Introduction and Commentary, Strategic Science Symposium: Human Papillomavirus Testing—Are You Ready for a New Era in Cervical Cancer Screening? Open Access

Cervical cytology screening represents the prototype of a successful cancer screening test. Mortality has decreased by more than 70% in the 5 decades since cytology screening became widely available in the United States. Cervical cancer, once the leading cause of cancer deaths in women in the United States, now ranks 13th,1 although it is still extremely common worldwide.

Despite these successes, a single cervical cytology test is associated with a significant false-negative rate. The Agency for Healthcare Research and Quality (formerly the Agency for Health Care Policy Research) estimated the sensitivity of a single conventional Papanicolaou test to be 51%.2 More recent well-controlled clinical trials with verification of positive and some negative results have found sensitivities of 70% to 80% for conventional Papanicolaou tests and 85% to 95% for liquid-based cytology tests.3,4 Such calculations are based on detection of biopsy-proven, high-grade lesions and cancers when a cytology result of at least atypical squamous cells is considered positive.

Another significant issue is the large number of women with borderline atypical cytology findings who require additional management to determine the presence of a significant lesion. The only way to maintain sufficient sensitivity with cytology screening is to set a low threshold for abnormal results. In the past, women with atypical squamous cells of undetermined significance (ASC-US) on cytology were generally managed by either several repeat cytology examinations or by colposcopic examination. The former represented both an inconvenience and a risk to both patients and clinicians; women had to return 2 or 3 times for cytology to achieve sufficient sensitivity, and some women would inevitably be lost to follow-up. Colposcopy was seen as a safer, more immediate management approach, yet very expensive and inconvenient, causing anxiety to many women.

Human papillomavirus (HPV) DNA testing has emerged as another testing modality for cervical cancer and precursor lesions. Human papillomavirus testing can be used as an adjunct for atypical cytology results, as a primary screening test, as a follow-up test after colposcopy or treatment, and as a quality assurance measure. The College of American Pathologists was aware of the tremendous impact HPV testing will have on future laboratory testing and cervical cancer screening. For this reason, a Strategic Science seminar, entitled “HPV Testing: Are You Ready for a New Era in Cervical Cancer Screening?,” convened September 21–22, 2002, in Rosemont, Ill.

Strategic Science is a new multimodal continuing medical education series integrating science with numerous aspects of laboratory management. This new education model, in which expert speakers address a new or evolving scientific issue from both the clinicians' and pathologists' perspectives, is offered in 3 continuing medical education–generating formats. These include an online preconference, a day-and-a-half live symposium, and peer-reviewed publications. Strategic Science is designed for the practicing pathologist and comprehensively addresses pertinent aspects of new technology assessment, practice guidelines, quality assurance, regulatory compliance, risk and liability, billing and coding, cost analysis, consultation, utilization and information management, and results reporting. This issue of Archives of Pathology & Laboratory Medicine contains 12 additional papers from this Strategic Science seminar on HPV testing. Our summary conclusion is that this conference demonstrated that HPV testing has the potential to lead to a major paradigm shift in cervical cancer testing methods.

In these proceedings, the role of HPV in the epidemiology and pathogenesis of cervical cancer is addressed by Mark Schiffman, MD, and Philip E. Castle, PhD,5 and by Mark Stoler, MD.6 It is now recognized that HPV infection always precedes cervical cancer and is a necessary but insufficient agent for the development of cervical cancer. Many women become infected with HPV, but only a fraction of them develop high-grade squamous intraepithelial lesions, and even fewer progress to cancer. Other cofactors include smoking, oral contraceptive use, multiparity, and possibly inflammation. The current impact of insufficient screening worldwide translates into a major risk factor, as the majority of cancers occur in women with no screening in the previous 5 years. Those experiencing persistent infection with oncogenic or high-risk types of HPV are highly likely to progress to a significant high-grade lesion or cancer. Proteins encoded by HPV genes E6 and E7 are known to interact with the tumor suppressor genes p53 and pRb, causing cell transformation. Trials for HPV vaccines based on viruslike particles are underway, but it will likely be years until these vaccines influence cancer mortality.

The basis of HPV testing methods are discussed by Dr Stoler6 and Roger Hubbard, PhD,7 and David Bolick, MD,8 discusses business aspects of the most commonly used methods. Important considerations in any laboratory test include biologic correlation, clinical validity, and utility; analytic sensitivity, specificity, accuracy, and precision; and technical feasibility, costs, validation, and quality control/quality assurance mechanisms. Early testing methods included nonamplified techniques, such as Southern blot and dot-blot. In situ hybridization is an example of a nonamplified method that is currently applicable for both tissue and cytology specimens, and that can be useful for cytologic/histologic correlation. Hybrid capture and polymerase chain reaction (PCR) methods represent sensitive amplified methods currently in use. Surrogate methods, such as immunohistochemical assessment of p16 are also being investigated.

The HPV Hybrid Capture (HC) assay is the only HPV test with Food and Drug Administration (FDA) approval8 at this time, and it has the most clinical experience, as many of the large investigational trials, including the ASC-US/Low-Grade Squamous Intraepithelial Lesion (LSIL) Triage Study (ALTS) used this method. The current HC2 assay detects 13 oncogenic high-risk types, and a successor HC3 assay is in development. Although HC2 is very sensitive, specificity remains a concern. Young women and those with LSILs are frequently positive for HPV, and there are also borderline positive results that need to be repeated. This test is labor intensive, requiring skilled testing personnel. Reimbursement by third-party payers is marginal in some regions of the country and may require a significant volume to achieve a return of costs.

Ventana has developed an automated in situ hybridization method for use on both cytology and histology samples. While this system provides uniform assay specifications and probes, there are few published data available on efficacy and test performance. Laboratory costs are slightly higher than for HC, but immunocytochemistry Current Procedural Terminology codes provide higher Medicare payments for the global technical and professional charges. The economics of testing may be more favorable for smaller volumes, but a pathologist must review all slide results.

Polymerase chain reaction represents an extremely versatile technique, but is currently a “home-brew” assay for HPV. Consensus primers allow detection of a panel of HPV types, but individual genotypes can also be assayed by a variety of methods. Real-time PCR can be used to estimate viral load, although this does not have known prognostic value at present, and many future testing methods will likely incorporate PCR methods. Specially trained personnel and equipment are needed, but per test disposable costs are generally lower, allowing fewer tests to break even. Coding is identical as for HC.

Results from ALTS provide an abundance of data to justify the use of HPV testing as a triage for ASC-US and as a follow-up test.9 The scientific director for ALTS, Mark Schiffman, MD, states that HPV testing is at least as sensitive as immediate colposcopy in detection of biopsy-proven, high-grade lesions, with about 55% of women with ASC-US referred for colposcopy. In contrast, 2 repeat cytology tests are needed to achieve equivalent sensitivity, with a decrement in specificity. Human papillomavirus testing is an even more efficient triage method in women older than 29 years; however, it is not an efficient triage method for LSIL, as most patients have positive results. Human papillomavirus–positive ASC-US and LSIL can be followed similarly after colposcopy with 12-month HPV testing, as both have a 26% to 28% cumulative risk of developing high-grade lesions at 2 years.

J. Thomas Cox, MD, describes how HPV testing is used according to the recently released Consensus Guidelines for the Management of Women With Cervical Cytological Abnormalities, developed under the leadership of the American Society for Colposcopy and Cervical Pathology (ASCCP). Human papillomavirus testing is 1 of 3 options for managing women with ASC-US, but it is the preferred management when liquid cytology tests are used.10 Only a high-risk HPV test panel should be used for patient management. In the ASCCP algorithm, women with positive high-risk HPV tests proceed to colposcopy, and those with negative tests return for a repeat cytology at 12 months. A 12-month HPV test is also useful in follow-up of LSIL and other lesions after colposcopy. Currently insufficient data exist to rely on routine HPV triage for cervical cytology results reported as atypical endocervical cells.

Human papillomavirus is also promising as a primary screening test, as Attila Lörincz, PhD, and Ralph Richart, MD, report.11 Human papillomavirus sensitivity exceeds cytology sensitivity in all studies, although specificity is lower, especially in women younger than 30 years. The combined cytology and HPV test is nearly 100% sensitive for detection of high-grade lesions in some studies, and a double-negative test has nearly 100% negative predictive value. Human papillomavirus testing recently received FDA approval as a primary screening test in women 30 years and older, and this approval may allow lengthening of screening intervals in mature women.

According to the 2001 Bethesda System, HPV tests should be reported as positive or negative, and the test method should be provided in the report. Stephen Raab, MD,12 Bethesda Ancillary Testing Forum moderator, discusses various reporting options in his article, including probabilistic and interpretive reporting. An integrated cytology/HPV report is ideal, but laboratories that refer out HPV tests may not find this practical.

R. Marshall Austin, MD, PhD,13 discusses risk management and safety issues of cervical screening in his article. The public has high expectations for cervical cancer screening that far exceed the reality of Papanicolaou test performance. In a medicolegal context, it appears that HPV tests may be more reproducible with no durable evidence that would remain for biased retrospective interpretation. In contrast, acceptable cytology sensitivity is dependent on a poorly reproducible atypical interpretation threshold. A clinical approach with immediate reflex HPV triage for ASC-US could avoid delays and loss of patients to follow-up. If initial data are born out in larger studies, HPV screening may also prove to be useful for adenocarcinoma detection, an area where the efficacy of cytology screening has proven disappointing.

Jacqueline Seabrook, BSc, and Roger Hubbard, PhD,14 and Richard Lozano, MD,15 discuss laboratory implementation, quality assurance, and regulatory aspects of HPV testing. As a high-complexity test, validation studies are required for non–FDA-cleared and modified HPV assay methods. It is clear that in anticipation of HPV testing, laboratories will need to be comprehensive in addressing new specimen collection methods, specimen adequacy, and storage requirements, as they may relate to false-positive and false-negative results. New testing comes with additional requirements for updated procedure manuals, accreditation compliance, and proficiency testing. Pathologists will also be faced with a significant role to educate and train their own staff, clinicians, and other pathologists about HPV testing. Issues of reflex testing and documentation of clinicians' orders, even if standing orders for ASC-US triage, must be addressed. Laboratories should periodically monitor the ASC-US rate to guard against overuse of this interpretive category; other quality improvement tools include monitors of HPV-positive ASC-US cases and correlation of HPV results with follow-up biopsies. Many factors may influence a laboratory's decision to consider whether HPV testing should be performed in house or sent to a reference laboratory. These factors would include consideration of workflow, volumes, regional payments for specific Current Procedural Terminology codes, payer mix, local clinical practices, and expectations of clinicians.8 

Patient and public health education are major issues that were discussed but not resolved at this seminar. Many women have little knowledge of the role of HPV in cervical carcinogenesis, as patient advocate Nancy Roach emphasized.16 Educational materials should provide a simple, clear message with concrete information that may require cultural tailoring. Human papillomavirus is considered a sexually transmitted disease, so it follows that widespread use of HPV tests could lead to anxiety and turmoil among women and their families. What information should women receive concerning HPV tests, and how can we de-emphasize the sexually transmitted disease connection? How do we educate patients and providers on the significance of a positive HPV result when examination and cytology are negative, and what is the proper follow-up for such women? While pathologists and laboratories may not directly deal with patient questions, we do have a responsibility for educating providers on the significance of HPV test results. Pathology can and should play a major advocacy role by provision of educational materials and by joint efforts with other medical organizations.