In CAE, diagnosis is known to be difficult and to depend on clinical experience (Bäuerle et al.2013). The patient’s history combined with the clinical suspicion of the embolism due to the initial neurological findings and the direct temporal relation of potentially emboligenic procedures are the most important diagnostic clues (Muth and Shank2000). Early in the course of the disease, air emboli can be detected by computed tomography (CT) (Valentino et al.2003). In both reported patients, air could be verified without doubt, as imaging by CT was performed early after the accident. Despite immediate critical care support the course of the disease was lethal in both cases. Because a do-not-resuscitate order existed for patient 1, mechanical ventilation and resuscitation were omitted. In patient 2 intensive care unit therapy was performed. Hyperbaric oxygen therapy (HBO) was desired as it is recommended, in particular within the first 6 h after the accident (Muth and Shank2000; van Hulst et al.2003; O’Dowd2013). However HBO was not available in our hospital at this time and transport to the next HBO facility was considered to be high risk and of questionable benefit due to the patient’s global clinical condition. Recently therapeutic hypothermia was successfully reported in CAE (Bäuerle et al.2013). However, due to septic conditions and a severe coagulation disorder in patient 2, hypothermia and even hemicraniectomy for decompression were judged to be unreasonable.
As reported in the literature, in CAE the right hemisphere seems to be more affected due to the course of the right brachiocephalic trunc which is the first main artery rising from the aortic trunc (Heckmann et al.2000). In both reported cases, the right hemisphere was more severely damaged supporting the hypothesis that venous embolism transits into the arterial system and primarily affects the right-sided great vessels from the aortic branch.
Pathophysiologically in patient 1, a mechanism comparable to CAE associated with central venous lines was assumed (Heckmann et al.2000; Han et al.2010)). Air entered the venous system in this case by the venous port which had been inserted a year earlier with access to the left cephalic vein. Despite no documentation of a right-left shunt via a persistent foramen ovale, shunts from the venous system to the arterial system are quite frequent occurring via the lung (Han et al.2010).
In patient 2 the pathophysiology is not as clear. As in earlier reports on ERCP or esophagogastroduodenoscopy-associated CAE, it is speculated that inflammation of the bile duct and its surrounding veins, previous intervention of the bile duct system, and insufflation of air during endoscopy may have contributed to air entering the vascular system either via portal and liver veins or via mucosal disruptions (Pandurangadu et al.2012; Nern et al.2012; Koster et al.2012). If air entered the arterial cerebral circulation, gas bubbles obstruct the cerebral vessels and cause distal ischemia (Muth and Shank2000) via a complex cascade of damage (van Hulst et al.2003) which resembles damage resulting from hypoxic brain injury (Bäuerle et al.2013).
The reported case vignettes are remarkable for two reasons. First, a lethal cerebral air embolism associated with the use of a port catheter under home care conditions has hitherto not been reported according to a literature search. Unfortunately, the exact circumstances which led to the air entry into the port catheter could not be resolved. As a rule, however, a port catheter should be handled with the same caution as a central venous line. Second, a lethal cerebral air embolism in association with ERCP or even esophagogastroscopy is sparsely reported. Including our case and reviews from 2010 and 2012, 11 cases are documented in the literature (Finsterer et al.2010; Nern et al.2012). Even the finding of air inclusions within the superficial cortex and the superior sagittal sinus in our second patient is quite unique. Such a finding has been reported earlier in only two patients with ERCP associated CAE (Nern et al.2012).