Objectives <p>To identify imaging features that differentiate malignant lesions from non-malignant conditions associated with incidental uptake in the esophagus and gastroesophageal junction (GEJ) on <sup>18</sup>F-FDG PET/CT.</p> Methods <p>Patients were included if they (1) underwent whole-body <sup>18</sup>F-FDG PET/CT between January 2009 and July 2023, with reports documenting incidental uptake in the esophagus or GEJ and recommending upper endoscopy; and (2) had upper endoscopy within one year of the PET scan. Patients with known esophageal or GEJ cancer were excluded. <sup>18</sup>F-FDG PET/CT images were assessed for uptake location, pattern (diffuse, segmental, or focal), eccentricity (circumferential or eccentric), maximal and mean standardized uptake values (SUVs), and CT findings (normal, wall thickening, or mass). Patients were categorized into benign conditions, Barrett’s esophagus, and malignancies based on endoscopic and pathological diagnoses.</p> Results <p>A total of 106 patients were included, comprising 82 cases of benign conditions, 11 Barrett’s esophagus, and 13 malignancies. <sup>18</sup>F-FDG PET/CT features were not significantly different between Barrett’s esophagus and benign conditions. Compared with the Barrett’s esophagus and benign conditions, malignant lesions showed a higher frequency of isolated mid-esophageal uptake (2.2% vs. 30.8%, <i>P</i> = 0.048), eccentric uptake (2.2% vs. 38.5%, <i>P</i> &lt; 0.001), and wall thickening or mass on CT (15.1% vs. 92.4%, <i>P</i> &lt; 0.001). Lesion SUV<sub>max</sub>/liver SUV<sub>mean</sub> demonstrated excellent diagnostic performance in predicting malignancies (area under the curve 0.94). The presence of at least one of the following five features yielded 100% sensitivity, 65.6% specificity, and 28.9% positive predictive value for malignancies: focal uptake, eccentric uptake, isolated mid-esophageal uptake, lesion SUV<sub>max</sub>/liver SUV<sub>mean</sub> ≥ 2.8, and wall thickening or mass on CT. The presence of at least two of these features increased specificity to 94.6% while maintaining 100% sensitivity.</p> Conclusion <p>Any of the five incidental <sup>18</sup>F-FDG PET/CT imaging features in the esophagus or GEJ warrant endoscopy.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Incidental esophageal and gastroesophageal junction uptake on ¹8F-FDG PET/CT: when is upper endoscopy warranted?

  • Yihan Cao,
  • Bashar Kako,
  • Ashkan Eighaei Sedeh,
  • Caitlin Deffler,
  • Thomas S. C. Ng,
  • David Z. Chow,
  • Shahein H. Tajmir,
  • Samuel J. Klempner,
  • Matthew R. Strickland,
  • Umar Mahmood,
  • Pedram Heidari,
  • Shadi A. Esfahani

摘要

Objectives

To identify imaging features that differentiate malignant lesions from non-malignant conditions associated with incidental uptake in the esophagus and gastroesophageal junction (GEJ) on 18F-FDG PET/CT.

Methods

Patients were included if they (1) underwent whole-body 18F-FDG PET/CT between January 2009 and July 2023, with reports documenting incidental uptake in the esophagus or GEJ and recommending upper endoscopy; and (2) had upper endoscopy within one year of the PET scan. Patients with known esophageal or GEJ cancer were excluded. 18F-FDG PET/CT images were assessed for uptake location, pattern (diffuse, segmental, or focal), eccentricity (circumferential or eccentric), maximal and mean standardized uptake values (SUVs), and CT findings (normal, wall thickening, or mass). Patients were categorized into benign conditions, Barrett’s esophagus, and malignancies based on endoscopic and pathological diagnoses.

Results

A total of 106 patients were included, comprising 82 cases of benign conditions, 11 Barrett’s esophagus, and 13 malignancies. 18F-FDG PET/CT features were not significantly different between Barrett’s esophagus and benign conditions. Compared with the Barrett’s esophagus and benign conditions, malignant lesions showed a higher frequency of isolated mid-esophageal uptake (2.2% vs. 30.8%, P = 0.048), eccentric uptake (2.2% vs. 38.5%, P < 0.001), and wall thickening or mass on CT (15.1% vs. 92.4%, P < 0.001). Lesion SUVmax/liver SUVmean demonstrated excellent diagnostic performance in predicting malignancies (area under the curve 0.94). The presence of at least one of the following five features yielded 100% sensitivity, 65.6% specificity, and 28.9% positive predictive value for malignancies: focal uptake, eccentric uptake, isolated mid-esophageal uptake, lesion SUVmax/liver SUVmean ≥ 2.8, and wall thickening or mass on CT. The presence of at least two of these features increased specificity to 94.6% while maintaining 100% sensitivity.

Conclusion

Any of the five incidental 18F-FDG PET/CT imaging features in the esophagus or GEJ warrant endoscopy.