Objective. To investigate the biodistribution and kinetic constants of 68Ga-DOTATATE in normal organs through dynamic total-body positron emission tomography/computed tomography (PET/CT). Methods. Seven patients who experienced endoscopic resection of gastric neuroendocrine tumor were enrolled. Dynamic total-body PET/CT scans over 60 min were performed. Time-activity curves were obtained by drawing regions of interest in normal organs. Rate constants, including K1, k2, k3, and vB, were computed using a two-tissue compartment model. Factor analysis was used to compare the rate constants among subjects and regions. Hierarchical cluster analysis was performed to identify organs with similar kinetic characteristics. Results. The highest uptake of 68Ga-DOTATATE was observed in the spleen followed by kidneys, adrenals, liver, pituitary gland, pancreas head, prostate, pancreas body, and thyroid, parotid, and submandibular glands. Low background level of 68Ga-DOTATATE uptake was observed in the nasal mucosa, bone, blood pool, and cerebrum. In addition, the uptake in the pancreas head was noted to be higher than the pancreas body () on the basis of each time point of dynamic PET. There were differences of rate constants among different organs. The mean K1 ranged from 0.0507 min−1 in the left nasal mucosa to 1.21 min−1 in the left kidney, and mean k2 ranged from 0.0174 min−1 in the spleen to 4.4487 min−1 in the left cerebrum. The mean k3 ranged from 0.0563 min−1 in the right cerebrum to 4.6309 min−1 in the left adrenal, and mean vB ranged from 0.0001 in the left cerebrum to 0.2489 in the right adrenal. However, none of the rate constants was significantly different among subjects or among different sites within a single organ. Three groups of organs with similar kinetic characteristics were identified: (1) cerebrum; (2) pituitary gland, liver, adrenal, and prostate; and (3) nasal mucosa, parotid and submandibular glands, thyroid, spleen, pancreas, kidney, and bone. Conclusion. Uptake and clearance of 68Ga-DOTATATE, in terms of kinetic constants, were different in different organs. The kinetic parameters of 68Ga-DOTATATE in different organs provide a reference for future dynamic PET imaging.

1. Introduction

Neuroendocrine tumors (NET) are heterogeneous neoplasms arising from neuroendocrine cells in different organs [1]. High somatostatin receptor (SSTR) expression of NETs allows 68Ga-somatostatin analog (SSA) positron emission tomography/computed tomography (PET/CT) imaging for diagnosis and treatment. Standardized uptake value (SUV) using in the traditional static PET image is a widely used semiquantitative marker of SSTR expression. However, changes of tumor SUV in 68Ga-DOTATOC PET imaging have been found no correlation with the treatment outcome in peptide receptor radionuclide therapy [2, 3].

By contrast, net uptake rate (Ki) determined by tracer kinetic analysis of dynamic PET image appeared to be a more accurate measurement tool for quantification of SSTR density and therapeutic evaluation than those of SUV [4]. However, 68Ga-SSA dynamic PET/CT imaging studies have been limited mostly to tumors, such as meningiomas [5] and metastatic neuroendocrine tumors [4, 68]. As yet, few systematic studies about kinetic parameters of normal organs have been reported. This is mainly because previous dynamic protocols capture data within a very limited scan range, equivalent to the axial field-of-view of the traditional PET scanner (typically 15–30 cm). The recent introduction of a 194 cm long total-body PET/CT scanner allows simultaneous total-body dynamic imaging and high-quality tracer input function from major arteries, avoiding invasive arterial blood sampling [9].

A study of 2-deoxy-2-[18F]fluoro-D-glucose ([18F]FDG) kinetics in normal organs using total-body PET served as a reference for pharmacokinetic studies in other radiotracers [9]. Considering the ultra-high sensitivity of about 174 kcps/MBq [10], the author further demonstrated that dynamic [18F]FDG total-body PET scan with ultra-low activity obtained relevant kinetic metrics of full activity with the advantage of decrease in radiation dose and data size [11]. Although the administered dose of 68Ga-SSA is 2 MBq/kg of body weight [12], 1 MBq/kg of body weight was used in this study based on our previous total-body dynamic PET data. The aim of this study was to investigate the biodistribution and kinetic parameters of 68Ga-DOTATATE in normal organs through dynamic total-body PET/CT imaging with half-dose activity.

2. Materials and Methods

2.1. Patients

This prospective study was approved by the Institutional Review Board of Zhongshan Hospital, Fudan University (approval number IRB-B2020-186R), and all patients signed informed consent. The patients had only undergone endoscopic resection for gastric neuroendocrine tumor (gNET) earlier without other therapy and had been referred for 68Ga-DOTATATE PET/CT to search for evidence of residual or metastatic disease from August to December 2020. Seven patients were included in this study based on the following criteria: (1) no evidence of recurrent disease on conventional imaging before 68Ga-DOTATATE PET/CT, (2) all patients underwent 60 min dynamic PET imaging, and (3) no tracer-avid recurrence and metastases on PET scan. Detailed information of the seven patients is shown in Table 1.

2.2. Total-Body Dynamic PET/CT Imaging

68Ga-DOTATATE was prepared as described previously [13]. Total-body PET/CT (uEXPLORER, United Imaging Healthcare, Shanghai, China) with an axial field-of-view of 194 cm was used in this study. The patient first underwent a low-dose CT scan (120 kV; 10 mAs) over total body for attenuation correction. Then, 60 min dynamic PET scanning was performed simultaneously after a bolus injection of 48.1–73.6 MBq 68Ga-DOTATATE by hand into a vein near the ankle. At the end of the dynamic acquisition, a high-dose CT scan from head to proximal femoral was performed (120 kV; automatic tube current modulation) for diagnosis. The PET images were corrected for attenuation, scatter, alignment, decay, normalization, and randoms. Then, they were reconstructed using list-mode ordered subset expectation maximization (three iterations, 20 subsets) algorithm incorporating time-of-flight and point spread function modeling applying a 3.0 mm Gaussian filter. The image matrix size was 192 × 192 pixels. The dataset was divided into 55 frames: 36 × 5 s, 19 × 180 s. Representative frames of PET images are shown in Figure 1.

2.3. Time–Activity Curves (TACs)

Reconstructed PET and CT images were transferred to the vendor-provided workstation (uWS-MI R001; United Imaging Healthcare, Shanghai, China) for dynamic analysis. Two-dimensional regions of interest (ROIs) were drawn in organs according to the method described previously [9, 11]. ROIs were drawn at the same locations on the total-body images and on the last time frame of the dynamic scans, which were transferred to all earlier time frames. ROIs as large as possible were drawn within the limits of the activity distribution on PET, avoiding any obvious anatomical abnormalities or blood vessels on CT. ROI placement in each organ is shown in Figure 2.

The input function was obtained by drawing ROI in the ascending aorta as described by Liu et al. [9] because the ascending aorta is the closest major artery to the heart, and it is not affected by breathing and cardiac movements and the spillover of activity. Cerebrum ROIs were drawn in the temporal lobe on both sides to represent brain tissue as described previously [5]. For pituitary gland, ROI was placed in the middle of the gland. For the nasal mucosa, parotid and submandibular glands, thyroid, adrenal, and kidney cortex, ROIs were placed in the middle of each side. Liver ROIs were drawn in the upper, middle, and lower areas of the right lobe and middle area of the left lobe. Pancreas ROIs were drawn in the head and body part. Spleen ROI was placed in the area with the largest transaxial size. Prostate ROI was drawn in the middle of the gland. Bone ROI was placed in the third lumbar body. The average radioactivity concentration for each ROI was obtained at Bq/mL from each PET frames and TACs were generated for all organs.

2.4. Data Analysis

TAC data were uploaded to the PMOD software (version 3.201, PMOD Technologies Ltd., Zurich, Switzerland) [57]. The TAC data from the ROI in the descending aorta were used for image-derived input function (IDIF). A two-tissue compartment model was used for model fitting according to previous studies regarding dynamic 68Ga-DOTATOC PET imaging [57]. The default setting of PMOD regarding the lower and upper bounds of estimated kinetic parameter was applied, namely, 0–1 for vB and 0–8 for K1, k2, and k3. A constant weighting method with a factor of 0.05 was used. The rate constants K1, k2, and k3 and vB were computed. The compartmental configuration of this model is given in Figure 3. K1 is associated with the receptor binding, k2 with the displacement from the receptor, k3 with the cellular internalization, and k4 with the externalization. As k4 is commonly close to zero, it was not analyzed in this study [5, 14].

2.5. Statistical Analysis

Statistical analyses were conducted using SPSS software version 23.0 (IBM Corp., New York, NY, USA). If necessary, the summary statistic is expressed as mean ± standard deviation and coefficient of variation. Factor analysis was used to compare the rate constants of organs among subjects and regions. Variances were compared using mixed models. Hierarchical cluster analysis was performed to identify organs with similar kinetic characteristics. Significance was set at and all P-values reported were two-sided.

3. Results

3.1. Patient Characteristics

Seven patients (five males, two females; age 39–68 years) were imaged. All tumors were low- to intermediate-grade gNET, according to the World Health Organization 2019 grading of gastroenteropancreatic NETs. Table 1 summarizes the patient characteristics. The mean injected activity of 68Ga-DOTATATE for all subjects was 65.6 ± 8.8 MBq.

3.2. Biodistribution

Figures 1 and 4 illustrate the biodistribution of 68Ga-DOTATATE as a function of time in normal organs. In the first few seconds after injection into a leg vein, the tracer traveled to the heart and was then distributed through the arteries to all the organs of the body. Gradual accumulation of the tracer could be seen in the kidneys, spleen, adrenals, liver, and pituitary, thyroid, and salivary glands. Clearance of 68Ga-DOTATATE from the blood was fast. Radioactivity in the blood decreased rapidly to 2.1% of the peak level within 60 min of the dynamic scanning after reaching the peak at 40 s. The highest uptake was observed in the spleen, adrenals, and pituitary gland, due to SSTR-specific uptake. High 68Ga-DOTATATE uptake in the kidneys and liver was also observed. Radioactivity in the spleen, adrenals, pituitary gland, and liver increased with time and no excretion trend was seen until 60 min. The rapid presence in the kidneys, followed by a passage toward the urinary bladder, illustrates the mainly renal excretion of the tracer. Limited uptake was also observed in the prostate and in the thyroid. Low background uptake was observed in the cerebrum and bone.

In addition, the uptake in the pancreas head was noted to be higher than the rest of the gland in the static PET image. Radioactivity of the pancreas head was significantly higher than that of the pancreas body () on the basis of each time point of dynamic PET.

3.3. Kinetic Constants

The descriptive statistics of K1, k2, k3, and vB are shown in Tables 2 and 3. Representative fitted curves are shown in Figure 4. The rate constants varied widely between different organs. Mean K1 ranged from 0.0507 min−1 in the left nasal mucosa to 1.21 min−1 in the left kidney, and mean k2 ranged from 0.0174 min−1 in the spleen to 4.4487 min−1 in the left cerebrum. Mean k3 ranged from 0.0563 min−1 in the right cerebrum to 4.6309 min−1 in the left adrenal, and mean vB ranged from 0.0001 in the left cerebrum to 0.2489 in the right adrenal. There was no significant difference in the rate constants between different sites within an organ. The intersubject variances of K1 in various organs are acceptable. However, from k2 to k3, the intersubject variances showed an increasing trend. The largest intersubject variances in terms of CV for k2 and k3 were 264.6% in the left adrenal and 229.8% in the right middle area of liver, respectively. Furthermore, compared with the intersite variances of rate constants, the intersubject variances were larger and their contributions to the overall variances were remarkable, but was not statistically significant in any organ (). Mean K1, k2, and k3 values of the nasal mucosa were 0.0572, 0.1025, and 0.0771 min−1, respectively.

Figure 5 shows the distribution of organs in three-dimensional rate constant space and illustrates the maximum likelihood grouping. An optimum of three groups of organs was obtained, as shown in Table 4. The first group had the lowest K1 and k3, and the highest k2, representing low binding and internalization but high displacement of 68Ga-DOTATATE, and it consisted of the cerebrum. The second group consisted of the pituitary gland, liver, adrenal, and prostate, which had similar K1 values to those of organs from group 1, but with the lowest k2 and the highest k3, indicating low displacement and high internalization of 68Ga-DOTATATE. The third group had the highest K1 but a relatively lower k2 and k3, including the nasal mucosa, parotid and submandibular glands, thyroid, spleen, pancreas, kidney, and bone, indicating high binding but low internalization of 68Ga-DOTATATE.

4. Discussion

There are some important findings in this study. First, the biodistribution of 68Ga-DOTATATE in normal organs using dynamic total-body PET/CT imaging showed that the uptake and clearance in different normal organ types were different. Second, different kinetic characteristics were found in normal organs, which may reflect variable expression levels of SSTR. Besides, similar kinetic metrics were also found in some organs. The exact rationale for these findings is hard to explain, but there may be a reflection of the interorgan relationships in the kinetic metrics of 68Ga-DOTATATE in normal organs. As 68Ga-DOTATATE was considered stable without obvious metabolites in short time [4, 8], the kinetic metrics as identified in this study could be obtained with an IDIF noninvasively.

The 194 cm axial field-of-view of total-body PET/CT scanner has unique advantage over dynamic imaging. First, the scanner covers all organs in the body, and permits total-body dynamic imaging and kinetic analyses of the physiologies of all organs of interest simultaneously and systematically. Second, it has the ability to derive high-quality tracer input function from large arterial vessels in PET images with high accuracy for pharmacokinetic studies, avoiding invasive arterial blood sampling [9]. However, the challenge for single-bed dynamic PET imaging is that large arterial vessels are not always present in the field-of-view, and one may need to utilize other blood pools such as the carotid arteries, ascending aorta, descending aorta, or abdominal aorta [15, 16]. These approaches may involve more difficult ROI placement, and suffer from partial volume effects for smaller blood pools, such as the carotid arteries [15]. We chose the ascending aorta for obtaining the input function, as it had been determined to have the strongest correlation with the results from arterial sampling [16]. Third, the large increase in sensitivity arising from total-body coverage allows for very small amounts of activity for imaging the entire body [17, 18]. Recent study demonstrated that ultra-low activity dynamic [18F]FDG total-body PET scan not only obtained relevant kinetic metrics of full activity, but also reduced radiation dose and data size, making it more acceptable and easier for data reconstruction, transmission, and storage for clinical practice [11]. Therefore, half-dose 68Ga-DOTATATE was used in this study based on previous total-body dynamic PET data.

We found overall similar biodistribution and generally overlap of the range of 68Ga-DOTATATE uptake compared with previous data [1922]. 68Ga-SSA distribution is limited to those tissues that actually accumulate the tracer. The high uptake of 68Ga-DOTATATE in the spleen, adrenals, pituitary gland, and kidneys may be associated with the high expression of SSTR2, and has been proven in previous in vitro studies [23, 24]. Besides, 68Ga-DOTATATE is excreted in the renal tract, and, therefore, renal parenchymal uptake might be overestimated due to intense excretory tracer activity in the medullary pyramids and possible signal crossover from the renal pelvis. Although in vitro studies have shown that hepatocytes and hepatic stellate cells of the normal liver parenchyma do not express any of the SSTR subtypes [25], high 68Ga-DOTATATE uptake in the liver was observed in this study. However, as an organ related to metabolism and elimination of radiotracer, nonsomatostatin receptor-mediated uptake in liver [26] may have contributed to this phenomenon.

Previous studies regarding 68Ga-DOTATATE biodistribution focused mainly on static imaging [21, 22], but rarely on dynamic imaging. In this study, we systematically evaluated the simultaneous biodistribution of 68Ga-DOTATATE over the entire body through dynamic total-body PET/CT imaging, which is the first up till now. Blood activity measurements showed that the radiopeptide was rapidly eliminated. Dynamic PET studies demonstrated that radioactivity in the blood decreased rapidly to 2.1% of the peak level within 60 min of the dynamic scan. Besides, the rapid tracer accumulation in the kidneys, followed by a passage into the bladder, indicated a main clearance way of the tracer through kidney. The 68Ga-DOTATATE uptake of spleen and liver increased with time, while the uptake of kidney reached the peak at 6 min followed by gradual decrease and then slow increase, which were similar to that of 68Ga-DOTATATE and 68Ga-DOTATOC in previous studies [4, 20], reflecting the combination of specific receptor binding and nonspecific tissue handling of the peptide. Unlike the protocol of 45 min dynamic scans followed by three whole-body PET/CT scans at 1, 2, and 3 hr in the literature [4, 20], 60 min dynamic protocol without longer times was used in this study. Most 68Ga-DOTATATE TACs of normal organs in this study were ascending, while [18F]FDG TACs of normal organs were descending [9]. There were essential differences in the imaging principles between 68Ga-DOTATATE and [18F]FDG, which could also be reflected by TAC characteristics.

Physiological uptake in the pancreas head has been described in 68Ga-SSA PET/CT imaging previously [21, 27, 28], which was also observed in this study. As shown in Figure 4, the radioactivity concentration of pancreas head was significantly higher than that of pancreas body on the basis of each time point of dynamic PET. A possible explanation is that the pancreas head is rich in pancreatic polypeptide cells, which express SSTR subtypes 1–4 on their surface [29]. This phenomenon makes differentiation between physiological and pathological 68Ga-SSA uptake in pancreas head challenge [21, 27]. Moreover, quantification using SUV and pancreas/liver ratio had no additional value to aid diagnosis. Whole-body dynamic 68Ga-DOTATOC PET/CT protocol was used to differentiate physiological uptake of the pancreatic uncinate process from pathological uptake, showing the excellent diagnostic performance of the Ki (90.9%) [30]. Dynamic PET scan can acquire the kinetic parameters, and might aid in differential diagnosis.

Studies on the kinetic parameters of 68Ga-SSA mostly focus on tumors, but rarely on normal organs. Henze et al. [5] evaluated 68Ga-DOTATOC kinetics in patients with meningiomas, in which the nasal mucosa was selected as the reference tissue given that it had mild to moderate physiological SSTR density expression, with the mean vB, K1, k2, and k3 values of 0.110, 0.401, 0.556, and 0.061 min−1, respectively. In this study, the vB (0.0009), K1 (0.057 min−1), and k2 (0.103 min−1) of nasal mucosa were much smaller, while k3 (0.077 min−1) was close to the above result. Because of the size of nasal mucosa, partial volume correction might account for this point.

Dynamic total-body PET imaging provides conditions for real-time and systematic visualization and quantification of various organs in the body, especially for the heart–brain axis [31]. In this study, similar kinetic characteristics among different organs were found through the hierarchical clustering analysis, which may the internal connections among these organs. However, extended study, perhaps based on related disease models, is required in the future for further explanation of the interactions and relationships of 68Ga-DOTATATE among different organs.

This study bears limitations. Given the standard method has not been established for clinical practice, time delay correction was not performed in this study although its importance for total-body PET kinetic modeling has been identified recently [3234]. Second, selected patients were not healthy volunteers, but those with gNETs endoscopic resection. Therefore, discrepancy of the kinetic rates may exist between the results in this study and those from totally healthy volunteers. On the other hand, patients with gNETs endoscopic resection might have resulted in a representative referral population in current practice of patients undergoing 68Ga-DOTATATE PET imaging for suspected or for staging/restaging in the case of an already identified NET.

5. Conclusion

68Ga-DOTATATE uptake and clearance in different organs were different. The distribution and normal range of 68Ga-DOTATATE kinetic metrics in normal organs identified in this study could provide a reference for assessing tracer kinetics in future dynamic PET imaging.

Data Availability

The data that support the findings of this study are available from the corresponding author.

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the Ethical Standards of the Institutional and/or National Research Committee and with the Declaration of Helsinki 1964 and its later amendments or comparable ethical standards.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

Hongyan Yin and Guobing Liu contributed equally to this article. All authors have read and approved the final manuscript.


This study was funded by the Shanghai Municipal Key Clinical Specialty Project (grant number SHSLCZDZK03401), the Major Science and Technology Projects for Major New Drug Creation (grant number 2019ZX09302001), the Shanghai Science and Technology Committee Program (grant number 20DZ2201800), the Three-Year Action Plan of Clinical Skills and Innovation of Shanghai Hospital Development Center (grant number SHDC2020CR3079B), and the Next Generation Information Infrastructure Construction Project funded by Shanghai Municipal Commission of Economy and Informatization (grant number 201901014).