Tom Maguire always figured that, if he ever developed cancer, he would pursue the toughest treatment available. “You destroy yourself, and then you can come back,” he said.

His view was tested this year when he was told he had bladder cancer that had invaded the muscle wall of the organ. The standard of care, he learned, usually involves removing the bladder. He would have the choice of wearing a bag to collect his urine or having a difficult surgery to fashion a new bladder from his intestines. Both prospects filled the 63-year-old avid hiker and scuba diver with dread.

Then doctors at Philadelphia’s Fox Chase Cancer Center told him about a new clinical trial designed to allow people with certain types of tumors to keep their bladders while being closely monitored.

Since getting into the trial a few months ago, “I have been walking on air,” he said. “I guess you don’t always have to take an all-in, nuclear approach.”

For decades, “Big C” diagnoses sent patients and doctors scrambling for the most aggressive weapons to take into battle. The severity of the threat, as well as the limited knowledge of the disease and the pervasive use of military language to describe cancer, justified the blunt-force approach.

“Our focus historically on the ‘war on cancer’ implied that more is better and decimation is desired,” said Justin Bekelman, a radiation oncologist at the University of Pennsylvania.

But today, the “fighting cancer patient” metaphor is falling out of favor, not only because it subtly blames patients who “lose the fight” but also because it doesn’t capture a world of new biological insights, improved treatments and molecular tests that are transforming how cancer is treated.

At the root of the change is the recognition that not all cancers are the same. Equipped with new tools and evidence, oncologists are “de-escalating,” cutting back on toxic and costly approaches likely to do more harm than good. “Knowing when not to treat is great medicine,” Bekelman said.

Yet for many patients, and even some doctors, doing less in the face of danger is emotionally difficult. Reshma Jagsi, a radiation oncologist at the University of Michigan, said she sees women every day who look for the most aggressive treatment for breast cancer, even if it isn’t needed. “They say, ‘I need to be there for my kids, my students, my fill-in-the-blank,’ ” she said.

Increasingly, however, strong evidence for de-escalation is spurring cultural and medical practice shifts. A landmark clinical trial published in June found that more than two-thirds of women with early stage breast cancer can safely avoid chemotherapy. Men with early stage, low-risk prostate cancer are rapidly embracing “active surveillance” over surgery — and avoiding possible complications such as incontinence and sexual dysfunction. And throat cancer caused by human papillomavirus, doctors now know, is different from other types of the disease, allowing a cutback in a brutal treatment regimen and reducing the risk of potentially devastating disfigurement.

In lung cancer, immunotherapy, which is usually less toxic than chemotherapy, has emerged as a first-line treatment for many patients. And people with advanced kidney cancer can skip surgery to have their kidneys removed and instead go right to drug treatment, a recent study showed.

The de-escalation trend hasn’t spread to all diagnoses. The most common form of thyroid cancer, which poses little risk, is often still treated with unnecessary surgery, experts say. And some malignancies, such as pancreatic cancer, are so lethal that doctors are racing to find ways to ramp up treatment.

Selecting the right patients for less-intensive therapies is critical. A prostate cancer patient with a mutation of the BRCA gene — an indication that the cancer will probably spread more quickly — needs treatment, not monitoring, experts say. Finding more biological indications, or “biomarkers,” that predict a tumor’s course or how it is likely to respond to specific treatments is crucial for guiding targeted treatments.

“It’s a precision oncology story,” said Norman “Ned” Sharpless, the director of the National Cancer Institute.

But even proponents of de-escalation acknowledge it can be a challenging change for physicians who have been trained that more intense treatment is preferable. “Better safe than sorry is a paradigm,” said oncologist Elizabeth Plimack.

That’s why biomarkers are so important to help guide therapy decisionsand build acceptance for them. In 2015, Plimack and her colleagues discovered that certain genetic mutations in bladder-cancer tumors predict that chemotherapy will wipe out the cancer and make it unlikely to recur.

These biomarkers, which occur in only a minority of patients, suggest that treatment can be based partly on patients’ individual risks, said Daniel Geynisman, the oncologist leading the bladder-cancer trial. The way medicine is currently practiced, “we are clearly overtreating a lot of patients, and we are probably undertreating some patients.”

Bladder cancer is the sixth-most common cancer in the U.S., said to the National Cancer Institute. Many of the cases are confined to the lining of the bladder and those the tumors are removed. But cancer that has grown into the muscle wall of the bladder is considered invasive. In most of those cases, the standard of care involves extensive surgery to remove the bladder and prostate in men and the bladder and uterus in women. “Nobody wants that surgery,” Geynisman said.

Maguire said he was resigned “if that was going to be the price for staying alive.” But then he learned he had the biomarkers that made him eligible for the bladder-preservation trial. “I have never been lucky,” he said.

If the techniques work, and are confirmed in larger trials, it will change the traditional standard of care, Geynisman said.