Immunotherapy for bowel cancer

Your immune system is constantly inspecting cells in your body. When it spots a cell that looks foreign or damaged — such as one infected with a virus — it attacks it. Cancer cells are abnormal, and the immune system should in principle recognise them. Sometimes it does, and the cancer is quietly cleared before anyone knows it was there. But many cancers grow large enough to cause illness, and one of the reasons is that they have found ways to hide from the immune system.

One of the ways cancers hide is by displaying a molecule called PD-L1 on their surface. This molecule pushes the brakes on any immune cell that comes close, telling it "leave me alone". The most widely used immunotherapy drugs in bowel cancer are checkpoint inhibitors — antibodies that block PD-1 (the brake on the immune cell) or PD-L1 (the brake on the cancer cell). With the brakes lifted, the immune system can attack the cancer.

This is fundamentally different from chemotherapy. Chemotherapy kills cancer cells directly. Immunotherapy does not — it removes the cloak that is hiding the cancer from your own immune system, and lets your body do the work.

Immunotherapy works in bowel cancer because of one specific piece of biology, and understanding it is the most useful thing on this page.

Every time a cell divides, it copies its DNA. The copying machinery makes occasional mistakes. Normal cells have a built-in spell-check called the mismatch repair (MMR) system — four proteins (MLH1, MSH2, MSH6, PMS2) that find and fix these copy errors. When the spell-check is working, errors are corrected before they cause problems.

In a minority of bowel cancers, this spell-check is broken. Errors accumulate at very high rates, particularly in stretches of repetitive DNA called microsatellites. The cancer ends up with an unusually large number of abnormal proteins on its surface, which the immune system reads as "this looks foreign". These cancers are described as:

• Mismatch-repair-deficient (dMMR) — based on staining for the four proteins.
• Microsatellite-instability-high (MSI-H) — based on a DNA test of the microsatellites.

The terms are used interchangeably; they describe the same underlying problem from two angles. The opposite — a normal, working spell-check — is called mismatch-repair-proficient (pMMR) or microsatellite-stable (MSS).

How big is each group?

• Early-stage bowel cancer (stage I to III). About 12 to 15 per cent of cancers are dMMR/MSI-H — more common in right-sided colon cancers and in younger patients.
• Metastatic bowel cancer (stage IV). Only about 4 to 5 per cent are dMMR/MSI-H. (The reason for the lower rate at this stage is partly that dMMR cancers tend to spread less aggressively in the first place.)

About a third of dMMR bowel cancers occur in people with Lynch syndrome — an inherited condition where one of the four mismatch-repair genes is faulty from birth. The other two thirds are sporadic: the dMMR change has happened only in the cancer, not in the rest of the body, usually because the MLH1 gene has been switched off by a process called hypermethylation.

Because a dMMR result has implications beyond the cancer itself — it may be the first clue to Lynch syndrome — pathology testing usually leads on to:

• A BRAF test and an MLH1 methylation test on the tumour, to work out whether the dMMR is likely sporadic or possibly Lynch.
• If Lynch is suspected, a referral to a family cancer clinic for genetic counselling and, where appropriate, blood-based germline testing.
• If Lynch is confirmed, screening and surveillance recommendations for relatives — see family screening for bowel cancer.

In Australia, testing the mismatch-repair status of every new bowel cancer is now routine. Two tests are used:

• Immunohistochemistry (IHC). The four mismatch-repair proteins are stained for on a slice of tumour. If one or more is missing, the cancer is dMMR. Done on the biopsy at colonoscopy or on tissue from surgery.
• Microsatellite instability (MSI) testing. A DNA test of the repetitive microsatellite stretches. A high level of instability (MSI-H) indicates dMMR. Used when IHC is borderline or unavailable, and increasingly built into broader molecular panels.

The result is in your pathology report. It is one of the first things the multidisciplinary team looks at when planning treatment.

The checkpoint-inhibitor drugs used for dMMR/MSI-H bowel cancer in Australia and internationally are:

• Pembrolizumab (Keytruda) — anti-PD-1 antibody. Given as a short intravenous infusion every 3 or 6 weeks.
• Nivolumab (Opdivo) — anti-PD-1 antibody. Given intravenously, sometimes combined with ipilimumab (anti-CTLA-4) for an extra-strong effect.
• Dostarlimab (Jemperli) — anti-PD-1 antibody. Featured in the early rectal-cancer organ-preservation trials. Currently used in trial settings and in specialist multidisciplinary discussion.

Infusions usually take 30 to 60 minutes. There is no take-home pump, no daily tablets, and no chemotherapy unit overnight stays. Many people drive home the same day. Treatment continues for as long as it is helping and being tolerated — for metastatic disease, often up to 2 years for a planned course, or longer if response is ongoing.

The role of immunotherapy in bowel cancer is expanding, but the established and emerging settings can be grouped as follows:

For the pMMR/MSS majority — which is roughly 95 per cent of metastatic bowel cancer — checkpoint inhibitors used alone have not been shown to improve outcomes, and chemotherapy remains the mainstay. Research into combinations that might make immunotherapy work in this group is ongoing, but at present it is not a routine option.

Most people tolerate immunotherapy infusions themselves well — there is no nausea, no hair loss, and no bone-marrow suppression of the kind chemotherapy causes. The side effects come from a different direction. By lifting the brakes on the immune system, immunotherapy can cause it to attack normal tissues. These reactions are called immune-related adverse events (irAEs).

Any organ can be affected, but the most common are:

• Skin — itch, rash, dryness. Usually mild.
• Bowel — diarrhoea, sometimes severe (immune colitis). Can develop weeks to months into treatment.
• Thyroid — under-active or over-active thyroid. Often picked up on routine bloods. May need lifelong thyroid hormone replacement.
• Liver — inflammation (hepatitis) — usually picked up on routine bloods.
• Lungs — inflammation (pneumonitis) causing breathlessness or new cough.
• Pituitary gland — leading to fatigue, low blood pressure, hormonal problems.
• Joints — arthritis-like pain and stiffness.
• Rarely — diabetes, heart inflammation, kidney inflammation, nerve and muscle effects.

One important point: immune-related effects can occur during treatment or weeks to months after the last dose. Mention to any doctor you see, for years afterwards, that you have had immunotherapy — it changes the way some symptoms need to be investigated.

Immunotherapy for bowel cancer is one of the most promising developments of the last decade — but it is not, at present, a general-purpose treatment. The honest summary:

• It works well for the dMMR/MSI-H subgroup — a small minority of bowel cancers. For these patients, it can produce responses that chemotherapy cannot.
• It does not work, on its own, for the pMMR/MSS majority. Research is active, but there is no current evidence to support replacing chemotherapy with immunotherapy for the 85 to 95 per cent of bowel cancers that fall into this group.
• It is not without cost. The immune-related side effects can be serious and occasionally permanent. The decision to use immunotherapy weighs the chance of response against this side-effect profile.
• Long-term follow-up is still being gathered. The dramatic early results of the dMMR rectal-cancer organ-preservation trials, in particular, need years more follow-up before they can be considered standard care.

If your cancer is dMMR/MSI-H, immunotherapy will be a substantial part of the conversation with your medical oncologist. If your cancer is pMMR/MSS, immunotherapy is unlikely to be part of your standard plan, but it may still be raised in the context of clinical trials. Either way, the mismatch-repair result on your pathology report is the key piece of information that determines this. Ask about it.