Invasive cervical cancer remains an important public health problem as a leading cause of cancer-related death in women in many developing countries ( 1 ) ( http://globocan.iarc.fr/ ). Despite the promise of a highly efficacious prophylactic human papillomavirus (HPV) vaccine ( 2–4 ), HPV vaccination will not prevent approximately 30% of invasive cervical cancer nor does it treat preexisting infections and precancerous lesions, and it requires cold chain delivery and is expensive to administer. Even if vaccines were made available at a greatly reduced cost and widespread vaccination could be rapidly implemented globally (which is not possible in the near-to-mid term), the primary prevention of invasive cervical cancer by prophylactic HPV vaccination would not be realized for another 30 or more years, during which time an estimated 20 million additional cervical cancers will occur in less developed countries ( 1 ) ( http://globocan.iarc.fr/ ). Therefore, secondary prevention, that is, screening and treatment of precancerous lesions, will be needed for a long time to come.
In this issue of the Journal, Denny et al. ( 5 ) present long-term follow-up results of a randomized controlled trial to compare carcinogenic HPV DNA testing with visual inspection of the cervix after acetic acid staining (VIA) as the primary screen for precancerous cervical lesions in a screen-and-treat strategy that eliminates the need for confirmation by biopsy. This landmark study was conducted in a low-resource setting where such strategies are needed to prevent the majority of invasive cervical cancer. The authors made the following observations: 1) HPV testing with immediate cryotherapy of HPV DNA–positive women prevented 72.5% of cervical intraepithelial neoplasia grade 2 or worse (CIN2+) and 77.4% CIN grade 3 or worse (CIN3+) over 36 months, more than the percentages of CIN2+ and CIN3+ prevented by the VIA-based screen-and-treat approach (31.0% and 38.1%, respectively); 2) The sensitivity for 3-year cumulative CIN2+ was statistically significantly greater for HPV testing vs VIA (90% vs 53%, respectively), whereas the specificity was virtually identical (83% vs 78%, respectively); 3) Cryotherapy was 75%–77% effective at ablating CIN2+ lesions; and 4) Cryotherapy of HPV-positive women non-statistically significantly increased the incidence of HIV infection. Importantly, Denny et al. have demonstrated that a simplified screen-and-treat prevention program that is based on HPV DNA screening, which eliminates costly and time-consuming colposcopically directed biopsy for diagnosis, can effectively and rapidly reduce precancerous lesions in a low-resource setting.
The superior performance of HPV DNA testing compared with VIA in a screen-and-treat program is predicted by studies ( 6–13 ) that have shown that HPV DNA screening is more sensitive for detection of precancerous lesions and early-stage cancers compared with VIA or cytology. A highly sensitive test, as demonstrated in this study ( 5 ), can detect precancerous lesions sooner, thus allowing for early treatment. Therefore, a single round of HPV DNA screening can rapidly reduce the incidence of invasive cervical cancer and related mortality within a few years ( 9 , 11 , 12 ). Also important, a negative HPV DNA test performed at the correct age past the peak of new infections among young women provides 5–10 years of reassurance against precancerous lesions ( 10 , 14 ) and against invasive cervical cancer and related mortality ( 11 ). Because a single HPV DNA screen effectively distinguishes between the risks for precancerous lesions and cancer for the subsequent years, resources for cervical cancer prevention can be focused on the fraction of the population at risk. Broad population coverage with one or two lifetime screens ( 15 ), using low-cost HPV DNA testing ( 16 ) with immediate treatment by cryotherapy for those with positive findings, could substantially reduce the burden of cervical cancer in a low-resource setting.
Of concern, however, was the observed, albeit non-statistically significant, 1.75-fold increase in the incidence of HIV infection in the HPV DNA testing arm compared with the VIA and control arms ( 5 ). As the authors note, because of small numbers and other limitations, it is not clear whether this difference is meaningful or a chance finding. If real, the increase in HIV infection in populations from HIV-endemic regions may offset the health benefits of cervical cancer prevention. An important secondary analysis from this trial will be to compare the risk of HIV infection in HPV DNA–positive vs HPV DNA–negative women in the VIA arm to see whether the effect is HPV-related rather than a generic (but possible) association with treatment. One possible explanation is that cryotherapy in women with HPV infections triggered an increase in the clearance of HPV that is biologically comparable to natural HPV clearance, which has been recently linked to an increase in HIV acquisition ( 17–20 ). Further investigation of this possible serious adverse consequence is warranted.
Denny et al. estimated that approximately 75% of CIN2+ lesions were prevented by cryotherapy. However, this study used nitrous oxide gas as a cryogen, which is difficult to obtain in developing countries and approximately 5–10 times more costly than carbon dioxide, a possible alternative cryogen that is more readily available. Carbon dioxide needs to be validated immediately as a safe and effective alternative cryogen to nitrous oxide for cryotherapy to facilitate the roll out of screen-and-treat programs. Treatment failures (including one cancer) were caused by either technical failures of cryotherapy or incorrect visual assessment before enrollment. Conversely, it is likely that some women were incorrectly excluded before randomization, when cryotherapy would have effectively treated their lesion. All visual assessments of the cervix, even colposcopy, are prone to errors in determining the size and location of lesions ( 21 ). Relying on VIA for triage of which HPV DNA–positive women can be treated will be useful but not perfect. In countries that lack the capacity for surgical treatment or cancer management, a more extreme and controversial approach might be to treat all HPV DNA–positive women without subsequent visual inspection to ensure the rapid treatment of all “treatable” lesions.
The concern about overtreating women without prevalent CIN2+ in an HPV-based screen-and-treat program has several important consequences regarding the acceptability of such a program but should be put into context. Implementing cancer screening means accepting some overtreatment. In many developed countries, the frequent use of imperfect screening tests and subsequent overtreatment is a routine and accepted trade-off to prevent invasive cervical cancer. In the United States, women repeatedly undergo cervical cytology throughout their lifetimes to achieve high programmatic sensitivity. Although no numbers are readily available, it seems probable that repeated screening has led to a large fraction of women undergoing colposcopy and diagnostic and treatment procedures, many of which were likely unnecessary. Many precancerous lesions detected and treated are CIN2, an equivocal precancerous diagnosis that often regresses without therapy because it represents an admixture of CIN3 ( 22–24 ) and acute HPV infections that will never become invasive cervical cancer. Finally, restriction of screen-and-treat programs to the proper age range would decrease overtreatment. HPV DNA testing works best for women aged approximately 30–35 years or older, or women at least 15–20 years older than the median age of sexual initiation in the population. In this age group, cervical cancer is still preventable, and a greater proportion of the detected HPV infections are likely to be persistent and strongly linked to the eventual development of cancer compared with transient infections found in younger women ( 25 ).
With the advent of accurate low-cost HPV DNA testing ( 16 ), we can anticipate offering low-income women a highly sensitive screen once or twice in a lifetime. The exact strategy used will vary. In this South African trial ( 5 ), nurses collected the specimens and treated screen-positive women with cryotherapy after a visual triage that determined indication for cryotherapy. The menu of secondary prevention options, each with trade-offs of effectiveness, costs, and acceptability, will include which test to use (as more become available), which sampling method, and the management of HPV DNA–positive women and their treatment ( Figure 1 ). Substituting pelvic examinations with self-collection of a cervicovaginal specimen [albeit with some loss of sensitivity ( 26 )] might reduce programmatic costs, clinic burden, and visit times while increasing screening coverage. Some countries will demand a high level of diagnostic accuracy before treatment. However, colposcopic evaluation with biopsy diagnosis is limited by the availability of trained colposcopists ( 27 ) and pathologists; more fundamentally, a single colposcopic examination as typically practiced ( 28 ) is less accurate than commonly realized for determining if treatment is warranted. When advanced diagnostic services are in short supply, it may be easier to establish and maintain a reliable treatment threshold based on an objective biomarker (eg, p16 INK4a immunohistochemical staining) instead of a morphologic assessment ( 29 ).
More than a quarter million women continue to die each year from cervical cancer; we cannot justify waiting for HPV vaccines nor insist on replicating unsustainable first-world screening strategies. As demonstrated in the South African trial ( 5 ), a simplified program of HPV screen and treat could effectively reduce the population risk of cervical cancer, including women already infected with HIV ( 30 ). With low-cost accurate HPV screening tests coming online, cervical cancer prevention is becoming more effective, affordable, and feasible for low-resource settings. Future research should focus on the development of low-cost, topical therapeutics to simplify treatment. Nevertheless, we are now poised to implement large-scale prevention programs to reduce the unequal burden of invasive cervical cancer with the prevention tools that are now available to us. All that missing is the political will and the monetary investment to do so.
J.C.G. and P.E.C. were supported (in part) by the Intramural Research Program of the National Institutes of Health, National Cancer Institute.