant recurrences and 14% for multiple recurrences”. Factors predicting for risk of recurrences were FIGO stage >2, positive lymph nodes and vascular space invasion, which were all statistically significant.17,18 The low recurrence rate in our chemoradiotherapy group may be a reflection of selection bias and short follow-up period.

Despite lower rates of infection, wound breakdown and lymphedema observed in our study, surgery is associated with significant morbidity. To address this issue, tertiary centres have been offering sentinel node evaluation. Only 6% of our patients underwent sentinel node evaluation. Chronic lymphedema lasting beyond six months has been described to be worse among those with greater extent of lymphadenectomy, sartorius transposition and adjuvant groin irradiation.19,20 Sentinel lymph node biopsy, although not yet the standard of care, is safe and recommended in patients with early stage disease to reduce complications.6,21

Evidence surrounding the choice of chemotherapy used is based on Phase 2 studies. Moore et al evaluated radiation treatment and weekly cisplatin (40mg/m2)in patients not amenable to surgery. Although 64% complete response was observed, 31% did not complete treatment with a 3.4% death rate, 16% Grade 3 or higher toxicities and 7% refused treatment.22 Another study of 28 patients with locally advanced vulvar cancer that evaluated chemoradiotherapy with 5FU-mitomycin observed an 85% overall response, with vulvar desquamation seen in 93% of patients.23 We observed that weekly cisplatin was much better tolerated compared to 5FU-mitomycin. One may argue that our patients receiving weekly cisplatin were much younger (median 50 years) with a lower ACCIS compared to the 5FU-mitomycin cohort (median 69 years). Older and more comorbid patients may have been treated with 5FU-mitomycin perhaps due to age selection bias. Data surrounding tolerability and selection criteria of chemotherapy in patients above 65 years of age is scarce. One study assessing chemoradiation using cisplatin and paclitaxel in elderly patients (aged >70 years) with oesophageal cancer showed that 67% completed treatment as scheduled, with the expense of significant haematological toxicity.24

We observed that higher comorbidity score among cancer patients impacted on their overall survival. This finding is supported by a retrospective analysis in vulvar cancer, which showed that Charlson score 2 or greater was associated with a decreased overall survival (hazard ratio 3.03).25 Sogaard et al further highlight that the comorbidities among cancer patients influence the receipt of standard of care, with higher treatment associated complication rate and overall survival.26 Although ACCIS has been demonstrated as a reliable risk measure for perioperative cancer surgery,27 literature on its routine use to optimise care in either vulvar cancer management or multidisciplinary cancer meetings are limited.

The limitations of our study include its small sample size, single centre and retrospective nature. HPV reporting was poor and treatments altered during the time. In advanced disease, chemoradiotherapy demonstrated excellent local control, with cisplatin better tolerated compared to 5FU-mitomycin, requires validation in larger prospective studies and longer follow-up. The impact and importance of expert MDM review is emphasised and recommended. We propose that ACCIS could be incorporated as a tool to reduce treatment morbidity and mortality and may potentially further assist with accurate decision making in vulvar cancer MDMs. This requires further review in prospective larger studies.

Conclusions

We present the “real world” outcomes of a rare malignancy. Treatment is associated with high morbidity in this generally older and comorbid population. Larger prospective multi-centre studies are proposed.

‍