Clinical characteristics and diagnostic performance of vaginal squamous cell carcinoma after hysterectomy.

Observational OA: gold CC-BY-NC-ND-4.0
⚙ AI-generated summary by qwen3.7-flash, 2026-11-21 ⓘ

This retrospective cohort study of 227 patients found that vaginal squamous cell carcinoma after hysterectomy is predominantly HPV-driven, with regular surveillance associated with earlier stage diagnosis and higher cytology yields.

One-sentence paraphrase of the abstract; not a substitute for reading it. No clinical advice. How this works

⚙ AI-generated deep summary by qwen3.7-flash, 2026-11-21 · read from full text ⓘ

This retrospective cohort study analyzed clinical characteristics and diagnostic performance in 227 patients with primary vaginal squamous cell carcinoma at a Chinese tertiary referral center. The researchers found that the majority of cases occurred in women who had previously undergone hysterectomy, with most tumors strongly associated with high-risk human papillomavirus infection. Cytology and colposcopy demonstrated high diagnostic yields, and regular post-hysterectomy surveillance was linked to earlier-stage diagnosis, although the study acknowledges limitations regarding the lack of comprehensive pre-hysterectomy screening records. Relevance to endometriosis: adenomyosis is listed as one indication for previous hysterectomy in this patient cohort, though the paper's main focus is on vaginal cancer diagnostics rather than these benign or malignant gynecologic conditions themselves.

Read from the paper's body, not the abstract. Not a substitute for reading the paper. No clinical advice. How this works

Abstract

BackgroundVaginal squamous cell carcinoma (SCC) is a rare gynecological malignancy. Current guidelines do not recommend routine cancer screening after hysterectomy for non-cervical conditions based on the assumption of negligible risk. However, clinical characteristics of vaginal SCC in post-hysterectomy women remain inadequately defined, particularly for those with non-cervical indications.MethodsWe conducted a retrospective cohort study of 227 patients with vaginal SCC diagnosed at a Chinese tertiary center between 2017 and 2025. Clinical characteristics, high-risk human papillomavirus (HPV) status, diagnostic yields of cytology, colposcopy, ultrasound and MRI, and latency from hysterectomy to cancer diagnosis were analyzed.ResultsWe found that 66.5% (151/227) of vaginal SCC cases occurred in women with prior hysterectomy; 86.1% had surgery for cervical lesions and 13.9% for non-cervical indications. The overall high-risk HPV positivity rate was 94.7%, with HPV16 predominating (74.9%). Regular post-hysterectomy surveillance was associated with earlier stage at diagnosis: 90.6% of surveilled patients presented with stage I disease versus 65.8% of non-surveilled patients. Latency to diagnosis was significantly shorter for women with prior cervical lesions (median 46.5 versus 120 months). Cytology and colposcopy demonstrated diagnostic yields exceeding 90%, while imaging showed lower detection rates.ConclusionsVaginal SCC after hysterectomy is predominantly HPV-driven, including in women with non-cervical indications. Regular post-hysterectomy surveillance is strongly associated with earlier stage at diagnosis. While universal screening is not warranted due to disease rarity, these findings support individualized HPV-informed risk counseling for sexually active women after hysterectomy and reinforce current surveillance recommendations for those with prior cervical lesions.
Full text 50,833 characters · extracted from pmc-nxml · 5 sections · click to expand

Methods

The study was conducted at the Obstetrics and Gynecology Hospital of Fudan University, a major national referral center. Inclusion criteria: All patients diagnosed with vaginal SCC at our institution between May 2017 and May 2025 were consecutively enrolled. All cases underwent histologic confirmation by two independent gynecologic pathologists to establish the diagnosis, following the detailed pathological criteria described below. Exclusion criteria: Patients with concurrent invasive cervical or vulvar cancer, or a prior history of these cancers within 5 years, were excluded. Notably, prior hysterectomy was not an inclusion or exclusion criterion; patients were enrolled irrespective of hysterectomy status. Therefore, the high proportion of post-hysterectomy patients in the vaginal SCC subgroup reflects the natural composition of our diagnosed case series, though we recognize that referral patterns to a tertiary center may influence this composition. To ensure diagnostic consistency, all outside referrals received a centralized pathological review by a senior pathologist and a comprehensive colposcopic reassessment by a senior colposcopist to evaluate for invasive disease and delineate the lesion area. All available clinical data were recorded; collected variables encompassed demographics, medical history, histologic information, cytology, HR-HPV testing results, colposcopic impressions, and findings from B-ultrasound and pelvic MRI. It should be noted that comprehensive pre-hysterectomy screening records (especially HPV status) were not consistently available for all referred patients, which is a recognized constraint of the retrospective design. From an initial cohort of 341 patients with vaginal malignant tumors, 236 had a history of hysterectomy, and 105 had not undergone hysterectomy. Subsequent analysis focused on the 227 cases with primary vaginal SCC, with the indications for previous hysterectomy summarized in Table  1 . Among the 227 vaginal SCC patients, only one patient in the post-hysterectomy group had a documented condition (immune thrombocytopenia) potentially associated with altered immune status; no cases of HIV (human immunodeficiency virus), transplant, or chronic immunosuppressive therapy were identified. Table 1 Indications and rates of previous hysterectomy in 151 post-hysterectomy vaginal SCC patients Indications for hysterectomy Number % Cervical lesions Cervical cancer 10 6.6 CIN2/3 120 79.5 Non-cervical lesions Uterine cancer 2 1.3 Ovarian cancer 2 1.3 Myoma 15 9.9 Adenomyosis* 1 0.7 Pelvic organ prolapse 2 1.3 Total 151 100 *This one patient had a previous hysterectomy for both myoma and adenomyosis. CIN : cervical intraepithelial neoplasia. Indications and rates of previous hysterectomy in 151 post-hysterectomy vaginal SCC patients *This one patient had a previous hysterectomy for both myoma and adenomyosis. CIN : cervical intraepithelial neoplasia. This retrospective cohort study was designed to characterize patients who were diagnosed with vaginal SCC. As such, it cannot provide estimates of the prevalence or absolute risk of vaginal SCC in the general female population, the post-hysterectomy population, or subgroups thereof. Its strength lies in the detailed phenotyping of affected individuals. All cases underwent histologic confirmation by two independent gynecologic pathologists specializing in lower genital tract pathology. The diagnosis of primary vaginal SCC was established based on the following criteria: Tumor site: The tumor was confirmed to originate primarily in the vagina, with the epicenter located in the vaginal wall. Cases with concurrent cervical or vulvar involvement, or a prior history of these cancers within 5 years, were excluded. Histologic features: Diagnosis was based on examination of hematoxylin and eosin (H&E)-stained sections from colposcopically directed biopsies or surgical specimens. The histologic diagnosis of SCC required the presence of malignant epithelial cells demonstrating squamous differentiation, characterized by intercellular bridges, keratin pearl formation, or individual cell keratinization. Tumors were graded as well, moderately, or poorly differentiated based on the degree of keratinization, nuclear pleomorphism, and mitotic activity, according to standard World Health Organization (WHO) classification criteria for tumors of the female genital tract 10 . Immunohistochemistry (when indicated): In diagnostically challenging cases (e.g., poorly differentiated tumors or distinction from other histologic types), immunohistochemical staining was performed. Squamous differentiation was supported by positivity for p63, CK5/6, or p40, and negativity for markers of other lineages (e.g., PAX8, GATA3 for adenocarcinoma; S100, HMB45 for melanoma). P16 immunohistochemistry was also performed in selected cases to assess HPV association, with strong and diffuse block-positive staining (nuclear and cytoplasmic) considered indicative of HPV-driven etiology 11 . Diagnostic review process: All outside referral slides underwent centralized pathological review by a senior pathologist at our institution. Any discrepancies between the original diagnosis and the review were resolved through consensus discussion, with a third expert pathologist consulted if necessary. The final diagnosis was recorded only after complete agreement. Cytology was assessed via liquid-based Pap tests (SurePath TM ; Becton, Dickinson, USA). All results were reported using the third edition of the Bethesda System. The spectrum of diagnoses included: negative for intraepithelial lesion or malignancy (NILM), atypical squamous cells of undetermined significance (ASC-US), low-grade squamous intraepithelial lesion (LSIL), high-grade squamous intraepithelial lesion (HSIL), atypical squamous cells that cannot exclude HSIL (ASC-H), squamous cell carcinoma (SCC), atypical glandular cells (AGC), and adenocarcinoma (AC). Cervical or vaginal swab samples for HPV DNA testing were collected by experienced gynecologists during pelvic examination prior to any surgical intervention or colposcopic biopsy. The collection site was determined based on the patient’s anatomical status and clinical presentation: for patients with prior hysterectomy, samples were collected from the vaginal vault, with particular attention to the cuff and suture lines, using a swab rotated against the vaginal walls; for patients without prior hysterectomy (intact cervix), samples were collected from the cervix as part of routine cervical cancer screening. In these patients, a separate dedicated vaginal swab was not routinely obtained unless vaginal pathology was specifically suspected. It is important to note that in patients without prior hysterectomy, the HPV test results reflect cervical HPV status, which may not necessarily represent the HPV status of the vaginal lesion itself. This distinction is acknowledged as a study limitation. High-risk (HR) HPV DNA testing was conducted using the Cobas 4800 HPV assay (Roche Diagnostics, Indianapolis, IN, USA), hybrid capture-2 testing (Digene, Maryland, USA) or a fluorescence-based multiplex real-time HPV DNA genotyping kit (Bioperfectus, Jiangsu, China) 12 , which can detect HR-HPV types 16, 18, 31,33, 35, 39, 45, 51, 52, 56, 58, 66, and 68 13 . A patient was considered HR–HPV positive if any of these tests yielded a positive result from a cervical or vaginal swab sample. Conversely, a patient was classified as HR-HPV negative if all performed swab-based tests returned a negative result. It should be noted that HPV status was determined solely from exfoliative cell samples (swabs), and confirmatory testing for HPV DNA (e.g., PCR or in-situ hybridization) on the tumor tissue itself was not routinely performed as part of this retrospective study. Colposcopy was performed by experienced colposcopists within 6 months after cytology and/or HR-HPV testing. Postmenopausal patients received topical estrogen prior to the procedure. Indications for colposcopy included: abnormal cytology results (LSIL, HSIL, ASC-H, AGC, SCC, AC, and repeat ASC-US); HPV16, 18 infection; persistent infection of other HR-HPV subtypes; symptoms or signs of possible cervical, vaginal, or vulvar cancer, including any suspicious cervical abnormality found during a pelvic examination, abnormal genital tract bleeding, or unexplained cervico-vaginal discharge; first follow-up visit after cervical conization, or hysterectomy for cervical cancer or HSIL. Colposcopy was performed by experienced colposcopists following the 2011 International Federation of Cervical Pathology and Colposcopy (IFCPC) terminology 14 . A detailed description of colposcopic findings and initial diagnosis was recorded. Colposcopic impressions were categorized as normal or benign lesions, LSIL, HSIL, or suspected malignancy. Utilizing 5% acetic acid and Lugol’s solution, targeted biopsies were conducted on suspected lesions of the cervix, vagina, and vulva, which can reveal a range of conditions from inflammations and precancerous lesions to malignant tumors. For post-hysterectomy patients, the vaginal vault suture ends were meticulously inspected to avoid missing hidden lesions in the dimples. It is important to note that cytology, HR–HPV testing, and colposcopy were not performed as independent, blinded evaluations. In clinical practice, they form a sequential diagnostic pathway. Specifically, an abnormal cytology result and/or a positive HR–HPV test served as the primary indication for referral to colposcopy. Consequently, the colposcopist was typically aware of the patient’s prior screening results at the time of the examination, which could influence the scrutiny applied. This reflects real-world clinical management but may introduce incorporation bias, potentially enhancing the apparent sensitivity of colposcopy when performed in this triage context. The MRI examination was performed with a 1.5-T MR unit (Avanto; Siemens, Erlangen, Germany) with an eight-channel pelvic phased-array coil. For all examinations, the patients lay supine and freely breathed. The protocol routinely included T2-weighted imaging, diffusion-weighted imaging (DWI) with apparent diffusion coefficient (ADC) mapping, and dynamic contrast-enhanced T1-weighted imaging. The slice thickness for high-resolution sequences was 3–4 mm. Vaginal gel opacification was not routinely administered. Due to the retrospective nature of this study, the timing of various diagnostic examinations varied among patients. To standardize our analysis, we applied the following inclusion criteria for each modality: cytology and HR-HPV test results were included if obtained within 6 months prior to histologic diagnosis; colposcopic examinations were included if performed as part of the diagnostic workup leading to biopsy (typically within 6 weeks of diagnosis); B-ultrasound and pelvic MRI findings were included if performed within 3 months prior to diagnosis. These time windows were chosen to maximize the relevance of each test to the disease status at diagnosis while accommodating real-world clinical variability. The performance of cytology, HR-HPV testing, and colposcopy was evaluated for their efficacy in detecting vaginal neoplasia. For B-ultrasound and pelvic MRI, which are primarily indicated for staging established cancers or evaluating symptoms, we assessed their incidental lesion detection rate; that is, the frequency with which they identified a suspicious vaginal mass or abnormality that led to or was consistent with the subsequent histological diagnosis of SCC. This comparison aims to describe the real-world diagnostic journey rather than to equate the roles of these modalities. Statistical analyses were mainly performed using SPSS 27.0 software (IBM, Armonk, New York, USA). Continuous variables following a normal distribution are presented as mean ± standard deviation (SD) and were compared using the independent samples t -test. As the time interval from hysterectomy to vaginal SCC diagnosis was not normally distributed, comparisons between groups were performed using the Mann–Whitney U test. Categorical variables were presented as numbers and percentages and were compared using the Chi-square test or Fisher’s exact test, as appropriate. This included comparisons of detection rates between diagnostic modalities, HPV genotype distributions between groups, and histologic type frequencies based on hysterectomy status or indication. To evaluate factors associated with the latency period from hysterectomy to vaginal SCC diagnosis, we performed Kaplan-Meier survival analysis with the log-rank test. The “time” variable was defined as the interval (in months) between the first hysterectomy and the secondary surgery for vaginal SCC; the “event” was defined as the occurrence of vaginal SCC. Given that all patients with available follow-up data were alive at the time of analysis, traditional post-diagnosis survival analysis was not feasible. Multivariate Cox proportional hazards regression analysis was performed to assess the independent effects of HPV infection status (positive versus negative) and the cause of hysterectomy (cervical lesions versus non-cervical lesions) on the latency period. Hazard ratios (HR) with 95% confidence intervals (CI) were calculated. To account for potential differences in follow-up duration, we conducted separate analyses for all patients post-hysterectomy and those with latency >60 months post-hysterectomy. A two-tailed P value < 0.05 was considered to be statistically significant. Kaplan–Meier survival analysis and Cox proportional hazards regression were performed using R software (version 4.5.2, Boston, Massachusetts, USA). Due to the retrospective case-series design lacking a disease-negative control group, we were unable to calculate standard diagnostic accuracy metrics such as sensitivity, specificity, predictive values, or AUC. Instead, we assessed the diagnostic yield of each modality, defined as the proportion of patients with confirmed vaginal SCC in whom the test result was abnormal (e.g., ≥ASC-US for cytology, positive for HR–HPV, malignant impression on colposcopy, or suspicious findings on imaging) prior to histologic diagnosis. These yields reflect the clinical detection capability of each test within the diagnostic pathway of affected patients. This study was carried out according to the principles in the Declaration of Helsinki 1964 and all subsequent revisions. It was approved by the institutional ethics review board of the Obstetrics and Gynecology Hospital of Fudan University (2024-23-X1). Informed consent was obtained from all individual participants included in the study.

Results

A total of 227 patients with primary vaginal SCC were included, of whom 151 (66.5%) had undergone prior hysterectomy (post-hysterectomy group), and 76 (33.5%) had an intact uterus (no hysterectomy group). Among the post-hysterectomy patients, 130 (86.1%) had undergone hysterectomy for cervical lesions (cervical HSIL or cancer), and 21 (13.9%) for non-cervical indications (e.g., uterine myoma, adenomyosis, pelvic organ prolapse, Table  1 ). The mean age at vaginal SCC diagnosis was 54.4 ± 10.1 years, with no significant difference between the two groups (post-hysterectomy: 55.2 ± 8.2 years; no hysterectomy: 52.8 ± 12.9 years; P  = 0.089). The FIGO stage distribution at diagnosis was comparable between groups. In the post-hysterectomy group ( n  = 151), stage I accounted for 82.1% (124/151), stage II 6.0% (9/151), stage III 7.9% (12/151), and stage IV 4.0% (6/151). In the no hysterectomy group ( n  = 76), stage I accounted for 89.5% (68/76), stage II 3.9% (3/76), stage III 3.9% (3/76), and stage IV 2.6% (2/76). No significant difference in stage distribution was observed between the two groups ( χ ² = 2.193, P  = 0.533). This finding suggests that vaginal SCC may present at similar disease stages regardless of prior hysterectomy status. After excluding cases with missing data (20 of 227 patients), 207 patients were included in the analysis of multifocal disease. Overall, 47 patients (22.7%) had multifocal lesions involving multiple vaginal sites. Stratified by hysterectomy status, multifocal disease was present in 28 of 140 (20.0%) post-hysterectomy patients and 19 of 67 (28.4%) patients without prior hysterectomy. This difference was not statistically significant ( χ ² = 1.80, P  = 0.179). Of the 151 vaginal SCC cases after hysterectomy, 130 (86.1%) and 21 (13.9%) had the surgery for cervical and non-cervical lesions, respectively (Table  1 ). The overall interval ranged from 3 to 312 months (mean 64.0 ± 58.5 months, median 48 months, IQR 24–84 months). This interval was significantly shorter in patients with prior cervical lesions (51.9 ± 44.4 months, median 46.5 months, IQR 20–72 months) than in those with non-cervical lesions (138.6 ± 78.4 months, median 120 months, IQR 72–177.5 months, P  < 0.001, Fig.  1a ). We further stratified cases by a 5-year (60-month) threshold. Most cases (63.6%, 96/151) occurred within 5 years of hysterectomy, and the vast majority of these (94.8%, 91/96) had a history of cervical lesions. In contrast, among the 36.4% (55/151) of cases diagnosed after 5 years, 70.9% (39/55) had prior cervical lesions ( P  < 0.001). Fig. 1 The time span from the previous hysterectomy and distribution of HR-HPV in vaginal squamous cancer (SCC). a Box plot showing the time interval from hysterectomy to vaginal SCC diagnosis, stratified by surgical indication. Boxes represent interquartile ranges (IQR); whiskers represent 1.5× IQR. P value calculated by Mann–Whitney U test. *** P  < 0.001. b Overall HR-HPV genotype distribution in the entire vaginal SCC cohort ( n  = 179 HPV-positive cases). c HR-HPV genotype distribution in post-hysterectomy vaginal SCC patients ( n  = 122 HPV-positive cases). d HR-HPV genotype distribution in vaginal SCC patients without prior hysterectomy ( n  = 57 HPV-positive cases). e HR-HPV genotype distribution in post-hysterectomy vaginal SCC patients whose hysterectomy was performed for cervical lesions ( n  = 104 HPV-positive cases). f HR-HPV genotype distribution in post-hysterectomy vaginal SCC patients whose hysterectomy was performed for non-cervical lesions ( n  = 18 HPV-positive cases). a Box plot showing the time interval from hysterectomy to vaginal SCC diagnosis, stratified by surgical indication. Boxes represent interquartile ranges (IQR); whiskers represent 1.5× IQR. P value calculated by Mann–Whitney U test. *** P  < 0.001. b Overall HR-HPV genotype distribution in the entire vaginal SCC cohort ( n  = 179 HPV-positive cases). c HR-HPV genotype distribution in post-hysterectomy vaginal SCC patients ( n  = 122 HPV-positive cases). d HR-HPV genotype distribution in vaginal SCC patients without prior hysterectomy ( n  = 57 HPV-positive cases). e HR-HPV genotype distribution in post-hysterectomy vaginal SCC patients whose hysterectomy was performed for cervical lesions ( n  = 104 HPV-positive cases). f HR-HPV genotype distribution in post-hysterectomy vaginal SCC patients whose hysterectomy was performed for non-cervical lesions ( n  = 18 HPV-positive cases). Among the 151 post-hysterectomy patients, the margin status of the original hysterectomy specimens was available for 140 patients (92.7%). Positive margins, all of which were high-grade VaIN (no invasive carcinoma), were identified in 12 patients (8.6%). Of these 12 patients with documented residual VaIN at the margin, 9 (75%) developed vaginal SCC within five years, while only 3 (25%) did so after five years. These 9 cases account for 10.3% of all early-onset (≤5 years) vaginal SCCs in this cohort. It is noteworthy that the majority of early-onset cases (78/87, 89.7%) had no documented margin involvement. However, due to the lack of systematic preoperative colposcopic evaluation of the vagina in most patients prior to hysterectomy, the presence of undiagnosed, synchronous VaIN at the time of surgery cannot be excluded in these cases. To further investigate factors influencing the time interval from hysterectomy to vaginal SCC diagnosis, we performed Kaplan-Meier analysis and multivariate Cox regression, focusing on patients with follow-up durations of >60 months post-hysterectomy. Kaplan-Meier analysis (Fig.  2a, b ) showed that in all patients post-hysterectomy ( n  = 151), those with prior hysterectomy for non-cervical lesions ( n  = 21) had a significantly longer mean latency period compared to those with cervical lesions ( n  = 130) (log-rank P  60-month cohort (non-cervical: n  = 16; cervical: n  = 39; log-rank P  = 0.001; Fig.  2b ). Multivariate Cox regression found that in the post-hysterectomy cohort, multivariate Cox regression revealed that the cause of hysterectomy was independently associated with latency to vaginal SCC diagnosis. Compared to patients with cervical indications, those with non-cervical indications had a significantly longer latency (HR = 3.20, 95% CI: 1.97–5.21, P  60-month cohort: hysterectomy cause remained significant (HR = 2.63, 95% CI: 1.43–4.83, P  = 0.002), while HPV infection remained non-significant (HR = 0.70, 95% CI: 0.34–1.46, P  = 0.340). Fig. 2 Kaplan–Meier analysis of the latency period from hysterectomy to vaginal SCC diagnosis, stratified by surgical indication. a Patients post-hysterectomy (non-cervical: n  = 21; cervical: n  = 130; log-rank P  60 months post-hysterectomy (non-cervical: n  = 16; cervical: n  = 39; log-rank P  = 0.001). a Patients post-hysterectomy (non-cervical: n  = 21; cervical: n  = 130; log-rank P  60 months post-hysterectomy (non-cervical: n  = 16; cervical: n  = 39; log-rank P  = 0.001). Among the 151 post-hysterectomy patients, 142 (94.0%) had clearly documented information on post-surgical surveillance. Of these, 94 patients (66.2%) had regular follow-up (cytology/HPV testing) after hysterectomy, while 48 patients (33.8%) did not. The FIGO stage distribution differed significantly between the two groups. Early-stage (I) disease was observed in 90.4% (85/94) of surveilled patients compared to 64.6% (31/48) of non-surveilled patients ( χ ² = 14.19, P  < 0.001). Conversely, advanced-stage (III–IV) disease was present in only 5.3% (5/94) of the surveilled group versus 27.1% (13/48) of the non-surveilled group ( χ ² = 13.60, P  < 0.001). Statistical analysis confirmed a significant difference in stage distribution (full four-stage distribution: χ ² = 18.51, P  = 0.000). These findings indicate that regular post-hysterectomy surveillance was associated with significantly earlier stage at diagnosis for vaginal SCC. We further analyzed the relationship between post-hysterectomy surveillance and FIGO stage at diagnosis, stratified by the indication for hysterectomy. Among the 142 patients with documented follow-up information, 123 had prior cervical lesions, and 19 had non-cervical indications. Among patients with cervical indications, 69.1% (85/123) had regular follow-up, compared to 47.4% (9/19) in the non-cervical group. In the cervical lesion group, patients with regular surveillance (n = 85) presented with significantly earlier stage disease compared to those without surveillance (n = 38): stage I disease was 90.6% vs. 65.8% ( P  = 0.001), and advanced-stage (III–IV) disease was 5.9% versus 28.9% ( P  = 0.000). In the non-cervical lesion group, a similar trend was observed: stage I disease was 88.9% in surveilled patients ( n  = 9) versus 60.0% in non-surveilled patients ( n  = 10, P  = 0.153), and advanced-stage disease was 0% versus 20.0% ( P  = 0.156). However, due to the small sample size (n = 19), these differences did not reach statistical significance. Importantly, when comparing cervical and non-cervical groups directly, no significant differences in stage distribution were observed among surveilled patients ( P  = 0.442) or among non-surveilled patients ( P  = 0.503), suggesting that the benefit of surveillance on stage at diagnosis is consistent across both groups, though statistically underpowered in the non-cervical subgroup. These findings support the potential value of post-hysterectomy surveillance for early detection, including in women with benign surgical indications, and underscore the need for larger studies in this population. HR-HPV status was available for 189 of the 227 women with vaginal SCC, with an overall positivity rate of 94.7% (179/189), some patients were with multiple infections. HPV16 was the predominant genotype (74.9%, 134/179), followed by HPV58 (7.3%, 13/179), HPV52 (5.0%, 9/179), HPV18 (5.0%, 9/179), HPV33 (3.4%, 6/179), and HPV51 (3.4%, 6/179) (Fig.  1b ). Among the 151 patients with a prior hysterectomy, 126 were tested, of whom 96.8% (122/126) were HR-HPV positive. In this subgroup, HPV16 was also dominant (78.7%, 96/122), with HPV58 (5.7%, 7/122), HPV18 (3.3%, 4/122), HPV33 (2.5%, 3/122), and HPV52 (1.6%, 2/122) being the next most common (Fig.  1c ). The HR–HPV positivity and HPV16 prevalence did not differ significantly between patients whose hysterectomy was performed for cervical (97.2%, 104/107; HPV16: 79.8%, 83/104) versus non-cervical lesions (94.7%, 18/19; HPV16: 72.2%, 13/18; P  = 0.573 and P  = 0.468, respectively; Fig.  1e, f ). In the 76 patients without a prior hysterectomy, 90.5% (57/63) were HR–HPV positive. HPV16 remained the most prevalent genotype (66.7%, 38/57), though other types, notably HPV52 (12.3%, 7/57), contributed more substantially ( P  = 0.002) than in the post-hysterectomy group, followed by HPV58 (10.5%, 6/57), HPV18 (8.8%, 5/57), HPV51 (7.0%, 4/57), and HPV33 (5.3%, 3/57) (Fig.  1d ). This genotype distribution aligns with the established role of HPV16 and 18 in vaginal carcinogenesis 15 , 16 , while also highlighting the notable contribution of HPV52 and 58, which is a characteristic feature in Chinese populations 7 . The full ranking of HR-HPV genotypes across all subgroups is detailed in Fig.  1b–f . Cytology results were available for 173 of the 227 vaginal SCC patients, with an overall abnormality rate (≥ASC-US) of 90.8% (157/173). The positivity rate was comparable between patients with and without a previous hysterectomy ( P  = 0.804). Among those post-hysterectomy, the rate was 92.6% (88/95) for patients with a history of cervical lesions and 83.3% (15/18) for those with non-cervical lesions, a difference that was not statistically significant ( P  = 0.203). Detailed cytology findings are provided in Table  2 . Table 2 Detailed cytology test results of 209 women with vaginal squamous cell carcinoma Cytology test results NILM ASC-US LSIL HSIL ASC-H SCC Total Positivity Post-hysterectomy 10 17 4 46 23 13 113 91.2% For cervical lesions 7 13 4 41 22 8 95 92.6% For non-cervical lesions 3 4 0 5 1 5 18 83.3% No previous hysterectomy 6 20 4 14 12 4 60 90.0% Sum 16 37 8 60 35 17 173 90.8% Positivity indicates abnormal cytology test results (ASC-US or more severe)/total cytology. NILM negative for intraepithelial lesion or malignancy, ASC-US atypical squamous cells of undetermined significance, ASC-H atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion, LSIL low-grade squamous intraepithelial lesion, HSIL high-grade squamous intraepithelial lesion, SCC squamous cell carcinoma. Detailed cytology test results of 209 women with vaginal squamous cell carcinoma Positivity indicates abnormal cytology test results (ASC-US or more severe)/total cytology. NILM negative for intraepithelial lesion or malignancy, ASC-US atypical squamous cells of undetermined significance, ASC-H atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion, LSIL low-grade squamous intraepithelial lesion, HSIL high-grade squamous intraepithelial lesion, SCC squamous cell carcinoma. Colposcopic impression suggested malignancy in 94.0% (79/84) of vaginal SCC patients who underwent the examination prior to diagnosis. This high detection rate was consistent across patient groups, showing no significant difference based on hysterectomy status ( P  = 0.698). Similarly, among post-hysterectomy patients, a malignant impression was frequently reported regardless of the original surgical indication, with 93.9% (46/49) for prior cervical lesions and 100% (8/8) for non-cervical lesions ( P  = 0.472). Detailed findings are summarized in Table  3 . Table 3 Comparative colposcopic findings of vaginal squamous cell carcinoma Colposcopic impression HSIL Malignant tumor Total Positivity Post-hysterectomy 3 54 57 94.7% For cervical lesions 3 46 49 93.9% For non-cervical lesions 0 8 8 100% No previous hysterectomy 2 25 27 92.6% Positivity indicates colposcopic impression of malignant tumor/total colposcopic impression. HSIL high-grade squamous intraepithelial lesion. Comparative colposcopic findings of vaginal squamous cell carcinoma Positivity indicates colposcopic impression of malignant tumor/total colposcopic impression. HSIL high-grade squamous intraepithelial lesion. As imaging modalities are primarily used for staging or symptom evaluation rather than primary screening, we assessed their incidental lesion detection rate—the proportion of patients in whom a suspicious vaginal abnormality was identified prior to histologic diagnosis. B-ultrasound was performed in 158 patients prior to vaginal SCC diagnosis, with abnormal findings (such as vaginal mass, wall thickening, or hypervascularity) detected in 35.4% (56/158). The detection rate was significantly influenced by hysterectomy status, being more than twice as high in patients with a prior hysterectomy (42.3%, 47/111) than in those without (19.1%, 9/47; P  = 0.005). However, among post-hysterectomy patients, the detection rate did not differ significantly based on the original surgical indication (cervical lesions: 43.2% (41/95) vs. non-cervical lesions: 37.5% (6/16); P  = 0.672). Pelvic enhanced MRI was performed in 121 vaginal SCC patients prior to diagnosis, with suggestive findings (such as vaginal mass, thickening, or enhancement) in 68.6% (83/121). The detection rate was comparable between patients with (71.3%, 62/87) and without (61.8%, 21/34) a prior hysterectomy ( P  = 0.312). Among post-hysterectomy patients, the rate was similar regardless of the surgical indication (cervical lesions: 72.0% (54/75) vs. non-cervical lesions: 66.7% (8/12); P  = 0.705). MRI, as the primary modality for staging, demonstrated a significantly higher lesion detection rate (68.6%) compared to B-ultrasound (35.4%, P  = 0.000), yet it remained markedly lower than the detection rates achieved by cytology (90.8%), and colposcopic impression (94.0%) (all P- values = 0.000), confirming that mucosal-based methods are superior for primary detection.

Discussion

Vaginal SCC is a rare malignancy, and its low prevalence in both the general population and among post-hysterectomy women does not justify population-based screening. The value of this large, single-center cohort study lies not in quantifying population risk, but in providing a detailed clinical and etiological characterization of the women who do develop this disease. Our primary objective was to assess the clinical characteristics and diagnostic pathways of vaginal SCC in patients with prior hysterectomy, with a particular focus on those whose surgery was performed for non-cervical indications. Our study yielded several key findings. First, the majority of vaginal SCC cases in our cohort occurred in women with a prior hysterectomy (66.5%, 151/227), predominantly performed for cervical lesions (86.1%, 130/151). However, a non-negligible subset (13.9%, 21/151) had a hysterectomy for non-cervical indications. Second, the overall HR-HPV positivity rate was 94.7% (179/189), with HPV16 being the predominant genotype (74.9%, 134/179), followed by HPV58 (7.3%) and HPV52 (5.0%), a distribution consistent with both global reports and the distinct epidemiological profile of Chinese populations. Third, cytology and colposcopy demonstrated high diagnostic yields (>90%), significantly outperforming imaging modalities (B-ultrasound: 35.4%; MRI: 68.6%). Fourth, our Kaplan–Meier and Cox regression analyses revealed that the latency period from hysterectomy to vaginal SCC diagnosis was significantly shorter in women with prior cervical lesions compared to those with non-cervical indications, and this association remained robust after adjusting for HPV status and restricting to longer follow-up cohorts. Fifth, and importantly, among the 142 post-hysterectomy patients with documented follow-up information (94.0% of the cohort), regular surveillance was strongly associated with earlier stage at diagnosis, with 90.4% (85/94) of surveilled patients presenting with stage I disease compared to only 64.6% (31/48) of non-surveilled patients ( P  < 0.001). The significantly shorter latency observed in women with prior cervical lesions (median 46.5 months) compared to those with non-cervical indications (median 120 months) is notable ( P  < 0.001). This finding persisted in both Kaplan–Meier analysis and multivariate Cox regression. Several factors may explain this difference. First, women with prior cervical lesions were more likely to undergo regular post-hysterectomy surveillance: among patients with cervical indications, 69.1% (85/123) had regular follow-up, compared to 47.4% (9/19) in the non-cervical group. This surveillance intensity likely contributed to the earlier detection of vaginal abnormalities. Second, these women may harbor pre-existing, undiagnosed HPV persistence or vaginal intraepithelial neoplasia (VaIN) at the time of hysterectomy. The finding that 75% (9/12) of patients with histologically confirmed positive vaginal margins (VaIN) progressed to invasive SCC within five years provides direct evidence that residual high-grade VaIN at the time of hysterectomy carries a high risk of rapid progression to invasive cancer. This subset, however, represents only 10.3% (9/87) of all early-onset post-hysterectomy vaginal SCCs in our cohort. For the remaining 89.7% (78/87) of early-onset cases without documented margin involvement, it is plausible that many had synchronous, undiagnosed VaIN at the time of surgery, given that preoperative colposcopic assessment of the vagina was not routinely performed in most patients undergoing hysterectomy, particularly those with benign indications. Notably, pre-hysterectomy HPV status was available for the majority of patients, supporting the interpretation that many of these infections were long-standing rather than newly acquired. For women with non-cervical-indication hysterectomy, the lower rate of routine surveillance (47.4%) likely contributes to delayed diagnosis until symptoms develop, artificially prolonging the apparent latency. This interpretation is supported by our finding that among non-surveilled patients, advanced-stage (III-IV) disease was present in 27.1% (13/48) compared to only 5.3% (5/94) in surveilled patients. The consistent findings in longer-term (>60 months) cohorts suggest that this association is robust and not merely an artifact of differential follow-up duration. It is important to contextualize these findings within the established natural history of HPV-related carcinogenesis. HPV infection is most commonly acquired soon after sexual debut, and progression from persistent infection to invasive carcinoma typically spans 10–20 years. Although pre-hysterectomy HPV status was not available for most patients, we may surmise that vaginal SCCs diagnosed after hysterectomy, particularly those occurring within 5 years of surgery, more probably arise from pre-existing, undiagnosed HPV persistence or subclinical VaIN present at the time of hysterectomy, rather than rapid progression from de novo postoperative infections. This aligns with our observation that the shortest latencies occurred in women with prior cervical lesions, a population already at increased risk for persistent HPV infection. A methodological consideration when interpreting latency differences is surveillance intensity bias: women with prior cervical lesions underwent more regular follow-up (69.1% versus 47.4%), potentially leading to earlier detection and artificially shortened observed latency. Conversely, women with non-cervical indications had lower surveillance rates, with cancers more often diagnosed after symptoms developed, potentially prolonging apparent latency despite similar disease biology. Therefore, while our data demonstrate a clear association between surgical indication and time to diagnosis, the precise biological latency cannot be definitively determined from this retrospective cohort. A critical finding of our study is that HR-HPV positivity was uniformly high across all subgroups, regardless of the original indication for hysterectomy. In patients with prior hysterectomy for non-cervical indications, 94.7% (18/19) of tested cases were HR-HPV positive—a rate comparable to that in the cervical lesion group (97.2%, 104/107). This challenges the prevailing assumption that women undergoing hysterectomy for benign conditions are at negligible risk for HPV-related vaginal pathology. The lack of association between HPV status and latency in Cox regression analysis (HR = 0.86, 95% CI: 0.57–1.29 for the post-hysterectomy cohort; HR = 0.70, 95% CI: 0.34–1.46 for >60-month cohort) may initially seem counterintuitive. However, when >90% of cases are HPV-positive, HPV infection becomes a constant rather than a variable, and its power to discriminate between groups is necessarily limited. HPV16 was the predominant genotype, followed by HPV58, HPV52, HPV18, and HPV33. The dominance of HPV16 aligns with global reports 6 , 17 , 18 , while the notable prevalence of HPV58 and HPV52 represents a distinct characteristic of the Chinese population 19 , 20 . For patients without prior hysterectomy, HPV testing was performed on cervical samples as part of routine screening. While cervical and vaginal HPV infections are often concordant due to shared exposure and field carcinogenesis, we cannot definitively assume that the HPV genotype detected in the cervix is identical to that driving the vaginal SCC. This represents a limitation of our retrospective design. However, the extremely high HPV positivity rate (94.7%) and the predominance of HPV16 across all subgroups suggest a consistent etiologic role of HPV in vaginal SCC, regardless of sampling site. A further consideration is that HPV-negative status was determined solely by swab testing without routine tissue-based confirmation. While swab-based testing is standard in clinical practice, we cannot entirely exclude the possibility that rare HPV-driven tumors with low viral load or focal integration were misclassified as HPV-negative. Current guidelines recommend up to 25 years of surveillance for women with a history of CIN2+ following hysterectomy 5 . For women whose hysterectomy was performed for benign, non-cervical indications, current guidelines generally do not recommend any form of vaginal surveillance 21 – 23 . In our cohort, post-hysterectomy surveillance was strongly associated with earlier stage at diagnosis among women with a history of cervical lesions. A similar trend was observed in the smaller subgroup of women whose hysterectomy was performed for non-cervical indications, where surveilled patients also showed a higher proportion of early-stage disease. Although these differences did not reach statistical significance due to the limited sample size (n = 19), the absolute reduction in advanced-stage disease (20.0 percentage points) is clinically meaningful. Notably, among surveilled patients, stage distribution did not differ between the cervical and non-cervical indication groups, suggesting that the potential benefit of regular follow-up may extend to all post-hysterectomy women, regardless of the original surgical indication. These findings reinforce current guidelines that recommend continued surveillance for at least 25 years after treatment for high-grade cervical lesions. They also raise the question of whether selected women with benign-indication hysterectomy, particularly those who remain sexually active or have persistent HPV infection, might benefit from individualized risk assessment. However, the small number of non-cervical cases in our study precludes definitive conclusions, and larger prospective investigations are needed to establish the optimal surveillance strategy for this understudied population. Furthermore, while our data suggest a stage shift associated with surveillance, we cannot exclude the influence of lead-time bias or unmeasured confounders inherent to retrospective designs. Consistent with the superficial mucosal origin of vaginal SCC, cytology (90.8%, 157/173) and colposcopy (94.0%, 79/84) demonstrated exceptionally high diagnostic yields, significantly outperforming imaging modalities. These findings confirm that methods targeting exfoliated cells or directly visualizing mucosal abnormalities are inherently well-suited for detecting vaginal neoplasia. In contrast, B-ultrasound and MRI, which rely on the presence of mass lesions or deep tissue invasion, showed substantially lower detection rates (35.4%, 56/158 and 68.6%, 83/121, respectively). It is important to emphasize that these imaging modalities are not routine screening tools for vaginal SCC; their primary roles are in staging (MRI) and initial symptom evaluation (ultrasound). The moderate detection rate of MRI, while superior to ultrasound, remains inferior to direct tissue-based methods. Additionally, the lack of routine vaginal gel opacification in our MRI protocol may have further limited sensitivity for detecting small or superficial lesions. These findings underscore that abnormal vaginal cytology or HR-HPV positivity in post-hysterectomy women should prompt direct colposcopic evaluation rather than reliance on imaging for diagnosis. Our findings have several implications for clinical practice and future research. First, for women with a history of CIN2+ who undergo hysterectomy, the strong association between surveillance and early-stage diagnosis (90.6% stage I in surveilled versus 65.8% in non-surveilled cervical indication patients) underscores the value of regular assessment. Second, our data challenge the current guideline assumption of negligible risk in women undergoing hysterectomy for non-cervical indications. While the absolute risk for any individual woman in this group remains very low, the finding that 94.7% of their vaginal SCCs were HPV-positive, combined with the observation that surveilled patients in this subgroup showed a clinically meaningful trend toward earlier stage at diagnosis (88.9% stage I versus 60.0%), indicates that the biological potential for HPV-related carcinogenesis persists after surgery. This does not justify universal screening, but it does support a paradigm shift towards individualized, HPV-informed risk counseling for sexually active women post-hysterectomy. Clinicians should consider discussing the possibility of persistent HPV exposure and its oncogenic implications with these patients, particularly those with new or persistent HR-HPV infections detected incidentally. Third, the observation that 75% of patients with positive vaginal margins progressed to invasive SCC within five years underscores that for patients with known cervical lesions, thorough preoperative colposcopic evaluation of the entire vagina is essential to detect and treat synchronous VaIN prior to or at the time of hysterectomy. Fourth, the uniformly high HPV positivity across all subgroups suggests that HPV-based testing may serve as a more sensitive and etiology-driven approach to risk stratification than reliance on surgical history alone. Prospective studies are urgently needed to determine whether targeted surveillance of high-risk subgroups (e.g., women with persistent HR-HPV infection after hysterectomy) could be clinically meaningful and cost-effective. One major strength of our study is its large sample size of 227 primary vaginal SCC cases from a national referral center, enabling detailed analysis of clinical features and risk factors in this understudied population. The availability of post-hysterectomy follow-up data for 94.0% (142/151) of patients strengthens the reliability of our findings. The use of multiple diagnostic modalities also allowed a comprehensive comparison of their efficacy. Several limitations should be acknowledged. The retrospective design may introduce selection and information bias. The single-institution setting may limit generalizability. The predominance of post-hysterectomy patients in our cohort (66.5%) likely reflects both a true biological association and referral patterns to a tertiary center. Second, the absence of pre-hysterectomy HPV testing data limits our ability to determine whether post-hysterectomy SCCs arose from pre-existing infections or de novo acquisition. Third, for patients without prior hysterectomy, HPV testing was performed on cervical rather than vaginal samples. While cervical and vaginal HPV infections are often concordant, this represents a methodological consideration when interpreting HPV results for this subgroup. Finally, the relatively small number of patients with non-cervical indications ( n  = 21) limits statistical power for subgroup analyses, though the consistency of findings across multiple analytical approaches (Kaplan–Meier, Cox regression, stage distribution trends) strengthens our conclusions. In summary, this large clinical cohort study demonstrates that vaginal SCC in post-hysterectomy patients is overwhelmingly an HR–HPV-driven disease, occurring not only in women with prior cervical lesions but also in a non-negligible subset of those with non-cervical indications. The latency to cancer diagnosis is significantly shorter in women with prior cervical lesions, likely reflecting a combination of closer surveillance and pre-existing disease. Regular post-hysterectomy surveillance is strongly associated with earlier stage at diagnosis, reinforcing the importance of completing recommended follow-up in women with a history of CIN2+. While the rarity of vaginal cancer precludes recommendations for universal screening, our findings challenge the assumption of negligible risk in women with non-cervical-indication hysterectomy and support the consideration of individualized, HPV-informed risk assessment for sexually active women post-hysterectomy. Prospective studies with detailed longitudinal follow-up are urgently needed to define optimal, evidence-based surveillance strategies for this understudied population.

Introduction

Accounting for only about 2% of gynecological tumors, primary vaginal cancer is a rare malignancy. Nevertheless, it presents a serious health risk due to diagnostic difficulties and limited therapeutic choice 1 . The predominant histology is squamous cell carcinoma (SCC), with adenocarcinoma, melanoma, and sarcoma being less common 2 . Although the implementation of cervical cancer screening has successfully curbed cervical cancer rates, a considerable clinical gap persists in the screening and prevention of vaginal SCC, particularly after hysterectomy 3 , 4 . Current guidelines from the American Society for Colposcopy and Cervical Pathology (ASCCP) generally do not recommend continued vaginal cancer screening after hysterectomy in patients without a history of cervical intraepithelial neoplasia 2 (CIN2)+ within the past 25 years or who have completed the 25-year surveillance period 5 . This can lead to the misconception that no further gynecologic screening is necessary. However, these women may remain sexually active and thus still at risk for new or persistent high-risk human papillomavirus (HR-HPV) infection, which can lead to vaginal intraepithelial neoplasia (VaIN) or vaginal SCC 6 . It is important to note that HPV acquisition typically occurs near sexual debut, and the development of invasive carcinoma usually requires persistent infection over many years or decades. Therefore, vaginal SCC diagnosed after hysterectomy is more likely to reflect the progression of a pre-existing, long-standing HPV infection, possibly present at the time of surgery, rather than rapid carcinogenesis from a recently acquired infection. Nonetheless, the presence of an intact, susceptible vaginal mucosa post-hysterectomy underscores the importance of understanding the full clinical trajectory of HPV-related disease in this population. This oversight underscores the critical need to understand the clinical characteristics of post-hysterectomy vaginal SCC to improve patient outcomes. Vaginal SCC, the predominant histologic type, typically arises from the superficial squamous epithelium and progresses through vaginal intraepithelial neoplasia (VaIN) before becoming invasive. Even at early invasive stages, the disease often presents as mucosal abnormalities (e.g., plaques, erosions, or exophytic lesions) rather than deep tissue masses. This superficial origin has important implications for diagnostic approaches, as methods targeting the mucosal surface may achieve higher detection rates than those relying on deep tissue assessment. Although the epidemiology and prognostic factors of vaginal cancer have been studied 7 , 8 , the clinical characteristics and risk factors specific to vaginal SCC following hysterectomy remain inadequately characterized. Furthermore, the role of HR-HPV and the diagnostic yield of cytology and colposcopy in this context, particularly among women whose hysterectomy was performed for non-cervical indications, are yet to be fully elucidated 3 , 9 . To address these gaps, we conducted a retrospective cohort study of 227 patients with primary vaginal SCC at a Chinese tertiary referral center. We found that most cases occurred after hysterectomy and were strongly associated with HR-HPV infection. Cytology and colposcopy showed high diagnostic yields, while regular post-hysterectomy surveillance was associated with earlier-stage diagnosis. These findings support careful characterization of post-hysterectomy vaginal SCC and individualized, HPV-informed risk counseling without implying a need for universal screening.

Supplementary Material

Description of Additional Supplementary Data Supplementary Dataset 1 Description of Additional Supplementary Data Supplementary Dataset 1

Text is read by the "Ask this paper" AI Q&A widget below. Extraction quality varies by source — PMC NXML preserves structure cleanly, OA-HTML may include some navigation residue, and OA-PDF can have broken hyphenation. The publisher copy (via DOI) is the canonical version.

My notes (saved in your browser only)

⚙ Ask this paper AI returns verbatim quotes from the full text · source: pmc-nxml ⓘ

Answers must be backed by verbatim quotes from this paper's full text. Hallucinated quotes are dropped automatically; if no verbatim passage answers the question, we say so. How this works

Citation neighborhood (no data yet)

We don't have any in-corpus citations linked to this paper yet. This is a recent paper (2026) — citers typically take a year or two to land, and the OpenAlex reference graph may still be filling in.

SciLite annotations

organisms 16
human papillomavirus siv/hiv human immunodeficiency virus human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus type 16 human papillomavirus human papillomavirus type 16
chemicals 3
haematoxylin estrogen acetic acid

Source provenance

europepmc
last seen: 2026-11-22T09:57:54.571072+00:00
scilite
last seen: 2026-11-22T10:17:46.714424+00:00
unpaywall
last seen: 2026-11-16T06:37:28.113292+00:00
License: CC-BY-NC-ND-4.0