AJR 2005; 184:574-578
© American Roentgen Ray Society
High-Frequency Contrast Harmonic Imaging of Ophthalmic Tumor Perfusion
Klaus Schlottmann1,
Barbara Fuchs-Koelwel2,
Martina Demmler-Hackenberg2,
Veit-Peter Gabel2 and
Jürgen Schölmerich1
1 Department of Internal Medicine I and Interdisciplinary Department of
Ultrasound, University Hospital of Regensburg, FJS-Allee 11, Regensburg,
Germany.
2 Department of Ophthalmology, University Hospital of Regensburg, Regensburg,
Germany.
Received April 14, 2004;
accepted after revision June 2, 2004.
Presented in part at the 2004 European Symposium on Contrast Ultrasound
Imaging, Rotterdam, The Netherlands.
Address correspondence to K. Schlottmann.
Introduction
The evaluation of symptomatic or incidentally diagnosed ophthalmic tumors
usually is performed using funduscopy and A- and B-mode sonography. Additional
techniques are Doppler modes
[1], fluorescence or
indocyanine green angiography
[2], CT
[3], and MR tomography
[4]. A- and B-mode sonography
allows the measurement of tumor depth and size
[5], and A-mode and Doppler
sonography estimates tumor vascularization and perfusion
[6]. Histologic data indicate
the importance of tumor vascularization as a determinant of the biologic
behavior and the response to radiation therapy of choroidal melanoma. Hence,
the analysis of tumor perfusion is crucial to discriminate solid from liquid
or degenerative lesions and for the evaluation of response of choroidal
melanoma to radiation therapy
[7]. Uveal melanoma vascularity
also is correlated to its metastatic potential
[8].
A new sonographic technique, contrast harmonic imaging, uses
IV-administered gas-filled microbubbles that remain intravascular after IV
injection [9]. The nonlinear
harmonic backscatter signals from such resonating microbubbles now can be
visualized continuously at very low transmit power (low mechanical index). We
have adapted contrast harmonic imaging at a low mechanical index using the
sonographic contrast agent BR1 and a high-frequency transducer to analyze
perfusion of ophthalmic tumors. In this feasibility study, we show the
analysis of perfusion of choroidal tumors is possible on contrast harmonic
imaging, even when other sonographic techniques fail to detect tumor
perfusion.
Subjects and Methods
From July 2002 to August 2003, 16 patients with choroidal melanomas (Figs.
1A,
1B,
1C,
1D,
1E,
1F and
2A,
2B) and four patients with
choroidal metastases were included in our study after giving informed consent.
The study was approved by the local ethics committee of the University
Hospital of Regensburg.
View larger version (94K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —35-year-old woman with progressive loss of vision. B-mode
(A) and power Doppler (B) sonograms show typical choroidal
melanoma. Several vessels can be detected on power Doppler sonogram (level
4).
|
|
View larger version (30K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —35-year-old woman with progressive loss of vision. B-mode
(A) and power Doppler (B) sonograms show typical choroidal
melanoma. Several vessels can be detected on power Doppler sonogram (level
4).
|
|
View larger version (65K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —35-year-old woman with progressive loss of vision. Contrast
harmonic images show time course of tumor perfusion at 0 (C), 10
(D), 33 (E), and 50 (F) sec after injection of contrast
agent. Entire lesion is filled with bubbles, which is representative of
hyperperfusion of melanoma (level 4). Fifty seconds after injection of
contrast bolus, decline of bubble signals in melanoma is evident.
|
|
View larger version (68K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D. —35-year-old woman with progressive loss of vision. Contrast
harmonic images show time course of tumor perfusion at 0 (C), 10
(D), 33 (E), and 50 (F) sec after injection of contrast
agent. Entire lesion is filled with bubbles, which is representative of
hyperperfusion of melanoma (level 4). Fifty seconds after injection of
contrast bolus, decline of bubble signals in melanoma is evident.
|
|
View larger version (75K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1E. —35-year-old woman with progressive loss of vision. Contrast
harmonic images show time course of tumor perfusion at 0 (C), 10
(D), 33 (E), and 50 (F) sec after injection of contrast
agent. Entire lesion is filled with bubbles, which is representative of
hyperperfusion of melanoma (level 4). Fifty seconds after injection of
contrast bolus, decline of bubble signals in melanoma is evident.
|
|
View larger version (69K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1F. —35-year-old woman with progressive loss of vision. Contrast
harmonic images show time course of tumor perfusion at 0 (C), 10
(D), 33 (E), and 50 (F) sec after injection of contrast
agent. Entire lesion is filled with bubbles, which is representative of
hyperperfusion of melanoma (level 4). Fifty seconds after injection of
contrast bolus, decline of bubble signals in melanoma is evident.
|
|
View larger version (103K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —73-year-old woman with choroidal melanoma that was irradiated
before contrast harmonic imaging. Contrast harmonic image at 23 sec after
contrast injection shows lesion that is completely filled with bubble signals
(level 4).
|
|
In the 16 patients with melanomas, 18 sonographic investigations were
performed: 10 naïve melanomas and eight pretreated by radiation therapy.
Two patients were examined before and after radiation therapy. Of the four
patients with choroidal metastases, two had metastasized breast carcinoma, and
two had metastasized bronchial carcinoma. All patients with choroidal
metastases also showed metastases in other locations.
Choroidal melanomas were diagnosed using funduscopy and A-mode sonography.
In one patient with melanoma, histology was obtained by enucleation. For
funduscopic analysis, the following criteria were used to define the diagnosis
of melanoma: pigmentation, tumor height, orange pigment accompanying serous
retinal detachment, localization, secondary glaucoma, cataract, and uveitis.
On A-mode sonography, the diagnostic criteria were shape, prominence,
reflectivity, and internal structure. The criterion for the diagnosis of
melanoma on B-mode sonography was tumor of the posterior pole, and it was
level 2–4 perfusion at maximum sensitivity for power Doppler signal
detection on power Doppler sonography.
Contrast sonography was performed in all patients after an initial
investigation that included funduscopy and A-mode sonography in the outpatient
clinic of the department of ophthalmology.
We used a sonography machine (Elegra, Siemens) with a 7.5-MHz transducer
(7.5L40, Siemens) adapted to contrast harmonic imaging conditions at a low
mechanical index. All patients were evaluated with standard settings; B-mode
sonography was performed with or without tissue harmonic imaging. We did not
use a mechanical index higher than 0.8, and continuous insonation of one
region of the bulbus for longer than 10 sec was avoided. Power Doppler signal
was regulated for the most sensitive visualization of tumor vessels without
aliasing at low persistence. Contrast harmonic imaging was set to a low
mechanical index of 0.1–0.2, the receive amplification was set to 60 dB,
and the transmit center frequency was set to 3.3 MHz. These contrast harmonic
imaging–specific settings proved to give sufficient microbubble signals
with the 7.5-MHz transducer.
After the maximum tumor size (thickness and diameter) was measured, power
Doppler sonography was performed. The image with maximum vessel intensity on
power Doppler sonography was analyzed later to quantify vascularization. The
sonographic contrast agent BR1 (SonoVue [sulfur hexafluoride], Bracco) was
prepared according to the manufacturer's recommendations. For each
investigation, 4.8 mL of BR1 was injected IV as a bolus within 3–5 sec
followed by a bolus injection of 10 mL of saline.
The investigator who performed B-mode, power Doppler, and contrast harmonic
imaging was blinded to the ophthalmologists' diagnosis. The entire
investigation was recorded on VHS videotapes, and additional digital pictures
were stored on the hard disk for later analysis. Both power Doppler signals
and bubble signals were quantified arbitrarily by division into five levels of
intensity (Table 1). After the
investigations, one of the authors proposed the most likely diagnosis from the
information gained by the sonographic investigations. After the
investigations, two experienced sonographers analyzed the videos and assigned
the power Doppler and contrast harmonic imaging data of tumor perfusion to the
respective levels. All patients were monitored for retinal changes after the
contrast harmonic imaging investigation. No bleeding or other type of
complication was detected.
Results
From September 2002 until July 2003, a total of 16 choroidal melanomas and
four metastases were investigated. Ten patients presented with untreated
melanomas, and eight patients were treated by radiation therapy, two of whom
were examined before and after radiation therapy. One melanoma could not be
identified on funduscopy because of its peripheral location, and one tumor
could not be analyzed on A-mode sonography. In contrast, all lesions were
visualized on B-mode sonography. The mean lesion thickness was 7.8 mm (range,
2.4–19.3 mm) for the un-treated melanomas, 3.8 mm (range, 1.7–9.8
mm) for the irradiated melanomas, and 3.5 mm (range, 2.8–4 mm) for the
metastases. The minimum tumor thickness to which power Doppler and contrast
harmonic imaging were allowed to apply proved to be 1.5 mm. Tumors with a
thickness of less than 1.5 mm were too thin to spatially project blood vessels
or particular microbubbles into the tumor. In such lesions, the signals
deriving from choroidal vessels could not be distinguished reliably from tumor
vessels.
As we expected, the B-mode images alone did not allow discrimination of
melanomas from metastases. Power Doppler signals tended to be at higher levels
in untreated melanomas than in irradiated melanomas or metastases
(Table 1). All untreated
melanomas showed level 2–4 power Doppler signals, whereas five of eight
irradiated melanomas were devoid of power Doppler signals and only two showed
level 4–5 vascularization. In two of eight irradiated melanomas, no
bubble signals were detected, but four of eight irradiated melanomas revealed
strong perfusion. None of the metastases was positive for power Doppler
signals, and all metastases proved to be perfused on contrast harmonic
imaging. However, only one metastasis from breast cancer was filled almost
completely with microbubbles (level 3), whereas the others revealed weak
perfusion with particular bubbles (level 1). All untreated melanomas were
perfused heavily on contrast harmonic imaging, with nine of 10 being almost or
completely filled with bubbles (levels 3 and 4). A single melanoma that was
less perfused (level 2) almost completely occupied the bulbus. In this
melanoma, tumor necrosis was most likely responsible for the incomplete
filling with contrast microbubbles.
In patients with level 2 or higher tumors, the microbubble inflow into the
tumor vessels occurred at a mean time of 21 sec (range, 11–34 sec) after
injection of BR1. The maximum bubble intensity in the tumors was reached after
26 sec (range, 14–42 sec). In lesions with level 1 signals, bubble
inflow and a signal maximum could not be defined because bubble signals were
sparse. Maximum intralesional vascular microbubble intensity lasted for only
several seconds, which can be explained by the high microbubble concentration
in the inflowing arterial blood after administration of the contrast agent as
an IV bolus. In all investigations, the microbubble concentration
significantly decreased over time. After a mean of 59 sec (range, 44–129
sec), only sparse signals in all lesion and ophthalmic vessels could be
detected.
Discussion
The aim of our study was to determine whether high-frequency contrast
harmonic imaging at a low mechanical index can be performed in the eye to
visualize contrast flow in choroidal tumor vessels. Furthermore, we
investigated whether irradiated melanomas are depleted of tumor vessels, which
is one of the major criteria used to estimate response to radiation
therapy.
We were able to show that the technique of contrast harmonic imaging with
BR1 at low mechanical index is feasible. Furthermore, we found that irradiated
choroidal melanomas in which no power Doppler signals were detected were still
massively perfused by contrast microbubbles. At present, we have no data that
support the hypothesis that these patients will develop local recurrence of
the choroidal melanomas. For this purpose, further studies are being
performed.
Recently, other groups have used sonographic contrast agents to increase
the sensitivity for the detection of ophthalmic tumor vessels and tumor
perfusion with Doppler techniques
[10–12].
The transmit energy in these studies was likely higher than a mechanical index
of more than 1. In all studies, the contrast agent SH U 508A (Levovist,
Schering) was used, which is insufficient for low-mechanical-index contrast
harmonic imaging because of the rigid shell. These bubbles are destroyed
immediately after insonation at a mechanical index greater than 0.7, which can
result in destruction of the vascular endothelium and extravasation
[13,
14]. However, no ophthalmic
side effects or complications were reported in the investigations
described.
In our study, we did not use BR1 to increase color Doppler signals for two
reasons. First, the transmitted energy is higher than that recommended by the
Center for Devices and Radiological Health of the U.S. Food and Drug
Administration (recommended mechanical index, 
0.23) when sonographic
contrast agents are used at a high mechanical index with color or spectral
Doppler sonography [15]. For
safety reasons, we exclusively used low-mechanical-index contrast harmonic
imaging in the eye with an mechanical index of 0.2 or less to avoid side
effects or complications. Because neither symptoms nor signs other than those
caused by the underlying disease were observed on funduscopy after the
sonographic investigation, it seems that the technique is safe. The second
reason is that we expected disadvantages of color Doppler sonography over
contrast harmonic imaging to increase the sensitivity to detect very small
vessels with low perfusion velocity and few intravascular corpuscles that
cause the Doppler shift. Contrast harmonic imaging visualizes single bubble
signals moving slowly through tumor vessels down to capillary size. Hence, as
we have shown, contrast harmonic imaging is currently the most sensitive
method with which to detect perfusion of choroidal tumors. Contrast harmonic
imaging seems to be helpful especially when ophthalmoscopic examinations fail
because of opaque ophthalmic structures such as the cornea, the lens, or the
vitreous.
Contrast harmonic imaging at a low mechanical index could become a powerful
tool with which to estimate ophthalmic melanoma response to radiation therapy
if irradiated melanomas highly positive for bubble signals prove to be
insufficiently irradiated.
References
- Vecsei PV, Kircher K, Nagel G, et al. Ocular arterial blood flow of
choroidal melanoma eyes before and after stereotactic radiotherapy using
Leksell gamma knife: 2 year follow up. Br J Ophthalmol1999; 83:1324
–1328[Abstract/Free Full Text]
- Shields CL, Shields JA, De Potter P. Patterns of indocyanine green
videoangiography of choroidal tumors. Br J Ophthalmol1995; 79:237
–245[Abstract/Free Full Text]
- Peyster RG, Augsburger JJ, Shields JA, et al. Choroidal melanoma:
comparison of CT, funduscopy, and CT. Radiology1985; 156:675
–680[Abstract/Free Full Text]
- Seiler T, Bende T, Schilling A, Wollensak J. Magnetic resonance
tomography in ophthalmology. I. Choroid melanoma [in German]. Klin
Monatsbl Augenheilkd1987; 191:203
–210[Medline]
- Ossoinig KC. Standardized echography: basic principles, clinical
applications, and results. Int Ophthalmol Clin1979; 19:127
–210[Medline]
- Tranquart F, Berges O, Koskas P, et al. Color Doppler imaging of
orbital vessels: personal experience and literature review. J Clin
Ultrasound 2003;3:258
–273
- Wolff-Kormann PG, Kormann BA, Riedel KG, et al. Quantitative color
Doppler imaging in un-treated and irradiated choroidal melanoma.
Invest Ophthalmol Vis Sci1992; 33:1928
–1933[Abstract/Free Full Text]
- Lassau N, Koscielny S, Avril MF, et al. Prognostic value of
angiogenesis evaluated with high-frequency and color Doppler sonography for
preoperative assessment of melanomas. AJR2002; 178:1547
–1551[Abstract/Free Full Text]
- Blomley MJ, Cooke JC, Unger EC, Monaghan MJ, Cosgrove DO.
Microbubble contrast agents: a new era in ultrasound.
BMJ 2001;322:1222
–1225[Free Full Text]
- Cennamo G, Rosa N, Vallone GF, Smaltino F. First experience with a
new echographic contrast agent. Br J Ophthalmol1994; 78:823
–826[Abstract/Free Full Text]
- Brabrand K, Kerty E, Jakobsen JA. Contrast-enhanced ultrasound
Doppler examination of the retrobulbar arteries. Acta
Radiol 2001;42:135
–139[Medline]
- Lemke AJ, Hosten N, Richter M, et al. Contrast-enhanced color
Doppler sonography of uveal melanomas. J Clin
Ultrasound 2001;29:205
–211[Medline]
- [No authors listed]. Section 4: bioeffects in tissues with gas
bodies—American Institute of Ultrasound in Medicine. J
Ultrasound Med 2000;19:97
–108[Abstract/Free Full Text]
- Wible JH Jr, Galen KP, Wojdyla JK, Hughes MS, Klibanov AL,
Brandenburger GH. Microbubbles induce renal hemorrhage when exposed to
diagnostic ultrasound in anesthetized rats. Ultrasound Med
Biol 2002;28:1535
–1546[Medline]
- Computed Imaging Devices Branch, Division of Reproductive,
Abdominal, Ear, Nose, and Throat and Radiological Devices, Office of Device
Evaluation. Information for manufacturers seeking market clearance
of diagnostic ultrasound systems and transducers. Washington, DC:
Center for Devices and Radiological Health, U.S. Food and Drug Administration,
September 30, 1997, pp.1
–74
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Introduction
Subjects and Methods
