Determining when and how to use neuroimaging for migraine is an important issue that physicians face when seeing the 40 million Americans with migraine. Neuroimaging for suspected migraine is used for many reasons, including exclusion of secondary conditions that mimic migraine. Other reasons include medicolegal issues, busy practice conditions in which tests are ordered as a shortcut, providing neuroimaging to appease patient requests, and addressing concerns and expectations of referring clinicians.

Recommendations on the role of neuroimaging in diagnosing headache vary by specialty. An American Academy of Neurology (AAN) evidence‐based review published in 2000 gave a Grade B recommendation indicating that neuroimaging is not usually warranted for patients with migraine and normal neurological examination. A Grade C recommendation from AAN states that a lower threshold for neuroimaging may be applied for patients with atypical headache features or patients who do not fulfill the strict definition of migraine or have some additional risk factor.

The American Headache Society’s (AHS) “Choosing Wisely in Headache Medicine” campaign advises against performing neuroimaging studies in patients with stable headaches that meet criteria for migraine. The American College of Radiology’s “Choosing Wisely” campaign also recommends against performing imaging for uncomplicated headache. “This is an important issue to address because deciding on neuroimaging is a daily issue for clinicians who manage migraine,” says Randolph W. Evans, MD.

Addressing an Unmet Need

Dr. Evans and colleagues at AHS conducted a systematic review, published in Headache, in which the authors gathered evidence about the diagnostic utility of neuroimaging—specifically MRI and CT—in adults seeking outpatient treatment for episodic, chronic, and progressive migraine and in migraineurs with and without aura. For the analysis, 23 articles met inclusion criteria and were included in the final review. Articles were included in the study if they evaluated adults 18 and over who sought outpatient treatment for any type of migraine and who underwent neuroimaging.

“Incidental findings from neuroimaging are common in both migraineurs and control patients,” Dr. Evans says. “People who meet ICHD-3 criteria for migraine are no more likely to have significant neuroimaging findings than the general healthy population. We gave a Grade A recommendation stating that it is not necessary to do neuroimaging in patients with headaches consistent with migraine who have a normal neurologic examination.” The systemic review noted that neuroimaging may be considered for presumed migraine for a myriad of other reasons, but this was only given a Grade C recommendation due to little or no literature support (Table).

Proceeding With Caution

Patients with concerning clinical or exam features may have abnormalities that require attention and should still be imaged, but Dr. Evans says to proceed with caution when considering neuroimaging in other situations. “Even when the yield is low, neuroimaging is often done in migraineurs because of patient expectations or concerns from patients and their family,” he says. “Unfortunately, even when physicians follow guidelines, they are not indemnified in medical malpractice lawsuits when the rare migraine patient has significant incidental pathology.”

Rather than routinely neuroimaging patients with migraine, an alternative approach is to educate patients about the low yield of neuroimaging. Patients should also be reassured that neuroimaging can be performed at a later date if new signs or symptoms develop. “In addition, it is important to recognize that if a scan is performed, MRI is preferred over CT,” Dr. Evans adds. “MRI without contrast is usually adequate for an initial study unless there is a specific indication for contrast, such as concern for neoplastic disease or spontaneous intracranial hypotension.”

Although there are no plans for an update to this systematic review in the near future, Dr. Evans says a large prospective or retrospective study of the incidence of significant pathology in cluster headaches would be of interest in future research.