Visual Abstract
Abstract
The current study aimed to compare 68Ga-NODAGA-Cpa-cyclo(d-Cys-amino-Phe-hydroorotic acid-d-4-amino-Phe(carbamoyl)-Lys-Thr-Cys)-d-Tyr-NH2 (JR11) and 68Ga-DOTATATE PET/CT in patients with metastatic, well-differentiated neuroendocrine tumors. Methods: A prospective bicenter study aimed at enrolling 100 patients with histologically proven, metastatic or unresectable, well-differentiated neuroendocrine tumors was conducted. The first 48 patients represented the study cohort. Each patient received 68Ga-DOTATATE on the first day and 68Ga-NODAGA-JR11 on the second day. Whole-body PET/CT scans were performed at 40–60 min after injection. Normal-organ uptake, lesion numbers, lesion uptake, and sensitivity were compared. The potential impact on clinical management was also determined. Results: Overall, 68Ga-NODAGA-JR11 demonstrated lower background uptake in normal organs. Compared with 68Ga-DOTATATE, 68Ga-NODAGA-JR11 detected significantly more liver lesions (673 vs. 584, P = 0.002). The target-to-background ratio of liver lesions was significantly higher on 68Ga-NODAGA-JR11 (6.4 ± 8.7 vs. 3.1 ±2.6, P = 0.000). Comparable uptake was observed for primary tumors, bone lesions, and lymph node metastases. In total, 180 lesions were detected on conventional imaging in 15 patients; 165 and 139 lesions of them were positive on 68Ga-NODAGA-JR11 and 68Ga-DOTATATE, leading to a sensitivity of 91.7% and 77.2%, respectively. In 14.5% (7/48) of patients, 68Ga-NODAGA-JR11 PET might have a potential impact on clinical management. Conclusion: 68Ga-NODAGA-JR11 shows better sensitivity and a higher target-to-background ratio than 68Ga-DOTATATE. The detection of more lesions by the antagonist may have a potential impact on clinical management in a subgroup of patients.
Neuroendocrine tumors (NETs) are a heterogeneous group of neoplasms arising from endocrine cells. Most NETs overexpress somatostatin receptor (SSTR), a G-protein–coupled membrane receptor that can be targeted for molecular imaging and radionuclide therapy (1). SSTR imaging plays a critical role in the management of NETs, including staging, restaging, prognosis, therapy decision-making, and response monitoring (2). However, the detection ability of 68Ga-labeled SSTR agonists such as 68Ga-DOTATATE is still limited. There remains a clinical need to improve SSTR imaging.
Recently, an important improvement in the field of SSTR imaging was the introduction of SSTR antagonists (3–8). Bass et al. described the first radiolabeled SSTR antagonists in 1996 (9). Cpa-cyclo(d-Cys-amino-Phe-hydroorotic acid-d-4-amino-Phe(carbamoyl)-Lys-Thr-Cys)-d-Tyr-NH2 (JR11) is a novel SSTR2 antagonist with promising results in preclinical studies, but the clinical evidence is scarce (10). We reported a case of 68Ga-DOTATATE–negative gastric NET (grade 2). The gastric lesion was successfully imaged with 68Ga-NODAGA-JR11 and subsequently treated with endoscopic submucosal dissection (11). Nicolas et al. found that 68Ga-NODAGA-JR11 was superior to 68Ga-DOTATOC in sensitivity, lesion detection, and image contrast in patients with low- or intermediate-grade gastroenteropancreatic NETs (12).
The current article describes an interim analysis of a prospective bicenter study aimed at comparing 68Ga-NODAGA-JR11 and 68Ga-DOTATATE PET/CT in patients with metastatic, well-differentiated NETs.
MATERIALS AND METHODS
Study Design and Patient Population
This was a prospective bicenter study conducted at the First Affiliated Hospital of Fujian Medical University and Peking Union Medical College Hospital and aimed to enroll 100 patients. The study was registered at ClinicalTrials.gov (NCT04897542) and approved by the institutional review board of the First Affiliated Hospital of Fujian Medical University and Peking Union Medical College Hospital. All subjects gave written informed consent before study participation. Patients with histologically proven, metastatic or unresectable, well-differentiated NETs (grade 1 or 2) were prospectively and consecutively recruited to this study. If the patient was on a long-acting somatostatin analog, a washout phase of 28 d was required before study participation. No NET-specific treatment was allowed between the 2 scans. This interim analysis comprises 48 patients enrolled between August 2020 and November 2021, 24 in each center.
68Ga-DOTATATE and 68Ga-NODAGA-JR11 Preparation
Good-manufacturing-grade NODAGA-JR11 and DOTATATE precursors were supplied by CS Bio Co. and ABX GmbH, respectively. The radiolabeling was performed manually in a hot cell. Briefly, 68GaCl3 was eluted from a 68Ge/68Ga generator (Eckert & Ziegler GalliaPharm) using 5 mL of 0.1 M hydrochloride acid. The eluent (about 250 MBq) was added to a reaction vial containing the precursor (for NODAGA-JR11, 40 μg; for DOTATATE, 40 μg) and dissolved in sodium acetate buffer, for a final pH of 4 for the reaction mixture. The mixture was heated to 100°C for 10 min. After cooling at room temperature, the reaction mixture was diluted with 5 mL of saline and then loaded onto an Oasis HLB cartridge (Waters) (preconditioned with 1 mL of ethanol and 5 mL of saline) and washed with saline to remove unincorporated radionuclide. Finally, the product was eluted off the cartridge with 75% ethanol solution, diluted with saline, and passed through a Millipore filer (0.22 μm, 25 mm) into a sterile product vial. The radiochemical purity of the final product was more than 99%, and the ethanol amount was not more than 10% for injection. The non–decay-corrected radiochemical yield of 68Ga-NODAGA-JR11 and 68Ga-DOTATATE was about 65% for a 30-min radiolabeling process. The radioactivity of the final injection would be about 150 MBq after 40 min of 68Ga elution (30 min for labeling + 10 min for patient preparation). The injected peptide dose was 40 μg per patient, and the specific activity was about 3.75 MBq/μg.
68Ga-DOTATATE and 68Ga-NODAGA-JR11 PET/CT Imaging
The study was performed on a time-of-flight PET/CT scanner (Biograph mCT64; Siemens) at both centers. The patients received an intravenous injection of 68Ga-DOTATATE (150 ± 55 MBq) on the first day and 68Ga-NODAGA-JR11 (148 ± 52 MBq) on the second day. The CT scans were performed with tube voltage of 120 kV, effective tube current of 70–120 mA (CareDose 4D; Siemens), and a slice thickness of 3 mm. PET data were reconstructed using ordered-subsets expectation maximization (2 iterations and 21 subsets).
Adverse Event Monitoring
Vital signs (blood pressure and heart rate) and clinical symptoms were monitored and recorded for up to 2 h after injection. Adverse events were recorded according to version 4.03 of the Common Terminology Criteria for Adverse Events.
Image Interpretation and Data Analysis
68Ga-DOTATATE and 68Ga-NODAGA-JR11 PET/CT images were analyzed on a Syngo MultiModality Workplace (Siemens). The images were reviewed by a board-certified nuclear medicine physician. The reader was masked to patients’ medical history and radiopharmaceutical. When the reader reviewed the images, the lesion number, location, and uptake were recorded in the report form. After all measurements were finished, the reader was permitted to reopen the 2 scans again and compare the data. However, any revision to the original data was not allowed in this phase; only comparison and reconciliation were allowed.
For normal tissues, the physiologic uptake of 68Ga-DOTATATE and 68Ga-NODAGA-JR11 in the following organs was recorded: pituitary gland, lungs, spleen, renal cortex, adrenal glands, normal liver parenchyma, stomach, small intestine, and pancreas (uncinate process). Regions of interest were drawn over the organs, excluding focal lesions. Any activity from adjacent organs such as renal pelvis and urinary bladder was avoided. The SUVmax (using body weight normalization) of the regions of interest in normal organs was recorded. In the case of bilateral organs such as the adrenal glands and renal cortex, the average SUVmax was calculated.
Any focal accumulation of 68Ga-DOTATATE and 68Ga-NODAGA-JR11 not explained by physiologic uptake or benign lesions, such as bone trauma, hemangioma, and degeneration disease, was defined as a focal lesion. CT correlation was used to help characterize the lesions. Regions of interest were drawn around the lesions on transverse slices for semiquantitative analysis. The number and SUVmax of focal lesions were recorded. For liver lesions, relative uptake was quantified using target-to-background ratio, defined as the SUVmax of the lesion divided by the SUVmax of the normal liver parenchyma.
Sensitivity
To calculate the sensitivity of 68Ga-DOTATATE and 68Ga-NODAGA-JR11 PET/CT, comparison with conventional imaging using contrast-enhanced MRI or CT was performed. Patients with available conventional imaging (MRI was preferred if both MRI and CT were available) within 1 y of the PET study were included in the analysis.
The lesion-based sensitivity of the PET studies was calculated as the percentage of lesions on conventional imaging positive for PET uptake. An alternative combination strategy was also used to calculate the sensitivity; that is, lesions visible on any of the 3 imaging modalities (conventional imaging, 68Ga-DOTATATE PET, or 68Ga-NODAGA-JR11 PET) were considered true lesions for NET. This was part of an exploratory extension and not part of the original study design.
Potential Impact on Clinical Management
To answer whether the imaging findings on 68Ga-NODAGA-JR11 PET/CT would change the patients’ clinical management, an imaging-based post hoc analysis was done. The 68Ga-NODAGA-JR11 PET results would be considered to potentially impact clinical management if any of the criteria shown in Table 1 were fulfilled. This was part of an exploratory extension and not part of the original study design.
Criteria to Define Potential Impact on Clinical Management of 68Ga-NODAGA-JR11 over 68Ga-DOTATATE
Statistical Analysis
Data were expressed as mean ± SD. Differences in SUVmax and target-to-background ratio between 68Ga-NODAGA-JR11 and 68Ga-DOTATATE were evaluated using paired t tests (SPSS, version 23) and only for matched lesions. The lesion numbers were statistically compared using sign tests. The McNemar test was used to compare sensitivity. A P value of less than 0.05 was considered to indicate a significant difference.
RESULTS
The clinical characteristics are summarized in Supplemental Table 1 (supplemental materials are available at http://jnm.snmjournals.org). Both tracers were tolerated well by all patients. No adverse events were reported.
Biodistribution
Mild to moderate uptake was noted for 68Ga-NODAGA-JR11 in almost all organs except the urinary tract (Fig. 1; Table 2). Compared with 68Ga-DOTATATE, 68Ga-NODAGA-JR11 showed lower background uptake except for the lung.
Comparison of whole-body maximum-intensity projections of 6 representative patients (patients 26, 32, 33, 36, 43, and 45 from left to right). Physiologic uptake is seen at pituitary gland, salivary glands, thyroids, adrenal glands, spleen, and bowel on 68Ga-DOTATATE images (top row). Nevertheless, these normal organs show minimal or mild uptake on 68Ga-NODAGA-JR11 images (bottom row). Besides, 68Ga-NODAGA-JR11 depicts more liver lesions than 68Ga-DOTATATE, with lower liver background uptake.
Comparison of Normal-Organ SUVmax Between 68Ga-NODAGA-JR11 and 68Ga-DOTATATE PET/CT
Tumor Detection
In total, 1,095 and 1,003 focal lesions were depicted on 68Ga-NODAGA-JR11 and 68Ga-DOTATATE PET/CT, respectively (P = 0.007; Supplemental Table 2).
On patient-based comparison, 68Ga-NODAGA-JR11 demonstrated a better detection ability for liver lesions (Fig. 1). Of 37 patients with liver metastases, 54.1% (20/37) showed more liver metastases on 68Ga-NODAGA-JR11 than on 68Ga-DOTATATE, whereas 35.1% (13/37) demonstrated comparable results. Only 4 patients had fewer liver lesions detected on 68Ga-NODAGA-JR11 PET/CT (Fig. 2).
Patient-based comparison of lesion detection.
On lesion-based comparison, 68Ga-NODAGA-JR11 detected significantly more liver lesions (673 vs. 584, P = 0.002; Fig. 3).
Lesion-based comparison of lesion detection. 68Ga-DOTATATE and 68Ga-NODAGA-JR11 PET/CT images are shown of postoperative gastric NET patient (patient 36) with multiple liver metastases. (A) 68Ga-DOTATATE maximum-intensity projection shows several liver lesions with high uptake. (B) Transaxial fusion image appears to show abnormal moderate-uptake lesion (arrow) in high liver background uptake. (C and D) One transaxial fusion image (C) shows liver metastasis (arrow), whereas no liver lesion was found in another (D). (H) 68Ga-NODAGA-JR11 maximum-intensity projection shows many more liver lesions than 68Ga-DOTATATE. (E–G) Transaxial fusion images at same 3 levels as in B–D similarly found liver lesions that matched those on 68Ga-DOTATATE (solid arrows). 68Ga-NODAGA-JR11 PET/CT shows lower liver background uptake and additional liver metastases (dashed arrows) at sites that were negative on 68Ga-DOTATATE PET/CT.
For primary tumors, bone metastases, and lymph node metastases, the 2 tracers were comparable on both patient-based and lesion-based comparison.
Tumor Uptake
For matched lesions, 68Ga-NODAGA-JR11 demonstrated comparable uptake except that uptake by bone lesions was significantly lower (Table 3). The target-to-background ratio for liver lesions, however, was significantly higher on 68Ga-NODAGA-JR11 than on 68Ga-DOTATATE (6.4 ± 8.7 vs. 3.1 ± 2.6, P = 0.000).
Uptake of Matched Lesions on 68Ga-NODAGA-JR11 and 68Ga-DOTATATE PET/CT
Sensitivity Estimation
There were 15 patients (6 from one center and 9 from the other) with available contrast-enhanced MRI or CT within 1 y of the study. The median interval between PET and conventional imaging was 3 mo (range, 0–12 mo). Eight of the follow-up conventional imaging studies were contrast-enhanced MRI, and 7 were contrast-enhanced CT. There were a total of 180 lesions detected on conventional imaging; 165 and 139 of the lesions were positive on 68Ga-NODAGA-JR11 and 68Ga-DOTATATE, leading to a mean sensitivity of 91.7% (range, 87.6%–95.7%) and 77.2% (range, 71.0%–83.4%), respectively. All lesions missed by 68Ga-NODAGA-JR11 were small, usually less than 0.5 cm. Bone and other lesions outside the scan field of view on conventional imaging were not included in the analysis.
Notably, 26 lesions within the scan field of view on conventional imaging were identifiable on PET but not on conventional imaging. All were positive for uptake on both PET scans, including 1 pancreatic lesion missed on MRI, 1 pancreatic lesion missed on CT, and 24 hepatic lesions missed on CT. All 26 lesions were in locations typical of NET and had a markedly elevated SUVmax (mean, 21.5 and 18.7 for 68Ga-NODAGA-JR11 and 68Ga-DOTATATE, respectively). After discussion with a radiologist and an oncologist, these 26 lesions were all considered NET lesions. There were 206 lesions identified using the combination strategy. The sensitivity of 68Ga-NODAGA-JR11, 68Ga-DOTATATE, and conventional imaging was 92.7% (range, 89.1%–96.3%), 79.6% (range, 74.1%–85.2%), and 87.4% (range, 82.8%–92.0%), respectively.
Potential Impact on Clinical Management
In 56% (27/48) of patients, there were discrepancies between 68Ga-NODAGA-JR11 and 68Ga-DOTATATE PET. Among them, 14.5% (7/48) of patients showed that 68Ga-NODAGA-JR11 PET might have potentially changed clinical management by detecting more lesions than 68Ga-DOTATATE according to the criteria. In 1 patient (patient 41), 68Ga-NODAGA-JR11 revealed fewer liver lesions on PET than did 68Ga-DOTATATE, a finding that could lead to insufficient management if standard 68Ga-DOTATATE PET were omitted. In the remaining 19 patients, the discrepancies were not significant enough to change clinical management (Table 4).
Potential Impact on Clinical Management of 68Ga-NODAGA-JR11 PET/CT Imaging
DISCUSSION
Our study prospectively compared lesion detection ability between 68Ga-NODAGA-JR11 and 68Ga-DOTATATE in a single group of patients. The strengths of the study include a prospective, bicenter design and a large cohort. The results favor 68Ga-NODAGA-JR11 because of a higher detection ability and better image contrast for liver metastases, as well as a potential impact on clinical management.
Despite lack of tracer internalization into tumor cells, preclinical and initial clinical studies have facilitated a shift from SSTR agonists to antagonists in recent years (3–5,13,14). In our previous study, the antagonist 68Ga-DOTA-JR11 showed an overall lesion detection ability comparable to that of 68Ga-DOTATATE and lower uptake (13). This finding contrasts somewhat with the present findings. In the current study, 68Ga-NODAGA-JR11 showed a higher lesion detection ability and comparable tumor uptake. The divergence might be explained by the different SSTR2 affinity (50% inhibitory concentrations of 29, 1.2, and 0.2 nmol/L for 68Ga-DOTA-JR11, 68Ga-NODAGA-JR11, and 68Ga-DOTATATE, respectively) (15).
In addition, we did notice a statistically significant difference in bone lesion uptake between 68Ga-NODAGA-JR11 and 68Ga-DOTATATE PET/CT. The difference, however, was small and did not affect the detection of bone lesions by 68Ga-NODAGA-JR11. It can in part be explained by the low bone marrow uptake of 68Ga-NODAGA-JR11. Therefore, we do not think the difference is marked enough to generate clinical implications.
We observed that the bone marrow uptake of 68Ga-NODAGA-JR11 was lower than that of 68Ga-DOTATATE. This finding somewhat contradicts already published literature showing potentially increased hematologic toxicity with the therapeutic counterpart (177Lu-DOTA-JR11) in comparison to 177Lu-DOTATATE (16). The underlying mechanism of the hematologic toxicity of antagonist peptide receptor radionuclide therapy is currently not well understood. Bone marrow uptake is low using either a 68Ga- or a 177Lu-antagonist. The reason should not be the radiation to the general bone marrow, which was considered the main reason for radioactive hematologic toxicity. We agree with Reidy-Lagunes et al. that antagonists might be able to target SSTR expressed on bone marrow stem cells or progenitor cells, causing more severe and prolonged bone marrow suppression (16).
68Ga-NODAGA-JR11 and 68Ga-DOTATATE have a sensitivity of 91.7% (range, 87.6%–95.7%) and 77.2% (range, 71.0%–83.4%), respectively, using conventional imaging as a reference. Nevertheless, whereas dedicated contrast-enhanced MRI might be the best imaging modality for the detection of liver metastasis, conventional imaging may not serve as the perfect reference for whole-body SSTR PET imaging (17,18). That is to say, conventional imaging—contrast-enhanced CT in particular—is not sensitive enough for NET detection. Some SSTR-positive lesions observed on PET might not be visualized on conventional imaging (19). This is especially true for bone lesions, which are usually out of the scan field of view; nonosseous changing soft-tissue lesions that are embedded in surrounding soft tissues; and lymph nodes that are too small to fit the size criteria. This hypothesis was supported by our observation that 68Ga-NODAGA-JR11 had higher sensitivity than conventional imaging using a combination strategy. However, we must emphasize the importance of the modality chosen as the standard of truth. Of the 26 lesions missed on conventional imaging, 25 were missed by conventional CT. The results could be different if only MRI is involved.
The potential impact of antagonist PET on clinical management is essential for clinicians to know but has not been previously reported. Our post hoc analysis showed, in a subgroup (14.5%) of patients, that detection of more lesions by 68Ga-NODAGA-JR11 could have a potential impact on the patients’ clinical management. The ability to detect more liver metastases with antagonists may also change the therapeutic strategy if local therapy for hepatic lesions were to be considered. Partial liver resection would be contraindicated if additional bilobar liver lesions are identified (20). Besides, the lower background uptake of 68Ga-NODAGA-JR11 also helps to differentiate between physiologic uptake and true lesions. The intrapancreatic accessory spleen is a congenital entity often incidentally detected by imaging and can be misdiagnosed as NET. High uptake of 111In-diethylenetriamine pentaacetate octreotide and 68Ga-DOTATATE in the intrapancreatic accessory spleen has been described in some case reports, making it an important pitfall of agonist imaging (21,22). Compared with 68Ga-DOTATATE, 68Ga-NODAGA-JR11 had much lower spleen uptake. The intrapancreatic accessory spleen tends to have uptake comparable to or lower than that in the spleen parenchyma on 68Ga-NODAGA-JR11 PET/CT, whereas well-differentiated NET tends to have higher uptake than in the spleen (Supplemental Fig. 1). In that sense, antagonist imaging might provide a novel strategy to distinguish NET from an intrapancreatic accessory spleen.
The advantage of antagonists lies not only in better imaging of NETs but also in treatment decision-making. As a diagnostic companion for 177Lu-DOTA-JR11, 68Ga-NODAGA-JR11 is essential to decide whether patients are eligible for peptide receptor radionuclide therapy with 177Lu-DOTA-JR11. In the phase I trial of 177Lu-DOTA-JR11, the overall response rate was 45% (9/20 patients) after only 2 cycles of treatment (16). In particular, peptide receptor radionuclide therapy with antagonists might be a better choice in patients with agonist-negative but antagonist-positive lesions (20). Zhang et al. provided preliminary evidence of efficacy using 177Lu-DOTA-LM3 treatment in a patient with 68Ga-DOTATOC–negative high-grade liver metastases (23). The patient had a nearly complete remission after 3 cycles of peptide receptor radionuclide therapy, with a total of 20.4 GBq of 177Lu-DOTA-LM3. Besides, higher tumor uptake and prolonged retention of 68Ga-NODAGA-JR11 may result in favorable tumor-to-organ ratios and thus be an advantage for therapy with SSTR2 antagonists, even if all the lesions are visible on SSTR2 agonist PET/CT imaging.
Our study had some limitations. A crossover design to reduce the impact of the sequence of the 2 PET studies could further strengthen our study. Because of the nature of post hoc analysis, the sensitivity calculation was limited by the small population. The image files of some patients were not available and thus could not be analyzed. Furthermore, the comparison was limited to the scan field of view of conventional imaging, and lesions outside this field, such as bone metastases, were not included in the analysis. In addition, the potential impact of 68Ga-NODAGA-JR11 PET on clinical management was analyzed using imaging-based criteria. The clinician’s opinion should also be included in future studies.
CONCLUSION
68Ga-NODAGA-JR11 shows better sensitivity and a higher target-to-background ratio than 68Ga-DOTATATE. The detection of more lesions by antagonists has a potential impact on clinical management in a subgroup of patients.
DISCLOSURE
This study was funded in part by the National Natural Science Foundation of China (82171982 and 82071967), the CAMS innovation fund for medical science (CIFMS-2021-I2 M-1-025), the Joint Funds for the Innovation of Science and Technology of Fujian Province (2020Y9101), scientific research project funds from the Education Department of Fujian Province (JAT210089), and the Startup Fund for Scientific Research of Fujian Medical University (2021QH1062). No other potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: Is PET/CT with the SSTR antagonist 68Ga-NODAGA-JR11 more efficient at detecting lesions than the agonist 68Ga-DOTATATE in patients with metastatic, well-differentiated NETs?
PERTINENT FINDINGS: Forty-eight patients with metastatic, well-differentiated NETs were prospectively enrolled in a bicenter study to compare lesion detection ability between 68Ga-NODAGA-JR11 and 68Ga-DOTATATE PET/CT. 68Ga-NODAGA-JR11 was better able to detect liver metastases and had a higher target-to-background ratio than 68Ga-DOTATATE.
IMPLICATIONS FOR PATIENT CARE: The new radiopharmaceutical 68Ga-NODAGA-JR11 is more valuable than SSTR agonists in patients with NETs, especially in patients with liver-dominant metastases.
Footnotes
Published online Jul. 20, 2023.
- © 2023 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication September 7, 2022.
- Revision received April 20, 2023.