A Systematic Review of the Predictive Value of 18FDG-PET in Esophageal and Esophagogastric Junction Cancer After Neoadjuvant Chemoradiation on the Survival Outcome Stratification

• Published in: Journal of gastrointestinal surgery Vol. 18; no. 5; pp. 894 - 905
• Main Authors: Schollaert, Pascaline, Crott, Ralph, Bertrand, Claude, D’Hondt, Lionel, Borght, Thierry Vander, Krug, Bruno
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01.05.2014
• Subjects: Gastroenterology; Medicine; Medicine & Public Health; Original Article; Surgery; FDG-PET; Overall survival; Esophageal cancer; Disease-free survival
• ISSN: 1091-255X; 1873-4626
• DOI: 10.1007/s11605-014-2488-2

Purpose We studied the predictive value of [ 18  F]fluorodeoxyglucose-positron emission tomography ( 18 FDG-PET) for assessing disease-free (DFS) and overall survival (OS) in esophageal and esophagogastric junction cancer. Materials and methods A literature search (PUBMED/MEDLINE, EMBASE, Cochrane) was performed to identify full papers with 18 FDG-PET and survival data, using indexing terms and free text words. Studies with >10 patients with locally advanced esophageal cancer, presenting sequential or at least one post-adjuvant treatment 18 FDG-PET data and Kaplan–Meier survival curves with >6 months median follow-up period were included. We performed a meta-analysis for DFS and OS using the hazard ratio (HRs) as outcome measure. Sources of heterogeneity study were also explored. Results We identified 26 eligible studies including a total of 1,544 patients (average age 62 years, 82 % males). The TNM distribution was as follows: stage I 7 %, II 24 %, III 53 % and IV 15 %. The pooled HRs for complete metabolic response versus no response were 0.51 for OS (95 % CI, 0.4–0.64; P < 0.00001) and 0.47 for DFS (95 % CI, 0.38–0.57; P < 0.00001), respectively. No statistical heterogeneity was present. To explore sources of clinical heterogeneity, we also realised subgroup and regression analyses. Taken into account the moderate correlation between OS and DFS (ρ = 0.54), we used joint bivariate random regression model. These analyses did not show a statistically significant impact of study characteristics and PET modalities on the pooled outcome estimates. Conclusion Despite methodological and clinical heterogeneity, metabolic response on 18 FDG-PET is a significant predictor of long-term survival data.