© 2005 Oxford University Press
CORRESPONDENCE |
RESPONSE: Re: A Study of the Impact of Adding HPV Types to Cervical Cancer Screening and Triage Tests
Affiliations of authors: National Cancer Institute, National Institutes of Health, U.S. Departments of Health and Human Service, Bethesda, MD (MS, PEC, DS); Department of Pathology, University of Virginia, Charlottesville, VA (MS); Departments of Molecular Genetics and Microbiology and Obstetrics and Gynecology, University of New Mexico, Albuquerque, NM (CMW)
Correspondence to: Mark Schiffman, MD, MPH, Hormonal and Reproductive Epidemiology Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD (e-mail: schiffmm{at}exchange.nih.gov).
The use of human papillomavirus (HPV) DNA testing for triage of atypical squamous cells (ASCs or equivocal cytologic specimens) is already widespread, and its promise for general screening is gaining acceptance (1). The value of impending widespread HPV testing will depend on how thoughtfully it is used.
As supported by Franceschi and Clifford's statistically powerful contribution, epidemiologists are reaching agreement concerning which HPV genotypes should be included in diagnostic kits. There is little residual controversy over which types convey highest, intermediate, low, or virtually no risk of cancer. Also, as concluded recently by an IARC expert panel (2), not all potentially carcinogenic types of HPV automatically merit inclusion in test kits because the slight gains in sensitivity from the marginally oncogenic types are offset by substantial numbers of unnecessary referrals (3). Formal cost–benefit analyses considering societal preferences must guide the choices of which HPV types to target in cancer prevention.
Moving forward, it is worth considering which other HPV diagnostic issues are resolved, which answers are emerging, and which questions remain.
It would clearly be desirable to have additional Food and Drug Administration (FDA)–approved HPV tests. However, new diagnostic assays must be validated using data regarding prediction of risk of cancer and cervical intraepithelial neoplasia grade 3 (CIN3) from large, representative study populations. In addition to targeting the correct genotypes, HPV tests must have clinically validated viral load cut points. The detection of infections at very low viral loads substantially decreases the predictive value of a positive test while providing only a minimal increase in reassurance against cancer risk (4).
Test performance cannot be predicted by laboratory experiments or small demonstrations (5). For example, without performance data, it would have been impossible to predict that the prototype polymerase chain reaction assay that we analyzed for its ability to triage ASCs (3) would be slightly less sensitive (i.e., detecting fewer CIN3 or cancer) and more specific than the FDA-approved assay (Table 1), not the opposite as would have been expected from laboratory experiments. Any claims about new HPV assays should be based on published or openly available data; in particular, laboratories offering "home-brew" tests must meet this standard for their claims to be regarded as credible given the complexity of development of HPV diagnostics (5).
|
There is an emerging consensus that the next generation of HPV tests should provide some degree of typing (rather than collective detection of pooled types) to permit monitoring of type-specific HPV persistence, the necessary risk factor for cervical cancer. At a minimum, it will probably be important to distinguish HPV16 and HPV18 from the remaining carcinogenic types to permit more aggressive management commensurate with the higher risk of cancer that these types confer (6).
Important unresolved diagnostics and management issues center first on how to define viral persistence. How long should a woman (
30 years old) with oncogenic HPV infection and normal cytology be followed before referral to colposcopy (7)? If colposcopy shows no CIN2, how long should a woman who has a persistent oncogenic HPV infection be followed without treatment?
More generally, we need practical protocols for combining HPV detection, cytology, and colposcopy given that each has clear deficiencies for screening and patient management. These protocols must be adapted to low-resource settings in the United States and internationally, where cervical cancer is a major problem.
Finally, a future challenge is to develop assays that permit better prediction of viral persistence and cancer risk than HPV DNA detection alone. This would revolutionize the already dynamic HPV diagnostic field.
REFERENCES
(1) IARC Handbooks of Cancer Prevention. Cervical Cancer Screening. Lyon (France): IARC Press, 2005.
(2) Cogliano V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, et al. Carcinogenicity of human papillomaviruses. Lancet Oncol 2005;6:204.[CrossRef][ISI][Medline]
(3) Schiffman M, Khan MJ, Solomon D, Herrero R, Wacholder S, Hildesheim A, et al. A study of the impact of adding HPV types to cervical cancer screening and triage tests. J Natl Cancer Inst 2005;97:147–150.
(4) Schiffman M, Herrero R, Hildesheim A, Sherman ME, Bratti M, Wacholder S, et al. HPV DNA testing in cervical cancer screening: results from women in a high-risk province of Costa Rica. JAMA 2000;283:87–93.
(5) Stoler M. HPV testing in cervical cytology practice: it's all about choice. Acta Cytol 2005;49:117–119.[ISI][Medline]
(6) Castle PE, Solomon D, Schiffman M, Wheeler CM. HPV16 infection and the two-year absolute risk of cervical precancer in women with equivocal or mild cytologic abnormalities. J Natl Cancer Inst. In press.
(7) Wright TC Jr., Schiffman M. Adding a test for human papillomavirus DNA to cervical-cancer screening. N Engl J Med 2003;348:489–490.
Related Correspondence
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2005 97: 938-939.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||