In Sweden, approximately 400 patients per year are newly diagnosed with cancer of the esophagus and an additional 200 with cancer of the cardia [
CT has the disadvantage that it is less able to diagnose overgrowth of the tumor to adjacent organs and to detect small tumors <1 cm and growth to different layers in the esophageal wall. PET and the combination of 18F-FDG-PET/CT are both more accurate than CT alone with respect to diagnosing lymph node metastasis close to the tumor, and therefore staging [
While the surface of a cell usually includes some amount of somatostatin receptors, these receptors are over-expressed to a very high degree in multiple neoplastic diseases such as neuroendocrine tumors [
When this study was initiated, there were several radiopharmaceuticals available for somatostatin receptor scintigraphy, including 99mTc-depreotide. This tracer is 99mTc-labeled, and demonstrates good imaging characteristics with a short investigation protocol. 99mTc-depreotide has proven valuable in diagnosing pulmonary nodules [
The aim of the present study was to find out whether esophageal cancer can be imaged scintigraphically with 99mTc-depreotide and to determine the uptake characteristics of 99mTc-depreotide in the two main cancer types of the esophagus and relate these to results in patients with benign lesions (Barrett’s esophagus).
The study was approved by the Regional Ethical Review Board in Stockholm, Sweden and the Radiation Safety Committee at Karolinska University Hospital, Huddinge.
34 patients with dysphagia were referred to the Surgery Department at Huddinge University Hospital and further examined with gastroscopy, EUS, and CT. Nine of these were female and 25 male, with a median age of 63-64 years (range: 33–85 years). Among the 34 patients, 21 had cancer of the esophagus and 13 had Barrett’s esophagus. Five of the 21 patients with esophageal cancer had a Barrett’s esophagus too. The cancer diagnosis was established by histopathological examination of biopsy specimens in 19 cases and with EUS and cytological confirmation of diagnosis in 2 cases. All patients with Barrett’s esophagus were diagnosed via endoscope and subsequent multiple biopsies.
Locoregional lymph nodes were evaluated with EUS and histological examination of surgical specimen.
99mTc-depreotide (740 MBq) was administered via an antecubital vein. Single-photon emission computed tomography (SPECT) of the thorax was performed at 2 and 4 hours after injection, with the arms elevated, using three different gamma cameras. Most of the patients (25 of 34) were examined with a double-headed gamma camera (E-Cam, Siemens, Erlangen, Germany) and low-energy high-resolution parallel-hole collimators, using a 128 × 128 matrix, 64 projections through 360° rotation, and an acquisition time of 40 s per projection. An additional 5 patients were examined with a double-headed gamma camera (DST-XL; Sopha Medical Vision Scandinavia AB, Gif-sur-Yvette, France) and low-energy ultra-high-resolution parallel-hole collimators, using the same acquisition parameters as above. Finally, 4 patients were examined with a three-headed gamma camera (Picker IRIX, Cleveland, Ohio, USA) and low-energy high-resolution parallel-hole collimators, using a 128 × 128 matrix, 60 projections through 360° rotation, and an acquisition time of 64 s per projection. Transverse slices were reconstructed with an iterative algorithm (HOSEM v 3.5 iterative program; Hermes/NUD, Stockholm, Sweden) and formatted as a 128 × 128 matrix without attenuation correction. Images were postfiltered with a three-dimensional Fourier filter (Butterworth filter) with a cut-off frequency of 1.1 cycles/cm (order 5.00).
The results were evaluated both through visual assessment and through quantitative calculations in the 2-hour and 4-hour images. CT scans were used for an accurate localization of the 99mTc-depreotide uptake and for placement of the region of interest (ROI). On visual assessment, any focal 99mTc-depreotide uptake in the region of the known esophageal lesion was considered pathological. The quantitative evaluation of 99mTc-depreotide uptake was performed retrospectively on SPECT images in all 34 patients. First, an ROI was drawn manually around the esophageal tumor on each slice, using small margins. Next, a background ROI was drawn in healthy lung parenchyma (Figure
Evaluation of scintigraphic images with 99mTc-depreotide. Region of interest (ROI) was drawn manually around the esophageal tumor on each slice, using small margins, and a background ROI was drawn in healthy lung parenchyma.
For patients with negative uptake (i.e., tumor count density lower than the lung background count density), the uptake was scored as zero. Due to the small number of patients in each group, a nonparametric test was chosen. A two-sided Mann-Whitney
Among the 21 patients with cancer of the esophagus, 8 had squamous cell carcinoma, 11 had adenocarcinoma, 1 had an undifferentiated cancer, and 1 had an intramucosal cancer. Tumor size varied from 5 mm to 11 cm. The position of the tumor was in the proximal esophagus in 2 cases, in the middle part in 5 cases, and in the distal part in the remaining 14 cases.
Visual assessment revealed a pathological 99mTc-depreotide uptake in 16 of the 21 cancer patients (true-positive 76%) and an absence of pathological uptake in the remaining 5 (false-negative 24%). Six of the eight patients with squamous cell carcinoma and nine of the eleven patients with adenocarcinoma showed a pathological 99mTc-depreotide uptake. The remaining patient with pathological uptake had an 11 cm undifferentiated cancer in the mid-esophagus with a very high 99mTc-depreotide uptake. Among the false-negative cases, one had a small (5 mm) squamous cell cancer located in the middle part of the esophagus. The remaining four undetectable cancers were above 1 cm in size (varying from 12 × 9 mm to 12 × 38 mm) and were located in the distal part of the esophagus. The details of all recordings are given in Table
Tumor type, size, and location, CT result, and 99mTc-depreotide uptake in 21 esophageal cancer patients.
Diagnosis | Tumor size in mm | Location | CT | 99mTc-depreotide | |
---|---|---|---|---|---|
1 | Sqcc | 5 | Middle | Neg. | Neg. |
2 | Ac | 17 × 45 | Distal | Pos. | Pos. |
3 | Sqcc | 55 × 40 | proximal | Pos. | Pos. |
4 | Ac in B. | 65 × 55 × 13 | Distal | Pos. | Pos. |
5 | Ac | 90 × 75 × 25 | Distal | Pos. | Pos. |
6 | Imc and B. | 12 × 9 | Distal | Neg. | Neg. |
7 | Sqcc | 30 × 10 | Middle | Pos | Pos |
8 | Ac | 20 × 90 | Middle | Pos. | Pos. |
9 | Ac | 60 × 25 × 9 | Distal | Pos. | Pos. |
10 | Sqcc | 50 × 45 | Distal | Pos. | Pos. |
11 | Small cell cancer | 110 × 24 | Middle | Pos. | Pos. |
12 | Ac in B | 60 × 65 | Distal | Pos. | Pos. |
13 | Ac | 25 × 15 | Distal | Neg. | Pos. |
14 | Ac in B | 20 × 25 | Distal | Pos. | Pos. |
15 | Sqcc | 60 × 10 | proximal | Pos. | Pos. |
16 | Sqcc | 15 × 55 | Distal | Pos. | Pos. |
17 | Ac | 15 × 50 | Distal | Pos. | Pos. |
18 | Ac | 15 × 15 | Distal | Neg. | Neg. |
19 | Ac in B | 20 × 25 | Distal | Neg. | Neg. |
20 | Sqcc | 14 × 5 | Middle | Neg. | Pos. |
21 | Sqcc | 23 × 38 × 12 | Distal | Pos. | Neg. |
Ac: adenocarcinoma; Sqcc: squamous cell carcinoma; B: Barrett's esophagus; Imc: intramucosal cancer. 99mTc-depreotide uptake classified as negative or positive based on visual assessment. CT pos.: tumor is visible on the CT images; CT neg.: tumor is not visible on the CT images.
There was no 99mTc-depreotide uptake in the columnar metaplastic mucosa in any of the 13 Barrett’s patients, irrespective of the presence of low and high-grade dysplasia in the metaplastic epithelium (Figure
99mTc-depreotide uptake measured 2 hours after injection in patients with esophageal cancer and Barrett’s esophagus.
There were no significant differences between the ROI delineation and quantitative measurement of 99mTc-depreotide performed on the 2-hour acquisitions and those performed on the 4-hour acquisitions. A corresponding second ROI delineation and quantification, performed 6 months later, gave consistent results. Both intraobserver and interobserver variability was low with
A statistically significant difference (
In the 13 patients who had lymph node metastases at the final examination of the surgical specimen and with EUS only 5 showed 99mTc-depreotide uptake in the area of the lymph nodes.
In this study, we have shown that the imaging of esophageal cancer by means of somatostatin receptor scintigraphy with 99mTc-depreotide is feasible. Our hypothesis was based on two facts: first, that the physiological 99mTc-depreotide uptake in the thorax is low. Therefore, this could be a suitable area for tumor detection in most cases, and, second, that esophageal cancer has the same main histopathological types as lung cancer such as adenocarcinoma and squamous cell carcinoma. As scintigraphy with 99mTc-depreotide is useful for lung cancer detection, this second fact suggested that it could also be applied in esophageal cancer.
The majority of tumors (16/21) displayed a significant uptake of the tracer which could be clearly distinguished from that in the surrounding tissue. It was not unexpected that tumors under or near 10 mm in size were missed on the scintigraphic images. The detection limit of conventional gamma camera due to poor spatial resolution is well known, and according to widespread consensus scintigraphic methods are not suitable for screening purposes for any cancer types. Another observation is that even larger tumors in the distal part of the esophagus, 4 of 13 in the present study, could be missed with this method. Uptake of 99mTc-depreotide in lung cancers located in the lowest part of the right low lobe [
Our sensitivity figure of 76% is only an approximate value, due to the small number of patients in this study. Still, this is somewhat lower than both the sensitivity for detecting lung cancers [
As a control group, we used patients with Barrett’s esophagus. Barrett’s esophagus refers to an abnormal change (metaplasia) in the cells in the lower end of the esophagus. It is thought to be caused by damage from chronic acid exposure or reflux oesophagitis. This metaplasia confers an increased risk of adenocarcinoma. None of the 13 cancer-free Barrett’s esophagus patients in this study showed an increased 99mTc-depreotide uptake. Meanwhile, only 3 of 5 patients with both cancer and Barrett’s esophagus showed an increased 99mTc-depreotide uptake, leaving 2 false-negative results. The specificity of 100% for the applied method is high but should be used with caution, as the number of patients was relatively low and the spectra of different benign conditions in the esophagus was not fully represented in this pilot study.
Our results in the detection of loco-regional lymph node metastases were unsatisfactory. Only 5 of 13 patients with metastases seen with EUS and confirmed by histological examination were clearly detected by 99mTc-depreotide scintigraphy. It is probably caused by the close location to the primary tumor, where a high depreotide uptake cannot be separated from the uptake in the metastatic lymph nodes. It was disappointing to note that very few of the local, metastatic lymph nodes could be detected by this method. Through this pilot observation it can be envisioned that this technology cannot add to the available methods, such as EUS, in determining the node status in oesophageal cancer during the diagnostic and therapeutic workup.
As this study is the first of its kind, we considered it important to explore whether the quantitative assessment was reliable between different investigators and over time. Both intraobserver and interobserver variability was very low meaning that the applied calculations have good reliability.
We applied a somatostatin receptor scintigraphy with 99mTc-depreotide in a previously nonexplored cancer type where the optimal acquisition time was unknown. We used the same starting point for the imaging session as for the standard procedure in the detection of lung cancer, that is, 2 hours after injection [
Although future immunohistochemical studies will be needed to carefully map the density, detailed distribution, and localization of somatostatin receptors in the squamous cell esophageal carcinoma and adenocarcinomas, our data indicate that there is no major difference as reflected by the similarity in tracer accumulation between these two major tumor types. It is, however, of particular interest that in patients with Barrett’s esophagus, no accumulation of tracer was observed either in those with or in those without dysplastic histomorphologic changes in the columnar epithelium. Since there is no corresponding preneoplastic condition, concerning the squamous cell carcinoma development, it can be hypothesized that the somatostatin receptor expression reaches far higher levels in infiltrative neoplastic growth than in the intraepithelial neoplastic disease states. If so, this observation may be potentially very important and offer unique clinical opportunities, for example, when PET/CT or somatostatin receptor scintigraphy with 99mTc-depreotide technologies are applied.
Results of this pilot study showed feasibility of imaging oesophageal carcinoma with labeled somatostatin receptor analogue. We used a single-photon emitting tracer, but the results should be applicable or even better with positron emitting tracers. Use of PET/CT cameras combines a better spatial resolution of functional PET imaging with detailed anatomical information leading to a higher sensitivity. A multitude of new PET analogs are applied, whereas
After the successful visualization of somatostatin receptor positive tumors, a logical next step would be to use radiolabelled somatostatin analogues as a treatment of these patients. Such attempts were undertaken in patients with inoperable and/ormetastatic neuroendocrine tumors. While the objective responses for chemotherapy with the median time to progression is reported to be less than 18 months, PRRT with 90Y-octreotide or 177Lu-octreotate performs considerably better with a median time to progression of 30 and 40 months, respectively [
What could the future clinical application of our results be? Obviously, the method is not suitable neither for screening or primary diagnosis, because of methods modest sensitivity. Could this method be used for the detection of distant metastases expressing somatostatin receptors with somatostatin receptor-mediated radionuclide therapy as a consequent result? Does the uptake of the 99mTc-depreotide could be related to the prognosis of the oesophageal tumor? May this method be used in evaluation of treatment response in patients with tracer uptake? Does the natural history of Barrett’s esophagus and its malignisation could be predicted by tracer uptake? These issues are still to be answered.
Scintigraphic examination with 99mTc-depreotide is feasible for imaging esophageal cancer, but the method is not suitable neither for screening or primary diagnosis, because of methods modest sensitivity. Our first results showed high specificity which should be used with caution, as the number of patients was relatively low. Acquisitions starting 2 hours after injection are optimal and suffice for imaging. Further studies are needed to explore and determine the role of somatostatin receptor scintigraphy in clinical practice.
This research made use of the SMILE medical imaging laboratory at Karolinska University Hospital, Stockholm, Sweden.