Introduction
A neuropeptide first described in 1973,1 somatostatin generally serves to inhibit the release of other hormones such as growth hormone, insulin, and prolactin in addition to its role as a neurotransmitter.2 Somatostatin exerts its effects by binding to somatostatin receptors (SSTR), which are expressed on the surface of neuroendocrine cells. The five isoforms of SSTR, identified as SSTR1-5, have different patterns of expression depending on anatomic location. For example, SSTR2 is highly expressed in the spleen,3 SSTR1 and SSTR2 are expressed in the thyroid gland,4 while varying levels of SSTR1-5 are expressed by the pancreas.5
Because most neuroendocrine tumors (NETs) contain high levels of SSTR2, scintigraphy with the somatostatin analogue 111In-DTPA-octreotide, which binds to SSTR2 with high affinity as well as to SSTR3 and SSTR5 to a lesser degree,6 has been successfully employed for imaging of NETs since the first clinical trials in 1992.7 Recently, positron-emitting SSTR radiotracers labeled with 68Ga or 64Cu and imaged with positron emission tomography/computed tomography (PET/CT) were able to achieve superior image quality with lower patient radiation exposure given their higher affinity to SSTR2 and shorter radionuclide half-lives compared to 111In-DTPA-octreotide.8 Of the PET tracers commonly used, DOTATATE binds to SSTR2 with high specificity; DOTATOC, SSTR2 and SSTR5; DOTANOC, SSTR2, SSTR3, and SSTR5.9 When labeled with the beta-emitting 177Lu, DOTATATE can act as a theranostic agent for the treatment of well-differentiated NETs by exposing tumor cells to reactive oxygen species, resulting in oxidative damage and cell death.10
Given the variable expression of SSTR in normal tissues, physiologic uptake can occasionally mimic pathology. Alternatively, SSTR activity associated with benign disorders could lead to new indications for SSTR PET imaging in patients without known or suspected NET.
In this article, we review the sources of potential false positive findings and discuss new potential non-neoplastic indications for PET/CT imaging with radiolabeled DOTATATE, DOTATOC, and DOTANOC.
Methods
A literature search of publications between January 1, 1980, and August 31, 2023, was performed using PubMed. The SSTR PET tracers in clinical use “DOTATATE,” “DOTATOC,” and “DOTANOC” were used separately as keywords, and results pertaining to neoplastic etiologies, including NET, pheochromocytoma, paraganglioma, and meningioma, were excluded. References contained within the selected articles also received consideration. No restrictions were applied regarding sample size, study design, or outcome measures. Editorials, commentaries, and letters were excluded, as well as pre-clinical studies and those which fell outside of the scope defined by physiologic and benign causes of increased SSTR activity.
Finally, a comprehensive review was created, focusing on false-positive considerations during image interpretation and the potential for imaging of inflammation with SSTR PET.
Results
Physiologic Uptake and Pitfalls
The highest physiologic uptake of 68Ga-DOTATATE is generally in the spleen (Figure 1), where increased SSTR2 expression has been found specifically in the red pulp.11 Uptake levels in the adrenal glands, pituitary gland, and kidneys demonstrate high intensity, with activity in the kidneys and urinary tract being nonspecific for SSTR expression given the presence of renal excretion.
68Ga-DOTATATE localization to the salivary glands, thyroid, and liver is moderate, with nonspecificity of hepatic activity similarly due to excretion.12 Similar patterns of intense splenic activity followed by prominent renal and hepatic activity are present in PET/CT images obtained with 68Ga-DOTATOC and 68Ga-DOTANOC.13,14 Although labeling SSTR tracers with either 68Ga or 64Cu does not affect their biodistribution, the improvement in spatial resolution resulting from the lower positron range of 64Cu suggests that 64Cu-labeled agents may achieve wider clinical use in the future.15,16
In a study of 120 patients by Boy et al, the authors observed significantly higher 68Ga-DOTATOC uptake in the uncinate process of the pancreas compared to that in the pancreatic head, body, and tail.4
This common finding among all SSTR PET agents has been described as a classic pitfall,17 which is often more diffuse and less intense compared to the well-defined focal uptake associated with NETs in this region (Figure 2).18 Lakhotia et al considered 775 patients imaged with 111In-DTPA-octreotide, 68Ga-DOTATATE, 68Ga-DOTATOC, and 68Ga-DOTANOC from eight studies and found increased physiologic uptake in the pancreatic head and uncinate process in 229 patients (29.5%).18-26 Similarly, Tabacchi et al found increased 68Ga-DOTANOC uptake at the uncinate process in 77 out of 172 patients considered (44.8%).17 The uncinate process in particular has been found to contain an increased numbers of pancreatic polypeptide cells, which highly express SSTR.27 Brabander et al observed that increased 111In-DTPA-octreotide localization to the uncinate process was more common in patients with diabetes mellitus, which is associated with increased serum pancreatic polypeptide, further supporting pancreatic polypeptide cell hyperplasia as a possible mechanism to explain uncinate process activity.20 Although Kroiss et al report successfully using standardized uptake values (SUVs) to differentiate physiologic and pathologic uptake,28 Krascz found that even in patients with suspiciously high SUVs in the uncinate process, correlative imaging with CT and MRI often reveals no underlying lesion to suggest presence of tumor.22 Thus, cautious image interpretation taking into account this potential pitfall and consideration of additional evaluation with contrast-enhanced CT or MRI are necessary for accurate interpretation of pancreatic uptake in SSTR PET.
Although SSTR radiotracers are not known to localize to the pancreatic tail to the same degree as the uncinate process, variability in this region could similarly lead to diagnostic uncertainty. Focal islet cell hyperplasia has been reported as a possible mechanism causing increased SSTR expression in the pancreatic tail in patients with high physiologic activity (Figure 3).29 A study in 35 patients by Delbeke et al found that physiologic uptake within the pancreatic tail is typically equal or less than that of the liver.29 Therefore, quantification with SUVs, as well as correlation with CT or MRI, could play a role in discriminating normal and malignant activity in the pancreatic tail.
Related to the intense physiologic uptake demonstrated by the spleen, accessory spleen SSTR activity is similarly high, which can acquire an appearance suspicious for NET involvement, especially if intrapancreatic (Figure 4).30-32 Splenosis, often occurring after trauma or splenectomy, represents another atypical splenic tissue distribution that can resemble malignant activity.33 By assessing whether 99mTc-labeled sulfur colloid or heat-denatured red blood cell scintigraphy is able to confirm the presence of splenic tissue rather than tumor, misdiagnosis and subsequent high-risk intervention are potentially avoidable.34
Imaging of Inflammation
White blood cells such as macrophages are known to express SSTR2, resulting in low grade uptake associated with inflammation. As a result, a variety of inflammatory conditions imaged by SSTR PET/CT have been reported in the literature, including endometriosis, large vessel vasculitis, idiopathic pulmonary fibrosis, and pulmonary tuberculosis.35-39
SSTR PET has shown preliminary success in assessing systemic sarcoidosis. 68Ga-DOTATOC PET/CT was compared to 67Ga scintigraphy in a study of 20 sarcoidosis patients by Nobashi et al, who noted 68Ga-DOTATOC-positive lesions in 19 patients and 67Ga-positive lesions in 17 patients.40 These results not only indicate that SSTR PET is a feasible method of assessing sarcoidosis-related inflammation, but also that its performance may be similar or better than that of 67Ga, which is an established marker of infection and inflammation.
A study which corroborates this finding used 68Ga-DOTANOC PET/CT in 39 patients and observed a sensitivity of 93% and a specificity of 83% in the diagnosis of sarcoidosis, associated with decreased activity after treatment with symptomatic improvement.41 The described correlation between imaging and clinical improvement thus provides convergent validity for this approach.
A case report showing focal intracranial 68Ga-DOTATATE in the cavernous sinus of a symptomatic patient suggests a role for SSTR PET/CT in neurosarcoidosis, which was confirmed via biopsy.42 In addition to new potential clinical indications for imaging, these early observations also indicate that presence of sarcoidosis could complicate image interpretation in patients undergoing assessment for SSTR-positive neoplasm.
In light of the growing role of 18F-fluorodeoxyglucose (FDG) PET/CT in cardiac sarcoidosis, experiences using SSTR PET/CT for the same indication have started to appear.43-48 A 2016 study by Gormsen et al compared PET/CT with 18F-FDG and 68Ga-DOTANOC in a prospective analysis of 19 patients with suspected cardiac sarcoidosis.49 The authors found that while 18F-FDG findings were inconclusive in 11 patients, there were no inconclusive 68Ga-DOTANOC studies. The overall diagnostic accuracy of 18F-FDG for cardiac sarcoidosis was reported as 79%, compared to an overall accuracy of 100% for 68Ga-DOTANOC in this small cohort.
A more recent study published in 2021 by Bravo et al showed that in 13 subjects with suspected cardiac sarcoidosis, all of whom had positive 18F-FDG findings, only 7 subjects showed definite or probable abnormal cardiac uptake of 68Ga-DOTATATE.50 However, there was 100% concordance between 18F-FDG and 68Ga-DOTATATE in positive thoracic nodal involvement, suggesting that the role of SSTR PET in cardiac sarcoidosis is less clear compared to the stronger evidence supporting its use in systemic manifestations.
Atherosclerotic disease has also been shown to correlate with SSTR tracer activity due to the presence of macrophages, potentially allowing for the early identification of vulnerable plaques in patients with risk factors.51,52 By focusing on symptomatic carotid artery plaques in 10 patients planning to undergo carotid endarterectomy, Pedersen et al found that uptake of 64Cu-DOTATATE was higher in symptomatic plaques compared to the contralateral side.53 After analyzing gene expression in the plaque specimens, the investigators found that tracer activity was correlated with the presence of alternatively activated macrophages.
The advantage of using 68Ga-DOTATATE rather than 18F-FDG for atherosclerosis imaging was illustrated in a study that showed feasible coronary artery disease assessment in all 42 patients assessed with 68Ga-DOTATATE, which was not possible with 18F-FDG in most cases due to high adjacent myocardial activity.54 SSTR PET has also shown decreased inflammation in atheromatous plaques in response to medical intervention. Specifically, 22 subjects with type 2 diabetes were imaged with 68Ga-DOTATATE PET/CT before and 3 months after initiating atorvastatin therapy, with a resulting 31% decrease in target-to-background ratio.55 Similarly, 64Cu-DOTATATE coronary uptake was found to decrease after 26 weeks of liraglutide therapy in 30 patients with type 2 diabetes.56 Therefore, SSTR PET could help guide clinical decision making by identifying severity of disease and assessing efficacy of treatment.
Discussion
SSTR PET is currently being performed clinically for the primary purposes of staging and following NETs. Although the uptake mechanism of SSTR tracers is more specific than that of 18F-FDG, many nontumor sources of uptake are still present. Familiarity with normal tissues which naturally express SSTR and common variations such as elevated uptake in the pancreatic uncinate process is necessary to avoid interpretive error. Normal splenic anatomy would be unlikely to confuse a reader, but the presence of an accessory spleen in an ambiguous or deceptive location could result in a diagnostic dilemma, warranting identification of splenic tissue with scintigraphic techniques or correlation with additional anatomic imaging.
Splenosis would be of particular concern in patients having undergone resection of NET, which is sometimes accompanied by splenectomy in cases with pancreatic involvement. Subtle morphological signs of splenosis such as the presence of smooth, round nodules combined with knowledge of the patient’s treatment course may alert the reader to the possibility of splenosis rather than more concerning pathology such as peritoneal metastasis, additional testing would often be necessary given the lack of specificity of SSTR PET in such cases.
Similar to physiologic phenomena, benign and inflammatory disorders can hinder accurate interpretation if alternative explanations are not considered. For example, a focal 68Ga-DOTATATE-avid lesion in the cavernous sinus may appear classic for meningioma, but in rare instances this may instead represent neurosarcoidosis. SSTR-positive mediastinal or hilar lymph nodes may represent either nodal metastases or reactive, infectious, or inflammatory lymph nodes depending on exposures and co-morbidities.
Although the distinctive vascular pattern and low-level uptake associated with atherosclerosis would be unlikely to pose a diagnostic challenge, the strong association between SSTR activity and presence of activated macrophages in atherosclerotic plaques suggests a possible role in directing therapy.
Conclusion
SSTR PET imaging has quickly revolutionized diagnosis, treatment, and surveillance of NETs and other SSTR-expressing tumors. A nuanced understanding of tracer behavior is necessary for precise image interpretation and optimal utilization.
