Patients with a type of liver cancer known as hepatocellular carcinoma (HCC) often face an anxious wait after their tumor is removed. In up to half of these people, the cancer will return within 2 years after surgery or a treatment that destroys their tumors with heat. Researchers haven’t identified any therapies that can stop it from coming back.

That could change thanks to a class of drugs once seen as a failed revolution in cancer therapy. Angiogenesis inhibitors, which throttle tumors by cutting off their blood supply, never measured up to expectations when they reached clinical trials more than 2 decades ago. But they are now getting a boost from a newer set of drugs with a more impressive track record: checkpoint inhibitors, which unleash the immune system’s T cells to attack tumors. At the American Association for Cancer Research (AACR) meeting last month, researchers presented evidence that the combination of an angiogenesis inhibitor and a checkpoint inhibitor delayed recurrence in patients with HCC, a first for this cancer type.

If the duo receives approval from the U.S. Food and Drug Administration (FDA), it will be the eighth pairing of the drug types OK’d in the past 4 years. More than 200 clinical trials are now testing this approach in various kinds of cancers, spurred by evidence that angiogenesis inhibitors help cancer-fighting T cells find their way deep into tumors. “There’s a very clear rationale” for combining these two classes of drugs, says Robert Kerbel, a tumor immunologist at the University of Toronto.

This resurgence for angiogenesis inhibitors comes after astronomical expectations for the drugs were dashed in the late 1990s. Their premise was compelling. Tumors, hungry for oxygen and nutrients, would have to stimulate new blood vessels to sprout and grow to them; by preventing these vessels from forming, angiogenesis inhibitors should starve tumors and curb their growth. But the drugs fell flat for various reasons, including tumors becoming resistant or using other mechanisms to obtain blood. “It was an almighty crash” when the drugs didn’t deliver, says University of Oxford tumor pathologist Francesco Pezzella, whose team discovered that tumors can commandeer existing vessels instead of growing new ones.

Still, angiogenesis inhibitors have recorded some successes. The monoclonal antibody bevacizumab gained regulatory approval in 2004 for patients with advanced colon cancer, and more than a dozen other angiogenesis inhibitors have since joined the antitumor arsenal. But on their own, angiogenesis inhibitors typically add only a few months to patients’ lives. They “are very good at controlling disease, but by themselves they are not curative,” says medical oncologist Brian Rini of Vanderbilt University Medical Center.

Scientists pair approved cancer drugs all the time to test whether they work better in concert, but a phenomenon discovered more than 25 years ago may explain why the checkpoint inhibitor-angiogenesis inhibitor partnership is productive. The blood vessels inside a tumor are a mess—bloated, twisting, and leaky. In the late 1990s, Rakesh Jain, a tumor biologist at Harvard Medical School (HMS), noticed that angiogenesis inhibitors had a surprising effect on tumors. The drugs “normalized” the vessels, spurring them to narrow, straighten, and become less porous. “What I saw was that the blood vessels [inside tumors] were getting better,” says Jain, who published his “vascular normalization” hypothesis in 2001. Although controversial at first, the explanation is now widely accepted, says Oxford cancer biologist Anette Magnussen.

Abnormal blood vessels leave the interior of a tumor low on oxygen, inhibiting any T cells that enter to attack the tumor cells. That handicaps checkpoint inhibitors, which work by keeping cancer cells from flipping the inhibitory switches on T cells. By restoring normal circulation in the tumor, angiogenesis inhibitors may reverse the immune-suppressing conditions and permit tumor-targeting T cells to pounce. Jain says normalizing tumor vessels may provide another benefit—reining in metastasis, the spread of tumor cells to other parts of the body. An oxygen-starved tumor “is like a wounded tiger,” he says. It is more dangerous in that state because it is prone to release cells that can relocate and establish new tumors elsewhere.

So far, dozens of clinical trials have evaluated checkpoint inhibitor-angiogenesis inhibitor combos. Some proved toxic, and others failed. However, the studies also led FDA to approve new therapies for types of liver, kidney, lung, and endometrial cancers. The drug combinations are not curative, but they restrain tumor growth. And some extend patients’ lives by many months over angiogenesis inhibitors alone.

Trials have revealed other promising results as well, including those for HCC, which liver surgeon Pierce Chow of Duke-NUS Medical School in Singapore presented at the AACR meeting. In a phase 3 trial, he revealed, a combination of bevacizumab and the checkpoint inhibitor atezolizumab cut the odds that tumors would return after surgery or heat treatment by 28% in patients whose disease was diagnosed early. Researchers are still following the patients to determine whether these benefits persist and the combination increases survival.

Scientists are not just mixing and matching existing options. At least one startup, DynamiCure, is looking to develop new ­angiogenesis-thwarting drugs that are safer and can be coupled with checkpoint inhibitors. The company has already started a clinical trial of one antibody that stimulates vessel normalization. “When new companies are founded, that tells you the field is not dead, it’s blossoming,” Jain says.

Still, the strategy of combining the two drug types faces many challenges. For one thing, most angiogenesis inhibitors directly or indirectly stymie vascular endothelial growth factor (VEGF), a molecule that incites blood vessel growth. However, normal tissues also require VEGF, and the drugs can cause side effects such as bleeding, high blood pressure, and strokes.

Beyond side effects, researchers need better ways to determine whether tumors are responding to the drug duos, Magnussen says. “The challenge for using this [strategy] in humans is how you monitor blood vessel normalization in a cost-effective and time effective way.”

HMS cancer biologist Dan Duda worries that researchers and pharmaceutical companies are rushing drug combinations into trials without considering the type of cancer and variables such as timing and schedule of treatments. “We are making the same mistakes we made the first time” with angiogenesis inhibitors, he says, adding, “We should take our time and think a little bit.”