Creutzfeldt-Jakob disease (CJD) causes a characteristic pattern on MRI that can distinguish it from other rapidly progressing dementias not caused by prions, according to a new study in Neurology.
“Many radiologists are not familiar with these findings, so it is imperative that neurologists read their MRIs on their own, and use the knowledge from this study to interpret them,” according to lead author Michael Geschwind, MD, PhD, a neurologist at the Memory and Aging Center and associate professor of neurology at the University of California-San Francisco (UCSF).
The pattern is a relatively simple one, comprising a higher signal on DWI (diffusion-weighted imaging) than FLAIR (fluid attenuation inversion recovery) in areas classically associated with CJD, including cingulate, striatum, and multiple neocortical gyri, in the context of clinical cues, suggests that CJD is a possible diagnosis. Previous criteria indicated only that CJD was characterized by DWI and FLAIR hyperintensity, or DWI hyperintensity alone.
A major benefit of these new criteria, Dr. Geschwind said, is that they allow the neurologist to distinguish between CJD and other disorders that may present similarly, including ones for which good treatments are available.
“CJD can be hard to diagnose, because it can look like a lot of other neurologic and psychiatric illnesses,” he said. While the classical presentation of rapidly progressive dementia with ataxia is a clear pointer to CJD, “people don't always present classically, and even when they do, there are other things in the differential,” including encephalitis, paraneoplastic or non-paraneoplastic autoimmune disorders, and rapid presentations of other neurodegenerative disease, such as Alzheimer disease and Parkinson disease.
CJD can be misdiagnosed as one of these other conditions, or the mistake can be made in the other direction. In either case, there may be weeks to months of uncertainty, and potentially lost treatment opportunity, until the correct diagnosis is made.
Diagnostic criteria for CJD from the World Health Organization date back to 1998, “but those criteria are really designed for epidemiologic purposes, to look back after death if there is no autopsy,” because they include features not seen until late in the disease. “They are not really designed for diagnosing the patient you are seeing right in front of you,” Dr. Geschwind said.
Recent studies of spinal fluid markers indicate some promise for earlier diagnosis, “but they are still not as good as MRI.”
To refine and test the MRI criteria, Dr. Geschwind had two neuroradiologists blinded to diagnosis examine images from 90 patients, including 29 with non-prion rapidly progressing dementias (npRPD), 48 with sporadic CJD, 6 with familial CJD, and 7 with another genetic prion disorder, Gerstmann-Straussler-Scheinker (GSS) disease. Scans included FLAIR and DWI images, and an ADC (apparent diffusion coefficient) map. All the patients in the study were referred to the Center as having CJD. “I think this is a better control group” than previous studies, Dr. Geschwind said, with controls who were not initially misdiagnosed.
Each radiologist classified each patient as definitely, probably, probably not, or definitely not CJD, using the proposed criteria. Their diagnoses were compared to autopsy results for most cases, or, absent that, diagnoses were based on extensive clinical examinations.
“Pathologically proven CJD is the ultimate diagnosis, but in ten years at our Center, we have never diagnosed someone as having CJD that didn't have CJD at autopsy. If we call someone CJD, we are as about as positive as you can be.”
The results identified four distinguishing MRI features. “I think everyone should look at the MRI with these criteria in mind,” Dr. Geschwind said.
First, in sporadic CJD, hyperintensity on DWI was greater than on FLAIR, while the opposite was true for npRPD. Second, in those sporadic CJD patients with subcortical DWI hyperintensity, the ADC map in those regions was always hypointense, indicating restricted diffusion. Third, limbic involvement alone was never seen in sporadic CJD, but was often seen in npRPD. And fourth, sporadic CJD showed characteristic DWI patterns of gray matter involvement, with specific gyri more likely to be involved.
The higher DWI intensity in CJD may be related to the pathology, although more work must be done to confirm that correlation. “We have looked at what is causing these diffusion-related changes,” by comparing the MRI to the autopsy findings, Dr. Geschwind explained. “The MRI changes seem to correlate pretty much equally with vacuolation and deposition of prions.” Since the vacuoles are filled with fluid, it makes sense they would restrict water movement, he said.
Dr. Geschwind noted that “vacuolation” is a more accurate term than the older “spongiform” description of gross pathology. He has also begun to reverse the order of the eponyms to “Jakob-Creutzfeldt.” Creutzfeldt described one patient, while his student Jakob described five others. “We still have the brains, and Creutzfeldt's patient didn't have CJD, while two of Jakob's patients did. It really should be Jakob disease,” he suggested.
While the MRI can be crucial in differentiating CJD from non-prion dementias, the scan alone will only take you so far, Dr. Geschwind cautioned. “Since we completed study on that cohort, we have found patients who don't have CJD, but have almost identical MRI findings,” including one with extrapontine myelinolysis, and another with hyperglycemia and uncontrolled seizures. In each case, as the patient improved, the MRI abnormalities abated.
“Even using what we think are pretty darn good criteria, we are still going to be fooled in some cases if we just relied on MRI, so it's important to have other factors in the criteria,” most importantly the history, Dr. Geschwind added.
The most important value of the criteria, he said, is that some of the conditions misdiagnosed as CJD are instead disorders in which patients may stabilize or even recover, if they get the right treatment. “I think we are missing some of the diagnoses in these patients.”
“The neurologic community is badly in need of better ante mortem diagnostic modalities for CJD, and this study assists in shedding some light on that,” said James Sejvar, MD, a neuroepidemiologist at the National Center for Emerging and Zoonotic Infectious Diseases, part of the Centers for Disease Control and Prevention in Atlanta, GA. “There has been quite a bit of work in recent years on using some of these more advanced radiologic modalities for prion disease, and I think this an important contribution to that overall understanding.”
“One does not render a diagnosis of CJD lightly, because of the catastrophic nature of it. So being able to differentiate early what may be an alternative diagnosis, and one that may be treatable, is critical, probably the most critical thing in terms of diagnosis of CJD. I think this study is going to alert the neurologic community to the potential utility of MRI in the diagnosis, and that should make it even easier to gather additional data to substantiate these findings.”