Dr. Oliver Sartor has a provocative question for patients who are running out of time.
Most are dying of prostate cancer. They have tried every standard treatment, to no avail. New immunotherapy drugs, which can work miracles against a few types of cancer, are not known to work for this kind.
Still, Dr. Sartor, assistant dean for oncology at Tulane Medical School, asks a diplomatic version of this: Do you want to try an immunotherapy drug before you die?
More from the New York Times:
A lynching memorial is opening. The country has never seen anything like it.
A lifesaving pump for cancer patients is being phased out
F.D.A. cracks down on 'juuling' among teenagers
The chance such a drug will help is vanishingly small — but not zero. "Under rules of desperation oncology, you engage in a different kind of oncology than the rational guideline thought," Dr. Sartor said.
The promise of immunotherapy has drawn cancer specialists into a conundrum. When the drugs work, a cancer may seem to melt away overnight. But little is known about which patients might benefit, and from which drugs.
Some oncologists choose not to mention immunotherapy to dying patients, arguing that scientists first must gather rigorous evidence about the benefits and pitfalls, and that treating patients experimentally outside a clinical trial is perilous business.
But others, like Dr. Sartor, are offering the drugs to some terminal patients as a roll of the dice. If the patient is dying and there's a remote chance the drug will help, then why not?
"Immunotherapy is a particularly nuanced problem," said Dr. Paul Helft, an ethicist and oncologist at Indiana University School of Medicine.
Cancer doctors are well aware of the pitfalls of treating patients before all the evidence is in.
Many still shudder at the fiasco that unfolded in the 1980s and 1990s, when doctors started giving women with breast cancer extremely high doses of chemotherapy and radiation on the theory that more must be better. The doctors did not systematically collect data; instead, they reported patient anecdotes claiming success.
Then a clinical trial found that this treatment was much worse than the conventional one — the cancers remained just as deadly when treated with high doses, and the regimen itself killed or maimed women.
But immunotherapy is like no cancer treatment ever seen. It can work no matter what kind of tumor a person has. All that matters is that the immune system be trained to see the tumor as a foreign invader.
Tumors have mutations that stud them with bizarre proteins. The white blood cells of the immune system try to attack but are repelled by a molecular shield created by the tumors. The new drugs allow white blood cells to pierce that shield and destroy the tumors.
Last week brought a yet another example of the surprising power of this approach. Lung cancer patients who normally would receive only chemotherapy lived longer when immunotherapy was added, researchers reported in a clinical trial.
But the drugs are exorbitantly expensive. One that Dr. Sartor often uses costs $9,000 per dose if used once every three weeks, and $7,000 if used once every two weeks. Often, he and other doctors persuade a patient's insurer to pay. If that fails, sometimes the maker will provide the drug free of charge.
Immunotherapy drugs can have severe side effects that can even lead to death. Once the immune system is activated, it may attack normal tissues as well as tumors. The result can be holes in the intestines, liver failure, nerve damage that can cause paralysis, serious rashes and eye problems, and problems with the pituitary, adrenal or thyroid glands. Side effects can arise during treatment or after the treatment is finished.
For most patients, though, there are no side effects or only minor ones. That makes giving an immunotherapy drug to a dying patient different from trying a harsh experimental chemotherapy or a treatment like intense radiation.
The problem is deciding ahead of time if an immunotherapy drug will help. Doctors check biomarkers, chemical signals like proteins that arise when the immune system is trying to attack. But they are not very reliable.
"A positive biomarker does not guarantee that a patient will benefit, and a negative biomarker does not mean a patient will not benefit," said Dr. Richard Schilsky, senior vice president and chief medical officer of the American Society of Clinical Oncology. "You don't have a solid biology to go on."
It was this problem, described at a medical conference a couple of years ago, that led Dr. Sartor to begin offering immunotherapy to dying patients.
"I was thinking, 'My God, these tests that are used to drive clinical decision making are not worth a damn,'" he said. "These are peoples' lives here. We are playing with the highest of stakes."
"For certain people it is like, bingo, you give the drug to them and they have a long-lasting and positive benefit," he added. "When our knowledge is not sufficient to inform our decisions, then we have an ethical conundrum."
Out of curiosity, Dr. Sartor emailed eight prominent prostate cancer specialists asking if they, too, offered immunotherapy drugs to patients on the off-chance the treatments would help.
Five said they offer it, with a variety of provisos, offering comments like, "If I was a patient, I want my doc to do everything."
Dr. Daniel George, at Duke University, said he does not offer immunotherapy to every man who is dying of prostate cancer. But, he said, "for those patients who want to do everything they possibly can, that's the group where we try checkpoint inhibitors," a type of immunotherapy.
To the others — the majority of his patients with metastatic prostate cancer — he does not mention immunotherapy.
"We have to balance between hope and reality," he said. "The most difficult conversation we have with patients is when we have to tell them that more treatment is actually hurting them more than the cancer."
Dr. Daniel Petrylak, a prostate cancer specialist at Yale, said his inclination was to offer immunotherapy only to those rare patients whose tumors have a genetic marker indicating the immune system is trying to attack — already an approved indication for prostate cancer, he noted. But this strategy gives him a rationale for trying the drugs on patients with other cancers.
With the possibility of a dramatic and prolonged response, he said in an interview, "how can you ethically deny this to patients?"
At the Dana-Farber Cancer Institute in Boston, Dr. Christopher Sweeney said he petitions an insurance company to get an immunotherapy drug when the patient has a genetic marker predicting a possible response — an indicator the drug might work even if there is as yet no clinical trial evidence that it will — and is strong enough to tolerate the treatment.
But if those conditions do not apply, as is usually the case, Dr. Sweeney only gives the drugs to patients if he can do so as part of a clinical trial, where something can be learned from their experience.
And if there is no clinical trial for the patient? "I basically say I don't have any approved therapies," Dr. Sweeney said. "Here's the truth — most patients don't benefit from these drugs."
He tells patients that just because he has no more drugs to give does not mean he has abandoned them. Supportive care can make patients more comfortable, even prolong their lives.
Dr. Sartor disagreed with the approach. "I would love for every patient to be on a clinical trial," he said. "But does that mean I shouldn't try because I don't have a trial?"
One of the first patients Dr. Sartor treated with immunotherapy was George Villere, a retired investment adviser who lived in New Orleans.
Mr. Villere had bladder cancer and had tried chemotherapy. It didn't work, so Dr. Sartor told Mr. Villere that he had run out of conventional options and asked if he wanted to try immunotherapy. At the time, the drugs had not been approved for bladder cancer.
Mr. Villere and his wife, Fran Villere, thought it over, asking themselves whether they would regret it if they did not try. "I thought we would," Mrs. Villere recalled in an interview.
Their insurance agreed to pay, and Mr. Villere took the drug for several months. Nonetheless, he died on November 15, 2016, at age 72.
"He had no side effects," Mrs. Villere said. "But the drug didn't do a damn thing."
Then there is Clark Gordin, 67, who lives in Ocean Springs, Miss. He had metastatic prostate cancer, "a bad deck of cards," he said in an interview.
Dr. Sartor tried conventional treatments, but they didn't work for Mr. Gordin. Finally, the doctor suggested immunotherapy.
Mr. Gordin's insurer refused. But then the lab that had analyzed his tumor discovered it had made a mistake.
There was a chance Mr. Gordin might respond to immunotherapy, because he had a rare mutation. So his insurer agreed to pay.
Immediately after taking the drugs, Mr. Gordin's PSA level — an indicator of the cancer's presence — went down to nearly zero.
"Makes my heart nearly stop every time I think about it," Dr. Sartor said. "Life sometimes hangs on a thin thread."