Role of Combined [68Ga]Ga-DOTA-SST Analogues and [18F]FDG PET/CT in the Management of GEP-NENs: A Systematic Review

Gastro-entero-pancreatic neuroendocrine neoplasia (GEP-NENs) are rare tumors, but their frequency is increasing. Neuroendocrine tumors normally express somatostatin (SST) receptors (SSTR) on cell surface, especially G1 and G2 stage tumors, but they can show a dedifferentiation in their clinical history as they become more aggressive. Somatostatin receptor imaging has previously been performed with a gamma camera using [111In]In or [99mTc]Tc-labelled compounds, while [68Ga]Ga-labelled compounds and PET/CT imaging has recently become the gold standard for the diagnosis and management of these tumors. Moreover, in the last few years 18F-fluorodeoxyglucose ([18F]FDG) PET/CT has emerged as an important tool to define tumor aggressiveness and give relevant prognostic information, particularly when coupled with [68Ga]Ga-labelled SST analogues PET/CT. This review focuses on the importance of combined imaging with [68Ga]Ga-labelled SST analogues and [18F]FDG for the management of GEP-NENs.


Introduction
Gastro-entero-pancreatic neuroendocrine neoplasia (GEP-NENs) or neuroendocrine tumors (NET) are rare and heterogeneous diseases, with increasing incidence and prevalence over the last decades [1]. Their prognosis is affected by a number of factors, including the primary tumor site, grading and staging [2][3][4].
The vast majority of these tumors express somatostatin (SST) receptors (SSTR) on tumor the cell surface, a feature that may be used for both diagnostic and therapeutic purposes. In fact, in addition to conventional cross-sectional imaging procedures, somatostatin receptors functional imaging tests (FITs) (i.e., 68 Ga-labelled SST analogues, 111 In-labelled SST analogues and 99m Tc-labelled SST analogues) are recommended in GEP-NEN patients at the time of disease diagnosis, as well as during patient follow-up [5].
Hybrid positron emission tomography and computed tomography with 68 Ga-labelled SST analogues ([ 68 Ga]Ga-DOTA-SST PET/CT) is considered the gold standard technique in GEP-NENs, due to its high sensitivity and specificity, which are reported to vary between 91-95% and 82-97%, respectively [6].
Positron emission tomography/computed tomography with 18 F-fluorodeoxyglucose ([ 18 F]FDG PET/CT) has been suggested as an alternative tool for tissue sampling for the assessment of the aggressiveness of tumors, and it has shown prognostic value in NENs [7].
In the present review, the role of [ 68 Ga]Ga-DOTA-SST PET/CT and [ 18 F]FDG PET/CT in NENs was investigated, focusing on the impact of their combined use in the clinical practice.

Conventional Somatostatin Receptors Imaging
Somatostatin receptors (SSTRs) are highly expressed on neuroendocrine cells. Octreotide was the first synthetic somatostatin analogue used in nuclear medicine to image SSTRs expression. For many years, SSTR imaging has been performed as octreotide scan with a gamma camera, although, in the last few years, PET/CT has been established as the best technique to image SSTRs.
Somatostatin receptor scintigraphy (SRS) with [ 111 In]In-pentetreotide (Octreoscan ® , Mallinckrodt Medical, St. Louis, MO, USA), first performed in 1989, is based on the specific binding of a radiolabeled somatostatin analogue (octreotide) to high affinity somatostatin receptors (mainly type 2) expressed by most neuroendocrine tumors. Imaging is generally performed at four and 24 h after the i.v. administration of Octreoscan ® with a two-headed gamma camera equipped with a medium energy collimator. SRS shows a good accuracy for whole body imaging and has been routinely used for the diagnosis and follow-up of neuroendocrine tumors [8,9].
The principal limitation of the use of SRS with Octreoscan ® is the low spatial resolution of the gamma camera that mainly affects the evaluation of small lesions, particularly in organs with a high physiological uptake (for example the liver). The introduction of SPECT/CT hybrid imaging has improved the accuracy of SRS-increasing its spatial resolution and the anatomic localization of pathologic sites of increased uptake due to CT co-registration, with a reduction of false positive results [10].

PET/CT with 68 Ga-labelled Peptides
SRS-PET with 68 Ga-labelled peptides has recently become the gold standard in the diagnosis and management of well differentiated neuroendocrine tumors. It has proven to be better than SRS imaging with gamma-emitting isotopes because of its higher sensitivity compared to Octreoscan. [28,29] Furthermore, the use of PET/CT, as compared to SPECT/CT, offers higher spatial resolution, lower scan times and a lower patient radiation exposure per scan [30,31].
[ 68 Ga]Ga-DOTA-NOC, [ 68 Ga]Ga-DOTA-TOC, and [ 68 Ga]Ga-DOTA-TATE, despite their different affinity for somatostatin receptors SSRs (they bind especially type 2, but also type 3 and type 5), have shown no differences in clinical practice; the three of them are widely used [32]. DOTA-TATE, DOTA-TOC and DOTA-NOC can also be labelled with Lutetium-177 ( 177 Lu) and Yttrium-90 ( 90 Y) for peptide receptor radionuclide therapy (PRRT), although DOTA-TATE is the most frequently used in clinical practice.
For SRS-PET, fasting is not required, and the scan starts generally 45-60 min after the intravenous administration of the labelled compound. Some authors suggest discontinuing "cold" octreotide therapy to avoid a possible SSTR blockade [32]. Physiological [ 68 Ga]Ga-DOTA-SST analogue biodistribution includes the uncinate process of the pancreas, spleen, liver, adrenal glands, pituitary gland and urinary tracts. Possible pitfalls include inflammatory processes (leucocytes and macrophages express SSR2), splenosis, osteoblastic activity (degenerative bone disease, fracture, hemangioma, epiphyseal growth plates), and meningiomas. In addition, other tumors can overexpress SSTR, such as pheochromocytoma, paraganglioma and neuroblastoma [33]. Previous studies found that SUV max values were significantly higher in patients with well-differentiated neuroendocrine carcinomas than with poorly differentiated carcinomas, whereas no correlation was found between SUV max and Ki67 [34].
Overall, almost 90% of G1-G2 GEP-NENs present a positive finding due to the high SSTR expression on cell surface of these tumors. Recently, a negative correlation between tumor proliferative activity, expressed by Ki67, and [ 68 Ga]Ga-DOTA-TATE, expressed as SUV max , has been observed in a retrospective analysis including 126 GEP-NENs [35]. The proposed SUV max cut-off able to discriminate between G1-G2 vs G3 GEP-NENs was 11.2, thus suggesting that patients with a lower SUV max value present a significant higher risk of having a more aggressive and highly proliferating G3 NEN.
A possible correlation between SUV max and Ki67 has also been investigated by other studies, however, with heterogeneous findings. In the paper by Partelli and coworkers, the median SUV max value varied between 31.5 and 53.5 in G1-G2 pancreatic NENs, whereas it was 16.5 in the G3 subgroup [36]. A prognostic role of SUV max was proposed by Ambrosini et al., who observed a significantly longer progression-free survival in patients with an SUV max value > 38 compared with those with a lower SUV max value, in a series of G1-G2 pancreatic NENs (p = 0.002) [37]. Indeed, a lower cut-off value was proposed in the study by Sharma et al., in which the authors observed significantly different progression-free survival curves when patients were stratified according to SUV max values with a cut-off of 14.5 [38].
All together, these results show that, although it is clear that a high SUV max value means a better clinical outcome, the best cut-off value that is able to definitively stratify patients in different subgroups with different prognosis still needs to be identified.
Nevertheless, there is solid scientific evidence showing the positive impact of PET/CT with [ 68 Ga]Ga-DOTA-SST analogues on GEP-NEN clinical management, particularly in detecting distant metastases [39]. A recent systematic review and metanalysis, including 14 studies, showed that PET/CT findings resulted in management change in 44% of patients, thus confirming the pivotal role of this technique in the management of NEN patients [30].

PET/CT Imaging with [ 18 F]FDG
[ 18 F]FDG (fluorodeoxyglucose) is the most common radiopharmaceutical used in PET imaging, especially for oncologic purposes. FDG is a glucose analogue that is actively transported by specific glucose transport proteins (GLUT, particularly GLUT1 and 3) into the cell and then phosphorylated by hexokinase. After phosphorylation, glucose normally enters the glycolysis pathway, but FDG cannot and remains trapped into the cell. Tumor cells, due to their higher metabolic activity, show an increased number of glucose transporters and hexokinase, leading to a higher FDG uptake than normal tissues [40]. PET images are generally acquired one hour after the intravenous injection of [ 18 F]FDG ( 18 F-half life: 109 min). Patients are required to fast at least six hours before injection, and blood glucose levels must not exceed 200 mg/dL, although 180 mg/dL is desirable. [ 18 F]FDG PET/CT plays a very small role and has low sensitivity in small growing well differentiated neuroendocrine tumors (G1 and G2), but its role is emerging in the evaluation and management of high-grade NENs (G3). Indeed, with time, NENs may show a de-differentiation, losing their ability to express somatostatin receptors and increasing their metabolism and FDG avidity. Many studies have demonstrated a positive correlation between Ki67 expression and [ 18 F]FDG SUV max [41,42]. This finding suggests that [ 18 F]FDG PET/CT has an important prognostic value in high grade NENs.
[ 18 F]FDG PET/CT is commonly performed in aggressive NENs, as there is emerging evidence that the presence of increased glucose metabolism in NENs correlates with the aggressiveness of tumors and bad prognosis [7]. Though a clear relationship between [ 18 F]FDG-PET/CT positivity and an unfavorable clinical outcome has been reported, particularly in aggressive G3 tumors, its role in more indolent G1 and G2 GEP-NENs still remains unclear. The European Neuroendocrine Tumors Society guidelines do not recommend [ 18 F]FDG-PET/CT in NEN patients unless a G3 grading is present [43,44]. Conversely, several papers have reported its utility in patients with low grade (G1-G2) tumors-in whom it may provide useful prognostic information-in order to improve patients' clinical management. [ 18 F]FDG PET/CT sensitivity in G1-G2 GEP NENs is reported to range between 40% and 60%, whereas it increases to almost 95% in G3 tumors [42,45]. However, [ 18 F]FDG PET/CT positivity in GEP-NEN lesions does not depend on grading alone-it also depends on tumor aggressiveness, differentiation, and GLUT expression [45].
According with European Neuroendocrine Society grading system [46,47], the grade 1 NENs group includes tumors with very low proliferative activity, their Ki67 limit being <3% or ≤2% in pancreatic or gastrointestinal primaries, respectively. Grade 2 (G2) NENs are diagnosed when tumors have a higher Ki67 value, though it still remains below 20%. A Ki67 value of > 20% leads to a diagnosis of G3 tumors (in the pancreatic primaries, G3 NENs are further classified according with tumor morphology in NEN G3 if well-differentiated or NEC G3 if poorly-differentiated) [48].
Patients' prognosis dramatically depends on grading systems. In fact, G1 tumors are considered indolent diseases with an excellent long-term clinical outcome even in the setting of advanced disease. Their five-year survival rates are 60-80% and 40-60% in pancreatic and gastrointestinal primaries stage IV diseases, respectively. On the opposite side, G3 NENs, particularly if poorly-differentiated morphology is present, need to be considered aggressive diseases, with five-year survival rates ranging between 10% and 30% [2,4].
Based on these considerations, these tumors are expected to have a very different glycolytic activity, resulting in an extremely different probability to have [ 18 F]FDG PET/CT positivity, which is common in G3 NENs, whereas it is infrequent in less proliferating tumors.
The availability of an accurate non-invasive diagnostic tool, such as [ 18 F]FDG PET/CT, which is able to predict tumor behavior, may help in the early identification of those patients with unfavorable clinical outcome. This is even more significant considering that when disease progression is documented, a tumor grading increase may occur throughout the disease course in up to 25% of patients [49,50].

Combined [ 68 Ga]Ga-DOTA-SST Analogues and [ 18 F]FDG PET/CT in GEP-NENs
In the last few years, many authors have suggested to combining both [ 18 F]FDG PET/CT and SRS-PET/CT for the management of neuroendocrine tumors, particularly G2 and G3. This combined analysis can provide useful information on tumor heterogeneity, the characterization of SSTRs expression, and tumor grade, thus guiding clinicians to the proper treatment options.
Our aim was to clarify, according to data available in the literature, if combined imaging should be recommended in the routine clinical practice for the management of GEP-NENs. For this review, we included only studies regarding GEP-NENs in which combined imaging, with both [ 68 Ga]Ga-DOTA-SST analogues and [ 18 F]FDG-PET/CT, was performed. All other papers about neuroendocrine neoplasms that did not focus on GEP-NENs were excluded. Only original articles were included, while other types of publications (reviews, case reports, and others) were excluded, such as papers in languages other than English.
Searching on PubMed and the Scopus database for papers that analyze the role of combined imaging in GEP-NEN using the following terms [FDG and 68GA and (NEN or NET or GEP)], we found 134 papers; 29 were duplicates, 43 were reviews, book chapters, editorials, comments, case reports, or other non-eligible types, 22 were not about GEP-NEN, three were not in English, and eight did not compare [ 68 Ga]Ga-DOTA-SST analogues and [ 18 F]FDG PET/CT. We overviewed the remaining seven works, plus one that we added after searching from the references of these papers, for a total of eight original articles (Figure 1).  showed a FDG positivity whose percentage was higher in the G2 than in the G1 group. Moreover, the SUV max values at FDG were greater in G2 than in G1 tumors, and FDG positive lesions tended to be larger than negative ones and showed a higher probability to have locoregional lymph nodes involvement and distant metastases. With these data, authors suggested that a pre-operative evaluation with combined imaging, in cases without consensus for surgical therapy, can distinguish patients who can benefit from other treatment options.
Recently, Chan et al. [55] proposed a new grading scheme for metastatic NEN by using both SRS-PET/CT and [ 18  Authors found that the NETPET Score significantly correlates with tumor grade and overall survival, gives important prognostic information, and can identify patients that may benefit from PRRT. Based on these considerations, dual radiopharmaceutical PET imaging might be used as a prognostic biomarker. However, this needs to be validated in a larger prospective series of patients.
Zhang et al. [56] analyzed 83 patients with GEP-NEN divided in three groups based on the Ki67 index and mitotic count: Well-differentiated NENs group A (Ki67 < 10%), group B (Ki67 > 10%), and poorly differentiated NECs group C. Patients underwent both [ 68 Ga]Ga-DOTA-TATE and [ 18 F]FDG PET/CT. This retrospective study showed that while [ 68 Ga]Ga-DOTA-TATE was useful for staging and follow-up, [ 18 F]FDG showed a correlation with the aggressiveness of the tumor. [ 68 Ga]Ga-DOTA-TATE uptake was also related to a good prognosis. By contrast, tumors with a worse prognosis were those solely with [ 18 F]FDG uptake. Furthermore, they found that the sensitivity of the dual tracer (94%) was higher than that solely with [ 68 Ga]Ga-DOTATATE and [ 18 F]FDG. This finding was more evident in NET (Ki67 > 10%) than in NEC.
Taking all the reported data together, it appears clear that both techniques together play an important prognostic role in NENs, the clinical outcome being more favorable in patients with positive SRS-PET/CT and negative [ 18 F]FDG PET/CT (Tables 1 and 2).

Conclusions
The combined use of SRS-PET/CT and [ 18 F]FDG PET/CT still remains a challenge for physicians dealing with NENs. Assessing dual radiopharmaceutical PET/CT as a single parameter allows one to consider the two techniques as complementary rather than competitors, as previously suggested [57]. Using the dual tracer FITs, two different aspects of tumor biology may be explored: SSTR expression and glucose metabolism. SRS-PET/CT and [ 18 F]FDG PET/CT may be used for precise staging of patients with metastatic tumors, in which metastatic lesions may present heterogeneous metabolic activity and somatostatin receptor expression [42].
Though, there is solid scientific evidence confirming the clinical role of these procedures in patient management, the optimal selection of patients who would benefit from their combined use still remains debated. Large prospective studies with homogeneous series of NEN patients are required to resolve this unmet need.

Conclusions
The combined use of SRS-PET/CT and [ 18 F]FDG PET/CT still remains a challenge for physicians dealing with NENs. Assessing dual radiopharmaceutical PET/CT as a single parameter allows one to consider the two techniques as complementary rather than competitors, as previously suggested [57]. Using the dual tracer FITs, two different aspects of tumor biology may be explored: SSTR expression and glucose metabolism. SRS-PET/CT and [ 18 F]FDG PET/CT may be used for precise staging of patients with metastatic tumors, in which metastatic lesions may present heterogeneous metabolic activity and somatostatin receptor expression [42].
Though, there is solid scientific evidence confirming the clinical role of these procedures in patient management, the optimal selection of patients who would benefit from their combined use still remains debated. Large prospective studies with homogeneous series of NEN patients are required to resolve this unmet need.

Conclusions
The combined use of SRS-PET/CT and [ 18 F]FDG PET/CT still remains a challenge for physicians dealing with NENs. Assessing dual radiopharmaceutical PET/CT as a single parameter allows one to consider the two techniques as complementary rather than competitors, as previously suggested [57]. Using the dual tracer FITs, two different aspects of tumor biology may be explored: SSTR expression and glucose metabolism. SRS-PET/CT and [ 18 F]FDG PET/CT may be used for precise staging of patients with metastatic tumors, in which metastatic lesions may present heterogeneous metabolic activity and somatostatin receptor expression [42].
Though, there is solid scientific evidence confirming the clinical role of these procedures in patient management, the optimal selection of patients who would benefit from their combined use still remains debated. Large prospective studies with homogeneous series of NEN patients are required to resolve this unmet need.
Nevertheless, based on major findings from this systematic review, it emerges that the combined use of SRS-PET/CT and [ 18 F]FDG PET/CT should be considered in the following clinical scenarios: (i) At the time of initial diagnosis: In those patients with intermediate tumor proliferative activity (i.e., G2 tumors); if there is a heterogeneous SSTR expression among different tumor lesions; and in non-functioning tumors when patients have tumor-related symptoms (i.e., pain and weight loss). (ii) During follow-up: In addition to conventional radiological imaging at the time of first disease restaging after changing anti-proliferative medical treatment; at the time of disease progression after prolonged stable disease; and in case of a discrepancy between conventional radiological evaluation and clinical/biochemical assessment.