Vaginal intraepithelial neoplasia (VAIN) is rare, with an incidence of approximately 0.2–0.3 cases per 100,000 women in the United States.1 Its diagnosis is typically made through colposcopic-directed biopsy after an abnormal Pap test result. Most patients are asymptomatic, but many women report postcoital spotting or unusual vaginal discharge. Risk factors for development of vaginal dysplasia and carcinoma include older age,2 low socioeconomic status,3,4 smoking,4 history of genital warts,3 immunosuppression,5 history of diethylstilbestrol exposure,6 human papillomavirus infection,4 previous or concomitant cervical dysplasia,7,8 previous pelvic radiation therapy,9 and hysterectomy.10 Many similarities between the development of cervical intraepithelial neoplasia (CIN) and vaginal dysplasia have been documented.
Unlike the well-characterized natural history of CIN, the low prevalence of VAIN has limited evaluation of its natural history, optimal treatment, and recommended surveillance. As the precursor lesion to invasive vaginal carcinoma, recent interest has been placed on establishing optimal screening recommendations for individuals at high risk and on identifying risk factors affecting recurrence.10–12 There is no current consensus regarding the optimal management of VAIN. A wide variety of therapeutic modalities have been used with varying success,13–17 and the necessity of treating low-grade vaginal dysplasia remains unknown.10 Additionally, the recommended surveillance of women with VAIN remains controversial. The objective of this study was to evaluate the natural history of VAIN and to identify risk factors for increased time to normalization, persistence, recurrence, and progression to invasive vaginal carcinoma.
MATERIALS AND METHODS
After obtaining approval from the Health Policy and Services of the Department of Defense and Institutional Review Board approval from the Walter Reed National Military Medical Center, we conducted a Department of Defense–wide search of electronic medical records and identified all women with VAIN examined at military treatment facilities between January 2002 and January 2012. Patients were identified by International Classification of Diseases, 9th Revision, Clinical Modification codes for vaginal dysplasia using the Patient Administration Systems and Biostatistics Activity database. Study inclusion required patients to have a histologic diagnosis of VAIN, a minimum of a 12 months of follow-up from the time of initial diagnosis, and at least one postdiagnosis or posttreatment evaluation. Women who had a primary diagnosis of invasive vaginal carcinoma with concomitant primary VAIN were excluded because the goal of this study was to examine the time from diagnosis of VAIN to vaginal cancer. The data analysis start time of 2002 was used because this marks the initiation of the current electronic medical record system utilized by the Department of Defense.
Eligible electronic medical records were reviewed, and the following demographic data were collected: age, gravidity, parity, body mass index, and smoking status. Pertinent comorbidities, such as history of diethylstilbestrol exposure, immunosuppression, history of cervical dysplasia, previous treatments for cervical dysplasia, or history of other gynecologic malignancies, were noted. For patients who had undergone a hysterectomy, the date of surgery, indication for procedure, and pathologic diagnosis for cervix and uterus were documented. The date and results of all Pap tests, colposcopic findings, histologic diagnoses, and the treatments administered were then chronologically recorded, in addition to the date of last follow-up examination. In women with multiple grades of VAIN on initial diagnosis, the highest grade of VAIN was assigned as the final histologic diagnosis. Low-grade dysplasia was defined as VAIN 1, and high-grade dysplasia was defined as VAIN 2 and 3. To better-evaluate whether time to normalization was affected by differences in surveillance intervals, we also analyzed the intensity of follow-up in patients with low-grade and high-grade dysplasia by evaluating the number of surveillance visits and histologic biopsies performed after the time of diagnosis.
All patients were reviewed by two separate investigators blinded to each other's results to ensure accurate and complete data. After compiling the database, a third reviewer performed quality assurance for one-third of cases selected at random and adjudicated minor disagreements, finalizing a de-identified database. Additionally, a secondary search was performed to identify all patients who received International Classification of Diseases, 9th Revision, Clinical Modification codes for vaginal cancer in the Patient Administration Systems and Biostatistics Activity database. This search was performed as a check to make sure that patients with vaginal cancer and a preceding diagnosis of VAIN were not missed during the data query in error.
The normalization, persistence, and recurrence rates of low-grade and high-grade vaginal dysplasia were calculated. Normalization was defined as negative cytology, colposcopic examination, or vaginal biopsy after a diagnosis of VAIN. Persistence was diagnosed when repeat vaginal biopsy confirmed the same or higher histologic grade of vaginal dysplasia. Recurrence was defined as regression of VAIN with subsequent redevelopment of histologically confirmed vaginal dysplasia. The time to normalization was calculated from the date of diagnosis in patients who did not undergo treatment and from the date of treatment for patients who underwent treatment.
Continuous data are described using the mean (±standard deviation) and compared between groups using the two-sample t test. Categorical variables are presented as counts with percentages and compared between groups using Fisher exact test. Ordinal data are presented as medians (with minimum–maximum ranges) and compared using the Wilcoxon rank-sum test.
Univariate analyses of the associations of the grade of dysplasia, age, smoking status, history of cervical dysplasia, previous hysterectomy, and treatment with time to normalization were performed using Kaplan-Meier survival curves, the log rank test, and Cox proportional hazards models. Multivariate analysis of these risk factors was then explored using Cox regression. Results are presented as median time to normalization with 95% confidence intervals (95% CIs) and hazard ratios (HRs). P<.05 was considered statistically significant. All statistical analyses were performed using SPSS for Windows 19.0.
Our search identified 188 patients with a reported diagnosis of VAIN. Of these, 61 patients (32%) were excluded; 35 were excluded for lack of pathologic confirmation of disease and 26 were excluded for not fulfilling the 12 months of follow-up for various reasons. Thus, the final study population comprised 127 patients with median follow-up of 34 months (range 12–169 months). The initial histopathology results diagnostic of VAIN were determined at 23 different military treatment facilities.
The mean age was 47.4 years, and the mean body mass index (calculated as weight (kg)/[height (m)]2) was 26.4. Overall, there were no statistically significant differences in patient characteristics among the different grades of VAIN (Table 1).
A history of cervical dysplasia was documented in 61 patients (48%) and cervical cancer was documented in 11 patients (9%); neither of these were predictors of VAIN severity (P=.99 and P=.52, respectively). Two patients (2%) reported diethylstilbestrol exposure in utero, one patient (1%) had transplant-associated immunosuppression, and one patient (1%) was positive for human immunodeficiency virus. Hysterectomy preceded diagnosis of VAIN in 96 patients (76%), and in 37 (39%) of these patients there was a documented history of cervical dysplasia or carcinoma. The median interval between hysterectomy and VAIN diagnosis was 11 years (range 0–38 years). There was no statistically significant difference in history of hysterectomy among the different grades of VAIN (P=.68).
Of the 127 patients, 75 (59%) had low-grade vaginal dysplasia and 52 patients (41%) had high-grade vaginal dysplasia. Initial follow-up visit for the patients after histologic diagnosis of vaginal dysplasia included cytology alone (79 patients [62%]), cytology and colposcopy (23 patients [18%]), colposcopy with biopsy (23 patients [18%]), and colposcopy without biopsy (two patients [2%]). All patients were followed-up with cytology at some point during additional follow-up visits, whereas 81 patients (64%) also underwent subsequent colposcopic evaluations.
Fifty-three patients (42%) underwent treatment of vaginal dysplasia: 15 of 75 (20%) with low-grade vaginal dysplasia and 38 of 52 (73%) with high-grade vaginal dysplasia. The various therapeutic modalities used included surgery (wide local excision or vaginectomy), laser ablation, topical 5-fluorouracil, estrogen, and Imiquimod. The two most common therapies used were laser ablation (27 patients [51%]) and surgery (12 patients [23%]).
To evaluate the natural history of VAIN, we analyzed patients who underwent no treatment. Of the 74 patients who received no treatment, 64 (87%) demonstrated normalization, seven (9%) demonstrated persistence, and three (4%) experienced recurrence of disease. No patients had progression to vaginal cancer. Two patients with persistent disease had progression from VAIN 1 to VAIN 2 over a 3-year period without treatment; both were treated (one with surgery and one with laser ablation) and subsequently demonstrated normalization.
Of the 53 patients who underwent treatment, 49 (92%) demonstrated normalization and four (8%) demonstrated persistence of disease. No patients with VAIN 1 who underwent treatment had progression to a higher grade of dysplasia. Additionally, no patients demonstrated recurrence of disease or progression to vaginal cancer after treatment. Normalization, persistence, and recurrence rates did not differ by grade of dysplasia or treatment status (Table 2).
In sum, 113 patients (89%) demonstrated normalization of disease. Among these patients who experienced normalization, 54 patients (48%) met normalization criteria based on cytology alone, 23 patients (20%) met criteria based on cytology and colposcopy without biopsy, 22 patients (19%) met criteria based on cytology and colposcopy with biopsy, 10 patients (9%) met criteria based on histology alone, and four patients (3.5%) met criteria based on only colposcopic impression.
Eleven patients (9%) had persistent disease and three patients (2%) had recurrent disease. Four of the 11 patients with persistent dysplasia (two patients with VAIN 1 and two patients with VAIN 3) had undergone treatment, whereas seven patients (six with VAIN 1 and one with VAIN 2) had not (Table 3). Median follow-up was 34 months (range 12–169 months) from the time of diagnosis. All three patients (2%) who demonstrated recurrence of disease had untreated VAIN 1. Recurrence occurred at a median of 17 months from the time of initial diagnosis (Table 4).
We next evaluated the time to normalization of patients with low-grade and high-grade vaginal dysplasia. When comparing all patients (those treated and untreated), median time to normalization was longer in patients with low-grade dysplasia than in those with high-grade dysplasia (15.9 months compared with 10.0 months; HR 1.5; 95% CI 1.004–2.1; P=.045) (Fig. 1). Thus, patients with high-grade dysplasia experienced normalization 37% sooner than those with low-grade dysplasia.
When evaluating only patients with low-grade vaginal dysplasia, longer median times to normalization were noted among patients who underwent treatment, although this finding did not reach statistical significance: median time to normalization was 15.9 months for patients who were not treated compared with 27.1 months for patients who underwent treatment (HR 1.0; CI 0.5–1.8; P=.99) (Fig. 2). In a similar fashion, the median time to normalization for patients with high-grade vaginal dysplasia who did not undergo treatment was 9.0 months in comparison with 10.0 months for those with high-grade vaginal dysplasia who underwent treatment (HR 1.1; CI 0.6–2.2; P=.69 (Fig. 3).
The univariate association of risk factors with time to normalization is presented in Table 5. A higher grade of vaginal dysplasia was significantly associated with a shorter time to normalization. No other factors were found to be significant in univariate or multivariate analysis.
To determine other factors that may have had a role in differences in normalization between high-grade and low-grade VAIN, we analyzed the intensity of follow-up visits for these patients. Although the median interval of visits was longer in patients with low-grade VAIN, this difference was not statistically significant (6.9 months compared with 4.5 months; P=.958). Additionally, there were no significant differences when comparing patients with high-grade and low-grade dysplasia regarding the number of visits until normalization (three compared with two; P=.3) or regarding the interval between initial diagnosis and first follow-up visit (5 months compared with 6 months; P=.053). However, there was a statistically significant difference in total number of vaginal biopsies performed during all follow-up visits; patients with high-grade VAIN underwent an average of 3.3 biopsies, whereas patients with low-grade VAIN underwent 2.5 biopsies (P=.045).
In our secondary search, six patients were found to have a history of vaginal cancer. The mean age at diagnosis was 60 years (range 36–72 years). One patient had stage I vaginal cancer diagnosed and the other patients had stage II or stage III diagnosed. Only one patient had Pap tests results for the 5 years preceding the diagnosis of vaginal cancer; all Pap test results were normal until her diagnosis of stage II vaginal cancer. Two other patients had a Pap test result that demonstrated vaginal cancer, but there were no records of results before these tests. The remaining three patients had no records available before diagnoses because their diagnoses were performed at the initiation of the current electronic medical record system or they were referred to and received care outside the military health care system.
In our study, 89% of patients demonstrated normalization of VAIN and none had progression to invasive vaginal cancer during a median surveillance of almost 3 years. Normalization, persistence, and recurrence rates did not differ by grade of dysplasia or treatment. However, dysplasia grade was found to be an independent prognostic factor of time to normalization (P=.045) (Table 5). Interestingly, although follow-up intervals were similar among patients with low-grade and high-grade dysplasia, the mean number of biopsies was higher for patients with high-grade dysplasia. Because lesions can often be small, this “biopsy effect” may have played some role in the difference in time to normalization.
Another possible explanation for this difference may be attributable to the underlying processes causing low-grade and high-grade dysplasia. Previous studies evaluating the prevalence of HPV in VAIN found that 90–100% of patients with VAIN have HPV.18 Interestingly, HPV types differ between low-grade and high-grade vaginal dysplasia, which may partially explain the different clinical behaviors between these groups.
The National Cancer Institute reports that the mean age of diagnosis of vaginal cancer is 70 years.19 The mean ages of patients with VAIN and vaginal cancer in our study were 47.4 and 60 years, respectively. Although we did not identify any patients with a diagnosis of vaginal cancer that was preceded by VAIN, we feel our data regarding patients with vaginal cancer are too limited to make conclusions. Previous studies evaluating VAIN have produced conflicting results regarding progression and normalization.10,20–29 Aho et al20 reported a rate as high as 9% for progression of VAIN to vaginal cancer. Similar to the study performed by Massad,21 we found no progression of VAIN to invasive vaginal cancer. However, our lack of progression may be attributed to the freely accessible military health care system, but it also may be attributable to the likely 10–20 years of lag time between VAIN and the development of vaginal cancer, based on median ages of diagnoses. Although our median surveillance of 34 months is lengthy, it may have been too short of an interval to detect patients who will later experience development of vaginal cancer.
Previously published regression rates of VAIN are similar to those of our study and range from 69% to 95%.20–28 Similar to our data, Stillman et al10 reported that a history of CIN did not correlate with persistence or progression of VAIN, whereas Dodge et al28 found that VAIN occurred most often among women with CIN or vulvar intraepithelial neoplasia. Additionally, previous investigators have noted multifocality as a risk factor for recurrence.26,28 A recent study by Gunderson et al29 demonstrated fairly high recurrence rates among all grades of VAIN. Although not statistically significant, patients with VAIN 1 had a recurrence or progression rate of 71% compared with patients with VAIN 3 who had a recurrence or progression rate of 31%.29 Recurrence in this study included patients with follow-up abnormal Pap test results, and our study limited this diagnosis to only patients with a documented histologic recurrence of VAIN, which may explain their high reported rate of recurrence.
Although our data and those of other studies did not demonstrate a high rate of recurrence or progression, we acknowledge that some reports have. Our data add to the limited literature about a rare diagnosis and highlight its likely indolent course. As such, surveillance intervals should mirror the natural history of the disease process. Therefore, we feel it is reasonable to practice a more liberal surveillance schedule for patients with VAIN with annual cytology and colposcopy. A majority of our patients underwent colposcopy at their follow-up visits, and we feel it is reasonable to perform both cytology and colposcopy at these visits because assessing the entire vagina with a spatula may not be adequate. The interval of 12 months is reasonable because we found that the median time to normalization was 10 months for high-grade dysplasia and 16 months for low-grade dysplasia. Furthermore, this is in line with recent changes to Pap screening with less frequent intervals.30 Finally, patients with VAIN should be followed-up indefinitely. Although our dataset represents one of the longest follow-up periods in the literature, the average age of vaginal cancer occurrence is 70 years, which is more than 20 years past the mean age of diagnosis of VAIN. It should be noted that closer follow-up for patients with immunosuppression may be warranted, because they were not well-represented in our study.
Clinicians must determine whether treatment of vaginal dysplasia is warranted or if surveillance will suffice. We continue to support treatment of high-grade vaginal dysplasia given previous reports of progression to vaginal carcinoma. Although there were similar rates of regression irrespective of treatment, there might have been other clinical details not captured that would have explained the decision to treat, such as size and characteristics of the individual lesion.
Limitations of our study include its retrospective nature and possible differences in the military population and that of the general public. There was no standard protocol for surveillance, but follow-up intervals of 4.5–6.9 months likely are in line with very conservative management. Additionally, we did not examine p16 staining because this was not routinely performed. Strengths of our study include the large number of patients with VAIN and complete medical records, and a median follow-up of almost 3 years.
Many of the previous studies of the natural history of VAIN have been limited in both number of patients and duration of follow-up. Our investigation of 127 patients with frequent follow-up visits over a median surveillance of 3 years highlights the indolent nature of this disease, despite grade and treatment. As such, treatment decisions for VAIN should be individualized, but more liberal surveillance intervals are likely adequate. Because 11% of patients with VAIN will either experience recurrence or have persistent disease, lifetime surveillance is recommended.
1. Henson D, Tarone R. An epidemiologic study of cancer of the cervix, vagina, and vulva based on the Third National Cancer Survey in the United States. Am J Obstet Gynecol 1977;129:525–32.
2. Johnston GA Jr, Klotz J, Boutselis JG. Primary invasive carcinoma of the vagina. Surg Gynecol Obstet 1983;156:34–40.
3. Brinton LA, Nasca PC, Mallin K, Schairer C, Rosenthal J, Rothenberg R, et al.. Case-control study of in situ and invasive carcinoma of the vagina. Gynecol Oncol 1990;38:49–54.
4. Madsen BS, Jensen HL, van den Brule AJ, Wolfhart J, Frisch M. Risk factors for invasive squamous cell carcinoma of the vulva and vagina—population-based case-control study in Denmark. Int J Cancer 2008;122:2827–34.
5. Hankins CA, Lamont JA, Handley MA. Cervicovaginal screening in women with HIV infection: a need for increased vigilance? CMAJ 1994;150:681–6.
6. Richart RM. The incidence of cervical and vaginal dysplasia after exposure to DES. JAMA 1986;255:36–7.
7. Choo YC, Anderson DG. Neoplasms of the vagina following cervical carcinoma. Gynecol Oncol 1982;14:125–32.
8. Geelhoed GW, Henson DE, Taylor PT, Ketcham AS. Carcinoma in situ of the vagina following treatment for carcinoma of the cervix: a distinctive clinical entity. Am J Obstet Gynecol 1976;124:510–16.
9. Liao JB, Jean S, Wilkinson-Ryan I, Ford AE, Tanyi JL, Hagemann AR, et al.. Vaginal intraepithelial neoplasia (VAIN) after radiation therapy for gynecologic malignancies: a clinically recalcitrant entity. Gynecol Oncol 2011;120:108–12.
10. Sillman FH, Fruchter RG, Chen YS, Camilien L, Sedlis A, McTigue E. Vaginal intraepithelial neoplasia: risk factors for persistence, recurrence, and invasion and its management. Am J Obstet Gynecol 1997;176:93–9.
11. So KA, Hong JH, Hwang JH, Song SH, Lee JK, Lee NW, et al.. The utility of the human papillomavirus DNA load for the diagnosis and prediction of persistent vaginal intraepithelial neoplasia. J Gynecol Oncol 2009;20:232–7.
12. Frega A. Predictors of persistent vaginal intraepithelial neoplasia in previously hysterectomized women by high-risk HPV DNA detection. Cancer Lett 2007;249:235–41.
13. Wee WW, Chia YN, Yam PK. Diagnosis and treatment of vaginal intraepithelial neoplasia. Int J Gynaecol Obstet 2012;117:15–7.
14. Murta EF, Neves MA, Sempionato LR, Costa MC, Maluf PJ. Vaginal intraepithelial neoplasia: clinical-therapeutic analysis of 33 cases. Arch Gynecol Obstet 2005;272:261–4.
15. Krebs HB. Treatment of vaginal intraepithelial neoplasia with laser and topical 5-fluoruracil. Obstet Gynecol 1989;73:657–60.
16. Kim HS. Risk factors for recurrence of vaginal intraepithelial neoplasia in the vaginal vault after laser vaporization. Lasers Surg Med 2009;41:196–202.
17. Yalcin OT. Vaginal intraepithelial neoplasia: treatment by carbon dioxide laser and risk factors for failure. Eur J Obstet Gynecol Reprod Biol 2003;106:64–8.
18. Insinga RP, Liaw KL, Johnson LG, Madeleine MM. A systematic review of the prevalence and attribution of human papillomavirus types among cervical, vaginal, and vulvar precancers and cancers in the United States. Cancer Epidemiol Biomarkers Prev 2008;17:1611–22.
19. Kosary CL. Cancer of the vagina. SEER survival monograph: Cancer survival among adults: U.S. SEER Program, 1988-2001, patient and tumor characteristics. Bethesda (MD): National Cancer Institute; 2007.
20. Aho M, Vesterinen E, Meyer B, Purola E, Paavonen J. Natural history of vaginal intraepithelial neoplasia. Cancer 1991;68:195–7.
21. Massad LS. Outcomes after diagnosis of vaginal intraepithelial neoplasia. J Low Genit Tract Dis 2008;12:16–9.
22. Benedet JL, Sanders BH. Carcinoma in situ of the vagina. Am J Obstet Gynecol 1984;148:695–700.
23. Audet-Lapointe P, Body G, Vauclair R, Drouin P, Ayoub J. Vaginal intraepithelial neoplasia. Gynecol Oncol 1990;36:232–9.
24. Schockaert S, Poppe W, Arbyn M, Verguts T, Verguts J. Incidence of vaginal intraepithelial neoplasia after hysterectomy for cervical intraepithelial neoplasia: a retrospective study. Am J Obstet Gynecol 2008;199:113.e1–113e5.
25. Kalogirou D, Antoniou G, Karakitsos P, Botsis D, Papadimitriou A, Giannikos L. Vaginal intraepithelial neoplasia (VAIN) following hysterectomy in patients treated for carcinoma in situ of the cervix. Eur J Gynaecol Oncol 1997;18:188–91.
26. Diakomanolis E, Stefanidis K, Rodolakis A, Haidopoulos D, Sindos M, Chatzipappas I, et al.. Vaginal intraepithelial neoplasia: report of 102 cases. Eur J Gynaecol Oncol 2002;23:457–9.
27. Rome RM, England PG. Management of vaginal intraepithelial neoplasia: a series of 132 cases with long-term follow-up. Int J Gynecol Cancer 2000;10:382–90.
28. Dodge JA, Eltabbakh GH, Mount SL, Walker RP, Morgan A. Clinical features and risk of recurrences among patients with vaginal intraepithelial neoplasia. Gynecol Oncol 2001;83:363–9.
29. Gunderson CC, Nugent EK, Elfrink SH, Gold MA, Moore KN. A contemporary analysis of epidemiology and management of vaginal intraepithelial neoplasia. Am J Obstet Gynecol 2013;208:410.e1–6.
© 2013 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
30. Screening for cervical cancer. Practice Bulletin No. 131. American College of Obstetricians and Gynecologists. Obstet Gynecol 2012;120:1222–38.