PAP/NGC Program Review
Emily E. Volk, MD
The publication of the Bethesda 2001 guidelines
for cervicovaginal cytology recommended changes in cytology categories and terminology. Some of the more striking changes were the modification of the atypical squamous cell category (ASC) and atypical glandular cell category (AGC) to delete the modifiers "favor reactive" and "favor neoplastic." Other changes included eliminating the intermediate adequacy category "satisfactory but limited" and replacing it with more objective criteria for unsatisfactory specimens. Concurrent with the publication of the Bethesda 2001 guidelines has been the widespread replacement of conventional cervical smear techniques with liquid-based cytology, or LBC, which is increasingly used in conjunction with human papillomavirus triage testing for atypical squamous cells.
The CAP has distributed numerous questionnaires on cervical cytology practice
as part of the Interlaboratory Comparison Program in Cervicovaginal Cytopathology,
or Pap program. The results of these surveys continue to provide important benchmarking
data. Diane Davey, MD, past chair of the CAP Cytopathology Committee, and other
members of the Cytopathology Committee, including current chair Dina Mody, MD,
have analyzed the results of the Pap 2003 supplementary questionnaire. Their
findings are published in the November 2004 issue of the Archives of Pathology
& Laboratory Medicine (Davey DD, et al. Bethesda 2001 implementation and
reporting rates: 2003 practices of participants in the College of American Pathologists
Interlaboratory Comparison Program in Cervicovaginal Cytology. Arch
Pathol Lab Med. 2004;128:1224-1229).
The questionnaire was mailed to 1,751 laboratories that participated in the Pap program in mid-2003. Of the 759 responding laboratories, the vast majority (85.5 percent) had implemented Bethesda 2001 terminology. In addition, most laboratories had eliminated the use of the benign cellular changes category (77.9 percent) and the "satisfactory but limited" adequacy category (73.6 percent). The minimum squamous cellularity criteria had been adopted by 85.3 percent of laboratories. However, only little over half (56 percent) of reporting laboratories had adopted the use of an education note in cervicovaginal cytology reports. The majority (79.6 percent) of laboratories had implemented Bethesda 2001 terminology during 2002.
The questionnaire also included queries about the use of conventional versus LBC, specimen volumes, and abnormal rates. The majority of responding laboratories (66.3 percent) offered both conventional cervicovaginal smears and LBC. A minority of laboratories processed conventional smears alone (24.4 percent) or LBC alone (9.3 percent). Comparison of squamous epithelial abnormality rates revealed higher rates for those laboratories using LBC compared with conventional smears.
The group compared ThinPrep and SurePath (as a single LBC category) with conventional smears; rates for LSIL, HSIL, ASC-US, and ASH were significantly higher for LBC than conventional smears (P<0.01 for all comparisons). The median LSIL rates for LBC were 2.4 percent, compared with 1.4 percent for conventional smears. The median HSIL rates for LBC were 0.5 percent, compared with 0.4 percent for conventional smears. The median ASC-US rates for LBC were 4.1 percent, compared with 3.1 percent for conventional smears. The median ASC-H rates for LBC were 0.2 percent, compared with 0.1 percent for conventional smears. Due to higher LSIL rates, the median ASC/SIL was somewhat lower for LBC (1.3 compared with 1.4 for conventional smears).
When the group compared the 2003 data combining information from all specimens (LBC and conventional smears) with the median reporting rates from the 1996 survey (conventional smears only), they identified a slight increase in LSIL detection, from 1.6 percent in 1996 to 2.1 percent in 2002. In contrast, HSIL detection rates showed little change, with a 0.5 percent median in both surveys. The 1996 median reporting rate for ASC was 4.5 percent. The 2003 survey divided ASC into ASC-US and ASC-H categories, with corresponding median reporting rates of 3.9 percent and 0.2 percent. If these categories are combined into a single ASC category, the median reporting rate is 4.0 percent. The median ASC/SIL ratio was 1.4 in the 2003 data, compared with the 1996 ratio of 2.0. In fact, the 1996 median ratio of 2.0 falls above the 95 percent confidence interval of 1.36-1.56 for the recent data. The median unsatisfactory rate for all specimen types was 0.5 percent in the recent data, identical to 1996 data.
The survey also revealed that adjunctive human papillomavirus testing was offered by 79.9 percent of laboratories. Most (62.2 percent) reported that HPV testing was referred to a reference laboratory and that Digene Hybrid Capture on LBC vial was the most frequently used method (85.9 percent). Interestingly, 47 percent of laboratories that reported performing HPV testing did not know the percentage of their ASC-US cases that test positive for high-risk HPV DNA. Of the laboratories that were aware of the positive rates, most (32 percent) report a rate between 41 percent and 60 percent of ASC-US cases. Because of the relatively small numbers of responding laboratories that were aware of HPV positivity rates, no statistically significant data were available to determine with certainty whether the HPV-positive ASC-US rate was related to a laboratory's ASC/SIL ratio.
This work demonstrates that the laboratory community has widely adopted Bethesda 2001 terminology and HPV DNA testing. More important, it provides new benchmarking data for cervicovaginal cytology that pathologists and cytotechnologists will find invaluable for performance improvement.
Dr. Volk, a member of the CAP Cytopathology Committee, is in the Department
of Pathology and Laboratory Medicine, William Beaumont Hospital, Troy, Mich.
Endoscopic ultrasound-guided FNA
Dina R. Mody, MD
An article on endoscopic ultrasound-guided fine-needle aspiration by Jhala, et al (Endoscopic ultrasound-guided fine needle aspiration biopsy: a powerful tool to obtain samples from small lesions. Cancer Cytopathol. 2004;102: 239- 246), reviews the utility of this procedure in small and large lesions. Small lesions are defined as those less than or equal to 25 mm, and large lesions are those larger than 25 mm. A single endosonographer performed all aspirates, thus eliminating operator bias. The conclusion of the authors was that small and large lesions can be sampled easily with high sensitivity (96 percent), specificity (100 percent), and diagnostic accuracy (96 percent). However, to obtain this level of accuracy, a pathologist has to be present in the endoscopy suite for immediate adequacy evaluation. Based on the cumulative frequency curve of the number of passes needed to achieve specimen adequacy, it was demonstrated that 90 percent of adequate samples were obtained within six passes. Only a marginal increase in adequacy was attained with additional passes. Lesions less than or equal to 25 mm require a mean of 2.5 passes (one to nine range) to obtain adequate samples. Larger lesions require 4.5 passes (range: one to 11 passes). The site of the lesion was also an important factor in determining adequacy. Pancreatic EUS-FNAs consistently yielded lower adequacy rates in comparison with other sites such as liver, biliary tree, or lymph nodes. A mean of 15 DQ stained slides (range of four to 34) were required for pancreatic adequacy assessment.
The excellent editorial by Mary Schwartz (Endoscopic ultrasound-guided fine needle aspiration: time, diagnostic challenges and clinical impact. Cancer Cytopathol. 2004;102:203-206) reviews the subject of EUS-FNAs for different sites and diagnostic pitfalls in evaluation of these aspirations, depending upon site. She points out the time constraints (one to two hours per case) and the low reimbursement levels related to the pathologists being present in the endoscopic suite.
Dr. Mody, chair of the Cytopathology Committee, is professor of pathology, Baylor College of Medicine, Houston, and director of cytopathology at Baylor and The Methodist Hospital, Houston.