FROM THE DEPARTMENTS OF DIAGNOSTIC RADIOLOGY, UROLOGY AND RADIATION PHYSICS, UNIVERSITY HOSPITAL, S-90185 UMEA, SWEDEN.

MAGNETIC RESONANCE IMAGING OF RENAL MASSES AT 0.02 TESLA

L. EKELUND, S . - 0 . HIETALA, B. LJUNGBERG, H. LOTZ and G. WICKMAN

Abstract To determine the usefulness of extremely low field magnetic

resonance imaging (MRI) in the evaluation of renal masses 19 patients with 15 tumors (13 renal and 2 renal pelvic carcinomas) and 8 cysts were examined in a 0.02 tesla MRI unit. The findings were compared with results of computed tomography (CT) and ultrasound. Cavography was performed in 6 patients. MRI en- abled differentiation between cysts and solid tumors. Tumor extension into the inferior vena cava could be demonstrated in one case and liver metastases in two patients. The image quality was inferior to that reported at higher field strengths and the tumors were more precisely staged by CT and ultrasound. At present, it is unlikely that low field MRI will play any substantial role in the evaluation of renal masses.

Key words: Kidney neoplasms, MR diagnosis; MRI, low field strength.

There have been several reports on magnetic resonance imaging (MRI) of the kidneys during the last few years (1-10). Advantages of MRI include differentiation of solid masses from benign cystic lesions, and identification of major blood vessels and vascular invasion without admin- istration of contrast media. The experience so far has been limited to high and intermediate field systems and no reports have dealt with the results of MRI at extremely low field strengths (0.02 tesla). We report on a comparison of low field MRI with other imaging modalities for evalu- ating renal masses.

Material and Methods

Nineteen patients, 12 men and 7 women, with a mean age of 62 years (racge 39-76 years) were examined. Thir- teen had renal cell carcinoma and 2 carcinoma of the renal pelvis. Two patients with renal carcinoma also had renal cysts. Four patients had simple renal cysts. Bilateral cysts occurred in two patients.

MR images were obtained in a 0.02 tesla (T) magnetic field (Acutscan, Instrumentarium Corp., Finland). All pa-

tients were studied in a close-coupled body coil with a saturation-recovery (SR) pulse-sequence utilizing gradient reversal echo generation. The echo time could be varied from 40 to 80 ms.

The shortest repetition time (TR 250 ms) and echo time available (TE 40 ms) were always utilized. This sequence creates T1 weighted images with the most favourable signal to noise ratio. Most of the patients had a second SR scan with a longer repetition time (TR 1000 ms) and echo time (TE 80 ms) in order to enhance the T2 differentiation. In 4 patients additional images were obtained using the spin-echo (SE) sequence. The SE pulse sequence is a conventional Can-Purcell sequence by which the echo time can be varied from 100 to 300 ms. Potential improve- ments in the images due to a better T2 separation by this sequence were, however, counter-weighted by an unfa- vourable acquisition time and signal to noise ratio.

The slice thickness was 10 mm and coronal and trans- verse imaging planes were routinely used. The total ex- amination time varied between 45 and 90 min and typical acquisition times were 10 to 15 min.

The results of the MR studies were correlated to those of computed tomography (CT) (Siemens Somatom 2) in 14 and to those of ultrasonography in all 15 patients with renal tumors. Cavography was performed in 6 patients.

All diagnoses were confirmed histopathologically, ex- cept for renal cysts, for which ultrasound or CT findings were accepted as diagnostic.

Results

Renal cysts (2-10 cm in diameter) were identified on T1 contrast scans as well demarcated, smooth lesions of uniform low signal intensity in contrast to surrounding

Accepted for publication 19 August 1987.

61

Act

a R

adio

l Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

QU

T Q

ueen

slan

d U

nive

rsity

of

Tec

h on

10/

31/1

4Fo

r pe

rson

al u

se o

nly.

62 L. EKELUND, S . -0 . HIETALA, B. LJUNGBERG, H. LOTZ AND G . WICKMAN

a

Fig. 1. Simple cyst of the left kidney. Transverse (a) and coronal (b) SR 250140 MR images show round mass of homogeneous low signal intensity.

a

Fig. 2. a) Contrast enhanced CT scan demonstrating a partly necrotic carcinoma posteriorly in the right kidney. Normal renal veins are seen entering the inferior vena cava. b) Transverse SR 250140 MR image through the same level showing the tumor as an ill-defined mass of intermediate signal intensity. The renal artery can be seen. c) On a T2 weighted image (SR 1000180) there is an increase in signal intensity within the tumor and in normal renal parenchyma.

renal parenchyma (Fig. 1). With prolonged TR and TE values an increase in signal intensity was seen.

In the 15 patients with renal or renal pelvic carcinomas tumor size ranged between 5 and 15 cm in diameter. Abnormal findings were detected by MRI in all but one patient, who did not manage to go through a complete examination. Most neoplasms appeared as lesions that deformed the renal contour and demonstrated a varying signal intensity pattern. With short TR, tumor signal in- tensity ranged from low to intermediate; on T2 contrast scans tumor signal intensity showed a varying degree of increase (Fig. 2). Saturation recovery scans were found to be more informative than spin-echo images.

In one patient with a necrotic carcinoma, CT demon-

strated a thick-walled cystic lesion, suspected to be a carcinoma. At MRI the mass had an inhomogeneous sig- nal intensity, reinforcing the diagnosis of malignancy (Fig. 3).

When CT and MRI were compared in evaluation of tumor extension CT was found superior in delineating tumor masses, in revealing adenopathy, and for assessing the perirenal and pararenal tissues (Fig. 4). The shortcom- ings of MRI were felt to depend on poor spatial resolution and motion artefacts due to long imaging times.

One patient had tumor invasion of the renal vein and the inferior vena cava as shown by both ultrasound and CT and verified at cavography and subsequent surgery. In 7 patients it was possible to see the inferior vena cava on

Act

a R

adio

l Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

QU

T Q

ueen

slan

d U

nive

rsity

of

Tec

h on

10/

31/1

4Fo

r pe

rson

al u

se o

nly.

MAGNETIC RESONANCE IMAGING OF RENAL MASSES AT 0.02 TESLA 63

a b

Fig. 3. a) Contrast enhanced CT scan showing a cystic lesion in a h - I

Fig. 5. a) Coronal SR 1000/80 MR image shows areas of in- creased signal intensity within the inferior vena cava indicating tumor extension (+). b) The tumor thrombus is well delineated at cavography.

the upper pole of the right kidney with a thick, partly calcified wall. No contrast enhancement was seen in the lesion which had an attenuation value of 32 HU. b) Transverse SR 1ooo/80 MR image through the same level. The calcified wall can be seen as a rim of low signal intensity. The contents of the cystic lesion have - an inhomogeneous signal intensity, indicating that this is not a simple cyst. Histopathologic examination of the resected speci- men showed necrotic renal carcinoma.

a b Fig. 4. a) Contrast enhanced CT scan showing large, partly necrotic carcinoma of the left kidney with considerable extension medially. b) Transverse SR 1000/80 MR image at the same level. Inhomogeneous signal intensity within the tumor. The margins of the tumor are not as well appreciated as on the corresponding CT scan.

MR images, and the tumor thrombus that was present in one of these patients could be demonstrated on coronal images (Fig. 5).

Three patients had liver metastases as demonstrated by CT and ultrasound. On MR images the liver parenchyma was usually poorly defined and the structures difficult to identify. However, in two patients it was possible to distinguish the hepatic metastases as lesions of increased signal intensity on both transverse and coronal images (Fig. 6).

Discussion

MRI has demonstrated its ability to display normal renal parenchyma and to differentiate between cystic and solid renal masses (1, 4, 5-10). HRICAK et coll. (3), using a superconductive 0.5 T magnet operating at 0.35 T, cor- rectly staged 26 of 27 tumors. Of particular interest was

a b

Fig. 6. Same patient as in Fig. 5. a) Non-enhanced CT scan showing multiple liver metastases. b) Coronal SR 1000/80 MR image demonstrating bright tumor in the lateral portion of the left kidney. One of the hepatic metastases is seen as an area of high signal intensity in the right lobe of the liver.

the ability of MRI to demonstrate vascular structures without the need for contrast enhancement, and MRI was found more sensitive than CT for determining the extent of tumor thrombus and for showing invasion of the wall of the inferior vena cava. However, tumor invasion of the wall of the inferior vena cava could not be confirmed in a recent report by FEIN et coll. (2), who compared MRI and CT in diagnosing and staging 27 patients with proven renal carcinoma. They used a 0.5 T magnet and found MRI to be similar to CT in the staging procedure.

KARSTAEDT et coll. (5) examined 34 patients with renal masses (25 with renal carcinoma) with a 0.15 T resistive magnet and compared the results with those of ultra- sound, CT, angiography and phlebography. Abnormalities were detected by MRI in all patients. Differentiation be- tween solid and cystic lesions was made with MRI, ultra- sound and CT. Tumor invasion of the renal vein and the inferior vena cava was demonstrated in 7 patients by MRI, ultrasound, CT and phlebography.

Despite these promising results, MRI has not yet been established as an indispensible, commonly available and

Act

a R

adio

l Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

QU

T Q

ueen

slan

d U

nive

rsity

of

Tec

h on

10/

31/1

4Fo

r pe

rson

al u

se o

nly.

64 L. EKELUND, S . - 0 . HIETALA, B. LJUNGBERG, H. LOTZ AND G. WICKMAN

cost-effective technique in the evaluation of renal masses as it has in neuroradiology. Accordingly, it was recently concluded that ultrasound and CT should continue to be the primary imaging modalities for the evaluation of renal masses because of superior spatial resolution, wider avail- ability, and lower cost (8).

There are no reports available about the application of extremely low field MRI in the imaging of renal masses. If this technique could provide information comparable to that obtained from systems with higher field strengths, this would have important implications from a cost-bene- fit point of view as very low field MRI units are cheaper than current CT scanners and the annual operational costs are much lower than for MR equipment with higher field strength.

With our low field MRI unit it was possible to differenti- ate cysts from tumors and we were able to demonstrate 14 out of 15 renal neoplasms. In principle the MRI features with this system were the same as those reported for units with higher magnetic field strengths (1-10). We also found a potential for the demonstration of tumor extension into the inferior vena cava as well as for the detection of liver metastases. Disadvantages of low field MRI include poor spatial resolution, long imaging times and motion arte- facts, resulting in an image quality definitely surpassed by that obtained at higher field strengths. The signal to noise ratio (SNR) of very low field MRI, though relatively low, is sufficient for the detection of abnormal signals from a variety of pathologic lesions. An increased SNR might provide a more esthetic image but is not necessarily more efficacious for diagnosing the lesion itself. On the other hand a low SNR reduces the possibility for obtaining more detailed morphologic information, e.g. about the extent of a renal carcinoma and possible invasion of adjacent or- gans. In our material most renal tumors were more ade- quately staged by both ultrasound and CT. Low field MRI, when available, may be used as a supplement in equivocal cases, i.e. where caval vein invasion is suspect-

ed, but not proven, or for further evaluation of mass lesions appearing as complex cysts at ultrasound or CT imaging.

ACKNOWLEDGEMENT This investigation was supported by grants from Lion’s Re-

search Foundation, University of Umel, Sweden.

Request for reprints: Professor Leif Ekelund, Department of Diagnostic Radiology, University Hospital, S-90185 Umel, Swe- den.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

REFERENCES CHOYKE P. L., KRESSEL H. Y ., POLLACK H. M., ARGER P. M., AXEL L. and MAMOURIAN A. C.: Focal renal masses. Magnet- ic resonance imaging. Radiology 152 (1984), 471. FEIN A. B., LEE J. K. T., BALFE D. M. et coll.: Diagnosis and staging of renal cell carcinoma. A comparison of MR imaging and CT. Amer. J. Roentgenol. 148 (1987), 749. HRICAK H., DEMAS B. E., WILLIAMS R. D. et coll: Magnetic resonance imaging in the diagnosis and staging of renal and penrenal neoplasms. Radiology 154 (1989, 709. - and NEWHOUSE J. H.: MR imaging of the kidney. Radiol. Clin. N. Amer. 22 (1984), 287. KARSTAEDT N., MCCULLOUGH D. L., WOLFMAN N. T. and DYERR. B.: Magnetic resonance imaging of the renal mass. J . Urol. 136 (1986), 566. KULKARNI M. V., SHAFF M. I., SANDLER M. P. et coll.: Evaluation of renal masses by MR imaging. J. Comput. Assist. Tomogr. 8 (1984), 861. LIPUMA J. P.: Magnetic resonance imaging of the kidney. Radiol. Clin. N. Amer. 22 (1984), 925. MAROTTI M., HRICAK H., FRITZSCHE P., CROOKS L. E., HEDG- COCK M. W. and TANAGHO E. A.: Complex and simple renal cysts. Comparative evaluation with MR imaging. Radiology 162 (1987), 679. NEWHOUSE J. H., MARKEZ J. A. and KAZAM E.: Magnetic resonance imaging of the kidneys, Cardiovasc. Intervent. Radiol. 8 (1986), 351. PAPANICOLAOU N., HAHN P. F., EDELMAN R. R. et coll.: Magnetic resonance imaging of the kidney. Urol. Radiol. 8 (1986), 139.

Act

a R

adio

l Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

QU

T Q

ueen

slan

d U

nive

rsity

of

Tec

h on

10/

31/1

4Fo

r pe

rson

al u

se o

nly.