18F-FDG PET/CT oncologic imaging at extended injection-to-scan acquisition time intervals derived from a single-institution 18F-FDG-directed surgery experience: feasibility and quantification of 18F-FDG accumulation within 18F-FDG-avid lesions and background tissues

Background 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) is a well-established imaging modality for a wide variety of solid malignancies. Currently, only limited data exists regarding the utility of PET/CT imaging at very extended injection-to-scan acquisition times. The current retrospective data analysis assessed the feasibility and quantification of diagnostic 18F-FDG PET/CT oncologic imaging at extended injection-to-scan acquisition time intervals. Methods 18F-FDG-avid lesions (not surgically manipulated or altered during 18F-FDG-directed surgery, and visualized both on preoperative and postoperative 18F-FDG PET/CT imaging) and corresponding background tissues were assessed for 18F-FDG accumulation on same-day preoperative and postoperative 18F-FDG PET/CT imaging. Multiple patient variables and 18F-FDG-avid lesion variables were examined. Results For the 32 18F-FDG-avid lesions making up the final 18F-FDG-avid lesion data set (from among 7 patients), the mean injection-to-scan times of the preoperative and postoperative 18F-FDG PET/CT scans were 73 (±3, 70-78) and 530 (±79, 413-739) minutes, respectively (P < 0.001). The preoperative and postoperative mean 18F-FDG-avid lesion SUVmax values were 7.7 (±4.0, 3.6-19.5) and 11.3 (±6.0, 4.1-29.2), respectively (P < 0.001). The preoperative and postoperative mean background SUVmax values were 2.3 (±0.6, 1.0-3.2) and 2.1 (±0.6, 1.0-3.3), respectively (P = 0.017). The preoperative and postoperative mean lesion-to-background SUVmax ratios were 3.7 (±2.3, 1.5-9.8) and 5.8 (±3.6, 1.6-16.2), respectively, (P < 0.001). Conclusions 18F-FDG PET/CT oncologic imaging can be successfully performed at extended injection-to-scan acquisition time intervals of up to approximately 5 half-lives for 18F-FDG while maintaining good/adequate diagnostic image quality. The resultant increase in the 18F-FDG-avid lesion SUVmax values, decreased background SUVmax values, and increased lesion-to-background SUVmax ratios seen from preoperative to postoperative 18F-FDG PET/CT imaging have great potential for allowing for the integrated, real-time use of 18F-FDG PET/CT imaging in conjunction with 18F-FDG-directed interventional radiology biopsy and ablation procedures and 18F-FDG-directed surgical procedures, as well as have far-reaching impact on potentially re-shaping future thinking regarding the “most optimal” injection-to-scan acquisition time interval for all routine diagnostic 18F-FDG PET/CT oncologic imaging.

(Continued on next page) (Continued from previous page) Conclusions: 18 F-FDG PET/CT oncologic imaging can be successfully performed at extended injection-to-scan acquisition time intervals of up to approximately 5 half-lives for 18 F-FDG while maintaining good/adequate diagnostic image quality. The resultant increase in the 18 F-FDG-avid lesion SUV max values, decreased background SUV max values, and increased lesion-to-background SUV max ratios seen from preoperative to postoperative 18 F-FDG PET/CT imaging have great potential for allowing for the integrated, real-time use of 18 F-FDG PET/CT imaging in conjunction with 18 F-FDG-directed interventional radiology biopsy and ablation procedures and 18 F-FDG-directed surgical procedures, as well as have far-reaching impact on potentially re-shaping future thinking regarding the "most optimal" injection-toscan acquisition time interval for all routine diagnostic 18 F-FDG PET/CT oncologic imaging.
Keywords: 18 F-FDG, PET/CT, SUV max , Injection-to-scan acquisition time, Delayed imaging, Lesion-to-background ratio, Tumor-to-background ratio, 18 F-FDG-directed surgery, Real-time, Oncologic Background 18 F-fluorodeoxyglucose ( 18 F-FDG) positron emission tomography/computed tomography (PET/CT) is a well-established imaging modality for a wide variety of solid malignancies [1][2][3][4][5]. Its utilities have included initial cancer diagnostics, staging, restaging, therapy planning, therapy response monitoring, surveillance, and cancer screening for at-risk populations. Beyond these utilities, there has been growing interest in evaluating the feasibility of utilizing 18 F-FDG and PET/CT technology for providing real-time information within the operative room and perioperative arena .
As part of an effort to provide surgeons with improved intraoperative tumor localization and image-based verification of completeness of resection, our collaborative group at The Ohio State University has previously described a novel, multimodal imaging and detection strategy involving perioperative patient and ex vivo surgical specimen 18 F-FDG PET/CT imaging performed in combination with intraoperative 18 F-FDG gamma detection [51]. As part of this schema, patients could undergo both a same-day preoperative diagnostic whole-body 18 F-FDG PET/CT and a same-day postoperative diagnostic limited field-of-view 18 F-FDG PET/CT, utilizing a single preoperative dose of 18 F-FDG. This has provided our group with a unique dual-set of diagnostic 18 F-FDG PET/CT images, in which the initial same-day preoperative diagnostic whole-body 18 F-FDG PET/CT images were acquired within the injection-to-scan acquisition time interval generally recommended for diagnostic whole-body 18 F-FDG PET/CT imaging [63], and in which the second set of same-day diagnostic limited field-of-view 18 F-FDG PET/CT images were acquired after the completion of the surgical procedure, once the patient had completed standard postoperative recovery in the post-anesthesia care unit. This second set of same-day diagnostic limited field-of-view 18 F-FDG PET/ CT images was highly dependent upon the length of the surgical procedures performed, thus creating injectionto-scan acquisition time intervals for that second set of same-day diagnostic limited field-of-view 18 F-FDG PET/ CT images at time points far beyond what is generally described.
The current retrospective data analysis was undertaken to examine 18 F-FDG-avid lesions and corresponding background tissues on same-day preoperative and postoperative 18 F-FDG PET/CT scans to assess the feasibility and quantification of diagnostic 18 F-FDG PET/CT oncologic imaging at extended injection-to-scan acquisition time intervals. Herein, we have: (1) demonstrated the ability to acquire diagnostic quality images at extended injection-to-scan acquisition times; (2) identified and quantified the amount of 18 F-FDG accumulation in 18 F-FDG-avid lesions and in corresponding background tissues at these extended injection-to-scan acquisition times; and (3) compared the amount of 18 F-FDG accumulation in 18 F-FDG-avid lesions and in corresponding background tissues at these extended injection-to-scan acquisition times to that of the corresponding injectionto-scan acquisition time interval generally recommended for diagnostic whole-body 18 F-FDG PET/CT oncologic imaging.

Methods
All aspects of the current retrospective analysis were approved by the Cancer Institutional Review Board (IRB) at The Ohio State University Wexner Medical Center. The data for the current retrospective analysis were acquired from a master prospectively-maintained database (with database inclusion dates from June 2005 to June 2012), which were generated from the combination of several Cancer IRB-approved protocols, and which involved a multimodal imaging and detection approach to 18 F-FDG-directed surgery for the localization and resection of 18 F-FDG-avid lesions in patients with known and suspected malignancies. Depending upon the clinical scenario, these 18 F-FDG-directed surgical procedures were performed with either the intent for curative resection, for palliation, or for making a definitive tissue diagnosis, as based upon the standard of care management for any given disease presentation.
All patients who were eligible to be included in this current retrospective analysis consisted of those individuals who: (1) received a same-day single-dose preoperative intravenous injection of 18 F-FDG; (2) underwent same-day preoperative diagnostic whole-body 18 F-FDG PET/CT scan (usually consisting of 6 to 8 field-of-view PET bed positions, and with 2 minutes of PET imaging for each field-of-view PET bed position); (3) proceeded to the operating room for their anticipated surgical procedure and completed standard postoperative recovery in the post-anesthesia care unit; and (4) underwent a same-day postoperative diagnostic limited field-of-view 18 F-FDG PET/CT scan (which was limited only to the immediate area of the surgical resection field, usually consisting of 1 to 3 field-of-view PET bed positions, in order to limit overall patient radiation exposure for the CT portion of the PET/CT, and with 10 minutes of PET imaging for each field-of-view PET bed position). All patients fasted for a minimum of 6 hours before undergoing the same-day preoperative diagnostic whole-body 18 F-FDG PET/CT scan. Only a single intravenous dose of 18 F-FDG was used on the day of surgery, and was attempted to be administered approximately 75 minutes prior to the planned time of the same-day preoperative diagnostic whole-body 18 F-FDG PET/CT scan, which was performed within the time frame recognized by the Society of Nuclear Medicine for 18 F-FDG PET/CT image acquisition [63]. The 18 F-FDG PET/CT images were acquired on one of three clinical diagnostic scanners: (1) Siemens Biograph 16 (Siemens, Knoxville, Tennessee); (2) Phillips Gemini TF (Philips, Amsterdam, Netherlands); and (3) Siemens Biograph mCT (Siemens, Knoxville, Tennessee). Only those patients with 18 F-FDG-avid lesions seen on both same-day preoperative diagnostic whole-body 18 F-FDG PET/CT scan and same-day postoperative diagnostic limited field-of-view 18 F-FDG PET/CT scan were used in the current retrospective analysis. For any individual patient, the same-day preoperative diagnostic whole-body 18 F-FDG PET/CT scan and same-day postoperative diagnostic limited field-of-view 18 F-FDG PET/CT scan were performed on the same clinical diagnostic scanner.
The same-day preoperative diagnostic whole-body 18 F-FDG PET/CT images and same-day postoperative diagnostic limited field-of-view 18 F-FDG PET/CT images were evaluated by two nuclear medicine physicians who were initially blinded to all clinical information related to each set of preoperative and postoperative 18 F-FDG PET/CT images. The two nuclear medicine physician readers first judged the quality of the preoperative and postoperative 18 F-FDG PET/CT images as either being of diagnostic image quality or of non-diagnostic image quality, based upon criteria that were previously reported [64]. The two readers evaluated each set of preoperative and postoperative 18 F-FDG PET/CT images for identification of all 18 F-FDG-avid lesions that were considered suspicious for or consistent with malignancy. The location and maximum standard uptake value (SUV max ) of each 18 F-FDG-avid lesion were recorded. Likewise, a corresponding background SUV max was obtained either from (1) an area of tissue deemed as normal within the same organ as the 18 F-FDG-avid lesion; (2) an area of tissue deemed as normal in a location adjacent to the 18 F-FDG-avid lesion; or (3) within a single area of tissue deemed as normal elsewhere within the body when multiple 18 F-FDG-avid lesions were being evaluated in an individual case. The corresponding background SUV max values were taken from the same location on both the preoperative and postoperative 18 F-FDG PET/CT scans. Finally, the two readers were given access to the operative report for each case corresponding to each preoperative and postoperative 18 F-FDG PET/CT images data set, in order to determine which 18 F-FDG-avid lesions had been: (1) completely surgically resected; (2) partially surgically resected or biopsied; or (3) not surgically manipulated or altered (i.e., intentionally left in situ within the patient at the time of the 18 F-FDG-directed surgical procedure). The 18 F-FDG PET/CT images were all analyzed/processed on a Philips Extended Brilliance Work Station (Philips, Amsterdam, Netherlands).
All continuous variables were expressed as mean (±SD, range). The software program IBM SPSS® 21 for Windows® (SPSS, Inc., Chicago, Illinois) was used for the data analysis. All mean value comparisons for continuous variables (including the comparisons for 18 F-FDGavid lesion SUV max values, background SUV max values, and lesion-to-background SUV max ratios) from the preoperative 18 F-FDG PET/CT image group and the postoperative 18 F-FDG PET/CT image group were performed by using the 2-tailed paired samples t-test. All categorical variable comparisons were made using 2 × 2 contingency tables that were analyzed by either the Pearson chi-square test or the Fisher exact test, when appropriate. P-values determined to be 0.05 or less were considered to be statistically significant.

Derivation of the final 18 F-FDG-avid lesion data set
From a total of 166 patients who gave consent to participate in one of the IRB-approved protocols, a total of 157 patients were taken to the operating room for 18 F-FDG-directed surgery. A total of 31 of the 157 patients underwent both a same-day preoperative diagnostic whole-body 18 F-FDG PET/CT scan and a same-day postoperative diagnostic 18 F-FDG PET/CT scan utilizing a single same-day preoperative intravenous injection of 18 F-FDG. These 31 sets of preoperative and postoperative 18 F-FDG PET/CT images were evaluated by two nuclear medicine physicians for determination of diagnostic image quality versus non-diagnostic image quality. All of the 31 preoperative 18 F-FDG PET/CT imaging studies were determined to be of diagnostic image quality. A total of 5 of the 31 postoperative 18 F-FDG PET/CT imaging studies were determined to be of non-diagnostic image quality. The average injection-to-scan time for these 5 postoperative 18 F-FDG PET/CT studies with non-diagnostic image quality was of significantly longer duration, at 719 minutes (±90, 612-853), as compared to 530 minutes (±79, 413-739) for the remaining 26 postoperative 18 F-FDG PET/CT studies with diagnostic image quality (P < 0.001), suggesting that the finding of non-diagnostic image quality on a postoperative 18 F-FDG PET/CT scan was a direct consequence of any given postoperative 18 F-FDG PET/CT scan being performed at the extreme outer-limit of the extended injection-to-scan acquisition time interval. No other 18 F-FDG PET/CT imaging variables or any patient variables were significantly different for the postoperative non-diagnostic image quality group as compared to the postoperative diagnostic image quality group.
From the 26 remaining matching sets of preoperative and postoperative 18 F-FDG PET/CT studies that were determined to be of diagnostic image quality, a total of 87 individual 18 F-FDG-avid lesions were identified on the preoperative 18 F-FDG PET/CT images. There were 30 18 F-FDG-avid lesions identified on the preoperative 18 F-FDG PET/CT images that were completely surgical resected, 10 18 F-FDG-avid lesions that were partially surgically resected or biopsied, and 12 18 F-FDG-avid lesions were not within the field of view that was utilized on the postoperative 18 F-FDG PET/CT images (as the postoperative 18 F-FDG PET/CT scan was performed in a limited fashion to only to the bed of the surgical resection field). Therefore, these 52 of the original 87 individual 18 F-FDGavid lesions identified on the preoperative 18 F-FDG PET/ CT images were not considered for further data analysis.
The remaining 35 18 F-FDG-avid lesions identified on the preoperative 18 F-FDG PET/CT images were determined to represent preoperative 18 F-FDG-avid lesions that had not been surgically manipulated and were left in situ within the patient at the time of the surgical procedure, and were within the field of view on the postoperative 18 F-FDG PET/CT images. There were 3 of these remaining 35 preoperative 18 F-FDG-avid lesions that were not 18 F-FDG-avid on the postoperative 18 F-FDG PET/CT images. Of the 3 preoperative 18 F-FDG-avid lesions not found to be 18 F-FDG-avid on the postoperative 18 F-FDG PET/CT images, 2 preoperative 18 F-FDG-avid lesions were located within the bilateral tonsils in a patient who was later confirmed to have recurrent thyroid cancer within the mediastinum, but without any evidence of metastatic spread to the tonsils. These 2 areas of preoperative mild focal 18 F-FDG-avidity seen within the bilateral palatine tonsils (SUV max 4.3 on the left and 4.0 on the right), but not found to be 18 F-FDG-avid on the postoperative PET/CT images, were determined to be secondary to nonmalignant inflammation, a well-known pitfall of diagnostic 18 F-FDG PET/CT imaging of the tonsillar region. The third preoperative 18 F-FDG-avid lesion was located within the stomach region of a patient with diffuse metastatic serous ovarian cancer. This area of preoperative focal 18 F-FDG-avidity seen within the stomach region (SUV max 10.0), but not found to be 18 F-FDG-avid on the postoperative 18 F-FDG PET/CT images, has not been further evaluated to date secondary to the lack of performance of any subsequent follow-up diagnostic 18 F-FDG PET/CT imaging. As such, these 3 18 F-FDG-avid lesions were not considered for further data analysis. In the end, a total of 32 of the original 87 individual 18 F-FDG-avid lesions identified on the preoperative 18 F-FDG PET/CT images were considered as the final 18 F-FDG-avid lesion data set for the current retrospective data analysis comparing the preoperative and postoperative 18 F-FDG PET/CT images. The region of the body in which these 32 18 F-FDG-avid lesions were located was designated as the thorax for 12 lesions, abdomen/pelvis for 11 lesions, neck for 5 lesions, and axilla for 4 lesions.

Patient variables
The 32 18 F-FDG-avid lesions, constituting the final 18 F-FDG-avid lesion data set, originated from a total 7 patients (5 females and 2 males) from among the initial group of 31 patients who had undergone both a sameday preoperative diagnostic whole-body 18 F-FDG PET/CT scan and a same-day postoperative diagnostic 18 F-FDG PET/CT scan. For those 7 patients, the mean patient age was 65 (±12, 43-80) years, the mean patient weight was 80.3 (±28.1, 56.7-136.1) kilograms, the mean preoperative blood glucose level of 103 (±15, 82-121) milligrams/deciliter, and the mean intravenous 18 F-FDG dose used on the day of surgery was 559 (±104, 437-755) megabecquerels. A histologic diagnosis of malignancy was known to be lymphoma in 3 cases, colorectal carcinoma in 2, breast carcinoma in 1, and ovarian carcinoma in 1.  Table 1).
Two representative example cases of an 18 F-FDG-avid lesion seen on both same-day preoperative diagnostic whole-body 18 F-FDG PET/CT scan and same-day postoperative diagnostic limited field-of-view 18 F-FDG PET/ CT scan are shown in Figures 1 and 2.
Of the 32 18 F-FDG-avid lesions examined, only 1 18 F-FDG-avid lesion demonstrated a reduction in the lesion-to-background SUV max ratio from the preoperative to the postoperative 18 F-FDG PET/CT images. This particular 18 F-FDG-avid lesion was located in the ascending colon of a patient with colorectal carcinoma, having a preoperative 18 F-FDG-avid lesion SUV max of 7.9 (with a preoperative background SUV max of 1.0) and a postoperative 18 F-FDG-avid lesion SUV max of 7.5 (with a postoperative background SUV max of 1.2), resulting in a change in the lesion-to-background SUV max ratio of -1.7 from the preoperative to the postoperative study. Interestingly, on a subsequent followup diagnostic whole-body 18 F-FDG PET/CT scan performed 9 months after 18 F-FDG-directed surgery, the same area of this particular former 18 F-FDG-avid lesion in the ascending colon was no longer characterized as 18 F-FDG-avid, demonstrating a SUV max of 2.1 (with a background SUV max of 1.7).

Discussion
The results of the current retrospective data analysis, comparing preoperative and postoperative 18 F-FDG PET/CT imaging for 32 individual 18 F-FDG-avid lesions (not surgically manipulated or altered during 18 F-FDG-directed surgery, and for which all such 18 F-FDG-avid lesions were visualized on both preoperative and postoperative 18 F-FDG PET/CT imaging), yielded several very important observations. First, 18 F-FDG PET/CT imaging performed at extended injection-to-scan acquisition times of up to a mean time of 530 minutes (i.e., approximately 5 half-lives for 18 F-FDG) was able to maintain a designation of good/adequate diagnostic image quality deemed necessary for clinical interpretation. Second, the mean 18 F-FDG-avid lesion SUV max value increased significantly from preoperative to postoperative 18 F-FDG PET/CT imaging (7.7 to 11.3; P < 0.001). Third, mean background SUV max value decreased significantly from preoperative to postoperative 18 F-FDG PET/CT imaging (2.3 to 2.1; P = 0.017). Fourth, the mean lesion-to-background SUV max ratio increased significantly from preoperative to postoperative 18 F-FDG PET/CT imaging (3.7 to 5.8; P < 0.001). These collective observations from our current analysis have potential far-reaching implications regarding the currently held premises related to 18 F-FDG PET/CT oncologic imaging. Multiple investigators  have evaluated the concepts of delayed phase and dual-time-point diagnostic 18 F-FDG PET imaging approaches. In these numerous studies, attempts have been made to qualify and quantify the impact of the length of the injection-toscan time interval on differentiating malignant processes from benign processes. As one might expect, the findings reported amongst these various investigators have been highly variable, with some supporting the use of delayed phase and dual-time-point diagnostic 18 [65,78,89,90,94,101,102,109,115,116,123,129,130,132,135,137,139,140,147,148,150,151,156,158,164,166,168].
The inherent difference in intracellular glucose-6phosphatase levels, as it relates to benign cells and tumor cells, can be used to support the notion that the delayed phase and dual-time-point diagnostic 18 F-FDG PET/CT imaging approaches are advantageous [36,100,111,154,159,[170][171][172][173][174][175][176]. Initially, benign cells, such as in the case of inflammatory processes, may appear hypermetabolic as they transport increased number of glucose molecules into their cytoplasm. However, the glucose is not indefinitely retained secondary to the fact that those benign cells contain normal levels of intracellular glucose-6-phosphatase, thus allowing glucose molecules to subsequently exit the cytoplasm of those cells via glucose transporter membrane proteins. On the other hand, tumor cells have decreased levels of intracellular glucose-6-phosphatase, thus allowing for a continuous accumulation of 18 F-FDG into tumor cell over time. Therefore, methodologies that use a delayed phase in their diagnostic 18 F-FDG PET imaging approach should allow for an expected gradual decline in intracellular 18 F-FDG retention within initially hypermetabolicappearing benign tissues as compared to the continued accumulation of intracellular 18 F-FDG within malignant tissues [100,111,154,159]. Nevertheless, there are several reasons why the notion that delayed phase and dual-time-point diagnostic 18 F-FDG PET imaging approaches are advantageous may not be so simple and clear cut. First, it is well-recognized that there can be a significant degree of overlap in the pattern of 18 F-FDG uptake between benign tissues and various malignant tissues [154,159]. Second, there are substantial inherent variations in the methodology used in various delayed phase and dual-time-point diagnostic 18 F-FDG PET imaging protocols from institution to institution, with great variability in the timing of the initial scan and the delayed scan, as well as a general paucity of data where the delayed scan is performed at very extended injection-to-scan acquisition time intervals after the initial time of 18 F-FDG injection. Collectively, the vast majority of the reported series within the literature performed their delayed scan within approximately 1.5 to 2.5 hours from the initial time of 18 F-FDG injection [65,67,[70][71][72][73][74]79,82,83,[85][86][87][88][89][90][91][92][93][97][98][99][100][102][103][104]106,107,109,110,112,113,[115][116][117][118][119][120][121][122][123][124][125][126][127][129][130][131][132][133][134][135][137][138][139][140][141][142][143], and with far fewer series reporting their delayed scan at injection-to-scan acquisition times of approximately 3 hours or more from the initial time of 18 F-FDG injection [66,68,69,75-78,80,81,84,94-96, 101,105,108,111,114,128,136,144].
There  [144], who all performed delayed phase diagnostic F-FDG PET imaging at ultra-extended injection-to-scan acquisition time intervals, for which their clinical findings are particularly noteworthy of further discussion.
As pertaining specifically to 18 F-FDG PET imaging for brain tumors, there have been 3 clinical series that have reported successful delayed imaging extending out to ultra-extended injection-to-scan acquisition time intervals [81,136,144] [144]. In both series [81,144], they reported better tumor identification and delineation, and advocated the use of delayed intervals imaging. Horky et al. reported dualtime-point diagnostic 18 F-FDG PET imaging in patients treated with radiation for brain metastases, with delayed scans performed at a mean time of 225 minutes (range of 118 to 343 minutes) after the early scan done at 45 to 60 minutes after 18 F-FDG injection in a series of 32 patients [136]. They found that although the early and late SUV max values of the lesions alone did not differentiate residual tumor from post-radiation necrosis, the change in the lesion-to-gray matter early SUV max ratio to late SUV max ratio did.
Along similar lines for 18 F-FDG PET imaging of soft tissues masses, Lodge et al. reported a series of 29 patients in which a 6-hour 18 F-FDG PET imaging protocol was used [68]. In this protocol, a 2-hour dynamic emission data acquisition was performed after 18 F-FDG administration, followed by 2 further 30-minute static scans, which were started at 4 hours and 6 hours after 18 F-FDG administration. They found that the SUV value for high-grade sarcomas increased with time, reaching a peak SUV value at approximately 4 hours after initial 18 F-FDG administration, while benign soft tissue lesions reached a maximum SUV value within approximately 30 minutes after initial 18 F-FDG administration. They concluded that improved differentiation of high-grade sarcomas from benign soft tissue lesions was aided by SUV values derived from delayed intervals imaging.
Likewise, for 18 F-FDG PET imaging of non-small cell lung cancer, Basu et al. reported on 3 patients in whom an 8-hour 18 F-FDG PET imaging protocol was used [111]. In this protocol, 18 F-FDG PET imaging was performed, starting at 5 minutes, and continuing at 1, 2, 4, 6, and 8 hours after initial 18 F-FDG administration. They found that sites of non-small cell lung cancer showed a progressive increase in 18 F-FDG uptake over the 8-hour course, while surrounding normal tissues demonstrated either a declining or stable pattern of 18 F-FDG uptake with time. They concluded that delayed injection-toscan acquisition time intervals had "implications in detecting malignant lesions with greater degree of certainty"…"due to better contrast between the abnormal site and the surrounding background".
Of last mention, similar recommendations for the use of delayed injection-to-scan acquisition time interval imaging have been made by other investigators at somewhat less extended injection-to-scan acquisition time intervals of approximately 3 hours in breast cancer [66,105,128], cervical cancer [76,77], hepatocellular cancer [84], biliary malignancies [95], lung cancer [75,96,108], and thymic epithelial tumors [114].
The results of the previously reported series demonstrating their ability to successfully perform delayed imaging at extended injection-to-scan acquisition time intervals of approximately 3 hours or more from the initial time of 18 F- FDG injection [66,68,69,[75][76][77][78]80,81,84,[94][95][96]101,105,108,111,114,128,136,144], as well as those demonstrating the added value to performing delayed imaging at extended injection-to-scan acquisition time intervals of approximately 3 hours or more from the initial time of 18 F-FDG injection [66,68,[75][76][77]81,84,95,96,105,108,111,114,128,136,144], are all highly consistent with the results of our current retrospective data analysis. It is clear that our currently presented data, demonstrating increasing 18 F-FDG-avid lesion SUV max values, decreasing background SUV max values, and increasing lesion-tobackground SUV max ratios from preoperative to postoperative 18 F-FDG PET/CT imaging, supports the potential utility of delayed phase and dual-time-point diagnostic 18 F-FDG PET/CT imaging. This suggests that delayed scans performed at an appropriately selected extended injectionto-scan acquisition times can potentially minimize or alleviate the issue of overlap in the pattern of 18 F-FDG uptake between benign tissues versus malignant tissues, as well as between background tissues versus malignant tissues. This phenomenon appears to be the temporal outcome of a resultant gradual accumulation of 18 F-FDG within malignant tissues and continued decreased background level of 18 F-FDG within the surrounding normal tissues, thus leading to a progressive increase in the lesion-to-background SUV max ratio. A key element to this overall line of reasoning, as it relates to the proper use of 18 F-FDG in molecular imaging, is the recognition of the negative impact of "background" issues, and "not signal", as recently eloquently described by Frangioni [177], but which was recognized early on in the evolution of PET imaging by Hoffman and Phelps [178]. This time-dependent phenomenon observed in our current retrospective analysis is consistent with our previously reported findings regarding same-day preoperative diagnostic whole-body 18 F-FDG PET/CT images and sameday perioperative ex vivo surgical specimen 18 F-FDG PET/ CT imaging, in which we observed similar trends of increased 18 F-FDG accumulation in 18 F-FDG-avid lesions within ex-vivo surgical specimens and of decreased 18 F-FDG activity within adjacent normal tissues [37]. However, we fully acknowledge and recognize that significant further investigations are warranted to better assess this phenomenon and to formally evaluate the clinical usefulness of extended injection-to-scan acquisition time intervals in various diagnostic 18 F-FDG PET/CT oncologic imaging applications.
Analogous to our current discussions regarding the evaluation and quantification of 18 F-FDG-avid lesions and corresponding background tissues at these extended injection-to-scan acquisition time intervals for 18 F-FDG PET imaging approaches, there have been two groups of investigators utilizing 18 F-FDG-directed surgery [11,17,21], other than our own collaborative group [51], who have previously examined the equivalent question as it pertains to the impact of the length of time from injection of 18 F-FDG to the performance of intraoperative gamma detection probing [11,17,21]. One such group [17,21] recognized that there was an increased tumor-to-background ratio of 18 F-FDG seen during intraoperative gamma detection probing when there was a longer duration (i.e., up to 6 hours of time) from injection of the 18 F-FDG dose to intraoperative probing. However, they did not endorse lengthening the duration from injection of the 18 F-FDG dose to performing intraoperative gamma detection probing or to performing perioperative 18 F-FDG PET imaging [21]. Instead, they specifically commented that lengthening the duration from injection of the 18 F-FDG dose "might compromise image quality as a result of lower count rates" [21]. The other such group [11], as based upon the evaluation of 18 F-FDG count rates for only three patients, concluded that intraoperative gamma detection probing was "more suitable" at 1 to 3 hours post-injection of 18 F-FDG as compared to 6 to 7 hours post-injection of 18 F-FDG. In both instances, these two groups of investigators fell short of recognizing the potential efficacies of extended injection-to-scan acquisition time intervals.
Although we clearly recognize that the current retrospective data analysis is based upon only 32 individual 18 F-FDG-avid lesions, the potential significance of our current collective observations is far-reaching for 18 F-FDG PET/CT oncologic imaging. While the possibility of ultra-extended injection-to-scan acquisition time intervals of up to approximately 5 half-lives for 18 F-FDG was first alluded to in the dose uptake ratio simulation studies by Hamberg et al. in 1994 [179] and was later clinically examined by Lodge [144], its potential future impact has not previously been fully realized within the nuclear medicine or surgical literature. The ability to maintain good/adequate diagnostic image quality for 18 F-FDG PET/CT imaging at extended injection-to-scan acquisition time intervals of up to approximately 5 halflives and the resultant time-dependent increase in the observed 18 F-FDG-avid lesion SUV max values, decrease in the observed background SUV max values, and increase in the lesion-to-background SUV max ratios allow for and justify the more widespread and integrated, real-time use of diagnostic 18 F-FDG PET/CT imaging in conjunction with 18 F-FDG-directed interventional radiology biopsy procedures and ablation procedures, as well as with 18 F-FDG-directed surgical procedures. Such integrated, real-time utilities for diagnostic 18 F-FDG PET/CT imaging would facilitate periprocedural verification of appropriate tissue targeting during 18 F-FDG-directed interventional radiology biopsy procedures and ablation procedures and for perioperative verification of appropriate tissue targeting and completeness of resection during 18 F-FDG-directed surgical procedures. Furthermore, these resultant timedependent observations could have far-reaching impact on potentially re-shaping future thinking regarding what represents the "most optimal" injection-to-scan acquisition time interval for all routine diagnostic 18 F-FDG PET/CT oncologic imaging, as the current procedure guideline for tumor imaging with 18 F-FDG PET/CT, as published by the Society of Nuclear Medicine, simply states that "emission images should be obtained at least 45 minutes after radiopharmaceutical injection" [63].

Conclusions
Our current retrospective data analysis demonstrates that 18 F-FDG PET/CT oncologic imaging can be successfully performed at extended injection-to-scan acquisition time intervals of up to approximately 5 half-lives for 18 F-FDG while maintaining good/adequate diagnostic image quality. The resultant increased 18 F-FDG-avid lesion SUV max values, decreased background SUV max values, and increased lesion-to-background SUV max ratios seen from preoperative to postoperative 18 F-FDG PET/CT imaging have great potential for allowing for the integrated, real-time use of 18 F-FDG PET/CT imaging in conjunction with 18 F-FDG-directed interventional radiology biopsy and ablation procedures and 18 F-FDG-directed surgical procedures, as well as have far-reaching impact on potentially re-shaping future thinking regarding the "most optimal" injection-to-scan acquisition time interval for all routine diagnostic 18 F-FDG PET/CT oncologic imaging. In these regards, we fully acknowledge and recognize the need for further investigations to better assess and formally evaluate the clinical utility of extended injectionto-scan acquisition time intervals in various diagnostic 18 F-FDG PET/CT oncologic imaging applications.

Competing interests
All the authors declare that they have no competing interests to report.
Authors' contributions SPP was responsible for the overall study design, data collection, data organization, data analysis/interpretation, writing of all drafts of the manuscript, and has approved final version of the submitted manuscript. DAM was involved in study design, data collection, data organization, data analysis/interpretation, writing portions of the manuscript, and has approved final version of the submitted manuscript. SMS was involved in data organization, data analysis, and has approved final version of the submitted manuscript. EWM was involved in discussion about study design, data analysis/interpretation, critiquing drafts of the manuscript, and has approved final version of the submitted manuscript. NCH was involved in study design, discussion about data analysis/interpretation, editing portions of the manuscript, and has approved final version of the submitted manuscript.