The Impact of Metformin on Survival Outcomes in Patients with Comorbid Cancer and Type 2 Diabetes Mellitus

Type 2 diabetes mellitus (T2DM) is a common condition that affects hundreds of millions of people worldwide. It occurs when the body does not use insulin properly, leading to high blood sugar. Insulin is a hormone made by the pancreas that helps move sugar from the blood into the body’s cells, where it can be used for energy. Many people with T2DM take metformin, a safe and affordable medication that lowers blood sugar and improving how the body responds to insulin. Because of these benefits, it is one of the most widely prescribed drugs for diabetes. In recent years, researchers have discovered that metformin may have health benefits beyond diabetes. It may help with obesity (excess weight that increases the risk of illness) and may help protect against heart disease.

There is also interest in whether metformin might improve outcomes for people who have both cancer and diabetes. Many patients live with both conditions, which makes treatment decisions more complex. Some studies have suggested that patients taking metformin may live longer, but the results have not been consistent. Some of these studies may be misleading because of a problem called immortal time bias.

In cancer clinical trials, patients are not randomly assigned to "metformin use" or "non-use" at the start of the trial. Instead, metformin is prescribed by physicians during the trial follow-up visits when medically needed to treat diabetes. Because patients must remain alive long enough to reach the follow-up visits and start the medication, certain study designs may unintentionally make the treatment appear more beneficial than it truly is. This issue is known as “immortal time bias.”

Before planning this research, we reviewed 125 studies on metformin and survival in cancer patients with T2DM. About 55 studies that did not deal properly with immortal time bias reported stronger benefits. However, many of those studies used insurance records or summary-level results rather than detailed information for each person. That means it remains unclear whether the reported benefits were real or caused by bias or other differences between patients.

We will use individual participant data from clinical trials shared through Vivli. Clinical trials are carefully run studies in which treatments, medicines, and follow-up visits are recorded in detail. These data will let us see exactly when a person was diagnosed, when they started metformin, and what happened to their cancer over time.

We will combine individual patient data from multiple trials to study the true relationship between metformin use and survival in cancer patients with diabetes. We will also test statistical methods that can reduce the impact of immortal time bias, so that our results are not exaggerated or misleading.

This research is important because it will clarify whether metformin has real benefits for cancer patients with diabetes. The findings may guide doctors in making better treatment decisions and highlight the importance of using correct methods to avoid bias in future medical studies.