Background: Liquid-based cervical cytology was developed to improve the diagnostic reliability of Papanicolaou (Pap) smears. Conventional Pap smears can have false-negative and false-positive results because of inadequate sampling and slide preparation, and errors in laboratory detection and interpretation. However, liquid-based cytology rinses cervical cells in preservatives so that blood and other potentially obscuring material can be separated. It also allows for additional testing of the sample, such as for human papillomavirus (HPV). The comparative accuracy of each technique has been studied extensively and has yielded conflicting results; recent systematic reviews reported that there is no convincing evidence to recommend one technique over the other. Siebers and colleagues designed this prospective study to compare the histologic detection rates and positive predictive values of conventional Pap smears and liquid-based cervical cytology.
The Study: This randomized controlled trial involved Dutch women 30 to 60 years of age. The Dutch cervical screening program tests women every five years; sampling is done by family physicians. Family medicine practices associated with the clinical study sites were randomly assigned to perform conventional Pap smears or liquid-based cytology (Thin-Prep) for all participating patients. Informed refusal was offered at each site; otherwise, all women screened at the participating practices were included in the study. Women with abnormal screening results were followed for 18 months, with clinically appropriate histologic testing (with colposcopy and biopsy) or repeat cervical cytology. The primary outcome was the ratio of detection rates (relative sensitivity) of histologically confirmed cervical intraepithelial neoplasia (CIN) or cervical carcinoma between the two screening tests. Secondary outcomes addressed relative specificity, which was determined by the ratios of the positive predictive values of each test for increasing degrees of dysplasia (CIN 1 or low-grade squamous intraepithelial lesions [SIL]; CIN 2 or high-grade SIL).
Family practice personnel were instructed in the proper collection and handling techniques for their assigned screening test, and cytotechnologists and pathologists were trained in processing and evaluating the liquid-based cytology tests. Gynecologists, pathologists, and laboratory technicians involved in the follow-up and review of histology and cytology were all blinded to the type of cytology used and initial test result.
Results: Approximately 84,000 women were evaluated from 246 family practices participating in the trial. In this intention-to-treat analysis, 45,818 liquid-based cytology tests and 38,504 conventional Pap smears were reviewed. Several of the large, urban practices were assigned to the liquid-based cytology group, which accounted for the difference in sample sizes. Among 2,474 persons with abnormal results, 1,918 (77.5 percent) had low-grade lesions and 556 (22.5 percent) had high-grade SIL or above. Abnormalities were detected at statistically similar rates between the two screening tests at all levels.
Conclusion: The authors conclude that conventional Pap smears and liquid-based cytology screening tests have equivalent detection ratios and positive and negative predictive values for detecting CIN or cervical cancer.
editor's note: In an accompanying editorial,1 Schiffman and Solomon acknowledge that the study by Siebers and colleagues helps clarify an important quality issue: that when performed correctly, conventional Pap smears and liquid-based cytology are equivalent for detecting cervical abnormalities. They add that liquid-based cytology often results in fewer unsatisfactory specimens and allows for HPV testing on the same sample. Choice of screening test also depends on the cytology laboratory in which the samples are processed; although liquid-based cytology is more expensive, its ease of use allows laboratories to process slides more quickly and efficiently. For these reasons, liquid-based cytology has virtually replaced conventional Pap smears in the United States. The bigger question, however, is how cytology screening itself will fit into the broader picture of cervical cancer prevention and detection in a rapidly changing environment that includes HPV vaccines and increasingly specific tests for HPV subtypes.—a.c.f.