Why a New Immunotherapy for Lung Cancer Works for Only Some People

Share
Cancer genomics researcher Timothy Chan

Genomics researcher Timothy Chan was surprised to discover that lung tumors with a lot of smoking-induced mutations tend to respond better to anti-PD-1 immunotherapy.

When the US Food and Drug Administration approved the drug nivolumab (Opdivo®) last month for use in patients with advanced non-small cell lung cancer (NSCLC), it marked yet another milestone for a new generation of cancer therapies that unleash the immune system to destroy cancer cells.

The drug — called a PD-1 inhibitor — removes a natural brake on the immune system. Memorial Sloan Kettering physician-scientists played a major role in developing this approach, which is producing impressive results when combined with standard anticancer therapies. The effectiveness of the PD-1 inhibitor against lung cancer is especially promising, as the disease is very common and urgently needs better treatments.

The therapy has produced remarkable results, completely eliminating metastatic cancer in some patients. But PD-1 inhibitors are effective in only 20 to 30 percent of people with NSCLC — and so far doctors haven’t been able to predict which patients are most likely to benefit.

Now researchers have gained clarity about what type of tumors this treatment tends to work against. In studying DNA changes in tumors of NSCLC patients who received pembrolizumab, another PD-1 inhibitor, MSK scientists identified a genetic pattern that correlates with successful treatment. This pattern is characteristic of tumors that contain a lot of mutations caused by smoking.

For the first time, a mutational pattern has been linked to immunotherapy outcome.

The findings, published in the April 3 issue of the journal Science, could guide the use of PD-1 inhibitors in lung cancer patients and also influence clinical trial approaches to investigate the drugs for other cancer types.

“The link between smoking-related mutations and immunotherapy responsiveness was totally unexpected,” says MSK cancer scientist Timothy Chan, who led the research. “And this is the first time anyone has shown that a widespread mutational landscape clearly affects the outcome of an immunotherapy.”

The landscape of mutations Dr. Chan describes is typical of cells whose DNA has been damaged by exposure to certain chemicals — including those present in cigarette smoke — or radiation. The damage often leads to a type of genetic error called transversion.

Dr. Chan and his colleagues, including medical oncologist Matthew Hellmann who specializes in lung cancer and is a co-first author on the paper, performed a thorough DNA sequencing analysis of tumors from 34 NSCLC patients who had been treated with pembrolizumab, which is presently being used for melanoma treatment and studied in clinical trials for lung cancer and several other cancers. They discovered that patients are more likely to respond to the therapy if their tumor DNA displays a high number of transversions, a characteristic feature of smoking-related cancers.

Among patients whose tumors had this genetic signature, 72 percent had a lasting benefit from the treatment that continued for six months or more. In comparison, these responses happened in only 13 percent of people whose tumors had low transversion numbers.

A Genomic Test to Guide Treatment

While results indicate that both smokers and nonsmokers with NSCLC may benefit from anti-PD-1 immunotherapy, smokers are much more likely to respond. “It’s relatively easy to find out whether a patient has the ‘smoking signature’ by sequencing their tumor DNA,” Dr. Chan says.

“Interestingly,” he adds, “we found that performing the genomic test was a much better way of identifying responders than collecting data about the patients’ smoking history.” This may be because susceptibility to tobacco exposure varies among people, and patients with similar smoking histories may not have the same amount of mutations in their tumor DNA.

Genetic testing could help doctors and patients make better-informed choices about whether to pursue anti-PD-1 therapy for NSCLC. This means patients with disease that isn’t treatable with immunotherapy could be spared from getting the drug and suffering from its side effects, which often are mild but in some cases can be serious or even life threatening.

Implications for a Range of Cancers

The impact of the findings extends well beyond NSCLC patients with this specific mutational profile. Smoking can lead to many types of cancer other than lung, including head and neck, bladder, and esophageal cancer. “All these smoking-related cancers have a similar mutational landscape,” Dr. Chan explains. He posits that any cancer, regardless of where in the body it first arises, is more likely to be sensitive to the drug if it carries the genetic signature.

Presently, researchers are launching clinical trials of PD-1 inhibitors in a range of cancer types, hoping to figure out why some people respond and others don’t. “Based on our data,” Dr. Chan says, “there may be a rationale for testing these drugs primarily in patients whose tumors have a high transversion rate” and hence may have a better chance of responding.

New Insights about How Checkpoint Inhibitors Work

In November, Dr. Chan and his colleagues reported on findings that explain why another immunotherapy drug, ipilimumab (Yervoy®), works for only about one in five patients with advanced melanoma. A conclusion from that study is that the more mutations a tumor has, the more likely it is to respond to drugs known as checkpoint inhibitors, such as ipilimumab, nivolumab, and pembrolizumab.

In particular, the presence of mutations that make cancer cells express new antigens — substances that the immune system is able to “see” and trigger a response against — appears to be linked to immunotherapy responsiveness.

The new lung cancer study supports this hypothesis. Transversions often lead to changes in a cell’s genetic code, which in some cases may result in the production of new antigens. One explanation as to why lung tumors with high transversion numbers might be more sensitive to a checkpoint inhibitor could be that they are easier for the immune system to detect.

Some features of immunotherapy responsiveness may apply to various cancers.

“It is very important to see that the initial observations we made regarding mutations and immunotherapy response in melanoma are also applicable to other cancers,” notes physician-scientist Jedd Wolchok, a leading immunotherapy expert and a co-author of the study. “This reinforces our belief that there are common features of the relationship between the immune system and various cancers.”

The researchers are hopeful these insights will make it possible to achieve even better results for more patients with different cancer types. Dr. Chan notes that scientists are only beginning to understand the relationship between a tumor’s genetic makeup and its responsiveness to checkpoint inhibitors. Future studies might reveal whether gene changes induced by factors other than smoking, such as UV light or aging, contribute to shaping a tumor’s responsiveness to immunotherapy.