The science behind new screening guidelines for cervical cancer

Poster of Jane Kim

August 22, 2018 – New guidelines for cervical cancer screening, based in part on research by Harvard Chan School’s Jane Kim, were issued by a national task force on August 21. Kim, professor of health decision science, describes the key role that mathematical modeling played in updating the guidelines.

Q: What prompted the updating of guidelines?

A: The previous cervical cancer screening guidelines were issued in 2012 by the U.S. Preventive Services Task Force, an independent body of experts. The guidelines recommended that women get Pap tests—screening procedures that test for precancerous or cancerous cells on the cervix—every three years from ages 21 to 65 years. For women who want to lengthen the screening interval from every three to every five years, the guidelines included a second option: switching at age 30 to co-testing, which comprises a Pap test along with a test for human papillomavirus (HPV), a sexually transmitted infection responsible for most cervical cancers.

Cervical cancer screening guidelines, along with guidelines for other types of cancer screenings and preventive services, are periodically updated by the task force. Recently there’s been new evidence suggesting that it’s really the HPV test that provides the biggest benefit in terms of finding cervical precancers and cancers. The question for the task force was: Is there enough evidence to recommend that the HPV test be administered alone? If so, how often should the test be done, and at what age should women switch from Pap testing to HPV testing?

The task force commissioned us to analyze various screening scenarios using mathematical modeling. This type of model integrates information on how cervical cancers typically progress and the impact of interventions, based on empirical evidence from clinical trials and observational studies. We looked at 19 different screening options, such as Pap testing alone, HPV testing alone, and co-testing, and we considered different starting ages and different screening intervals. We looked at which options would provide the biggest health benefit to women— such as fewer cervical cancer-related deaths—as well as which would cause the most harm, such as additional and perhaps unnecessary invasive testing. We used colposcopies, diagnostic tests that are typically recommended if Pap tests or HPV tests turn up something abnormal and which may cause discomfort or bleeding, as a measure of harm in our modeling.

Q: What did the modeling show?

A: Compared to the 2012 recommendations, we found that there were strategies that could lead to near equivalent or even better health benefits for women with HPV testing alone. One scenario that gives similar benefits at a lower rate of harm is initiating Pap testing every three years starting at age 21 years and switching to HPV testing alone every five years at age 30 years. Under this scenario, we estimated a cervical cancer-related death rate of 0.29 per 1,000 women over the course of a lifetime of screenings. Under the 2012 guidelines, the cervical cancer death rate with only the Pap test every three years is 0.76 per 1,000 women, and with co-testing—the Pap test combined with the HPV test every five years beginning at age 30—is 0.30 per 1,000 women.

There were other screening scenarios in which the estimated cervical cancer death rate over the course of a woman’s life was slightly lower than 0.29 per 1,000, but those scenarios also involved higher rates of colposcopies. At a certain point, if you keep stepping up how much testing you do, you get diminishing returns—more harm for only slight increases in health benefits.

Q: With the changing of the guidelines, do you think there will be pushback from some people who worry that cutting back on the Pap test, which has long been the ‘gold standard’ for cervical cancer screening, will result in some cancers being missed?

A: The recommendations were posted in draft form last year on the task force’s website and included strategies involving Pap testing alone and HPV testing alone. During a subsequent public comment period there were comments from women, patient advocacy groups, and providers who were concerned about missing cancers with HPV testing alone. Some expressed support for using only the HPV test in women over age 30 but thought it should be done more frequently than the drafted recommendations. Others were concerned because, at the moment, there is only one FDA-approved HPV test available on the market, possibly resulting in reduced access to cervical screening. But the final task force recommendations are quite inclusive, with strategies involving Pap testing alone, HPV testing alone, as well as co-testing.

The bottom line is that all of these various screening strategies provide a huge benefit. If we didn’t do any screening, the rate of cervical cancer deaths would be over 8 per 1,000 women over a lifetime of screening—much higher than any of the screening strategies we looked at. We are lucky to have these amazing screening technologies and great ways to use them. The question is how to use them in the most accessible, efficient, and cost-effective way.

Karen Feldscher