Interobserver Variability in HPV Test Results in Cervicovaginal Specimens
Interobserver Variability in HPV Test Results in Cervicovaginal Specimens
We studied interobserver variability in the proportions of human papillomavirus (HPV)-positive results for atypical squamous cells of undetermined significance (ASCUS) and atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion (ASC-H) diagnoses among 5 pathologists from the same laboratory during a 2-year period. These proportions were compared with individual pathologist's ASCUS/squamous intraepithelial lesion (SIL) ratios.
Of 1,299 ASCUS diagnoses, 32.3% had HPV testing; 49.4% were HPV+. Positive findings by individual pathologists ranged from 38% to 67% (P = .057). There was a difference in the proportions of high-risk HPV results for individual pathologists (P < .001). For the pathologist who diagnosed 38% (23/61) of samples as HPV+, the ASCUS/SIL was 0.58; the pathologist who diagnosed 67% (28/42) as HPV+ had a ratio of 1.02. Of the ASC-H diagnoses, 32.9% were tested for HPV; 63% (46/73) were positive. Although the HPV+ proportion by pathologist ranged from 54% to 83%, no significant differences were identified.
Within the same laboratory, interobserver variability exists in the proportions of HPV positivity for ASCUS and ASC-H interpretations.
According to the Bethesda System, atypical squamous cells (ASC) is a term for which the definition has evolved over time. Generally, ASC is used to designate squamous epithelial cells in cervicovaginal cytologic samples that appear abnormal and have features suggestive but not fully diagnostic of squamous cell dysplasia (cervical intraepithelial neoplasia). Currently, the Bethesda System recognizes two levels of ASC: atypical squamous cells of undetermined significance (ASCUS) and atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion (ASC-H). Commonly, ASCUS refers to specimens that approach or resemble low-grade squamous intraepithelial lesions (LSILs), whereas in ASC-H, lesions have cellular features suggestive of high-grade SIL (HSIL). Although morphologic criteria have been promulgated for both categories, the actual application in day-to-day cytologic practice manifests considerable variability.
The central carcinogenic role of human papillomavirus (HPV) in the development of cervical epithelial abnormalities has been recognized for some time. Accordingly, a variety of tests have been created to identify the presence of HPV and the associated cancer risk type in cervicovaginal samples. Although some clinicians use these tests as part of the primary screening package for women, most professionals have used HPV testing in women with a cytologic interpretation of ASCUS.
As manifested by the results of the ASCUS/LSIL Triage Study and other investigations, viral testing has proven useful for determining whether women with such a diagnosis should be referred for colposcopy and biopsy or be followed up in screening programs. This determination is based on the knowledge that HPV, especially high-risk HPV types, is associated with invasive cervical carcinoma and its precursors that, if untreated, are more likely to develop into malignancy at a relatively high rate.
However, the value of HPV DNA in the quality assurance process of cervical cytology is almost totally unknown. Because HPV DNA should be positive in a prominent but as yet undetermined proportion of specimens interpreted as ASCUS and ASC-H, the results of concurrently performed cytologic and HPV DNA testing may provide a measure of validity of such morphologic interpretations by individual pathologists and laboratories. We acknowledge that low-risk HPV types are generally not associated with the development of HSILs and carcinomas, and, thus, testing for them on clinical grounds is usually not indicated. On the other hand, low-risk viruses may result in cytomorphologic features of ASCUS and, to a lesser extent, of ASC-H, which are indistinguishable from the features of high-risk HPV and, thus, detection of low-risk viruses is desirable and valid from a quality assurance perspective. This measure should prove to be more objective than cytologic and histologic interpretations, in which subjectivity and, thus, variability are known to exist.
We studied interobserver variability in the proportions of human papillomavirus (HPV)-positive results for atypical squamous cells of undetermined significance (ASCUS) and atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion (ASC-H) diagnoses among 5 pathologists from the same laboratory during a 2-year period. These proportions were compared with individual pathologist's ASCUS/squamous intraepithelial lesion (SIL) ratios.
Of 1,299 ASCUS diagnoses, 32.3% had HPV testing; 49.4% were HPV+. Positive findings by individual pathologists ranged from 38% to 67% (P = .057). There was a difference in the proportions of high-risk HPV results for individual pathologists (P < .001). For the pathologist who diagnosed 38% (23/61) of samples as HPV+, the ASCUS/SIL was 0.58; the pathologist who diagnosed 67% (28/42) as HPV+ had a ratio of 1.02. Of the ASC-H diagnoses, 32.9% were tested for HPV; 63% (46/73) were positive. Although the HPV+ proportion by pathologist ranged from 54% to 83%, no significant differences were identified.
Within the same laboratory, interobserver variability exists in the proportions of HPV positivity for ASCUS and ASC-H interpretations.
According to the Bethesda System, atypical squamous cells (ASC) is a term for which the definition has evolved over time. Generally, ASC is used to designate squamous epithelial cells in cervicovaginal cytologic samples that appear abnormal and have features suggestive but not fully diagnostic of squamous cell dysplasia (cervical intraepithelial neoplasia). Currently, the Bethesda System recognizes two levels of ASC: atypical squamous cells of undetermined significance (ASCUS) and atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion (ASC-H). Commonly, ASCUS refers to specimens that approach or resemble low-grade squamous intraepithelial lesions (LSILs), whereas in ASC-H, lesions have cellular features suggestive of high-grade SIL (HSIL). Although morphologic criteria have been promulgated for both categories, the actual application in day-to-day cytologic practice manifests considerable variability.
The central carcinogenic role of human papillomavirus (HPV) in the development of cervical epithelial abnormalities has been recognized for some time. Accordingly, a variety of tests have been created to identify the presence of HPV and the associated cancer risk type in cervicovaginal samples. Although some clinicians use these tests as part of the primary screening package for women, most professionals have used HPV testing in women with a cytologic interpretation of ASCUS.
As manifested by the results of the ASCUS/LSIL Triage Study and other investigations, viral testing has proven useful for determining whether women with such a diagnosis should be referred for colposcopy and biopsy or be followed up in screening programs. This determination is based on the knowledge that HPV, especially high-risk HPV types, is associated with invasive cervical carcinoma and its precursors that, if untreated, are more likely to develop into malignancy at a relatively high rate.
However, the value of HPV DNA in the quality assurance process of cervical cytology is almost totally unknown. Because HPV DNA should be positive in a prominent but as yet undetermined proportion of specimens interpreted as ASCUS and ASC-H, the results of concurrently performed cytologic and HPV DNA testing may provide a measure of validity of such morphologic interpretations by individual pathologists and laboratories. We acknowledge that low-risk HPV types are generally not associated with the development of HSILs and carcinomas, and, thus, testing for them on clinical grounds is usually not indicated. On the other hand, low-risk viruses may result in cytomorphologic features of ASCUS and, to a lesser extent, of ASC-H, which are indistinguishable from the features of high-risk HPV and, thus, detection of low-risk viruses is desirable and valid from a quality assurance perspective. This measure should prove to be more objective than cytologic and histologic interpretations, in which subjectivity and, thus, variability are known to exist.