Measuring Tumor Hypoxia with 18F-FETNIM PET in Esophageal Squamous Cell Carcinoma: A Pilot Clinical Study
J. Yue,J. Yu,A. R. Cabrera,X. Sun,S. Zhao,L. Ma
DOI: https://doi.org/10.1016/j.ijrobp.2011.06.520
2011-01-01
Abstract:Purpose/Objective(s)To evaluate hypoxia in esophageal squamous cell carcinoma (SCC) with (18F-FETNIM PET/CT. We determined an imaging threshold for hypoxia, quantified the spatiotemporal variability of hypoxia in untreated tumor, and evaluated the ability of 18F-FETNIM PET to predict clinical response following concurrent chemoradiotherapy (CCRT).Materials/MethodsTwenty-eight consecutive patients with inoperable SCC of the esophagus were consecutively accrued between April 2007 and June 2010. The first 10 patients received two pretreatment 18F-FETNIM-PET/CT scans on separate days. The remaining 18 patients only underwent 18F-FETNIM-PET/CT once before CCRT. The ratio of the maximum standardized uptake value (SUVmax) of 336 normal tissue regions (i.e., heart, lung, brain or muscle) to the mean standardized uptake value (SUVmean) of the respective patient's spleen was calculated, and the imaging threshold for hypoxia defined as the level of uptake demonstrated by less than 5% of tissue regions. Among the patients with two pretreatment scans, each pair of scans was compared with respect to location and intensity of uptake to assess for baseline spatiotemporal variability. Logistic regression analysis was used to determine whether pretreatment imaging characteristics are predictive of clinical response.ResultsThe mean and median ratios of the SUVmax of tissue:SUVmean of spleen were nearly identical, and 95% of the ratios fell below 1.3. The mean Dice similarity coefficient for the hypoxic volumes on pretreatment PET scans acquired in the same patient on different days was 0.12 (range, 0.05-0.21). Individuals' tumor SUVmax and SUVmean did not vary significantly, but on average, the geometric centers of hypoxic regions shifted 15 mm (range, 8-20 mm) from the first pretreatment scan to the second. SUVmax was the imaging characteristic most predictive of treatment response (p = 0.041), with high SUVmax associated with poor clinical response.Conclusions18F-FETNIM PET/CT scan depict hypoxia in esophageal SCC. Prior to CCRT, tumor hypoxia demonstrates spatial variability on different days, though overall 18F-FETNIM uptake remains similar. Baseline SUVmax may be predictive of treatment response. Purpose/Objective(s)To evaluate hypoxia in esophageal squamous cell carcinoma (SCC) with (18F-FETNIM PET/CT. We determined an imaging threshold for hypoxia, quantified the spatiotemporal variability of hypoxia in untreated tumor, and evaluated the ability of 18F-FETNIM PET to predict clinical response following concurrent chemoradiotherapy (CCRT). To evaluate hypoxia in esophageal squamous cell carcinoma (SCC) with (18F-FETNIM PET/CT. We determined an imaging threshold for hypoxia, quantified the spatiotemporal variability of hypoxia in untreated tumor, and evaluated the ability of 18F-FETNIM PET to predict clinical response following concurrent chemoradiotherapy (CCRT). Materials/MethodsTwenty-eight consecutive patients with inoperable SCC of the esophagus were consecutively accrued between April 2007 and June 2010. The first 10 patients received two pretreatment 18F-FETNIM-PET/CT scans on separate days. The remaining 18 patients only underwent 18F-FETNIM-PET/CT once before CCRT. The ratio of the maximum standardized uptake value (SUVmax) of 336 normal tissue regions (i.e., heart, lung, brain or muscle) to the mean standardized uptake value (SUVmean) of the respective patient's spleen was calculated, and the imaging threshold for hypoxia defined as the level of uptake demonstrated by less than 5% of tissue regions. Among the patients with two pretreatment scans, each pair of scans was compared with respect to location and intensity of uptake to assess for baseline spatiotemporal variability. Logistic regression analysis was used to determine whether pretreatment imaging characteristics are predictive of clinical response. Twenty-eight consecutive patients with inoperable SCC of the esophagus were consecutively accrued between April 2007 and June 2010. The first 10 patients received two pretreatment 18F-FETNIM-PET/CT scans on separate days. The remaining 18 patients only underwent 18F-FETNIM-PET/CT once before CCRT. The ratio of the maximum standardized uptake value (SUVmax) of 336 normal tissue regions (i.e., heart, lung, brain or muscle) to the mean standardized uptake value (SUVmean) of the respective patient's spleen was calculated, and the imaging threshold for hypoxia defined as the level of uptake demonstrated by less than 5% of tissue regions. Among the patients with two pretreatment scans, each pair of scans was compared with respect to location and intensity of uptake to assess for baseline spatiotemporal variability. Logistic regression analysis was used to determine whether pretreatment imaging characteristics are predictive of clinical response. ResultsThe mean and median ratios of the SUVmax of tissue:SUVmean of spleen were nearly identical, and 95% of the ratios fell below 1.3. The mean Dice similarity coefficient for the hypoxic volumes on pretreatment PET scans acquired in the same patient on different days was 0.12 (range, 0.05-0.21). Individuals' tumor SUVmax and SUVmean did not vary significantly, but on average, the geometric centers of hypoxic regions shifted 15 mm (range, 8-20 mm) from the first pretreatment scan to the second. SUVmax was the imaging characteristic most predictive of treatment response (p = 0.041), with high SUVmax associated with poor clinical response. The mean and median ratios of the SUVmax of tissue:SUVmean of spleen were nearly identical, and 95% of the ratios fell below 1.3. The mean Dice similarity coefficient for the hypoxic volumes on pretreatment PET scans acquired in the same patient on different days was 0.12 (range, 0.05-0.21). Individuals' tumor SUVmax and SUVmean did not vary significantly, but on average, the geometric centers of hypoxic regions shifted 15 mm (range, 8-20 mm) from the first pretreatment scan to the second. SUVmax was the imaging characteristic most predictive of treatment response (p = 0.041), with high SUVmax associated with poor clinical response. Conclusions18F-FETNIM PET/CT scan depict hypoxia in esophageal SCC. Prior to CCRT, tumor hypoxia demonstrates spatial variability on different days, though overall 18F-FETNIM uptake remains similar. Baseline SUVmax may be predictive of treatment response. 18F-FETNIM PET/CT scan depict hypoxia in esophageal SCC. Prior to CCRT, tumor hypoxia demonstrates spatial variability on different days, though overall 18F-FETNIM uptake remains similar. Baseline SUVmax may be predictive of treatment response.