Cancer Epidemiology 39 (2015) 237–241
Contents lists available at ScienceDirect
Cancer Epidemiology
The International Journal of Cancer Epidemiology, Detection, and Prevention
journal homepage: www.cancerepidemiology.net
Random biopsy in colposcopy-negative quadrant is not effective
in women with positive colposcopy in practice
Yan Song a,1, Yu-Qian Zhao b,1, Xun Zhang a,*, Xiao-Yang Liu a, Ling Li a, Qin-Jing Pan a,
Gui-Hua Shen a, Fang-Hui Zhao b, Feng Chen b, Wen Chen b,**, You-Lin Qiao b
a
b
Department of Pathology, Cancer Institute & Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
Department of Cancer Epidemiology, Cancer Institute & Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
A R T I C L E I N F O
A B S T R A C T
Article history:
Received 22 August 2014
Received in revised form 7 January 2015
Accepted 11 January 2015
Available online 12 February 2015
Aim: To assess the efficacy of random biopsy in diagnosing those high-grade squamous intraepithelial
lesions or carcinomas (HSIL+) missed by colposcopy-directed biopsy, and to identify the scenarios of
cervical cancer screening when random biopsy is necessary.
Patients/interventions: Data from 1997 women who participated in the Shanxi Province Cervical Cancer
Screening Study I (SPOCCS I) were reviewed. Each woman received human papillomavirus (HPV) testing
with the second-generation hybrid capture, liquid-based cytology, four-quadrant biopsy and
endocervical curettage. The final diagnosis was based on the most severe pathological result obtained.
The efficacy of random biopsy and colposcopy-directed biopsy was evaluated on the basis of the final
pathological results.
Results: For women with severe cytological abnormalities (HSIL+) and negative colposcopy, the yield of
HSIL+ diagnosed by random biopsy was 25%. On the other hand, the yield of HSIL+ diagnosed by random
biopsies in the negative quadrant was no more than 4% when the colposcopy was positive, regardless of
the cytological findings. For women with negative HPV, no HSIL+ was found by random biopsy. For
women with severe cytological abnormalities (HSIL+) and positive HPV, the yield of HSIL+ diagnosed by
random biopsy was 35% when colposcopy was negative. For women with low-grade intraepithelial
lesion (LSIL) and positive HPV, the yield of HSIL+ diagnosed by random biopsy was 12.5% when
colposcopy was negative.
Conclusion: Random biopsy is not effective in the negative quadrant in women with positive colposcopy,
but should be performed in women with cytological HSIL+ but negative colposcopy, or in those with
cytological LSIL or HGSL+ and positive HPV but negative colposcopy.
ß 2015 Elsevier Ltd. All rights reserved.
Keywords:
Colposcopy
Random biopsy
Colposcopic-directed biopsy
Cervical intraepithelial neoplasia
1. Introduction
The introduction of cervical screening – including cytological
diagnosis, human papillomavirus (HPV) testing, and colposcopy
with biopsy – has markedly reduced the number of deaths from
cervical cancer in recent years [1,2]. It is generally agreed that
* Corresponding author at: Department of Pathology, Cancer Institute & Hospital,
Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing
100021, China. Tel.: +86 10 8778 7508.
** Corresponding author at: Department of Cancer Epidemiology, Cancer Institute
& Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College,
Beijing 100021, China. Tel.: +86 10 8778 7411.
E-mail addresses: [email protected] (X. Zhang), [email protected]
(W. Chen).
1
These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.canep.2015.01.008
1877-7821/ß 2015 Elsevier Ltd. All rights reserved.
women with cytological high-grade squamous intraepithelial
lesion (HSIL) should be referred for colposcopy. Women with
cytological low-grade intraepithelial lesions (LSILs) should also
receive colposcopy after a single abnormal result. In women with
cytological borderline changes, i.e. atypical squamous cells of
undetermined significance (ASCUS), acceptable options include
testing for high-risk HPV, repeat cervical cytology, and/or
immediate colposcopy [3]. In colposcopy, acetic acid is applied
to the cervix under magnification, the grade of lesion is assessed,
and the lesion is biopsied under colposcopic guidance (colposcopydirected biopsy) [4]. Colposcopic-directed biopsy is the most
commonly used diagnostic method for high-grade squamous
intraepithelial lesions or carcinoma (HSIL+) in cervical screening
[5,6], while HSIL+ cases may be missed by colposcopy-directed
biopsies when lesions are not obvious [7,8]. Based on the results of
364 HSIL+ cases, it was reported that 37.4% of HSIL+ cases were
238
Y. Song et al. / Cancer Epidemiology 39 (2015) 237–241
detected by random biopsy [9]. Another paper reported that
colposcopy-directed biopsy augmented with random biopsy could
markedly increase the yield of HSIL+, regardless of skill [10]. As a
result, some authors recommend applying random biopsies to the
squamocolumnar junction in practice if cytology is high-grade
[9,11].
In current clinical practice, the improved sensitivity of
cytological screening with the addition of HPV testing has led to
increasing numbers of referrals for colposcopy. Is it necessary to
perform random biopsy in every negative quadrant in colposcopy if
cytology is high-grade? What is the best way to combine random
biopsy with colposcopy-directed biopsy to detect more HSIL+ after
triaging with cytology and/or HPV testing? Being the only study in
which all participants received identical screening tests –
including colposcopy, four-quadrant biopsies and endocervical
curettage (ECC) – Shanxi Province Cervical Cancer Screening Study
I (SPOCCS I) provided us with a unique opportunity to assess the
efficacy of random biopsy in every quadrant in diagnosing HSIL+ in
practice.
squamocolumnar junction was seen. The cervix was divided
visually into four quadrants by lines drawn from 12 to 6 o’clock
and from 3 to 9 o’clock positions. Each quadrant of the cervix
was graded separately as negative (no lesions seen), low-grade
squamous intraepithelial lesion (LSIL) suggestive of HPV or cervical
intraepithelial neoplasia 1 (CIN1), high-grade squamous intraepithelial lesion (HSIL) suggestive of intraepithelial neoplasia 2 or 3
(CIN2 or 3), or invasive cervical cancer (ICC). All abnormalities
diagnosed by colposcopy were biopsied. If colposcopic examination
showed no lesions in any quadrant, a random biopsy was obtained at
the squamocolumnar junction in four quadrants at the 2, 4, 8, or
10 o’clock positions, respectively. It was acceptable to take more
than one biopsy per quadrant depending on the colposcopic
impression. Unlike traditional biopsy forceps with 5-mm jaws, all
biopsies were performed with a bronchoscopy biopsy instrument
that has 2-mm jaws and is virtually painless for most patients. ECC
was also performed on every patient with a Kevorkian curette.
2. Subjects and methods
The histological diagnosis was based on the consensus of two
gynecological pathologists of the Cancer Institute and Hospital in
Beijing who independently reviewed every biopsy. In specimens
with discordant diagnoses, the final diagnosis was based on the
majority assessment after a third gynecological pathologist
examined the specimen. The pathologists were not aware of the
results of any of the other tests. The final diagnosis was based on
the most severe biopsy result obtained (i.e. colposcopy-directed,
random, or ECC). The biopsies were also sent to the Cleveland Clinic
for a similar review. No significant differences in diagnoses were
found between the Beijing and the Cleveland Clinics.
2.1. Subjects
Data from 1997 women who participated in SPOCCS I between
June 1999 and July 1999 were reviewed. The Institutional Review
Boards for human research subjects of the Cleveland Clinic
Foundation and the Cancer Institute/Hospital of the Chinese
Academy of Medical Sciences (CICAMS) approved this study. Nonpregnant women between the ages of 35 and 45 with no history of
cervical screening, pelvic radiation, or hysterectomy were eligible.
Only women who voluntarily signed the informed consent after
the doctors explained the study procedure and potential side
effects in detail were enrolled in our study. All women were
screened for cervical lesions/neoplasia. Each woman underwent
cervical screening with cytological analysis and visual inspection
with acetic acid (VIA). Physicians gathered cervical samples for
HPV DNA tests and liquid-based cytology. Colposcopy was
performed on every woman and biopsy was performed in every
quadrant. An ECC was also done on every patient.
2.2. Liquid-based cytology and HPV testing
Specimens for HPV testing were obtained from the endocervix
with a conical-shaped brush placed high in the vagina and rotated
three to four times. The conical brushes were placed into a liquid
medium and tested for a mixture of 13 intermediate- and high-risk
types of HPV (HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 5658, 59 and 68)
with the second-generation hybrid capture microplate-based HPV
test (HC2 test, Digene, Corp.). A cutoff value of 1.0 pg HPV DNA was
regarded as positive.
Specimens for liquid-based cytology were obtained with a
plastic spatula and an endocervical brush and placed in transport
medium. Liquid-based cytology was prepared via the ThinPrep
method (CytycCorp., Boxborough, MA) and interpreted using the
Bethesda classification system by cytopathologists who were
blinded to the results of the screening tests from the Cancer
Institute and Hospital in Beijing. Cytopathologists at the Cleveland
Clinic reviewed all abnormal slides and 5% of the normal slides. No
significant differences in diagnoses were seen between the Beijing
and the Cleveland Clinics.
2.4. Pathological diagnosis
2.5. Statistical analysis
Statistical analyses were performed using SAS software (SAS
Institute, Cary, NC). All continuous variables were tested for
normal distribution and presented by mean standard deviation.
Sample proportions and corresponding 95% Clopper–Pearson confidence intervals (95%CI) were estimated under binomial assumption.
Comparisons between proportions were analyzed using the Fisher
exact test. All tests were two-sided, and p-values <0.05 were
considered to be statistically significant.
3. Results
3.1. Participant characteristics
The mean age of participants was 39.1 3.16 years. The mean
number of pregnancies was 3.1 1.27 and the mean number of births
was 2.6 0.93. Among the 1997 women, 98.5% were currently
married, 93.3% had never smoked, and 0.3% had a history of
condyloma. Of the enrolled 1997 women, 1784 (89.3%) were negative
for squamous intraepithelial lesion, 127 (6.4%) had LSIL (CIN1), 74
(3.8%) had HSIL (43 CIN2 and 31 CIN3), and 12 (0.6%) had ICC.
Therefore, 4.3% (86 of 1997) had HSIL+. Of these 86 HSIL+ cases, 74.4%
(28 CIN2, 24 CIN3 and 12 ICC) were diagnosed by colposcopy-directed
biopsy, 20.9% (14 CIN2 and 4 CIN3) were diagnosed by random
biopsy, and 4.7% (one CIN2 and three CIN3) were diagnosed by ECC
alone. In order to analyze the efficacy of biopsy, four HSIL+ cases
found by ECC alone were excluded from this paper.
2.3. Colposcopy and biopsy
3.2. Efficacy of colposcopy-directed biopsies and random biopsies and
correlation with results of cytology (ThinPrep Pap)
Gynecological oncologists performed all colposcopies and
biopsies. Colposcopies were considered satisfactory if the entire
Of the ThinPrep Pap smears, four were unsatisfactory,
1480 were negative, 314 were ASCUS, 91 were LSIL, 96 were HSIL,
Y. Song et al. / Cancer Epidemiology 39 (2015) 237–241
and 12 were ICC. The sensitivity of ThinPrep Pap (>ASCUS) in the
detection of HSIL+ in this trial was 89.0% (95%CI: 80.9–94.5%) and
the specificity was 93.4% (95%CI: 92.3–94.4%).
Among women with positive colposcopy, colposcopy-directed
biopsies detected more HSIL+ with increasing severity of
cytological diagnosis (12.5% for cytological LSIL to 66.7% for
cytological HSIL+), while random biopsies did not detect more
cases of HSIL+ regardless of cytological results (3.1% for cytological
LSIL and 2.4% for cytological HSIL+). Moreover, most of the HSIL+
cases diagnosed by random biopsy in these conditions were found
in patients when only one or two quadrants were positive on
colposcopy.
Among women with negative colposcopy, the yield of HSIL+
diagnosed by random biopsy increased with the severity of
cytological diagnosis. The yield of HSIL+ diagnosed by random
biopsy was 6.8% (95%CI: 2.2–15.6%) among women with cytological LGIL and 25% (95%CI: 10.8–44.9%) among women with
cytological HSIL+ respectively (Table 1).
3.3. Efficacy of colposcopy-directed biopsies and random biopsies and
correlation with results of HPV test
Of the 1997 women tested for HPV with HC2, 364 were positive,
1576 were negative, and 57 were of unknown status. The
sensitivity of HC2 for detecting HSIL+ was 97.6% (95%CI: 94.4–
100%) and specificity was 84.8% (95%CI: 83.2–86.4%).
For women with positive HC2, the yield of HSIL+ diagnosed by
colposcopy-directed biopsy was 37.5% (95%CI: 30.3–45.2%) while
the yield of HSIL+ diagnosed by random biopsy was 3.1% (95%CI: 1.2–
6.8%) when there was a positive colposcopy. When the colposcopy
was negative, the yield of HSIL+ diagnosed by random biopsy was
6.4% (95%CI: 2.2–15.6%). For women with negative HC2, the yield of
HSIL+ diagnosed by colposcopy-directed biopsy was 0.6% (0.1–1.9%)
and no HSIL+ was diagnosed by random biopsy.
3.4. Efficacy of colposcopy-directed biopsies and random biopsies and
correlation with the combined results of HPV test and cytology
(ThinPrep Pap)
The combined results of cytology and HPV test showed that all
subjects with HSIL+ had either an abnormal cytology or a positive
239
HPV. No case of HSIL+ was found in women with both negative
cytology and negative HPV.
In the 99 women with cytological HSIL+ and positive HPV, the
yield of HSIL+ diagnosed by colposcopy-directed biopsy was 67.1%
(53/79, 56.2–76.8%) if the colposcopy was positive. An additional 12/
79 (15.2%, 8.5–24.4%) HSIL+ were detected in both colposcopydirected biopsy and random biopsy, and 2/79 (2.5%, 0.4–8.1%) were
detected only by random biopsy. Among the 20 women with
cytological HSIL and positive HPV but negative colposcopy, random
biopsy detected seven HSILs (35%, 16.7–57.3%) (Table 2). In the
53 women with cytological LSIL and positive HPV, if the colposcopy
was positive (21/53), the yield of HSIL+ diagnosed was 4/21 (19.0%,
6.4–39.8%) by colposcopy-directed biopsy and 1/21 (4.8%, 0.2–
21.3%) by random biopsy. Among these 53 women, when the
colposcopy was negative (32/53), random biopsy detected 4/32
(12.5%, 4.1–27.5%) HSILs. In the 54 women with ASCUS and positive
HPV, the yield of HSIL+ was 5/18 (27.7%, three by colposcopydirected biopsy, one by random biopsy, one by both methods) and 1/
36 (2.8%, 0.1–13.0%, by random biopsy) with positive and negative
colposcopy, respectively. Among 158 patients with positive HPV test
and normal cytology, the yield of HSIL+ detected by colposcopydirected biopsy was 2.4% (0.1–11.2%). The yield of HSIL+ detected by
random biopsy was 1.7% (0.3–5.6%) and 4.8% (0.8–12.9%) when
colposcopy was negative or positive, respectively. Regardless of
cytological results, no HGIL+ was found by random biopsy among
women with negative HPV.
4. Discussion
The inadequacy of the colposcopy-directed biopsy in diagnosing HSIL+ has long been a concern [12]. The efficacy of colposcopydirected biopsy depends not only on the efficacy of colposcopy
itself but also on the thickness of the cervical epithelium and the
number of biopsies [13,14]. As the efficacy of colposcopy-directed
biopsy increases, the efficacy of random biopsy is sure to decrease.
Of the 86 HSIL+ cases in this study, 74.4% were diagnosed by
colposcopy-directed biopsy and 20.9% were diagnosed by random
biopsy, indicating that random biopsy as a supplement to
colposcopy-directed biopsy can discover HSIL+ cases missed by
conventional colposcopy-directed biopsy.
Table 1
Efficacy of colposcopic-directed biopsies and random biopsies to detect HGIL+ with correlation to cytology (ThinPrep Pap).
Cytology = HSIL+ (108)
Scopy+ (84)
1 or 2 quadrants (46)
3 or 4 quadrants (38)
Scopy (24)
Cytology = LGIL (91)
Scopy+ (32)
1 or 2 quadrants (26)
3 or 4 quadrants (6)
Scopy (59)
Cytology = ASCUS (314)a
Scopy+ (81)
1 or 2 quadrants (69)
3 or 4 quadrants (12)
Scopy (231)
Cytology normal (1480)
Scopy+ (321)
1 or 2 quadrants (275)
3 or f 4 quadrants (46)
Scopy (1159)
a
2 scopy unknown.
HSIL+ detected by colposcopic
directed biopsy
HSIL+ detected by
random biopsy only
Total
N (percent, 95%CI)
N (percent, 95%CI)
N (percent, 95%CI)
56/84 (66.7, 56.1–76.1)
20/46 (43.5, 30.0–58.0)
36/38 (94.7, 83.7–99.1)
2/84 (2.4, 0.4–7.6)
2/46 (4.3, 0.7–13.6)
0
6/24 (25, 10.8–44.9)
58/84 (69.1, 58.6–78.2)
22/46 (47.8, 33.8–62.2)
36/38 (94.7, 83.7–99.1)
6/24 (25, 10.8–44.9)
4/32 (12.5, 4.1–27.5)
2/26 (7.7, 1.3–23.2)
2/6 (33.3, 6.0–73.8)
1/32 (3.1, 0.2–14.5)
1/26 (3.8, 0.2–17.5)
0
4/59 (6.8, 2.2–15.6)
5/32 (15.6, 6.0–31.3)
3/26 (11.5, 3.0–28.3)
2/6 (33.3, 6.0–73.8)
4/59 (6.8, 2.2–15.6)
3/81 (3.7, 0.95–9.7)
2/69 (2.9, 0.5–9.2)
1/12 (8.3, 0.4–34.7)
1/81 (1.2, 0.06–5.9)
0
1/12 (8.3, 0.4–34.7)
1/231 (0.4, 0.02–2.1)
4/81 (4.9, 1.6–11.5)
2/69 (2.9, 0.5–9.2)
2/12 (16.7, 2.9–45.1)
1/231 (0.4, 0.02–2.1)
1/321 (0.3, 0.02–1.5)
1/275 (0.4, 0.02–1.8)
0
1/321 (0.3, 0.02–1.5)
0
0
2/1159 (0.2, 0.03–0.6)
2/321 (0.6, 0.10–2.0)
1/275 (0.4, 0.02–1.8)
0
2/1159 (0.2, 0.03–0.6)
Y. Song et al. / Cancer Epidemiology 39 (2015) 237–241
240
Table 2
Efficacy of colposcopic-directed biopsies and random biopsies to detect HSIL+ with correlation to combination results of cytology (ThinPrep pap) and HPV.
Cytology = HSIL+ (108)a
HPV+ (99)
Scopy+ (79)
Scopy (20)
HPV (6)
Scopy+ (2)
Scopy (4)
Cytology = LSIL (91)b
HPV+ (53)
Scopy+ (21)
Scopy (32)
HPV (37)
Scopy+ (11)
Scopy (26)
Cytology = ASCUS (314)c
HPV+ (54)
Scopy+ (18)
Scopy (36)
HPV (246)
Scopy+ (61)
Scopy (185)
Cytology normal (1480)d
HPV+ (158)
Scopy+ (42)
Scopy (116)
HPV (1281)
Scopy+ (272)
Scopy (1009)
a
b
c
d
HSIL+ detected by
colposcopic directed
biopsy only
HSIL+ detected by
colposcopic directed
biopsy and random
biospsy
HSIL+ detected by
random biopsy only
N (percent, 95%CI)
N (percent, 95%CI)
N (percent, 95%CI)
53/79 (67.1, 56.2–76.8)
0
12/79 (15.2 8.5–24.4)
0
2/79 (2.5, 0.4–8.1)
7/20 (35, 16.7–57.3)
1/2 (50, 2.5–97.5)
0
0
0
0
0
4/21 (19.0, 6.4–39.8)
0
0
0
1/21 (4.8, 0.2–21.3)
4/32 (12.5, 4.1–27.5)
0
0
0
0
0
0
3/18 (16.7, 4.4–39.0)
0
1/18 (5.6, 0.3–24.5)
0
1/18 (5.6, 0.3–24.5)
1/36 (2.8, 0.1–13.0)
1/61 (1.6, 0.08–7.8)
0
0
0
0
0
1/42 (2.4, 0.1–11.2)
0
0
0
2/42 (4.8, 0.8–12.9)
2/116 (1.7, 0.3–5.6)
0
0
0
0
0
0
3 HPV unknown/not done.
1 HPV unknown/not done.
12 HPV unknown/not done. 2 coloscopy unknown.
41 HPV unknown/not done.
In clinical practice, cytology and HPV testing are the most
important triage tools for colposcopic referral [2]. With increasing
numbers of abnormal smears and HPV tests, colposcopy is being
used more frequently. There is clearly a need to identify the best way
to combine random biopsy with colposcopy-directed biopsy to
detect more HSIL+ cases. In areas where only cytology is used as
triage for colposcopic referral, our results showed that the ability of
random biopsy to find HSIL+ increased with the severity of the
cytological abnormality in women with negative colposcopy. When
there were severe cytological abnormalities (HSIL+) but a negative
colposcopy, the ability of random biopsy to discover HSIL+ was 25%,
supporting the use of random biopsy in this condition. In a
subsequent study, SPOCCS II, Pretorius et al. came to the same
conclusion, but they did not discuss the efficacy of random biopsy in
the negative quadrants after colposcopy-directed biopsies were
performed on colposcopy-visible lesions [9]. Our results have shown
that random biopsies in addition to colposcopy-directed biopsy did
detect a few more HSILs, although the detection rate was similar
between cytological LSIL (3.1%) and cytological HSIL cases (2.4%),
indicating that random biopsy in positive colposcopy was not as
important as that in negative colposcopy. Our data also showed that
directed colposcopic biopsy in more quadrants was associated with
an increased yield of HSIL detection, a result similar to that reported
by Belinson et al. [15]. In their report, the sensitivity of colposcopy
for detecting HSIL+ involving one quadrant was 64%, and for
detecting lesions involving three or four quadrants it was 89%. In
women with mild cytological abnormality (LSIL) and negative
colposcopy, the yield of random biopsy in detecting HSIL+ was 6.8%,
indicating that there is probably no strong need to perform random
biopsy in patients with LSIL if cytology is used as the only triage tool.
The levels of diagnoses experienced vary largely among
cytologists in different countries, and false-negative fractions
reported or calculated from data in the literature range from
approximately 2% to 28% [11]. Especially in some rural areas in
China, few experienced cytologists can be found, so additional
methods besides cytology must be used in practice to increase the
sensitivity to detect HSIL+. The HPV test was such an additional
tool. Here we showed that among the patients with cytological
LSIL, HSIL+ was seen in 6.8% patients with negative colposcopy by
random biopsy, and this was increased to 12.5% if there was a
positive HPV test, indicating that cytological LSIL in the presence of
positive HPV should prompt for random biopsy even if there is a
negative colposcopic examination. In our later study, SPOCCS II, the
efficacy of random biopsy in detecting HSIL in women with
positive HPV and cytological LSIL was 8.2% (37/450) (data not
shown). Kelly et al. followed up 965 HPV-positive women with
negative colposcopy but with ASCUS and LSIL for 3 years, and they
found 4.4% (95%CI 4.0–7.0%) HSIL+, further indicating that HSIL+
can be found in patients with negative colposcopy (they did not
separate ASCUS from LSIL) [16]. No HSIL+ was found in women
with negative HPV, regardless of cytological results.
In conclusion, random biopsy is a necessary supplement to
colposcopy-directed biopsy for detecting HSIL+ cases. Random
biopsy can increase HSIL+ detection only among women with an
HPV-positive test. In the absence of HPV testing but with abnormal
Pap, colposcopy images provide the best selection for sampling to
detect HSIL+. In patients with positive colposcopy, random biopsy
is not effective in increasing the detection rate of HSIL+; in contrast,
it is much more effective in patients with negative colposcopy but
with mild or severe cytology (LSIL or HSIL+) and a positive HPV test.
Y. Song et al. / Cancer Epidemiology 39 (2015) 237–241
Conflict of interest statement
All authors declared no conflict of interest.
Authorship contribution
You-Lin Qiao involved in study conception and design.
Xiao-Yang Liu, Qin-Jing Pan, Ling Li, Gui-Hua Shen, Feng Chen
involved in acquisition of data. Fang-Hui Zhao involved in analysis
and interpretation of data. Yan Song, Yu-Qian Zhao involved in
drafting of manuscript. Xun Zhang, Wen Chen involved in critical
revision.
Acknowledgments
We thank all collaborators in this project, especially Robert G.
Pretorius from the department of Obstetrics and Gynecology,
Southern California Permanente Medical Group, and Jerome L.
Belinson from the department of Obstetrics and Gynecology, the
Cleveland Clinic Foundation. We also thank the field collaborators in
Xiangyuan County Women and Children Hospital, Shanxi Province.
We thank Irene Joan Chang for assistance in editorial review of this
manuscript. We thank Dengfeng Cao and Peng Li from International
Agency for Research of Cancer, France, for assistance in editorial
review of this manuscript. Funding for this study was provided by
the Taussig Cancer Center Cleveland Clinic Foundation, Cancer
Institute/Hospital, Chinese Academy of Medical Sciences, Terry Fox
Foundation, Transamerica Corporation, Digene Corp., Cytyc Corp.,
Optical Biopsy Tech, LLC, and Carl Zeiss, Inc.
References
[1] Deodhar KK. Screening for cervical cancer and human papilloma virus: Indian
context. Clin Lab Med 2012;32:193–205.
241
[2] Saleh FH. Cervical cancer: a missed health priority in Tanzania. East Afr J Public
Health 2011;8:247–9.
[3] Wright Jr TC, Massad LS, Dunton CJ, Spitzer M, Wilkinson EJ, Solomon D.
2006 consensus guidelines for the management of women with abnormal
cervical screening tests. J Low Genit Tract Dis 2007;11:201–22.
[4] Nazeer S, Shafi MI. Objective perspective in colposcopy. Best Pract Res Clin
Obstet Gynaecol 2011;25:631–40.
[5] Sankaranarayanan R, Wesley R, Thara S, Dhakad N, Chandralekha B, Sebastian
P, et al. Test characteristics of visual inspection with 4% acetic acid (VIA) and
Lugol’s iodine (VILI) in cervical cancer screening in Kerala, India. Int J Cancer
2003;106:404–8.
[6] Pretorius RG, Kim RJ, Belinson JL, Elson P, Qiao YL. Inflation of sensitivity of
cervical cancer screening tests secondary to correlated error in colposcopy. J
Low Genit Tract Dis 2006;10:5–9.
[7] Stoler MH, Vichnin MD, Ferenczy A, Ferris DG, Perez G, Paavonen J, et al. The
accuracy of colposcopic biopsy: analyses from the placebo arm of the Gardasil
clinical trials. Int J Cancer 2011;128:1354–62.
[8] Massad LS, Collins YC. Strength of correlations between colposcopic impression and biopsy histology. Gynecol Oncol 2003;89:424–8.
[9] Pretorius RG, Zhang WH, Belinson JL, Huang MN, Wu LY, Zhang X, et al.
Colposcopically directed biopsy, random cervical biopsy, and endocervical
curettage in the diagnosis of cervical intraepithelial neoplasia II or worse. Am J
Obstet Gynecol 2004;191:430–4.
[10] Pretorius RG, Belinson JL, Burchette RJ, Hu S, Zhang X, Qiao YL. Regardless of
skill, performing more biopsies increases the sensitivity of colposcopy. J Low
Genit Tract Dis 2011;15:180–8.
[11] Naryshkin S. The false-negative fraction for Papanicolaou smears: how often
are abnormal smears not detected by a standard screening cytologist? Arch
Pathol Lab Med 1997;121(March):270–2.
[12] Underwood M, Arbyn M, Parry-Smith W, De Bellis-Ayres S, Todd R, Redman
CW, et al. Accuracy of colposcopy-directed punch biopsies: a systematic
review and meta-analysis. BJOG 2012;119:1293–301.
[13] Yang B, Pretorius RG, Belinson JL, Zhang X, Burchette R, Qiao YL. False negative
colposcopy is associated with thinner cervical intraepithelial neoplasia 2 and
3. Gynecol Oncol 2008;110:32–6.
[14] Gage JC, Hanson VW, Abbey K, Dippery S, Gardner S, Kubota J, et al. Number of
cervical biopsies and sensitivity of colposcopy. Obstet Gynecol 2006;108:264–72.
[15] Belinson J, Qiao YL, Pretorius R, Zhang WH, Elson P, Li L, et al. Shanxi Province
Cervical Cancer Screening Study: a cross-sectional comparative trial of multiple techniques to detect cervical neoplasia. Gynecol Oncol 2001;83:439–44.
[16] Kelly RS, Walker P, Kitchener H, Moss SM. Incidence of cervical intraepithelial
neoplasia grade 2 or worse in colposcopy-negative/human papillomaviruspositive women with low-grade cytological abnormalities. BJOG 2012;119:
20–25.