P12-07Population-based HPV Testing Performance: Potential False-negative HPV Results in a Cervical Cancer Surveillance Population

10. HPV testing
H. Zhou 1, R. Bassett 1, A. Khanna 1, M. Dawlett 1, N. Sneige 1, K. Schmeler 1, A. Milbourne 1, T. Bevers 1, Y. Gong 1, G. Staerkel 1, M. Guo 1.
1MD Anderson Cancer Center (United States)

Background / Objectives

Human papillomavirus (HPV) testing has been recommended for primary cervical cancer screening in the United States. HPV testing performance in different patient populations has not been systemically evaluated.

 


Methods

We retrospectively searched our institution’s database for women aged 30 years and older who underwent HPV/Pap cytology co-testing in our Cancer Prevention Center (CPC, for general screening population) or Gynecology Clinics (GYN, for cancer surveillance) (2007-2014). SurePath Pap tests and Hybrid Capture 2 or Cervista HPV assays were used in both centers. A total of 22,005 cases were analyzed, including 13,951 from CPC (mean age, 54.9 years; 30-91 years) and 8,054 from GYN clinics (mean age, 50.9 years; 30-96 years). HPV testing results were compared between the two populations. The differences were analyzed by the Fisher exact test.


Results

Women with abnormal Pap test results (≥ASC-US) accounted for 6.5% in the CPC and 27.2% in the GYN clinics (Table 1). The distribution of HPV positivity is illustrated in Table 2. The overall HPV-positive rates were significantly different in women tested at the CPC (5.7%) vs. the GYN clinics (14.8%) (P<0.0001). Pap tests with a high-grade squamous intraepithelial lesion (HSIL) or squamous carcinoma result were positive for HPV in 100% (21/21) of cases from CPC and 82.8% (130/157) of cases from GYN clinics (P=0.05). In 25 women from the GYN clinics with HSIL/squamous carcinoma Pap cytology and a negative HPV result, the follow-up biopsy showed dysplasia in 20 cases, including high-grade dysplasia or squamous carcinoma in 14.

Table 1. Pap Cytologic Diagnosis in Women Screened at the Cancer Prevention Center (CPC) or Gynecology Clinics (GYN)

  NILM     (%) AGC (%) ASC-US (%) LSIL (%) ASC-H (%)

HSIL/SqCa (%)

Total    (%)
CPC 13044 (93.5) 55  (0.4) 589     (4.2) 199 (1.4) 43     (0.3)

          21            (0.2)

13951 (100)
GYN 5866 (72.8) 150 (1.9) 1311 (16.3) 463 (5.7) 107  (1.3)

          157             (1.9)

8054 (100)

 

 

 

 

 

 

 

 

Table 2. Positive HPV Test Rates in Populations at the Cancer Prevention Center (CPC) or Gynecology Clinics (GYN)

  NILM AGC ASC-US LSIL ASC-H HSIL/SqCa
CPC HPV + (%) 3.8* 3.6 20.2 67.3 37.2 100**
GYN HPV + (%) 6.8* 9.3 21.9 68.5 39.3 82.8**

* P < 0/0001; ** P = 0.05


Conclusion

HPV testing performance for cancer screening is population dependent. Potentially higher false-negative HPV test results can occur in a cancer surveillance population compared to a cancer screening population. Further study is required to determine the cause of false negative HPV test results in cancer surveillance population.


References