Int J Clin Oncol DOI 10.1007/s10147-015-0817-4

ORIGINAL ARTICLE

Fluorodeoxyglucose positron emission tomography/computed tomography for diagnosis of upper urinary tract urothelial carcinoma Seiji Asai1 · Tetsuya Fukumoto1 · Nozomu Tanji1 · Noriyoshi Miura1 · Masao Miyagawa2 · Kenichi Nishimura1 · Yutaka Yanagihara1 · Akitomi Shirato1 · Yuki Miyauchi1 · Tadahiko Kikugawa1 · Masayoshi Yokoyama1 

Received: 9 October 2014 / Accepted: 3 March 2015 © Japan Society of Clinical Oncology 2015

Abstract  Background  The purpose of this study was to assess the ability of fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) to detect upper urinary tract urothelial carcinomas (UTUC) compared with pathological examination of tissues obtained by ureteroscopic biopsy and split cytologic analysis of urine obtained after retrograde pyelography. Methods  Clinicopathological records of patients at our institution were retrospectively reviewed. Fifty patients with clinically suspected UTUC, who were histologically diagnosed by nephroureterectomy, partial ureterectomy, or endoscopic biopsy, were included. The patient cohort included 42 men and 8 women, with a median age of 73 (range 54–92) years. Results  Only 27 % of 49 patients with UTUC had positive voided urine cytology, and 33 % of 40 patients had positive split urine cytology. In addition, 40 % of 10 patients had a positive endoscopic biopsy. However, 83 % of 48 patients with UTUC had positive results from FDG-PET/

CT examination. The positive predictive value of FDGPET/CT was 95 %. There were no correlations between sensitivity and tumor stage or tumor grade. Sensitivity of FDG-PET/CT for patients with and without diabetes mellitus was 60 and 89 %, respectively. Conclusions  These preliminary results from a small number of patients revealed that FDG-PET/CT enabled effective detection of UTUC.

* Seiji Asai [email protected]

Akitomi Shirato [email protected]‑u.ac.jp

Tetsuya Fukumoto [email protected]

Yuki Miyauchi [email protected]

Nozomu Tanji [email protected]‑u.ac.jp

Tadahiko Kikugawa [email protected]‑u.ac.jp

Noriyoshi Miura [email protected]

Masayoshi Yokoyama [email protected]‑u.ac.jp

Masao Miyagawa [email protected]‑u.ac.jp

1



Department of Urology, Ehime University Graduate School of Medicine, Shitsukawa, Toon, Ehime 791‑0295, Japan

Kenichi Nishimura [email protected]

2



Department of Diagnostic and Therapeutic Radiology, Ehime University Graduate School of Medicine, Toon, Ehime 791‑0295, Japan

Yutaka Yanagihara [email protected]‑u.ac.jp

Keywords  FDG-PET/CT · Upper urinary tract urothelial carcinoma · Urine cytology · Ureteroscopy

Introduction Upper urinary tract urothelial carcinomas (UTUC) account for approximately 5–10 % of all urothelial carcinomas, with an annual incidence in Japan of 1.5 cases per 100,000 population [1, 2]. Anatomically, UTUC include tumors located within the ureter, renal pelvis, and calyces. Cytological or

13



pathological confirmation is needed to proceed to nephroureterectomy. Retrograde pyelography (RP) with split cytological study or endoscopic biopsy is often performed to confirm diagnosis of UTUC. However, the accuracy and sensitivity of urine cytology for detection of UTUC are quite disappointing. As a result of technical improvements, including miniaturization and increased scope flexibility, ureteroscopy enables investigation of the entire upper urinary tract (UUT) which increases detection of lesions not revealed by a filling defect on RP [3]. In addition, coupled with biopsies, ureteroscopy enables more accurate diagnosis than is possible by conventional urine cytology analysis [4]. Despite these advantages, the accuracy of ureteroscopic biopsy is still inherently limited, because of inadequate tissue volume, absence of papillary fronds, presence of crush artifact, and unrepresentative tumor sampling, all of which may obscure histopathological diagnosis. According to some reports, an estimated 20 % or more of ureteroscopic biopsies may be nondiagnostic [5, 6]. Further cause for concern is the invasive nature of these examinations, which are associated with such complications as pyelonephritis and ureteral injury, and, in particular, because general anesthesia is usually used for ureteroscopy. Positron-emission tomography (PET) with 18F-fluorodeoxyglucose (FDG) has developed rapidly and is becoming widely used in clinical oncology. For many cancers, including lung, breast, colon, and lymphoma, it is useful for diagnosis, staging, and prognosis [7]. In contrast, use of FDG/ PET in urological oncology has developed more slowly, because the radiotracer is excreted into the urine, making structures and original tumors difficult to see against the tracer background. Recently, efforts have been made to eliminate FDG excreted into the bladder more quickly, by forced diuresis [8] or bladder irrigation [9, 10], and PET has been coupled with computed tomography (CT) [8] to reduce the incidence of false-positive results. The objectives of this study were to evaluate the diagnostic accuracy of FDG-PET/CT for patients with UTUC, in comparison with pathological examination of tissues obtained by ureteroscopic biopsy and split cytologic analysis of urine samples obtained after retrograde pyelography (RP). Patients and methods All available clinicopathological data for 50 patients with clinically suspected UTUC who were examined by FDGPET/CT and subsequently histologically diagnosed by nephroureterectomy, partial ureterectomy, or endoscopic biopsy at our institution from September 2010 to March 2014 were retrospectively reviewed. Patient characteristics are listed in Table 1. The patient cohort consisted of 42 men and 8 women, with a median age of 73 (range 54–92) years. Of the 50 total patients, 11 had already received

13

Int J Clin Oncol

appropriate medication for diabetes mellitus (DM), for example insulin or oral medications, and 12 patients had serum HbA1c levels of more than 6.2 % on FDG-PET/CT examination. The tumor originated in the renal pelvis for 25 patients (50 %) and in the ureter for 25 (50 %) (U1: 6, U2: 6, U3: 13). The patients underwent a standard diagnostic evaluation with RP (43 patients) or urine cytologic analysis (50 patients). Cytology results were obtained from voided urine (49 patients) and upper tract washings (40 patients). Endoscopic examination was performed under general anesthesia with a flexible or rigid ureterorenoscope for 14 patients. Follow-up ureterorenoscopic biopsy was not performed in four of those patients, however, because obvious tumors were visible. Under direct vision, the tumor

Table 1  Patient characteristics Characteristic Age (years)  Median  Range Gender  Male  Female Diabetes mellitus Tx  −  + Tumor laterality  Right  Left Primary site  R  U1  U2  U3 Pathological stagea  pTis  pTa  pT1  pT2  pT3  pT4 Pathological gradea  G1  G2  G3

No. of patients (%)

73 54–92 42 (84) 8 (16) 39 (78) 11 (22) 18 (36) 32 (64) 25 (50) 6 (12) 6 (12) 13 (26) 3 (6) 11 (22) 9 (18) 4 (8) 17 (34) 4 (8) 2 (4) 23 (46) 23 (46)

R renal pelvis, U1 upper ureter, U2 middle ureter, U3 lower ureter, Tx appropriate medications for diabetes mellitus a

  One patient diagnosed as IgG4-related ureteritis and 1 patient diagnosed as ureteral amyloidosis by final pathological examination were omitted

Int J Clin Oncol

was grasped by use of 3 Fr. grasping forceps, and the tissue was pulled out with the scope. FDG-PET/CT imaging was performed by use of a multislice PET/CT camera (discovery PET/CT 600; GE Healthcare, Milwaukee, WI, USA). All patients fasted for a minimum of 6 h with a blood glucose level of 80–120 mg/dl before intravenous administration of 18 F-FDG. Approximately 500 ml water was given to promote excretion of FDG. A whole-body image was obtained exactly 90 min after intravenous administration of 222–370 MBq (6–10 mCi) 18F-FDG. The PET emission images were corrected for measured attention and reconstructed by use of an ordered-subset expectation maximization iterative algorithm in accordance with the manufacturer’s instructions. Integrated PET and CT images were reviewed on advantage workstations (GE Healthcare). The display field of view was 60 × 60 cm, which consisted of 192 × 192 matrixes. Voxel size was 3.125 × 3.125 × 3.27 mm3. On the axial fusion images of PET and CT, a volumetric region of interest encompassing the entire tumor was drawn to ensure correct identification of the maximum counts, and maximum standardized uptake value (SUVmax) was calculated. Two experienced radiologists given such clinical information as suspicious diagnosis and the diseased site independently interpreted FDG-PET/CT images. If there was a discrepancy between them, they discussed the interpretation by use of images, with dynamic CT or MRI if applicable, and obtained a final diagnosis. FDG-PET/CT examination was usually performed after RP. However, it was performed without such conventional examinations as RP and ureterorenoscopy for 5 patients. Final pathological diagnosis was performed with the specimen obtained from nephroureterectomy for 48 cases, partial ureterectomy for one case, and endoscopic biopsy for one case. The patient who received a partial ureterectomy was preoperatively suspected of ureteritis. The patient who received only an endoscopic biopsy was diagnosed with ureteral amyloidosis on endoscopic examination and did not receive subsequent open surgery. In addition, lymphadenectomy in conjunction with nephroureterectomy was performed for 28 patients. Pathologic evaluation and specimen grading were in accordance with the general rules of the Japanese Urological Association and the Japanese Society of Pathology for clinical and pathological studies on renal pelvic and ureteral cancer [11]. Tumor stage was determined by use of TNM criteria, which quantify the extent of the tumor (T), involvement of nearby lymph nodes (N), and the presence of metastasis (M). Histologic evaluation of the specimens revealed IgG4-related ureteritis for one patient (2 %), ureteral amyloidosis for one patient (2 %), and urothelial carcinoma grade 1 for two patients (4 %), grade 2 for 23 patients (46 %), and grade 3 for 23

patients (46 %). Pathologic stage was pTis for three, pTa for 11, pT1 for nine, pT2 for four, pT3 for 17, and pT4 for four patients. The Mann–Whitney U test or the chi-squared test was used to assess the association between clinicopathological features and FDG-PET/CT findings. Binary logistic regression analysis was performed in the multivariate analysis. A p value of  6.2 7 eGFR (median) 50.9 Tumor laterality  Right 14  Left 26 cT