The use of fluorodeoxyglucose positron emission tomography (FDG-PET), for diagnosing lung cancer may have to be rethought. According to an analysis of data from a national prospective trial, FDG-PET has substantially lower sensitivity and specificity than in previously published studies.
The use of fluorodeoxyglucose positron emission tomography (FDG-PET), for diagnosing lung cancer may have to be rethought. According to an analysis of data from a national prospective trial, FDG-PET has substantially lower sensitivity and specificity than in previously published studies.
“The current National Comprehensive Cancer Network guidelines recommend FDG-PET scans for the diagnosis of NSCLC,” said Eric L. Grogan, MD, MPH, of Vanderbilt University Medical Center. “These recommendations are based on studies that demonstrate a high degree of accuracy with this diagnostic tool.”
Meta-analyses of FDG-PET’s ability to diagnose lung cancer have suggested excellent sensitivity (94%) and specificity (83%), but the procedures were restricted to select centers that may have more experience than many. Grogan and colleagues conducted a secondary analysis of the national ACOSOG Z4031 trial due its use of widespread centers.
A total of 1,074 patients with stage I known or suspected NSCLC enrolled in the trial between 2004 and 2006; all patients underwent surgical resection, and the cancer rate in all patients was 80% vs 20% benign lesions. The present analysis compared differences in FDG-PET results by enrolling site. There were 51 different sites in 39 cities, and preoperative FDG-PET results were available for a total of 682 patients.
The lung cancer prevalence was 83% in this group of patients. FDG-PET sensitivity was found to be 82%, and specificity was 31%, substantially lower than in previous analyses. Positive predictive value of FDG-PET was 85%, and negative predictive value was 26%. Grogan noted that accuracy of the scans increased with lesion size.
There were a total of 80 false positive scans, 69% of which were granulomas. There were 101 patients who had a false negative scan; the most common false negative was caused by adenocarcinomas (64%), followed by squamous tumors (12%), and others.
“In the eight cities with greater than twenty-five enrollees, sensitivity varied significantly,” Grogan said. In Los Angeles, for example, the sensitivity was 67%, while in Durham it was 91%.
“It is generalizable to clinical practice, because multiple FDG-PET scanners were used, the scans were performed at both community and academic centers, and interpreted by multiple radiologists,” Grogan said, though he noted the intrinsic limitations to conducting a secondary analysis of a prospective trial. A further limitation lay in the lack of access to original images; instead, the investigators relied on written radiology reports, meaning that no additional review of images was possible.
Tetsuya Mitsudomi, MD, PhD, of the Aichi Cancer Center in Nagoya, Japan, who was not involved with the study, noted that “it is not possible to remove all the false positives if you use FDG.” He added that new tracers are being developed that may improve diagnostic accuracy in NSCLC, including C-methionine, fluorothymidine, and many others.
Dr. Grogan concluded that there is clearly much to be desired with FDG-PET as a diagnostic tool. “The results of PET scans in this population should be interpreted cautiously, and reasons for the poor test performance should be explored in other studies.
Neoadjuvant Capecitabine Plus Temozolomide in Atypical Lung NETs
Read about a woman with well-differentiated atypical carcinoid who experienced a 21% regression in primary tumor size after 12 months on neoadjuvant capecitabine and temozolomide.