Original Research
Genitourinary Imaging
September 2007

Incidental Findings on Renal MR Angiography

Abstract

OBJECTIVE. The purpose of this article is to assess the incidence of incidental vascular and nonvascular findings in patients undergoing renal MR angiography and to determine the extent to which these findings alter patient management.
MATERIALS AND METHODS. Reports from 380 consecutive renal MR angiography examinations performed at a single institution over a 12-month interval were examined. The presence of incidental vascular (i.e., nonrenal artery) and nonvascular findings was noted. Clinical records of patients with significant incidental findings were examined to determine whether additional imaging, biopsy, or surgery was performed.
RESULTS. Overall, 151 (40%) of 380 patients had one or more additional vascular findings not related to the renal arteries, and 221 (58%) of 380 patients had one or more additional nonvascular findings. Vascular findings included mesenteric artery stenosis or occlusion in 33% of patients, moderate to severe aortic atherosclerosis in 17%, aortic aneurysms in 7%, and aortic dissection in 2%. Incidental malignancies were detected in 10 patients (3%), and indeterminate lesions requiring follow-up imaging, biopsy, or surgery were noted in 18 patients (5%). Overall, management in 5% of patients was significantly altered (i.e., required biopsy, surgery, or other intervention) by incidental findings detected on renal MR angiography. Benign lesions not requiring additional imaging or follow-up occurred in 54% of patients and consisted predominantly of renal cysts.
CONCLUSION. Incidental findings on renal MR angiography are common. Most incidental lesions can be adequately detected and characterized with the addition of a few pulse sequences to the standard renal MR angiography protocol at a minimal cost in imaging time. The high incidence of incidental findings emphasizes the importance of performance and interpretation of these examinations by physicians with training in abdominal cross-sectional imaging.

Introduction

Renal MR angiography is one of the most commonly performed examinations in body and vascular MRI and is generally regarded as an accurate and effective technique in the assessment of renal artery stenosis [16]. A number of different protocols are used in academic centers and private practice, but nearly all include 3D contrast-enhanced MR angiography as the primary technique. However, there is more variability in the amount of additional imaging performed before and after 3D contrast-enhanced MR angiography. Some practices simply perform scout sequences followed by 3D contrast-enhanced MR angiography, with the goal of rapid throughput and efficiency. Others include additional sequences before and after contrast administration to characterize incidental renal and adrenal lesions.
The incidence and importance of incidental findings on renal MR angiography is not generally known, and the purpose of this study is to determine the frequency of these findings, categorize their importance, and assess their impact on patient treatment.

Materials and Methods

Reports from 380 consecutive renal MR angiography examinations performed at a single institution over a 12-month interval from May 1, 2004, through May 1, 2005, were examined. Renal MR angiography images were interpreted by 17 fellowship-trained radiologists, including 15 abdominal radiologists, who interpreted 374 examinations, and two musculoskeletal radiologists, who interpreted six examinations. Experience in MR angiography interpretation ranged from 1 to 16 years (average, 8 years). Renal MR angiography examinations evaluated consisted of those requested for indications of hypertension, renal insufficiency, or renovascular disease. Specifically excluded were examinations performed for staging known renal cell carcinoma, characterization of renal masses, and all research protocols.
Reports were evaluated for the following information: presence or absence of significant (> 50% diameter luminal narrowing, or moderate to severe) renal artery stenosis and significant vascular disease involving the abdominal aorta, iliac arteries, and mesenteric arteries. Incidental nonvascular findings were also noted. Patient age and sex were recorded. Patient records were assessed to determine whether subsequent intervention was performed for renal artery stenosis or other vascular findings. Records were also examined in cases in which an incidental finding of significance was noted to determine whether additional diagnostic imaging was performed and whether subsequent surgery or biopsy was performed and, if so, the pathologic diagnosis.
Patient records were evaluated to determine whether cross-sectional imaging (sonography, CT, or MRI) had been performed within 1 year before or after renal MR angiography. Reports of these examinations were then examined for discrepancies with the MR angiography reports. Vascular discrepancies in renal sonography were noted for renal arteries and the aorta; mesenteric arteries were generally not examined on these studies. These findings were correlated with the results of renal MR angiography and conventional renal angiography when available. In cases in which nonvascular discrepancies were noted, the MR angiography and additional examinations were reviewed by the author to assess which report was correct.
The final 50 renal MR angiography examinations were assessed to determine the average examination time and the percentage of the examination required to obtain the sequences not directly related to MR angiography (single-shot fast spin-echo, in-phase and out-of-phase spoiled gradient-echo, and fat-saturated contrast-enhanced spoiled gradient-echo sequences). This was done by noting the time stamp on each sequence, which occurs at the initiation of the sequence. The total examination time represented the time difference between the first and last acquisitions plus the acquisition time of the last sequence. The acquisition time for a given pulse sequence (including prescription, download, and time before scanning) was estimated by the time from the end of the previous sequence plus the acquisition time of the given sequence.
All renal MR angiography examinations were performed on a 1.5-T TwinSpeed EXCITE system (GE Healthcare). The renal MR angiography examination consisted of a three-plane localizer acquired using either a single-shot fast spin-echo or a fast spoiled gradient-echo sequence. This was followed by coronal single-shot fast spin-echo images (TE, 80; bandwidth, 83 kHz; flip angle, 110°; matrix, 256 × 256; 0.5 excitations; 5-mm slice thickness with 1-mm gap; 40- to 44-cm field of view). Axial in-phase and out-of-phase spoiled gradient-echo images were then obtained, with coverage including the adrenals and kidneys (TR range/first-echo TE, second-echo TE, 100–200/2.1, 4.2; flip angle, 70°; bandwidth, 32 kHz; matrix, 256 × 192; 1 excitation; 6-mm slice thickness with 1-mm gap; field of view, 32–44 cm, with 0.75 partial-phase field of view). Renal MR angiography was performed using a coronal oblique 3D spoiled gradient-echo sequence with elliptic centric view ordering and the following parameters: 26- to 30-cm field of view; 0.75 phase field of view; 256 × 224 in-plane matrix; 1.6-mm section thickness; 35° flip angle; 62- to 83-kHz bandwidth; and TR/TE, 3.5/1.4. Typically 40 sections were acquired, and zero filling was performed to obtain a reconstructed 512 × 512 in-plane matrix with 50% overlapping reconstructions along the z-axis. Parameters were adjusted according to patient size and breath-hold capacity, with acquisition times generally in the range of 16–22 seconds.
A contrast dose of 0.1–0.2 mmol/L/kg (gadodiamide [Omniscan, GE Healthcare]) was injected at 3 mL/s with an automatic injector (Spectris Solaris, Medrad) and a scanning delay time was selected on the basis of a preceding test bolus acquisition using 1–2 mL of contrast material. A single arterial phase MR angiography acquisition was acquired without additional venous phase imaging. After MR angiography, an axial fat-saturated 2D spoiled gradient-echo sequence was performed (TR range/TE, 100–200/2.6; flip angle, 70°; bandwidth, 32 kHz; matrix, 256 × 192; 0.75 excitations; 32- to 44-cm field of view with 0.75 phase field of view; 6-mm slice thickness; and 1-mm gap).

Results

Renal MR Angiography Findings

Renal MR angiography was performed in 380 patients over 1 year. The patient population consisted of 192 women and 188 men. Patient ages ranged from 20 to 93 years (mean, 66 ± 15 [SD] years). Moderate to severe renal artery stenosis (i.e., luminal diameter narrowing of > 50%) was found in 173 patients (45%). Twenty patients had fibromuscular dysplasia (FMD) with significant renal artery narrowing, and 11 patients had renal artery occlusion (Table 1). Renal artery aneurysms were detected in seven patients. Renal transplant arteries were examined in 13 patients. Conventional renal angiography was performed in 68 patients after MR angiography results suggesting significant renal artery stenosis, resulting in renal stent placement in 52 patients and percutaneous transluminal angioplasty (PTA) in 12 patients. In two patients, conventional angiography was negative despite positive MR angiography. Two patients had conventional angiography performed despite negative MR angiography because of a strong clinical suspicion of renovascular hypertension; both of these studies were negative. Conventional renal angiography was performed in seven of 20 patients with suspected FMD, confirming the diagnosis and resulting in PTA in all cases. One of the negative conventional angiograms occurred in a patient with suspected FMD.
TABLE 1: Vascular Findings on Renal MR Angiography in 380 Patients
FindingNo.%
Renal arteries  
   Renal artery stenosisa17345
   Fibromuscular dysplasia205
   Renal artery occlusion113
   Total20454
Additional vascular findings  
   Mesenteric artery stenosis or occlusion12733
   Aortic atherosclerosisa6617
   Iliac stenosis or occlusiona287
   Abdominal aortic aneurysm257
   Penetrating atheromatous ulcer164
   Dissection92
   Aortic stenosis3 
   Aortic occlusion1 
   Aortic vasculitis, retroperitoneal fibrosis1 
   Nonocclusive thrombus in inferior vena cava
1

a
Moderate to severe

Additional Vascular Findings

Significant vascular findings outside the renal artery territory were common (Figs. 1, 2A, and 2B). One hundred twenty-seven patients (33%) had significant stenosis (> 50%) or occlusion of one or more mesenteric arteries, consisting of 31 patients with celiac stenosis or occlusion, 12 patients with superior mesenteric artery stenosis or occlusion, 39 patients with inferior mesenteric artery stenosis or occlusion, and 45 patients with stenosis or occlusion of more than one mesenteric artery. These lesions were known in 26 patients and unknown in 101 patients. Sixty-six patients had moderate to severe aortic arteriosclerosis (as described in MR angiography reports), 25 had abdominal aortic aneurysms measuring greater than 3 cm in diameter (14 known and 11 unknown), 16 had penetrating atheromatous ulcers (3 known, 13 unknown), and nine had dissections of the aorta or mesenteric vessels (3 known, 6 unknown) (Table 1).
Fig. 1 Incidental vascular findings in 70-year-old woman. Sagittal oblique subvolume maximum-intensity-projection image from renal MR angiography data reveals 4.2-cm infrarenal abdominal aortic aneurysm. Note severe stenosis of origins of celiac and superior mesenteric arteries. Inferior mesenteric artery was not identified and was presumed to be thrombosed.
Fig. 2A Incidental vascular findings in 82-year-old man. Subvolume maximum-intensity-projection image from renal MR angiography reveals ulcerating plaques in infrarenal abdominal aorta and right common iliac artery. Note also focal severe stenosis of left renal artery and diffusely small right renal artery.
Fig. 2B Incidental vascular findings in 82-year-old man. Axial fat-saturated contrast-enhanced spoiled gradient-echo image shows ulcerating plaque (arrow) and right renal atrophy.
Four patients underwent subsequent surgical repair of abdominal aortic aneurysms. In two cases the aneurysm was initially detected on renal MR angiography, and in two cases the aneurysm had significantly increased in diameter since previous examinations. One patient underwent subsequent angiography and stent placement in the superior mesenteric artery, and another patient underwent angiography and stent placement in the right external iliac artery (both of these findings were initially detected on renal MR angiography).

Nonvascular Findings

Ten patients had malignancies detected (Table 2), eight of which were unknown before renal MR angiography. These included renal cell carcinoma in three patients; lymphoma in two patients; and metastatic adenocarcinoma from an unknown primary source, small cell lung carcinoma, pancreatic adenocarcinoma, pancreatic islet cell carcinoma, and pheochromocytoma in one patient each (Figs. 3A, 3B, 4A, and 4B).
TABLE 2: Malignant Lesions Incidentally Found on Renal MR Angiography in 380 Patients
FindingNo.Previously Diagnosed?Additional ImagingOutcome
Renal cell carcinoma3NoChest CT in 1 patientPartial nephrectomy in 2 patients, radiofrequency ablation in 1
Lymphoma21/2CT of chest, abdomen, pelvisBone marrow biopsy followed by chemotherapy
Adrenal pheochromocytoma1NoNoneAdrenalectomy
Metastatic pancreatic adenocarcinoma1NoNoneSonographically guided biopsy of hepatic metastasis; patient died
Pancreatic islet cell carcinoma1NoCT of chest, abdomen, pelvis; endoscopic sonographyEndoscopic sonographically guided biopsy, surgery deferred
Metastatic adenocarcinoma from unknown primary source1NoCT of chest, abdomen, pelvis; bone scanningCT-guided biopsy; patient died
Small cell lung carcinoma
1
Yes
CT of chest, abdomen
CT-guided biopsy; patient died
Total
10 (3%)



Indeterminate lesions requiring biopsy or follow-up imaging (as noted in the MR angiography report) were detected in 18 patients (Table 3). In five patients biopsy or surgery was performed, with a resulting benign diagnosis. These included an atypical adrenal adenoma (without signal dropout on out-of-phase images), hepatic focal nodular hyperplasia, sarcoidosis in a patient with retroperitoneal adenopathy, sclerosing mesenteritis, and ovarian serous cystadenoma (Figs. 5A and 5B). In 13 patients, lesions requiring follow-up imaging were detected, including moderately complex renal cysts in four patients (follow-up MRI performed at 6 months in three patients revealed no significant change; the fourth patient had no follow-up imaging performed at our institution), ovarian cysts in two postmenopausal patients (serial pelvic sonography was performed in one patient, no additional imaging in the second patient), small pancreatic cystic lesions in five patients (two patients had follow-up MRI without significant change), a possible pancreatic lesion in one patient (no abnormality detected on follow-up CT), and an indeterminate adrenal lesion with no follow-up imaging in one patient.
TABLE 3: Indeterminate Lesions Incidentally Found on Renal MR Angiography in 380 Patients
FindingNo.Previously Diagnosed?Additional ImagingOutcome
Cystic pancreatic lesion5No6- to 12-month follow-up MRI in 2/5 patientsNo significant change in 2 patients with follow-up
Complex renal cyst4No6-month follow-up MRI in 3/4 patientsNo significant change in complex cysts
Ovarian cyst2NoPelvic sonography in 1/2 patientsSimple cyst, no significant change on 6-month follow-up sonography
Adnexal mass1NoPelvic sonographySurgery: serous cystadenoma
Adrenal mass2NoNoneSurgery in 1 patient: adenoma; no follow-up in 1 patient
Hepatic mass1NoSonographySonographically guided biopsy: focal nodular hyperplasia
Retroperitoneal adenopathy1NoCT of chest, abdomen, pelvisCT-guided biopsy: sarcoid
Mesenteric mass1NoNoneCT-guided biopsy: sclerosing mesenteritis
Possible pancreatic mass
1
No
CT
No lesion on dual-phase contrast-enhanced CT
Total
18 (5%)



Benign lesions requiring no additional investigation were detected in 207 patients, or 54% (many patients had more than one type of benign lesion). Renal cysts were by far the most common lesion, detected in 176 patients (46%). Other benign lesions included hepatic and splenic cysts, hepatic and splenic hemangiomas, adrenal adenomas, hepatic focal nodular hyperplasia, and renal angiomyolipomas (Table 4). Significant findings not involving a focal mass or lesion were detected in 32 patients, including biliary and pancreatic ductal dilatation, moderate to large pleural effusions, moderate ascites, and pulmonary atelectasis or consolidation (Table 4). Miscellaneous findings of minor importance included cholelithiasis, hepatic fatty infiltration, ventral hernia, and transplant-related lymphoceles (Table 4). Renal findings related to renovascular disease were noted in 60 patients, including renal scarring or atrophy and renal infarcts.
TABLE 4: Benign Masses and Miscellaneous Incidental Findings on Renal MR Angiography in 380 Patients
FindingNo.%
Benign masses  
   Renal cyst17646
   Hepatic cyst4311
   Adrenal adenoma195
   Hepatic hemangioma123
   Splenic hemangioma2 
   Renal angiomyolipoma2 
   Hepatic focal nodular hyperplasia2 
   Sclerosing mesenteritis1 
   Ovarian serous cystadenoma1 
   Sarcoid1 
Miscellaneous, moderate significance  
   Moderate to large pleural effusion82
   Biliary ductal dilatation82
   Pulmonary atelectasis or consolidation3 
   Pancreatic ductal dilatation3 
   Moderate ascites2 
   Moderate to severe cardiomegaly2 
   Choledocholithiasis2 
Miscellaneous, minor significance  
   Cholelithiasis246
   Hepatic steatosis123
   Nonobstructive ventral hernia2 
   Transplant-related lymphocele2 
   Gallbladder adenomyomatosis1 
   Hip effusion
1

Fig. 3A Incidental renal cell carcinoma in 75-year-old man. Coronal single-shot fast spin-echo image (A) and axial contrast-enhanced fat-saturated spoiled gradient-echo image (B) from renal MR angiography reveal focal enhancing mass in right kidney. Renal cell carcinoma was confirmed at surgery.
Fig. 3B Incidental renal cell carcinoma in 75-year-old man. Coronal single-shot fast spin-echo image (A) and axial contrast-enhanced fat-saturated spoiled gradient-echo image (B) from renal MR angiography reveal focal enhancing mass in right kidney. Renal cell carcinoma was confirmed at surgery.
Overall, 151 (40%) of 380 patients had one or more additional vascular findings not related to the renal arteries, and 221 (58%) of 380 patients had one or more additional nonvascular findings. Thirteen patients underwent biopsy or surgery for an incidental nonvascular finding not previously detected, and six patients underwent surgical or interventional treatment of nonrenal vascular lesions initially detected or worsening on renal MR angiography, resulting in 19 patients (5%) with significant changes in management after detection of incidental findings on renal MR angiography.

Additional Imaging and Discrepancies

Of the 380 renal MR angiography patients, 192 had no additional cross-sectional imaging performed within 1 year of the MR angiography examination. One hundred eighty-eight patients (49%) had one or more additional cross-sectional examinations, as follows: 129 renal duplex sonography examinations, 26 abdominal or pelvic sonography, 64 abdominal or abdominopelvic CT (38 without IV contrast material), and five abdominal MRI. These numbers include examinations performed for further evaluation of findings detected on renal MR angiography.
Forty-eight nonvascular discrepancies were noted in reports of MR angiography and other cross-sectional examinations in 45 patients. Twenty-two of these discrepancies were findings missed or not visible on MR angiography. These included 10 patients with nonobstructive renal calculi seen on sonography or CT and not visualized on MR angiography, three patients with cholelithiasis present on CT or sonography and in retrospect present on MR angiography, two patients with adrenal adenomas detected on CT and in retrospect present on MR angiography, two patients with hepatic steatosis on abdominal sonography and in retrospect present on MR angiography, two patients with probable hepatic focal nodular hyperplasia on CT not detected on MR angiography, one patient with a possible pancreatic mass on MR angiography not detected on CT, one patient with a small peritransplant fluid collection on renal sonography and in retrospect present on MR angiography, and one patient with a subcentimeter renal angiomyolipoma on renal sonography and in retrospect present on MR angiography.
Fig. 4A Incidental pheochromocytoma in 75-year-old woman. Coronal single-shot fast spin-echo (A) and axial contrast-enhanced fat-saturated spoiled gradient-echo (B) images from renal MR angiography reveal heterogeneous enhancing right adrenal mass. In-phase and out-of-phase images were not consistent with adrenal adenoma. Pheochromocytoma was found at surgery.
Fig. 4B Incidental pheochromocytoma in 75-year-old woman. Coronal single-shot fast spin-echo (A) and axial contrast-enhanced fat-saturated spoiled gradient-echo (B) images from renal MR angiography reveal heterogeneous enhancing right adrenal mass. In-phase and out-of-phase images were not consistent with adrenal adenoma. Pheochromocytoma was found at surgery.
Eleven nonvascular discrepancies represented missed or undetected findings on CT. Eight patients had renal cysts on MR angiography that were not reported on CT (unenhanced CT in five patients), one patient had an adrenal adenoma described on MR angiography and in retrospect present on CT, one patient had a dilated common bile duct in retrospect present on CT, and one patient with a splenic hemangioma reported on CT and not detected on MR angiography that in retrospect probably represented normal late arterial phase perfusion. Eleven nonvascular discrepancies represented missed or undetected findings on sonography: seven patients with renal cysts on MR angiography not reported on sonography, two patients with minimal abdominal ascites in retrospect present on abdominal sonography, one patient with a hepatic hemangioma not detected on abdominal sonography, and one patient with a renal angiomyolipoma reported on renal sonography that was not seen on MR angiography and in retrospect likely represented fat in a cortical defect. Extrarenal lesions not noted on renal sonography were not considered missed.
Thirty-eight renal vascular discrepancies in 33 patients were noted between renal duplex sonography and renal MR angiography reports. In 15 patients, multiple renal arteries were detected on MR angiography and not described on renal sonography (angiography performed in four of these patients confirmed the MR angiography reports). In 18 patients, renal MR angiography detected moderate to severe stenosis of one or more renal arteries not described on sonography (angiography performed in seven patients confirmed the MR angiography diagnosis in six cases and sonography results in one case). In five patients, significant renal artery stenosis was detected on sonography but not on MR angiography. Renal angiography performed in one case confirmed the MR angiography findings. No CT vascular discrepancies were noted, although renal vascular findings were described in only three cases, and a significant percentage of the CT examinations were performed without contrast material.
Fig. 5A Incidental ovarian serous cystadenoma in 44-year-old woman. Coronal single-shot fast spin-echo image from renal transplant MR angiography examination reveals complex right adnexal cystic lesion adjacent to transplanted kidney.
Fig. 5B Incidental ovarian serous cystadenoma in 44-year-old woman. Axial fat-saturated contrast-enhanced spoiled gradient-echo image reveals cystic adnexal lesion with small solid enhancing nodule (arrow). Serous cystadenoma was found at surgery.

Renal MR Angiography Examination Times

The 50 final renal MR angiography examinations evaluated had an average total examination time of 21 ± 9 minutes, with minimal and maximal values of 10 and 62 minutes. These numbers underestimate the true examination times because they do not include time spent on initial patient positioning or patient removal at the end of the examination. Estimation of the time per sequence is also limited by the fact that some sequences were acquired in two or more breath-holds, and therefore the true acquisition time for a particular sequence could be longer than the calculated time. The average time spent on non–MR angiography sequences was 10 ± 4 minutes, with maximal and minimal values of 4 and 26 minutes. The average percentage of the examination time occupied by non–MR angiography sequences was 57% ± 16%.

Discussion

Renal MR angiography has become a common examination. Significant renal artery stenosis can be detected noninvasively with high accuracy without exposing patients to ionizing radiation or iodinated contrast media [16]. Three-dimensional contrast-enhanced MR angiography is the key component of any renal MR angiography examination, and it is reasonable to envision a streamlined approach consisting only of a scout sequence followed by renal MR angiography. This minimizes the total examination time and allows improved patient throughput and perhaps a reduction in the cost of the examination.
On the other hand, incidental findings in renal MR angiography are common. Although most vascular lesions are adequately depicted on 3D contrast-enhanced renal MR angiography, nonvascular findings often require additional sequences for visualization and characterization. To some extent this is a reflection of choices made in our protocol: A small field of view with considerable wrap-around artifact is used to maximize spatial resolution, and only arterial phase images are acquired (both arterial and venous phase images are acquired in some protocols). Although most incidental nonvascular findings are benign, predominantly consisting of renal and hepatic cysts, these are often not completely characterized on the basis of scout images and the MR angiography sequence alone, and therefore might require additional imaging to classify as benign.
The additional sequences, then, are important not only to detect significant lesions but also to identify benign lesions and thereby avoid the expense of additional imaging. In-phase and out-of-phase spoiled gradient-echo sequences, for example, allow characterization of incidental adrenal adenomas without the need for follow-up imaging. The combination of T2-weighted single-shot fast spin-echo images and contrast-enhanced fat-saturated T1-weighted spoiled gradient-echo images allows confident identification of simple cysts. Hepatic hemangiomas could also be identified with confidence in most cases, on the basis of high signal intensity on T2-weighted images and uniform enhancement on contrast-enhanced spoiled gradient-echo images. Unenhanced T1-weighted spoiled gradient-echo images were not obtained in our protocol but could be helpful to assess for enhancement in complex lesions in the kidneys and elsewhere—for example, hemorrhagic cysts. T2-weighted images, in addition to identifying cysts and hemangiomas, also help to detect gallstones, dilated biliary and pancreatic ducts, and renal collecting system obstruction. All of the additional non–MR angiography sequences performed in our protocol are short acquisitions, obtained in one or two breath-holds.
The non–MR angiography pulse sequences acquired in the renal MR angiography protocol required a relatively small increment in additional imaging time (10 minutes on average); however, they constituted a significant percentage (57%) of the total examination time. This is a reflection of the short average examination time of 21 minutes (an underestimation of the true examination time, as noted previously).
Indeterminate or malignant lesions were often characterized with relatively little additional imaging. The incidentally discovered renal cell carcinomas, for example, proceeded to surgery without additional imaging (except unenhanced chest CT in one patient). This was also the case for the patient with pheochromocytoma. Imaging-guided biopsy was performed after renal MR angiography in one patient with lymphoma and in patients with metastatic pancreatic adenocarcinoma and metastatic adenocarcinoma from an unknown primary source. However, additional imaging was required in several patients, and biopsy or surgery was performed on benign lesions in five patients. Furthermore, in our practice, a staff radiologist, resident, or fellow is on-site and reviews all examinations before the patient is dismissed. If an indeterminate lesion or other finding is noted, additional imaging can be performed immediately, a luxury that is not available in many private practice groups and some academic centers.
The question of whether significant findings were missed or incorrectly characterized by renal MR angiography was only partially addressed in this study. In those patients with additional cross-sectional studies performed within 1 year of renal MR angiography, a relatively small number of discrepancies were noted, and many of the findings not detected on the initial MR angiography interpretation were present in retrospect. However, renal duplex sonography constituted a significant percentage of the additional cross-sectional examinations (58%), and these studies provided no information regarding extrarenal incidental findings because only the kidneys and renal arteries were evaluated.
Several studies have examined the incidence and importance of incidental findings in other abdominopelvic imaging techniques [719], most notably CT colonography and CT for suspected renal colic. Yee et al. [7], for example, examined 500 men undergoing CT colonography, finding extracolonic findings in 63% of patients and clinically significant findings in 9%. Gluecker et al. [8] evaluated 681 asymptomatic patients undergoing screening colonoscopy and concurrent CT colonography and found a 10% incidence of extracolonic findings of high importance. Xiong et al. [9] reviewed 17 CT colonography studies comprising 3,488 patients, finding that 40% of patients had extracolonic findings, 14% had further investigations, 2.7% had an extracolonic neoplasm, and 0.9% had an abdominal aortic aneurysm. Katz et al. [10] examined 1,000 consecutive unenhanced helical CT examinations performed for suspected renal colic and noted significant alternative genitourinary or nongenitourinary diagnoses established or suspected in 10% of patients.
The occurrence of incidental nonvascular findings in renal MR angiography is similar to those just noted, and the percentage with additional vascular findings is higher. This is not surprising because the CT studies did not use IV contrast material, and the patient population with suspected renovascular disease is probably more likely to have concurrent aortic, mesenteric, and iliac atherosclerosis. The relatively high percentage of patients with mesenteric artery stenosis could be an overestimation because patients with celiac artery narrowing on the basis of arcuate ligament compression were not distinguished from those with atherosclerotic narrowing, and the diagnostic accuracy of MR angiography for inferior mesenteric artery stenosis might be limited on the basis of the small size of this vessel.
Renal MR angiography has both advantages and limitations with respect to CT in identifying and characterizing incidental findings. In general, MRI has superior soft-tissue contrast, and because IV contrast material is used, unenhanced and contrast-enhanced images can be obtained to evaluate the enhancement characteristics of incidental lesions. In this regard, one might expect that additional imaging to characterize indeterminate lesions is less likely to be necessary after renal MR angiography. On the other hand, anatomic coverage is usually greater in CT colonography, which generally includes the entire abdomen and pelvis, and visualization of small pulmonary nodules in the lung bases is probably superior.
This study has several limitations. The study was retrospective and relied on the initial interpretations without additional review of the images, except in cases in which discrepancies were noted between renal MR angiography and additional cross-sectional studies. The patient population was restricted to a single institution, and it is likely that the incidence and importance of incidental findings will vary depending on the characteristics of patients undergoing renal MR angiography. Long-term follow-up of patients with indeterminate lesions is relatively limited. The question of whether significant findings in the abdomen were not detected on renal MR angiography was incompletely addressed. This is most likely to be relevant with regard to the liver, because anatomic coverage of the liver is generally incomplete in our protocol, and dynamic enhancement characteristics of hepatic lesions are not obtained. Nearly all renal MR angiography examinations in this study were interpreted by abdominal radiologists, which is not typical of most practices. The success in detecting and characterizing incidental lesions may have benefited from interpretations by radiologists with extensive experience in abdominal imaging, just as it is possible that vascular radiologists would provide an added benefit in the interpretation of vascular findings. Finally, our success in characterizing incidentally detected lesions likely benefits from reviewing all studies before the patient is dismissed and performing additional sequences as necessary.
In summary, incidental vascular and nonvascular findings on renal MR angiography are common. Benign lesions, most notably renal and hepatic cysts, are by far the most common nonvascular findings; however, a small but significant percentage of patients have malignant or indeterminate lesions. The addition of a small number of additional pulse sequences to a standard renal MR angiography protocol allows detection and characterization of most incidental abdominal lesions at a relatively small cost in imaging time. The relatively high incidence of nonvascular incidental findings accentuates the importance of performance and interpretation of these examinations by physicians with training in abdominal imaging.

Footnote

Address correspondence to J. F. Glockner ([email protected]).

References

1.
Schoenberg SO, Rieger J, Weber CH, et al. High spatial resolution MR angiography of renal arteries with integrated parallel acquisitions: comparison with digital subtraction angiography and US. Radiology 2005; 235:687-698
2.
Tan KT, Van Beek EJR, Brown PWG, Van Delden OM, Tijssen J, Ramsay LE. Magnetic resonance angiography for the diagnosis of renal artery stenosis: a meta-analysis. Clin Radiol 2002; 57:617-624
3.
Thornton J, O'Callaghan J, Walshe J, O'Brien E, Varghese JC, Lee MJ. Comparison of digital subtraction angiography with gadolinium-enhanced magnetic resonance angiography in the diagnosis of renal artery stenosis. Eur Radiol 1999; 9:930-934
4.
Leung DA, Hagspiel KD, Angle F, Spinosa DJ, Matsumoto AH, Butty S. MR angiography of the renal arteries. Radiol Clin N Am 2002; 40:847-865
5.
Fain SB, King BF, Breen JF, Kruger DG, Riederer SJ. High spatial resolution contrast-enhanced MR angiography of the renal arteries: a prospective comparison with digital subtraction angiography. Radiology 2001; 218:481-490
6.
Prince MR, Yucel EK, Kaufman JA, Harrison DC, Geller SC. Dynamic gadolinium-enhanced three-dimensional abdominal MR arteriography. J Magn Reson Imaging 1993; 3:877-881
7.
Yee J, Kumar NN, Godara S, et al. Extracolonic abnormalities discovered incidentally at CT colonography in a male population. Radiology 2005; 236:519-526
8.
Gluecker TM, Johnson CD, Wilson LA, et al. Extracolonic findings at CT colonography: evaluation of prevalence and cost in a screening population. Gastroenterology 2003; 124:911-916
9.
Xiong T, Richardson M, Woodroffe R, Halligan S, Morton D, Lilford RJ. Incidental lesions found on CT colonography: their nature and frequency. Br J Radiol 2005; 78:22-29
10.
Katz DS, Scheer M, Lumerman JH, Mellinger BC, Stillman CA, Lane MJ. Alternative or additional diagnoses on unenhanced helical computed tomography for suspected renal colic: experience with 1000 consecutive examinations. Urology 2000; 56:53-57
11.
Messersmith WA, Brown DFM, Barry MJ. The prevalence and implications of incidental findings on ED abdominal CT scans. Am J Emerg Med 2001; 19:479-481
12.
Ahmad NA, Ather MH, Rees J. Incidental diagnosis of diseases on un-enhanced helical computed tomography performed for ureteric colic. BMC Urol 2003; 3:2
13.
Osman MM, Cohade C, Fishman EK, Wahl RL. Clinically significant incidental findings on the unenhanced CT portion of PET/CT studies: frequency in 250 patients. J Nucl Med 2005; 46:1352-1355
14.
Ishimori T, Patel PV, Wahl RL. Detection of unexpected additional primary malignancies with PET/CT. J Nucl Med 2005; 46:752-757
15.
Ng CS, Doyle TC, Courtney HM, Campbell GA, Freeman AH, Dixon AK. Extracolonic findings in patients undergoing abdomino-pelvic CT for suspected colorectal carcinoma in the frail and disabled patient. Clin Radiol 2004; 59:421-430
16.
Hellstrom M, Svensson MH, Lasson A. Extracolonic and incidental findings on CT colonography (virtual colonoscopy). AJR 2004; 182:631-638
17.
Leslie JA, Prihoda T, Thompson IM. Serendipitous renal cell carcinoma in the post-CT era: continued evidence in improved outcomes. Urol Oncol 2003; 21:39-44
18.
Green DE, Woodward PJ. The management of indeterminate incidental findings detected at abdominal CT. Semin Ultrasound CT MR 2005; 26:2-13
19.
Hara AK. Extracolonic findings at CT colonography. Semin Ultrasound CT MR 2005; 26:24-27

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: 693 - 700
PubMed: 17715119

History

Submitted: October 23, 2006
Accepted: April 19, 2007

Keywords

  1. incidental findings
  2. kidney
  3. MR angiography

Authors

Affiliations

James F. Glockner
Department of Radiology, Mayo Clinic, 200 First St. SW, Rochester, MN 55905.

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options

PDF

View PDF

PDF Download

Download PDF

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media