Matt Silverman, MS4
Article: Kew J, Ahuja A, Loftus WK, Scott PMJ, & Metreweli C. “Peritonsillar Abscess Appearance on Intra-oral Ultrasonography.” Clinical Radiology (1998), 53, 143-146.
Goal: To assess whether ultrasound can correctly distinguish peritonsillar abscesses from peritonsillar cellulitis, and determine if there is a consistent appearance on ultrasound of peritonsillar abscess that will make them easier to identify. Further, the researchers hoped to correlate abscess size and duration of symptoms with the echogenic appearance on ultrasound.
Experimental Design: A prospective double-blind study of 15 patients who had symptoms of peritonsillar infection was performed using a sector scanner, 7-9 MHz 2x2x10 cm intracavitary transducer. The tip of the probe, covered in gel and rubber, was placed directly on the palatine tonsils by one of two trained technicians. After the ultrasound, a contrast enhanced 5x5 mm axial CT scan of the neck was performed and interpreted by an independent radiologist. 13 of the 15 patients then underwent surgical confirmation of abscess. It was assumed that the central non-enhancing, low density areas of the CT-scans were areas of necrosis or pus and were measured as a percentage of the abscess, which was then correlated with ultrasound.
Results: Patients were symptomatic for an average of 4 days with an average volume of necrosis on CT of 16.4 ml. One patient was excluded from the study due to trismus and inability to ultrasound. 11 of the remaining patients had abscesses that correlated on US, CT and surgery. There was one false negative on ultrasound, where it appeared as cellulitis, but CT and surgery confirmed it was an abscess. Another patient had what appeared to be an abscess on ultrasound, but CT was negative and the patient improved without surgery. Another patient was removed out of inability to confirm abscess through US, CT, or surgery due to an enlarged phlegmonous tonsil.
On ultrasound, a peritonsillar abscess appeared as an isoechoic rim with a central hypoechoic area for all but two patients, who instead had a homogenous isoechoic pattern. There was no correlation of the echogenic pattern on ultrasound and the number of symptomatic days. The more common appearance of a hypoechoic abscess on ultrasound was present when the amount of necrosis was greater than 10% on CT. Further, ultrasound incorrectly localized abscesses in four patients. The authors suggest that differing appearances of echopatterns of abscess on ultrasound may be due to the smaller volume of necrosis having a less acoustic transmission and thus a more isoechoic pattern.
Room for Improvement: The authors cite that 30% of peritonsillar abscesses occur in the lower pole of the tonsils, which are difficult to access and therefore assess via ultrasound. The tonsil is often distorted and displaced medially and caudally when there is an infection, which can be difficult to find on ultrasound. They acknowledge that the mis-localization of the collections that occurred in their study happened in the earlier patients that they examined, which suggests a learning period and a more experienced technician may have an easier time identifying the proper location of the collection. They could improve the study by using more experienced technicians, which should improve their ability to localize the abscesses. Further, the power of the study is on the lower side with only 15 study participants, 1 of which was unable to participate due to trismus. They could improve on this by increasing the population.
Conclusion: Intra-oral ultrasound is a relatively safe, short, easy procedure to perform that can correctly localize and identify the presence of a peritonsillar abscess once the technician performing the exam is experienced. The different ultrasonic appearances did not correlate with the number of symptomatic days, but abscesses with a central hypoechoic area and a surrounding isoechoic rim pattern were less likely if the volume of pus relative to the whole abscess was less than 10% on CT. They were unable to predict the ultrasound appearances by the duration of symptoms.