Rising and Falling Trends in Incidence Rates of Vulvar and Vaginal Cancers in the United States, 2000-2016

Background: Few studies focus on incidence trends in vulvar and vaginal by histological type, race, age, and region. We aimed to evaluate the temporal incidence trends and survival for the two cancers. Methods: Cases with primary vulvar or vaginal carcinoma from 2000-2016 were identied from the Surveillance, Epidemiology, and End Results 18 registries. Annual percent change (APC) and average APC (AAPC) were calculated to evaluated trends. Relative survival was estimated to compare survival outcomes. Results: Vulvar cancer incidence rates were highest among non-Hispanic whites (20.9 per 1,000,000 person-years) and in the Midwest (27.7), while vaginal cancer rates were highest among non-Hispanic blacks (8.2) and in the South (6.7). Overall, vulvar cancer rate increased 0.4% annually (AAPC, 0.4%; 95% CI, 0.1% to 0.6%), but vaginal cancer rate decreased 1.4% annually (AAPC, -1.4%; 95% CI, -2.0% to -0.5%). The increased vulvar cancer rate was only observed for squamous cell carcinoma (APC, 1.0%), while the decreased vaginal cancer rate was only observed for other malignancies (APC, -1.7%). Vulvar cancer rates increased only among whites (APC, 1.2%), but vaginal cancer rates decreased only among whites (AAPC, -1.3%) and Asians (APC, -2.7%). Vulvar cancer rates increased among patients aged 50-59 (APC, 1.1%), 60-69 (APC, 1.3%), and 70-79 (APC, 1.2%) years, while vaginal cancer rates decreased among patients aged 50-59 (APC, -1.4%) and 80 (APC, -1.4%) years or older. Overall, whites had the highest ve-year relative survival for both vulvar (72.22%) and vaginal (49.03%) cancers. Conclusion: There in vulvar and vaginal cancer incidence rates attributable to histological type, race, age, and region. The increased vulvar cancer rates and the declined vaginal cancer rate highlights the importance of further studies to investigate the reason for the disparities.

During 1999-2015 in the US, the incidence rate of vulvar SCC increased by 1.3% per year in women aged 50-69 years old, and the incidence rate of vaginal SCC decreased by 0.3% per year in women aged younger than 40 years and 70 years or older (5). Moreover, no studies are characterizing the incidence and survival rates of ADE and OM for vulvar and vaginal carcinomas due to their rareness. Furthermore, although the trends in vulvar and vaginal cancers have been documented, it is unclear whether these changes are equal or not when classi ed by histological type, race, and region. To the best of our knowledge, previous studies have not taken these features into account when evaluating the incidence and survival rates of these two cancers. Racial disparities of incidence and survival rates were well documented in several common carcinomas such as malignant meningioma, hepatocellular carcinoma, uterine corpus carcinoma, prostate carcinoma, and pancreatic carcinoma (6)(7)(8)(9)(10)(11). However, little is known about the racial disparities in the incidence and survival rates of vulvar and vaginal carcinomas.
The primary aim of this study was to evaluate the age-adjusted incidence and ve-year survival of vulvar and vagina carcinomas by race, histological subtype, and region using the Surveillance, Epidemiology, and End Results (SEER) database of the US. The secondary aim was to estimate the temporal trends of incidence rate over time by race, histological subtype, and region.

Study population
We identi ed the incidence rates in patients with microscopically con rmed cancers of the vulva (labium majus, C51.0; labium minus, C51.1; clitoris, C51.2; overlapping lesion of the vulva, C51.8; vulva NOS, C51.9) and vagina (C52.9) diagnosed between 2000 and 2016 from the SEER 18 database, representing approximately 28% of the US population, using the third edition of the International Classi cation of Diseases for Oncology (ICD-O-3) codes. We included patients above 20 years old and excluded those younger than 20 years due to the rareness of these cancers in this age group. Cases were categorized by race into non-Hispanic white (hereafter referred to as whites), non-Hispanic black (blacks), non-Hispanic American Indian/Alaska Native (Indians), Non-Hispanic Asian or Paci c Islander (Asians), and Hispanics. Incidence rates were age-adjusted to that of the 2000 US standard population and expressed per 1,000,000 woman-years. Rate ratios were estimated, and the 95% con dence intervals (CI) were calculated using the Tiwari modi cation method.
Trends in incidence rates were calculated and plotted using the National Cancer Institute Joinpoint regression software (version 4.7.0.0). In Joinpoint software, we calculated annual percent changes (APCs) and 95% CIs using the Empirical Quantile method (12,13). The software program chose the besttting linear regression model in a logarithmic scale based on the least-weighted Bayesian Information Criterion to identify years when APCs signi cantly changed, allowing a maximum of three joinpoints, while providing a minimum number of joinpoints necessary to t the data. In most situations, a single segment ts the data best (single APC). If more than one APC was estimated, trends were summarized by the average APC (AAPC).
We estimated overall ve-year relative survival in the patients, and then by race and histological type diagnosed during 2000-2016. Cases not microscopically con rmed and had missing data of survival time were excluded. Expected survival was estimated using the Ederer II method. Cases in the survival cohort were matched to the socioeconomic, geographic, and race annual life tables (SES/Geography/Race Annual Life Tables) by age, sex, race, calendar year, and county of residence at the time of cancer diagnosis (14). Relative survival was estimated using the Actuarial method and age-adjusted to the International Cancer Survival Standard Population, which is for cancer sites with increasing incidence by age, including vulva and vagina cancers. The relative survival was de ned as the ratio of the observed to the expected survival rate. All statistical tests were two-sided, and a P-value of less than 0.05 was considered to indicate statistical signi cance.

Results
Age-adjusted incidence rates (2000 to 2016) We identi ed 18,681 primary vulvar cancer cases and 5,691 primary vaginal cancer cases from 2000 to 2016 in the SEER 18 database. Overall, the age-adjusted incidence rate was 18.1 per 1,000,000 womanyears for vulva carcinoma and 5.5 per 1,000,000 woman-years for vaginal carcinoma. The incidence rates varied considerably by race, with the highest vulvar cancer incidence rate in whites (20.9) and the highest vaginal cancer rates in blacks (8.2). Moreover, the lowest incidence rates for vulvar and vaginal cancers were all demonstrated in Asians (7.1 and 3.6, respectively). The incidence rate varied widely by histological type and was highest in SCC (13.7 and 3.5 for vulvar and vaginal cancers, respectively), followed by OM (4.1 and 1.1), and ADE (0.3 and 0.9) ( Table 1).
The incidence rate overall varied regionally and was highest in the Midwest for vulvar cancer (23.7), and in the South for vaginal cancer (6.7). However, after stratifying by histological type and race, no marked regional difference was observed in most subsets ( Table 2).
The incidence rates increased monotonously with age and were highest among patients aged 80 years or older overall and by histological type (Table 3).
Five-year relative survival overall and by histologic subtype and race The overall ve-year relative survival rates were 72.3% and 49.2% in women with vulvar and vaginal cancers, respectively. The ve-year relative survival of women with vulvar and vaginal cancer was 69.1% and 52.4% for SCC, 63.5% and 53. 8 for ADE, and 83.6% and 37.6% for OM, respectively. The highest relative survival rates for both vulvar (72.7%) and vaginal (50.2%) cancers were in the whites, but the pattern was inconsistent after stratifying by histological type (Table 4 and Fig. 3).

Discussion
The Principal ndings of this study are that vulvar cancer incidence rates were highest and increasing among non-Hispanic whites, while vaginal cancer rates were highest and stable among non-Hispanic blacks but declining among whites and Asians. Whites have the highest ve-year relative survival rate for both vulvar and vaginal cancers.
As we discovered in this study, the vulvar SCC incidence rates have been increasing, supporting previous observations evaluating vulvar SCC incidence in the United States (15,16), consistent with the ndings of previous studies suggesting increasing vulvar cancer incidence rates in Norway, Germany, and Australia (17)(18)(19). We also found that the incidence rate of vaginal OM had been declining, but the incidence rate of vulvar OM was stable, which was consistent with the ndings of previous studies(20, 21). Although our research ruled out the probability that increased vulvar incidence was attributable to misclassi cation of vaginal cancer to vulvar cancer because of rate changes in different histological types and races for the two cancers, it remained unclear why vulvar cancer and vaginal cancer incidence rates had contrasting temporal trends. It is likely that unknown risk factors drove the increasing incidence for vulvar SCC and the decreasing incidence of vaginal OM, for example, life-style and genetic susceptibility.
Consistent with a nding reported in a previous study(5) based on the SEER database, our study con rmed the increasing incidence rates in vulvar SCC patients aged 50-59 and 60-69 years in the US during the past decades. Additionally, we observed an increased vulvar SCC incidence among patients aged 70-79 years. Moreover, we found a stable vaginal SCC incidence in all age groups, which was in contrast to a published study that identi ed a decreased vaginal SCC incidence in women younger than 40 years and 70 years or older(5). We found a decreased vaginal OM incidence among patients aged 50-59, 60-69, and 80 years or older.
To the best of our knowledge, this study was the rst to evaluate vulvar and vaginal cancer incidence rates by histologic type among Asians and Hispanics. Although the incidence rates for the two cancers were lower in these groups, vaginal cancer incidence has decreased rapidly among Asians, which underscores the need for a study investigating the potential protective factors among those women.
Another unique characteristic of this study was assessing incidence by geographic region. Overall, the vulvar incidence rate was highest in the Midwest, and the vaginal incidence rate was highest in the South. However, after stratifying by histological type and race, no signi cant regional difference in rates was observed within most subsets. It was likely that the small size in each subset limited the statistical e cacy.
The strength of our study is the detailed incidence rate trends by histological type, race, age, and region. The stability of this population-based cohort (SEER 18) ensured that the same geographic regions were represented over the entire duration of the study. Further, the detailed incidence records of newly diagnosed cases, as well as an extended period covered, allowed for APC calculations and trends observation. Furthermore, a large sample size of patients aged from 20 to 80 years or older, almost complete follow-up data, and high-quality control of the SEER program, make our results generalizable. Our study was subjected to some limitations. It was a retrospective SEER-based study. SEER did not provide information on established risk factors for vulvar and vaginal cancer, such as human papillomavirus. Therefore, we could not directly assess the variation in vulvar and vaginal cancer incidence and trends between histological types, race, and geographical regions to these risk factors. Finally, although SEER registries use standardized codes and progress to classify race and ethnicity data, the initial collection of this information is carried out by health care facilities and practitioners, and misclassi cation in a small proportion of cases is possible.
Given the racial disparities, clinicians would pay more attention to patients in special groups with higher incidence rates and worse survival, which is in favor of making clear the reason for the disparities. The results of this study would promote policymakers to inspect whether the health care delivery needs of patients in different racial groups are fully met or not, which is of great bene t. Ongoing surveillance for the two cancers using high-quality population-based registries is crucial to monitor cancer rates and trends.

Conclusions
There were profound disparities in vulvar and vaginal cancer incidence rates attributable to histological types, race, age, and region. Overall, the vulvar cancer incidence rate increased by 0.4% annually from 2000 to 2016, and vaginal cancer incidence rate decreased by 1.4% annually in the same period. Vulvar cancer incidence only increased for SCC, while vaginal cancer incidence only decreased for OM. The vulvar cancer incidence rate was highest and increased among non-Hispanic whites, while vaginal cancer incidence rate was highest among blacks but decreased among whites and Asians. This study demonstrates the most recent and comprehensive evaluation of differences in incidence trends and survival for vulvar and vaginal cancers in the United States. Identifying the differences can provide critical information to improve survival, explore etiology, and satisfy health care delivery needs.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and material
The datasets analyzed in this study are available in the SEER database, https://seer.cancer.gov/.

Competing interests
The authors declare that they have no competing interests. Tables Table 1 Age-adjusted incidence rates of vulvar and vaginal cancers and by histological subtype and race