|
|
||||||||
Nuclear Medicine |
-iodovinyl Estradiol Scintigraphy1
1 From the Departments of Nuclear Medicine (R.J.B., L.J.R., G.W.S.), Radiotherapy (G.v.T.), and Pathology (L.A.N.), Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands; and Nycomed Amersham Cygne and Eindhoven University of Technology, Eindhoven, the Netherlands (L.J.R., A.G.J.). From the 2000 RSNA scientific assembly. Received October 10, 2000; revision requested November 25; final revision received March 30, 2001; accepted April 4. Address correspondence to R.J.B. (e-mail: r.bennink@amc.uva.nl).
| ABSTRACT |
|---|
|
|
|---|
-iodovinyl estradiol (Z-MIVE) scintigraphy for the detection of estrogen receptors in patients with primary breast carcinoma. MATERIALS AND METHODS: In 22 patients, estrogen receptor status was assessed with planar scintigraphy and single photon emission computed tomography (SPECT) 4 hours after the injection of 185 MBq 123I-labeled Z-MIVE. For histologic and estrogen receptor immunohistochemical analysis, breast carcinoma tissue was obtained in all patients by means of biopsy or resection of the primary tumor. Two experienced physicians semiquantitatively scored the scintigraphic and immunohistochemical findings. The uptake ratio at scintigraphy and the immunohistologic staining intensity were scored as negative, weak, intermediate, or strong.
RESULTS: All patients had histologically proven breast cancer. Immunohistologic staining for estrogen receptors yielded negative findings in four patients and positive findings in 18 (weak staining, n = 2; intermediate staining, n = 6; strong staining, n = 10). Planar 123I-labeled Z-MIVE scintigraphic findings were negative in five patients and positive in 17 (weak uptake, n = 2; intermediate uptake, n = 10; strong uptake, n = 5), resulting in one false-negative finding. Findings at 123I-labeled Z-MIVE SPECT were negative in four patients and positive in 18. The sensitivities of 123I-labeled Z-MIVE scintigraphy for estrogen receptors were 100% with SPECT and 94% with planar scintigraphy. The correlation between immunohistologic and planar scintigraphic scores of estrogen receptor status was 0.72 (P < .01).
CONCLUSION: 123I-labeled Z-MIVE scintigraphy is a sensitive noninvasive tool for the detection of estrogen receptors in patients with breast cancer.
Index terms: Breast neoplasms, radionuclide studies, 00.12161, 00.32 Breast neoplasms, SPECT, 00.12162, 00.32 Hormones
| INTRODUCTION |
|---|
|
|
|---|
To overcome the dilemma of treating patients with breast cancer, especially those with metastatic disease, when knowledge of local or metastatic estrogen receptor expression is limited, several studies (917) were performed to investigate the possibility of in vivo imaging of estrogen receptors with receptor-specific radioligands at positron emission tomography (PET) or single photon emission computed tomography (SPECT). We reported (18,19) on an iodine 123 (123I)labeled estrogen receptor ligand, cis-11ß-methoxy-17
-iodovinyl estradiol (Z-MIVE), which shows high binding affinity for estrogen receptors both in rat and human mammary-tumor tissue preparations. A biodistribution study (20) revealed an acceptable effective dose equivalent for the amount of 123I-labeled Z-MIVE required for imaging of estrogen receptors in primary and metastatic breast carcinoma. In a clinical pilot study (21) that focused on the feasibility of 123I-labeled Z-MIVE scintigraphy, the preliminary results indicated that 123I-labeled Z-MIVE findings show good agreement with estrogen receptor immunohistochemical findings.
The purpose of our study was to determine the sensitivity of 123I-labeled Z-MIVE scintigraphy in the detection of estrogen receptors in a group of patients with primary breast carcinoma.
| MATERIALS AND METHODS |
|---|
|
|
|---|
|
-tri-n-butylstannylvinylestradiol precursor, as previously described (19). For the oxidative radio-iododestannylation, [123I]sodium iodide in methanol, which was produced with the indirect proton reaction on 99.8% enriched xenon 124 (22), was obtained from Nycomed Amersham Cygne. The specific activity of 123I-labeled Z-MIVE was 185 TBq/mmol or greater. The radionuclidic purity was greater than 99.9% and was determined by using a germanium semiconductor detector (Canberra Industries, Greenwich, Conn) with a multichannel analyzer connected to a gamma spectrum analyzing computer. The radiolabeled product, which was dissolved in ethanol, was diluted with 0.9% saline to a maximal ethanol concentration of 10%. Prior to its use, the product was checked at thin-layer chromatography for the absence of free radioactive iodide. The amount of free iodide was 2% or less. The final product was filtered through a sterile 0.22-µm membrane (Millex GV; Millipore, Milford, Mass).
Imaging Studies
All patients received approximately 300 mg of potassium iodide orally, in three doses daily during 2 days, to block thyroid uptake of free radioactive iodide. Imaging was performed after a single intravenous injection of approximately 185 MBq 123I-labeled Z-MIVE in the arm opposite the side of the known breast lesion to avoid any false-positive uptake in the axillary lymph nodes. Patients were positioned supine, and when the thoracic region was imaged, the arms were placed alongside the head. Planar acquisitions were performed by using a dual-headed gamma camera (Body Scan; Siemens Medical Systems, Hoffman Estates, Ill) interfaced to a Sun (Sun Microsystems, Mountain View, Calif) workstation (Hermes; Nuclear Diagnostics, Stockholm, Sweden). SPECT was performed with a three-headed SPECT camera (Siemens Medical Systems). With both cameras, a medium-energy collimator was used, with the energy peak centered at 159 keV and a 15% window.
In all patients, whole-body images were obtained 4 hours after injection. Whole-body images were simultaneously acquired in anterior and posterior planes (scanning speed, 10 cm/min; matrix, 256 x 512 pixel). If necessary, additional spot planar images were acquired for 10 minutes in a 256 x 256pixel matrix. SPECT data were acquired immediately after planar scintigraphy during 15 minutes in 20 views (45 seconds per detector; 60 angles; 6°/angle; matrix, 64 x 64 pixels). Transverse, coronal, and sagittal sections (thickness, 7.13 mm) were reconstructed by using Wiener-filter preprocessing. This was followed by filtered backprojection on the Sun workstation.
Data Analysis
Two experienced nuclear medicine physicians (R.J.B., G.W.S.) independently evaluated all images on the basis of the normal distribution of 123I-labeled Z-MIVE in healthy volunteers (20). Only one (G.W.S.) of the observers was blinded to the clinical information and the findings with the other imaging modalities, except for the notion of primary breast cancer. Both observers were blinded with respect to the estrogen receptor status.
Planar and tomographic images were reviewed for the presence of focally increased uptake in the breast. Tracer uptake was semiquantitatively scored on planar images as follows: 0 = negative, 1 = weak, 2 = intermediate, and 3 = strong. Disagreements about uptake intensity were resolved by consensus.
Quantification of abnormal 123I-labeled Z-MIVE uptake seen on the whole-body and/or spot planar images was performed by using the region-of-interest method. One observer (R.J.B.) drew regions of interest over the entire area with increased 123I-labeled Z-MIVE uptake (ie, tumor). Nonspecific uptake was estimated by drawing a similar region of interest in the contralateral breast or, if this was not feasible, in the surrounding normal-appearing breast tissue. For each region of interest, the average counts per pixel were calculated on only the anterior images of the primary breast carcinoma. For each tumor, 123I-labeled Z-MIVE uptake was quantified as the ratio of total tumor uptake to nonspecific uptake.
In Vitro Estrogen Receptor Determination
Estrogen receptor immunohistochemical analysis was performed essentially as described by Sannino and Shousha (23). Briefly, sections were preincubated for 15 minutes with 10% normal goat serum in phosphate-buffered saline. After incubation with the primary antibody 1D5 (Dako; Patts, Glostrup, Denmark), the secondary biotinylated immunoglobulin, rabbit anti-mouse (1:200 dilution; Dako), was applied for 30 minutes. Subsequently, after incubation with horseradish peroxidaseconjugated streptavidin complex (1:200 dilution; Dako), a solution of 0.1% 3,3-diaminobenzidine (Sigma, St Louis, Mo) and 0.01% hydrogen peroxide was used to develop the peroxidase activity. A previously identified strongly estrogen receptorpositive tumor was used as a positive control. Negative controls were derived by omitting the primary antibody. Staining intensity was assessed as negative, weak, intermediate, or strong.
Statistical Evaluation
A Kruskal-Wallis test was used to compare differences between uptake ratios in immunohistochemically defined estrogen receptor subgroups. A Cuzick Wilcoxon-type test for trend was used to demonstrate the relation between the subgroups and the uptake ratio. The Spearman rank test corrected for ties was used to analyze the correlation between the subgroups and visual scintigraphic uptake scores. Differences were considered significant when the P value was less than .05.
| RESULTS |
|---|
|
|
|---|
|
|
|
|
|
|
| DISCUSSION |
|---|
|
|
|---|
A clinical pilot study (21) performed with 123I-labeled Z-MIVE had promising results for noninvasive in vivo evaluation of estrogen receptor status in both primary and metastatic breast cancer. Based on these results, this study had the aim of determining the sensitivity of 123I-labeled Z-MIVE scintigraphy for the detection of estrogen receptors in a larger group of patients with primary breast carcinoma.
In our series, 82% of the patients had immunohistochemically proven estrogen receptors. This percentage is somewhat higher than the 60%70% reported in the literature (4). However, statistical evaluation did not reveal a significant difference between this finding in our study population and the reported incidence of estrogen receptor positivity in breast cancer. For 17 of 18 breast tumors that were estrogen receptor positive at immunohistochemical staining, whole-body or spot planar scintigrams showed detectable 123I-labeled Z-MIVE uptake. At SPECT, the estrogen receptor status of all lesions was identified correctly, increasing the sensitivity from 94% to 100%. This high sensitivity of 123I-labeled Z-MIVE scintigraphy confirms previous preliminary observations (21), and with the good agreement regarding estrogen receptor status at scintigraphy and immunohistochemical staining, it demonstrates the potency of 123I-labeled Z-MIVE scintigraphy for the detection and characterization of estrogen receptorpositive breast tumors.
Scintigraphic detection of estrogen receptorpositive tumors will largely depend on tumor size and receptor density, among other factors. The smallest estrogen receptorpositive tumor in our series, a T1 tumor, showed strong immunohistochemical staining. This tumor (in patient 8) showed clear 123I-labeled Z-MIVE uptake. In this study, the use of SPECT further improved contrast and lesion localization, with the identification of one tumor (in patient 20) that was missed at planar scintigraphy (Fig 4). This T2 tumor showed weak immunohistochemical estrogen receptor staining. False-negative cases might be expected in small tumors exhibiting weak estrogen receptor positivity. However, for those small tumors (ie, T1 breast cancer with a low likelihood of distant metastatic disease), surgery is the primary treatment. Therefore, antiestrogen treatment is not indicated.
Although the correlation between the immunohistochemical and visual scintigraphic scores was high, 123I-labeled Z-MIVE scintigraphy seemed to cause underestimation of estrogen receptor positivity in our series. This finding can be explained by the fact that estrogen receptor staining within the tumor was often inhomogeneous; this finding is consistent with findings in the literature (5). Tissue attenuation of 123I-labeled Z-MIVE radioactivity is another explanation for the underestimation of estrogen receptormediated 123I-labeled Z-MIVE uptake. Since a clear distinction between estrogen receptornegative and estrogen receptorpositive lesions could be made at scintigraphy, corrections for attenuation were not attempted. The calculation of uptake ratios confirmed the visual interpretation of the images. Since the difference in uptake ratios between estrogen receptornegative and estrogen receptorpositive lesions was significant, a difference in the uptake of a lesion of more than 10% compared with the uptake in a mirrored region of interest in the contralateral breast (ratio, 1.1) can be considered positive. Since count density on SPECT sections is subject to multiple factors, such as reconstruction filters and parameters, depth dependency, and nonuniform attenuation, SPECT uptake ratios were not calculated, although SPECT improved contrast and lesion detection.
123I-labeled Z-MIVE scintigraphy had no false-positive results in this trial, yielding optimal specificity. However, with the relatively low number of patients with estrogen receptornegative tumors, this finding should not be regarded as representative. Estrogen receptor specificity was previously illustrated with almost complete blockade of tumor uptake in patients treated with antiestrogen (tamoxifen citrate) therapy (21) and with blocking experiments with diethylstilbestrol in estrogen receptorpositive tumor-bearing rats (19).
In view of the optimal selection of patients for therapy to block estrogen receptors, it is important to characterize not only the primary tumor but also the possible metastases, since the estrogen receptor status may be different between the primary tumor and the metastases or among the metastases (6,7). Scintigraphic and histopathologic correlation of metastases is often more difficult, since metastases can be difficult to reach for biopsy. In the present study, we confirmed a high sensitivity for estrogen receptors in primary breast neoplasms. Taking into account the observations reported previously (21), we are confident to report the estrogen status of known metastases of estrogen statuspositive breast carcinoma, which aids the optimal treatment and follow-up of breast carcinoma patients.
PET with [fluorine 18(18F)]fluoroestradiol also has been shown to have a high sensitivity and no false-positive cases for the detection of estrogen receptorpositive primary, as well as metastatic, human breast cancer (10,11). Agreement rates between the results of [18F]fluoroestradiol PET and estrogen receptor assays in primary and metastatic lesions have been reported (12) to be 82% and 94%, respectively.
Recently, [18F]fluoroestradiol PET demonstrated the heterogeneity of estrogen receptor expression in metastatic breast cancer, again showing the potential use of estrogen receptor scintigraphy as a noninvasive tool (25). Although [18F]fluoroestradiol PET is a sensitive imaging method for the detection of estrogen receptorpositive breast cancer, its application in daily clinical situations is hindered by its more limited availability and high costs. Therefore, a conventional nuclear medicine imaging procedure with SPECT could have a role in the work-up of patients with breast carcinoma.
In conclusion, 123I-labeled Z-MIVE scintigraphy has a high sensitivity for the detection of estrogen receptorpositive primary breast carcinoma. The high correlation between the amount of 123I-labeled Z-MIVE uptake and the level of estrogen receptor expression indicates that conventional scintigraphy is a reliable, noninvasive tool for the assessment of the estrogen receptor status in breast carcinoma. Knowledge of this status may facilitate the choice of systemic therapy (hormonal or cytotoxic) in many patients with breast cancer.
| FOOTNOTES |
|---|
-iodovinyl estradiol Author contributions: Guarantors of integrity of entire study, R.J.B., L.J.R., G.W.S.; study concepts, L.J.R., G.v.T., A.G.J.; study design, R.J.B., G.W.S., L.J.R.; literature research, R.J.B., L.J.R.; clinical studies, R.J.B., L.J.R.; data acquisition, R.J.B., L.J.R., G.W.S.; data analysis/interpretation, R.J.B., G.W.S., L.A.N.; statistical analysis, R.J.B., G.W.S.; manuscript preparation and definition of intellectual content, R.J.B., G.W.S., L.J.R.; manuscript editing, revision/review, and final version approval, all authors.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
D. A. Mankoff, J. M. Link, H. M. Linden, L. Sundararajan, and K. A. Krohn Tumor Receptor Imaging J. Nucl. Med., June 1, 2008; 49(Suppl_2): 149S - 163S. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Van Den Bossche and C. Van de Wiele Receptor Imaging in Oncology by Means of Nuclear Medicine: Current Status J. Clin. Oncol., September 1, 2004; 22(17): 3593 - 3607. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. J. Bennink, G. van Tienhoven, L. J. Rijks, A. L. Noorduyn, A. G. Janssen, and G. W. Sloof In Vivo Prediction of Response to Antiestrogen Treatment in Estrogen Receptor-Positive Breast Cancer J. Nucl. Med., January 1, 2004; 45(1): 1 - 7. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |