Clinical characteristics, treatment patterns, and prognosis in vulvar cancer and vaginal cancer: a national multicenter retrospective study in China

Abstract

Background

Vulvar and vaginal cancers significantly impact women’s health and quality of life. This study analyzes the prevalence of human papillomavirus (HPV), clinicopathological characteristics, treatment patterns, and survival outcomes, to understand their relationships and better guide prevention and therapy for Chinese patients with vulvar or vaginal cancer.

Methods

We conducted a retrospective, multicenter study across nine tertiary hospitals located in seven regions of China, including patients diagnosed and treated between 2017 and 2021. Data were collected from medical records and standardized questionnaires, with routine clinical follow-up. Continuous variables were analyzed using the Wilcoxon rank-sum test, while categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. Survival outcomes, including overall survival and disease-free survival, were evaluated using the Kaplan-Meier method and multivariable cox proportional hazards regression.

Results

A total of 218 patients (131 vulvar, 87 vaginal) were analyzed. HPV infection was significantly higher in vaginal than vulvar cancer (76.1% vs. 33.8%, p < 0.001). A non-significant trend toward younger age at diagnosis has been observed in HPV-positive vulvar cancer patients (p = 0.072). Both vulvar and vaginal cancers were predominantly early-stage (FIGO I: 57.3% vs. 51.7%), but vulvar tumors were more often well-differentiated (46.7%), while vaginal cancers were mainly moderately (39.2%) or poorly differentiated (23.5%). Receiving surgery during primary treatment phase was more common in vulvar cancer (84.6% vs. 50.6%, p < 0.001), often without adjuvant therapy (p = 0.016). Primary surgery was associated with longer overall survival and disease-free survival in vulvar cancer (p < 0.001), but not in vaginal cancer. A non-significant trend toward longer disease-free survival was observed in HPV-positive vulvar cancer patients (p = 0.107).

Conclusions

Vulvar and vaginal cancers exhibit distinct profiles in histological differentiation, HPV prevalence and age association. Surgery is central to vulvar cancer management and correlates with survival benefit, whereas vaginal cancer treatment is more heterogeneous. HPV status may have modest prognostic relevance in vulvar cancer, warranting further study.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-27384-3.

Background

Human papillomavirus (HPV) infection is a common sexually transmitted agent, with persistent high-risk types driving nearly all cervical cancers and enabling cervical cancer elimination efforts [1–3]. Beyond the cervix, HPV also contributes to vulvar and vaginal cancers, whose incidence is rising globally and in China [1, 4, 5]. Although guidelines recommend integrated prevention, the implementation and effectiveness of such strategies remain insufficiently evaluated [6].

Informing evidence-based policy requires understanding HPV prevalence, clinical profiles, treatment, and prognosis for these cancers. Systematic reviews indicate significant geographic variation in HPV attribution, with pooled prevalence in vulvar cancer ranging from 29.5% to 73.3% worldwide, and 43.5% in Asia [7]. Similarly, in vaginal squamous cell carcinoma, HPV prevalence varies from 65.1% in Europe and 72.1% in North America to 45.0% in Asia [8]. These disparities underscore the need for region-specific epidemiological studies. In China, however, robust nationwide data are scarce, with existing studies being single-center, small-scale, and inconsistent [9, 10].

Treatment of vulvar and vaginal cancers also lacks standardization, which may influence patient prognosis [11–13]. In China, dedicated clinical guidelines for these malignancies have been established only in recent years [14, 15]. Before their introduction, vaginal cancer management often relied on extrapolation from cervical or vulvar cancer protocols, particularly in advanced stages [16]. In addition, while HPV status is recognized as a prognostic biomarker in other HPV-related cancers such as head and neck cancer [17, 18], its role in predicting outcomes in vulvar and vaginal cancers remains inadequately characterized [19–21]. Thus, real-world investigations into treatment practices and their interplay with HPV status are urgently needed to guide therapeutic optimization.

This multicenter retrospective study therefore aims to determine the HPV attribution rate, delineate clinicopathological features, and document real-world treatment patterns for vulvar and vaginal cancers across China. It will also evaluate the impact of HPV status and other key factors on survival to inform tailored prevention and treatment strategies.

Methods

Study design and population

We conducted a multicenter retrospective study across nine tertiary hospitals from seven geographical regions in China (North, Northeast, East, South, Southwest, Northwest, Central). Patients diagnosed and treated for primary vulvar (ICD-10: C51) or primary vaginal (ICD-10: C52) cancer between January 1, 2017, and June 30, 2021, were included. Those with severe comorbidities (e.g., coronary heart disease, organ transplantation) or inability to provide informed consent were excluded.

Data was collected via hospital information systems and structured questionnaires. Informed consent was obtained during routine follow-ups; for deceased or unreachable patients, consent was waived by the ethics committee. The study protocol was approved by the Ethics Committee of the National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (Approval No.: NCC 20/041-2237).

Data collection

A structured questionnaire, which underwent expert review and was validated for reliability and validity in a pilot study, was developed for this study. Questionnaire with same structures have been used in previously published study [22]. It comprised four sections: (1) demographic characteristics (region, occupation, education, marital status, etc.); (2) risk factors (HPV infection status, history of cervical intraepithelial neoplasia [CIN], genital warts, other sexually transmitted infections, family or personal cancer history, tobacco and alcohol use, etc.); (3) clinicopathological features (age at diagnosis, tumor site, initial symptoms, Federation of Gynecology and Obstetrics [FIGO] stage, histologic type and differentiation); and (4) treatment and outcome data (primary treatment details, types of chemotherapy/radiotherapy, surgical procedures, recurrence status and date [disease-free survival, DFS], and vital status with date of death [overall survival, OS]).

All staging was conducted in accordance with the FIGO 2009 system. For primary treatment modalities, patients were first stratified according to whether surgery was included as part of the primary treatment. Among patients who received surgery during the initial treatment phase, further stratification was conducted based on the use of additional treatment modalities. HPV testing in this study was conducted as part of routine clinical practice at each participating center using vulva or vaginal swab specimens. All assays included in the analysis were approved by the National Medical Products Administration and all tests were conducted in accredited tertiary hospital laboratories. High-risk HPV status was categorized based on following rules: patients were defined as HPV-positive if at least one high-risk HPV type was detected, including HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68; those tested and negative were HPV-negative; and those without documented testing were considered HPV-untested.

Follow-up and prognostic outcomes

All follow-up was conducted through the hospitals’ routine post-treatment surveillance without additional intervention from the study team. The last follow-up date was December 22, 2022. OS was defined as the time from the start of primary treatment to all-cause death or last follow-up, whichever occurred first. DFS was defined as the time from the start of primary treatment to first recurrence, all-cause death, or last follow-up, whichever occurred first.

Statistical analysis

All statistical analyses were conducted using R version 4.2.3. Data visualization was performed with the “ggplot2” and “ggsurvplot” packages. Continuous variables are presented as median and interquartile range (IQR), and categorical variables as frequency and proportion. Group comparisons were made using the Wilcoxon rank-sum test (continuous variables) or the Chi-square/Fisher’s exact test (categorical variables). Survival analyses were assessed with the Kaplan–Meier method and multivariable cox proportional hazards regression. All analyses were conducted using complete-case analysis, and no imputation was performed for missing variables. All tests were two-sided, with p < 0.05 considered statistically significant.

Results

Basic characteristics of the involved cases

Figure S1 presents the flow diagram of this study. A total of 218 patients were included: 131 with vulvar cancer and 87 with vaginal cancer (Table 1). Among vulvar and vaginal cancer patients, 65 (49.6%) and 46 (52.9%) individuals underwent HPV testing, respectively. Meanwhile, 107 (81.7%) and 51 (58.6%) patients had available data on pathological differentiation, 130 (99.2%) and 87 (100%) patients had available data on primary treatment information, and 129 (98.5%) and 84 (96.6%) patients had available data on prognostic outcomes, respectively. All patients had complete information on FIGO stage and histological classification.

Table 1.

Clinicopathological characteristics of patients with vulvar and vaginal cancer

Clinical characteristics [n (%)]a	Vulvar cancer (n = 131)	Vaginal cancer (n = 87)	p-value
Age at diagnosis
<=54	44 (33.6)	32 (36.8)	0.367b
55–64	35 (26.7)	28 (32.2)
65–74	38 (29.0)	23 (26.4)
>=75	14 (10.7)	4 (4.6)
Median (IQR)	60.0 (50.5, 69.0)	58.0 (50.0, 66.5)	0.247c
FIGO stage
I	75 (57.3)	45 (51.7)	0.589b
II	19 (14.5)	18 (20.7)
III	29 (22.1)	17 (19.5)
IV	8 (6.1)	7 (8.0)
HPV status
Untested	66 (50.4)	41 (47.1)
Tested	65 (49.6)	46 (52.9)	< 0.001b
Positive	22 (33.8)	35 (76.1)
Negative	43 (66.2)	11 (23.9)
Histological Classification
Squamous cell carcinoma	114 (87.0)	72 (82.8)	0.242d
Adenocarcinoma	6 (4.6)	2 (2.3)
Other types	11 (8.4)	13 (14.9)
Differentiation
Not assessed	24 (18.3)	36 (41.4)
Assessed	107 (81.7)	51 (58.6)
Well differentiated	50 (46.7)	3 (5.9)	< 0.001d
Moderately to well differentiated	16 (15.0)	6 (11.8)
Moderately differentiated	24 (22.4)	20 (39.2)
Moderately to poorly differentiated	9 (8.4)	8 (15.7)
Poorly differentiated	8 (7.5)	12 (23.5)
Undifferentiated	0 (0.0)	2 (3.9)

a. IQR interquartile range, FIGO International Federation of Gynecology and Obstetrics, HPV human papillomavirus

b. Chi-square test

c. Wilcoxon rank-sum test

d. Fisher’s exact test

Most demographic characteristics were similar between vulvar and vaginal cancer patients (Table S1). However, a history of CIN was significantly more common in patients with vaginal cancer (33.8% vs. 9.3%, p < 0.001). Regarding clinicopathological features, the median age at diagnosis was comparable between vulvar (60.0 years) and vaginal (58.0 years) cancer patients, and both were predominantly diagnosed at FIGO stage I (57.3% vs. 51.7%, respectively). A marked difference was observed in HPV prevalence: among tested patients, the infection rate was significantly higher in the vaginal cancer cohort (76.1%) than in the vulvar cancer cohort (33.8%) (p < 0.001). Histologically, squamous cell carcinoma was the predominant type in both groups (87.0% and 82.8%, respectively). However, histological differentiation differed significantly (p < 0.001). Well-differentiated tumors were most common in vulvar cancer (46.7%), whereas moderately differentiated tumors predominated in vaginal cancer (39.2%).

The baseline comparison between patients with and without assessed differentiation showed that, in both vulvar and vaginal cancers, those with assessed differentiation had a higher proportion of squamous cell carcinoma (vulvar: p < 0.001; vaginal: p < 0.001; Table S2). Additionally, among vulvar cancer patients, the distribution of marriage status, Body Mass Index (BMI), and history of CIN differed between those with and without assessed differentiation (p = 0.015, p = 0.088, and p = 0.070, respectively; Table S2).

For HPV testing, baseline comparisons indicated that, except for job, all clinical and demographic characteristics were balanced between vulvar cancer patients with and without HPV testing (Table S3). In contrast, among vaginal cancer patients, those who underwent HPV testing were characterized by a younger age at diagnosis, lower FIGO stage, and a higher proportion with a history of CIN (p = 0.016, p = 0.058, and p < 0.001, respectively; Table S3).

Treatment strategies of the enrolled cohorts

Surgery served as the primary treatment for most vulvar cancer patients (84.6%), compared to only half of vaginal cancer patients (50.6%) (p < 0.001, Table 2). Notably, a significantly higher proportion of vulvar cancer patients underwent surgery without adjuvant therapy (70.0% vs. 47.7%, p = 0.016, Table 2).

Table 2.

Treatment strategies of patients with vulvar and vaginal cancer

Primary treatment/FIGO staging [n (%)]a	Total	p-valuec	I	II	III	IV	p-value
Vulvar cancer	n = 130b		n = 74b	n = 19	n = 29	n = 8
RT	6 (4.6)	< 0.001c	1 (1.4)	1 (5.3)	3 (10.3)	1 (12.5)	0.010d
CT	8 (6.2)		3 (4.1)	1 (5.3)	1 (3.4)	3 (37.5)
RT + CT	6 (4.6)		2 (2.7)	2 (10.5)	2 (6.9)	0 (0.0)
Surgical treatment	110 (84.6)		68 (91.9)	15 (78.9)	23 (79.3)	4 (50.0)
without RT/CT	77 (70.0)	0.016c	60 (88.2)	10 (66.7)	8 (34.8)	0 (0.0)	< 0.001d
combine with RT/CT	33 (30.0)		8 (11.8)	5 (33.3)	15 (65.2)	4 (100.0)
Vaginal cancer	n = 87		n = 45	n = 18	n = 17	n = 7
RT	9 (10.3)		1 (2.2)	2 (11.1)	5 (29.4)	1 (14.3)	< 0.001d
CT	17 (19.5)		6 (13.3)	4 (22.2)	4 (23.5)	3 (42.9)
RT + CT	17 (19.5)		5 (11.1)	6 (33.3)	3 (17.6)	3 (42.9)
Surgical treatment	44 (50.6)		33 (73.3)	6 (33.3)	5 (29.4)	0 (0.0)
without RT/CT	21 (47.7)		19 (57.6)	0 (0.0)	2 (40.0)	-	0.028d
combine with RT/CT	23 (52.3)		14 (42.4)	6 (100.0)	3 (60.0)	-

a. RT radiotherapy, CT chemotherapy, FIGO International Federation of Gynecology and Obstetrics 

b. One patient was excluded because of not reporting any treatment information

c. Chi-square test, comparations between vulvar and vaginal cancer patients

d. Fisher’s exact test

Primary treatment selection was strongly associated with FIGO stage in both cancers (vulvar: p = 0.010; vaginal: p < 0.001; Table 2), particularly in vaginal cancer where no stage IV patients underwent primary surgery. Furthermore, among patients who received surgery during the primary treatment phase, the use of adjuvant therapy varied significantly by FIGO stage in both vulvar (p < 0.001) and vaginal (p = 0.028) cancer cohorts.

Stage-by-stage comparisons confirmed that vulvar cancer patients were more likely to receive primary surgery at most stages (I-III: p < 0.05, Table S4). Additionally, among stage I and II patients who underwent surgery as primary treatment, vulvar cancer patients more frequently received additional adjuvant treatments than vaginal cancer patients (p = 0.001 and p = 0.012, respectively; Table S4).

Clinical characteristics and treatment patterns in patients with different HPV status

Given the established role of HPV infection in vulvar and vaginal carcinogenesis, we further analyzed differences in clinical characteristics and treatment patterns among HPV tested patients. For vulvar cancer cases, a non-significant trend toward younger age at diagnosis has been observed in HPV-positive patients (p = 0.072, Table 3). No other significant differences in clinical characteristics were observed for either cancer.

Table 3.

Clinical characteristics in patients who underwent HPV testing

Indicators/HPV status [n (%)]a	Positive	Negative	p-value
Vulvar cancer	n = 22	n = 43
Age at diagnosis
<=54	11 (50.0)	14 (32.6)	0.072b
55–64	1 (4.5)	14 (32.6)
65–74	9 (40.9)	12 (27.9)
>=75	1 (4.5)	3 (7.0)
Median (IQR)	58.5 (50.0, 67.5)	59.0 (45.5, 69.0)	0.939c
FIGO staging
I	8 (36.4)	28 (65.1)	0.122b
II	4 (18.2)	3 (7.0)
III	7 (31.8)	10 (23.3)
IV	3 (13.6)	2 (4.7)
Classification
Squamous cell carcinoma	20 (90.9)	35 (81.4)	0.730b
Adenocarcinoma	1 (4.5)	2 (4.7)
Other types	1 (4.5)	6 (14.0)
Differentiation
Not assessed	6 (27.3)	5 (11.6)
Assessed	16 (72.7)	38 (88.4)
Well differentiated	4 (25.0)	25 (65.8)	0.027b
Moderately to well differentiated	2 (12.5)	5 (13.2)
Moderately differentiated	4 (25.0)	3 (7.9)
Moderately to poorly differentiated	3 (18.8)	3 (7.9)
Poorly differentiated	3 (18.8)	2 (5.3)
Undifferentiated	0 (0.0)	0 (0.0)
Vaginal cancer	n = 35	n = 11
Age at diagnosis
<=54	16 (45.7)	6 (54.5)	0.235b
55–64	11 (31.4)	4 (36.4)
65–74	8 (22.9)	1 (9.1)
>=75	0 (0.0)	0 (0.0)
Median (IQR)	56.0 (48.5, 63.0)	53.0 (48.0, 59.0)	0.470c
FIGO staging
I	21 (60.0)	7 (63.6)	0.863b
II	9 (25.7)	2 (18.2)
III	3 (8.6)	2 (18.2)
IV	2 (5.7)	0 (0.0)
Classification
Squamous cell carcinoma	32 (91.4)	8 (72.7)	0.138b
Adenocarcinoma	0 (0.0)	0 (0.0)
Other types	3 (8.6)	3 (27.3)
Differentiation
Not assessed	16 (45.7)	3 (27.3)
Assessed	19 (54.3)	8 (72.7)
Well differentiated	1 (5.3)	1 (12.5)	0.122b
Moderately to well differentiated	3 (15.8)	1 (12.5)
Moderately differentiated	8 (42.1)	0 (0.0)
Moderately to poorly differentiated	2 (10.5)	3 (37.5)
Poorly differentiated	5 (26.3)	3 (37.5)
Undifferentiated	0 (0.0)	0 (0.0)

a. HPV human papillomavirus, IQR interquartile range, FIGO International Federation of Gynecology and Obstetrics 

b. Fisher’s exact test

c. Wilcoxon rank-sum test

Age distribution analysis revealed that over half of HPV-positive vulvar cancer patients were diagnosed within the 50–54 years and 65–69 years age groups, a pattern distinct from HPV-negative patients (Figure S2A, S2B). For vaginal cancer, HPV-positive diagnoses were more frequent between ages 45–69 (Figure S2C, S2D). Regarding treatment, no statistically significant differences were found in primary treatment modalities based on HPV status for either vulvar or vaginal cancer (Table 4).

Table 4.

Treatment patterns in patients who underwent HPV testing

Primary treatment/HPV status [n (%)]a	Positive	Negative	p-value
Vulvar cancer	n = 22	n = 43
RT	1 (4.5)	2 (4.7)	0.906b
CT	0 (0.0)	2 (4.7)
RT + CT	1 (4.5)	2 (4.7)
Surgical treatment	20 (90.9)	37 (86.0)
without RT/CT	12 (60.0)	26 (70.3)	0.624c
combine with RT/CT	8 (40.0)	11 (29.7)
Vaginal cancer	n = 35	n = 11
RT	4 (11.4)	1 (9.1)	0.575b
CT	5 (14.3)	1 (9.1)
RT + CT	8 (22.9)	5 (45.5)
Surgical treatment	18 (51.4)	4 (36.4)
without RT/CT	10 (55.6)	0 (0.0)	0.097b
combine with RT/CT	8 (44.4)	4 (100.0)

a. HPV human papillomavirus, RT radiotherapy, CT chemotherapy  

b. Fisher’s exact test

c. Chi-square test

The association between treatment patterns, HPV status, and prognosis

Among vulvar cancer patients, the median follow-up time was 17.4 months for OS and 18.7 months for DFS. During follow-up, 8 OS events were observed, with 122 patients censored at the time of last follow-up. For DFS, 20 events (recurrence or death) were recorded, while 110 patients were censored. Meanwhile, among vaginal cancer patients, the median follow-up time was 18.1 months for OS and DFS. During follow-up, 5 OS events were observed, with 79 patients censored at the time of last follow-up. For DFS, 8 events (recurrence or death) were recorded, while 76 patients were censored. Detailed results of the KM analyses are presented in Table S5-S6, with the corresponding survival curves shown in Figs. 1 and 2.

Fig. 1.

OS and DFS of vulvar cancer patients with different primary treatment. A-C OS survival curves for all/stage I-II/stage III-IV vulvar cancer patients stratified by primary treatment (surgery vs. non-surgery). D-F DFS survival curves for all/stage I-II/stage III-IV vulvar cancer patients stratified by primary treatment. ‘Surgery’ indicates patients who received surgical treatment as primary therapy, regardless of adjuvant radiotherapy/chemotherapy. ‘Median OS/DFS: NR’ indicates median survival not reached. NR: not reached, OS: overall survival, DFS: disease-free survival

Fig. 2.

OS and DFS of vaginal cancer patients with different treatment patterns. A-C OS survival curves for all/stage I-II/stage III-IV vaginal cancer patients stratified by primary treatment (surgery vs. non-surgery). D-F DFS survival curves for all/stage I-II/stage III-IV vaginal cancer patients stratified by primary treatment. ‘Surgery’ indicates patients who received surgical treatment as primary therapy, regardless of adjuvant radiotherapy/chemotherapy. ‘Median OS/DFS: NR’ indicates median survival not reached. NR: not reached, OS: overall survival, DFS: disease-free survival

KM analyses revealed that primary surgical treatment was associated with significantly longer OS in vulvar cancer (p < 0.001; Fig. 1A). This association, though attenuated, still existed in stage-stratified analyses (Stage I/II: p = 0.079; Stage III/IV: p = 0.053; Fig. 1B and C). A consistent association was also observed in the analysis of DFS (all patients: p < 0.001; stage I/II: p = 0.560; stage III/IV: p = 0.020; Fig. 1D and F). Meanwhile, in vaginal cancer, a non-significant trend was observed between primary surgical treatment and longer OS and DFS (OS: p = 0.142, DFS: p = 0.104; Fig. 2A and D). However, this association disappeared after further stratified analyses (Fig. 2).

Regarding HPV status, no significant differences in OS were observed based on HPV status for either cancer type (Figure S3A-S3B). A non-significant trend toward longer DFS was noted in HPV-positive vulvar cancer patients (p = 0.107; Figure S3D). These findings remained consistent in age-stratified analyses (Figure S4-S5).

Multivariable Cox regression analyses further demonstrated that, after adjusting for age at diagnosis, FIGO stage, histological subtype, and differentiation, and accounting for clustering effects by center, primary surgery remained associated with longer OS and DFS in vulvar cancer (OS: HR = 0.141 [95% CI: 0.048–0.416], p < 0.001; DFS: HR = 0.258 [95% CI: 0.133–0.501], p < 0.001; Table 5). In contrast, no significant association between primary treatment modality and survival outcomes was observed among vaginal cancer patients (Table 5).

Table 5.

Multivariable cox regression analyses among vulvar and vaginal cancer patients

Complete case analysisa	adjusted HR (95%CI)	adjusted p-value
Vulvar cancer (n = 106)
OS (events = 8)
Primary treatment (Surgical vs. Unsurgical)	0.141 (0.048–0.416)	< 0.001
Age at diagnosis (> 64 vs. < = 64)	4.858 (1.062–22.218)	0.042
FIGO staging (III/IV vs. I/II)	3.596 (0.401–32.241)	0.253
Classification (all other types vs. SCC)	4.265 (0.874–20.806)	0.073
Differentiation (Undifferentiated to Moderately to poorly differentiated vs.  Moderately differentiated to Well differentiated)	0.781 (0.351–1.740)	0.545
DFS (events = 20)
Primary treatment (Surgical vs. Unsurgical)	0.258 (0.133–0.501)	< 0.001
Age at diagnosis (> 64 vs. < = 64)	3.000 (1.048–8.592)	0.041
FIGO staging (III/IV vs. I/II)	4.828 (1.920-12.145)	< 0.001
Classification (all other types vs. SCC)	7.193 (3.765–13.743)	< 0.001
Differentiation (Undifferentiated to Moderately to poorly differentiated vs.  Moderately differentiated to Well differentiated)	0.276 (0.077–0.986)	0.047
Vaginal cancer (n = 51)
OS (events = 4)
Primary treatment (Surgical vs. Unsurgical)	1.405 (0.126–1.570)	0.782
Age at diagnosis (> 64 vs. < = 64)	0 (0-Inf)	< 0.001
FIGO staging (III/IV vs. I/II)	14.918 (5.606–39.700)	< 0.001
Classification (all other types vs. SCC)	2.963 (0.979–8.974)	0.055
Differentiation (Undifferentiated to Moderately to poorly differentiated vs.  Moderately differentiated to Well differentiated)	3.222 (0.490-21.202)	0.223
DFS (events = 5)
Primary treatment (Surgical vs. Unsurgical)	0.290 (0.047–1.806)	0.185
Age at diagnosis (> 64 vs. < = 64)	2.044 (1.305–3.202)	0.002
FIGO staging (III/IV vs. I/II)	8.757 (1.640-46.768)	0.011
Classification (all other types vs. SCC)	6.174 (2.456–15.520)	< 0.001
Differentiation (Undifferentiated to Moderately to poorly differentiated vs.  Moderately differentiated to Well differentiated)	1.394 (0.643–3.021)	0.400

a. Only patients with complete data on all variables were included in the multivariable Cox regression analysis; HRs and p-values have been adjusted for center clustering effect

HR hazards ratio, CI confidence interval, OS overall survival, DFS disease-free survival, FIGO Federation of Gynecology and Obstetrics, SCC squamous cell carcinoma

Discussion

To our knowledge, this multicenter retrospective study represents the first comprehensive analysis of clinicopathological characteristics, treatment patterns, and prognostic factors among vulvar and vaginal cancer patients within a Chinese population. Key findings reveal distinct profiles for these two malignancies: a significantly higher HPV prevalence in vaginal cancer (76.1%) compared to vulvar cancer (33.8%), a predominant use of surgery as primary treatment for vulvar cancer (84.6%) contrasting with more varied approaches for vaginal cancer, and a significant association between primary surgical treatment and longer OS specifically in vulvar cancer patients. Furthermore, we observed that HPV-positive vulvar cancer patients were diagnosed at a younger age, and a trend suggesting a potential survival advantage for HPV-positive status emerged. These findings contribute novel epidemiological and clinical insights specific to the Chinese context, where data on these rare gynecological cancers remain scarce.

The marked difference in HPV prevalence between vaginal (76.1%) and vulvar (33.8%) cancers in our cohort confirms the established etiological gradient seen internationally, while providing crucial population-specific data for China. Globally, vaginal cancers are over three times more likely to be HPV-attributable than vulvar cancers (77.8% vs. 25.0%) [1]. Notably, the HPV positivity in our vulvar cancer cases is lower than that reported in some prior Chinese studies (range: 39.8%–74.1%) [23–25]. This discrepancy may be attributed to genuine geographical variations in HPV attribution, differences in study periods reflecting evolving epidemiological patterns, or methodological heterogeneity. Although based on retrospective data with heterogeneous HPV detection methods, our study may better capture the “real-world” diagnostic landscape, as it incorporates data from real clinical HPV testing process, in contrast to prior studies that predominantly relied on highly sensitive PCR-based assays of tissue specimens [23–25]. For vaginal cancer, our finding aligns with the high attribution rate (83.3%) reported by Huang et al. [9], reinforcing its status as a predominantly HPV-driven malignancy in China. This etiological dichotomy underscores fundamentally different pathogenic pathways and may have downstream implications for prevention (e.g., HPV vaccination impact) and the exploration of biomarker-driven therapies.

Due to the rarity of vulvar and vaginal cancers, well-established treatment protocols remain limited. Our analysis shows that while both cancers present predominantly at early stages (FIGO I: 57.3% vs. 51.7%), their primary treatment strategies differ substantially. Surgery was the main primary treatment for vulvar cancer (84.6%), consistent with international guidelines that consider resection the cornerstone of management [12–15]. The frequent use of surgery alone in early-stage disease aligns with global standards. In contrast, treatment for vaginal cancer was more heterogeneous, with primary surgery used in only 50.6% of patients. While this figure underscores a substantial reliance on non-surgical primary modalities compared to vulvar cancer, it is notable that major guidelines, such as those from ESGO/ESTRO, emphasize definitive radiotherapy as the primary modality for most vaginal cancers due to superior organ preservation and comparable oncologic outcomes in light of the anatomical challenges [13]. The stage distribution in the present study aligns with some international reports [26, 27], and comprehensive contemporary data on these rare malignancies specifically within Chinese populations are notably scarce. However, limited data availability and sample size precluded a detailed characterization of primary treatment strategies and a comprehensive depiction of clinical management for vulvar and vaginal cancers in China. Nonetheless, given the rarity of these malignancies and the paucity of data, our findings help fill critical evidence gaps and describe the current practice for the vulvar and vaginal cancers treatment in China, identifying areas where local practice aligns with or diverges from international patterns.

The most compelling clinical insight from our survival analysis is the significant association between primary surgical treatment and longer OS and DFS specifically for vulvar cancer, a relationship that persisted after stage stratification and adjusted for other clinical characteristics in multivariable cox regression analyses. This finding reinforces the cornerstone role of surgery in vulvar cancer management and aligns with the survival advantage reported in another Chinese study [28]. The contrast with some Western data, such as the UK study by Smith et al. which showed a less definitive benefit, invites important questions [11]. However, in that study, the researchers did not perform effective analyses stratified by cancer stage, leaving the potential impact of staging unclarified. In our study, even after stratifying patients by stage, surgical treatment remained associated with longer OS, strongly support the continued central role of radical surgery within the context of Chinese practice, Similarly, a non-significant trend was observed between primary surgical treatment and longer OS and DFS in vaginal cancer (OS: p = 0.142, DFS: p = 0.104). Despite this interpretation should be tempered by missing key variables and baseline imbalances, which required a complete-case analysis that reduced sample size and may limit generalizability, these findings still contribute meaningful evidence and offer new perspectives for current clinical practice. Further studies in various populations from different countries and regions could better validate the above findings.

The prognostic implications of HPV infection are well-documented in several cancers but remain less defined in vulvar and vaginal malignancies [17, 18]. Our study found that the median age at diagnosis for vulvar (60.0 years) and vaginal (58.0 years) cancers aligned with national registry data. Meanwhile, a non-significant trend toward younger age at diagnosis has been observed in HPV-positive vulvar cancer patients. This observation adds a layer to the dual-pathway model of vulvar carcinogenesis, suggesting that HPV-positive cases may represent a distinct clinical phenotype with potentially different risk factors or host-tumor interactions. While our study was underpowered to confirm a statistically significant independent prognostic effect of HPV status on survival, the consistent trend towards longer DFS in HPV-positive vulvar patients echoes the well-known favorable prognostic signature of HPV in oropharyngeal and anal cancers [17, 18, 29, 30]. This suggestive signal positions HPV status as a potential candidate biomarker for risk stratification in these cancers, meriting investigation in larger, prospective cohorts. Nevertheless, the relatively low rate of HPV testing, together with the current lack of association between HPV status and treatment choice indicates that de-intensification or intensification strategies based on HPV are not yet part of clinical decision-making for vulvar or vaginal cancers. Our study lays the groundwork for investigating whether such biomarker-driven strategies could be relevant in these gynecologic malignancies.

Positioning our findings within the global cervical cancer elimination initiative highlights a critical opportunity for synergistic prevention of other HPV-associated cancers, notably vulvar and vaginal malignancies. The substantial HPV-attributable fraction, particularly in vaginal cancer as confirmed in this study, underscores the foundational role of primary prevention through HPV vaccination. Meta-analyses and real-world evidence corroborate that current vaccines offer high efficacy against high-grade precancerous lesions of the vulva and vagina [31]. Concurrently, secondary prevention should be enhanced by leveraging existing cervical cancer screening platforms. As recommended by expert consensus in China, opportunistic visual inspection of the vulva and vagina during routine cervical screening can improve early detection rates for these rare cancers [6]. This integrated approach, leveraging vaccination for future risk reduction and optimizing current screening platforms for early detection, is essential to decrease the burden and improve outcomes for these rare yet preventable gynecological cancers.

Several limitations must be acknowledged. The retrospective design, limited detailed treatment information, and non-standardized HPV testing inherently limit the precision of our prevalence estimates and introduce potential information bias. The sample size, though substantial for a study of these rare cancers, constrained the statistical power for detailed subgroup and survival analyses by HPV status. Additionally, baseline imbalances between patients with and without assessed differentiation and HPV testing further reduced the effective sample size in the complete-case analysis and limited the generalizability of the survival analyses findings. Meanwhile, the exclusion of patients with severe comorbidities or inability to provide informed consent may have preferentially removed older and frailer individuals, also limiting the generalizability of the survival analyses findings. Future studies with larger sample size from diverse populations across different countries and regions are warranted to overcome these limitations. Despite these limitations, the strengths of our study are significant: its multicenter design enhances generalizability within China, the direct comparative framework offers unique insights into differential management and outcomes, and the analysis of real-world practice fills a critical gap in the literature.

In conclusion, this multicenter study provides a comprehensive profile of vulvar and vaginal cancers in China, highlighting distinct etiological and clinical pathways. Our findings confirm the higher HPV-attributable fraction in vaginal cancer, reinforce the critical importance of surgical management in vulvar cancer, and identify promising avenues for future research, particularly regarding the prognostic implications of HPV status, offering evidence directly relevant to clinical practice in China. Furthermore, the common causative link to HPV positions the prevention of vulvar and vaginal cancers as an ideal extension of the global cervical cancer elimination agenda. A unified prevention framework, spanning primary, secondary, and tertiary levels, can thus be strategically deployed for synergistic impact across these HPV-associated diseases.

Abbreviations

HPV

Human papillomavirus

CIN

Cervical intraepithelial neoplasia

FIGO

Federation of Gynecology and Obstetrics

DFS

Disease‑free survival

OS

Overall survival

IQR

Interquartile range

HR

Hazards ratio

CI

Confidence interval

BMI

Body Mass Index

SCC

Squamous cell carcinoma

Authors' contributions

FZ was responsible for conceptualization and funding acquisition. FZ and CD participated in project administration, study design, supervision, data interpretation, and revised the manuscript. XZ, HW, YH and YZ performed the statistical analysis, participated in methodology and software development, and drafted the manuscript. XZ, HW, YH, QD, YZ, SH, MY, HY, YW, XC, YZ, SZ, WW, GH and LL carried out the studies, participated in data collection, curation and investigation. All authors read and approved the final manuscript.

Funding

This work was supported by a Merck Grant (# MISP 59476). The funding body had no role in the design of the study, data collection, analysis, interpretation, or manuscript writing.

Data availability

The data underlying this article is not publicly available. Deidentified data could be shared upon reasonable request to the corresponding author.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Ethics Committee of the National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (Approval No.: NCC 20/041 2237). All procedures were performed in accordance with the relevant guidelines. Written informed consent was obtained from all participating patients during routine follow-ups. For deceased or unreachable patients, the requirement for informed consent was waived by the ethics committee.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.