THURSDAY, June 29, 2017 (HealthDay News) -- Joann Elmore is a doctor, so when her dermatologist said her skin biopsy indicated possible melanoma, she knew just what to do -- get a follow-up biopsy to verify.
But she got two polar-opposite diagnoses, leaving her anxious and uncertain. One pathologist declared it benign, while the other called it suspicious for invasive melanoma, the deadliest form of skin cancer.
"It showed me what patients go through," said Elmore, a professor at the University of Washington School of Medicine. "It sort of made me realize that in much of what we do, there is an art and it is subjective."
Pathologists can vary widely when assessing skin biopsies for melanoma, particularly when the case is not clear-cut, according to a study led by Elmore.
When asked to repeatedly assess the same set of cases, pathologists often disagreed with one another and, sometimes, themselves.
Disagreements occurred more often when the skin biopsy fell in the middle ground between clearly benign or definitely melanoma, Elmore said.
"It's very hard classifying in those middle diagnostic categories," Elmore explained. "On the two extremes -- if it's totally normal or if it's a high-grade invasive melanoma -- those are easier to identify when you're looking at them under a microscope."
The only way to diagnose melanoma at this time is to place a skin sample under a microscope and compare it to textbook examples, Elmore said.
The problem is, these examples used for diagnosis have never been organized into one clear standard shared by everyone, she noted.
"The pathologists are not the problem," Elmore said. "The problem is we have not developed adequate classification systems."
In Elmore's case, it took a leading pathologist who wrote textbooks on the topic to provide a definitive third opinion -- she had an atypical lesion that is harmless but looks a lot like melanoma.
For their study, Elmore and her colleagues pulled together 240 different skin biopsies ranging from an ordinary mole to advanced melanoma.
The study authors sent off the glass slides, in groups of 36 or 48 slides, to 187 pathologists across the United States. The pathologists reviewed the same slides on two occasions, at least eight months apart.
Elmore's team then compared the pathologists against themselves and a "gold standard" consensus diagnosis for each slide reached by a panel of pathology experts.
Pathologists often agreed on their diagnosis in cases at either extreme. But lesions that were not that straightforward produced a lot of disagreement, the investigators found.
The pathologists agreed with the consensus diagnosis less than half of the time in these middle-ground cases, with agreement ranging from 25 percent to 43 percent. Agreement with themselves ranged from 35 percent to 63 percent, the findings showed.
One slide in particular produced 18 different diagnoses from 36 pathologists, representing the full range from common mole up to invasive and heavily pigmented melanoma, the researchers reported.
Dr. Ted Gansler said these results "may be shocking to patients, but should not be surprising to pathologists or dermatologists who have spent time peering at these samples through a microscope." Gansler is a pathology expert.
"Most renowned experts in this field sometimes disagree," said Gansler, strategic director of pathology research for the American Cancer Society. The reality, he said, is no other lab test is better than the consensus of these experts in predicting whether a spot on someone's skin is benign, malignant or atypical.
"As one would expect, ordinary pathologists are not as accurate as these experts," Gansler added.
Elmore said that patients facing one of these uncertain diagnoses need to remain calm, since the great majority of skin biopsies do not wind up diagnosed as melanoma.
"Most biopsies are benign. Even if they get one of these diagnoses, it doesn't mean they need to act immediately," she added. "They have time to communicate with one of their doctors and get educated. It's not an emergency that needs to be dealt with within the next five days."
Gansler pointed out that in "real world" practice, doctors often share difficult cases with colleagues who have more expertise in assessing skin biopsies, or forward them to recognized experts.
At the same time, Elmore said, there's a need for the field of skin cancer to come together and create a clearer classification system.
"We're at a point where we need to step back and figure out how to fix this," she concluded.
The study was published online June 28 in BMJ.
SOURCES: Joann Elmore, M.D., MPH, professor, University of Washington School of Medicine, Seattle; Ted Gansler, M.D., MPH, strategic director, pathology research, American Cancer Society; June 28, 2017, BMJ, online