URORADIOLOGY

COMPUTED TOMOGRAPHY OF PRIMARY TRANSITIONAL CELL CARCINOMA OF UPPER URINARY TRACTS* ROBERT A. BADALAMENT, M.D. WILLIAM E BENNETT, M.D. JAMES G. BOVA, D.O. PAUL R. KENWORTHY, M.D.

HENRY A. WISE, II, M.D. STEPHEN SMITH, M.D. JOHN PEREZ, D.O.

From the Department of Surgery, Division of Urology, and the Department. of Radiology, The Ohio State University; the Department of Urology, Riverside Methodist Hospital; and the Department of Urology, Mount Carmel Medical Center; and the Department of Urology, Doctors Hospital, Columbus, Ohio ABSTRACT-Preoperative computed tomography (CT) was utilized to evaluate 20 patients with primary transitional cell carcinoma of the upper urinary tracts. Of the 20 patients, 18 (90 %) had CT visualization of the tumor as either a discrete mass or local ureteral and/or renal pelvic wall thickening; 2 (10 %) had false-negative examinations. Seven of the 20 patients (35 %) had CT evidence of tumor extension demonstrated by frank tumor invasion beyond the urothelium or by perirenal pelvic and/or periureteral fat streaks. Of the 4 patients with fat streaks, 2 (50 %) had superficial tumors (TAT.& 1 had a T, (25 %) tumor, and 1 had a T3 (25 %) tumor. All 3 patients with CT findings of direct extension of tumor through the wall of the ureter or renal pelvis had T3 tumors. Among the 13 with localized noninvasive tumor on CT, 5 (38 %) had superficial tumors (TA, TIS, T,), 5 (38 %) had Tz tumors, and 3 (21%) had T3 tumors. Of the 5 patients with enlarged regional lymph nodes (~1.5 cm) on CT, 2 had tumor confirmed histologically, 2 had subsequent negative CT-guided biopsies, and 1 had a negative lymphadenectomy. Distant metastasis was discovered in 1 patient. The data suggest that when CT demonstrates direct tumor extension through the renal pelvic or ureteral wall, it is a sensitive indicator of high-stage disease. However, in the absence of this finding, CT is of limited value in staging patients with primary transitional cell carcinoma of the pyeloureteral systern.

In the past, clinical staging of patients with primary upper urinary tract transitional cell carcinomas did not have the potential impact on patient management as it does today. The majority of patients were treated by nephroureterectomy regardless of disease extent. 1,2Yet, today, accurate clinical staging may define a select group of patients with carcinomas which may be amenable to either segmental or endo*Supported by the Bremer Foundation, The Department of Surgery Research Foundation, and The Urologic Research and Development Fund.

UROLOGY

/ JULY 1992

I

VOLUME 40, NUMBER

1

scopic resection. 2,3 Additionally, the recent development of effective systemic chemotherapeutic regimens has altered the management of patients with advanced transitional cell carcinomas. Since adequate renal function is necessary for optimal chemotherapy, preoperative chemotherapy is often administered to take advantage of the function of the involved renal unit prior to nephroureterectomy.4,5 Although computed tomography (CT) permits the evaluation of retroperitoneal structures, the accuracy of this imaging modality in

71

FIGURE1. (A) Retrograde study of left ureter demonstrates obstruction by mass. (B) Ureteral wall thickening (arrow) and its eccentric lumen were seen on this CT scan done with IV contrast media. Pathologic examination of nephroureterectomy specimen revealed TA tumor. patients with primary transitional cell carcinoma of the upper urinary tracts remains uncertain.e-e The sensitivity of CT in staging patients with renal cell carcinoma is high, in most cases obviating the need for renal arteriography.lOsll The favorable experience of CT in patients with renal cell carcinoma prompted the current review of CT in 20 patients with primary transitional cell carcinoma of the upper urinary tracts. Material and Methods Between January 1985 and August 1989, 20 patients with primary transitional cell carcinomas of the upper urinary tracts were identified through retrospective review from The Ohio State University, Riverside Methodist, Mount Carmel East, and Doctors Hospitals of Columbus, Ohio. Each patient had CT evaluation reviewed by one of us (JGB) who correlated it with other radiologic studies but was blinded as to operative findings. The mean age of the patients was sixty-three years (range of 40-84 years). There were 15 male (75 %) and 5 female (25 % ) patients. The primary tumor was located in the renal pelvis in 9 patients (45 %), the ureter in 9 (45 %), and both renal pelvis and ureter in 2 (10 % ) . Tumors were graded using The Armed Forces Institute of Pathology Protocol: grade I, well-differentiated; grade II, moderately differentiated; and grade III, poorly differen-

72

TABLE

I.

Relationship of CT findings to pathologic stage

T*, Tis, CT Findings Normal or localized tumor Tumor extension by periureteral streaking Tumor extension through pyeloureteral wall Total

No.(%)

T, (%)

Te(%)

Ta(%)

13 (65)

5 (38)

5 (38)

3 (21)

.4 (20)

2 (50)

1 (25)

1 (25)

3 (15) 20 (100)

0 (0) 7 (35)

0 (0) 6 (30)

3 (100) 7 (35)

tiated.12 Two patients (10%) had grade I tumors, 7 (35 %) had grade II tumors, and 11 (55%) had grade III tumors. Using the TNM classification as proposed by The American Joint Committee on Cancer, the primary tumor category was TA in 3 (15 % ), TIS in 1 (5 % ), Tl in 3 (15%), T2 in 6 (30%), and T3 in 7 (35%) patients (Table I) .13 The patient’s treatment consisted of nephroureterectomy in 18 patients (90 %), radical nephrectomy in 1 patient (5 %), and percutaneous needle biopsy with subsequent systemic chemotherapy in 1 patient (5%). Computed tomography was done with a variety of third and fourth generation scanners. Scanning technique was individualized. In all scans oral contrast medium was used to opacify bowel loops. Most scans were obtained before

UROLOGY

/

JULY 1992

I VOLUME 40, NUMBER 1

FIGURE 2. (A) IV urogram study shows lobulated mass in right upper pole collecting system. (B) Contrastenhanced CT scan localizes this maSs (arrows) within collecting system; however, pathologic examination revealed extension into renal parenchyma consistent with T3 primary tumor.

and after intravenous contrast agents were administered. Additionally, the majority of scans were performed at 1 cm contiguous sections, and in some instances thinner sections were taken through areas of interest. Two patients had technically suboptimal scans because Z-cm serial cuts were obtained; nevertheless, tumor was identified in these patients.

Results Of the 20 patients studied, tumor was identified in 18 patients and not visualized in 2 patients. In the 2 patients with false-negative scans, one had TIS of the renal pelvis and the other had a 0.8 by 0.8 by 0.5 cm T3 tumor of the renal pelvis. In the group of patients with identifiable tumor, two distinct patterns were observed: thickening of the renal pelvis or ureteral wall or a distinct mass. Renal pyeloureteral wall thickening had attenuation values of

FIGURES. (A) Lobulated filling defect, with narrowing of right upper pole infundibulum was identified on conventional urogram. (B) Mass (arrows) was found to infiltrate upper pole renal sinus fat; however, on specimen TA tumor was found. pathologic examination of nephroureterectomy

UROLOGY

/ JULY1992 / VOLUME40,NUMBERl

73

FIGURE 4. (A) Left retrograde pyelogram demonstrated mass involving medial aspect of renal pelvis causing partial obstruction to calices in upper and mid kidney. (B) Large, locally invasive ma.ss (arrows) extending to aorta and psoas muscle is shown. Pathologic examination of nephroureterectomy demonstrated extension beyond pyeloureteral system.

soft tissue and sometimes caused eccentric positioning of the ureteral lumen (Fig. 1). Occasionally, increased density around the ureter or periureteral streaking suggestive of tumor extension beyond the ureteral wall was noted, Distinct masses were identified as tumor projections into the collecting system (Fig. 2) or tumor which appears to infiltrate the renal sinus fat or parenchyma (Fig. 3). Of the 7 patients with CT findings indicating tumor invasion, 4 patients had periureteral streaks and 3 patients had direct extension of tumor through the wall of the ureter or renal pelvis. Of the 4 patients with periureteral fat streaks, 2 (50 % ) had superficial tumors (TA and Tr), 1 (25 % ) had a Tz tumor, and 1 (25 % ) had a TI tumor. All 3 patients that had evidence of direct extension of the primary tumor beyond the urinary tract on CT had TS tumors (Fig. 4). Among the 13 patients with localized noninvasive tumor on CT, 5 (38 % ) had superficial tumors, 5 (38 %) had Tz tumors, and 3 (21%) had TI tumors. Among the 3 patients with primary renal pelvic T3 tumors without CT evidence of invasion, 1 had a false-negative CT as mentioned, the other 2 patients had maximum tumor diameters of 1.6 cm and 3 cm, respectively. CT not only detected abnormal lymph nodes in 5 patients, but also provided guidance for percutaneous biopsy in 3. Biopsy was negative in 2 patients (TIS and T, primary tumors) and 1

74

was positive (T3 primary tumor). Two other patients met the criteria for lymphadenopathy on CT (2 1.5 cm) and had formal lymphadenectomy performed. In 1 the lymph nodes were positive (T, primary tumor) and in the other, negative for metastases (T1 primary tumor). Additionally, evidence of distant metastatic disease (liver and bone) was seen in 1 patient. This patient had primary tumor invasion beyond the urinary tract, a T, primary tumor, and negative regional lymph nodes on CT. Thus, all 3 patients with N + or M + disease had direct tumor extension beyond the pyeloureteral system on CT scan, and had a T, primary tumor. Comment Transitional cell carcinoma is an uncommon tumor of the upper urinary tracts although it is the second most common primary malignancy to involve the kidneys. I4 Since 90 percent of patients (18120) were treated by nephroureterectomy, the present series demonstrates the practice of obtaining CT scan but not acting on the staging information it provides. Currently, other therapeutic options are considered acceptable for patients with low-stage and highstage disease. Limited resection may be performed in small, low-stage, distal ureteral tumors.3 Furthermore, endoscopic resection may be indicated for patients with small lowstage tumors and poor renal function, l5 Chemotherapy is now being utilized in patients with

UROLOGY

I JULY1992 I VOLUME40,NUMBERl

advanced stage disease . 5 Preoperative administration of systemic chemotherapy is often utilized so that the nephrotoxicity is distributed between two kidneys instead of one. Therefore, noninvasive preoperative evaluation with CT may enhance staging accuracy and define subsets of patients whose management may differ. Tumors were identified by CT in 18 of 20 cases (90%). As previously described by Baron et uZ.,~ CT findings varied from ureteral and renal pelvic wall thickening to a distinct tumor mass. Seven of the 20 (35 %) cases demonstrated CT findings of tumor invasion, which varied from periureteral stranding to mass lesions extending into the perirenal region. The streaks of increased density in the periureteral fat appear to be a nonspecific finding. Four patients had this finding. However, only 1 patient had tumor into the periureteral fat; the others had Tz, T1, and TA lesions. The 3 remaining patients with CT direct tumor extension outside the pyeloureteral system all had T, lesions. Additionally, lymph node enlargement should also be considered a nonspecific finding, since only 2 of 5 patients (40 % ) with enlarged lymph nodes had tumor within them. Although CT is sensitive enough to detect 90 percent of lesions in the current study, these are more commonly diagnosed by conventional studies, such as intravenous or retrograde pyelography. CT may be helpful to confirm subtle lesions when conventional studies are not definitive such as differentiating nonradiopaque stone from a papillary tumor.6,Q.16J7 In this clinical setting, noncontrast scans are necessary to detect the high attenuation coefficients of nonradiopaque stones. CT is more sensitive to slight increases of radiodensity than any other imaging method. It has been shown that a small number of transitional cell carcinomas may calcify; however, most of these calcifications are stippled or curvilinear and are associated with masses.16 None of our cases exhibited calcification. Although the specificity of CT in evaluating small nonradiopaque filling defects remains undetermined, most agree that CT is a good examination to differentiate stones from tumors. The optimal technique we have found for imaging primary transitional cell carcinomas of the upper urinary tract is 5 mm sections through the area of interest without intravenous contrast media followed by 10 mm sections throughout the abdomen with additional 5 mm sections through the area of interest after intravenous contrast media has been adminis-

UROLOGY

/ JULY 1992

/

VOLUME

40, NUMBER

1

tered. Bowel opacification is also necessary to separate bowel from tumor outside the urinary tract. In summary, our analysis shows that CT is of limited value in staging patients with primary transitional cell carcinoma of the upper urinary tract. The exception is when CT demonstrates direct tumor extension through the renal pelvic or ureteral wall. In this instance it is a sensitive indicator of high-stage tumor. Preoperative chemotherapy may be considered in this highstage group, but the efficacy of such an approach has not been clearly established. 456 West 10th Avenue Columbus, Ohio 43210 (DR. BADALAMENT) References 1. Whitmore WF Jr: Management of urothelial tumors of the upper collecting system, in Skinner DG (Ed): Urologic Cancer, New York, Grune & Stratton, Inc, 1983, pp 181-197. 2. Zincke H, and Neves RJ: Feasibility of conservative surgery for transitional cell cancer of the upper urinary tract, Urol Clin North Am 11: 717 (1981). 3. Badalament RA, et al: Prognostic factors in patients with primary transitional cell carcinoma of the upper urinary tract, J Urol 144: 859 (1990). 4. Drago JR: Metastatic ureteral carcinoma, in: Current Therapy in Genitourinary Surgery, Philadelphia, BC Decker Inc, 1987, pp 50-52. 5. Scher HI, et al: Neoadjuvant M-VAC (methotrexate, vinblastine, doxorubicin, and cisplatin) for extravesical urinary tract tumors, J Ural 139: 475 (1988). 6. Parienty RA, et aE: Diagnostic value of CT numbers in pelvocalyceal filling defects, Radiology 145: 743 (1982). 7. Kenny PJ. and Stanley RJ: Computed tomography of ureteral tumors, J Comput Assist Tomogr 1: 102 (1987). 8. Baghdassarian Gatewood OM, Goldman SM, Marshall FF, and Siegelman SS: Computerized tomography in the diagnosis of transitional cell carcinoma of the kidney, J Ural 127: 876 (1982). 9. Baron RL, McClennan BL, Lee JD, and Lawson TL: Computed tomography of transitional cell carcinoma of the renal pelvis and ureter, Radiology 144: 125 (1982). 10. Johnson CD, Dunnick NR, Cohan RH, and‘Il1escas FF: Rena1 adenocarcinoma: CT staging of 100 tumors, AJR 148: 59 (1987). 11. Fein AB, et al: Diagnosis and staging of renal cell carcinoma: a comparison of MR imaging and CT, AJR 148: 749 (1987). 12. Bennington JL, and Beckwith JB: Tumors of the Kidney, Renal Pelvis, and Ureter, Bethesda, Armed Forces Institute of Pathology, 1975, pp 266311. 13. Beahrs 0: American Taint Committee on Cancer (AJCC) Manual for Cancer Staging, 3rd ed, Philadelphia, J.B. Lippincott, 1988. 14. Peterson RO: Renal Pelvis, Urologic Pathology, Philadelphia, J.B. Lippincott Co, 1986, pp 181-278. 15. Huffman JL: Diagnostic and therapeutic approaches to upper tract urothelial tumors, in Huffman JL, Bagley DH, and Lyon ES (Eds): Ureteroscopy, Philadelphia, W.B. Saunders Co., 1988, pp 107-123. 16. Dinsmore BJ, Pollach HM, and Banner MP: Calcified transitional cell carcinoma of the renal pelvis, Radiology 167: 401 (1988). 17. Daniel WW Jr, et ~2: Calcified renal masses. A review of ten years experience at the Mayo Clinic, Radiology 3: 474 (1972).

75