Thank laurelshihxbl for 翻译 of the AGC abstract:

verification bias:

Example: We perform Pap screening test with biopsy follow up. 10% women with LSIL Pap have or develope CIN 2/3. Then we conclude the risk of women with LSIL Pap for CIN 2/3 is 10%. However, women with normal Pap also can develop CIN 2/3 in study period. Rare studies can include a control group----to do the cervical biopsy for the women with negative Pap. True risk for LSIL-CIN 2/3 may be less than 10%.

This is called verification bias.

In fact the only study in which the control group of women with negative Pap had colposcopic examination was performed in Shanxi, China several years ago. This is called verification bias-adjusted.

In our AGC study, we added the data about hysterctomy specimens.  We calculated the AGC rate in all women with hysterectomy (cervical tissue were examined). This data avioided verification bias-----called verification bias- adjusted.

It is a very complicated concept. I am not sure if I explain it clearly.

 Above are my Pap-HPV research results in the past three years. Wish our young pathologists can also do some research and summarize you data based on your situation.

Thank you for reading,

cz

2008 USCAP abstract (is working for paper)

Follow-up outcomes of cytological and histological abnormalities among women with negative computer-imaged liquid-based Pap and positive HPV DNA test results

C Zhao, RM Austin

Department of Pathology, Magee-Womens Hospital (MWH) of University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania

Background:

Limited data from the U.S. and overseas has been reported on the natural history of high risk(hr) HPV DNA positive women screened with negative conventional Pap smear results, but no reports have documented follow-up of cytology negative- HPV positive women routinely screened with newer methods of liquid-based cytology (LBC), computer-assisted screening, and HPV reflex or co-testing. The purpose of this study was to document the development of cytological and histological abnormalities among several hundred women who tested hrHPV DNA positive along with negative computer-imaged liquid-based Pap co-test results (DNAwithPap).  

Design:

The computerized hospital records of MWH were searched for patients reported as negative on ThinPrep Imaging System (TIS)-imaged ThinPrep Pap tests (TPPT) who also had positive Hybrid Capture 2 (HC2) hrHPV DNAwithPap co-test results over a 30 month span between July 2005 and December 2007. Cytologic and histologic follow-up outcomes were analyzed.

Results:

During the study period 402 women with negative TPPT and concurrent positive HC2 hrHPV DNA test results had documented cytologic and/or histologic follow-up. Histologic follow-up included 111 women who underwent cervical biopsy with or without ECC and 39 who underwent ECC alone. The mean age was 41.6 years (15-84 years). The average follow-up period was 13 months, ranging from 1 to 35 months (mean 10.6 m). Follow-up results documented that 8 of 402 (2.0%) women had tissue diagnoses of intraepithelial neoplasia 2+, including, four CIN2, 2 CIN3, one VAIN3, and one case of AIS on biopsy and microinvasive endocervical adenocarcinoma on subsequent conization. CIN1 or LSIL were confirmed in 61 (15.2%) and 82 (20.4%) had follow-up ASC-US Pap test results without biopsy findings of CIN1+. All CIN 2+ and 50 of 61 CIN 1 lesions were diagnosed based on histology. The interval between positive HPV and negative TPPT findings and an initial diagnosis of CIN 2+ ranged from 1 month to 19 months (median 14 months).

Conclusion:

CIN3, often proposed as a surrogate for invasive cervical cancer in cervical screening trials, was detected on initial follow-up (average 13 months) in 2 of 402 women (0.5%) with negative computer-imaged TPPT and positive HPV co-test results. Inclusion of two additional cases of histologically detected CIN2, one VAIN3, and one case of AIS with microinvasive endocervical adenocarcinoma potentially alters the risk profile of this cohort of over 400 cytology negative HPV positive patients. Additional natural history studies are needed on cytology negative HPV positive women routinely screened with modern methods which are now prevalent in the U.S. The potential benefits of routinely combining LBC and HPV co-testing for enhanced detection of endocervical neoplasia deserves special study, given the limited reported success of screening in this area to date.         

Table 1

Cytologic and histologic follow-up results among 402 women with positive HPV tests and negative Pap tests

The previous history of cytological and histological abnormalities among women with negative computer-imaged liquid-based Pap and positive HPV DNA test results

Anisa Kanbour, MD, Shuping Zhao, MD/PhD, Amal Kanbour-Shakir, MD/PhD, Chengquan Zhao Zhao, MD.
Magee Womens Hospital, UPMC, Pittsburgh, PA, USA.

Abstract:
Introduction:
No reports have documented the previous Pap test history for women with cytology negative- HPV positive testing results with methods of liquid-based cytology (LBC), computer-assisted screening, and HPV co-testing. The purpose of this study was to document the abnormal cytology and/or positive HPV testing results among women who tested hrHPV DNA positive along with negative computer-imaged liquid-based Pap co-test results.

Materials and methods:
The computerized hospital records of MWH were searched for patients reported as negative on ThinPrep Imaging System (TIS)-imaged ThinPrep Pap tests (TPPT) who also had positive Hybrid Capture 2 (HC2) hrHPV DNAwithPap co-test results between January 2008 and June 2008. The previous Pap cytology and HPV testing results for these women were analyzed. For women has two or more abnormal Pap tests or biopsy results, only the most abnormal diagnosis was recorded.

Results:
During the study period 202 women had negative TPPT and concurrent positive HC2 hrHPV DNA test results. 71 women showed no previous Pap tests and HPV testing in our files. 131 women had documented previous cytologic and/or HPV testing, or cervical biopsy results. Among these women, 55 (42%) showed normal Pap test and/or negative HPV testing history, 76 (58.0%) had at least one time of abnormal Pap tests or biopsy or Positive HPV testing results previously (Table 1).The interval between an initial abnormal Pap test and current positive HPV and negative TPPT findings ranged from 6 month to 13 years with average of 2.9 years. 71 of 76 cases (93.4%) had the first atypical Pap tests within 5 years.

Conclusion:
Our study indicates that more than half of the women (58%) with negative computer-imaged TPPT and positive HPV co-test results had previous abnormal Pap test or positive HPV testing history. HPV positivity predicts subsequent infection even though the risk of developing high grade dysplasia in women with negative Pap and positive HPV testing is very low. The previous Pap test, cervical biopsy, and HPV testing result should be considered for the risk assessment for cytology negative HPV positive women.

Histological Follow-up Findings in Adolescents with HSIL Cytology Results

Chengquan Zhao, MD,  Pal Kalposi-Novak, MD, R. Mashall Austin, MD, Dr.
Magee Womens Hospital, UPMC, Pittsburgh, PA, USA.

Abstract:
Introduction:
The incidence of cervical intraepithelial lesions is increased in adolescents and reflects the high prevalence of hrHPV infection in this special population. Recent follow-up guidelines emphasize conservative follow-up options. Furthermore, data from cohort studies suggest that regression of both low grade and high grade CIN are quite frequent in very young women. In this study we analyzed histological follow-up data for adolescent women who had HSIL cytology reports. We also assessed the effect of presence or absence of an adequate TZ/ECS in liquid-based Pap tests on the follow-up biopsy diagnoses.
Materials and methods:
The computerized records of a large academic women’s hospital were searched for cases reported as HSIL on TPPT in women age 20 or younger over a 6 year span between January 2002 and December, 2007. Histologic and Pap follow-up results, variations among age groups of adolescent women, and impact of presence or absence of TZ/ECS in Pap test were analyzed. Chi-Square test analysis was performed using SAS 9.1 System.
Results:
During the study period a total of 474 women age 20 or younger had HSIL Pap test results. 335 adolescent women with at least one cervical biopsy with or without endocervical curettage were included in the analysis. The average age was 18.6 years (13-20 years). The average follow-up period was 24 months (0 to 75 months) with a median of 22 months. The overall histologic CIN2/3 and detection rate was 44.2% and 47.8% for CIN1. The average period between the HSIL Pap test and an initial diagnosis of CIN2/3 was 5 months (0-62 months) with a median 2 months. The rates for histologic documentation of CIN in women age 19-20 compared to younger women were not statistically different. Detailed histologic findings are shown in Table 1. No invasive carcinomas or adenocarcinoma in situ cases were identified in this series of adolescents. The percentage of CIN 2/3 diagnosed on histologic follow up was not statistically significantly different when comparing women with and without a TZ/ECS in their preceding HSIL Pap tests (44.5% vs. 38.9%, p=0.642).
Conclusion:
Less than half of adolescent patients with HSIL cytology results had documented histologic CIN2/3 over an average follow-up period of 24 months, and no cases of invasive carcinoma were identified. CIN1 histologic follow-up findings were as common as CIN2/3 findings, likely reflecting both the increased likelihood of HSIL regression in younger women as well as the challenges of precise cytologic and histologic classification. High rates of hrHPV infection, only moderate rates of histologic CIN2/3 following HSIL cytology, and absence of invasive carcinoma all mark the adolescent group as a unique subset of patients deserving further study. Identification of additional biomarkers for HSIL progression would be useful.

 CAP Nov. 2009 abstract (will work on paper soon)

Significance of hrHPV Testing in Women Age 50 Years or Older with LSIL and HSIL Cytology

Chengquan Zhao, MD (zhaoc@upmc.edu); Amer Heider, MD; R Mashall Austin, MD/PhD. Department of Pathology, Magee Womens Hospital, UPMC, Pittsburgh, Pa.

Abstract:
Context:
Older women are considered a “special population”. 2006 ASCCP Guidelines support reflex hrHPV testing as an option for postmenopausal women with LSIL. The data of hrHPV infection and its association to histologic CIN 2/3 in older women with LSIL and HSIL cytology are limited.
Design:
A computer-based search of Copath files of Magee-Womens Hospital, UPMC was carried out between July 1, 2005 and June 30, 2008 to retrieve women aged 50 years or above reported as LSIL or HSIL cytology who also were tested for hrHPV DNA (HC2). The HPV testing result, Pap and histologic follow up results were recorded.
Results:
hrHPV DNA was detected in 25 of 28 (89.3%) women age 50 or older with HSIL cytology, and in 154 of 217 (71.0%) women age 50 or older with LSIL cytology. The average interval between SIL cytology and an initial diagnosis of CIN 2/3 was 2.6 months (0 to 22 months) in women with HSIL cytology, and was 10.4 months (0-35 months) in women with LSIL cytology.The histologic findings are listed in the Table 1.
Histologic CIN Lesions between hrHPV Positive and Negative Testing Results in Older Women with LSIL and HSIL Cytology

TABLE 1

Age-Associated Absence of an EC/TZS in 10 Year Age Intervals (All TBS 2001 Pap Result Categories Included)

EC/ TZS indicates endocervical/ transformation zone sample.

P value was calculated by comparing each group with reference group.

TABLE 2

Comparison of LSIL Prevalence in Women with and without an EC/TZS (10 Year Age Intervals)

LSIL indicates low grade squamous intraepithelial lesion; EC/TZS, endocervical/ transformation zone sample; TPPT, ThinPrep Pap test.

TABLE 3

Comparison of HSIL Prevalence in Women with and without an TZ/ECS (10 Year Age Intervals)

HSIL indicates high grade squamous intraepithelial lesion; EC/TZS, endocervical/ transformation zone sample; TPPT, ThinPrep Pap test.

TABLE 4

hrHPV DNA Detection Rates in 526 Women with LSIL with and without an EC/TZ (10 Year Age Intervals)

LSIL indicates low grade squamous intraepithelial lesion; EC/TZS, endocervical/ transformation zone sample; TPPT, ThinPrep Pap test, hrHPV, high risk human papillomavirus.

*P value was calculated by Fishers’ exact test.

TABLE 5

hrHPV DNA Detection Rates in 133 Women with HSIL with and without TZ/ECS (10 year age intervals)