Median TBR in lesions with tumor, 2.1; median TBR in lesions without tumor, 0.8; P = .01.
The area under the curve is 0.72 (P < .001).
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
El Lakis M, Gianakou A, Nockel P, et al. Radioguided Surgery With Gallium 68 Dotatate for Patients With Neuroendocrine Tumors. JAMA Surg. 2019;154(1):40–45. doi:10.1001/jamasurg.2018.3475
What intraoperative techniques and thresholds define positive lesions that warrant resection during gallium 68–dotatate radioguided surgery for neuroendocrine tumors?
In this cohort study, a tumor to background ratio of 2.5 or more was a highly sensitive threshold for indicating a lesion to be consistent with a neuroendocrine tumor on histologic findings. The omentum should be used as the background count activity for gallium 68–dotatate radioguided surgery for patients with abdominal neuroendocrine tumors.
Areas with a tumor to background ratio of 2.5 or more warrant surgical resection.
Neuroendocrine tumors (NETs) express somatostatin receptors, which can be targeted with radiolabeled peptides. In a variety of solid tumors, radioguided surgery (RGS) has been used to guide surgical resection. Gallium 68 (68Ga) dota peptides have been shown to be more accurate than other radioisotopes for detecting NETs. A pilot study previously demonstrated the feasibility and safety of 68Ga-dotatate RGS for patients with NETs.
To evaluate what intraoperative techniques and thresholds define positive lesions that warrant resection during 68Ga-dotatate RGS.
Design, Setting, and Participants
This prospective cohort study, conducted between October 23, 2013, and February 14, 2018, included 44 patients with NETs who underwent 68Ga-dotatate RGS.
Gallium 68–dotatate RGS.
Main Outcomes and Measures
The in vivo and ex vivo tumor to background ratio (TBR) was assessed for resected lesions and correlated with the histopathologic findings.
Forty-four patients (22 women and 22 men; mean [SD] age, 51.0 [13.7] years) had 133 lesions detected on preoperative imaging scans, with a diagnosis of a pancreatic NET (19 of 44 [43%]), gastrointestinal NET (22 of 44 [50%]), and pheochromocytoma or paraganglioma (3 of 44 [7%]). The TBR was obtained by normalizing to the omentum (106 of 133 [79.7%]) or other solid organs (27 of 133 [20.3%]). The omentum had a significantly lower mean (SD) count than other solid organs for background count activity 3 hours after injection (22.1 [17.0] vs 34.5 [39.0]; P < .001). The lesions containing NETs had a higher TBR than those that did not contain NETs (18.9 vs 4.4; P < .001). On a receiver operating characteristic curve analysis, a TBR of 2.5 had a sensitivity of 90% and a specificity of 25%, and a TBR of 16 had a sensitivity of 54% and a specificity of 81%.
Conclusions and Relevance
A TBR of 2.5 or greater is a highly sensitive threshold for indicating a lesion to be consistent with a NET on histologic findings and thus warranting surgical resection. The omentum should be used as the background count activity for 68Ga-dotatate RGS for patients with abdominal NETs.
A unique and important feature of neuroendocrine tumors (NETs) is that they express somatostatin receptors, which can be targeted with radiolabeled peptides for imaging and treatment.1,2 Positron emission tomography (PET) with computed tomography (CT) using somatostatin analogues labeled with positron-emitting gallium 68 (68Ga) dota peptides has been shown to be more accurate than other agents for detecting many types of NETs and is an important imaging modality that affects the clinical management of patients with NETs.3-5
Operative intervention in patients with gastroenteropancreatic NETs (GEPNETs) and pheochromocytoma (PC) or paraganglioma (PGL) can be challenging in several clinical scenarios. Some patients may have functional tumors too small to localize during operative exploration. Persistent or recurrent disease can result from missed locoregional metastatic disease at initial surgical exploration. Incomplete surgical resection or positive margins are important prognostic factors in patients with GEPNETs and PC or PGL.6,7 Furthermore, reoperation for recurrent or persistent locoregional GEPNETs and PC or PGL is challenging because of scar tissue, altered anatomy, and—in some cases—low volume of disease but functioning tumors.
Radioguided surgery (RGS) for a variety of solid tumors provides real-time information of the location, extent of disease, completeness of resection, and confirmation of the resected tumor.8-10 Radioguided surgery has been shown to be useful for sentinel lymph node biopsy in breast cancers and melanoma or when used as an adjunct tool to help find parathyroid adenomas in reoperative cases.10-12 Radioguided surgery takes advantage of radiopharmaceuticals that are preferentially taken up by the tumor tissue. Each radionuclide has its own physical properties to consider, such as physical half-life and photon yield (emission probability per decay). The choice of a radiopharmaceutical for an RGS procedure is done based on the radionuclide’s physical properties, the radiotracer’s pharmacokinetics, and the agent that yields the highest tumor to background ratio (TBR). A pilot study previously demonstrated that RGS with 68Ga dotatate is feasible and safe in patients with GEPNETs.13 In the current prospective cohort study, we evaluated what intraoperative criteria define positive lesions and warrant resection during 68Ga-dotatate RGS.
A total of 326 patients with NETs were evaluated at the National Institutes of Health Clinical Center in Bethesda, Maryland, and underwent 68Ga-dotatate–labeled PET or CT imaging between October 23, 2013, and February 14, 2018. Forty-four patients were surgical candidates and had RGS based on positive results of a preoperative 68Ga-dotatate–labeled PET or CT scan. At the time of the operation, 5 mCi of 68Ga dotatate was injected intravenously. A handheld gamma probe (Neoprobe Model 2300; Neoprobe Corporation) was used during the surgical explorations. Background count rates (counts per second) were obtained from the omentum or solid organs. The gamma probe was used to explore the abdomen in a systematic fashion, and in vivo count rates of each tumor area were obtained. For each lesion, the target count was determined based on the highest count rate and normalized to the background count rate to calculate the TBR. This prospective study was approved by the National Cancer Institute review board and the National Institutes of Health Radiation Safety Committee (NCT01967537). Patients provided written informed consent.
All patients in the cohort had focal uptake of 68Ga dotatate on PET or CT scans and underwent a preoperative CT scan of the abdomen and pelvis and/or magnetic resonance imaging. Computed tomographic scans of the chest, abdomen, and pelvis were obtained using the SOMATOM Definition AS and SOMATOM Definition Flash (Siemens Medical Solutions) and the Aquilion ONE (Toshiba Medical Systems). The section thickness for CT scanning was 2 mm, and all studies were performed with intravenous nonionic water-soluble contrast agent as well as oral contrast. Magnetic resonance imaging scans of the abdomen and pelvis were obtained with 1.5-T and 3-T scanners (Achieva 1.5T and 3T [Philips]; and Verio 1.5T [Siemens Medical Solutions]). The image thickness was 6 mm. Images in the axial plane were obtained before and after injection of gadolinium-diethylenetriamine pentaacetic acid contrast agent. Positron emission tomographic scans and CT scans from the upper thighs to the skull were acquired 60 minutes after intravenous injection of 68Ga dotatate (mean [SD] administered activity, 5 [0.57] mCi). All PET and CT scans were acquired on Biograph mCT 64 and Biograph mCT 128 PET and CT scanners (Siemens Medical Solutions). Positron emission tomographic images were reconstructed using an iterative algorithm provided by the manufacturer, which also uses point-spread function and time of flight. Low-dose CT studies for attenuation correction and anatomic coregistration were performed without contrast.
Demographic and clinical data, including primary disease site, age, sex, imaging information, and pathologic findings, were collected prospectively and analyzed. The surgically resected tumor and tissue specimens were divided into the following 2 groups: group 1, lesions with disease on final histologic examination; and group 2, lesions without evidence of disease on final histologic examination.
Two group comparisons were performed using t test, χ2 test, and Mann-Whitney test as appropriate. All P values were from 2-sided tests and results were deemed statistically significant at P < .05. A receiver operating characteristic curve analysis was plotted to select the optimal TBR threshold. Statistical analyses were conducted using SPSS, version 21.0, for Windows (SPSS Inc).
Forty-four patients (22 women and 22 men) underwent RGS with 68Ga dotatate. The demographic and clinical characteristics of the patients are summarized in Table 1. Most patients had an open laparotomy (40 [91%]), and 24 of the open laparotomies (60%) were redo operations. The omentum was most commonly used as the background for normalizing the uptake as compared with solid organs (106 of 133 lesions [79.7%] vs 27 of 133 [20.3%]) (Table 2). The omentum had a significantly lower mean (SD) count compared with solid organs normalized at 3 hours after injection (22.1 [17.0] vs 34.5 [39.0]; P < .001). Of 133 resected lesions, 100 (75.2%) were positive for a NET on histologic findings. Most of the NETs (76 of 100 [76%]) were World Health Organization grade 1 tumors, and approximately half were lymph node metastases (71 of 133 [53.4%]) (Table 2). The median 68Ga-dotatate TBR for all surgically resected lesions was 13 (range, 0.8-486).
The lesions with NETs (group 1) had a higher mean maximum standardized uptake value on results of preoperative 68Ga-dotatate–labeled PET or CT imaging than did the lesions without NETs (59.4 vs 35.2; P = .03) and had a higher intraoperative median TBR (18.9 vs 4.4; P < .001) (Table 3). In group 1, the median TBR was significantly higher in lesions with grade 1 NETs compared with grade 2 NETs (22.9 vs 6.45; P = .02,). There was no significant difference in the TBR in group 1 vs group 2 lesions by recurrent disease status (66 of 100 [66%] vs 20 of 33 [61%]; P = .36) or by primary tumor site of origin. Lesions with no NETs were mostly lymph nodes (21 of 33 [64%]) and the remaining were from the pancreas, liver, and gastrointestinal tract origin. The solid lesions originating from the liver, pancreas, gastrointestinal tract, and adrenal gland had similar median TBRs compared with lesions originating from the lymph nodes (13.0 vs 15.5; P = .24). The 33 lesions that did not contain tumor were present in 19 patients (43%). In 5 patients (11%) who were receiving cold somatostatin analogues preoperatively, the TBR was similar compared with patients who were not receiving cold somatostatin analogues (P = .07). The ex vivo count was assessed in 66 lesions, 56 of which contained a NET and had a significantly higher TBR compared with lesions without NETs on histologic findings (2.1 vs 0.8; P = .01) (Figure 1). On receiver operating characteristic curve analysis, a TBR of 2.5 had a sensitivity of 90% and a specificity of 25%, and a TBR of 16 had a sensitivity of 54% and specificity of 81%, with an area under the receiver operating characteristic curve of 0.72 (P < .01) (Figure 2). When evaluating whether lesions were visible (including intraoperative ultrasound) and/or palpable or only detected with the gamma probe in a subset of RGS cases, 5 of 39 lesions (13%) were detected only with the gamma probe.
Of the 49 of 133 lesions (36.8%) seen on preoperative magnetic resonance imaging scans, 43 (88%) were NETs and had a significantly higher TBR as compared with lesions seen on magnetic resonance imaging and negative on histologic findings (21.0 vs 2.7; P = .04). Of the 64 lesions seen on preoperative CT scans, 60 were NETs on histologic findings and had a higher TBR as compared with lesions seen on CT scan and negative on histologic findings (23.1 vs 6.4; P = .40).
In this study, we evaluated what threshold should be used for 68Ga-dotatate RGS. We found that a TBR of 2.5 had the highest sensitivity, of 90%. Furthermore, the best intra-abdominal organ for normalization of the target count (TBR) was the omentum. Radioguided surgery may be helpful to guide the resection of occult nodal metastasis, as well as intraoperative assessment and determination of the adequacy of surgical margins. It may also be helpful in identifying and removing small tumors that are not visible or palpable in patients with recurrent NETs where the operative field is obscured by scar tissue. Radioguided surgery allows surgeons to distinguish tumor tissue from scar tissue, thereby significantly aiding the extirpation of all sites of the tumor or avoiding the surgical exploration of a site with low TBR even if suspicious lesions are seen on results of preoperative imaging studies.
Gallium 68 dota peptides have been shown to be superior to other analogues such as indium 111 pentetreotide, iodine 123 metaiodobenzylguanidine, or technetium 99 sulfur in detecting NETs.4 Radioguided surgery with 68Ga dotatate in patients with GEPNETs has been demonstrated to be safe and feasible,13 but the characteristics—such as TBR cutoff as an indication to resect the tumor—have not been established, nor has the utility of RGS for confirming a resected NET as an adjunct to a frozen section. In this study, we identify an optimal cutoff of 2.5 for the TBR. This information can then be applied to guide the extent of surgical resection; localize tumor-bearing tissue, especially in reoperations with extensive scar tissue; and confirm the successful resection of NETs.
The premise of RGS is to facilitate the removal of all sites of disease and allow for complete tumor resection. In the case of redo laparotomy, a focused screening of the suspicious area could be performed, and the count would be done in vivo to localize the tissue that needs to be removed. In addition, after the removal of the lesion, an ex vivo count can be measured to confirm that the pathological tissue has been resected. Although the ex vivo count was not determined in every case during this study, the ex vivo median TBR was 2.1 in 56 of 66 lesions that were positive for NETs. Last, the area of interest after tumor resection can be reevaluated to check for potential residual activity. In 7 of 8 cases, the resection bed activity was below the median preresection TBR (2.13 vs 32.6).
Gallium 68 dota peptides have a relatively fast clearance, with a half-life of 68 minutes. The time to enter the abdomen and resect the target lesion(s) varies among patients for different considerations; even in the same surgery, different target lesions may be removed at different time intervals. To overcome this limitation, the target lesion count is normalized to the background at the same time from tracer dose injection, and the TBR is used. We found no significant difference in TBR in samples with pathologically confirmed NETs by time from injection. The background reflects the physiological uptake of normal tissue. In this study, we used the omentum as the background in most cases (106 of 133 lesions [79.7%]), as the activity was lowest in this tissue. Earlier during our investigation, solid organs were also used, but they had high activity. The use of the omentum has several advantages in addition to the lower activity. It is easily identified and accessible regardless of the type of incision and the extent of adhesions. The omentum uptake is minimal and homogeneous when compared with solid organs such as the liver, pancreas, kidneys, and spleen, which have high physiological uptake of 68Ga dotatate.14 Finally, the omentum is rarely involved with tumor tissue, as carcinomatosis is rare in patients with NETs.
Previous reports have concluded that a TBR of at least 1.5 is required for confirmation of the tissue localization using indium 111-DTPA-octreotide.15 In our series, the median TBR for all resected tumors was 13 and, as expected, lesions consistent with NETs had a significantly higher TBR than those negative for NET histologic characteristics (18.9 vs 4.4; P < .001). A total of 21 of 33 lesions (64%) not harboring a tumor had a TBR of less than 5, but 6 of 33 lesions (18%) had an elevated TBR. On receiver operating characteristic curve analysis, a TBR of 2.5 showed a sensitivity of 90% and specificity of 25%, and a TBR of 15 showed a sensitivity of 54% and specificity of 81%. When using the data for surgical management intraoperatively to facilitate complete resection, we recommend removing all tissues with a TBR of 2.5 or higher when possible. In addition, we recommend clinical judgment in addition to a lower TBR threshold for obtaining extra margins when screening the tumor bed after a tumor is removed.
Somatostatin analogues such as octreotide and lanreotide are an accepted option for the treatment of patients with symptomatic NETs. Like the PET tracer 68Ga dotatate, octreotide binds to somatostatin receptor subtype 2, where it acts as an agonist. However, it was shown that treatment with a long-acting somatostatin analogue did not significantly reduce 68Ga-dotatate binding in NETs.16 In fact, our results of not finding differences in the TBR in patients receiving and not receiving somatostatin analogue treatment during 68Ga-dotatate RGS surgery is consistent with the literature on 68Ga-dotatate–labeled PET and CT imaging.
The mean (SD) radiation exposure to the surgeons and operating room personnel during 68Ga-dotatate RGS reported in a previous pilot study was 13.2 (9.1) mellirem (mrem) (range, 4-27 mrem) (annual limit for occupational workers is 5000 mrem per year).13 Thus, we have stopped monitoring radiation exposure during 68Ga-dotatate RGS, as it was associated with safe levels of exposure according to the National Institutes of Health Radiation Safety Committee.
This study has some limitations, such as the small number of participants, but NETs are rare. Also, the omentum was not routinely used in all cases at the beginning of our experience; thus, a TBR cutoff of 2.5 may have an even higher accuracy. Last, the relatively short half-life of 68Ga may affect its utility in very long cases (>3 hours), but the trace-dose injection time can be modified to optimize target lesion activity detection within the optimal time window based on the progress of the surgical exploration.
For a more uniform TBR, the omentum is a reliable background for 68Ga-dotatate RGS, when free of disease. A TBR of 2.5 or more is a highly sensitive threshold for guided surgical resection of suspicious lesions. Long-term follow-up and comparison with patients who did not have 68Ga-dotatate RGS are needed to determine if such a surgical approach reduces the rate of persistent or recurrent disease in patients with GEPNETs and PC or PGL.
Accepted for Publication: June 2, 2018.
Corresponding Author: Electron Kebebew, MD, Department of Surgery, Stanford University, 300 Pasteur Dr, H3642, Stanford, CA 94305 (email@example.com).
Published Online: September 26, 2018. doi:10.1001/jamasurg.2018.3475
Author Contributions: Drs El Lakis and Kebebew had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: El Lakis, Nockel, Wiseman, Sadowski, Kebebew.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: El Lakis, Nockel, Wiseman, Kebebew.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: El Lakis, Giannakou.
Obtained funding: Kebebew.
Administrative, technical, or material support: Giannakou, Tirosh, Patel, Pacak, Kebebew.
Supervision: El Lakis, Nilubol, Sadowski, Kebebew.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by the intramural program of the Center for Cancer Research, National Cancer Institute Intramural Research Program.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Create a personal account or sign in to: