An experimental immune therapy temporarily wiped out tumors in the brain and spinal cord of a patient with recurrent glioblastoma, one of the most lethal of all cancers.

Researchers at the City of Hope in Duarte, Calif., genetically engineered 50-year-old Richard Grady's own T cells to attack his tumors, which shrank and disappeared over 7.5 months. Although the Seattle man's cancer then recurred, he remains alive 15 months after he was told he had only weeks to live.

The case is reported this week in the New England Journal of Medicine. Earlier this month, the journal published another remarkable but solitary case: National Cancer Institute researchers isolated rare, tumor-penetrating T cells from a metastatic colon cancer patient, then multiplied and marshaled them to target the gene mutation driving her disease — a formidable mutation previously called "undruggable."

While individual experiences usually do not carry much scientific weight, immunotherapy researchers say these cases bolster hopes that personalized T-cell treatment can work on solid malignancies, including the deadliest types.  Experimental therapies using these immune soldier cells have been highly effective only in blood cancers and melanoma.

"I think there has been a lot of skepticism about using T cells in solid tumors," said City of Hope cancer immunologist Christine E. Brown, lead author of the glioblastoma report. "Even one case is important for the field."

Stephen Gottschalk, a pediatric oncologist involved in T-cell research at Baylor College of Medicine in Houston, agreed. "The overall sentiment is that T cells don't work for solid tumors," he said. "This is one of the first examples where a patient with disseminated glioblastoma responds. I think it's quite dramatic."

Mark Gilbert, chief of the neuro-oncology branch at the National Cancer Institute, reacted more cautiously: "Whether this one case translates into clinical progress or is just an anomaly based on some unique aspect of this patient's cancer, time will tell. It's obviously very interesting."

Many questions remain about the best molecular targets for T cells, how to avoid toxic side effects, and the most effective way to give the therapy.

City of Hope has treated 10 patients with repeated injections of the bioengineered T cells through a catheter into a brain tumor cavity, the hole left by surgical removal of a tumor.

That protocol did not stop new tumors from appearing in Grady's brain as well as his spine. (His cancer was originally diagnosed in November 2014 and recurred just six months after surgery, chemotherapy, and radiation.)

So the City of Hope doctors modified their protocol, using an intracranial catheter to inject T cells into the cerebrospinal fluid that bathes the brain and flows down into the spinal cord.

Grady received 16 T-cell infusions. Seven tumors in his brain and eight in his spine ultimately became undetectable on imaging scans. About halfway through the 7.5 months of treatment, he felt so good that he was gradually weaned from a powerful steroid — used to reduce brain swelling as the cancer grows — and  "returned to normal life and work activities," Brown and her co-authors wrote.

To put Grady's case in context, the median survival for recurrent glioblastoma patients is three to six months, Brown said.

The researchers also said it was "particularly noteworthy" that Grady suffered relatively minor side effects — fever, fatigue, body aches — because many T-cell trials have been hampered by severe, even fatal, immune system overstimulation and toxic neurological effects. They now plan to modify their study so more patients can get the dual T-cell infusions.

Despite Grady's manageable side effects, University of Pennsylvania neurosurgeon Donald M. O'Rourke questioned using such an invasive and risky delivery approach.

O'Rourke is leading a study that has treated 10 glioblastoma patients with a single large dose of T cells given through a vein. The cells penetrated the protective barrier of the brain — an insurmountable obstacle for most conventional drugs — and infiltrated tumors without serious side effects. But, as with almost all pilot studies, it is too soon to tell if the treatment is effective or can be enhanced to become so.

"We're hopeful we can build on this," O'Rourke said.

Gottschalk said Baylor researchers also had seen "encouraging" signs in a T-cell trial for glioblastoma.

The trials at City of Hope, Baylor, and Penn are each targeting a different molecular marker, or antigen, on brain tumor cells.

The City of Hope's target antigen is active in about 60 percent of glioblastoma patients, studies suggest. The antigen was not uniformly present, however, and was only moderately abundant on Grady's tumor cells.

"One of our questions," Brown said, "is what level of antigen we need to see responses" to the T cells' attack.

Another question is whether T cells can be programmed to attack multiple antigens at once, because aggressive, fast-mutating cancers like glioblastoma can quickly evolve to make less of the telltale marker. Molecular tests indicated that was what happened to Grady, accounting for his amazing but transient remission.

What is a T cell?

A T cell, or T lymphocyte, is a type of white blood cell that plays a crucial role in the immune system. T cells scan the intracellular environment, then target and destroy invaders such as viruses by homing in on their distinctive molecules, called antigens. A new way to fight cancer involves heightening T cells' ability to recognize and hunt down the body's own renegade cells. To do this, T cells are separated from the patient's blood, genetically engineered to recognize an antigen found on malignant cells, then returned to the bloodstream to mount an attack.