Articles

Prevalence of human papillomavirus DNA and p16INK4a

positivity in vulvar cancer and vulvar intraepithelial
neoplasia: a systematic review and meta-analysis
Zhuang Li*, Penglin Liu*, Ziying Wang, Zhaoyang Zhang, Zhongshao Chen, Ran Chu, Guiju Li, Qiuyue Han, Yong Zhao, Li Li, Jinwei Miao†,
Beihua Kong†, Kun Song†

Summary
Background Human papillomavirus (HPV) DNA and p16INK4a positivity have crucial roles in the pathogenesis of vulvar Lancet Oncol 2023
cancer and vulvar intraepithelial neoplasia. We aimed to examine the pooled prevalence of HPV DNA and p16INK4a Published Online
positivity in vulvar cancer and vulvar intraepithelial neoplasia worldwide. March 15, 2023
https://doi.org/10.1016/
S1470-2045(23)00066-9
Methods In this systematic review and meta-analysis, we searched PubMed, Embase, and the Cochrane Library
*Contributed equally as joint
databases for studies published between Jan 1, 1986, and May 6, 2022, that reported the prevalence of HPV DNA, or first authors
p16INK4a positivity, or both, in histologically verified vulvar cancer or vulvar intraepithelial neoplasia. Studies on a †Contributed equally as joint last
minimum of five cases were included. Study-level data were extracted from the published studies. Random effect authors
models were used to examine the pooled prevalence of HPV DNA and p16INK4a positivity in both vulvar cancer and Department of Obstetrics and
vulvar intraepithelial neoplasia, which were further investigated using stratified analyses by histological subtype, Gynaecology, Qilu Hospital of
geographical region, HPV DNA or p16INK4a detection method, tissue sample type, HPV genotype, publication year, and Shandong University, Jinan,
Shandong, China (Z Li MSc,
age at diagnosis. Additionally, meta-regression was applied to explore sources of heterogeneity. Z Wang MSc, Z Zhang MSc,
Z Chen MD, R Chu MD, G Li Msc,
Findings We retrieved 6393 search results, of which 6233 were excluded for being duplicates or after application of our Q Han MSc, Y Zhao MSc, L Li PhD,
inclusion and exclusion criteria. We also identified two studies from manual searches of references lists. 162 studies Prof B Kong PhD,
Prof K Song PhD); Department
were eligible for inclusion in the systematic review and meta-analysis. The prevalence of HPV in vulvar cancer of Gynecological Oncology,
(91 studies; n=8200) was 39·1% (95% CI 35·3–42·9) and in vulvar intraepithelial neoplasia (60 studies; n=3140) was Beijing Obstetrics and
76·1% (70·7–81·1). The most predominant HPV genotype in vulvar cancer was HPV16 (78·1% [95% CI 73·5–82·3]), Gynaecology Hospital, Capital
followed by HPV33 (7·5% [4·9–10·7]). Similarly, HPV16 (80·8% [95% CI 75·9–85·2]) and HPV33 (6·3% [3·9–9·2]) Medical University, Beijing
Maternal and Child Health Care
were also the most two predominant HPV genotypes in vulvar intraepithelial neoplasia. The distribution of type- Hospital, Beijing, China
specific HPV genotypes in vulvar cancer among geographical regions was different, with HPV16 varying between (P Liu MD, Prof J Miao PhD)
regions, showing a high prevalence in Oceania (89·0% [95% CI 67·6–99·5]) and a low prevalence in South America Correspondence to:
(54·3% [30·2–77·4]). The prevalence of p16INK4a positivity in patients with vulvar cancer was 34·1% (95% CI 30·9–37·4; Dr Kun Song, Department of
52 studies; n=6352), and it was 65·7% (52·5–77·7; 23 studies; n=896) in patients with vulvar intraepithelial neoplasia. Obstetrics and Gynaecology, Qilu
Hospital of Shandong University,
Furthermore, among patients with HPV-positive vulvar cancer, p16INK4a positivity prevalence was 73·3% (95% CI Jinan, Shandong 250012, China
64·7–81·2), compared with 13·8% (10·0–18·1) in HPV-negative vulvar cancer. The prevalence of double positivity for songkun2001226@sdu.edu.cn
HPV and p16INK4a was 19·6% (95% CI 16·3–23·0) in vulvar cancer and 44·2% (26·3–62·8) in vulvar intraepithelial
neoplasia. Most analyses had large heterogeneity (I²>75%).

Interpretation The high prevalence of HPV16 and HPV33 in vulvar cancer and vulvar intraepithelial neoplasia
emphasised the importance of nine-valent HPV vaccination in preventing vulvar neoplasm. Additionally, this study
highlighted the potential clinical significance of double positivity for HPV DNA and p16INK4a in vulvar neoplasm.

Funding Taishan Scholar Youth Project of Shandong Province, China.

Copyright © 2023 Elsevier Ltd. All rights reserved.

Introduction female genital tumours.4 Vulvar cancer can be preceded

Vulvar cancer, as a serious gynaecological malignancy, by vulvar intraepithelial neoplasia, which is classified as
accounts for about 5% of all female genital tract cancers.1 low-grade squamous intraepithelial lesions, high-grade
Vulvar squamous cell carcinoma is the most predominant squamous intraepithelial lesions, and differentiated
histological subtype of vulvar cancer and accounts for vulvar intraepithelial neoplasia.5–7 Infection with high-risk
more than 90% of all cases, including keratinising, non- human papillomavirus (HPV) is a well established cause
keratinising, basaloid, warty, and verrucous carcinoma,2,3 of vulvar cancer and vulvar intraepithelial neoplasia.8
which can further be classified as HPV-associated or Additionally, low-risk HPV genotypes, such as HPV6 and
HPV-independent vulvar squamous cell carcinoma HPV11, have been identified as the potential cause of
according to the 2020 WHO classification criteria for vulvar cancer and vulvar intraepithelial neoplasia.4

www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9 1

 Articles

Research in context
Evidence before this study in both vulvar cancer and vulvar intraepithelial neoplasia.
We searched the PubMed, Embase, and Cochrane Library Additionally, this was the first study, to our knowledge,
databases to identify studies published in English between to investigate the pooled prevalence of specific HPV genotypes
Jan 1, 1986, and May 6, 2022, using the terms “vulvar cancer” according to different geographical regions and report a
OR “vulvar intraepithelial neoplasia” AND “p16” OR “HPV”. Four disparity in HPV prevalence among regions. To our best
meta-analyses estimated the pooled prevalence of HPV DNA in knowledge, this systematic review and meta-analysis was also
vulvar cancer and vulvar intraepithelial neoplasia. However, the first to estimate the prevalence of p16INK4a positivity in
these meta-analyses only included studies published up to vulvar cancer and vulvar intra­epithelial neoplasia, with a
2015, after which the 2020 edition of the WHO classification prevalence of p16INK4a positivity of 34·1% (95% CI 30·9–37·4) in
criteria for female genital tumours was published and vulvar cancer and of 65·7% (52·5–77·7) in vulvar intraepithelial
numerous relevant studies have been published. Furthermore, neoplasia. Furthermore, we also reported the prevalence of
to the best of our knowledge, no previous systematic reviews double positivity for HPV DNA and p16INK4a in vulvar cancer
and meta-analyses have estimated the prevalence of p16INK4a (19·6% [95% CI 16·3–23·0]) and vulvar intraepithelial neoplasia
positivity in vulvar cancer or vulvar intraepithelial neoplasia, (44·2% [26·3–62·8]).
and no meta-analyses have examined the prevalence of double
Implications of all the available evidence
positivity for HPV DNA and p16INK4a in vulvar intraepithelial
This meta-analysis reported a high prevalence of HPV DNA in
neoplasia or vulvar cancer.
vulvar cancer (about 40%) and vulvar intraepithelial neoplasia
Added value of this study (about 76%), which was consistent with the findings of four
In this systematic review and meta-analysis, we added more previous meta-analyses. Owing to the high prevalence of
than 3000 new cases and estimated an HPV DNA prevalence of HPV16 and HPV33 in vulvar cancer and vulvar intraepithelial
39·1% (95% CI 35·3–42·9) in vulvar cancer and 76·1% neoplasia, multi-valent HPV vaccination covering HPV16 and
(70·7–81·1) in vulvar intraepithelial neoplasia, contributing to a HPV33 seems to have a greater clinical potential in preventing
more robust examination compared with previous meta- vulvar neoplasm, especially in Africa, Oceania, and Europe.
analyses. Among the pooled prevalence of specific HPV Additionally, this study highlighted the potential clinical
genotypes (HPV types 6, 11, 16, 18, 31, 33, 45, 52, and 58), importance of double positivity for HPV DNA and p16INK4a in
HPV16 and HPV33 were the two predominant HPV genotypes vulvar neoplasm.

The presence of HPV DNA in tumour tissue can be imported into the reference manager software Endnote,
detected by various methods, including PCR or Hybrid and duplicates were removed. Full details of the search
Capture 2. Integration of HPV DNA into the host strategies adopted in each database are shown in the
See Online for appendix genome could cause the overexpression of the tumour appendix (p 1).
suppressor protein p16INK4a, which can be detected by We included studies that reported the prevalence
immuno­histochemical staining and has been proposed of p16INK4a positivity, HPV DNA, or both, for a minimum
as a surrogate marker of persistent HPV infections.9 of five cases of histologically verified vulvar cancer or
The combined detection of HPV DNA and p16INK4a has vulvar intraepithelial neoplasia. Only studies that applied
been proposed as an increasingly reliable method for PCR or the Hybrid Capture 2 method to detect HPV DNA
diagnosing true HPV-driven tumours in oropharyngeal and methy­ lation or immunohistochemistry to detect
squamous cell carcinomas.10 p16INK4a were eligible for inclusion. Additionally, references
Since publication of previous meta-analyses,11–14 cited in the included publi­cations were investigated to
numerous relevant studies have been published. identify additional relevant articles. Only peer-reviewed
Therefore, we aimed to provide updated data on the studies were included; we excluded conference abstracts
pooled prevalence of HPV DNA in vulvar cancer and and case reports. Study populations that were included in
vulvar intraepithelial neoplasia, and to examine the more than one article were enrolled only once, and we
prevalence of p16INK4a positivity, both overall and according used the data from the article that contained the most
to different HPV statuses. Additionally, we estimated the detailed information. We contacted primary study authors
prevalence of double positivity for HPV DNA and p16INK4a by email for clarifications if needed. The study protocol is
in vulvar cancer and vulvar intraepithelial neoplasia. shown in the appendix (p 2).
We did this study in accordance with the PRISMA
Methods guidelines.15
Search strategy and inclusion criteria
In this systematic review and meta-analysis, we searched Data analysis
the PubMed, Embase, and Cochrane Library databases to Data were extracted independently by two authors (ZL
identify all English-language studies published between and PL), with inconsistencies discussed by two authors
Jan 1, 1986, and May 6, 2022. All records identified were or a third author (ZZ) until a consensus was reached.

2 www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9

 Articles

We extracted data on the following variables: first author, primer PCR, and Hybrid Capture 2), tissue type (frozen,
year of publication and sample collection, sample size, fixed, unknown, and exfoliated cells) and publication
geographical region, age at diagnosis, tissue type, type year (1989–99, 2000–09, and 2010–22), were included
of lesion (vulvar intraepithelial neoplasia or vulvar
cancer), histological subtype of vulvar cancer and vulvar
intraepithelial neoplasia, the method and PCR primers 6393 records identified
used for HPV DNA detection, HPV genotypes, and 2230 PubMed
4000 Embase
overall and type-specific HPV prevalence. Furthermore, 163 Cochrane Library
data on the p16INK4a testing method, definition of p16INK4a
positivity, prevalence of p16INK4a positivity (among the
1990 duplicates excluded
whole population, the HPV-negative population, and the
HPV-positive population), prevalence of double positivity
for HPV DNA and p16INK4a were also extracted. In a 4403 screened by title
small observational study,16 which applied both immuno­
histochemistry and methylation methods to test p16INK4a,
3652 non-eligible records excluded
only the evaluation by immunohistochemistry was
included in the analyses.
After data extraction, ZC reviewed the data and relevant 751 abstracts reviewed
full-text papers to correctly classify the method and PCR
primers used for HPV DNA detection. The histological
418 excluded
subtypes of vulvar cancer and vulvar intra­ epithelial 191 had no primary data, or data were not
neoplasia were reclassified on the basis of the WHO peer-reviewed
Classification for Female Genital Tumors4 and the 23 included no patients with vulvar cancer
or vulvar intraepithelial neoplasia
International Society for the Study of Vulvovaginal Disease 143 had no testing for HPV or p16INK4a, or
Terminology of Vulvar Squamous Intraepithelial Lesions.5 HPV test was not PCR or Hybrid Capture 2
7 had no vulvar samples
We did not assess the quality of included studies. 25 fewer than five cases were tested
We used random-effect models to calculate the pooled 8 were not in English
21 included populations with a predisposed
prevalence of HPV DNA and p16INK4a positivity in both risk of HPV-positivity
vulvar intraepithelial neoplasia and vulvar cancer. We
used the DerSimonian–Laird estimator17 to determine
the variance across studies, and applied the arcsine 333 full-text records assessed for eligibility
transformation to stabilise the variance.18 The pooled
prevalence of HPV DNA and p16INK4a positivity in each 173 excluded
study were shown with 95% CIs. Statistical heterogeneity 18 had no primary data, or data were not
between studies was investigated using Higgins’ peer-reviewed
6 included no patients with vulvar cancer or
I² statistics, and the significance of the heterogeneity was vulvar intraepithelial neoplasia
evaluated using Cochran’s Q test.19 Additionally, potential 30 had no testing for HPV or p16INK4a, or HPV
test was not PCR or Hybrid Capture 2
sources of heterogeneity were explored through stratified 2 had no vulvar samples
analyses and meta-regression. In the meta-regression, 9 fewer than five cases were tested
each variable was estimated in a univariable test of a 61 contained data reported in more than
one study
moderator variable, which was based on the χ² test. 7 did not report data separately for benign,
For studies estimating the HPV DNA prevalence premalignant, or malignant conditions
40 included populations with a predisposed
in vulvar intraepithelial neoplasia and vulvar cancer, risk of HPV-positivity
we did subgroup analyses stratified by different histo­
logical subtypes of vulvar cancer (keratinising, non-
2 studies identified through a review of reference
keratinising, warty, basaloid, and verrucous carcinoma), lists
vulvar intra­ epithelial neoplasia (low-grade and high
grade squamous intraepithelial lesions and differ­
entiated vulvar intra­ epithelial neoplasia) and age at 162 studies included in the systematic review
diagnosis (<60 and ≥60 years). The defined variables,
including geographical region (Europe, North America, 162 studies included in the meta-analysis*
South America, Asia, Africa, Oceania, and multicountry 91 on HPV and vulvar cancer
[ie, a population from more than one of these 60 on HPV and vulvar intraepithelial neoplasia
52 on p16INK4a and vulvar cancer
regions]),the method used for HPV DNA detection (ie, 23 on p16INK4a and vulvar intraepithelial neoplasia
consensus primer PCR, type-specific primer PCR,
combination of consensus primers PCR, combination Figure 1: Study selection
of consensus and type-specific primers PCR, unknown HPV=human papillomavirus. *Some studies included data for more than one of the categories listed.

www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9 3

 Articles

in the univariate meta-regression. Furthermore, we method (immunohisto­chemistry and methy­lation) and

explored the pooled type-specific prevalence of HPV publication year (1998–2005, 2006–14, and 2015–22),
genotypes covered by the nine-valent HPV vaccine were selected for meta-regression analysis. Additionally,
(HPV types 6, 11, 16, 18, 31, 33, 45, 52, and 58),20 among the pooled p16INK4a positivity prevalence was calcu­lated
patients with HPV-positive vulvar intraepithelial among all patients, those who were HPV-negative,
neoplasia or vulvar cancer, stratifying the analyses to and those who were HPV-positive. We also calculated
either single or all HPV infections (ie, single and the pooled prevalence of p16INK4a positivity according
multiple infections). to the different cutoffs of p16INK4a positivity used in
For studies estimating the prevalence of p16INK4a immunohistochemistry.
positiv­
ity in vulvar cancer and vulvar intraepithelial To address the risk of bias between studies and to
neoplasia, we did subgroup analyses according to assess the robustness of pooled estimate, sensitivity
histological subtype of vulvar cancer and vulvar intra­ analyses were done by excluding studies that included
epithelial neoplasia. Variables, including geographical populations with an increased risk of HPV infection
region (Europe, North America, South America, Asia, (ie, women with HIV). All analyses were done using
Africa, Oceania, and multi­ country), p16INK4a detection the statistical software R (version 4.2.0) with the meta

HPV DNA prevalence p16INK4a positivity prevalence

Studies* Total Patients Pooled I² for p value for Studies* Total Patients Pooled I² for p value for
patients positive for prevalence heterogeneity heterogeneity patients positive prevalence heterogeneity heterogeneity
HPV DNA of HPV for p16INK4a of p16INK4a
DNA positivity
positivity (95% CI)
(95% CI)

Vulvar cancer
All 91 8200 2958 39·1% 90·9% <0·0001 52 6352 1957 34·1% 83·4% <0·0001
(35·3–42·9) (30·9–37·4)
Vulvar 79 5911 2174 37·3% 90·2% <0·0001 50 4538 1426 34·8% 83·8% <0·0001
squamous cell (33·2–41·5) (31·2–38·6)
carcinoma
Keratinising 16 811 142 12·0% 88·1% <0·0001 8 361 45 11·2% 85·8% <0·0001
carcinoma (6·1–19·6) (3·8–21·9)
Non- 7 55 25 31·1% 80·5% <0·0001 4 20 10 50·0% 0·0% 0·7648
keratinising (7·4–62·0) (28·8–71·3)
carcinoma
Basaloid 17 133 109 83·4% 32·2% 0·0984 7 66 54 95·9% 86·0% <0·0001
carcinoma (74·3–90·8) (73·0–100·0)
Warty 13 141 115 86·5% 32·0% 0·1265 7 35 17 85·6% 89·9% <0·0001
carcinoma (76·9–93·8) (31·2–100·0)
Verrucous 7 16 8 58·9% 71·1% 0·0020 3 11 0 0·0% 0·0% 1·0000
carcinoma (14·0–95·9) (0·0–8·5)
Vulvar intraepithelial neoplasia
All 60 3140 2466 76·1% 90·5% <0·0001 23 896 553 65·7% 93·7% <0·0001
(70·7–81·1) (52·5–77·7)
Low-grade 14 147 104 61·1% 85·9% <0·0001 9 110 19 26·6% 80·8% <0·0001
squamous (37·8–82·0) (8·9–49·7)
intraepithelial
lesion†
High-grade 48 1747 1437 83·2% 84·3% <0·0001 19 468 435 97·5% 78·8% <0·0001
squamous (78·3–87·6) (93·2–99·7)
intraepithelial
lesion‡
Differentiated- 12 149 7 2·5% 55·8% 0·0095 10 165 0 0·0% 0·0% 1·0000
type vulvar (0·0–9·2) (0·0–0·6)
intraepithelial
neoplasia
Date are n, % (95% CI), or %. HPV=human papillomavirus.*Studies could contribute cases to more than one histological subtype of vulvar squamous cell carcinoma and vulvar intraepithelial neoplasia.
Furthermore, some studies only contributed to the pooled analysis of vulvar cancer or vulvar intraepithelial neoplasia, as that specific histological subtype was not available. †Includes vulvar intraepithelial
neoplasia grade 1. ‡Includes vulvar intraepithelial neoplasia grade 2 and 3, and usual-type vulvar intraepithelial neoplasia.

Table 1: Pooled prevalence of HPV DNA and p16INK4a positivity in vulvar cancer and vulvar intraepithelial neoplasia stratified by histological subtypes

4 www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9

 Vulvar cancer Vulvar intraepithelial neoplasia

Studies Total Patients Pooled prevalence p value I² for p value for Studies Total Patients Pooled prevalence p value I² for p value for
patients positive of HPV DNA (univariable heterogeneity heterogeneity patients positive of HPV DNA (univariable heterogeneity heterogeneity
for HPV positivity (95% CI) test of for HPV positivity (95% CI) test of
DNA moderator DNA moderator
variable) variable)
Geographical <0·0001 0·6800
location
Europe 47 4171 1401 30·8% (26·1–35·7) ·· 90·6% <0·0001 34 1405 1067 76·3% (67·6–83·9) ·· 92·1% <0·0001
North America 19 911 492 52·9% (44·7–61·1) ·· 82·5% <0·0001 10 543 408 68·3% (51·3–83·1) ·· 92·3% <0·0001
South America 2 266 82 32·1% (17·1–49·3) ·· 87·8% 0·0041 2 41 22 57·4% (32·0–80·9) ·· 57·2% 0·1265
Asia 16 756 353 48·4% (39·2–57·8) ·· 79·3% <0·0001 9 391 320 82·3% (72·7–90·2) ·· 66·2% 0·0026
Africa 2 104 65 66·0% (45·5–83·8) ·· 74·3% 0·0485 0 ·· ·· ·· ·· ·· ··
Oceania 4 283 77 44·4% (20·6–69·7) ·· 93·5% <0·0001 3 111 88 77·3% (50·6–95·5) ·· 88·7% 0·0001
Multicountry* 1 1709 488 28·6% (26·4–30·8) ·· ·· ·· 2 649 561 86·5% (83·7–89·0) ·· 0·0% 0·5476

HPV test and 0·1148 0·4698

primers
PCR, consensus 50 5516 1824 35·6% (30·9–40·5) ·· 91·5% <0·0001 32 1991 1537 74·3% (66·0–81·9) ·· 92·7% <0·0001
primers
PCR, type- 11 613 228 40·6% (30·2–51·4) ·· 83·9% <0·0001 8 207 169 81·2% (67·4–91·9) ·· 77·8% <0·0001
specific primers
PCR, 9 1040 526 52·9% (40·9–64·7) ·· 91·4% <0·0001 5 200 182 89·1% (78·6–96·3) ·· 71·9% 0·0066
combination of
consensus
primers
PCR, 17 882 304 38·7% (29·6–48·3) ·· 87·2% <0·0001 9 442 352 73·2% (59·7–84·7) ·· 87·8% <0·0001
combination of

www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9

consensus and
type-specific
primers
PCR, unknown 3 114 61 53·5% (44·4–62·6) ·· 0·0% 0·6554 2 174 149 85·7% (80·1–90·5) ·· 0·0% 0·6257
primers
Hybrid Capture 2 1 35 15 42·9% (26·3–60·6) ·· ·· ·· 4 126 77 63·6% (43·8–81·2) ·· 80·1% 0·0018
Tissue sample type 0·2032 0·2840
Frozen 14 507 249 46·6% (34·9–58·6) ·· 85·1% <0·0001 10 226 181 84·7% (66·1–96·7) ·· 90·3% <0·0001
Fixed 73 7527 2665 38·1% (34·1–42·3) ·· 91·5% <0·0001 39 2364 1841 72·7% (65·7–79·2) ·· 91·6% <0·0001
Unknown 4 166 44 30·8% (16·6–47·3) ·· 76·7% 0·0050 9 456 379 82·1% (71·4–90·8) ·· 84·3% <0·0001
Exfoliated cells 0 ·· ·· ·· ·· ·· ·· 2 94 65 69·2% (59·5–78·0) ·· 0·0% 0·9521
Publication year 0·9740 0·0182
1989–99 20 876 303 37·8% (30·8–45·0) ·· 77·2% <0·0001 12 557 423 80·8% (69·4–90·1) ·· 86·8% <0·0001
2000–09 22 823 367 38·5% (26·8–50·8) ·· 91·7% <0·0001 26 770 509 66·3% (53·8–77·8) ·· 91·8% <0·0001
2010–22 49 6501 2288 39·2% (34·6–43·9) ·· 92·4% <0·0001 22 1813 1534 83·1% (77·9–87·7) ·· 84·5% <0·0001
Date are n, % (95% CI), or %. If only one study contributed to the analysis, then no pooled prevalence was estimated and the actual proportion reported in the study was provided. HPV=human papillomavirus. *Multicountry indicates a population from
Articles

two or more of the listed regions.

Table 2: Pooled prevalence of HPV DNA in vulvar cancer and vulvar intraepithelial neoplasia, by stratified variable

5
 Articles

Pooled prevelance Pooled prevelance Pooled prevelance

(95% CI) (95% CI) (95% CI)
A Global B Africa* C Asia

HPV6 0·6% (0·2–1·3) ·· 2·5% (0·0–9·6)

0·4% (0·1–0·9) ·· 1·8% (0·0–10·3)

HPV11 0·1% (0·0–0·5) ·· 0·3% (0·0–3·1)

0·0% (0·0–0·1) ·· 0·0% (0·0–1·8)

HPV16 78·1% (73·5–82·3) 72·4% (51·4–89·3) 66·4% (54·4–77·4)

78·6% (73·2–83·5) 60·9% (40·6–79·4) 64·7% (51·5–76·9)

HPV18 4·4% (2·5–6·8) 0·9% (0·0–8·6) 10·3% (2·6–22·2)

3·0% (1·4–5·2) 4·4% (0·0–16·2) 9·8% (2·2–22·0)
HPV genotype

HPV31 0·7% (0·2–1·5) 3·1% (0·0–26·4) 1·8% (0·0–9·7)

0·5% (0·1–1·3) 13·0% (2·7–29·5) 1·1% (0·0–10·3)

HPV33 7·5% (4·9–10·7) 9·1% (3·4–17·2) 3·4% (0·0–18·4)

5·2% (2·8–8·2) 13·0% (2·7–29·5) 1·8% (0·0–17·8)

HPV45 0·4% (0·1–0·8) 2·0% (0·0–17·7) 0·5% (0·0–3·8)

0·3% (0·1–0·9) 8·7% (0·9–23·3) 0·0% (0·0–2·2)

HPV52 0·5% (0·1–1·1) 1·5% (0·0–10·9) 0·3% (0·0–2·5)

0·2% (0·0–0·6) 0·0% (0·0–4·1) 0·5% (0·0–4·0)

HPV58 0·1% (0·0–0·4) 0·0% (0·0–1·5) 1·5% (0·1–4·9)

0·1% (0·0–0·4) 0·0% (0·0–4·1) 1·0% (0·0–4·2)

D Europe E North America F Oceania

HPV6 0·3% (0·0–0·9) 1·7% (0·1–5·1) 0·0% (0·0–5·9)

0·2% (0·0–0·9) 0·9% (0·0–3·0) 0·0% (0·0–6·3)

HPV11 0·2% (0·0–0·9) 0·0% (0·0–0·5) 0·0% (0·0–5·9)

0·0% (0·0–0·2) 0·0% (0·0–0·6) 0·0% (0·0–6·3)

HPV16 81·3% (74·6–87·1) 79·8% (72·9–86·0) 89·0% (67·6–99·5)

83·6% (76·8–89·4) 83·1% (75·2–89·8) 79·2% (66·0–89·9)

HPV18 2·6% (0·9–5·2) 5·2% (1·7–10·6) 2·1% (0·0–15·5)

1·2% (0·2–3·1) 3·5% (0·8–8·0) 1·3% (0·0–8·6)
HPV genotype

HPV31 0·6% (0·1–1·6) 0·1% (0·0–1·2) 0·0% (0·0–5·9)

0·4% (0·0–1·2) 0·2% (0·0–1·4) 0·0% (0·0–6·3)

HPV33 8·3% (4·9–12·3) 7·1% (2·6–13·5) 12·5% (1·4–32·5)

6·3% (3·2–10·4) 2·3% (0·0–7·8) 13·3% (1·5–34·4)

HPV45 0·0% (0·0–0·3) 1·1% (0·0–4·7) 0·0% (0·0–5·9)

0·0% (0·0–0·4) 0·9% (0·0–3·5) 0·0% (0·0–6·3)

HPV52 0·2% (0·0–0·7) 1·2% (0·1–3·7) 6·3% (0·0–22·8)

0·1% (0·0–0·5) 0·6% (0·0–2·5) 6·7% (0·0–24·2)

HPV58 0·2% (0·0–0·7) 0·0% (0·0–0·6) 0·0% (0·0–5·9)

0·1% (0·0–0·5) 0·0% (0·0–0·7) 0·0% (0·0–6·3)
0 20 40 60 80 100 0 20 40 60 80 100
G South America Prevalence (%) Prevalence (%)

Among patients with vulvar cancer and HPV infection

HPV6 0·0% (0·0–2·2)
0·0% (0·0–2·7) Among patients with vulvar cancer and a single HPV genotype infection

HPV11 0·0% (0·0–2·2)

0·0% (0·0–2·7)

HPV16 54·3% (30·2–77·4)

50·3% (25·2–75·4)

HPV18 22·0% (1·9–55·5)

17·2% (0·4–50·9)
HPV genotype

HPV31 0·6% (0·0–5·1)

0·0% (0·0–1·4)

HPV33 7·6% (0·0–53·9)

3·7% (0·0–29·5)

HPV45 2·4% (0·2–6·9)

2·9% (0·3–8·2)

HPV52 0·0% (0·0–1·2)

0·0% (0·0–1·4)

HPV58 0·0% (0·0–1·2)

0·0% (0·0–1·4)
0 20 40 60 80 100
Prevalence (%)

Figure 2: Pooled prevalence of specific HPV genotypes in vulvar cancer, by geographical region
HPV=human papillomavirus. *Data on the prevalence of HPV6 and HPV11 were not available in the studies from Africa.

6 www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9

 Articles

package,21 and a two-sided p value of less than 0·05 was

Date are n, % (95% CI), or %. If only one study contributed to the analysis, then no pooled prevalence was estimated and the actual proportion reported in the study was provided. *Multicountry indicates a population from two or more of the listed regions.
heterogeneity
considered to be statistically significant.

p value for

<0·0001
<0·0001
<0·0001

<0·0001
<0·0001
0·0303

0·0638
0·0036
Role of the funding source

··

··
··
··
The funder of the study had no role in the study design,

heterogeneity
data collection, data analysis, data interpretation, or
writing of the report.

96·8%

88·2%

94·2%

95·4%
91·6%

49·7%
95·7%
78·7%
I² for

··
··
··

··
Results

moderator
A total of 6393 records were identified in the PubMed,

variable)
variable

0·4988

0·9048
0·4582
Pooled prevalence of p value

test of
Embase, and Cochrane Library database search. After the

(uni-

··
··

··

··
··
··
··

··
··
··

··
··
removal of 1990 duplicates, 4403 records were assessed by
title for eligibility and 3652 were excluded (figure 1). We

92·3% (43·7–100·0)

67·8% (44·6–87·1)
60·3% (27·8–88·3)
68·7% (60·0–76·7)
66·8% (51·7–80·3)
61·1% (24·9–91·4)

79·1% (67·4–88·1)

49·7% (17·7–81·8)
50·0% (27·2–72·8)

67·2% (53·2–79·7)
21·4% (8·3–41·0)
reviewed 751 abstracts and excluded 418. We assessed

p16INK4a positivity
333 full-text records and excluded 173, leaving 160 relevant
studies were eligible for inclusion. Two additional relevant
(95% CI)
studies were retrieved after reviewing the reference lists

··
of the included studies. Thus, 162 studies, published
Vulvar intraepithelial neoplasia

Patients
positive

between March, 1989, and May, 2022, were included in

p16INK4a

266
528
338

104
183
10
53

25
122
6
24
this meta-analysis. Details and references of all included
for

··
studies are shown in the appendix (pp 3–18; 20–29).
Patients

The prevalence of HPV DNA in vulvar cancer was

408
534
220

851

261
20
67

45
28
27

227
examined in 91 studies (n=8200). The prevalence of HPV

··
DNA ranged from 0% to 88·9%, with an overall prevalence
Studies

of 39·1% (95% CI 35·3–42·9) and a large heterogeneity

1
1

21

2
12
6
1

7
2
0

5
11
(I²=90·9%; p<0·0001; table 1; appendix p 31). For the

Table 3: Pooled prevalence of p16INK4a positivity in vulvar cancer and vulvar intraepithelial neoplasia, by stratified variable
sensitivity analysis, we excluded two studies22,23 that
heterogeneity
p value for

included populations with an increased risk of HPV

<0·0001
<0·0001

<0·0001
<0·0001
<0·0001
0·0478

0·0047
0·0030
0·7003
0·2849

infection (ie, women with HIV) and the pooled prevalence

··
··

of HPV DNA (38·4% [95% CI 34·7–42·3]) changed very

little, with an I² of 90·8% (p<0·0001). In the stratified
heterogeneity

analyses by the histological subtype of vulvar squamous

cell carcinoma, HPV DNA prevalence was highest in

80·0%
62·4%

83·4%

85·4%
79·5%
82·1%

12·6%

81·3%
67·1%
0·0%
I² for

··
··

warty carcinoma (86·5% [95% CI 76·9–93·8]) and lowest

in keratinising carcinoma (12·0% [6·1–19·6]; table 1).
moderator
variable)

When stratified by geographical region, the univariable

variable

0·0002

0·2016
0·2935
p value

test of
(uni-

meta-regression showed a significant difference in the

··

··

··
··

··
··
··
··

··
··
··

··
pooled prevalence of HPV DNA among vulvar cancer
36·2% (28·4–44·4)
34·5% (26·8–42·6)
52·9% (43·6–62·2)
73·3% (44·9–92·2)
43·5% (29·7–58·0)

41·9% (30·8–53·5)
27·8% (21·6–34·4)
43·9% (37·8–50·1)
positivity (95% CI)

29·5% (25·6–33·5)

41·7% (21·2–63·8)

32·1% (28·3–35·9)
33·7% (30·4–37·0)
Pooled prevalence

between geographical regions, which ranged from 30·8%

(95% CI 26·1–35·7) in Europe to 66·0% (45·5–83·8) in
of p16INK4a

Africa (p<0·0001; table 2). When stratified by age, the

pooled prevalence of HPV DNA in patients with vulvar
cancer who were younger than 60 years was 66·3%
Patients
positive

(95% CI 56·0–78·9) and was 30·9% (22·0–40·7) in those

p16INK4a

557

1913

44
935

165
316
52
23

265
11
63

1527

aged 60 years or older (appendix p 19).

for

The prevalence of HPV DNA in patients with

Patients

vulvar intraepithelial neoplasia was assessed in

6248
3404

1821

104

459
753
187
53

801
5092
15
119
Vulvar cancer

60 studies (3140 patients). The pooled prevalence of

HPV DNA in vulvar intraepithelial neoplasia was
Studies

76·1% (95% CI 70·7–81·1; appendix p 32), with a higher

49
3

3
31

7
11
3
2

15
1
1

30

prevalence in vulvar high-grade squamous intra­

p16INK4a methylation

epithelial lesions (83·2% [78·3–87·6]) and low-grade

Geographical region

p16 immuno-
South America
North America

histochemistry
Multicountry*

squamous intra­epithelial lesions (61·1% [37·8–82·0])

Publication year
1998–2005

than in differentiated vulvar intraepithelial neoplasia

p16INK4a test

2006–14
2015–22
Oceania
Europe

INK4a

(2·5% [0·0–9·2]; table 1). When stratified by age,

Africa
Asia

the pooled prevalence of HPV DNA in patients with

vulvar intraepithelial neoplasia who were younger

www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9 7

 Articles

A Patients who are HPV-positive

Events (n/N) Prevalence (95% CI)

Gasco et al (2002) 24
11/13 84·6% (54·6–98·1)
Osakabe et al (2007)25 4/5 80·0% (28·4–99·5)
Aulmann et al (2008)26 5/5 100·0% (47·8–100·0)
de Koning et al (2008)27 11/13 84·6% (54·6–98·1)
Alonso et al (2011)28 19/19 100·0% (82·4–100·0)
Lavorato-Rocha et al (2013)29 11/33 33·3% (18·0–51·8)
Reuschenbach et al (2013)30 39/74 52·7% (40·7–64·4)
Lee et al (2016)31 14/15 93·3% (68·1–99·8)
Sznurkowski et al (2016)32 25/38 65·8% (48·6–80·4)
Wakeham et al (2017)33 23/32 71·9% (53·3–86·3)
Pils et al (2017)34 32/40 80·0% (64·4–90·9)
Halec et al (2017)35 395/476 83·0% (79·3–86·3)
Hinten et al (2018)36 53/73 72·6% (60·9–82·4)
Arians et al (2019)37 13/16 81·2% (54·4–96·0)
Stefansson et al (2019)38 22/36 61·1% (43·5–76·9)
Thomas et al (2020)39 6/10 60·0% (26·2–87·8)
Lérias et al (2020)40 9/26 34·6% (17·2–55·7)
Allo et al (2020)41 30/33 90·9% (75·7–98·1)
Preti et al (2020)42 7/17 41·2% (18·4–67·1)
Gensthaler et al (2020)43 20/30 66·7% (47·2–82·7)
Kolitz et al (2022)44 15/22 68·2% (45·1–86·1)
Pooled prevalence 764/1026 73·3% (64·7–81·2)
Heterogeneity: I2=84·9%, p<0·0001

B Patients who are HPV-negative

Gasco et al (2002)24 12/23 52·2% (30·6–73·2)

van der Avoort et al (2006)45 2/16 12·5% (1·6–38·3)
Osakabe et al (2007)25 7/16 43·8% (19·8–70·1)
Aulmann et al (2008)26 15/23 65·2% (42·7–83·6)
de Koning et al (2008)27 1/26 3·8% (0·1–19·6)
Alonso et al (2011)28 1/79 1·3% (0·0–6·9)
Lavorato-Rocha et al (2013)29 18/68 26·5% (16·5–38·6)
Reuschenbach et al (2013)30 18/101 17·8% (10·9–26·7)
Lee et al (2016)31 6/41 14·6% (5·6–29·2)
Sznurkowski et al (2016)32 10/47 21·3% (10·7–35·7)
Wakeham et al (2017)33 2/27 7·4% (0·9–24·3)
Pils et al (2017)34 6/136 4·4% (1·6–9·4)
Halec et al (2017)35 103/1194 8·6% (7·1–10·4)
Hinten et al (2018)36 32/245 13·1% (9·1–17·9)
Arians et al (2019)37 6/59 10·2% (3·8–20·8)
Stefansson et al (2019)38 4/61 6·6% (1·8–15·9)
Thomas et al (2020)39 4/45 8·9% (2·5–21·2)
Lérias et al (2020)40 3/58 5·2% (1·1–14·4)
Allo et al (2020)41 12/77 15·6% (8·3–25·6)
Preti et al (2020)42 4/85 4·7% (1·3–11·6)
Gensthaler et al (2020)43 12/105 11·4% (6·0–19·1)
Kolitz et al (2022)44 5/14 35·7% (12·8–64·9)
Pooled prevalence 283/2546 13·8% (10·0–18·1)
Heterogeneity: I2=84·2%, p<0·0001

0 20 40 60 80 100
Prevalence (%)

Figure 3: Pooled prevalence of p16INK4a positivity in vulvar cancer globally, by HPV status
Grey squares represent the point estimate of the prevalence of p16INK4a positivity and horizontal lines represent 95% CIs. Details and references of all included studies
are shown in the appendix (pp 3–18; 20–29). HPV=human papillomavirus.

8 www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9

 Articles

than 60 years was 85·5% (95% CI 79·1–90·9) and positivity in vulvar intraepithelial neoplasia was 65·7%
46·8% (22·3–72·2) in those aged 60 years or older (95% CI 52·5–77·7; appendix p 35), with a higher
(appendix p 19). prevalence in high-grade squamous intraepithelial
The pooled prevalence of specific HPV genotypes (HPV lesions (97·5% [93·2–99·7]) and low-grade squamous
types 6, 11, 16, 18, 31, 33, 45, 52, and 58), globally and by intraepithelial lesions (26·6% [8·9–49·7]) than in
geographical region, was examined among patients with differ­
entiated vulvar intraepithelial neoplasia (0%
HPV-positive vulvar cancer (figure 2). Among all patients [0·0–0·6]; table 1). When estimating the pooled
with HPV DNA-positive vulvar cancer, HPV16 was the prevalence of p16INK4a positivity on the basis of the HPV
most frequent genotype, with a pooled prevalence of status of vulvar intraepithelial neoplasia, HPV-positive
78·1% (95% CI 73·5–82·3), followed by HPV33 (7·5% vulvar intraepithelial neoplasia had a higher prevalence
[4·9–10·7]), and HPV18 (4·4% [2·5–6·8]). When stratified of p16INK4a positivity (92·7% [95% CI 80·2–99·3]) than
by geographical regions, HPV16 was also the most HPV-negative vulvar intra­epithelial neoplasia (42·6%
predominant genotype in all regions, varying substantially [20·6–66·3]; appendix p 36).
between regions, with a high prevalence in Oceania 22 studies (3572 patients) estimating the prevalence of
(89·0% [95% CI 67·6–99·5]) and a low prevalence in double positivity for HPV DNA and p16INK4a in vulvar
South America (54·3% [30·2–77·4]). However, despite cancer were included. The overall prevalence of double
HPV33 being the second most predominant type in positivity for HPV DNA and p16INK4a in vulvar cancer
vulvar cancer worldwide, HPV18 was the second most ranged from 0% to 41·7%, yielding an overall pooled
frequent type in South America and Asia. A reasonably prevalence of 19·6% (95% CI 16·3–23·0; appendix p 37).
high pooled prevalence of HPV52 was observed Additionally, eight studies (381 patients) estimated the
in Oceania. prevalence of double positivity for HPV DNA and p16INK4a
Among all patients with HPV DNA-positive vulvar in vulvar intraepithelial neoplasia. The overall prevalence
intraepithelial neoplasia, HPV16 was the most common of double positivity for HPV DNA and p16INK4a in vulvar
genotype, with a pooled prevalence of 80·8% (95% CI intraepithelial neoplasia was 44·2% (95% CI 26·3–62·8;
75·9–85·2; appendix p 33). Additionally, the prevalence appendix p 37).
of HPV33 was 6·3% (95% CI 3·9–9·2), and HPV6
accounted for 2·9% (95% CI 0·8–6·3). When stratified Discussion
by geographical region, HPV16 was also the most In this systematic review and meta-analysis, we estimated
predominant genotype in all regions. Additionally, a pooled HPV DNA prevalence of approximately 40% in
HPV33 was the second most prevalent genotype in vulvar cancer and 76% in vulvar intraepithelial neoplasia.
Europe and North America, whereas HPV18 was the This meta-analysis included numerous new studies and
second most common genotype in Asia (appendix p 33). cases, contributing to increasingly robust and reliable
52 studies (6352 patients) detected the prevalence of findings. To our knowledge, this study was the first to
p16INK4a positivity in vulvar cancer. The pooled prevalence of estimate an overall p16INK4a positivity prevalence of about
p16INK4a positivity in vulvar cancer was 34·1% (95% CI 34% in vulvar cancer and 66% in vulvar intraepithelial
30·9–37·4), with a considerable heterogeneity (I²=83·4%; neoplasia. When stratified by age, we observed a higher
p<0·0001; table 1; appendix p 34). In the stratified analyses pooled prevalence of HPV DNA in patients younger
according to the histological subtype of vulvar squamous than 60 years than in those aged 60 years or older, in
cell carcinoma, the pooled prevalence of p16INK4a positivity both vulvar cancer and vulvar intraepithelial neoplasia.
differed among the different histologic subtypes, with the Furthermore, we also were the first to report a pooled
highest prevalence of 95·9% (95% CI 73·0–100·0) in double positivity prevalence of HPV DNA and p16INK4a,
basaloid carcinoma and the lowest prevalence of 0% of about 20% in vulvar cancer and 44% in vulvar
(0·0–8·5) in verrucous carcinoma (table 1). In the intraepithelial neoplasia.
meta-regression, the pooled prevalence of p16INK4a positivity The most recent related meta-analysis was published
in vulvar cancer was significantly different between in 2017.11 Our analysis included an additional 27 new
geographical regions (p=0·0002; table 3). Additionally, the studies concerning the prevalence of HPV DNA in vulvar
prevalence of p16INK4a positivity was similar among studies cancer, adding about 3200 patients to the pooled estimate.
that used different cutoffs to define p16INK4a positivity The pooled HPV DNA prevalence in our meta-analysis
(appendix p 30). When estimating the prevalence of p16INK4a was in agreement with the results of previous meta-
positivity on the basis of HPV status of vulvar cancer, analyses.11–13 When stratified by geographical regions, we
we observed a much higher pooled p16INK4a positivity found significant variation in the pooled prevalence of
prevalence among HPV-positive vulvar cancer (73·3% HPV DNA in vulvar cancer across regions, with a high
[95% CI 64·7–81·2]) than in HPV-negative vulvar cancer preva­lence in Africa (66·0%). To our knowledge, the
(13·8% [10·0–18·1]; figure 3). prevalence of HPV DNA in vulvar cancer in Africa was
We included 23 studies (896 patients) that estimated examined for the first time in our meta-analysis, and the
the prevalence of p16INK4a positivity in vulvar intra­ high prevalence showed the necessity of preventing
epithelial neoplasia. The pooled prevalence of p16INK4a HPV infection in Africa. Regarding the prevalence of

www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9 9

 Articles

HPV DNA in vulvar intraepithelial neoplasia, we found a vulvar cancer. In this meta-analysis, the definition
similar pooled prevalence (76·1%) to that in the previous of p16INK4a positivity cutoff varied across studies, with
meta-analysis (76·3%).11 no signifi­ cant differences in the pooled prevalence
In our meta-analysis, the type-specific HPV DNA of p16INK4a positivity.
preva­lence was estimated separately in both vulvar When stratified by different subtypes of vulvar
cancer and vulvar intraepithelial neoplasia, according to squamous cell carcinoma, the prevalence of HPV DNA
the genotypes covered by the nine-valent HPV vaccine. was higher in basaloid and warty carcinoma than in
HPV16 and HPV18 are widely recognised as the crucial keratinising carcinoma. Additionally, the pooled HPV
oncogenic genotypes for cervical neoplasm,46,47 and DNA prevalence (86·5%) among warty carcinoma in
HPV16 and HPV33 are frequently associated with this study was higher than that (76·5%) in the previous
vulvar neoplasm.3,48 Similarly, we found that HPV16 and meta-analysis.11 Nevertheless, the additional studies
HPV33 were the two most prevalent HPV types in included in this meta-analysis might lead to increasingly
vulvar cancer and vulvar intraepithelial neoplasia, which robust results. We found a higher prevalence of p16INK4a
emphasised that the regular screening of the vulva for positivity in basaloid and warty carcinoma than in
precancerous lesions should not be ignored among keratinising carcinoma. Furthermore, verrucous carcin­
women with HPV16 or HPV33 infection. It is well oma (categorised as a kind of HPV-independent vulvar
established that high-risk HPV is a causative agent in carcinoma according to the latest WHO classification
vulvar carcinogenesis, and the pooled prevalence of criteria for female genital tumours, 2020),4 showed a
HPV (for both high-risk and low-risk HPV) in our study high HPV DNA prevalence of 58·9% and a low pooled
might lead to the overestimation of oncogenic effect of p16INK4a positivity prevalence of 0%. These results were
HPV on the vulvar neoplasm. Notably, we found that the consistent with a study of penile cancer49 that reported a
prevalence of HPV6 and HPV11 was low in vulvar somewhat high prevalence of HPV DNA (21·7%) and
cancer and vulvar intraepithelial neoplasia; however, a low pooled prevalence of p16INK4a positivity (4·8%).
some patients with vulvar cancer or vulvar intraepithelial However, these findings regarding verrucous carcinoma
neoplasia were only infected by HPV6 or HPV11, should be interpreted with caution, owing to the small
which indicated that the potential carcinogenic effect number of cases.
of HPV6 and HPV11 in vulvar neoplasm requires HPV E6 and E7 overexpression is a well established
further investigation. marker of the tumourigenesis driven by HPV.50
To our knowledge, this was the first meta-analysis to Nevertheless, the detection of HPV E6 and E7 gene
explore the pooled type-specific HPV DNA prevalence in products (ie, proteins or mRNA) from tumour tissue
vulvar cancer stratified by geographical region. Despite might not be easy to accomplish in routine diagnostic
HPV16 being the most predominant genotype in every laboratories. Hence, alternative markers of HPV
geographical region, the prevalence of other type-specific infection, such as the presence of HPV DNA or p16INK4a
HPV genotypes varied substantially between regions, overexpression, provide potential alternatives despite
with the second most frequent genotype being HPV18 in their respective concomitant drawbacks. However, the
Asia and South America, and HPV33 in the other existence of HPV DNA could reflect a transient infection
regions. We also observed a relatively high prevalence of rather than genuine HPV-driven carcinogenesis, which
HPV52, HPV45, and HPV31 in Oceania and Africa might lead to the overestimation of the proportion of
compared with other regions. Our findings reported truly HPV-driven tumours.51 Additionally, despite p16INK4a
a different pooled type-specific HPV distribution in overexpression being regarded as a surrogate marker
different regions and underlined the possibly greater for HPV-induced transformation, it can also be caused
potential of nine-valent HPV vaccination in preventing by an event other than HPV transformation.3,52,53 To
vulvar neoplasm. increase the specificity with minimal decreases in
This was the first meta-analysis to report the pooled sensitivity, some scholars have proposed that double
prevalence of p16INK4a positivity in vulvar cancer and positivity for HPV DNA and p16INK4a could be a well
vulvar intraepithelial neoplasia. Notably, the prevalence established marker for HPV-induced head and neck
of p16INK4a positivity across geographical regions in squamous cell carcinoma.54
this meta-analysis should be compared with caution, The double positivity for HPV DNA and p16INK4a is
because most studies on the prevalence of p16INK4a regarded as a better prognostic marker in patients with
positivity in vulvar cancer and vulvar intraepithelial tonsillar and base of tongue cancer, compared with
neoplasia originated from Europe and North America, either HPV or p16INK4a positivity alone.55 However, the
with only a small number of studies from other regions. clinical value of double positivity for HPV DNA and
As we expected, a much higher pooled prevalence p16INK4a in vulvar cancer has yet to be widely investigated.
of p16INK4a positivity was observed in HPV DNA- On the basis of previous findings in tonsillar and base
positive vulvar cancer compared with HPV DNA- of tongue cancer, we speculate that patients with vulvar
negative vulvar cancer (73·3% vs 13·8%), suggesting the cancer who have double positivity for HPV and p16INK4a
active involvement of HPV in the carcino­ genesis of might correlate with a better prognosis than those who

10 www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9

 Articles

are positive for either HPV or p16INK4a alone. Therefore, References

patients with vulvar cancer who are positive for either 1 Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022.
CA Cancer J Clin 2022; 72: 7–33.
HPV or p16INK4a alone might need more intensive 2 Del Pino M, Rodriguez-Carunchio L, Ordi J. Pathways of vulvar
treatment strategies and stricter follow-up surveillance intraepithelial neoplasia and squamous cell carcinoma.
compared with those with double positivity for HPV Histopathology 2013; 62: 161–75.
3 de Sanjosé S, Alemany L, Ordi J, et al. Worldwide human
and p16INK4a. Further prospective randomised controlled papillomavirus genotype attribution in over 2000 cases of
trials are warranted to investigate the association of intraepithelial and invasive lesions of the vulva. Eur J Cancer 2013;
HPV and p16INK4a positivity with prognosis in patients 49: 3450–61.
with vulvar cancer. 4 WHO Classification of Tumours Editorial Board. Tumours of the
vulva. In: Female genital tumours, WHO classification of tumours,
Our study had several limitations. First, considerable 5th edn, volume 4. Lyon: IARC Publications, 2020: 419–49.
heterogeneity among studies was observed, and despite 5 Bornstein J, Bogliatto F, Haefner HK, et al. The 2015 International
the investigation of sources of heterogeneity through Society for the Study of Vulvovaginal Disease (ISSVD) terminology
of vulvar squamous intraepithelial lesions. Obstet Gynecol 2016;
stratified analyses and meta-regressions, the sources 127: 264–68.
of heterogeneity could not be explained, which 6 Thuijs NB, van Beurden M, Bruggink AH, Steenbergen RDM,
might affect the generalisability of our findings. Berkhof J, Bleeker MCG. Vulvar intraepithelial neoplasia: incidence
and long-term risk of vulvar squamous cell carcinoma. Int J Cancer
Second, some of the estimates were based on small 2021; 148: 90–98.
samples and should be interpreted cautiously. Third, 7 Singh N, Gilks CB. Vulval squamous cell carcinoma and its
most of the studies included in this meta-analysis precursors. Histopathology 2020; 76: 128–38.
8 Hinten F, Molijn A, Eckhardt L, et al. Vulvar cancer: two pathways
reported a mixed prevalence of both high-risk and low- with different localization and prognosis. Gynecol Oncol 2018;
risk HPV, which limited the evaluation of pooled 149: 310–17.
prevalence for high-risk and low-risk HPV, respectively. 9 Carozzi F, Confortini M, Dalla Palma P, et al. Use of p16-INK4A
Finally, we included only English-language studies, overexpression to increase the specificity of human papillomavirus
testing: a nested substudy of the NTCC randomised controlled trial.
although it has been suggested that pooled estimates Lancet Oncol 2008; 9: 937–45.
do not seem to be affected by language restrictions in 10 Prigge E-S, Arbyn M, von Knebel Doeberitz M, Reuschenbach M.
meta-analyses.56 Diagnostic accuracy of p16INK4a immunohistochemistry in
oropharyngeal squamous cell carcinomas: a systematic review and
In conclusion, we found a high pooled prevalence of meta-analysis. Int J Cancer 2017; 140: 1186–98.
HPV DNA in vulvar intraepithelial neoplasia and 11 Faber MT, Sand FL, Albieri V, Norrild B, Kjaer SK, Verdoodt F.
vulvar cancer, with HPV16 and HPV33 being the most Prevalence and type distribution of human papillomavirus in
squamous cell carcinoma and intraepithelial neoplasia of the vulva.
predominant specific HPV genotypes, highlighting that Int J Cancer 2017; 141: 1161–69.
the regular screening of vulvar precancerous lesions 12 De Vuyst H, Clifford GM, Nascimento MC, Madeleine MM,
should be enhanced among women with HPV16 or Franceschi S. Prevalence and type distribution of human
papillomavirus in carcinoma and intraepithelial neoplasia of the vulva,
HPV33 infection. Different distribution of specific HPV vagina and anus: a meta-analysis. Int J Cancer 2009; 124: 1626–36.
genotypes among geographical regions was observed in 13 Smith JS, Backes DM, Hoots BE, Kurman RJ, Pimenta JM.
this meta-analysis. These findings emphasised the great Human papillomavirus type-distribution in vulvar and vaginal
cancers and their associated precursors. Obstet Gynecol 2009;
value of prophylactic nine-valent HPV vaccination in 113: 917–24.
preventing vulvar cancer and vulvar intraepithelial 14 Insinga RP, Liaw K-L, Johnson LG, Madeleine MM. A systematic
neoplasia, particularly in Africa, Oceania, and Europe. review of the prevalence and attribution of human papillomavirus
types among cervical, vaginal, and vulvar precancers and cancers
Additionally, this study highlighted the potential clinical in the United States. Cancer Epidemiol Biomarkers Prev 2008;
importance of double positivity for HPV DNA and p16INK4a 17: 1611–22.
in vulvar neoplasm. 15 Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020
statement: an updated guideline for reporting systematic reviews.
Contributors BMJ 2021; 372: n71.
KS, BK, and JM conceptualised and designed the study. ZL and PL wrote 16 Oonk MHM, Eijsink JJH, Volders HH, et al. Identification of
the original manuscript. ZL, PL, ZW, and ZZ collected the data. ZC, RC, inguinofemoral lymph node metastases by methylation markers in
and GL did the statistical analyses. ZL, PL, QH, and YZ accessed and vulvar cancer. Gynecol Oncol 2012; 125: 352–57.
verified the data. LL offered valuable suggestions and consultations on 17 Blettner M, Sauerbrei W, Schlehofer B, Scheuchenpflug T,
study design and data analyses. All authors reviewed and edited the Friedenreich C. Traditional reviews, meta-analyses and pooled
manuscript and approved the final version. All authors had full access to analyses in epidemiology. Int J Epidemiol 1999; 28: 1–9.
all the data in the study and had final responsibility for the decision to 18 Freeman MF, Tukey JW. Transformations related to the angular
submit for publication. and the square root. Ann Math Statist 1950; 21: 607–11.
Data sharing 19 Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring
inconsistency in meta-analyses. BMJ 2003; 327: 557–60.
This systematic review and meta-analysis used data that were available
in previously published studies. 20 Joura EA, Giuliano AR, Iversen O-E, et al. A 9-valent HPV vaccine
against infection and intraepithelial neoplasia in women.
Declaration of interests N Engl J Med 2015; 372: 711–23.
We declare no competing interests. 21 Balduzzi S, Rücker G, Schwarzer G. How to perform a meta-
analysis with R: a practical tutorial. Evid Based Ment Health
Acknowledgments 2019; 22: 153–60.
This study was supported by the Taishan Scholar Youth Project of
22 Rantshabeng PS, Moyo S, Moraka NO, et al. Prevalence of
Shandong Province (grant number tsqn201812130). We acknowledge all oncogenic human papillomavirus genotypes in patients diagnosed
the authors whose papers have been included in this meta-analysis for with anogenital malignancies in Botswana. BMC Infect Dis 2017;
their work, which made this analysis possible. 17: 731.

www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9 11

 Articles

23 Mpunga T, Chantal Umulisa M, Tenet V, et al. Human 40 Lérias S, Esteves S, Silva F, et al. CD274 (PD-L1), CDKN2A (p16),
papillomavirus genotypes in cervical and other HPV-related TP53, and EGFR immunohistochemical profile in primary,
anogenital cancer in Rwanda, according to HIV status. Int J Cancer recurrent and metastatic vulvar cancer. Mod Pathol 2020;
2020; 146: 1514–22. 33: 893–904.
24 Gasco M, Sullivan A, Repellin C, et al. Coincident inactivation of 41 Allo G, Yap ML, Cuartero J, et al. HPV-independent vulvar
14-3-3sigma and p16INK4a is an early event in vulval squamous squamous cell carcinoma is associated with significantly worse
neoplasia. Oncogene 2002; 21: 1876–81. prognosis compared with HPV-associated tumors.
25 Osakabe M, Hayashi M, Katayama Y, et al. Characteristics of vulvar Int J Gynecol Pathol 2020; 39: 391–99.
squamous cell carcinoma in Japanese women. Pathol Int 2007; 42 Preti M, Rotondo JC, Holzinger D, et al. Role of human
57: 322–27. papillomavirus infection in the etiology of vulvar cancer in Italian
26 Aulmann S, Schleibaum J, Penzel R, Schirmacher P, Gebauer G, women. Infect Agent Cancer 2020; 15: 20.
Sinn HP. Gains of chromosome region 3q26 in intraepithelial 43 Gensthaler L, Joura EA, Alemany L, Horvat R, de Sanjosé S, Pils S.
neoplasia and invasive squamous cell carcinoma of the vulva are The impact of p16(ink4a) positivity in invasive vulvar cancer on
frequent and independent of HPV status. J Clin Pathol 2008; disease-free and disease-specific survival, a retrospective study.
61: 1034–37. Arch Gynecol Obstet 2020; 301: 753–59.
27 de Koning MN, Quint WG, Pirog EC. Prevalence of mucosal and 44 Kolitz E, Lucas E, Hosler GA, et al. Human papillomavirus-positive
cutaneous human papillomaviruses in different histologic subtypes and -negative vulvar squamous cell carcinoma are biologically but
of vulvar carcinoma. Mod Pathol 2008; 21: 334–44. not clinically distinct. J Invest Dermatol 2022; 142: 1280–90.e7.
28 Alonso I, Fusté V, del Pino M, et al. Does human papillomavirus 45 van der Avoort IA, Shirango H, Hoevenaars BM, et al. Vulvar
infection imply a different prognosis in vulvar squamous cell squamous cell carcinoma is a multifactorial disease following two
carcinoma? Gynecol Oncol 2011; 122: 509–14. separate and independent pathways. Int J Gynecol Pathol 2006;
29 Lavorato-Rocha AM, Rodrigues IS, de Melo Maia B, et al. Cell cycle 25: 22–9.
suppressor proteins are not related to HPV status or clinical outcome 46 Goodman A. HPV testing as a screen for cervical cancer. BMJ 2015;
in patients with vulvar carcinoma. Tumour Biol 2013; 34: 3713–20. 350: h2372.
30 Reuschenbach M, Roos J, Panayotopoulos D, et al. Characterization 47 Serrano B, de Sanjosé S, Tous S, et al. Human papillomavirus
of squamous cell cancers of the vulvar anterior fourchette by genotype attribution for HPVs 6, 11, 16, 18, 31, 33, 45, 52 and 58 in
human papillomavirus, p16INK4a, and p53. J Low Genit Tract Dis 2013; female anogenital lesions. Eur J Cancer 2015; 51: 1732–41.
17: 289–97. 48 Garland SM, Joura EA, Ault KA, et al. Human papillomavirus
31 Lee LJ, Howitt B, Catalano P, et al. Prognostic importance of genotypes from vaginal and vulvar intraepithelial neoplasia in
human papillomavirus (HPV) and p16 positivity in squamous cell females 15–26 years of age. Obstet Gynecol 2018; 132: 261–70.
carcinoma of the vulva treated with radiotherapy. Gynecol Oncol 49 Olesen TB, Sand FL, Rasmussen CL, et al. Prevalence of human
2016; 142: 293–98. papillomavirus DNA and p16INK4a in penile cancer and penile
32 Sznurkowski JJ, Żawrocki A, Biernat W. The overexpression of p16 intraepithelial neoplasia: a systematic review and meta-analysis.
is not a surrogate marker for high-risk human papilloma virus Lancet Oncol 2019; 20: 145–58.
genotypes and predicts clinical outcomes for vulvar cancer. 50 von Knebel Doeberitz M, Oltersdorf T, Schwarz E, Gissmann L.
BMC Cancer 2016; 16: 465. Correlation of modified human papilloma virus early gene
33 Wakeham K, Kavanagh K, Cuschieri K, et al. HPV status and expression with altered growth properties in C4-1 cervical
favourable outcome in vulvar squamous cancer. Int J Cancer carcinoma cells. Cancer Res 1988; 48: 3780–86.
2017; 140: 1134–46. 51 Jung AC, Briolat J, Millon R, et al. Biological and clinical relevance
34 Pils S, Gensthaler L, Alemany L, Horvat R, de Sanjosé S, Joura EA. of transcriptionally active human papillomavirus (HPV) infection in
HPV prevalence in vulvar cancer in Austria. Wien Klin Wochenschr oropharynx squamous cell carcinoma. Int J Cancer 2010;
2017; 129: 805–9. 126: 1882–94.
35 Halec G, Alemany L, Quiros B, et al. Biological relevance of human 52 Alemany L, Saunier M, Tinoco L, et al. Large contribution of
papillomaviruses in vulvar cancer. Mod Pathol 2017; 30: 549–62. human papillomavirus in vaginal neoplastic lesions: a worldwide
36 Hinten F, Molijn A, Eckhardt L, et al. Vulvar cancer: two pathways study in 597 samples. Eur J Cancer 2014; 50: 2846–54.
with different localization and prognosis. Gynecol Oncol 2018; 53 Alemany L, Cubilla A, Halec G, et al. Role of human
149: 310–17. papillomavirus in penile carcinomas worldwide. Eur Urol 2016;
37 Arians N, Prigge ES, Nachtigall T, et al. Overexpression of p16(INK4a) 69: 953–61.
serves as prognostic marker in squamous cell vulvar cancer patients 54 Mena M, Taberna M, Tous S, et al. Double positivity for HPV-DNA/
treated with radiotherapy irrespective of HPV-status. Front Oncol p16 is the biomarker with strongest diagnostic accuracy and
2019; 9: 891. prognostic value for human papillomavirus related oropharyngeal
38 Stefansson K, Oda H, Öfverman C, Lundin E, Hedman H, cancer patients. Oral Oncol 2018; 78: 137–44.
Lindquist D. LRIG1-2 and LMO7 immunoreactivity in vulvar 55 Garnaes E, Frederiksen K, Kiss K, et al. Double positivity for HPV
squamous cell carcinoma: association with prognosis in relation to DNA/p16 in tonsillar and base of tongue cancer improves
HPV-DNA and p16INK4a status. Oncology Reports 2019; 42: 142–50. prognostication: Insights from a large population-based study.
39 Thomas J, Leufflen L, Chesnais V, et al. Identification of specific Int J Cancer 2016; 139: 2598–605.
tumor markers in vulvar carcinoma through extensive human 56 Moher D, Pham B, Klassen TP, et al. What contributions do
papillomavirus DNA characterization using next generation languages other than English make on the results of meta-analyses?
sequencing method. J Low Genit Tract Dis 2020; 24: 53–60. J Clin Epidemiol 2000; 53: 964–72.

12 www.thelancet.com/oncology Published online March 15, 2023 https://doi.org/10.1016/S1470-2045(23)00066-9