| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Departments of Radiology (E.S.P., E.S.S., P.R., M.P.B.) and Surgery, Division of Urology (T.C., M.P.B.), University of Pennsylvania Medical Center, 3400 Spruce St, Philadelphia, PA 19104. From the 1997 RSNA scientific assembly. Received March 25, 1998; revision requested June 19; revision received September 8; accepted December 9. Address reprint requests to E.S.S.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Between February 1991 and September 1997, 38 patients (41 lesions) who underwent partial nephrectomy at a single institution were preoperatively evaluated with contrast material–enhanced, multiplanar, surface-coil MR imaging. Imaging findings that could affect the decision to perform partial nephrectomy were retrospectively evaluated: tumor size; tumor location; presence of pseudocapsule; suspected tumor invasion of renal sinus fat, renal collecting system, renal vein, or perinephric fat; and morphologic and physiologic status of the contralateral kidney. Correlation was made with surgical and pathologic findings.
RESULTS: Thirty-three of 41 lesions (80%) were renal cell carcinomas, five were oncocytic neoplasms (12%), two were hemorrhagic cysts (5%), and one was an angiomyolipoma (2%). Twenty-four of 41 (59%) lesions had pseudocapsules. In most cases, the perinephric fat (n = 38 [93%]), the renal sinus fat (n = 31 [76%]), and the renal collecting system (n = 39 [95%]) were correctly interpreted as being uninvolved by tumor.
CONCLUSION: Renal neoplasms amenable to partial nephrectomy can be identified and characterized with contrast-enhanced, multiplanar, surface-coil MR imaging.
Index terms: Angiomyolipoma, 81.3141 • Kidney, cysts, 81.311 • Kidney neoplasms, 81.321, 81.324 • Kidney neoplasms, MR, 81.121412, 81.121415, 81.12142 • Oncocytoma, 81.3239 • von Hippel-Lindau disease
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Imaging findings that suggest that a renal lesion can be successfully and completely removed at partial nephrectomy include tumor smaller than 3 cm; peripheral location of the tumor; lack of invasion of the renal sinus fat, perinephric fat, or renal collecting system; presence of a pseudocapsule; lack of renal venous involvement; and absence of lymphadenopathy or distant metastases. Larger tumors and lesions that are locally invasive of the renal sinus fat, renal collecting system, or perinephric fat may be removed at partial nephrectomy if there are compelling reasons (ie, limited renal function, bilateral renal malignancies, or tumor in a solitary kidney).
The intrinsic soft-tissue contrast and direct multiplanar imaging capabilities of magnetic resonance (MR) imaging make it well suited for detecting and characterizing small renal lesions, as well as for determining the extent of local tumor invasion. To determine the MR imaging characteristics of lesions in patients who subsequently underwent technically successful partial nephrectomy, we retrospectively reviewed the MR images from 38 patients (41 lesions).
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The mean patient age at the time of surgery was 60 years (age range, 28–82 years). Two patients had von Hippel-Lindau syndrome. Nine additional patients had previously undergone nephrectomy: seven for renal cell carcinoma and two for transitional cell carcinoma. The mean preoperative serum creatinine level was 1.2 mg/dL (106 µmol/L; range, 0.6–4.3 mg/dL [53–380 µmol/L]). Five patients had a serum creatinine level of 1.5 mg/dL (133 µmol/L) or greater.
Our standard protocol for evaluating a renal lesion evolved over the time encompassed by this study. We obtained a series of coronal localizer images, followed by nonenhanced T1-weighted breath-hold spoiled gradient-recalled-echo (GRE) images (100–240/2.1 and 4.2 [repetition time msec/echo time msec]; 90° flip angle; 256 x 128–192 matrix; one signal acquired; field of view, three-fourths in the phase-encoding direction and sized to the patient; 6-mm section thickness; and 1-mm section interval, with a bandwidth of ±16 or ±32 kHz). For patients unable to undergo breath-hold imaging, we employed the following T1-weighted sequence: 300–600/9–14 (effective), two signals acquired. T2-weighted images were obtained with an axial fat-suppressed, respiratory-triggered fast spin-echo sequence (3,000–9,000/102–155 [effective] [the repetition time was dependent on the respiratory rate]; 256 x 192–256 matrix; two signals acquired; other parameters similar to those for the T1-weighted GRE sequence). Breath-hold coronal, T1-weighted, fat-suppressed, spoiled GRE imaging was performed before, immediately after, and then 45 seconds after the intravenous administration of 0.1 mmol of gadopentetate dimeglumine (Magnevist, Berlex Laboratories, Wayne, NJ; or Omniscan, Nycomed, Princeton, NJ) per kilogram of body weight. Imaging parameters included 68–130/1.6–2.9; 60°–90° flip angle; 256 x 128 matrix; two signals acquired (no phase wrap); 20–26-cm field of view; 4–6-mm section thickness; and a 1-mm section interval, with a bandwidth of ±32 kHz. Finally, we obtained delayed axial fat-suppressed spoiled GRE images by using parameters similar to those used to obtain the precontrast images.
All images were retrospectively reviewed by two or more conferring staff radiologists (E.S.S., P.R., M.P.B.) with an interest in urologic imaging who were unaware of the surgical and pathologic outcomes. Retrospective MR interpretation was by consensus.
Image reviewers noted the location of the lesion, the imaging plane in which the lesion was best delineated, whether the lesion was of primarily homogeneous or heterogeneous signal intensity, and whether the lesion appeared cystic or solid. Reviewers also compared the signal intensity of the lesion with that of the normal renal cortex on T1-weighted, T2-weighted, dynamic, and delayed postgadolinium images.
Imaging findings that could affect the decision to perform partial nephrectomy were evaluated. These included tumor size in three planes; tumor location within the kidney; presence of a pseudocapsule (a thin band of fibrous tissue and compressed renal parenchyma surrounding the lesion); tumor invasion of the renal sinus fat, collecting system, renal vein, or perinephric fat; presence of lymphadenopathy; and morphologic and physiologic status of the contralateral kidney. Partial nephrectomy was accomplished by means of either wedge resection, guillotine excision (removal of a renal pole), or tumor enucleation. Imaging findings were correlated with surgical and pathologic data.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Twenty of 38 (53%) patients had an absolute indication for partial nephrectomy: bilateral neoplasms (n = 11 [27%]) or tumor in a solitary kidney (n = 9 [22%]). Bilateral synchronous renal neoplasms were accurately diagnosed in 11 of the 11 patients (100%), including both patients with von Hippel-Lindau syndrome. All 11 of these patients underwent radical nephrectomy and contralateral partial nephrectomy. Of these 11 patients, 10 (91%) had renal cell carcinomas, while one (9%) had a renal cell carcinoma and a contralateral transitional cell carcinoma. The nine additional patients (22%) for whom partial nephrectomy was absolutely indicated had previously undergone nephrectomy and subsequently developed lesions in the remaining solitary kidney.
Of the remaining 18 (44%) patients in our study, two (5%) had renal insufficiency (serum creatinine level >1.5 mg/dL [>133 µmol/L]), a relative indication for partial nephrectomy. Partial nephrectomy in the remaining 16 (39%) patients was performed on an elective basis.
Prior to MR imaging, 25 of 38 patients (66%) had undergone computed tomography (CT). Most of these examinations were performed at other institutions, and many were not performed with a dedicated renal protocol: thin-section (5-mm–section) renal imaging before and after the power-injected, bolus administration of iodinated contrast material. CT images from 16 (64%) of the 25 patients were prospectively interpreted as being diagnostic of renal cell carcinoma, and MR imaging was performed to resolve issues of staging. In the nine (36%) remaining patients, MR imaging was performed to characterize lesions detected at earlier cross-sectional imaging.
Thirteen of our 38 patients (34%) had not previously undergone CT. These were largely patients who had had prior renal cell carcinomas who were unable or unwilling to receive iodinated contrast material.
Thirty-three of the lesions (80%) were renal cell carcinomas, five (12%) were oncocytic neoplasms, two (5%) were hemorrhagic cysts, and one (2%) was an angiomyolipoma (Fig 1).
|
|
On T1-weighted images, 21 of the 41 lesions (51%) were primarily hypointense to the normal renal cortex, 15 (37%) lesions were isointense, and five (12%) were hyperintense.
On T2-weighted images, 20 of the 41 lesions (49%) were hyperintense to the normal cortex, 15 lesions (37%) were hypointense, and six (15%) were isointense.
T1-weighted fat-suppressed spoiled GRE images obtained during dynamic gadopentetate dimeglumine injection showed that 30 of 40 lesions (75%) enhanced to a lesser degree than the normal renal parenchyma. Ten of the 40 lesions (25%) were isointense to the normal cortex on these images, and no lesions were hyperintense. Dynamic imaging of one lesion was not performed.
On delayed postgadolinium images, 33 of the 41 lesions (80%) continued to enhance less than normal renal parenchyma. Six lesions (15%) remained isointense, and two lesions (5%) became hyperintense.
On both T1- and T2-weighted images, both of the hemorrhagic cysts were hyperintense and enhanced less than the surrounding renal parenchyma.
The appearance of the oncocytomas was variable. Two of the five (40%) were primarily homogeneous in signal intensity, while three of the five (60%) were heterogeneous. On T1-weighted images, three oncocytomas (60%) were hypointense, and two (40%) were isointense; on T2-weighted images, three (60%) were hyperintense, and two (40%) were hypointense. In the dynamic postgadolinium phase, three (60%) were isointense, and two (40%) were hypointense. In the delayed phase, one lesion (20%) remained isointense, while four (80%) were hypointense.
The angiomyolipoma was heterogeneous in appearance and was isointense on T1-weighted images but hypointense on T2-weighted and on postgadolinium images (Fig 1).
Twenty-four of 41 lesions (59%) exhibited a pseudocapsule (Fig 2). Pseudocapsules were identified in 20 of 33 renal cell carcinomas (61%), two of five oncocytic neoplasms (40%), one of two hemorrhagic cysts (50%), and one of one angiomyolipoma (100%). Pseudocapsules were best seen on T2-weighted images (13 of 24 [54%]) and on postgadolinium spoiled GRE images (10 of 24 [42%]). One pseudocapsule (4%) was best seen on nonenhanced T1-weighted images.
|
|
|
|
|
|
Surgery was technically successful in all cases; tumor removal was performed by means of wedge resection in 25 of 41 cases (61%), guillotine excision in 14 cases (34%), and enucleation in two cases (5%). Surgical resection margins were clear in all cases.
Follow-up imaging was available in 34 of 35 patients (97%) who had renal cell carcinomas or oncocytic neoplasms: by means of MR imaging in 31 and by means of CT in three (mean follow-up, 19.7 months; range, 1–71 months). Thirty-one of these 34 patients (91%) had no evidence of residual tumor or tumor recurrence. One patient with von Hippel-Lindau syndrome developed an ipsilateral renal cell carcinoma 31 months after partial nephrectomy, which necessitated subsequent radical nephrectomy. Another patient developed a metachronous renal cell carcinoma in the contralateral kidney that was identified 21 months following partial nephrectomy. On the third patient's initial MR image, the renal neoplasm was complicated by extensive subcapsular hemorrhage. A baseline MR imaging examination following partial nephrectomy demonstrated multifocal renal cell carcinoma, which necessitated subsequent radical nephrectomy. In retrospect, the multifocality of this patient's disease was identifiable on the initial MR image (Fig 8).
|
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Recently, investigators in several studies have examined the efficacy of partial nephrectomy for select renal neoplasms, namely, those which occur in solitary kidneys, synchronously in both kidneys, or in those patients with poor renal function. Rates of local tumor recurrence following partial nephrectomy range from 4% to 10% (6), but overall patient survival has not substantially differed from that of patients with similar-stage disease who have undergone radical nephrectomy (7–12).
These findings have stimulated interest in performing elective partial nephrectomy in patients with small renal neoplasms and normal contralateral kidneys. Arguments against this therapeutic philosophy are based on the known 7%–11% incidence of tumor multifocality and the potential of incomplete resection of the malignancy (13,14). However, the rate of local tumor recurrence following partial nephrectomy has been shown to be approximately 2% at 3–6 years and is similar to the rate of metachronous renal cell carcinoma in the contralateral kidney following radical nephrectomy (15,16). Furthermore, the rate of multicentricity for tumors 3 cm or smaller has been shown to be less than 3% (14), which makes partial nephrectomy a reasonable treatment for patients with small renal neoplasms.
CT and MR imaging are both effective modalities for staging renal cell carcinoma. MR imaging has been shown to be slightly more accurate in staging than CT, especially for the evaluation of tumor involvement of the perinephric fat, the renal vein, and the adjacent organs (17,18). US, although useful in screening for the presence of renal tumors, is limited in the staging of renal cell carcinoma because of the difficulty in assessing retroperitoneal lymphadenopathy (19).
Thin-section, helical CT can be used to accurately stage renal cell carcinoma. Although CT cannot be used to reliably differentiate stage I disease (tumor confined by the renal capsule) from stage II disease (tumor spread to the perinephric fat), this distinction has not historically been critical, as the perinephric fat is routinely excised as part of a radical nephrectomy (19). If, however, partial nephrectomy is contemplated, it is desirable to know preoperatively if the perinephric fat is invaded by cancer, as this may affect the surgical approach. Findings of one study (20) have shown MR imaging to be slightly superior to CT in making this determination.
The detection of tumor thrombus involving the renal vein and inferior vena cava excludes the possibility of a curative partial nephrectomy. In one study, CT was 78% accurate for the detection of tumor in the renal vein and 96% accurate for the detection of involvement of the inferior vena cava (21). MR imaging is generally considered to be superior to CT in the evaluation of these venous structures (17,19,22), although achieved results are highly dependent on equipment and technique. Although tumor thrombus may be suspected on conventional T1- and T2-weighted images, nonenhanced or contrast-enhanced MR angiographic sequences can be used to confirm the presence of renal venous or inferior vena caval thrombus.
Identification of small synchronous renal cell carcinomas has proved difficult. Intraoperative US is no more sensitive than preoperative CT scanning for detection of small lesions (23). Tumors smaller than 1 cm have not been reliably detected and characterized with either CT or MR imaging (24), although there were limitations in several studies with high failure rates for the radiologic detection of synchronous renal cell carcinoma. Investigators in one such study found that neither MR nor CT could reliably depict multifocality if lesions were smaller than 2 cm, although dynamic postgadolinium images were not obtained (24). In another study, radiologically undetected multifocal disease was reported in half of all nephrectomized kidneys in patients who would have been candidates for partial nephrectomy; however, the authors (25) fail to mention which radiologic studies were obtained and how they were performed.
Contrast-enhanced, surface-coil MR imaging is well suited for evaluating renal lesions in which partial nephrectomy is being considered. The direct multiplanar imaging capabilities of MR imaging make tumor size easily measurable in three planes. In particular, the relationship of polar lesions to the normal adjacent renal parenchyma can be difficult to define on axial images, whether they be CT or MR images (Fig 3). Sagittal and coronal images considerably assist in defining the relationship of polar masses to the adjacent normal renal parenchyma.
Lesion conspicuity on dynamic and delayed postgadolinium images is generally very high, owing to both the differential enhancement of tumor and normal renal parenchyma and to the inherently high soft-tissue MR imaging contrast. Tumor heterogeneity, thought to represent hemorrhage or necrosis (26), also adds to lesion conspicuity.
In our study, MR imaging had reasonable accuracy for evaluating possible involvement of perinephric fat, renal sinus fat, and the renal collecting system. MR imaging was particularly accurate in identifying the true-negative findings in the group, correctly depicting the perinephric fat (38 of 38 cases [100%]), the renal sinus fat (31 of 34 cases [91%]), and the renal collecting system (39 of 39 cases [100%]) to be uninvolved by parenchymal tumor. Among our patients, microscopic infiltration of the renal sinus fat by renal cell carcinoma was not appreciated in one patient, who thus had a false-negative diagnosis. This patient nevertheless underwent a technically successful partial nephrectomy with clear tumor margins and underwent an unremarkable follow-up MR imaging examination at 6 months.
Only one false-negative finding was generated, and that was of microscopic renal sinus fat invasion not detected on MR images. Therefore, in our patient population, the negative predictive value of MR imaging was 100% for invasion of the perinephric fat (38 of 38 cases) and for the collecting system (39 of 39 cases) and 97% (31 of 32 cases) for involvement of the renal sinus fat.
Several false-positive diagnoses were made (perinephric fat, one of 41 cases [2%]; renal sinus fat, six of 41 cases [15%]; collecting system, one of 41 cases [2%]). As with all imaging modalities, it is extremely difficult with MR imaging to determine whether malignant tissue and normal tissue are merely adjacent to one another or local invasion has occurred.
No attempt was made to define the sensitivity and specificity of MR imaging for the determination of invasion of the perinephric fat, the renal sinus fat, and the collecting system. The population chosen for this study is not appropriate for such evaluation, as patients with clear invasion of these tissues on preoperative imaging studies would generally not have been considered for partial nephrectomy unless other circumstances forced consideration of partial nephrectomy. Also, at our institution, renal lesions preoperatively evaluated with MR imaging tend to be "problem" lesions, those in which diagnosis or staging is uncertain at contrast-enhanced CT.
Twenty-four of 41 lesions (59%) studied demonstrated a pseudocapsule, or a thin, hypointense rim surrounding the lesion (Fig 2). Pseudocapsules consist of fibrous tissue and compressed renal parenchyma and have been associated with lower tumor grades. Findings of several studies have demonstrated that pseudocapsules are most often seen on T2-weighted images (27,28). Our study findings confirm this. When a pseudocapsule was present, it was best seen on T2-weighted images (13 of 24 lesions [54%]). A minority of pseudocapsules were best seen on postgadolinium spoiled GRE images (10 of 24 lesions [42%]). One pseudocapsule (one of 24 lesions [4%]) was best delineated on nonenhanced T1-weighted images.
Although most of the lesions in our series were renal cell carcinomas (33 of 41 lesions [80%]), several other lesions were also excised. Five were oncocytic neoplasms, which are considered radiologically indistinguishable from renal cell carcinomas (3,29). The oncocytomas we studied displayed no MR imaging characteristics that would have allowed them to have been reliably distinguished from renal cell carcinomas. However, all of the oncocytic neoplasms in our study displayed cellular atypia, and partial nephrectomy has been advocated as the treatment of choice (22).
Two hemorrhagic cysts were removed by means of partial nephrectomy in our study. Both of these lesions were prospectively and retrospectively suspected to be hemorrhagic cysts at MR imaging, although a small focus of renal cell carcinoma could not be radiologically excluded. In the patients with these lesions, the contribution of MR imaging to patient treatment was the confirmation that these lesions could be excised by using partial nephrectomy.
Stage III or IV renal cell carcinoma is also well depicted on MR images (22). In particular, MR imaging has advantages over CT in evaluating the renal vein and inferior vena cava for tumor thrombus (17,30). In all of our cases, the renal vein was well seen and was correctly shown to be free of thrombus.
Lymph nodes larger than 1 cm were present in one case. Because these were inferior to the renal hilum and in the ipsilateral common iliac chain, they were thought to be in a distribution atypical for metastatic renal cell carcinoma. The patient's renal cell carcinoma was successfully extirpated at partial nephrectomy, and these nodes have been stable for the ensuing 4 years.
The contrast-enhanced, surface-coil MR imaging characteristics of lesions amenable to partial nephrectomy have been described. Although our evaluation is retrospective, in many cases the patients in this study underwent partial nephrectomy because preoperative MR images suggested that the renal lesions could be safely extirpated in this fashion. Heightened interest in the urologic community in partial nephrectomy will make it increasingly important for radiologists to determine preoperatively if a lesion can be successfully removed by means of partial nephrectomy. Contrast-enhanced MR imaging seems well suited for making that determination with a reasonable degree of accuracy.
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, E.S.P., E.S.S., P.R., M.P.B.; study concepts and design, E.S.P., E.S.S., P.R., T.C., M.P.B.; definition of intellectual content, E.S.P., E.S.S., P.R., T.C., M.P.B.; literature research, E.S.P., E.S.S., P.R., M.P.B.; clinical studies, E.S.P., E.S.S., P.R., M.P.B.; data acquisition and analysis, E.S.P., E.S.S., P.R., M.P.B.; manuscript preparation, editing, and review, E.S.P., E.S.S., P.R., M.P.B.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. Pedrosa, M. R. Sun, M. Spencer, E. M. Genega, A. F. Olumi, W. C. Dewolf, and N. M. Rofsky MR Imaging of Renal Masses: Correlation with Findings at Surgery and Pathologic Analysis RadioGraphics, July 1, 2008; 28(4): 985 - 1003. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Rosen and M. D. Schnall Dynamic Contrast-Enhanced Magnetic Resonance Imaging for Assessing Tumor Vascularity and Vascular Effects of Targeted Therapies in Renal Cell Carcinoma Clin. Cancer Res., January 15, 2007; 13(2): 770s - 776s. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Roy Sr., S. El Ghali, X. Buy, V. Lindner, H. Lang, C. Saussine, and D. Jacqmin Significance of the Pseudocapsule on MRI of Renal Neoplasms and Its Potential Application for Local Staging: A Retrospective Study Am. J. Roentgenol., January 1, 2005; 184(1): 113 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. E. Rha, J. Y. Byun, S. E. Jung, S. N. Oh, Y.-J. Choi, A. Lee, and J. M. Lee The Renal Sinus: Pathologic Spectrum and Multimodality Imaging Approach RadioGraphics, October 1, 2004; 24(suppl_1): S117 - S131. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Ascenti, M. Gaeta, C. Magno, S. Mazziotti, A. Blandino, D. Melloni, and G. Zimbaro Contrast-Enhanced Second-Harmonic Sonography in the Detection of Pseudocapsule in Renal Cell Carcinoma Am. J. Roentgenol., June 1, 2004; 182(6): 1525 - 1530. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
