Alternative Screening Could Cut Cervical Cancer Deaths in Poor Nations

In the right hands, a swab of vinegar and a flashlight may detect more cervical cancer around the world than the recommended cytological screening known as a Pap smear. At the right time, a single DNA test for the virus that causes cervical cancer may also outperform repeated Pap smears.

Photo courtesy of Sue Goldie

Screening all women once at age 35 could cut cervical cancer deaths in poor countries (such as Haiti, above) by up to 25 percent, if combined with treatment. A second test at age 40 could cut women’s deaths from the cancer by up to half, says a new modeling study led by Sue Goldie (second from right).

In fact, dime for dime, either test is a better deal for poor countries that want to emulate the success of wealthy countries in sharply reducing cervical cancer rates, says a cost-effectiveness analysis of the three screening methods in five developing countries.

If developing countries screened all women at least once in their mid-30s, they could cut the number of cervical cancer deaths by up to 25 percent, reports the paper in the Nov. 17 New England Journal of Medicine. With a second screening several years later, the two tests could potentially halve the number of women dying from cervical cancer.

“No matter which option—the high-tech DNA test or low-tech visual inspection—the most important point is the simple fact of being able to deliver the treatment with the screening in one or two visits,” said lead author Sue Goldie, HSPH associate professor of health decision science.

Annual cytological testing of exfoliated cervix cells has been the global standard for decades, but the numbers tell the story of its failure to curb a highly preventable cancer in many regions. Cervical cancer tops the cancer mortality charts in poorer countries, which account for 80 percent of all cervical cancer deaths. It tends to kill women in their 40s, leaving large young families to fend for themselves.

Happily, abnormal cells destined to become cervical cancer are easily detected and safely treated in the earliest stages.

Standard Costs
The power of the Pap smear in knocking down cervical cancer in North America and Europe comes from established laboratories with refined techniques and quality controls, highly trained pathology professionals, and the money and time for several follow-up visits to confirm and treat precancerous cells.

It may surprise women to learn that Pap smears are little better than coin flips at detecting precancerous changes. Annual visits are necessary to increase the odds that the slow--growing precancerous cells will show up in a microscopic sample the next year.

“In the United States and many Western democracies, the Pap is quite successful, because we throw a lot of money at it,” said Eduardo Franco, director of the cancer epidemiology division at McGill University in Montreal. “It is silly for all of us in the community of cancer prevention to insist on a screening test that doesn’t work in developing countries, especially when there is new technology that improves the accuracy, such as the [viral] DNA test, or decreases the complexity, such as visual inspection.”

The cost-effectiveness analysis by Goldie and her colleagues augments growing evidence that may change the global standard of cervical cancer screening. For now, the World Health Organization is hanging on to its thick instruction manuals of Pap smear protocols. But in one Thai province, where 20 years of cytology screening once reached less than 5 percent of eligible women, a newer one-stop screen-and-treat program funded by the Bill & Melinda Gates Foundation has reached about half of the women aged 30 to 45 in five years, said Paul Blumenthal, a professor of gynecology and obstetrics at Johns Hopkins University and collaborator on the Goldie study.

Pictures of Health
To figure out what screening strategies could deliver lower cancer rates on a tight budget, Goldie teamed up with research groups in India, Kenya, Peru, South Africa, and Thailand. The unequivocal findings are even more striking in the face of the diverse demographic and economic profiles of each country.

Goldie builds computer models to help health policymakers glimpse into futures they can create, well before a randomized clinical trial can make a definitive determination after the fact. This fall, she won a MacArthur grant “for genius and creativity” in applying the tools of decision science to evaluate the clinical benefits, public health impact, and cost-effectiveness of alternative preventive and treatment interventions for a wide range of important public health problems.

“In decision science, we synthesize a large body of work,” Goldie said. “We ask what is the dent you can make in the burden of disease—in this case, cervical cancer—with alternative prevention and treatment strategies, taking into account the natural history of the disease, population heterogeneity, screening test performance and cost, treatment efficacy, and operational logistic issues.”

Goldie’s models always start with the biology. In the last 10 years, scientists have become convinced that cervical cancer has one root cause, persistent infection by human papillomavirus (HPV). The sexually transmitted infection is so common that three quarters of all women will be infected at some point. Most oncogenic HPV infections and even the mild cellular abnormalities they cause are symptomless and go away on their own. The peak of the transient infections occurs among women in their teens and 20s. The problem lies in the relatively rare persistent HPV infections. Among women in their 30s, these infections and precancerous changes are much more likely to lead to cancer.

Goldie validated her preliminary natural history model by comparing it to the epidemiological data of each country and consulting the research teams. Layered on that, Goldie added reams of data about the strategies for screening and treatment, a topic of great passion and debate among health professionals, including her collaborators.

Image adapted by Rachel Eastwood
from original provided by Sue Goldie

Test failures. The success of the annual Pap smear in reducing cervical cancer deaths in the United States has not translated to poor countries, where cervical cancer is a leading cause of cancer deaths and accounts for 80 percent of cervical cancer deaths worldwide. This graph shows cervical cancer incidence in the U.S. and five diverse low-resource countries.

“Most of the value of this kind of work comes in the process,” Goldie said. “As we iterate and collaboratively review interim results, I find people agree more than they disagree. We define where exactly the points of agreement are and standardize our assumptions about those areas. When there is clear disagreement, the challenge is to understand where the uncertainties are, get better data when possible—including conducting new studies to address key questions—and to explicitly incorporate the uncertainty into the model. If the results of the analysis are robust despite disagreement about selected particular details, it provides a powerful message. It really allows different investigators and practitioners to own the results, and in many cases, over time, extreme views shift.”

As new data become available, the model could spit out different results. For example, the virtual tie between HPV DNA testing and visual inspection belies big differences between them. DNA testing is expensive and accurate, but a negative test provides considerable certainty that the virus is not lingering and wreaking havoc in the cells. Visual inspection, in which vinegar turns precancerous cells cloudy white against a pink cervix, is less accurate than DNA but much cheaper to perform. If HPV DNA test costs go down, it could be the clear winner in a rerun of the cost-effectiveness analysis, Goldie said.

Even then, a visual inspection must be performed to evaluate the next treatment step. “As far as we’re concerned, the purpose of the research is to change policy or programs or, in some cases, to initiate policy and programs [to prevent cervical cancer],” said Blumenthal, whose research has shown the value of visual inspection in identifying and treating precancerous lesions. “It’s not about which test is best, it’s about which is the best test you can do.”

—Carol Cruzan Morton

top