Glutamic acid decarboxylase antibodies may suggest underlying cancer

NEW YORK - High titers of glutamic acid decarboxylase (GAD) may suggest the presence of an underlying cancer, particularly in the setting of classic paraneoplastic neurological syndrome (PNS), according to a new study.

"In some patients the presence of GAD antibodies indicates a paraneoplastic origin, particularly those who are older than 50, males, or that have concomitant antibodies against GABA receptors," Dr. Francesc Graus from the University of Barcelona, Spain, told Reuters Health by email. "Underlying tumors are not only thymomas, as previously described, but also lung cancer and tumors of different sites with a neuroendocrine histology."

High levels of GAD antibodies are seen in stiff-person syndrome and in some patients with other neurological conditions; cancers have sometimes been found at the time of neurological diagnosis in some patients with elevated GAD antibodies.

Dr. Graus's team described the PNS and tumor types associated with GAD antibodies in a retrospective study of 15 patients with definite or possible PNS examined between 1995 and 2013. They compared these patients with 106 patients with GAD antibodies who did not have PNS.

Among the patients with PNS, six had lung cancer, four had neuroendocrine tumors (two pancreatic and two thymic carcinoids), two had thymoma, two had breast cancer, and one had non-Hodgkin lymphoma, the researchers report in JAMA Neurology, online June 22.

In most patients (10/15) the neurological syndrome predated the diagnosis of cancer, with a median delay of 2.7 months between the diagnosis of the neurological syndrome and the associated cancer.

At the latest follow-up, eight patients had clinically progressed (including five that died as a result of PNS), four were stable, and three had neurological improvement.

Serum and cerebrospinal fluid (CSF) levels of GAD antibodies did not differ significantly between patients with and without PNS.

All three tumors that were tested showed expression of GAD.

Compared with the 106 patients with nonparaneoplastic GAD-antibody-associated disorders, patients with PNS were significantly older, more frequently male, were more likely to have coexistent neuronal cell-surface autoantibodies, and were more likely to present with classic PNS than with stiff-person syndrome or cerebellar ataxia.

"If your patient has GAD antibodies and a neurological syndrome suggestive of a paraneoplastic origin, particularly limbic encephalitis, a tumor search should be done," Dr. Graus said.

"Considering the tumors identified in the current series and previous reported cases," the researchers write, "the tumor workup should include mammogram and chest CT or CT-PET scan, depending on the clinical setting."

"These patients should be reevaluated in six months," Dr. Graus added. "With the new techniques, that are very sensitive (PET-CT), no further studies are probably needed."

"We must stress that this study was done in patients who have high titers of GAD antibodies in the range usually found in patients with stiff person syndrome," Dr. Graus cautioned. "The finding of low titers is not uncommon and it is probably irrelevant."

"GAD antibodies are worth being studied as part of a broad antibody search in patients with suspected autoimmune or paraneoplastic neurological disorders," writes Dr. Christian G. Bien from Epilepsy Center Bethel, Krankenhaus Mara, in Bielefeld, Germany, in a related commentary.

"These antibodies may pave the way to the identification of a tumor," he notes. "Even if a patient does not have a paraneoplastic syndrome, GAD antibodies have a sufficient syndromic specificity to terminate an extensive diagnostic process if found in a patient with a typical GAD antibody syndrome."

Dr. Jerome Honnorat from Hopital Neurologique in Bron, France, said, "With the results of this article, we now need to add anti-GAD antibodies in the list of antibodies that must be tested in patients suspected of PNS."

The precise frequency of PNS with anti-GAD antibodies cannot be gleaned from the current study, he told Reuters Health by email, though it "is probably very, very rare among all the patients with GAD antibodies and neurological syndrome."

"A prospective study would be necessary," Dr. Honnorat said.


JAMA Neurol 2015.

References: Reuters Health
comments powered by Disqus