2017 Section 7 Green Book

Otolaryngology–Head and Neck Surgery 154(3)

proliferating neoplastic cells consume glucose at a higher rate than normal cells do, the glucose analogue 18-F- fluorodeoxyglucose (18-FDG) accumulates at higher rates within malignant cells. However, nonspecific increases in the uptake of FDG within cells can also occur in normal salivary glands and lymphoid tissues and in the setting of infection and inflammation, such as that which occurs after radiotherapy. The aim of our systematic review and meta-analysis was to evaluate the diagnostic accuracy of PET and PET/CT for detecting residual and/or recurrent local and regional disease and distant metastases in patients with HNSCCs following radiotherapy or chemoradiotherapy. Hereafter, PET and PET/CT will be collectively referred to as PET, with dis- tinctions made where necessary. We searched EMBASE, PREMEDLINE, MEDLINE, and GoogleScholar for studies evaluating the diagnostic perfor- mance of FDG-PET in head and neck cancers. Additional relevant studies were identified by reviewing the reference list of articles retrieved and searching the Cochrane Database for Systematic Reviews. We used a search strategy based on a variety of keywords and Medical Subject Heading (MeSH) terms, with the search algorithm modified as necessary for each database Search terms included positron emission tomography, head and neck neoplasm, squamous cell carcinoma, local neoplasm recurrence, residual neoplasm, squamous cell carcinoma, local neoplasm recurrence, sensitivity and specificity (see Supplementary Appendix 1 and 2 at www.otojournal.org/supplemental). There were no language restrictions for our search, and we included all prospective studies published until February 28, 2015. Study Selection and Eligibility Criteria Two reviewers were involved in the selection of studies, data collection, and quality assessment process; any dis- agreements were resolved by consensus or by discussion with a third reviewer. Citations were initially screened to determine whether they pertained to the use of imaging in head and neck can- cers. The abstracts were then assessed for eligibility for inclusion based on the following criteria: FDG-PET or FDG-PET/CT for posttreatment response assessment or surveillance for residual or recurrent head and neck cancer after treatment with radiotherapy or chemoradiotherapy Histopathological analysis and/or close clinical and imaging follow-up was used as the reference standard Data on the number of true-positive, true-negative, false-positive, and false-negative results were avail- able or could be extracted based on the sensitivity, Methods Search Methodology

specificity, positive predictive value (PPV), and negative predictive value (NPV) provided Minimum of 10 patients

The full texts of these potentially eligible articles were retrieved and evaluated to ensure that all inclusion criteria were satisfied. Review articles, case reports, commentaries, conference proceedings, and letters to the editor were excluded. Retrospective studies were also excluded as these may potentially overestimate the diagnostic accuracy. Only patients with HNSCCs were included in this review; in stud- ies in which the population had a mixture of histology, an attempt was made to extract just the data on patients with SCCs. Studies that did not specify SCCs or where it was not possible to separate out the data on SCCs from other none- pithelial tumors were excluded. Studies were also excluded if part of the study population received surgery alone as the treatment, if the primary treatment modality was not reported, or if dual head coincidence gamma cameras were used to capture the images. Data Collection Process and Data Items Data from each study were extracted onto a standardized data extraction form. One reviewer collected the data, and a second reviewer checked the extracted data. We recorded the author names; journal; year of publication; sample size; initial treatment modality; description of study population including age, gender, site, and stage of disease; time to ini- tial PET imaging; definition of positive PET scan (visual, semiquantitative); location of recurrence (local, nodal, dis- tant, all sites considered together); reference standard; and duration of follow-up. The number of true-positive, true- negative, false-positive, and false-negative results was recorded or extracted onto a 2 3 2 table based on the sensi- tivity, specificity, PPV, NPV, and sample size data pro- vided. Based on the data in these tables, the sensitivity, specificity, PPV, NPV, and overall accuracy were calculated for each study. Quality Assessment of Studies The quality of each article was assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool. This 14-item assessment tool was developed by the Centre for Reviews and Dissemination at the University of New York and the Academic Medical Centre at the University of Amsterdam to allow for the consistent and reliable assess- ment of the quality of diagnostic accuracy studies included a systematic review. 10 Specifically, the tool assists in assessing the risk of bias, sources of variation, and reporting quality of diagnostic accuracy studies. We weighed each of these items equally and attributed a summary score to each study based on the responses to each question, with 1 for ‘‘yes,’’ 0 for ‘‘no,’’ and 0.5 for ‘‘unclear.’’ Statistical Analysis The weighted mean pooled sensitivity, specificity, PPV, NPV, diagnostic odds ratio (DOR), and their 95% confidence

23