Abstract
Neuroendocrine tumors (NETs) can be visualized using radiolabeled somatostatin analogs. We have previously shown the clinical potential of 64Cu-DOTATATE in a small first-in-human feasibility study. The aim of the present study was, in a larger prospective design, to compare on a head-to-head basis the performance of 64Cu-DOTATATE and 111In-diethylenetriaminepentaacetic acid (DTPA)-octreotide (111In-DTPA-OC) as a basis for implementing 64Cu-DOTATATE as a routine. Methods: We prospectively enrolled 112 patients with pathologically confirmed NETs of gastroenteropancreatic or pulmonary origin. All patients underwent both PET/CT with 64Cu-DOTATATE and SPECT/CT with 111In-DTPA-OC within 60 d. PET scans were acquired 1 h after injection of 202 MBq (range, 183–232 MBq) of 64Cu-DOTATATE after a diagnostic contrast-enhanced CT scan. Patients were followed for 42–60 mo for evaluation of discrepant imaging findings. The McNemar test was used to compare the diagnostic performance. Results: Eighty-seven patients were congruently PET- and SPECT-positive. No SPECT-positive cases were PET-negative, whereas 10 false-negative SPECT cases were identified using PET. The diagnostic sensitivity and accuracy of 64Cu-DOTATATE (97% for both) were significantly better than those of 111In-DTPA-OC (87% and 88%, respectively, P = 0.017). In 84 patients (75%), 64Cu-DOTATATE identified more lesions than 111In-DTPA-OC and always at least as many. In total, twice as many lesions were detected with 64Cu-DOTATATE than with 111In-DTPA-OC. Moreover, in 40 of 112 cases (36%) lesions were detected by 64Cu-DOTATATE in organs not identified as disease-involved by 111In-DTPA-OC. Conclusion: With these results, we demonstrate that 64Cu-DOTATATE is far superior to 111In-DTPA-OC in diagnostic performance in NET patients. Therefore, we do not hesitate to recommend implementation of 64Cu-DOTATATE as a replacement for 111In-DTPA-OC.
- neuroendocrine tumors
- cancer
- somatostatin receptor imaging
- 64Cu-DOTATATE
- 111In-DTPA-octreotide
- PET
- PET/CT
- SPECT
- molecular imaging
- prospective
In well-differentiated neuroendocrine tumors (NETs), somatostatin receptors are consistently overexpressed (1–3). This key feature of NETs may be visualized using radiolabeled somatostatin receptor ligands and SPECT or PET. Still, 111In-diethylenetriaminepentaacetic acid (DTPA)-octreotide is considered the standard method according to current international guidelines (4,5). However, because of the better sensitivity, spatial resolution, and inherently quantitative nature of PET, somatostatin receptor imaging (SRI) with PET is increasingly used. Most common is the use of the 68Ga-labeled somatostatin analogs, for example, 68Ga-DOTATATE (6–9), 68Ga-DOTATOC (10–12), and 68Ga-DOTANOC (13). However, the high positron energy of 68Ga limits spatial resolution and the short half-life of 68 min may be challenging logistically, because production has to be aligned with patients and to be repeated several times a day.
To overcome some of these challenges, we recently introduced 64Cu-DOTATATE, which has a long half-life of 12.7 h and can be produced as a once-daily batch with a shelf life of more than 24 h (14). These advantages of 64Cu-DOTATATE make it possible to produce the tracer in a central location and distribute it to local PET centers. The positron energy of 64Cu is also much lower than that of 68Ga and therefore should translate into better spatial resolution (Table 1). In addition, by choosing the octreotate peptid for this tracer, we have a better match for the commonly used therapy ligand 177Lu-DOTATATE, enabling accurate diagnostic imaging of the organs treated with 177Lu-DOTATATE.
Characteristics of Somatostatin Receptor Radiotracers
The use of DOTA as a chelate has been challenged in several studies (15,16). However, we demonstrated in our phase 1 study sufficient in vivo stability of the 64Cu-DOTA complex to obtain good-quality images (14), which we have also confirmed in preclinical studies comparing DOTA with other chelates (17–19). Data from our first-in-human study of 14 patients demonstrated that imaging was feasible using an activity of 200 MBq, with a favorable dosimetry of only 6.3 mSv (14). Compared with conventional SRI using 111In-DTPA-octreotide (111In-DTPA-OC), additional lesions were detected in 6 of 14 patients and 5 of these additional lesions were found in organs not previously known as metastatic. In the same study, we also found that PET could be performed between 1 and 24 h, and we selected the 1-h time point for routine use. Taken together, data from our previous study allowed us to establish a standardized protocol for 64Cu-DOTATATE in NET.
The aim of the present study was therefore, on a larger scale, to perform a prospective study comparing on a head-to-head basis in 112 NET patients the performance of 64Cu-DOTATATE and 111In-DTPA-OC as a basis for implementing 64Cu-DOTATATE as a routine. To evaluate discrepant findings, patients were followed for 42–60 mo after the 2 scans were obtained.
MATERIALS AND METHODS
Patients and Inclusion Criteria
Eligible patients had histopathologically confirmed NETs of gastroenteropancreatic or pulmonary origin of all grades (20). They were enrolled in the study from November 2009 to May 2011, and follow-up ended on November 15, 2014, for evaluation of discrepant imaging findings. Accordingly, patients were followed for 42–60 mo. Patients were offered study inclusion in the case of referral to conventional SRI with 111In-DTPA-OC as part of their routine examinations. 64Cu-DOTATATE PET and 111In-DTPA-OC SPECT were performed within a maximum of 60 d (mean, 24 d) in a random order. The study group consisted of 63 (56%) men and 49 (44%) women with a mean age of 62 y (range, 30–84 y) (Table 2).
Characteristics of Study Population (n = 112)
All patients gave written informed consent before inclusion. The study was approved by the Regional Scientific Ethical Committee (reference no. H-D-2008-045).
Radiotracer, Image Acquisition, and Reconstruction
Radiotracers
64Cu-DOTATATE was produced in-house as previously described (14). The labeling efficiency was greater than 95% (determined with radio–reversed-phase high-performance liquid chromatography), and the specific activity was 4.78 MBq/mmol. 111In-DTPA-OC was purchased from Covidien and prepared in accordance with the instructions of the manufacturer.
PET/CT
All patients were PET-scanned using a Biograph 64 TruePoint PET/CT scanner (Siemens Medical Solutions) with an axial field of view of 216 mm and a transaxial field of view of 205 mm. Axial and transaxial resolutions were 4.7 and 4.2 mm, respectively. Emission scans were acquired 1 h after injection of 183–232 (mean, 202) MBq of 64Cu-DOTATATE, with an average of 44.6 mmol (40.2–50.9 mmol) of octreotate administered per dose. Whole-body PET scans (skull to mid thigh) were obtained in 3-dimensional mode, with an acquisition time of 3 min per bed position. PET data were reconstructed with the TrueX (Siemens Medical Solutions) algorithm using 3 iterations and 21 subsets and smoothed by gaussian filter (full width at half maximum, 2 mm). CT data were used for attenuation correction. Before the PET scan, a diagnostic CT scan was obtained with 3-mm slice thickness, 120 kV, and a quality reference of 225 mAs modulated by the Care Dose 4D automatic exposure control system (Siemens Medical Solutions). Unless contraindicated, 75 mL of iodine-containing contrast agent were administered using an automatic infusion system (Optiray 300; Covidien), with scan delays of 60 s (flow rate, 1.5 mL/s), followed by an infusion of 100 mL of NaCl (flow rate, 2.5 mL/s). Furthermore, patients had been asked to drink 500 mL of water 25 min before image acquisition. PET and CT images were fused and reviewed on a dedicated workstation.
SPECT/CT
Planar and tomographic images were acquired using dual-head hybrid SPECT/CT cameras (Precedence 16-slice scanner [Philips Healthcare]; VG Hawkeye [GE Healthcare]) after intravenous administration of 181–268 (mean, 218) MBq of 111In-DTPA-OC. Whole-body planar images (anterior and posterior, scan speed 5 cm/min, 512 × 1,024 matrix) were acquired at 24 h after injection and at 48 h after injection if relevant (15-min static planar image [256 × 256 matrix] of the abdomen using a large-field-of-view medium-energy collimator). During the same session, SPECT over the abdomen or chest per indication (20 s/step, 128 angles, 128 × 128 matrix) was performed. A low-dose CT (20 mA, 140 kV [Precedence] or 2.5 mA, 140 kV [Hawkeye]) was used as an anatomic guide and for attenuation correction. Scatter correction was used, and SPECT and CT were fused and reviewed on dedicated workstations (EBW [Philips Healthcare]; eNTEGRA [GE Healthcare]).
Visual Image Analysis and Activity Quantification
PET/CT and SPECT/CT were analyzed separately by 2 different teams consisting of 2 experienced interpreters. The 2 teams were masked to the images and readings of the other team. There had to be a clearly detectable lesion on the somatostatin receptor image to be counted as SRI-positive in the case of lesion discovery on the respective fused PET/CT or SPECT/CT images. CT data, masked for both SPECT and PET, were additionally evaluated by an experienced radiologist and used as a reference. The absolute number of lesions per organ system was obtained with a numeric limitation of 10 lesions per organ and 30 positive findings for lymph nodes per patient.
Discrepant SRI findings—that is, foci that were recognized only on 1 of the 2 SRI methods—were classified as true-positive if they were positive on the CT-only assessment performed masked to PET and SPECT results but using the coregistered CT or were confirmed during later follow-up on biopsies, by the other SRI modality, or on additional imaging (CT or ultrasonography).
Maximum standardized uptake values were calculated for the lesion with highest tracer uptake in each organ obtained by drawing spheric volumes of interest sufficiently large to encompass the whole lesion—that is, including a rim of surrounding normal tissue.
Comparisons between the 2 methods were performed on the basis of patients, organs, and lesions. Sensitivity, specificity, and accuracy were calculated on a patient basis.
Statistical Analyses
The McNemar test for paired proportions with continuity correction was applied to compare 64Cu-DOTATATE and 111In-DTPA-OC at the patient level. The 95% confidence intervals for sensitivity, specificity, accuracy, and predictive values were computed using the adjusted Wald method. A sign test was used to explore differences in lesion detection rates between the 2 SRI modalities. Two-sided P values of less than 0.05 were considered statistically significant.
RESULTS
Patient-Based Comparison
One hundred twelve patients underwent 64Cu-DOTATATE PET, 111In-DTPA-OC SPECT, and contrast-enhanced diagnostic CT in the inclusion period of 19 mo. In 100 of 112 NET patients, residual or recurrent disease was established on the basis of previous clinical evaluation, CT, and SRI within the scope of this study and prospective follow-up. Accordingly, 12 patients were negative for disease and of these 12, 8 were expectedly negative as they were newly operated on with removal of the known pathologic focus/foci and referred for evaluation of possible residual disease.
Eighty-seven patients were congruently positive on both 64Cu-DOTATATE PET and 111In-DTPA-OC SPECT scans. Ten patients with proven residual or recurrent disease were identified only by 64Cu-DOTATATE PET, leading to 97 true-positive 64Cu-DOTATATE PET cases and 87 true-positive 111In-DTPA-OC SPECT cases. 111In-DTPA-OC SPECT was false-negative in 13 patients, whereas 64Cu-DOTATATE PET was false-negative in 3 patients. Two of these 3 matching false-negative cases comprised patients with high-grade (G3) pancreatic NETs with liver metastases and Ki-67 indices of 40% and 30%, respectively. The third patient had been diagnosed with a bronchopulmonary carcinoid with a mitotic count of 1 and liver metastases. No false-positive results for either SRI modality were seen on a patient basis. Accordingly, sensitivity for 64Cu-DOTATATE and 111In-DTPA-OC was 97% and 87%, respectively. The comparison of diagnostic sensitivity, specificity, accuracy, positive predictive value, and negative predictive value and their 95% confidence intervals are shown in Table 3. The McNemar test revealed a statistically different performance of the modalities (P = 0.017), with 64Cu-DOTATATE having a higher sensitivity and accuracy.
Comparison of Diagnostic Performance of 64Cu-DOTATATE and 111In-DTPA-OC
Five of the 10 patients with residual or recurrent NETs, which were revealed only by 64Cu-DOTATATE PET, had liver metastases. A typical case with discrepant PET and SPECT findings is shown in Figure 1.
Patient with gastric NET, which was not seen on 111In-DTPA-OC SPECT (Precedence scanner) (A) but was clearly visible (arrows) on 64Cu-DOTATATE PET (B). Likewise, CT did not reveal gastric NET, but biopsy confirmed it to be a true-positive finding.
Another example in which PET compared with SPECT identified 3 additional organs with tumor involvement is shown in Figure 2.
Multiple small liver metastases (>10), peritoneal solitary tumor mass, and 3 lymph node metastases shown on 64Cu-DOTATATE PET/CT in patient with pancreatic NET. No foci were detected by 111In-DTPA-OC SPECT (Precedence scanner). All findings on PET were confirmed to be true-positive. (A) 111In-DTPA-OC planar images. (B) 64Cu-DOTATATE maximum-intensity-projection image with arrows pointing at liver and lymph node metastases. Insert is fused PET/CT of peritoneal solitary tumor mass. (C) Axial CT and SPECT of liver. (D) Axial CT and PET of liver revealing several small liver metastases.
Of 112 included patients, bone metastases were present in 38 cases (34%) of which PET detected 36 and CT the remaining 2. All SPECT-positive cases were also PET-positive. In contrast, of the 36 PET-positive cases, 13 cases (29%) were SPECT-negative. Figure 3 shows a patient in whom multiple bone metastases were revealed by 64Cu-DOTATATE in contrast to only 1 identified by 111In-DTPA-OC.
Patient in whom PET identified more than 30 bone metastases whereas SPECT (VG Hawkeye scanner) identified only 1, and no bone lesions were visible on CT. During follow-up, multiple bone lesions could be identified on subsequent CT and SPECT scans but not until 2 y after lesions were found on the PET scan. (A) 111In-DTPA-OC from 2010. (B) 64Cu-DOTATATE from 2010; arrow marks bone metastasis. (C) 111In-DTPA-OC from 2012. (D) 64Cu-DOTATATE from 2012.
In another case, PET identified the primary tumor in the terminal ileum and 1 abdominal lymph node metastasis whereas SPECT identified only the abdominal lymph node metastasis. The patient was subsequently operated on, and both foci were identified and removed.
Organ-Based Comparison
An organ-based analysis yielded concordant findings in 72 (64%) patients—that is, that both 111In-DTPA octreotide and 64Cu-DOTATATE PET consistently showed no lesions or consistently revealed at least 1 lesion for all organs/regions of the patient. In 40 of the 112 patients (36%), 64Cu-DOTATATE PET detected lesions in additional organs/regions. There did not seem to be a difference in performance between the 2 SPECT scanners used, because there were 6 of the 21 patients (29%) scanned using the VG Hawkeye with discrepant findings in comparison to 34 of the 91 patients (37%) scanned using the Precedence scanner with discrepant findings. In some patients, 2 or more additional organ involvements were identified by PET, giving a total of 58 additional organ involvements found on PET compared with SPECT (Table 4). During the 42–60 mo of follow-up, the additional organ involvements identified on PET were confirmed to be true-positive in 35 of the 40 patients. Details on discrepant organ findings and how they were confirmed are given in Tables 4 and 5.
Organ-Based Analysis, Subgroup of 40 Patients with Discrepant Findings
Cases Showing Discrepant Results Between 64Cu-DOTATATE PET and 111In-Octreotide SPECT Scans
The most common discrepant finding was lymph node metastases. Thus, in 14 of the 40 discrepant cases, no lymph node metastases were present on SPECT whereas PET identified up to 11 lymph node metastases. Only in 2 of the 40 patients did SPECT identify more than 5 lymph node metastases whereas PET revealed more than 5 lymph node metastases in 13 cases.
In 11 patients, no liver lesions were present on SPECT whereas PET identified up to 20 liver metastases. In 3 of these 11 patients, the liver metastases were also not evident on CT, but all 3 were later confirmed to be true-positive lesions. In general for all 40 discrepant cases, PET identified at least as many and often more lesions than both SPECT and CT.
All lesions detected on 111In-DTPA-OC SPECT were also detected on 64Cu-DOTATATE PET. In 10 of the 40 cases, no lesions were detected by SRS and in 5 of these patients the CT scans were also negative.
Lesion-Based Comparison
In total, 1,213 lesions were detected by 64Cu-DOTATATE PET in comparison to 603 lesions detected by 111In-DTPA-OC SPECT (Table 6). Of the 112 patients, 21 (19%) were scanned on the VG Hawkeye scanner, and in agreement with this, 123 (20%) of the 603 lesions detected by 111In-DTPA-OC were identified using the VG Hawkeye scanner. In 28 patients (25%), PET and SPECT identified the same number of lesions; in 52 patients (46%), PET identified 1–9 more lesions than SPECT; and in 32 patients (29%), PET identified 10 or more additional lesions than SPECT. Tables 5 and 6 provide further details.
Comparison of Absolute Number of Lesions Detected by PET and SPECT Based on Organ Location
The most frequent finding was liver metastases, for which 468 lesions were present on PET, compared with 320 on SPECT. The clearest difference was the detection rate of bone metastases for which 208 lesions were found on PET, compared with 90 on SPECT.
DISCUSSION
In this prospective cohort study of 112 patients, we compared a new PET tracer, 64Cu-DOTATATE, with the SPECT-based gold standard for imaging of somatostatin receptors in tumor lesions of NET patients (21,22). We found that the PET-based tracer was far superior to the current gold standard. With these results, we conclude that the 64Cu-DOTATATE PET tracer can safely be implemented as a routine for diagnostic imaging of NET patients.
Recently, we developed and tested in a phase 1 study in 14 patients a new 64Cu-based tracer, 64Cu-DOTATATE, targeting somatostatin receptors (14). Although these initial results were most promising, a larger study is needed before implementing the tracer into clinical routine as replacement for 111In-DTPA-OC (Octreoscan; Covidien). In line with this, we prospectively studied a larger series of 112 consecutive NET patients in a head-to-head comparison study. In the present study, we found that twice as many foci were identified. Still, whether 10 or 15 liver metastases are found is not necessarily important additional information. However, also when analyzing organ by organ, 58 additional organ systems were found to be involved in 40 patients when 64Cu-DOTATATE was used, compared with 111In-DTPA-OC, which could potentially be of importance for the prognosis (23,24) and treatment selection (25–28). Nevertheless, more foci found by PET are not necessarily representative of true-positive foci. Therefore, we performed a long follow-up to be able to evaluate whether foci were true-positive or false-positive. With a follow-up of 42–60 mo, we found that 46 (79%) of the 58 discrepant findings could be evaluated as true-positive on PET, and in no case were PET findings false-positive.
Accordingly, 64Cu-DOTATATE performs better than 111In-DTPA-OC because it detects more regions truly involved in the disease. When better performance is additionally accompanied by a lower radiation dose of 6.3 mSv (14,29) and an easier workflow, for example, a 1-d instead of 2-d procedure, we do not hesitate to recommend implementing this technique in our routine as a replacement for 111In-DTPA-OC.
It may be argued that 68Ga-based tracers such as 68Ga-DOTATATE, 68Ga-DOTATOC, or 68Ga-DOTANOC would equally perform better than 111In-DTPA-OC (30). Indeed, such comparative studies also found more foci identified with 68Ga-based tracers than with SPECT (27,31,32). These studies on average found 30% (27,32,33) more foci using PET than SPECT. Compared with our study in which we found a higher percentage of additional lesions (50%), it seems we could have a better detection rate. Moreover, 64Cu-DOTATATE found several lesions not detected on CT scans, which is less commonly reported for the 68Ga-labeled tracers (32,34,35), which could easily be explained by the better resolution obtained when using 64Cu than 68Ga due to the 4-fold-lower positron range of 64Cu. Moreover, in contrast to our study with 64Cu, most of the 68Ga studies did not rigorously perform a long-term follow-up to establish whether the additional foci were true-positive.
It could be asked how our tracer would perform in a head-to-head comparison with 68Ga-based somatostatin receptor tracers. Currently we do not know this, and because only a study also including a long-term follow-up can answer which tracer is best, such a study is not yet available. However, with 111In-DTPA-OC still being the most commonly used tracer, in particular in the United States, we find our current study timely and of clinical relevance.
CONCLUSION
With detection of twice as many lesions, identification of disease involvement in organs not previously identified in one third of enrolled patients, and a favorable dosimetry and workflow, we have demonstrated that 64Cu-DOTATATE is far superior to 111In-DTPA-OC. Accordingly, we do not hesitate to recommend implementation of this technique in our routine as a replacement for 111In-DTPA-OC.
DISCLOSURE
The costs of publication of this article were defrayed in part by the payment of page charges. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734. This work was in part made possible by the generous support of grants from the following funds, which is gratefully acknowledged: the Danish National Advanced Technology Foundation, the John and Birthe Meyer Foundation, the Danish Medical Research Council, the Rigshospitalets Research Foundation, the Svend Andersen Foundation, the AP Moller Foundation, the Novo Nordisk Foundation, the Lundbeck Foundation, and the Danish Cancer Society. No other potential conflict of interest relevant to this article was reported.
Footnotes
Published online May 7, 2015.
- © 2015 by the Society of Nuclear Medicine and Molecular Imaging, Inc.
REFERENCES
- Received for publication February 27, 2015.
- Accepted for publication April 10, 2015.
Jump to section
Related Articles
Cited By...
- Preclinical Comparison of the 64Cu- and 68Ga-Labeled GRPR-Targeted Compounds RM2 and AMTG, as Well as First-in-Humans [68Ga]Ga-AMTG PET/CT
- An Investigation of Lesion Detection Accuracy for Artificial Intelligence-Based Denoising of Low-Dose 64Cu-DOTATATE PET Imaging in Patients with Neuroendocrine Neoplasms
- An Investigation of Lesion Detection Accuracy for Artificial Intelligence-Based Denoising of Low-Dose 64Cu-DOTATATE PET Imaging in Patients with Neuroendocrine Neoplasms
- SNMMI Procedure Standard/EANM Practice Guideline for SSTR PET: Imaging Neuroendocrine Tumors
- Choice Is Good at Times: The Emergence of [64Cu]Cu-DOTATATE-Based Somatostatin Receptor Imaging in the Era of [68Ga]Ga-DOTATATE
- Diagnosing metastatic pheochromocytoma: Trick lies in attention to details!
- Semiautomatic Tumor Delineation for Evaluation of 64Cu-DOTATATE PET/CT in Patients with Neuroendocrine Neoplasms: Prognostication Based on Lowest Lesion Uptake and Total Tumor Volume
- Somatostatin Receptor Imaging and Theranostics: Current Practice and Future Prospects
- 18F-FDG PET is Superior to WHO Grading as a Prognostic Tool in Neuroendocrine Neoplasms and Useful in Guiding PRRT: A Prospective 10-Year Follow-up Study
- 64Cu-DOTATATE PET in Patients with Neuroendocrine Neoplasms: Prospective, Head-to-Head Comparison of Imaging at 1 Hour and 3 Hours After Injection
- Updates to the Appropriate-Use Criteria for Somatostatin Receptor PET
- 64Cu-DOTATATE PET/CT and Prediction of Overall and Progression-Free Survival in Patients with Neuroendocrine Neoplasms
- 64Cu-DOTATATE PET/CT for Imaging Patients with Known or Suspected Somatostatin Receptor-Positive Neuroendocrine Tumors: Results of the First U.S. Prospective, Reader-Masked Clinical Trial
- Improving Contrast and Detectability: Imaging with [55Co]Co-DOTATATE in Comparison with [64Cu]Cu-DOTATATE and [68Ga]Ga-DOTATATE
- 64Cu-SARTATE PET Imaging of Patients with Neuroendocrine Tumors Demonstrates High Tumor Uptake and Retention, Potentially Allowing Prospective Dosimetry for Peptide Receptor Radionuclide Therapy
- Preclinical Investigation of 212Pb-DOTAMTATE for Peptide Receptor Radionuclide Therapy in a Neuroendocrine Tumor Model
- Somatostatin Receptor 2-Targeting Compounds
- Head-to-Head Comparison of 64Cu-DOTATATE and 68Ga-DOTATOC PET/CT: A Prospective Study of 59 Patients with Neuroendocrine Tumors
- Molecular Imaging of Gastroenteropancreatic Neuroendocrine Tumors: Current Status and Future Directions