A new drug that unleashes the body’s immune system to attack tumors can prolong the lives of people with the most common form of lung cancer, doctors reported on Friday, the latest example of the significant results being achieved by this new class of medicines.

In a separate study, researchers said they had found that a particular genetic signature in the tumor can help predict which patients could benefit from the immune-boosting drugs.

The finding could potentially extend use of these drugs to some patients with colorectal cancer, prostate cancer and other tumors that have seemed almost impervious to the new drugs. Most of the substantial results so far with these expensive drugs have been in treating melanoma and lung cancer.

“If you have the signature, you should treat with these checkpoint inhibitors,” Dr. Luis A. Diaz Jr., an associate professor of oncology at Johns Hopkins University and the senior author of the study on the genetic marker, said in an interview, referring to the new drugs.

Both studies were discussed at the annual meeting of the American Society of Clinical Oncology, which began Friday in Chicago. The study on the genetic signature was also published online by the New England Journal of Medicine.

The checkpoint inhibitors work by releasing molecular brakes, or checkpoints, that prevent the body’s immune system from attacking tumors. The products on the market so far are Keytruda, which is made by Merck, and Opdivo and Yervoy from Bristol-Myers Squibb. They have been the center of attention at the oncology conference the last two years and this year should be no different.

All three drugs are approved to treat melanoma, a deadly skin cancer. Opdivo was also approved in March to treat the so-called squamous-cell form of non-small cell lung cancer, which accounts for about one-quarter of the cases of lung cancer.

The new study discussed on Friday at the oncology conference shows that Opdivo also prolongs survival for patients with nonsquamous cell lung cancer, which accounts for most of the rest of the cases.

Patients who received Opdivo lived a median of 12.2 months compared with 9.4 months for those treated with the chemotherapy drug docetaxel. Moreover, Opdivo, also known as nivolumab, had far fewer serious side effects.

Patients whose tumors produced a protein called PD-L1, which looks to be another predictor of success for certain checkpoint inhibitors, fared even better. Among just those patients, median survival was 17.2 months in the group receiving Opdivo and nine months for those getting docetaxel.

“In a randomized trial I don’t think people have ever been able to see this kind of difference,” said Dr. Pasi A. Janne, a lung cancer specialist at the Dana-Farber Cancer Institute in Boston, who was not involved in the study.

The trial, paid for by Bristol-Myers, involved 582 patients with advanced cancer who had already had treatment with platinum-containing chemotherapy like carboplatin. This summer, Bristol intends to apply to have Opdivo approved to treat nonsquamous lung cancer.

In other studies discussed at the conference on Friday, Opdivo significantly shrank tumors in 19 percent of patients with advanced liver cancer and Keytruda in 25 percent of patients with head and neck cancer. There is now only one drug approved for liver cancer, Nexavar from Bayer and Amgen.

Other studies have shown the checkpoint inhibitors can help some patients with kidney, bladder, stomach and other forms of cancer. But there have been few signs of effectiveness with colorectal, prostate or pancreatic cancers.

Why do the drugs work so well for some cancers and barely at all for others?

One thought is that lung cancer and melanoma are often caused by things that damage DNA — tobacco smoke and ultraviolet radiation. They thus “have tons of mutations, hundreds of mutations per tumor,” said Dr. Diaz, more than most other tumors. That might make it easier for the immune system to recognize the cancer cells as something to be destroyed.

But some cancers have even more mutations — those with a genetic defect known as mismatch repair deficiency. This deficiency prevents the repair of changes in DNA that can arise as cells divide, allowing mutations to accumulate.

The deficiency is found in Lynch syndrome, an inherited condition that puts people at a high risk of developing cancer, particularly colorectal. Lynch syndrome accounts for about 5 percent of colorectal cancers. The mismatch repair deficiency is also found in about 10 percent of colorectal cancers that are not inherited, and probably in a few percent of many other cancers, including prostate and pancreatic, Dr. Diaz said.

To test their hypothesis, the researchers tried Keytruda, also known as pembrolizumab, in a small study on patients with advanced cancer who were no longer being helped by other treatments.

Four of 10 colorectal cancer patients with the deficiency experienced a significant shrinkage of their tumors, according to the paper in the New England Journal of Medicine. That was seen in none of the 18 colorectal cancers without the deficiency. Tumors also shrank in five of seven, or 71 percent, of the patients with cancers other than colorectal that had the mismatch repair deficiency.

Colorectal cancer patients with the deficiency also lived longer than those without the deficiency, though by how much is not yet known because many are still living.

DNA sequencing revealed that tumors with the deficiency had an average of 1,782 mutations, compared with only 73 for the tumors without the defect.

One of the patients in the study, Stefanie Joho, has Lynch syndrome. Her mother had colon cancer at age 44, then uterine cancer six years later. Ms. Joho was even unluckier, diagnosed with colorectal cancer shortly after graduating from New York University.

Two surgeries and two different types of chemotherapy failed to stop the disease. Out of options and in terrible pain despite taking high doses of a narcotic, Ms. Joho said she quit her job in New York and returned to her parents’ home outside Philadelphia, where she was mostly bedridden.

But after she began receiving Keytruda in August, her tumor started shrinking. Her pain ebbed so much that she threw away her narcotics, then had to be treated briefly at the hospital for withdrawal symptoms.

“I haven’t felt this well in four years,” said Ms. Joho, who is now 25.

Genetic tests for mismatch repair deficiency are commercially available. But insurers might not pay for the drugs — Keytruda and Opdivo cost $150,000 a year — based on such a small study. The study was paid for by Swim Across America and other charities, and the National Institutes of Health. Some of the authors have a patent pending on technology used in the study.

© 2015 The New York Times Company