A patient in their 6^th decade of life with obesity and episodic migraines with aura, presented to hospital 10 hours after last seen normal with acute left pyramidal hemiparesis, hypertension, severe migraine headache, emesis, and photophobia. The National Institute of Health Stroke Scale score was 7 (left facial droop, arm and leg weakness, sensory loss [light touch to the left leg], dysarthria, and mild drowsiness). Facial sensation was sparred. The patient followed commands, cranial nerve function was intact, tongue was midline, and there was a left extensor plantar response.

Computed tomography (CT) angiography (CTA) did not reveal evidence of ischemia or hemorrhage, intracranial or extracranial stenosis or occlusion, reversible cerebrovascular constriction syndrome, aneurysm, dissection, or venous sinus thrombosis. Aspirin monotherapy was initiated.

On the subsequent day, the patient developed increasing burning pain over the torso and back, worsened left hemiparesis, new tremulousness in the right lower extremity, and bilateral extensor plantar responses. Magnetic resonance imaging (MRI) revealed a medial medullary infarct (Figure 1). At this juncture, clopidogrel was added, but the patient continued to progress to bilateral lower extremity weakness and increasing pain. On day three of admission, the patient developed new dyspnea, only had flickers of movement in the left lower extremity, and the right lower extremity no longer had antigravity strength. The patient subsequently developed respiratory distress concerning for central neurogenic breathing and required intubation.

Figure 1.

Magnetic Resonance Imaging sequences on day 2 of hospital admission demonstrating a small bilateral medial medullary infarct (blue arrow) on (A) Diffusion Weighted Imaging (DWI) and (B) Apparent Diffusion Coefficient (ADC) affecting the bilateral anterior medullary pyramidal tracts, right more than left.

On the fourth day of admission, an MRI was repeated (Figure 2), demonstrating a larger area of restricted diffusion in the bilateral medial lower medulla resembling an “air pod sign” and consistent with an acute bilateral medial medullary infarction.^{Reference Agarwal, Raz and Yaghi1} Furthermore, susceptibility weighted angiography (SWAN) revealed a focus of susceptibility inferior to the vertebrobasilar confluence – suggestive of intraluminal thrombus in the anterior spinal artery (Figure 2C). A repeat CTA was unremarkable. Given ongoing clinical deterioration, ticagrelor was initiated in place of clopidogrel. Not knowing the patient’s CYP2C19 allele status, we wanted to pursue an antiplatelet therapy that did not depend on CYP2C19 metabolism.^{Reference Wang, Meng and Wang2} Furthermore, the noted intraluminal thrombus in the anterior spinal artery motivated us to additionally initiate intravenous heparin. Despite maximal medical therapy, in the days that followed, the patient remained conscious but became quadriplegic with preserved extraocular movements. There was no noted nystagmus or contrapulsion on extraocular movement assessment. Considering the severity of anticipated disability, in keeping with the patient’s wishes, care was transitioned to a comfort-based approach.

Figure 2.

Magnetic Resonance Imaging sequences on day 4 of hospital admission demonstrating an evolved bilateral medial medullary infarct (blue arrow) on (A) Diffusion Weighted Imaging (DWI) and (B) Apparent Diffusion Coefficient (ADC) in the appearance of an air-pod sign. (C) Susceptibility Weighted Angiography (SWAN) revealing evidence of thrombus (orange arrow) within the anterior spinal artery.

The first descriptions of medial medullary syndromes were more than a century ago and outlined clinical manifestations of unilateral medial medullary syndromes with triads of unilateral hemiparesis, tongue weakness, and sensory loss.^{Reference Pergami, Imbesi, Ceroni and Simonetti3} Bilateral medial medullary syndrome is rarer, and its recognition is challenging and of paramount importance for vascular neurologists, as offering thrombolysis to those presenting early may be essential to mitigating its high morbidity and mortality.

There are two especially striking clinical features about this case: first, the progressive clinical course, starting with unilateral weakness progressing to bilateral weakness, and second, the emergence of a central neurogenic pain syndrome and respiratory distress. The traditional dogma that vascular etiologies present with sudden and maximal neurologic deficit at onset is challenged here. Considering the literature leads us also to propose that the recognition of early respiratory distress and pain are essential clues, but these are widely underappreciated features of bilateral medial medullary syndrome.

One of the largest single-center cohort studies of medial medullary stroke syndromes included 86 consecutive patients.^{Reference Kim and Han4} Overall, 12/86 (14.0%) participants had bilateral medial medullary infarcts, which were more likely to extend dorsally from the ventral surface compared to unilateral medial medullary infarcts.^{Reference Kim and Han4} Interestingly, 21/86 (24.4%) developed a central post-stroke pain syndrome, and most described this as numb, cold or painful, not burning.^{Reference Kim and Han4} While medial medullary involvement would most classically impair medial lemniscal sensory pathways over spinothalamic pathways, cohort studies have described disturbed pain and temperature sensation in medial medullary syndrome that has been posited to arise from involvement of the spinoreticulothalamic pathway, which may regulate spinothalamic systems.^{Reference Kim and Han4,Reference Bassetti, Mattle and Bernasconi5} A compromised medial medullary reticular formation may therefore aberrantly influence nociception.^{Reference Bassetti, Mattle and Bernasconi5} In a cohort of medial medullary stroke patients, the most common presenting symptoms included motor (91%), sensory (73%), dysarthria (63%), and vertigo (59%).^{Reference Kim and Han4} Headache (10%) and tongue weakness (3%) were less frequent, and there was no respiratory distress in this cohort.^{Reference Kim and Han4} Interestingly, those with central post-stroke pain were more likely to have worse outcomes (mRS ≥ 3) than those without.^{Reference Kim and Han4} A systematic review of exclusively bilateral medial medullary stroke syndromes identified 38 patients presenting with bilateral motor paresis (64.9%), dysarthria (48.6%), and hypoglossal paresis (40.5%).^{Reference Pongmoragot, Parthasarathy, Selchen and Saposnik6} In contrast to the 2% of lateral medullary infarcts and 5% of unilateral medial medullary infarcts previous reported,^{Reference Kameda, Kawanami and Kurita7} 24.3% of bilateral medial medullary infarcts in this review developed respiratory failure.^{Reference Pongmoragot, Parthasarathy, Selchen and Saposnik6} The mechanism of respiratory failure, which is more common in those with bilateral medial medullary syndrome compared to unilateral, has been posited to relate to the extensive corticospinal tract involvement leading to impairment of voluntary respiratory control.^{Reference Bassetti, Mattle and Bernasconi5} There may be an additional influence of autonomic respiratory compromise from peri-ischemic impairment of the medial medullary reticular formation and more lateral structures such as the nucleus ambiguous and the nucleus of the tractus solitarius.^{Reference Bassetti, Mattle and Bernasconi5} Death occurred in 5/38 (13.2%). Overall, most descriptions are sudden onset and, thus, the progressive decline over three days noted in our case is noteworthy.

With respect to etiology, bilateral medial medullary infarcts are more likely to be associated with branch atheromatous disease compared to unilateral infarcts, which have a greater association with small vessel disease.^{Reference Kim and Han4} The most commonly reported cerebrovascular imaging findings include vertebral artery atherosclerosis (38.5%), occlusion (15.4%), basilar atherosclerosis (19.2%), and, remarkably, 38.5% had normal appearances of the vertebral, basilar, and anterior spinal artery.^{Reference Pongmoragot, Parthasarathy, Selchen and Saposnik6} Another important consideration is that the air-pod sign may be an indicator of a cerebrovascular anatomic variant that increases predisposition to bilateral medial medullary infarction (a type II anterior spinal artery,^{Reference Agarwal, Raz and Yaghi1} where the anterior spinal artery arises from only one vertebral artery).^{Reference Er, Fraser and Lanzino8,Reference Tatu, Moulin, Vuillie, Bogousslavsky, LR and J9} In our case, even with repeated CTA, the vertebrobasilar system was unremarkable, though the anterior spinal artery was poorly visualized. Ultimately, it was the SWAN that revealed a thrombus in the anterior spinal artery. This case highlights the utility of obtaining MRI in the acute setting for brainstem stroke syndromes. Not only is MRI more sensitive than CT for brainstem ischemia, when CT-angiography is unrevealing, SWAN-MRI may also be able to detect small thrombi in the anterior spinal artery.

The temporal evolution of bilateral medial medullary ischemia on serial MRI and progressive clinical decline is also intriguing. This observation raises questions regarding the cerebrovascular physiology of brainstem ischemia and whether the progression of infarct arose due to a failing ischemic penumbra over time. Thrombus progression may have also promoted progressive ischemia, though the patient was initiated on dual-antiplatelet therapy.

Overall, the rapid recognition of this rare and, at times, progressive bilateral medial medullary stroke syndrome (even in the absence of classic features such as tongue deviation but accompanied by pain and respiratory distress) and coupled with urgent MRI can facilitate the identification of thrombolytic candidacy and expedite monitoring in intensive care units.