Prostate cancer cells that have specific mutations are more likely to respond to immunotherapy drugs. (Parviz M. Pour/Science Source)
Immunotherapy drugs called checkpoint inhibitors hold great promise for cancer treatment. But only a minority of people respond to them. Researchers are trying to develop better therapies and figure out ways to identify who will respond to the drugs that already exist.
Checkpoint inhibitors tend to work well in people whose cancer cells are deficient in a DNA-repair process called mismatch repair (MMR). MMR-deficient cells can have hundreds or even thousands of mutations. A typical cancer cell has just a few dozen. Tumors with an MMR deficiency frequently display a characteristic called high microsatellite instability (MSI-H). Cells with MSI-H have repeated sequences of DNA, showing that something has gone awry in the repair process.
In 2017, the US Food and Drug Administration approved the checkpoint inhibitor pembrolizumab (Keytruda®) for MSI-H cancers. It was the first time that a drug was approved based on a tumor’s genetic profile rather than where the cancer originated.
Now a team of Memorial Sloan Kettering researchers has published findings in JAMA Oncology that show how people with MSI-H prostate cancer respond to these drugs. We spoke with MSK medical oncologist Wassim Abida, who led the research, on what this means for people with this specific type of prostate cancer.
What did this study reveal about MSI-H in prostate cancer, and why is this important?
Checkpoint inhibitors have had less impact in prostate cancer compared with many other cancers. There was not a clear sense of how many prostate cancers harbor evidence of MSI-H.
We wanted to see if DNA sequencing was an effective way to detect the presence of MSI-H or MMR deficiency. We used MSK-IMPACT™, a genetic test that looks for cancer-related mutations. We sequenced tumors from slightly more than 1,000 people having treatment for prostate cancer at MSK. The goal was to identify people who could enroll in clinical trials based on specific mutations. In this group, only 32 had cancer cells with MSI-H. This comes to a frequency of just over 3%, so it’s a small subset.
Of those 32 people, 11 received pembrolizumab or a similar immunotherapy drug. Half of those receiving these drugs showed a lasting benefit. This was true even in some people who had to stop taking the drugs because of side effects. So though the overall number of people responding to the treatment is small, the benefit can be durable.
How could this discovery change treatment for people with MSI-H prostate cancer?
First, this finding suggests that everyone with advanced prostate cancer should be considered for testing for MMR deficiency or MSI-H status. This is now possible with sequencing tools such as MSK-IMPACT. One conventional way to look for MMR deficiency has been a process called immunohistochemical staining, which uses antibodies to test for certain markers in a tissue sample. It is not feasible to do that kind of testing to identify the relatively small number of men who have MMR-deficient high or MSI-H cancers.
But DNA sequencing can be done to look for different targetable mutations, and this study shows that it is a practical way to test for MSI-H cancers. We are already using MSK-IMPACT to sequence tumors from all people with advanced cancers, so adding MSI testing to the panel, which we did in 2017, was straightforward. This is a good way to identify the 3 to 4% of men who might respond to pembrolizumab or a similar drug.
Although a response rate of about 50% is good, we need to investigate why these drugs don’t work for everyone. This is true not just in prostate cancer but for all solid tumors. The presence of MSI-H alone doesn’t necessarily mean they will respond, so we need a better understanding of which other factors are driving the cancer.
Does the presence of MSI-H in someone with prostate cancer have any impact beyond immunotherapy treatment?
In our study, seven of the 32 MSI-high patients also had a mutation in a gene associated with Lynch syndrome. This condition runs in some families and has been shown to increase the risk of developing certain types of cancer. In 2018, an MSK-led study estimated that people with high MSI had a one in six chance of carrying Lynch syndrome mutations. Our study suggests that this rate could be even higher in people with MSI-H prostate cancer — about one in four or one in five. So testing for Lynch syndrome mutations should be considered for people with MSI-H prostate cancer and can have implications for their family members as well.
We also found that the MSI characteristic occasionally did not appear until the disease spread. It was not present in samples from the prostate but was found in a later metastasis in two people with MSI-H prostate cancer. This means that MSI can be acquired during the course of the disease, and it suggests that, if possible, it is preferable to test a new biopsy rather than an older tumor.
