Changes in screening guidelines that imply suppression of procedures once recommended are always controversial because of the perception that benefits are being curtailed. Prior to 2012, cervical cancer screening guidelines issued by US-based expert bodies differed in several decision areas, making clinicians essentially cherry-pick among recommendations. To some extent, this approach to screening practices also served to shield clinicians from litigation. It implied starting screening earlier, doing it more frequently, and stopping later in life than necessary. This state of affairs changed in 2012, when the most influential professional groups updated their cervical screening guidelines, and recommendations became essentially unified. All groups recommended that women older than 65 years of age discontinue cervical cancer screening on the basis of evidence that screening benefits in this age group were minor and far outweighed by harms. The guidelines are very specific about the exceptions, which ensure acceptable safety. It is expected that the new guidelines will permit less wasteful cervical screening, while fostering the opportunity to direct resources towards ensuring adequate coverage of high-risk women.
“Will you still need me, will you still feed me, When I'm sixty-four?”—Lennon/McCartney, 1967
The principle of screening in medicine is unique in that it targets apparently healthy individuals to reduce future morbidity and mortality associated with a particular disease. The objective is to identify people at risk early on in the disease process, when treatment is likely to be effective. Inarguably, one of the most successful examples of this strategy is cervical cancer screening. For decades, Papanicolaou (Pap) cytological screening has reduced the incidence of and mortality from cervical cancer in countries with adequate resources to deploy this public health intervention. In the last 2 decades, we have also learned of the causal link between sexually transmissible human papillomavirus (HPV) infection and cervical cancer and have gained a good understanding of HPV epidemiology and the natural history of cervical neoplasia. These advances have stimulated research on effective screening strategies—research that has, in turn, been continuously translated into guidelines by expert bodies for the early detection of precancerous cervical lesions and cancer (1–4). Ever conscious of the central tenet of medical practice, primum non nocere (“First, do no harm”), these expert bodies attempt to define screening algorithms with the best balance of benefits and harms. The ideal strategy should identify those cervical cancer precursors likely to progress to invasive cancer, while avoiding the detection and unnecessary treatment of transient HPV infections and associated lesions that are not destined to become cancerous (1–4).
In 2012, the American Cancer Society, the American Society for Colposcopy and Cervical Pathology, and the American Society for Clinical Pathology provided a unified update of screening recommendations based on systematic evidence reviews by 6 different working groups and a symposium (1). These latest guidelines were developed using a consensus process involving subject matter experts and key stakeholders. Principles of the Grading Recommendations Assessment, Development, and Evaluation (GRADE) system were employed for evaluating the available evidence (1, 5). Working groups were instructed to propose strategies that best served women, while focusing on the benefits and harms of screening (1). Notable changes from previous updates affected recommendations for ages at starting and stopping screening, age-specific screening strategies and follow-up (inclusive of vaccinated women), and considerations regarding HPV testing as a stand-alone primary screening approach.
In this issue of the Journal, Rustagi et al. (6) argue that the latest guidelines (1) fail to adequately consider evidence from observational studies regarding the efficacy of cervical cancer screening among older (>65 years) women. While the guidelines recommend discontinuing screening in this age group only for women with adequate negative prior screening and no history of cervical intraepithelial neoplasia (CIN), grade 2, or worse (CIN2+) lesions within 20 years, Rustagi et al. (6) fear that this recommendation may have been ill-founded. In this commentary, we counter Rustagi et al.'s arguments and lend credence to the stringency with which the new guidelines address the topic of discontinuing screening.
THE FINE BALANCE BETWEEN BENEFITS AND HARMS
Rustagi et al. (6) defend the value of cervical cancer screening in older women by citing studies from the United Kingdom, Sweden, and the United States (7–10). However, in at least one of these studies (8), they seem to confuse evidence for unscreened and inadequately screened women aged >65 years, who are at high risk of disease, with evidence for well-screened women in this age group, who are at very low risk of disease. More importantly, their arguments for screening older women justly warrant consideration of the associated harms of doing so—a paramount point that is essentially unaddressed in their article.
To inform their recommendations, Saslow et al. (1) used the number of colposcopies reported, both overall and relative to the CIN, grade 3, or worse (CIN3+) lesions and cancers detected, as the primary measure of harm in their analysis. To this end, Working Group 4—tasked with evaluating discontinuation of screening—used in its review a modeling study that in itself incorporated parameters from an exhaustive review of epidemiologic and clinical studies. This study demonstrated that among women who underwent cytological screening every 3 years prior to age 65 years, only 1.6 cancer cases and 0.5 cancer deaths per 1,000 women would be averted if screening were to continue until the age of 90 years (1, 11). Continued screening would extend life expectancy by only 1 year per 1,000 women, at the cost of 58 extra false-positive results, 127 extra colposcopies, and 13 extra diagnoses of CIN2/3 that would require treatment (1, 11).
Sawaya et al. (12) similarly quantified the harms associated with screening in a cohort of postmenopausal women who had been previously screened. Screening outcomes were determined for more than 2,500 women with an average age of 67 years who had a normal Pap smear at baseline. After 2 years of follow-up, 1 woman was ultimately diagnosed with CIN, grade 1 (CIN1) or grade 2 (CIN2). However, to identify this woman, physicians had performed over 5,000 Pap smears, 33 colposcopies, 30 cervical/vaginal biopsies, 35 endocervical curettages, 8 endometrial biopsies, 4 dilation-and-curettage procedures, and 9 cone biopsies or loop electrosurgical excision procedures (12, 13).
These 2 studies demonstrate the fact that the potential harms from screening inherently begin with the decision to screen; yet, by definition, the benefits of screening are attained only when lesions are treated via excisional, ablative, or more aggressive surgical procedures such as hysterectomy (13). As women age, however, the risks and complications associated with such procedures increase; cervical transformation zones become smaller and less accessible in older women, and comorbid conditions that would rule out recommended surgical procedures become more common. Additionally, one must consider whether cervical lesions, if detected, would ever have progressed to cancer. A significant proportion of undiagnosed CIN2 may, in fact, represent regressive lesions (14).
Indeed, in women older than age 65 years with adequate and uneventful screening histories, the prevalence of CIN2+ is low, and cervical cancer is rare (1, 2, 15). This prompts concern regarding the resulting low positive predictive values of screening tests and the risks and harms of overscreening and overtreatment that inevitably originate from a high rate of false-positive screening results. However, Bellizzi et al. (16) found that 53% of women aged 75–79 years and 38% of those aged 80 years or older reported receipt of a Pap test within the previous 3 years. While several factors are likely to be contributing to this practice (e.g., patient preference, providers’ desires to maintain good relationships with patients, and/or a lack of awareness regarding guidelines), older women electing to continue screening should be informed that the likelihood of obtaining false-positive results requiring further and invasive testing is far greater than the likelihood of developing cancer (12). System strategies supporting the delivery of screening in primary care (e.g., electronic medical records) could also be used to identify women eligible for cervical screening per guidelines and to implement patient and physician reminders (17).
Although the risks from acting on false-positive screening results are considerable, one must also appreciate that potential harms extend well beyond those mentioned. The psychological effects of cervical screening in women can be substantial but are often unmeasured in clinical trials. These effects include the anxiety associated with having an abnormal Pap test, shame or low self-esteem related to the diagnosis of a sexually transmitted infection, and the perception of an increased risk of developing cancer in the future, which may lead to a greater desire for continued and frequent testing (1, 13).
OTHER EVIDENCE FOR THE DISCONTINUATION OF SCREENING IN WOMEN AGED >65 YEARS
The vast majority of cervical cancers in older women are found among those not previously screened or underscreened (1, 2, 4, 12, 18–26). Moreover, the peak period of cervical cancer risk in unscreened populations is approximately 35–55 years of age (27), a distribution explained by the fact that cervical cancers originate predominantly from HPV infections acquired in late adolescence and early adulthood (28). While sparse evidence suggests that the natural history of HPV infection is unaffected by age at acquisition, the incidence of new infections markedly declines with increasing age. Nonetheless, even if a new infection were to be acquired as a result of an existing or new sexual partner, it would most likely be transient irrespective of the woman's age (29, 30).
The recent guidelines recommend HPV and cytological co-testing as the preferred screening method for women aged 30–65 years (1). The extremely high sensitivity afforded by HPV DNA testing for detection of CIN2+ and this test's ability to target the true origin of cervical carcinogenesis (i.e., infection with an oncogenic HPV type) provides greater assurance of safety following a negative result relative to the paradigm of exclusive Pap test-based screening. The definition endorsed for releasing women from screening is rather exacting (1). Specifically, women aged >65 years with either 3 consecutive negative cytology results or 2 consecutively negative co-tests within the 10 years prior to ceasing screening should not be screened. This recommendation does not apply to high-risk populations of women, including those with a history of cervical cancer, those who were exposed to diethylstilbestrol (DES) in utero, and those who are immune-compromised. Given the slow, natural progression of incident HPV infection to cancer (typically 15–20 years or more), it is improbable that newly acquired infections and precursor lesions beyond the age of 65 years will have sufficient time to progress to cancer during the woman's lifetime, particularly if she was adequately screened prior to reaching that age (1, 28).
THE ETHICAL OBLIGATION OF RE-TARGETING RESOURCES
Ethical debates on the topic of screening harms have increased in recent years (16, 31–33). Many of the arguments focus on eliminating procedures that fail to provide benefit to patients (31–33)—for example, the recent evidence review by the US Preventive Services Task Force of the value of prostate cancer screening (34). In the United States, this debate extends to one of resources. Tests producing too many false-positive results are not only harmful but wasteful (31). Unlike the United Kingdom, the Netherlands, Canada, Australia, and Scandinavian countries, which have organized screening programs targeting all eligible women at no direct financial cost to patients, cervical screening in the United States is exclusively opportunistic and guideline-driven. This strategy leads to inequalities in health care, leaving women of low socioeconomic status who bear a high risk of cervical cancer with few opportunities to benefit from screening. Redirecting resources to programs, therapies, and treatments (or to appropriate at-risk and underserved populations) proven to be of greater benefit could mean expanded medical coverage for all (31). The waste avoidance argument becomes critical when considering that, in the United States alone, the number of adults aged 65 years or older may nearly double by the year 2030 (16, 35). Even small changes in the growth of annual per capita health-care spending can have tremendous implications for the future of Medicare, without compromising quality of care or measurable health outcomes (36).
CONCLUSION: PRIMUM NON NOCERE … IN SE BENEFICIUM EST (“… IN ITSELF IS A BENEFIT”)
This article's opening motto, verses from a Beatles song so dear to Baby Boomers, encapsulates the plea made by Rustagi et al. (6). As they thoughtfully argued, extending the age range for cervical cancer screening to women aged >65 years would probably bring about an incremental reduction in cervical cancer incidence. Nonetheless, the downside of this modest gain should be just as cautiously reviewed. For the first time in the United States, writers of cervical screening guidelines have agreed on all relevant recommendations (1, 3, 4). Focusing future efforts and resources to expand screening coverage to groups of unscreened or underscreened women will be of greatest value for reducing the incidence of and mortality from this disease in women of all ages. As indicated in the new guidelines (1, 3, 4), it is reasonable to believe that the future of cervical cancer screening in high-resource settings will incorporate HPV DNA testing as a stand-alone primary screening approach. This positioning is likely to usher in ever-advancing molecular technologies for use in screening programs, a paradigm shift that should provide women—particularly older women on the cusp of discontinuing screening—greater assurance regarding their future health. Since no science is perfect and our understanding of the natural history of HPV infection and cervical cancer is continuously evolving, some level of uncertainty will always remain regarding population-level recommendations. All we can ask ourselves is whether or not this uncertainty is sufficiently strong to warrant making changes that may, in the end, do more harm than good, while preventing resources from being directed to places where real benefit can be obtained.
Author affiliations: Division of Cancer Epidemiology, Department of Oncology, McGill University, Montreal, Quebec, Canada (Eduardo L. Franco); and Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, Quebec, Canada (Sandra D. Isidean, Eduardo L. Franco).
This work was supported by the Canadian Institutes of Health Research (team grant CRN-83320 to E.L.F. and a doctoral award to S.D.I.).
“When I'm 64”—Written by: John Lennon & Paul McCartney. ©1967 Sony/ATV Music Publishing LLC. All rights administered by Sony/ATV Music Publishing LLC, 8 Music Square West, Nashville, TN 37203. All rights reserved. Used by permission.
Dr. Eduardo L. Franco was a member of Working Group 5 and a member of the Writing Committee for the American Cancer Society/American Society for Colposcopy and Cervical Pathology/American Society for Clinical Pathology cervical cancer screening guidelines.