When researchers talk about the new, mostly experimental form of cancer treatment known as immunotherapy, they often use glowing terms like revolutionary and transformative.
Last week's PBS documentary Cancer: The Emperor of All Maladies was another example of this. The third and final episode told how many scientists believe that the body's innate defense system is the only weapon adaptable enough to conquer the mutating malignant cells.
But harnessing the immune system to launch a self-attack can be highly toxic, even deadly. In 2013, two immunotherapy pioneers - the University of Pennsylvania's Carl June and the National Cancer Institute's Steven A. Rosenberg - each published a report of fatal side effects in an early clinical trial of bioengineered T cells, the soldiers of the immune system. The documentary prominently featured both men, but did not mention the deaths.
So how promising is immunotherapy, and how do its potential dangers compare with conventional treatments such as chemotherapy and radiation?
"We have tried to learn from the early days of gene therapy and avoid the hype. A lot of times, the patients are doing it. They become zealots," June said. "What I say in general is, it's the early days, so we don't know everything. We need to have decades of observations. But so far, the toxicity [of immunotherapy] has been less" than conventional oncology weapons.
The immune system is highly tolerant of cancer because, unlike infection, cancer arises from the body's own healthy tissue. Only in the last 50 years has molecular genetic technology enabled precise manipulation of immune cells, and only in the last five years has that manipulation become remarkably effective.
Consider that in 2007, the Food and Drug Administration delayed approval of the first immunotherapy - Provenge, the prostate cancer vaccine - pending more proof of effectiveness. Yet in the last three years, the agency has fast-tracked approval of immunotherapies such as Yervoy and Keytruda that work by taking the brakes off the immune system.
The FDA has given "breakthrough" designation - intended to expedite approval - to the T cell therapy that Penn, Children's Hospital of Philadelphia, and Novartis Pharmaceuticals are developing. (Their work is moving ahead without the cloud of litigation, due to a patent-dispute settlement announced this week.)
At least 40 companies are conducting clinical trials of immunotherapies, and researchers keep finding new ways to selectively activate the mind-bogglingly complex immune system. This week, for example, a team from Washington University School of Medicine in St. Louis reported a new way of identifying protein markers, or antigens, that are unique to cancer cells and likely to signal T cells to attack. The approach involved sequencing the entire genetic code of patients' tumor cells and healthy cells, then comparing the genomes to find the telltale mutated proteins of cancer.
"In the first three patients, the vaccine turned on the T cells and was safe," said Gerald P. Linette, who was part of the team. "Right now, it's tedious, but as the technology improves, it's going to become easier, cheaper, and faster."
Finding antigens, or combinations of antigens, that are on cancer cells - but not healthy cells - has been a big challenge. The Penn immunotherapy, for example, engineers T cells to target an antigen found on B cells, which turn malignant in certain blood cancers. The T cells also kill healthy B cells, but fortunately, patients can survive without B cells.
In the trials with fatal side effects, T cells were engineered to target an antigen called MAGE-A3.
Sophisticated preclinical studies - computer analyses, cell cultures, even tests in animals - suggested that the target was safe.
But it wasn't.
T cells engineered by June's team were given to a myeloma patient at Penn and a melanoma patient at Washington University, where Linette was a coinvestigator. Both patients developed cardiac problems and died within several days as the cells destroyed their hearts.
In the other trial, led by the National Cancer Institute, the MAGE-A3 T cells attacked the brains of a patient with melanoma and another with esophageal cancer; they suffered mental changes, lapsed into comas, and died.
That antigen target has been abandoned, and researchers are redoubling their efforts - with studies like the one at Washington - to predict and prevent off-target effects.
"It's always about balancing risks and benefits," Linette said. "It's very difficult. But we can't remain still. We have to make progress."
As new targeted drugs and immunotherapies become more common, they are giving patients a host of side effects that range from the embarrassing to the life-threatening. Doctors, nurses, and patients talk about how they are coping.