Soon after its introduction into clinical practice, Octreoscan imaging was reported to be very useful for patients with NENs as it explores the whole body and provides important clinical indications, allowing the accurate detection and localization of the primary tumor based on its expression of SSTRs. Moreover, this imaging technique provides valuable information for staging disease extent, for identifying metastases to soft tissue or bone, and for restaging NEN patients during follow-up. In this regard, it should be emphasized that SRS (with either single-photon or positron emitting radiopharmaceuticals) cannot be considered as a true “diagnostic” procedure, i.e., for differentiating NENs from other tumors; in fact, several other types of tumors can also express SSTRs with high density, besides the possibility of SSTR expression also in benign lesions (e.g., chronic granulomatous diseases). The diagnosis of NENs is mostly based on clinical/biochemical findings, while imaging plays an important role for localizing the disease site(s), which is not an easy task even with high-resolution imaging techniques such as CT and/or MRI for tumors that can be very small and distributed virtually throughout the entire body. Thus, the main role of SRS in the initial approach to patients with a newly diagnosed NEN is its ability to localize the exact site(s) of the tumor(s), also as a possible guide for surgical resection.

Nuclear medicine imaging of NENs is crucial for planning therapy also in patients with non-resectable lesions; in fact, the tumor expression and density of SSTRs can guide therapeutic decision-making in NEN patients, by identifying those who will benefit from therapy with “cold” vs. radiolabeled somatostatin analogues as well as those who are candidates for chemo-radiotherapy.

SRS with Octreoscan is highly sensitive for the detection and staging of PanNENs. In general, gastrinomas abundantly express SSTRs; in particular, about 50% of gastrinomas express sstr₂ and sstr₅, 33% express sstr₁, 17% express sstr₃, and 83% express sstr₄ [10]. Whole-body SRS has a pivotal role in these NENs [11, 12], identifying with 60–90% sensitivity both the primary lesion and metastases [13, 14]. However, few studies have been published supporting the usefulness of SRS in patients with somatostatinomas. Although some authors [15] reported that these tumors can be SSTR-negative in binding assays and in receptor autoradiography experiments (mainly because the high levels of circulating somatostatin induce receptor down-regulation), in vivo visualization of somatostatinomas with SRS has been reported to be effective both within the pancreas and at distant metastatic sites [16].

Among the SSTRr receptors, sstr₂ is expressed in 100% of the patients with glucagonomas, while sstr₁, sstr₃, sstr₄, and sstr₅ are expressed in approximately two-thirds [10]. In these patients, SRS has a sensitivity of approximately 70% for primary tumor detection and distant staging. Since most pancreatic VIPomas express all five SSTR subtypes [10], SRS in these patients (who usually present with a large tumor in the tail of the pancreas) is useful for lesion localization and characterization, and for the detection of distant metastases. The overall sensitivity of SRS is about 86%, being much lower, as expected, for lesions < 1 cm in size [17].

Insulinomas often present as a small lesion characterized by low sstr₂ expression [18]. About 50% of these tumors express sstr₁, while 15–20% express sstr₃ and sstr₄ [19]. Therefore the sensitivity of SRS for insulinomas can be as low as 50–60% [13, 20]. Although non-malignant insulinomas rarely express SSTRs, in malignant insulinomas functional evaluation with SRS is highly recommended for tumor detection and staging as well as for determining the indications for therapy based on somatostatin analogs [21]. Secondary insulinomas often have higher sstr₂ expression than the primary tumor [13, 22, 23].

In the initial approach to patients with newly diagnosed PanNENs, the integration of SRS data with those provided by anatomic imaging is crucial. Due to the recent introduction into clinical routine of SPECT/CT hybrid equipment, radionuclide imaging can now provide the surgeon with detailed functional and topographic information for tissue sampling and resectability; in addition, it can exclude false-positive findings due to accessory spleen, recent surgical scars, and any other cause of granulomatous-lymphoid infiltrate that may mimic a tumor. Moreover, correlative imaging can aid the nuclear physician in correctly identifying small lesions that might have reduced SSTR expression because of recent treatments, as well as de-differentiated disease, both of which may lead to false-negative results at SRS (Fig. 11.3).

The diagnostic impact of SRS is mainly during initial staging, as the technique has been shown to modify the therapeutic strategy in up to 53% of patients with NENs. In fact, because of its high sensitivity in detecting distant metastases (61–96%), SRS may prevent surgery with curative intent in those patients whose tumors have already metastasized [24, 25]. Moreover, SRS is the most accurate imaging modality for the “one-shot” detection of liver and extrahepatic metastases in patients with pancreatic NENs (Fig. 11.4 ), although sensitivity can be adversely affected by the small size typical of metastatic lesions.

SRS also may be useful in the follow-up of patients, in particular to monitor the efficacy of treatment. In this regard, changes in functional volume have been reported to be more reliable than RECIST (Response Evaluation Criteria in Solid Tumors) and correlate well with the long-term clinical response [26]. Moreover, SRS can aid in the identification of potentially responsive patients for treatment with unlabeled somatostatin analogues (such as octreotide acetate), or with tumor-targeted somatostatin analogues radiolabeled with either ⁹⁰Y or ¹⁷⁷Lu. In the former approach, it is still debated whether the sensitivity of SRS is reduced in patients concurrently receiving therapeutic doses of octreotide acetate; thus, consideration should be given to temporarily suspending therapy before administering the radiopharmaceutical.

As alluded to above, the introduction of receptor PET radiopharmaceuticals has prompted a new era in receptor imaging for NENs. Based on its unique pharmacokinetics, ⁶⁸Ga-DOTA-TOC achieves high tumor/non-tumor ratios within a short period after its administration and is even better than DOTATOC containing other metal ions. In particular, 80% of the activity is accumulated in the tumors within 30 min, renal clearance is fast, and the radioactivity concentration in tissues not expressing SSTRs is very low. The combination of high contrast and fast imaging, as well as the better spatial resolution of PET compared with single-photon imaging allows the detection of smaller lesions not identified at SRS; this feature becomes of paramount importance in tailoring the clinical management of NENs on a patient-to-patient basis [27]. Moreover, the ability of PET to accurately quantify uptake in each lesion (in terms of SUV) allows the use of such data for monitoring the response to therapy. Finally, PET imaging with ⁶⁸Ga-DOTA-somatostatin analogues involves less radiation exposure for the patient than is the case for imaging with ¹¹¹Inpentetreotide.

Current data show that, despite some differences in receptor binding affinity among the three different ⁶⁸Ga-DOTA-somatostatin analogues (-TOC, -NOC, -TATE), there is no direct clinical correlate regarding advantages in the clinical accuracy of one radiopharmaceutical over the others. All of them have in fact been reported to provide clinically accurate and valuable information for NEN imaging, as consistently demonstrated by high tumor/non-tumor ratios, and definitely higher sensitivity than SRS [28]. Moreover, ⁶⁸Ga-DOTA-TOC and ⁶⁸Ga-DOTA-TATE can mimic as closely as possible the in vivo pharmacokinetic patterns of their ⁹⁰Y- or ¹⁷⁷Lu-labeled counterparts used for peptide receptor radionuclide therapy ( PRRT) [29].

⁶⁸Ga-DOTA-peptide PET performs better than CT and/or single-photon SRS for locating well-differentiated NEN lesions, accurately detecting even small-sized tumors, in particular metastatic sites in lymph nodes or bone [30] or tumors with unusual anatomical locations [31]. However, it should be underlined that the detection of a greater number of lesions does not always impact disease stage or the therapeutic approach.

In 84 patients with NEN, ⁶⁸Ga-DOTA-TOC PET/CT was reported to have 97% sensitivity for the detection of SSTR-positive lesions, superior to both CT (61 %) and single -photon SRS ( 52%) [32]. In another series of 51 patients with well-differentiated NEN, ⁶⁸Ga-DOTA-TOC showed 97% sensitivity and 92% specificity in the early detection of bone metastases, much higher than CT and SRS [33].

Staging, clinical management, and the therapeutic approach can be changed when unsuspected metastatic disease or local relapse is identified, or SSTR expression on NEN cells by ⁶⁸Ga-DOTA-peptides is confirmed/excluded, unlike conventional imaging. In 50 out of 90 NEN patients, ⁶⁸Ga-DOTA-NOC PET/CT was reported to impact the therapeutic approach with PRRT or with unlabeled somatostatin analogue treatment in 36 patients, surgical treatment in six patients; surgical treatment was excluded in another six patients, radiotherapy in one patient, liver transplantation in one patient, and further diagnostic assessment in one patient [34]. In 51 patients with NENs, 35 with negative and 16 with equivocal ¹¹¹In-DTPA-octreotide uptake on SRS, ⁶⁸Ga- DOTA-TATE PET detected significantly more lesions than SRS and modified management in 70.6% of the patients, subsequently considered candidates for PRRT [35].

Non-invasive quantification of the receptor expression pattern is especially useful for selecting those patients eligible for targeted therapy with either radiolabeled or cold somatostatin analogs as the most appropriate therapeutic approach. High uptake of ⁶⁸Ga-DOTA-peptides reflects a high SSTR expression on well-differentiated NENs, associated with slower growth rate and a higher likelihood of response to targeted therapy with either hot or cold somatostatin analogs. The SUV_max derived from ⁶⁸Ga-DOTA-NOC PET/CT has recently been reported to be a helpful prognostic factor of outcome in this regard. Out of 44 patients with NENs, the SUV_max was significantly higher in those with stable disease/partial response and provided valuable information to distinguish those patients showing progressive disease at follow-up. In particular, an SUV_max > 19.3 allowed the selection of patients with slower disease progression [35]. Although there is no agreement on the use of ⁶⁸Ga-DOTApeptides for assessing response to PRRT [31], they are widely employed to quantify the presence of SSTRs, radiopharmaceutical biodistribution prior to treatment [36], and to predict and assess the response to peptide receptor therapies in NENs.

Compared with MIBG scintigraphy, radiolabeled somatostatin analogues have been shown to be more accurate for detecting gastro-entero-pancreatic NENs and their metastases. Similar results have been observed with [¹⁸F]DOPA PET. In patients with well-differentiated NEN, ⁶⁸Ga-DOTA-TATE showed marginally higher sensitivity than ¹⁸F-DOPA (96% vs. 56%) [37].

Preoperative localization of the lesions in patients with congenital hyperinsulinism is currently the only application of [¹⁸F]DOPA PET in pancreatic malignancy. Hyperfunctioning pancreatic lesions have greater L-DOPA uptake and conversion to dopamine than normally functioning pancreatic tissue, which expresses only low levels of aromatic amino acid decarboxylase. The high sensitivity of this imaging approach could allow mini-invasive laparoscopic surgery with limited resections, reducing therefore the risk of long-term diabetes [38]. On the other hand, euglycemic and hyperinsulinemic adult patients show [¹⁸F]DOPA uptake in the whole pancreas, the main limitation for identifying insulinomas or β-cell hyperplasia.

Several papers have demonstrated that NENs with a positive [¹⁸F]FDG PET have increased aggressiveness, irrespective of either grading or the Ki67 proliferation index [39]. In fact, high glucose metabolism of tumors depending on increased GLUT expression is directly related to cell proliferation and failed differentiation [40]. The dedifferentiation of neoplastic NEN cells typically tends to increase their uptake of [¹⁸F]FDG. Thus, a high glucose metabolic state can provide crucial prognostic information (Fig. 11.5).

⁶⁸Ga-DOTA-TOC has been reported to detect a higher number of tumors in well-differentiated NENs than is the case with [¹⁸F]FDG [41]. Compared with [¹⁸F]FDG PET/CT, ⁶⁸Ga-DOTA-TATE PET/CT has higher uptake by well-differentiated, low-grade NENs. [¹⁸F]FDG does not provide valuable information in the assessment of well-differentiated, relatively slow growing NENs, characterized by a low metabolic rate and therefore low glucose consumption. Instead, it has an important role in higher grade, poorly differentiated NENs, which have a high fraction of proliferating cells (positive Ki-67 staining in > 5% of cells); furthermore, for lesions with low SSTR expression [¹⁸F]FDG PET better demonstrates progression on morphological imaging over a period of < 6 months [42] than either CT or MRI.