Internal Carotid Artery Narrowing in Children with Retropharyngeal Lymphadenitis and Abscess

We read with interest the article of Hudgins et al (1) describing, in 13 consecutive children, narrowing of the internal carotid artery (ICA) ipsilateral to retropharyngeal lymphadenitis and abscess, and suggesting that such narrowing is a common, benign, and, most likely, incidental imaging finding. We would like to draw the authors' attention to our article on this subject that was published in the March 1998 issue of the AJNR in which we emphasized the importance of the MR findings of this entity (2). It was a case report of carotid involvement by retropharyngeal abscess in a 4-year-old boy investigated by MR imaging. MR imaging showed not only lumen narrowing of the ICA but also enhancement of its wall. Although narrowing was greatest at the level of the abscess, the wall enhancement was seen all along the course of the ICA up to the cavernous segment. This finding, revealed only on contrast-enhanced T1-weighted images, is probably unidentifiable on contrast-enhanced CT owing to the contrast resolution of this technique. When this finding is present, however, it might indicate a more severe carotid involvement than suspected on the basis of CT images only and Hudgins et al might, therefore, have underestimated the degree of carotid involvement. Unfortunately, we lack data in the literature and actual histologic proof to define the exact nature of such wall abnormalities; as the authors suggest, this could have been from spasm or actual involvement of the wall of the vessel by the inflammatory process itself. This process could presumably be referred to as arteritis, because this term not only refers to an infection, but also to an inflammation of the arterial wall, whatever its severity. Intuitively, wall thickening with enhancement, as we found in our case, would be more readily expected in a true arteritis than in a spasm. Because arteritis can weaken the wall of a vessel, it could also cause perforation, leading to fatal hemorrhage or a pseudoaneurysm. We would like to emphasize another potential manifestation of carotid involvement by sepsis apart from arterial rupture—occlusion. Indeed, it has also been suggested that, in children, arteritis produced by direct extension of infection of the neck or throat might be the more important risk factor in addition to trauma for cervical occlusion (3). Although children almost always tolerate carotid occlusion well clinically, it reduces the capabilities of collateralization of the cerebral vasculature in adulthood. In their study, the authors did not exclude this second rare complication because the follow-up of the children included neurologic examination (all patients recovered with no neurologic deficits) and a head CT (normal in all patients) but no direct imaging of the carotids (eg, color Doppler sonography). Because those carotid complications are unusual, given the widespread and early use of antibiotics, we think further MR studies of more patients are needed before stating that intrinsic abnormalities of the ICA adjacent to an abscess are always benign. At this stage of knowledge, one question remains unanswered: should we perform an MR examination with gadolinium once CT shows carotid narrowing near a cervical abscess? Until the exact significance of these carotid imaging findings is known, we would still consider that lumen narrowing associated with enhancement is a sign of severity that should lead to aggressive treatment. In our case, the abscess was drained surgically the day after admission and the follow-up MR examination 1 week later showed normalization of the affected ICA.