This story by the LA Times covers a study of PapSeek, which is a new three-prong genetic test for Pap smear, inner-uterus smear, and blood samples. The test aims to detect endometrial and ovarian cancers, which typically aren’t found until they’re advanced, limiting effective treatments options and contributing to the lethality.
The story takes care to note the shortfalls of the alternatives, and how an effective early-detection screening could help thousands of women per year in the US alone. However, the story falters in very important ways. There’s no mention that 27% of the study’s authors could benefit financially from the administration and sale of PapSeek tests, doesn’t appear to incorporate independent experts’ perspectives on the study and its significance, and skips over any discussion of costs or potential harms.
Pap smears have slashed cervical cancer death rates by more than 60% since their inception in 1955, bringing the death rate to about 4,000 American women per year. But endometrial cancers kill roughly 10,000 women and ovarian cancer nearly 14,000 women per year — and there’s no widely accepted or effective screening method for early detection. This is especially important for ovarian cancer, as the lack of screening options have meant the numbers haven’t nudged much. Even if a screening test is not as sensitive as direct biopsies of cancerous tissues, any test at all for endometrial or ovarian cancers may catch more cases earlier, help patients begin treatment sooner, and see longer remissions and lifespan.
No costs are discussed, though the US National Library of Medicine says individual genetic tests “can range from under $100 to more than $2,000” — and PapSeek appears to use three genetic tests (on Pap smear, Tao brush, and blood tissues).
The story explains how adding a genetic test to a routine (though uncomfortable) gynecological procedure for women — i.e. Pap smears — could help doctors catch endometrial and ovarian cancers at an early stage, for which there is not yet a standard test. The article also does a nice job showing how layer-caking genetic tests on other tissues (inner uterus and blood) can nearly double the detection rate of ovarian cancer, from 33% (Pap-smear genetic test) to 63% (a full PapSeek test), and improve endometrial cancer detection from 81% to 93%.
This was a close call. On one hand, the story does address harms from false-positive tests: “Just as important, the tests were remarkably good at avoiding false-positive signals, which would needlessly alarm women and lead to riskier and more-invasive tests to confirm a diagnosis. For both types of cancer, the PapSeek test’s “specificity” — its ability to accurately recognize the absence of disease — approached 100%.”
Yet the story glossed over the high false-negative rates and how a patient might (incorrectly) interpret a result — i.e. by testing negative, they may wrongly believe they do not have a uterine or ovarian cancer.
The story describes the difference between test sensitivity (i.e. ability to detect a cancer) and specificity (i.e. ability to avoid false-positives), and explaining the significance of the data in those terms. It also lists the study’s sample size and describes how it was done in broad strokes, including how 656 of the 1,658 women who participated had ovarian or endometrial cancer.
One shortfall is that it doesn’t dig into how small some of the sample sizes were for endometrial cancer (382) and ovarian cancer (245) — in a larger study population, the conclusions of this study may not hold up. It also doesn’t state how the tissues were blindly analyzed by the researchers.
But the biggest limitation here is that this study uses already-confirmed cancer cases to test against. The story needed some caveats about how the accuracy may change when tested on a population of people where their cancer status is unknown.
Cancers noted in this story are described as “killers” and “deadly.” Though most cancers are eventually lethal if untreated, the mortality rate of uterine cancer (and primarily endometrial cancer) is about 18.7% of women within five years of diagnosis; for ovarian cancer, the death rate within five years of diagnosis is far higher, around 53.5%.
The story didn’t disease-monger, but it would have been stronger had it actually described burden of disease of the two cancers, though. For example, while ovarian cancer is more lethal, it’s also more rare.
This is the story’s biggest shortcoming, as there are no independent sources, and it fails to note a number of significant financial conflicts of interest. The research funding appears to be spread among a number of foundations that wouldn’t necessarily profit from PapSeek tests. However, the “competing interests” section details how 10 of the 37 authors stand to profit from the genes and tests used, including through patents, royalties, and equity — including five of the six corresponding authors.
Scientists having a financial stake in study is not uncommon, and doesn’t necessarily weaken a study’s results, but it should serve as a red flag to journalists to seek more outside expert opinions than normal. But the LA Times does not appear to have interviewed even one researcher not involved in the work, let alone without financial stakes in it. (Dr. Nickolas Papadopoulos, the only person quoted in the piece, is listed under several areas of stated competing interests in the study’s notes.)
This information was easily found in the news release.
The story describes standard Pap smear tests, adding that they do a “poor job” of detecting endometrial and ovarian cancer. Blood testing for cancer-related biomarkers, such as CA-125, is also noted as well as ultrasound tests of the uterus. The story also notes these are most commonly late-stage tests.
The story doesn’t discuss if women can get this test — a vital piece of information for those with a high risk of endometrial or ovarian cancers. While individual parts of the PapSeek test are FDA-approved, e.g. Pap smears and Tao brushes, the PapSeek genetic analysis does not yet appear to be approved (though approval criteria for medical devices are not as strict as they are for medications).
Also, there’s been a reduction in frequency of Pap testing in recent years. It’d be good to have a discussion of how implementing a new screening test in this environment muddies a message that’s already been disseminating about how frequently women should be getting pelvic exams and Pap smears (i.e., less often, with some stopping altogether after a certain age). How would a test like this have to work upstream to be reincorporated into gynecological screening?
The story notes that Pap smears rely on using a microscope to hunt for cancer cells, whereas the newer methods described use genetic testing. The story also notes how ultrasound and genetic blood tests for uterine cancers typically are used only when symptoms arise — an advanced stage of the cancers — since the those tests “fail to detect many cancers, and also send up a lot of false alarms.” The article places extra emphasis on so-called “liquid biopsies,” saying thy “hold the promise of revolutionizing cancer screening” due to a lack of early screenings for these cancers. (That remains to be seen.)
We didn’t detect any copy-pasted quotations from a Johns Hopkins Medicine news release, nor other obvious signs of reliance on a press release.