The first tissue-agnostic cancer drug approval

Historical milestone — this page describes an event dated 2017. It is not current breaking news.

The first approval that ignored the organ

On May 23, 2017, the FDA granted accelerated approval to pembrolizumab, sold as Keytruda. The approved group was adults and children with solid tumors that are microsatellite instability-high or mismatch repair deficient. The tumor had to have grown after earlier treatment, with no satisfactory alternative left. Colorectal cancer was included, after a fluoropyrimidine, oxaliplatin and irinotecan had failed.

Until then every cancer drug approval named an organ. Lung cancer. Breast cancer. This one named a defect in the tumor's DNA and let the organ fall where it may. The FDA had never done that before.

What the two acronyms mean

Cells copy their DNA constantly, and the copying makes mistakes. A set of proteins called the mismatch repair system finds and corrects them, like a proofreader.

When that system is broken, the tumor is called mismatch repair deficient, or dMMR. Errors then pile up, especially in short repeated stretches of DNA called microsatellites, which shift length from cell to cell. A lab can measure that instability. A tumor with a lot of it is microsatellite instability-high, or MSI-H.

The two labels describe the same failure, found by two different tests. dMMR is usually found by immunohistochemistry, which stains for the repair proteins and looks for one that is missing. MSI-H is usually found by PCR, which measures the length of those repeated stretches.

Why a broken proofreader makes immunotherapy work

Here is the logic that made the approval possible.

A tumor that cannot fix copying errors accumulates enormous numbers of mutations. Mutations change proteins. Changed proteins look foreign to the immune system. So a dMMR tumor is unusually visible to T cells, whatever organ it grows in.

The problem is that tumors switch those T cells off, using a brake called PD-1. Pembrolizumab blocks PD-1 and releases the brake. Put those two facts together, and the tumor's DNA repair status predicts the response better than its location does. Our explainer on biomarker testing and precision medicine covers how these tests are ordered.

The evidence, and its shape

The label is specific about what the approval rested on. Efficacy came from 149 patients with MSI-H or dMMR solid tumors. They were pooled across five open-label, single-arm trials: KEYNOTE-016, 164, 012, 028 and 158. None of them had a comparison group.

Among those 149 patients, the median age was 55. Ninety-eight percent had metastatic disease. The median number of prior treatments was two. Most, 135 of 149, had their MSI-H or dMMR status confirmed before enrolling.

The response rate was 39.6%. Broken out, 7.4% had a complete response, meaning no detectable tumor left, and 32.2% a partial response. By tumor type it was 36% in the 90 patients with colorectal cancer and 46% in the 59 with other cancers.

Durability was the striking part. Median duration of response had not been reached when the label was written, and 78% of responses had lasted six months or longer.

What accelerated approval means

The label says the indication was "approved under accelerated approval based on tumor response rate and durability of response," and that continued approval could depend on confirming benefit in later trials.

That sentence carries weight. Response rate measures tumors shrinking on a scan. It is not the same as people living longer. Drugs have shrunk tumors without extending life. The FDA took the earlier measure here because the group had few options and the responses lasted.

Who this actually applies to

MSI-H and dMMR tumors are a minority. The feature turns up most often in colorectal and endometrial cancer, and it can be inherited. In Lynch syndrome, a mismatch repair gene is faulty from birth, which raises the risk of several cancers. Our page on Lynch syndrome explains the inherited version, and our guide to MSI-H colorectal cancer covers what the result means on a pathology report.

The practical consequence is that the result has to be looked for. A tumor is only eligible if someone ordered the test.

When to get checked

This page is about a laboratory result, not a symptom. Two things follow.

If you or a close relative has had colorectal or endometrial cancer, especially under 50, ask whether the tumor was tested for MSI or mismatch repair proteins, and whether genetic counseling for Lynch syndrome is appropriate.

For the cancers where this feature is most common, these symptoms warrant an appointment:

• Blood in the stool, or a change in bowel habit lasting over three weeks
• Bleeding between periods, or any vaginal bleeding after menopause
• Belly pain or cramping that does not settle
• Weight loss you cannot explain
• Iron-deficiency anemia with no obvious cause

Two or more relatives with colorectal, endometrial or related cancers, particularly at young ages, is worth raising as a pattern rather than a coincidence.

What this does not mean

• This was an accelerated approval based on response rate in single-arm studies. At that point it did not show that people lived longer.
• It applies only to tumors carrying MSI-H or dMMR, a minority of cancers. Most patients are not eligible.
• A 39.6% response rate means most patients in those trials did not respond.
• Immunotherapy carries real risks. The 2017 label lists immune-mediated pneumonitis, colitis, hepatitis, endocrine problems and kidney inflammation.
• The 2017 label quoted here is not the current one, and the indication has been revised since.

Sources

• FDA, KEYTRUDA (pembrolizumab) prescribing information, BLA 125514 supplement 14, approved 23 May 2017 — https://www.accessdata.fda.gov/drugsatfda_docs/label/2017/125514s014lbl.pdf
• FDA openFDA Drugs@FDA API record for BLA 125514 — https://api.fda.gov/drug/drugsfda.json?search=application_number:%22BLA125514%22&limit=1
• NCI, Biomarker Testing for Cancer Treatment — https://www.cancer.gov/about-cancer/treatment/types/biomarker-testing-cancer-treatment
• SEER Cancer Stat Facts, Colorectal Cancer — https://seer.cancer.gov/statfacts/html/colorect.html

How this article was prepared

An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed.

The National Cancer Information Foundation publishes Cancer Explained. This page is for learning. It is not medical advice and does not suggest a test or treatment.

See an error, old source, or unclear wording? Tell us.