Reply

We would like to thank Ide and colleagues for their letter drawing our attention to their case report, An Early Observation of Carotid Involvement by Retropharyngeal Abscess. Their article was published after our paper had been accepted, but before it was revised, which likely explains the oversight on our part. We would like to comment on several of the points addressed by the authors.

To summarize, Ide et al appear concerned that we have underestimated the seriousness of internal carotid artery (ICA) narrowing in children with retropharyngeal abscess (RTPA), and that this may portend ICA occlusion of pseudoaneurysm that is clinically occult.

Although we did not appreciate the abnormal enhancement in the ICA wall to the level of the cavernous segment on CT that was enhanced on MR scans, we think it is unlikely that we have underestimated the degree of carotid involvement. Indeed, we pointed out that the degree of ICA narrowing was quite severe in several of our patients. It does appear, however, that if the perivascular inflammatory process (RTPA) is treated expeditiously with surgery, the vascular findings in our investigation, based on clinical follow-up, appeared to have no sequelae in the 13 patients we studied (1).

Ide et al suggest that clinical follow-up is not reliable to exclude ICA disease, because “children almost always tolerate [ICA] occlusion well clinically.” Based on clinical experience in large trauma and sickle cell populations treated at our pediatric hospital, complete ICA occlusion in childhood is commonly a devastating event resulting in dramatic neurologic impairment. We think it is unlikely that a complete ICA occlusion would be asymptomatic, especially in 13 sequential children.

We maintain that the only imaging required acutely is contrast-enhanced CT. Children with RTPA are seriously ill. Spiral CT is fast, requires no sedation in a child who already may have airway compromise, and, even if distal ICA enhancement is not appreciated on CT as it is on contrast-enhanced MR imaging, the treatment is still surgery or close observation and IV antibiotics. What additional information is gained with contrast-enhanced MR imaging? Of what significance is visualization of perivascular enhancement remote from the infection? Even in the authors' limited experience, the findings resolve after surgery. Furthermore, MR imaging might delay definitive treatment.

Should postoperative MR imaging be performed to rule out ICA occlusion or pseudoaneurysm? The authors are correct in stating that ICA occlusion/pseudoaneurysm are potential complications of cervical adenitis in children (2–4). The frequency of this complication, however, cannot be determined, as cases are sporadic and reported as single episodes. If the authors are aware of any literature suggesting that ICA occlusion/pseudoaneurysm is a frequent complication of RTPA, we would appreciate seeing the data. Although ICA injury is definitely a complication of penetrating neck trauma, we are unaware of any large series reporting vascular injury common (or even uncommon) after RTPA. Therefore, ICA injury following surgically treated RTPA appears to be rare, and to recommend postsurgical MR imaging because of a speculation of potential ICA injury is not based on enough evidence to change our practice. In our broad clinical experience dealing with RTPA, none of our surgeons or pediatric neuroradiologists have seen any child return with symptoms referable to ICA injury or occlusion.

In summary, ICA narrowing ipsilateral to RTPA is a dramatic imaging finding that, together with the clinical status of the patient, implies a serious inflammatory process that should be treated expeditiously. In light of the absence of compelling data that suggests otherwise, and based on our clinical experience, we maintain that neurologic sequelae are rare and follow-up imaging is not routinely recommended.