Welch. "It's not just cancer screening. It happens with heart disease, osteoporosis. With all diseases there is this push to go earlier."

Many of the public messages about screening communicate two things, says Schwartz: "It's really dangerous if you don't" get screened and "This is what a good person does." Framing the decision to be screened or not in a moral context makes it "harder for people to make an informed decision," she maintains. "How much are we destroying whatever sense of health and well-being people have by continually scaring them?" she wonders.

But Americans have been scared of cancer for a long time—long before cancer screening became popular and celebrities like Katie Couric began pushing various tests (see here in this issue for a related story). Consider, for instance, the Neil Simon play Brighton Beach Memoirs, set in Brooklyn in 1937. The play centers on a working-class Jewish family, with the main character a baseballcrazed teenager named Eugene. Early in act one, Eugene addresses the audience:

Eugene: [To the audience.] If my mother knew I was writing all this down, she would stuff me like one of her chickens. I'd better explain what she meant by Aunt Blanche's "situation." You see, her husband, Uncle Dave, died six years ago from [he looks around] this thing—they never say the word. They always whisper it. It was [he whispers]—cancer! I think they're afraid if they said it out loud, God would say, "I heard that! You said the dread disease! [He points his finger down.] Just for that, I smite you down with it!!"

Yet even today, despite the fact that researchers have a much better understanding of the broad spectrum of cellular abnormalities that are considered cancer, many such attitudes about the disease persist. Black explains why: "Our knowledge about the natural history of cancer is based largely on historical cases." Even though "a tiny little nodule in a CT scan" goes by the same name as the large tumors that were diagnosed in previous centuries, "people often assume that all these tiny abnormalities . . . will lead to the serious outcomes we observed in the past."

Wendy Wells, M.B.B.S, a British-trained Dartmouth pathologist who specializes in breast cancer diagnosis, agrees that a lot of confusion and anxiety surround the C-word. "As Gil would say, there's cancer and there's cancer," Wells puts it. She explains that labeling abnormal cells as cancer or not can be a close call—a fact that many patients are not aware of. In some cases, "if you understand how close you are to either getting the C-word or not getting the C-word, it's actually

The graph above and the box below show the discrepancy between what the public believes is the responsible stance regarding cancer screenings, and what the experts recommend. That sense of obligation is not necessarily a good thing, according to several DMS faculty.

quite terrifying," she adds.

"We've hit the morphology wall for all those descriptive things that pathologists learn about, that define normal or hyperplastic [that is, abnormal but not definitely cancerous] or cancer," explains Wells. "We're very good at that, but we can't predict behavior. So the race now is to look for other prognostic indicators . . . not only at the protein and molecular level but also at the gene level."

Pathologists diagnose cancer based on the characteristics of cells —known as cytology—as well as on the way the cells arrange themselves—known as architecture. By analyzing the cytology and architecture of cells in a tissue sample, pathologists decide whether the cells are cancerous or not. Next, if they are cancerous, the pathologists decide whether they are very abnormal (high-grade) or slightly abnormal (low-grade) and whether they have invaded the surrounding tissue (invasive) or not (noninvasive).

Beyond these rather general categories, however, are more complicated classification systems. And as Wells discovered through research that she and Patty Carney did with the New Hampshire Mammography Network, pathologists often don't agree on how to classify a given tissue sample. In other words, one pathologist might diagnose the cells from

a breast biopsy as atypical ductal hyperplasia, which is not considered cancer, while another pathologist might label the very same sample as low-grade ductal carcinoma in situ (DCIS), which is considered to be cancer.

"If you get the diagnosis of atypia, nothing else is done—no further excision, no radiation. If you just nudge the criteria and make it a low-grade DCIS, suddenly: Excision! Radiation! Everything full hog! And yet, the criteria are so nebulous," says Wells. She adds, however, that only about 2% of the breast biopsies she evaluates fall into this difficult-to-classify category. And when she can't reach a conclusion on a sample, Wells will consult with her colleagues and sometimes with experts elsewhere. If a consensus on a diagnosis still can't be made, the patient is notified. This is "a huge onus on the patient because you're literally saying, 'We really can't make up our mind,'" she says. Until pathologists know more about how certain cancer cells behave over time, these uncertainties will likely remain.

Wells says she is neither an advocate for nor an opponent of screening—though, at age 45, she does get regular mammograms—because not enough is known about how certain cancers behave. "I don't think we have enough data," says Wells. "Why Patty's work is so interesting is that she's doing exactly that. She's trying to get all the data" through the mammography network. But, she adds, "I see where Gil is coming from. I do think we are harming some patients."

All sides seem to agree on that point: some patients are being harmed by screening. In his book, Welch offers some advice to those who are concerned about being overdiagnosed. "Probably the best way to minimize the harmful effects of screening is to be willing to take some time" with the small, questionable abnormalities, he writes. "Even when 'cancer' is agreed on, it may make sense to wait and be sure the cancer is really growing." Watchful waiting, as the wait-andsee strategy is called, can be hard for patients and doctors alike because in some ways, just like screening, watchful waiting is a gamble.

Many public messages about screening communicate two things, says Schwartz: "It's really dangerous if you don't" get screened and "this is what a good person does." Framing the decision to be screened or not in a moral context makes it "harder for people to make an informed decision," she maintains.

One DHMC patient who opted for watchful waiting believes he's won the "bet." Robert Aliber, Ph.D., an emeritus professor of international economics and finance at the University of Chicago who now lives in Hanover, was diagnosed with an early-stage, moderately aggressive prostate cancer in 2001, when he was in