497 CT Diagnosis of Renal Angiomyolipoma: The Importance of Detecting Small Amounts of Fat Morton A. Bosniak1 AlecJ. Megibow Donald H. Hulnick Steven Horil B. Nagesh Raghavendra Received February 18, 1988; accepted after re- vision May 16, 1988. Presented at the annual meeting of the Society of tkoradiology, Orlando, FL, January 1988, and one of two recipients of the Society of tkoradiology Award. 1 All authors: Department of Radiology, New York University Medical Center, 560 First Ave., New York, NY 10016. Address reprint requests to M. A. Bosniak. AJR 151:497-501, September 1988 0361 -803x/88/1 51 3-0497 © American Roentgen Ray Society Six patients were reviewed who had renal angiomyolipoma (1.2-4.0 cm) in which only minimal amounts of fat were evident on CT. The fat content of the lesion was appreciated because tissue attenuation measurements of small areas of low attenuation within the tumors were performed and because thin-section (5-mm) and nonenhanced CT scans were used. The fat content of the lesions could be identified on 10-mm sections in three cases but only on 5-mm sections in three others. In two cases, fat was seen only on the nonenhanced 5-mm thin sections. Careful sampling of low-density regions within the mass must be performed because a single region of interest over the entire tumor will produce an average attenuation in the soft-tissue range. The use of 5-mm thin sections and thin, nonenhanced CT sections increases spatial and density resolution and de- creases susceptibility to partial-volume effects. In a correlative study, no areas of fat were detected in a review of 100 well-circumscribed (4.0 cm or smaller) renal cell carcinomas. Detecting the existence of fat in a renal lesion will establish the diagnosis of angiomyolipoma and is the only radiologic finding that can differentiate it from renal cell carcinoma. Thus, unnecessary surgery will be avoided in these cases. Angiomyolipoma of the kidney, also called renal hamartoma, can be diagnosed noninvasively and with great accuracy by modern imaging techniques because fat within these tumors is usually shown readily by sonography and CT [i -61. However, some angiomyolipomas contain only tiny amounts of fat that can be easily over- looked unless searched for carefully in the CT study. Careful sampling of low- density regions within the mass must be performed, and the use of nonenhanced scans and thin sections (5 mm) will increase the chances of establishing the presence of fat within the tumor. Presentation of representative cases and a discussion of the techniques necessary to improve the detection of small amounts of fat are the purposes of our report. Materials and Methods Six patients were reviewed who had small angiomyolipomas (1 .2-4.0 cm) that contained tiny amounts of fat. A GE 8800 scanner was used in two patients and a GE 9800 in five; in one patient, one scan was obtained with each machine. Standard 1 0-mm-thick sections were used in all cases; four patients also were studied with 5-mm sections. In all patients, scans were obtained with IV contrast material; three patients had unenhanced scans as well. In total, 45 g of iodine were administered by the rapid bolus-infusion technique. Fatty tissue was considered to be present within a tumor if a region-of-interest value of - 10 H or lower was found within the tumor. Region-of-interest measurements were used that included at least a total of three adjacent pixels. In three cases, the region-of-interest measurement included nine pixels or more. CT scanners were calibrated daily with a phantom. Water- density structures did not measure less than 0 H. Internal checks in each case were used to test the validity of the CT numbers by measuring known fat and fluid areas within the body on each scan. Sonographic findings in two cases and angiographic findings in one case were correlated with the CT findings. Follow-up studies were available in three cases, and pathologic results were available in three cases.
CT Diagnosis of Renal Angiomyolipoma: The Importance of Detecting Small Amounts of Fat
Dec 09, 2022
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
497
CT Diagnosis of Renal Angiomyolipoma: The Importance of Detecting Small Amounts of Fat
Morton A. Bosniak1 AlecJ. Megibow
Donald H. Hulnick Steven Horil
B. Nagesh Raghavendra
Received February 18, 1988; accepted after re- vision May 16, 1988.
Presented at the annual meeting of the Society of tkoradiology, Orlando, FL, January 1988, and one of two recipients of the Society of tkoradiology Award.
1 All authors: Department of Radiology, New York University Medical Center, 560 First Ave., New York, NY 10016. Address reprint requests to M. A. Bosniak.
AJR 151:497-501, September 1988 0361 -803x/88/1 51 3-0497 © American Roentgen Ray Society
Six patients were reviewed who had renal angiomyolipoma (1.2-4.0 cm) in which only minimal amounts of fat were evident on CT. The fat content of the lesion was appreciated because tissue attenuation measurements of small areas of low attenuation within the tumors were performed and because thin-section (5-mm) and nonenhanced CT scans were used. The fat content of the lesions could be identified on 10-mm sections in three cases but only on 5-mm sections in three others. In two cases, fat was seen only on the nonenhanced 5-mm thin sections. Careful sampling of low-density regions within the mass must be performed because a single region of interest over the entire tumor will produce an average attenuation in the soft-tissue range. The use of 5-mm thin sections and thin, nonenhanced CT sections increases spatial and density resolution and de- creases susceptibility to partial-volume effects. In a correlative study, no areas of fat were detected in a review of 100 well-circumscribed (4.0 cm or smaller) renal cell carcinomas.
Detecting the existence of fat in a renal lesion will establish the diagnosis of
angiomyolipoma and is the only radiologic finding that can differentiate it from renal cell
carcinoma. Thus, unnecessary surgery will be avoided in these cases.
Angiomyolipoma of the kidney, also called renal hamartoma, can be diagnosed noninvasively and with great accuracy by modern imaging techniques because fat within these tumors is usually shown readily by sonography and CT [i -61. However, some angiomyolipomas contain only tiny amounts of fat that can be easily over- looked unless searched for carefully in the CT study. Careful sampling of low- density regions within the mass must be performed, and the use of nonenhanced scans and thin sections (5 mm) will increase the chances of establishing the presence of fat within the tumor. Presentation of representative cases and a discussion of the techniques necessary to improve the detection of small amounts of fat are the purposes of our report.
Materials and Methods
Six patients were reviewed who had small angiomyolipomas (1 .2-4.0 cm) that contained tiny amounts of fat. A GE 8800 scanner was used in two patients and a GE 9800 in five; in one patient, one scan was obtained with each machine. Standard 10-mm-thick sections were used in all cases; four patients also were studied with 5-mm sections. In all patients, scans were obtained with IV contrast material; three patients had unenhanced scans as well. In total, 45 g of iodine were administered by the rapid bolus-infusion technique. Fatty tissue was considered to be present within a tumor if a region-of-interest value of - 10 H or lower was found within the tumor. Region-of-interest measurements were used that included at least a total of three adjacent pixels. In three cases, the region-of-interest measurement included nine pixels or more. CT scanners were calibrated daily with a phantom. Water-
density structures did not measure less than 0 H. Internal checks in each case were used to test the validity of the CT numbers by measuring known fat and fluid areas within the body
on each scan. Sonographic findings in two cases and angiographic findings in one case were correlated with the CT findings. Follow-up studies were available in three cases, and pathologic results were available in three cases.
Fig. 1.-Case 1: Angiomyollpoma of the kidney with small areas of fat within the tumor. A 68-year-old man had a CT study to evaluate renal masses seen on urography performed for prostatism.
A, IV contrast-enhanced CT scan (10-mm section) shows 2-cm mass in anterior aspect of midpole of right kidney. Some areas within lesion are of lower attenuation. The entire lesion measured 41 H. Other areas of decreased attenuation posteriorly placed in parenchyma represent partial voluming of cysts seen on adjacent images.
B-D, IV contrast-enhanced CT scans (5-mm sections). Areas of lower attenuation within mass are defined more clearly than In A. Lesion measured with large cursor in C) was 35 H. Portion of lesion measured with small cursor (arrow in D) was -26 H, indicating fat content of tumor.
The diagnosis of anglomyolipoma was established. A 2-year follow-up scan showed no change in lesion size.
498 BOSNIAK ET AL. AJR:151, September 1988
The CT scans of 1 00 patients with pathologically confirmed well- circumscribed renal cell carcinomas 4 cm in diameter or smaller chosen to conform in size and shape with the angiomyolipomas were reviewed to determine the presence of areas of low attenuation within the tumors. Scans were obtained on the equipment already described with and without contrast enhancement. Ten-millimeter sections were made in all patients; 5-mm sections were available in 73 cases.
Results
Six patients with renal angiomyolipomas of i .2-4.0 cm were included in the study; the three men and three women were 43-68 years old (Table i). Fat was identifiable within the lesion on i 0-mm-thick contrast-enhanced CT sections in three cases (-i 0 H or lower), was suggestive in one case (0 to -9 H), and was not seen in two cases. In the one patient in whom i 0-mm contrast-enhanced CT sections were only suggestive of fat, the 5-mm enhanced scan indicated it was definitely present. In three patients, 5-mm CT sections were obtained without IV contrast material. In one of these, the fat content of the tumor was detected more readily than on the contrast-enhanced scan. In the other two, the nonenhanced
TABLE 1: Angiomyolipomas with CT Evidence of Small Amounts of Fat
Lowest A ttenuation Number (H) on
Case K 0.
Age Gender Size
5-mm 5-mm Enhanced Nonenhanced
1 68 M 2.0 -5 -26 -35 2 64 F 1.2 73 15 -17 3 45 M 2.9 -13 -25 NP 4 59 M 2.0 -12 NP NP 5 43 F 4.0 -26 NP NP 6 43 F 2.5 98 9 -10
Note-Sonography, performed only in cases 2 and 3, showed hyperechoic lesions. Cases 4 and 5 had nephrectomies and case 6 had a tumor- ectomy. Cases 1 , 2, and 3 had follow-ups of 2, 3, and 4 years; respectively. NP = not performed.
5-mm CT sections were the only scans in which the fat content of the tumor could be documented. CT, sonographic, and angiographic findings in all six patients are illustrated in Figures 1-6.
Two patients had sonography, which showed highly echo- genic tumors that led to further study of the lesions. One patient had angiography, which showed irregular vessels. Three patients had surgical resection, in two cases because of the insistence of the surgeon. In one of these two, a tumorectomy was performed because hamartoma was strongly suspected. In the third, the fat content of the tumor was not appreciated initially but only after the pathologic diagnosis was made. Three patients did not undergo surgery. Follow-ups of 2, 3, and 4 years, respectively, have shown no change in the size or character of the lesion. Surgery was avoided in two of the patients because fat in the tumor was clearly established, on both 5-mm sections (but not the i 0- mm section) in one and on only the 5-mm nonenhanced section in the other.
In the evaluation of the 1 00 cases of renal cell carcinoma, no tissue indicative of fat was found within the tumors. The lowest CT number obtained was 5 H, found in a tumor
containing cystic spaces.
Discussion
Renal angiomyolipoma is a fairly common incidental finding in the kidney. It can be diagnosed with great accuracy by modem imaging techniques, especially CT, because of the fat content of the tumor. Patients with asymptomatic lesions require no treatment. Treatment for symptomatic lesions often is surgical, but renal embolization has also been used, espe- cially to stop acute bleeding [7]. In the past, surgical removal was common because of the inability to distinguish the lesion from renal cell carcinoma.
We studied six patients with renal angiomyolipomas in which minimal amounts of fat were evident on CT, and we
A B
A
Fig. 2.-Case 2: Angiomyollpoma of the kidney not diagnosed on 10-mm sectIons but only on nonenhanced 5-mm sectIons. A 64-year-old woman had sonography to evaluate a left renal mass.
A, Longitudinal sonogram of right kidney re veals highly echogenic focus at upper pole (ar- row).
B, IV contrast-enhanced CT scan (10-mm sac- tieii) shows large cyst In upper pole of left kidney (12 H). A 1.2-cm-diameter area of decreased at- tenuatlon relative to enhancing renal nephrogram Is noted at lateral margin of upper pole right kidney (arrow). Lesion measured 73 H.
C, IV contrast-enhanced scan (5-mm section) also shows cyst at upper pole of left kidney. Area of decreased attenuation Is seen at upper pole of right kidney. However, within lesion Is area of lower attenuation (arrow) that measures 15 H. In vIew of sonographlc findings, nonenhanced CT was repeated.
D, Noncontrast CT scan (5-mm section) shows area of low attenuation in upper pole of right kid- nay (arrow). Measurements as low as -17 H were obtained.
Because of fat detected In lesion, surgery was avoided. Follow-up sonography and noncontrast CT over a 3-year period revealed no change In lesion size.
B
Fig. 3.-Case 3: Angiomyollpoma in kidney of patient wIth lymphoma. A 45-year-old man had a CT study to stage known lymphoma.
IV contrast-enhanced CT scan (5-mm sectIon) shows retroperitoneal adenopathy representing known lymphoma. Area of relatively decreased attenuation In right kidney measures 40 H with large cursor over entire lesion. However, at tiny region of interest (arrow), measurements as low as -25 H were obtained, indicating lesion repre- sented an anglomyolipoma.
Patient has been followed for lymphoma for 4 years. No change has been seen in the size of anglomyollpoma.
Fig. 4.-Case 4: Angiomyolipoma not diagnosed preoperatively. A 59-year-old man had a CT survey for upper abdominal discomfort.
A, IV contrast-enhanced CT scan (10-mm section) shows 2-cm lesion oft posterior surface of left kidney. Lesion measures 68 H. Subsequently, measurement of -12 H (at tip of arrow) was obtained, indicating fat content of lesion.
B, Selective left renal arteriogram with epinephrine enhancement reveals tiny cluster of abnormal- appearing vessels (arrow). (Fig. 48 courtesy of R. Bernstein, New York.)
Patient underwent angiography and surgery at another hospital. Pathology indicated anglornyoll- poma. CT scan was reevaluated and small area of fat was then appreciated.
A B C
Fig. 5.-Case 5: Anglomyolipoma proved at surgery. A 43-year-old woman had CT survey of abdomen.
IV contrast-enhanced CT scan (10- mm section) shows 4-cm mass extend- ing oft lateral aspect of lower pole of right kidney. Some areas of mass measured up to 80 H; area at arrow was -26 H, indicating fat in the tumor.
Although advised to do further stud- les, surgeon performed nephrectomy. Pathologic diagnosis was angiomyoll- poma. (Case courtesy of E. Kazam, New York.)
500 BOSNIAK ET AL. AJR:151 , September 1988
Fig. 6.-Case 6: Angiomyolipoma treated by local resection. A 43-year-old woman had CT to evaluate renal mass detected at previous gynecologic survey.
A, IV contrast-enhanced CT scan (10-mm section) shows well-circumscribed enhancing mass (measuring 98 H).
B, IV contrast-enhanced CT scan (5-mm section) shows tiny area of decreased attenuation in center of lesion. Lowest CT measurement was 9 H.
C, Nonenhanced CT scan (5-mm section) shows tiny area of decreased attenuation in center of lesion measuring -10 H (arrow).
Because fat measurement value was borderline, tumorectomy was performed. Pathologic diagnosis was hamartoma with mostiy myomatous tissue and a small focus of fat In lesion center.
analyzed the CT techniques that best show the fat. All six patients had negative attenuation (Hounsfield) values. If the lesion is measured with a large region of interest that includes higher attenuating portions of the lesion, positive attenuation
values will be obtained. The negative numbers indicative of fat will be obtained only if tissue attenuation measurements of small areas of low attenuation within the tumor are made. If there is suspicion that the lesion might contain fat, but negative numbers below -1 0 H are not found in the lesion on 1 0-mm thick sections, the existence of fat in the lesion can be investigated by using 5-mm sections as well as by rescan- ning the patient without contrast enhancement. The use of 5- mm thin sections and nonenhanced scans improves the chances of detecting and measuring these small areas of fat because of the increased spatial and density resolution af- forded by thinner sections and because the CT value is less susceptible to partial-volume effects (cases 1 , 2, and 6). Showing that an area of the tumor suspected of being fatty tissue measures even more negatively on thinner sections is
in itself further proof of the reliability of the negative measure- ment and the existence offat in the lesion. In fact, if a negative
CT number (particularly in the -1 to -1 0 H range) obtained on a 1 0-mm section cannot be reproduced and shown to be even lower on 5-mm sections, the accuracy of the CT number on the initial study must be seriously questioned and a diag- nosis of fat-containing tumor cannot be established. The use of non-contrast-enhanced 5-mm CT sections is particularly important in proving the presence of small amounts of fat in these tumors. These lesions often are vascular, and the increased attenuation of the enhancing tissues on the con-
trast-enhanced scan tends to obscure the fat and makes
negative measurements in these small areas of fat more difficult to obtain. This was seen in two of our patients in whom negative CT numbers were obtained only on the 5-mm nonenhanced scans.
Some authorities [8, 9] consider absolute CT numbers relatively unreliable. It has been shown that there is a signifi- cant difference in absolute CT numbers between scanners and that there may be a significant difference in absolute CT numbers on the same scanner on different days. Factors such as peak kilovoltage, orientation of the area of interest with respect to the scan aperture (the position of the structure within the body), and the CT numbers of surrounding tissues also can affect the absolute CT numbers. However, the use of CT to distinguish fatty tissue on the basis of negative attenuation is well established [1-6].
The characteristic highly echogenic appearance of angio- myolipomas on sonography is well known and described in
the literature [1 -3, 1 0-1 2]. Although this is not a specific finding (because other renal lesions occasionally will be highly echogenic), the presence of this pattern in a renal lesion should alert the radiologist to the possibility of an angiomyo- lipoma and should suggest more studies to look for a focus offat in the lesion if it is not appreciated initially. This occurred in cases 2 and 3. In both cases, the characteristic sonographic findings alerted the radiologist to the possibility of angiomyo- liporna, which was then proved by repeat examination with 5-mm sections and follow-up studies.
Other fat-containing renal tumors might be encountered such as lipomas, well-differentiated liposarcomas, and per-
AJR:151, September1988 CT OF RENAL ANGIOMYOLIPOMA 501
haps teratomas. Lipomas contain more fat and are treated the same way as an angiomyolipoma. Liposarcomas generally are larger, tend to be invasive, and probably would be treated surgically because of their size and tendency to be sympto- matic. We were unable to find a small, well-marginated lipo- sarcoma that could be confused with an angiomyolipoma in our own material or in a literature search. Teratoma is extraor- dinarily rare and can be recognized by its calcium content.
It is only a small percentage of hamartomas of the kidney that do not contain fat (angiomyomas) [1 3, 14]. One such case has been reported that was diagnosed correctly without biopsy because of its association with lymphangiomyomato- sis [1 5]. In patients with tuberous sclerosis, some of the multiple renal hamartomas will not show fatty areas within the tumors. However, in such cases, the diagnosis is assumed and rarely is a clinical problem. However, when tuberous sclerosis or lymphangiomyomatosis is not associated with the case, a hamartoma that does not contain fat cannot be diagnosed without biopsy.
CT of renal cell carcinoma does not reveal islands of fat in the tumor similar to those in angiomyolipoma. None of the CT scans in patients with well-marginated (4 cm or smaller) carcinomas in our series showed the presence of fat within the tumor, and we were unable to find reference in the literature to such an occurrence. Care must be taken to be certain that fat in the renal sinus or perirenal tissues, which occasionally becomes engulfed within the tumor, is not con- fused with the intrinsic tissues of the tumor. However, the nature of the growth pattern of the neoplasm and its position in the kidney should indicate whether any fat seen in the tumor is part of the neoplasm or encompasses adjacent fatty tissue, particularly in the comparatively small, marginated lesions described in our series. In one report of two children with Wilms tumor [1 6] fatty elements were found. This can occur because the nonepithelial stromal portion of the tumor may undergo metaplastic change to cartilage, bone, adipose tissue, and muscle [1 7]. However, these cases of Wilms tumor would not be confused with the kind of tumors dis- cussed in this article.
ACKNOWLEDGMENTS
Deborah Persily for secretarial assistance.
REFERENCES
1 . Bosniak MA. Angiomyolipoma (hamartoma) of the kidney: a preoperative diagnosis is possible in virtually every case. Urol Radiol 1981;3: 135-1 42
2. Raghavendra BN, Bosniak MA, Megibow AJ. Small angiomyolipoma of the kidney: sonographic-CT evaluation. AJR 1983;141 :575-578
3. Bret PM, Bretagnolle M, Gaillard D, et ai. Small, asymptomatic anglomyo- lipomas of the kidney. Radiology 1985;154:7-10
4. Hanson GE, Hoffman RB, Sample WF, Becker R. Computed tomography diagnosis of renal angiomyolipoma. Radiology 1978;128:789-791
5. Totty WG, McClennan BL, Melson GL, Patel R. Relative value of computed tomography and ultrasonography in the assessment of renal angiomyoli- poma. J ComputAssist Tomogr 1981;5:173-178
6. Friedman AC, Hartman DS, Sherman J, Lautin EM, Goldman M. Computed tomography of abdominal fatty masses. Radiology 1981;139:415-429
7. Oesterling JE, Fishman EK, Goldman SE, Marshall FF. Management of angiomyolipoma. J Urol 1986;135:1 121-1124
8. Levi C, Gray JE, McCullough EC, Hattery RR. The unreliability of CT numbers as absolute values. AJR 1982;139:443-447
9. Zerhouni EA, Stitik FP, Siegelman SS, et al. CT of the pulmonary nodule: a cooperative study. Radiology 1986;160:319-327
10. Scheible W, Ellenbogen PH, Leopold GR, Siao NT. Upomatous tumors of the kidney and adrenal: apparent echographic specificity. Radiology
1978;129:153-156 11 . Lee TG, Henderson SC, Freeny PC, Raskin MM, Benson EP, Pearse HD.
Ultrasound findings of renal angiomyolipoma. JCU 1978;6: 150-1 55 12. Hartman DS, Goldman SM, Friedman AC, Davis CL, Maxwell JE, Sherman
JL. Angiomyolipoma: ultrasonic-pathologic correlation. Radiology 1981; 139:451-458
13. Reese MM, Winstanley OP. The small tumor-like lesions of the kidney. Br J Cancer 1958;12:507-516
14. Bennington JL, Beckwith JB. Tumors ofthe kidney, renal pelvis and ureter. In: Atlas of tumorpathology, 2nd series, fasc. 12. Washington, DC: Armed
Forces Institute of Pathology, 1975:20 15. Rumancik WM, Bosniak MA, Rosen RJ, Hulnick OH. Atypical renal and
pararenal hamartomas associated with lymphangiomyomatosis. AiR
1984;142:971 -972
16. Parvey LS, Warner RM, Calhihan TR, Magill HL.…
CT Diagnosis of Renal Angiomyolipoma: The Importance of Detecting Small Amounts of Fat
Morton A. Bosniak1 AlecJ. Megibow
Donald H. Hulnick Steven Horil
B. Nagesh Raghavendra
Received February 18, 1988; accepted after re- vision May 16, 1988.
Presented at the annual meeting of the Society of tkoradiology, Orlando, FL, January 1988, and one of two recipients of the Society of tkoradiology Award.
1 All authors: Department of Radiology, New York University Medical Center, 560 First Ave., New York, NY 10016. Address reprint requests to M. A. Bosniak.
AJR 151:497-501, September 1988 0361 -803x/88/1 51 3-0497 © American Roentgen Ray Society
Six patients were reviewed who had renal angiomyolipoma (1.2-4.0 cm) in which only minimal amounts of fat were evident on CT. The fat content of the lesion was appreciated because tissue attenuation measurements of small areas of low attenuation within the tumors were performed and because thin-section (5-mm) and nonenhanced CT scans were used. The fat content of the lesions could be identified on 10-mm sections in three cases but only on 5-mm sections in three others. In two cases, fat was seen only on the nonenhanced 5-mm thin sections. Careful sampling of low-density regions within the mass must be performed because a single region of interest over the entire tumor will produce an average attenuation in the soft-tissue range. The use of 5-mm thin sections and thin, nonenhanced CT sections increases spatial and density resolution and de- creases susceptibility to partial-volume effects. In a correlative study, no areas of fat were detected in a review of 100 well-circumscribed (4.0 cm or smaller) renal cell carcinomas.
Detecting the existence of fat in a renal lesion will establish the diagnosis of
angiomyolipoma and is the only radiologic finding that can differentiate it from renal cell
carcinoma. Thus, unnecessary surgery will be avoided in these cases.
Angiomyolipoma of the kidney, also called renal hamartoma, can be diagnosed noninvasively and with great accuracy by modern imaging techniques because fat within these tumors is usually shown readily by sonography and CT [i -61. However, some angiomyolipomas contain only tiny amounts of fat that can be easily over- looked unless searched for carefully in the CT study. Careful sampling of low- density regions within the mass must be performed, and the use of nonenhanced scans and thin sections (5 mm) will increase the chances of establishing the presence of fat within the tumor. Presentation of representative cases and a discussion of the techniques necessary to improve the detection of small amounts of fat are the purposes of our report.
Materials and Methods
Six patients were reviewed who had small angiomyolipomas (1 .2-4.0 cm) that contained tiny amounts of fat. A GE 8800 scanner was used in two patients and a GE 9800 in five; in one patient, one scan was obtained with each machine. Standard 10-mm-thick sections were used in all cases; four patients also were studied with 5-mm sections. In all patients, scans were obtained with IV contrast material; three patients had unenhanced scans as well. In total, 45 g of iodine were administered by the rapid bolus-infusion technique. Fatty tissue was considered to be present within a tumor if a region-of-interest value of - 10 H or lower was found within the tumor. Region-of-interest measurements were used that included at least a total of three adjacent pixels. In three cases, the region-of-interest measurement included nine pixels or more. CT scanners were calibrated daily with a phantom. Water-
density structures did not measure less than 0 H. Internal checks in each case were used to test the validity of the CT numbers by measuring known fat and fluid areas within the body
on each scan. Sonographic findings in two cases and angiographic findings in one case were correlated with the CT findings. Follow-up studies were available in three cases, and pathologic results were available in three cases.
Fig. 1.-Case 1: Angiomyollpoma of the kidney with small areas of fat within the tumor. A 68-year-old man had a CT study to evaluate renal masses seen on urography performed for prostatism.
A, IV contrast-enhanced CT scan (10-mm section) shows 2-cm mass in anterior aspect of midpole of right kidney. Some areas within lesion are of lower attenuation. The entire lesion measured 41 H. Other areas of decreased attenuation posteriorly placed in parenchyma represent partial voluming of cysts seen on adjacent images.
B-D, IV contrast-enhanced CT scans (5-mm sections). Areas of lower attenuation within mass are defined more clearly than In A. Lesion measured with large cursor in C) was 35 H. Portion of lesion measured with small cursor (arrow in D) was -26 H, indicating fat content of tumor.
The diagnosis of anglomyolipoma was established. A 2-year follow-up scan showed no change in lesion size.
498 BOSNIAK ET AL. AJR:151, September 1988
The CT scans of 1 00 patients with pathologically confirmed well- circumscribed renal cell carcinomas 4 cm in diameter or smaller chosen to conform in size and shape with the angiomyolipomas were reviewed to determine the presence of areas of low attenuation within the tumors. Scans were obtained on the equipment already described with and without contrast enhancement. Ten-millimeter sections were made in all patients; 5-mm sections were available in 73 cases.
Results
Six patients with renal angiomyolipomas of i .2-4.0 cm were included in the study; the three men and three women were 43-68 years old (Table i). Fat was identifiable within the lesion on i 0-mm-thick contrast-enhanced CT sections in three cases (-i 0 H or lower), was suggestive in one case (0 to -9 H), and was not seen in two cases. In the one patient in whom i 0-mm contrast-enhanced CT sections were only suggestive of fat, the 5-mm enhanced scan indicated it was definitely present. In three patients, 5-mm CT sections were obtained without IV contrast material. In one of these, the fat content of the tumor was detected more readily than on the contrast-enhanced scan. In the other two, the nonenhanced
TABLE 1: Angiomyolipomas with CT Evidence of Small Amounts of Fat
Lowest A ttenuation Number (H) on
Case K 0.
Age Gender Size
5-mm 5-mm Enhanced Nonenhanced
1 68 M 2.0 -5 -26 -35 2 64 F 1.2 73 15 -17 3 45 M 2.9 -13 -25 NP 4 59 M 2.0 -12 NP NP 5 43 F 4.0 -26 NP NP 6 43 F 2.5 98 9 -10
Note-Sonography, performed only in cases 2 and 3, showed hyperechoic lesions. Cases 4 and 5 had nephrectomies and case 6 had a tumor- ectomy. Cases 1 , 2, and 3 had follow-ups of 2, 3, and 4 years; respectively. NP = not performed.
5-mm CT sections were the only scans in which the fat content of the tumor could be documented. CT, sonographic, and angiographic findings in all six patients are illustrated in Figures 1-6.
Two patients had sonography, which showed highly echo- genic tumors that led to further study of the lesions. One patient had angiography, which showed irregular vessels. Three patients had surgical resection, in two cases because of the insistence of the surgeon. In one of these two, a tumorectomy was performed because hamartoma was strongly suspected. In the third, the fat content of the tumor was not appreciated initially but only after the pathologic diagnosis was made. Three patients did not undergo surgery. Follow-ups of 2, 3, and 4 years, respectively, have shown no change in the size or character of the lesion. Surgery was avoided in two of the patients because fat in the tumor was clearly established, on both 5-mm sections (but not the i 0- mm section) in one and on only the 5-mm nonenhanced section in the other.
In the evaluation of the 1 00 cases of renal cell carcinoma, no tissue indicative of fat was found within the tumors. The lowest CT number obtained was 5 H, found in a tumor
containing cystic spaces.
Discussion
Renal angiomyolipoma is a fairly common incidental finding in the kidney. It can be diagnosed with great accuracy by modem imaging techniques, especially CT, because of the fat content of the tumor. Patients with asymptomatic lesions require no treatment. Treatment for symptomatic lesions often is surgical, but renal embolization has also been used, espe- cially to stop acute bleeding [7]. In the past, surgical removal was common because of the inability to distinguish the lesion from renal cell carcinoma.
We studied six patients with renal angiomyolipomas in which minimal amounts of fat were evident on CT, and we
A B
A
Fig. 2.-Case 2: Angiomyollpoma of the kidney not diagnosed on 10-mm sectIons but only on nonenhanced 5-mm sectIons. A 64-year-old woman had sonography to evaluate a left renal mass.
A, Longitudinal sonogram of right kidney re veals highly echogenic focus at upper pole (ar- row).
B, IV contrast-enhanced CT scan (10-mm sac- tieii) shows large cyst In upper pole of left kidney (12 H). A 1.2-cm-diameter area of decreased at- tenuatlon relative to enhancing renal nephrogram Is noted at lateral margin of upper pole right kidney (arrow). Lesion measured 73 H.
C, IV contrast-enhanced scan (5-mm section) also shows cyst at upper pole of left kidney. Area of decreased attenuation Is seen at upper pole of right kidney. However, within lesion Is area of lower attenuation (arrow) that measures 15 H. In vIew of sonographlc findings, nonenhanced CT was repeated.
D, Noncontrast CT scan (5-mm section) shows area of low attenuation in upper pole of right kid- nay (arrow). Measurements as low as -17 H were obtained.
Because of fat detected In lesion, surgery was avoided. Follow-up sonography and noncontrast CT over a 3-year period revealed no change In lesion size.
B
Fig. 3.-Case 3: Angiomyollpoma in kidney of patient wIth lymphoma. A 45-year-old man had a CT study to stage known lymphoma.
IV contrast-enhanced CT scan (5-mm sectIon) shows retroperitoneal adenopathy representing known lymphoma. Area of relatively decreased attenuation In right kidney measures 40 H with large cursor over entire lesion. However, at tiny region of interest (arrow), measurements as low as -25 H were obtained, indicating lesion repre- sented an anglomyolipoma.
Patient has been followed for lymphoma for 4 years. No change has been seen in the size of anglomyollpoma.
Fig. 4.-Case 4: Angiomyolipoma not diagnosed preoperatively. A 59-year-old man had a CT survey for upper abdominal discomfort.
A, IV contrast-enhanced CT scan (10-mm section) shows 2-cm lesion oft posterior surface of left kidney. Lesion measures 68 H. Subsequently, measurement of -12 H (at tip of arrow) was obtained, indicating fat content of lesion.
B, Selective left renal arteriogram with epinephrine enhancement reveals tiny cluster of abnormal- appearing vessels (arrow). (Fig. 48 courtesy of R. Bernstein, New York.)
Patient underwent angiography and surgery at another hospital. Pathology indicated anglornyoll- poma. CT scan was reevaluated and small area of fat was then appreciated.
A B C
Fig. 5.-Case 5: Anglomyolipoma proved at surgery. A 43-year-old woman had CT survey of abdomen.
IV contrast-enhanced CT scan (10- mm section) shows 4-cm mass extend- ing oft lateral aspect of lower pole of right kidney. Some areas of mass measured up to 80 H; area at arrow was -26 H, indicating fat in the tumor.
Although advised to do further stud- les, surgeon performed nephrectomy. Pathologic diagnosis was angiomyoll- poma. (Case courtesy of E. Kazam, New York.)
500 BOSNIAK ET AL. AJR:151 , September 1988
Fig. 6.-Case 6: Angiomyolipoma treated by local resection. A 43-year-old woman had CT to evaluate renal mass detected at previous gynecologic survey.
A, IV contrast-enhanced CT scan (10-mm section) shows well-circumscribed enhancing mass (measuring 98 H).
B, IV contrast-enhanced CT scan (5-mm section) shows tiny area of decreased attenuation in center of lesion. Lowest CT measurement was 9 H.
C, Nonenhanced CT scan (5-mm section) shows tiny area of decreased attenuation in center of lesion measuring -10 H (arrow).
Because fat measurement value was borderline, tumorectomy was performed. Pathologic diagnosis was hamartoma with mostiy myomatous tissue and a small focus of fat In lesion center.
analyzed the CT techniques that best show the fat. All six patients had negative attenuation (Hounsfield) values. If the lesion is measured with a large region of interest that includes higher attenuating portions of the lesion, positive attenuation
values will be obtained. The negative numbers indicative of fat will be obtained only if tissue attenuation measurements of small areas of low attenuation within the tumor are made. If there is suspicion that the lesion might contain fat, but negative numbers below -1 0 H are not found in the lesion on 1 0-mm thick sections, the existence of fat in the lesion can be investigated by using 5-mm sections as well as by rescan- ning the patient without contrast enhancement. The use of 5- mm thin sections and nonenhanced scans improves the chances of detecting and measuring these small areas of fat because of the increased spatial and density resolution af- forded by thinner sections and because the CT value is less susceptible to partial-volume effects (cases 1 , 2, and 6). Showing that an area of the tumor suspected of being fatty tissue measures even more negatively on thinner sections is
in itself further proof of the reliability of the negative measure- ment and the existence offat in the lesion. In fact, if a negative
CT number (particularly in the -1 to -1 0 H range) obtained on a 1 0-mm section cannot be reproduced and shown to be even lower on 5-mm sections, the accuracy of the CT number on the initial study must be seriously questioned and a diag- nosis of fat-containing tumor cannot be established. The use of non-contrast-enhanced 5-mm CT sections is particularly important in proving the presence of small amounts of fat in these tumors. These lesions often are vascular, and the increased attenuation of the enhancing tissues on the con-
trast-enhanced scan tends to obscure the fat and makes
negative measurements in these small areas of fat more difficult to obtain. This was seen in two of our patients in whom negative CT numbers were obtained only on the 5-mm nonenhanced scans.
Some authorities [8, 9] consider absolute CT numbers relatively unreliable. It has been shown that there is a signifi- cant difference in absolute CT numbers between scanners and that there may be a significant difference in absolute CT numbers on the same scanner on different days. Factors such as peak kilovoltage, orientation of the area of interest with respect to the scan aperture (the position of the structure within the body), and the CT numbers of surrounding tissues also can affect the absolute CT numbers. However, the use of CT to distinguish fatty tissue on the basis of negative attenuation is well established [1-6].
The characteristic highly echogenic appearance of angio- myolipomas on sonography is well known and described in
the literature [1 -3, 1 0-1 2]. Although this is not a specific finding (because other renal lesions occasionally will be highly echogenic), the presence of this pattern in a renal lesion should alert the radiologist to the possibility of an angiomyo- lipoma and should suggest more studies to look for a focus offat in the lesion if it is not appreciated initially. This occurred in cases 2 and 3. In both cases, the characteristic sonographic findings alerted the radiologist to the possibility of angiomyo- liporna, which was then proved by repeat examination with 5-mm sections and follow-up studies.
Other fat-containing renal tumors might be encountered such as lipomas, well-differentiated liposarcomas, and per-
AJR:151, September1988 CT OF RENAL ANGIOMYOLIPOMA 501
haps teratomas. Lipomas contain more fat and are treated the same way as an angiomyolipoma. Liposarcomas generally are larger, tend to be invasive, and probably would be treated surgically because of their size and tendency to be sympto- matic. We were unable to find a small, well-marginated lipo- sarcoma that could be confused with an angiomyolipoma in our own material or in a literature search. Teratoma is extraor- dinarily rare and can be recognized by its calcium content.
It is only a small percentage of hamartomas of the kidney that do not contain fat (angiomyomas) [1 3, 14]. One such case has been reported that was diagnosed correctly without biopsy because of its association with lymphangiomyomato- sis [1 5]. In patients with tuberous sclerosis, some of the multiple renal hamartomas will not show fatty areas within the tumors. However, in such cases, the diagnosis is assumed and rarely is a clinical problem. However, when tuberous sclerosis or lymphangiomyomatosis is not associated with the case, a hamartoma that does not contain fat cannot be diagnosed without biopsy.
CT of renal cell carcinoma does not reveal islands of fat in the tumor similar to those in angiomyolipoma. None of the CT scans in patients with well-marginated (4 cm or smaller) carcinomas in our series showed the presence of fat within the tumor, and we were unable to find reference in the literature to such an occurrence. Care must be taken to be certain that fat in the renal sinus or perirenal tissues, which occasionally becomes engulfed within the tumor, is not con- fused with the intrinsic tissues of the tumor. However, the nature of the growth pattern of the neoplasm and its position in the kidney should indicate whether any fat seen in the tumor is part of the neoplasm or encompasses adjacent fatty tissue, particularly in the comparatively small, marginated lesions described in our series. In one report of two children with Wilms tumor [1 6] fatty elements were found. This can occur because the nonepithelial stromal portion of the tumor may undergo metaplastic change to cartilage, bone, adipose tissue, and muscle [1 7]. However, these cases of Wilms tumor would not be confused with the kind of tumors dis- cussed in this article.
ACKNOWLEDGMENTS
Deborah Persily for secretarial assistance.
REFERENCES
1 . Bosniak MA. Angiomyolipoma (hamartoma) of the kidney: a preoperative diagnosis is possible in virtually every case. Urol Radiol 1981;3: 135-1 42
2. Raghavendra BN, Bosniak MA, Megibow AJ. Small angiomyolipoma of the kidney: sonographic-CT evaluation. AJR 1983;141 :575-578
3. Bret PM, Bretagnolle M, Gaillard D, et ai. Small, asymptomatic anglomyo- lipomas of the kidney. Radiology 1985;154:7-10
4. Hanson GE, Hoffman RB, Sample WF, Becker R. Computed tomography diagnosis of renal angiomyolipoma. Radiology 1978;128:789-791
5. Totty WG, McClennan BL, Melson GL, Patel R. Relative value of computed tomography and ultrasonography in the assessment of renal angiomyoli- poma. J ComputAssist Tomogr 1981;5:173-178
6. Friedman AC, Hartman DS, Sherman J, Lautin EM, Goldman M. Computed tomography of abdominal fatty masses. Radiology 1981;139:415-429
7. Oesterling JE, Fishman EK, Goldman SE, Marshall FF. Management of angiomyolipoma. J Urol 1986;135:1 121-1124
8. Levi C, Gray JE, McCullough EC, Hattery RR. The unreliability of CT numbers as absolute values. AJR 1982;139:443-447
9. Zerhouni EA, Stitik FP, Siegelman SS, et al. CT of the pulmonary nodule: a cooperative study. Radiology 1986;160:319-327
10. Scheible W, Ellenbogen PH, Leopold GR, Siao NT. Upomatous tumors of the kidney and adrenal: apparent echographic specificity. Radiology
1978;129:153-156 11 . Lee TG, Henderson SC, Freeny PC, Raskin MM, Benson EP, Pearse HD.
Ultrasound findings of renal angiomyolipoma. JCU 1978;6: 150-1 55 12. Hartman DS, Goldman SM, Friedman AC, Davis CL, Maxwell JE, Sherman
JL. Angiomyolipoma: ultrasonic-pathologic correlation. Radiology 1981; 139:451-458
13. Reese MM, Winstanley OP. The small tumor-like lesions of the kidney. Br J Cancer 1958;12:507-516
14. Bennington JL, Beckwith JB. Tumors ofthe kidney, renal pelvis and ureter. In: Atlas of tumorpathology, 2nd series, fasc. 12. Washington, DC: Armed
Forces Institute of Pathology, 1975:20 15. Rumancik WM, Bosniak MA, Rosen RJ, Hulnick OH. Atypical renal and
pararenal hamartomas associated with lymphangiomyomatosis. AiR
1984;142:971 -972
16. Parvey LS, Warner RM, Calhihan TR, Magill HL.…
Related Documents
