Cytopathology in Focus: Why not call everything ASCUS?

Create PDF

Barbara A. Crothers, DO
August 2018—Below is a question shared on the ASC listserv. My reply to the question follows.

A pathologist colleague who practiced previously as an obstetrician/gynecologist is of the opinion that categorizing the level of abnormality we observe on a Pap test is a waste of time. All the clinician needs to know, he says, is whether the test is normal or abnormal. The Pap test is a screening test, he says correctly, and its only relevance is in pointing out who needs a colposcopy and biopsy.

Being more specific gives the clinician guidance on what to look for on colposcopy, I argue. There is a significant false-negative rate with colposcopy, and providing a more specific diagnosis can help improve the number of successful biopsies, as well as help avoid unnecessary biopsies. The appropriate diagnosis also ensures the patient receives the correct follow-up. With HPV testing, however, my colleague’s argument is compelling. Can you help?

The main problem with the primary HPV testing model (yes or no to colposcopy) is that the gold standard—colposcopy and biopsy—is not gold at all.

Colposcopy’s sensitivity is about 52 percent, and it doesn’t have the high specificity that a Pap test does. That means that patients and pathologists are relying on colposcopists, who have a 50:50 chance of detecting a lesion, to find disease. As we all know from doing cytologic-histologic correlations for more than 40 years, the most common reason for non-correlation is sampling error; 70 percent of errors, in other words, are due to colposcopy error. Most colposcopists have no specific training requirements or certification for the procedure, no required proficiency test, no quality assurance or quality improvement metrics, and, until recently, no standardized terminology for reporting lesions.

Does that mean colposcopy is a total waste of time? That we should rely on HPV tests alone to detect disease and just proceed to LEEP?