Eli Lilly and Co. and Merck KGaA’s Erbitux, approved for advanced colon cancer, failed to slow tumors in a study designed to expand the medicine’s use to patients whose disease is in an earlier stage.

Erbitux, used to treat aggressive colorectal tumors that have reached other areas of the body, didn’t help stop the spread when added to chemotherapy, according to a study reported Sunday at the American Society of Clinical Oncology meeting in Chicago. Erbitux added to the side effects of chemotherapy.

The finding is the latest of at least three studies that have narrowed the scope of a drug that was the first of its kind for colon cancer when approved in 2004. Last year, Erbitux’s prescribing information was changed to say that patients whose tumors have a mutation to a gene called KRAS, found in about 40 percent of colon cancer cases, aren’t helped by the treatment.

“Unfortunately with cancer, things that are hoped to work don’t always end up working,” said Herbert Hurwitz, associate professor of medicine at Duke University Medical Center in Durham, North Carolina, who wasn’t involved in the study. “This was a well-run study that asked a good clean question and got a good clean answer.”

Erbitux had sales of about $1.4 billion last year, according to IMS Health, a drug research company. It continues to be a useful drug for patients with tumors that have spread, said Hurwitz, who spoke in a telephone interview. In those patients, Erbitux competes with Roche Holding AG’s Avastin, which has shown similarly poor performance in studies for mid-stage colon cancer that hasn’t spread.

“From a market-share perspective, this would have been a big group of new patients and would have distinguished it from” Avastin, Hurwitz said. “Most people would probably have expected this drug to work as well or better” in earlier stages as later stages, he said in a phone interview.

The trial involved about 1,800 participants. About half were given Erbitux along with a chemotherapy called Folfox, while 858 patients were given Folfox alone.

After a follow-up of 16 months, both groups had a similar number of patients who didn’t show signs of their cancer returning. Patients taking Erbitux were actually slightly less likely to survive, with 82 percent still alive at follow-up compared with 87 percent taking chemotherapy alone.

About 147,000 Americans are diagnosed with colon cancer every year, making it one of the most common forms of tumors, according to the National Cancer Institute.

About 25 percent of patients are first identified with so-called metastatic cancer, which means the disease has already spread to multiple areas of the body, said Steven Alberts, lead researcher on today’s study and professor of oncology at the Mayo Clinic College of Medicine in Rochester, Minnesota.

Another third of colon cancer patients are diagnosed with stage 3 tumors, meaning the cancer has begun to reach nearby lymph nodes but haven’t yet established themselves in new regions of the body, Alberts said. The Erbitux study was aimed at stage 3 colon cancer patients.

“Maybe it didn’t live up to some of the initial hopes or expectations, but at the end of the day it still is an important drug that clearly is benefiting a proportion of patients,” Alberts said in an interview at the conference. “It is a meaningful drug in the metastatic setting; just not in the earlier setting.

“The hard part is to figure out why it happened. That’s the focus of our research now,” Alberts said.

Erbitux is a man-made antibody, a substance naturally produced by the immune system in response to infection. The drug works by latching onto cancerous cells and blocking replication. The flaw in the KRAS gene, which holds the blueprint for part of a messaging system inside the cell, is thought to render Erbitux ineffective.

As many as 40 percent of patients with colorectal cancer have the mutation, according to previous research. Among those without the gene mutation, only about a third to half of patients respond to treatment, Alberts said.

In September, a study known as the COIN study found that Erbitux failed to improve the outcomes even for its target group of metastatic colon-cancer patients with normal KRAS genes. Researchers are still examining results from that study to determine why they contradicted other studies. Part of the difference may be due to a different type of chemotherapy and lower doses of treatment used in the COIN study, Alberts said.

Lilly, based in Indianapolis, and New York-based Bristol Myers market Erbitux in the U.S., while Merck KGaA, of Darmstadt, Germany, sells it elsewhere. It’s approved for use in colorectal as well as head and neck tumors. The European Union in July rejected the drug as a treatment for a form of lung cancer.