The Papanicolaou test (Pap test, Pap smear) is limited in part by characteristics of the cervical cells. The Pap test is used to evaluate cells only at the surface of the cervix, and there is evidence that precancerous cells may not exfoliate normally. In this article, evidence from clinical studies indicating that some lesions do not shed their dysplastic cells is presented. Also described is the PapSure process, which uses the Pap test and a visual test, such as speculoscopy, in cervical screening. Speculoscopy is advantageous because the identification of cervical disease and its precursors is not based on the presence of exfoliated cells. This article presents studies that have attempted to determine whether speculoscopy can be used to compensate for the deficiencies of the Pap test in finding cervical disease. Also discussed are the implications of and treatment algorithms for a 'new' category of cervical disease--the "positive screening test"-- which is characterized (according to the PapSure test) by a positive visual test result and a negative Pap test result.
(MCConsultant 2002;2:10-15)

The Papanicolaou test (Pap test, Pap smear), which has been the standard of cervical cancer screening for the past 50 years, is an in vitro test. It relies on the collection of a representative sample of cells that are prepared on a slide and viewed through a microscope. Reviews of studies in which cervical biopsy was used as the gold standard for determining the presence or absence of cervical disease indicate that the results of the Pap test are often inaccurate.^1,2 This research also suggests that the low sensitivity of the Pap test is minimally due to slide-reading error. Many erroneous results from Pap testing are caused by sampling errors that occur when cervical cells are collected or by the characteristics of the cervical cells themselves.

The Pap test requires the collection of cervical cells. The collection device (a wooden or plastic spatula, a small brush, or a cotton swab) can only collect cells that are available on the surface of the cervix; lesions below that surface cannot be sampled. In addition, dysplastic or precancerous cells may not exfoliate normally; as a result, precancerous lesions are not detected by the Pap test.

New technologies have created in vivo procedures for finding cervical cancer and its precursors. Those procedures do not require the collection of exfoliated cervical cells, but instead enable the clinician to look directly at the cervix; they include the measurement of electrical conductivity; spectrophotometry, or the study of the absorption and reflection of light; and direct visualization, which is also based on light absorption and reflection. PapSure (Watson Pharmaceuticals, Atlanta, GA), a direct visual procedure for use in detecting cervical cancer and its precursors, uses a chemiluminescent light source of specific wavelengths in the evaluation of cervical cells.

In vivo tests based on the absorption and reflection of light are useful because the cytoplasm and the nucleus of cells react differently to light. Light is absorbed by and then passes through cytoplasm. Nuclear proteins, which are biochemically and structurally different from those in cytoplasm, reflect light. As a cell mutates from abnormal to precancerous to cancerous, its nucleus becomes larger and larger, and the nuclear-to-cytoplasmic ratio increases (Figure 1). Thus, dysplastic cells reflect light and appear white, and normal cells, which have more cytoplasm than dysplastic cells, appear pink.

Figure 1. The Role of Visual Tests in Identifying Cervical Disease

The PapSure Test

PapSure is the combination of a traditional Pap test and a visual examination of the cervix via speculoscopy. The US Food and Drug Administration has approved PapSure for sale and supports the claim that it improves the Pap test detection rate of neoplastic cervical abnormalities such as dysplasia or squamous intraepithelial lesions from 40% (the Pap smear alone) to 90% (the PapSure test).

The PapSure process begins with the activation of a low-energy light source (Speculite [Trylon Corporation, Torrance, CA]), which is necessary to perform a speculoscopic examination. That light source is activated and attached by adhesive to the upper blade of the speculum. The speculum is then inserted into the cervical canal, and a Pap test is performed. The cervix is then washed with 5% acetic acid. After 1 minute, the cervix is examined with 4-6 power magnification optic (the focal length of this optic should be between 16 and 20 inches). Dysplastic cells reflect light and appear as acetowhite lesions with defined edges, but normal epithelial cells absorb light and appear dark bluish in color from the blue-white coloration of the Speculite.

Shedding and Nonshedding Lesions

The accuracy of the Pap test depends on the exfoliation of surface epithelial cells and the collection of those cells. In the normal cervix, cells form at the basement membrane and are held together by adhesion molecules throughout the basal layer. As those cells mature and move upward toward the surface of the epithelium, the gene that directs the production of adhesion molecules is downregulated. As a result, cells near the surface of the normal squamous epithelium are not held together by strong adhesion molecules and are exfoliated. Those exfoliated cells are collected for the Pap test.

The Pap test is based on the principle that dysplastic or diseased tissue functions like normal cervical cells. However, biochemical literature indicates that in many cases, dysplastic cells do not stop producing adhesion molecules as they extend upward through the epithelial layer.^3-5 The dysplastic cells are held together even at the surface and are not shed. Therefore, it can be hypothesized that those abnormal cells, which are held together by persisting adhesion molecules, may not be available for collection. As a result, the underlying dysplasia (the preneoplastic process) remains unidentified.

A recent study attempted to determine whether those nonshedding lesions limit the ability of the Pap test to detect the precursors of cervical cancer.⁶ Subjects in that study consisted of a group of women (n = 21) with biopsy-proven dysplasia who had visible exocervical lesions diagnosed via colposcopy. Those subjects had also undergone Pap testing, the results of which were mixed (some results were abnormal and others, normal). The researchers wanted to determine two factors: 1) whether the women who had abnormal (positive) Pap test results exhibited shedding lesions with dysplastic cells that were not strongly attached to adjacent cells and were therefore exfoliated, and 2) whether the women who had normal (negative) Pap test results had nonshedding lesions with adhesion molecules that were distributed throughout the cervical epithelium (eg, no dysplastic cells would be available for collection).

Biopsy samples were stained with antibodies to E-cadherin and ß-catenin, which are types of adhesion molecules. The hypothesis was that shedding lesions would exhibit staining only through the basal layer (the bottom half of the cervical epithelium), just as normal cells would have done. Nonshedding lesions, however, would exhibit staining throughout all layers of the epithelium, because adhesion molecules would be present from the basement membrane to the surface. Figure 2 indicates that 80% of women with positive Pap test results had adhesion molecules only at the base of the dysplastic cells (shedding lesions). Ninety-one percent of women with normal Pap test results had adhesion molecules throughout the dysplastic cells (nonshedding lesions).⁶

Half of the women with dysplasia in that study had nonshedding lesions and a normal (negative) Pap test result, which indicates that the ability of the Pap test to detect pathologic cervical conditions is limited by nonshedding lesions.⁶ The presence of nonshedding lesions may explain cases of cervical cancer preceded by multiple normal Pap test results. Research has shown that as lesions become cancerous, adhesion molecules may break down, which enables cancerous cells to exfoliate.

In vivo technologies such as PapSure, which do not depend on the exfoliation of cells, overcome the limitations of the Pap test. Those technologies enable the clinician to evaluate cells below the surface of the epithelium so that early and late-stage cervical disease can be detected. If in vivo technologies are straightforward and simple enough, they can be used for population screening.

Figure 2. Adhesion Molecule Pattern and Pap Test Result

PapSure Data

Data from prospective studies indicate whether speculoscopy (the visual test of PapSure) can be used to compensate for the deficiency of the Pap test in detecting true disease. Prospective studies have used the Pap test, colposcopy (which served as the gold standard), and speculoscopy to identify the cervical disease state in all subjects. Each of those studies compared the sensitivity of the Pap smear alone with that of the PapSure test.

It is important to understand the interpretation of the PapSure screening test by investigators and clinicians. If either the Pap test or the result of speculoscopy is positive, the screening for cervical cancer is considered positive. The term "positive screening test" now refers to those with a negative Pap test and a positive visual test. It is hypothesized that this group of patients has a preponderance of lesions that do not shed abnormal (dysplastic) cells.

A European Multicenter Trial compared the results of the Pap test and PapSure in 3300 women at 32 different clinical sites.⁷ In that study, all participants underwent colposcopy and biopsy, and all investigators were gynecologists. In the detection of low-grade squamous intraepithelial lesions (LGSIL), the sensitivity of the Pap test was 37%, and that of the PapSure test was 91%. The Pap test detected high-grade squamous intraepithelial lesions (HGSIL) in 76% of women who actually had high-grade disease. The PapSure test detected high-grade cervical disease in 100% of women who indeed had high-grade disease (Figure 3).⁷ In that study, a Pap test result indicating atypical squamous cells of undetermined significance (ASCUS) was not considered abnormal.

Figure 3. European Multicenter Trial of the Detection of Cervical Disease*⁷

In the Southern California Trial, nurse practitioners used the Pap test and speculoscopy to screen 689 patients at four clinical sites.⁸ If the result of the Pap test or speculoscopy was positive, gynecologist investigators then performed colposcopy and biopsy, when indicated. That method did create some bias. The number of cases of low-grade and high-grade cervical disease detected by the Pap test and PapSure are shown in Figure 4.⁸

The combined clinical results of published trials on the PapSure test reveal that the Pap test detected 236 total cases of dysplasia and the PapSure test detected 789. Those results were compared with the results of the gold standard, colposcopy and biopsy, which is specific and sensitive. When compared with the results of that gold standard, the Pap test had a sensitivity of 40% and the PapSure test had a sensitivity of 90%.⁹

Figure 4. Efficacy of the Pap Test and the PapSure Test in the Detection of Cervical Disease⁸

Overcoming the Limitations of the Pap Test

Although the result of a Pap test for a nonshedding lesion may be negative, the PapSure test can be used to identify those lesions by means of speculoscopy. In the published studies to date, 6% to 11% of a screening population will have a negative Pap test result and a positive result from speculoscopy. Fifty percent to 70% of women with those results will have some grade of cervical disease.^10-14 Furthermore, and perhaps most importantly, 30% to 40% of all high-grade lesions are found only by the additional use of speculoscopy.^8,15 This is a significant finding, because most clinicians agree that high-grade disease (if not low-grade disease) must be treated.

The Effects of Nonshedding Lesions on the Detection of Human Papillomavirus

Several studies support the shedder-nonshedder hypothesis for human papillomavirus (HPV). One study in particular involved performing an HPV deoxyribonucleic acid (DNA) test on exfoliated cervical epithelial cells from women with biopsy-proven HPV. That test detected HPV in only 60% of the women who had that virus.¹¹

A study in Arkansas indicated that the presence of HPV DNA detectible in the exfoliate correlates strongly with the Pap test status of the individual.⁹ In that study, 105 women with biopsy-proven dysplasia and HPV served as subjects. Those women also had a positive PapSure test result (either an abnormal Pap test, an abnormal visual test, or both). A hybrid-capture test was also performed on the exfoliated material. The Table shows the percent of women with different Pap test results whose exfoliated cells were positive for HPV. Of those with a normal Pap test (nonshedding lesions), only 35% tested positive for HPV. This seems to indicate that nonshedding lesions affect the detection of HPV in exfoliated cells.

Table. Ability of Tests to Detect HPV Deoxyribonucleic Acid in Shedding and Nonshedding Lesions

Does the PapSure Test Detect Atypical Squamous Cells of Undetermined Significance?

Atypical squamous cells of undetermined significance (ASCUS), which is the Bethesda System classification for "questionably" abnormal cells, is a confusing category of Pap test results. Some ASCUS results are merely the result of inflammation, and some indicate a precancerous condition of the cervix. In all the published studies related to PapSure, a total of 500 women with that cytologic diagnosis have been studied. In those studies, approximately two thirds of the patients with a Pap test result indicating ASCUS had a normal speculoscopic examination result. The remaining third of patients with a Pap test result indicating ASCUS had an abnormal speculoscopic examination result.¹⁶ A study of PapSure testing indicated that 97% of all clinical cervical disease occurred in the women with a Pap test result indicating ASCUS and an abnormal speculoscopic examination result.¹⁶ The cervical disease that did occur in the group of patients with a positive ASCUS Pap test result and a normal speculoscopic examination result was always low grade. It should be noted that there is the possibility of cases of ASCUS with a normal speculoscopy with high-grade lesions in the cervical canal, which cannot be seen with speculoscopy nor sampled by the Pap test. Thus, ASCUS cases with normal speculoscopy should not be dismissed too hastily.

Conclusion

The Pap test is reliable for finding precancerous lesions that are shedding. Unfortunately, there is no method of determining which lesions shed dysplastic cells and which do not. As a result, the Pap test will never identify precancerous lesions in some patients. Adding speculoscopic examination to the Pap test provides the greatest sensitivity. The combination of the Pap test and speculoscopy (in the PapSure test) creates an in vivo advantage. The PapSure test can detect disease even if lesions are nonexfoliating dysplastic cells.

The combination of an in vivo technology with an in vitro test allows a single examination by a gynecologist to be more efficient. It brings clinicians closer to finding all women at risk for the development of cervical cancer. The Pap test and other in vitro tests enable the collection of cells from areas of the cervix that cannot be seen, and an in vivo technology that in "real time" enables the almost immediate identification of cervical disease. The combination of in vivo and in vitro procedures will create new algorithms for managing precancerous cervical disease.

The population most affected by new algorithms for managing precancerous cervical disease are women with a negative Pap test result and a positive speculoscopic examination (women with nonshedding lesions). Approximately 9% of the PapSure screening population are in that category. The management algorithms for the total PapSure screening population and for the new category of women with nonshedding lesions are depicted in Figure 5. As a result of the accuracy of the PapSure test, a clinician can monitor patients with potentially precancerous lesions that would be undetected by the Pap test.

Figure 5. Management Algorithms for the PapSure Test ^16,17

True Population Screening: Is It Cost Effective? A Lesson from History

A review of the costs and outcomes associated with a screening test for tuberculosis, which was introduced in the 1970s, might be helpful in determining the cost effectiveness of a new screening test for cervical cancer.

During the 1950s and 1960s, the standard screening test for tuberculosis was an annual chest radiograph. If active tuberculosis was found, the patient would receive drug therapy and would be isolated in a sanatorium. A decade later, scientists suggested that a skin test could be used to identify all those infected with the bacterium that causes tuberculosis. Individuals found to be infected underwent a thorough workup, and those with a negative skin test did not. Only a few of those with a skin test that was positive for tuberculosis had active disease requiring treatment, and the remaining patients, who were considered to be "at risk" for the future development of tuberculosis, were treated expectantly. The idea of preventive treatment, or chemoprophylaxis, was then introduced. Although the drugs used to treat tuberculosis were not without side effects and were not 100% effective in destroying the bacteria that cause the disease, a dramatic change in the incidence, mortality, morbidity, and costs associated with tuberculosis in the United States was predicted. That suggestion met with warnings that too great a cost for treatment would be incurred, because a skin test would identify individuals with active tuberculosis as well as those with only dormant infection. Those with only infection would still receive a confirmatory chest radiograph and prophylactic treatment, both of which would result in additional costs. Clinicians opposed to conducting a skin test screening program predicted that the cost of screening/treating tuberculosis would double with little benefit. At that time, although many individuals had been infected with tuberculosis, the active disease would develop in only an estimated 1% to 3% of those with a positive annual skin test. In many ways, such individuals were viewed as having "low-grade lesions" and were not likely to require intervention.

However, as a result of diagnosing tuberculosis in an earlier predisease stage, several unexpected cost-effective measures were achieved. The incidence of active tuberculosis in the native US population who were not infected with the human immunodeficiency virus decreased to nearly zero during the first 15 years after the skin test for tuberculosis was introduced. All sanatoriums for the treatment of patients with tuberculosis had been closed by the late 1980s. The cost of tuberculosis screening and treatment programs decreased to 20% of the pre-1970 level. Diagnosing tuberculosis at the earliest stage proved to be the most cost-effective approach. Screening with skin tests at intermittent intervals is the accepted cost-effective method of screening for tuberculosis in the United States.

The natural history of cervical cancer progresses from infection with the human papillomavirus (HPV) to precancerous low-grade and high-grade lesions and then to cancer. Because the Pap test is not sensitive in the discovery of low-grade lesions, the current objective of cervical cancer screening with the Pap test is to detect high-grade lesions. This does not necessarily mean, however, that low-grade lesions are not important or that treating low-grade lesions is not cost effective.

From the standpoint of disease progression, the natural history of tuberculosis and that of cervical cancer are similar. The accuracy of the PPD skin test and the PapSure test in detecting earlier stages of disease is significantly greater than that of their predecessors. The PPD skin test has already proven its cost effectiveness in the screening for tuberculosis. The PapSure test has the potential to prove the same in cervical cancer screening. First, though, current opinion about the stage at which cervical cancer should be detected must change. Decisions about the best treatment for low-grade cervical lesions must also be made. Longitudinal comparison studies must be conducted to show that just as the identification of early infection with the bacterium that causes tuberculosis was not inordinately expensive, discovering the earliest stages of cervical dysplasia will save cost and improve the outcome of treatment.

References

Princeton Media Associates
37 Prodelin Way
Millstone Twp, NJ 07726
Phone: (609) 371-1137
FAX: (609) 371-2733

Pap Test	No. of Patients	*Digene HC (Positive)**	Percentage
LGSIL (positive)	20	16	80
Normal or benign	68	24	35
ASCUS (positive) or worse disease	37	20	54
*Digene HC is a human papillomavirus deoxyribonucleic acid test performed on exfoliated cells. All women had biopsy-proven dysplasia and human papillomavirus. LGSIL = Low-grade squamous intraepithelial lesions; ASCUS = atypical squamous cells of undetermined significance.

New Cervical Screening Technology: Clinical Benefits of the PapSure Test

Abstract