Upper Urinary Tract Transitional Cell Tumors Diagnosis: Role of CT- Urography

    

    

    Figure 7:  A) CTU excretory phase, axial: mass in the renal pelvis, without
signs of fat invasion; histological staging: TCC pTaG1. B) CTU nephrograpic
phase, coronal: intraluminal lesion occupying the entire upper pole calix and
infiltrating the adjacent parenchyma; histological staging: TCC pT3G2.
    
    

Discussion

Since its introduction into clinical practice, IVU was the primary imaging technique for the investigation of urinary tract [16] and it was used as first choice investigation for the patients with haematuria, when upper urinary tract TCC was suspected [17].

However, according to 2011 European Society of Urology guidelines, CTU has become the technique of choice for the UUTTCC investigation, replacing IVU [1]. Thanks to multidetector technology progress, CTU is an excellent imaging modality for urothelial neoplasms identification [2,3,18], with a sensitivity ranging between 88% and 100%, and a specificity between 93% and 100%, as reported in the literature review and meta-analysis by Chlapoutakis et al. [4]. Compared to IVU, CTU has also the advantage of being better accepted by patients, requiring only preliminary fasting without further preparations; it is also quicker and more comfortable than IVU. In addition, CTU is a panoramic examination that allows evaluating urinary tract, renal parenchyma and other abdominal structures in the same session.

In our study, CTU correctly identified 19/21 cases of TCC (TP). In 38/45 cancer free cases CTU was negative (TN), while in 7/45 cases CTU detected suspicious lesions: our FP cases have been caused by inflammatory thickening (n=5), small endoluminal mass (n=1) and inflammatory pseudotumor (n=1), radiologically indistinguishable from cancer [3]. Per patients sensitivity, specificity, diagnostic accuracy, PPV and NPV for the detection of UUT-TCC were 90.5%, 84.4%, 86.4%, 73.1% and 95% respectively. Considering data by anatomic segments according to Wang [2], our results were slightly different (sensitivity: 86.4%, specificity: 97%, diagnostic accuracy: 96.1%, PPV: 73% and NPV: 98.7%): there is particularly an increase of specificity and accuracy.

Our not optimum PPV – which confirms previous results reported in literature [7,8] - is due to chronic inflammation inducing a focal thickening of the wall of the ureter or renal pelvis; in addition several benign diseases - such as fibroepithelial polyp, nephrogenic adenoma, inflammatory pseudotumor - may mimic transitional tumor: in these cases the definitive diagnosis can be established only by histologic examination [9].

Pielography has high sensitivity and specificity (96% and 97% respectively) for UUT-TCC diagnosis, with a PPV of 87% and a NPV of 97%: these values are similar to those of CTU (sensitivity 97%; specificity 93%, PPV 79%, NPV 99%), as reported in the only published study that compared CTU and RP [6]. This procedure, furthermore, allows selective collection of a urine sample for cytological examination [14]. RP, finally, may guide urological forceps biopsy, as described by Carrafiello et al [10], as an alternative to more complex ureteroscopic biopsy, performed in surgery room.

In our experience, 52 patients underwent both CTU and RP: in this group the two imaging techniques produced the same sensitivity (85,7%) and a similar specificity (for CTU 84.4%, for RP 82.2%), as well as similar PPV (46.2% and 42.2%) and NPV (97.4% and 97.3%). In 51/52 patients the two tecniques agreed on probable malignancy (13 patients) or probable benignity (38 patients) diagnosis. The two techniques differed in a single case, with a correct diagnosis for CTU (TN) and wrong diagnosis for RP (FP) (Figure 6). Our results are lower than those reported by Cowan et al. [6], especially PPV.

If we compare our entire series (66 patients) and that of patients who underwent CTU and RP (52 patients), it emerges that only 7/21 UUT-TCCs underwent both imaging studies; the remaining 14/21 UUT-TCCs underwent only CTU, which was considered adequate - in combination with voided urinary cytology and cystoscopy - to conclude the diagnostic work-flow in 12 cases, because suspicious CTU findings - corroborated by positive or suspicious urinary cytology - were highly indicative of upper urinary cancer [7]; in one case RP was unsuccessful, as the ureteral orifice was occluded by concomitant recurrent bladder tumor, while in the other case ureteroscopic-guided biopsy of a mass smaller than 1cm, followed by laser ablation, was added. In the group of seven patients submitted to both examinations, RP - used with only diagnostic aim - did not add any useful information compared to CTU; however, in two small masses and one focal wall thickening – indeterminate at CTU – RPguided biopsy confirmed the tumor presence.

On the other hand, all 45 patients negative for UUT-TCC underwent both CTU and RP, because CTU was never considered sufficient to rule out the suspicion of malignancy. Also in this group diagnostic RP did not add any information compared to CTU; however, in 6 cases in which CTU findings were suspicious, RPguided biopsy allowed a diagnosis of tumor absence (Figure 4,5); in another case – although US-guided biopsy of a kidney mass was negative for neoplastic cell – nephrectomy was performed with inflammatory pseudotumor removal, nevertheless avoiding RNU.

Therefore, RP should be performed only when CTU is inconclusive – that is in cases of incomplete ureteric opacification or of functionally excluded and inadequately excreting kidneys, as well as if intravenous iodinated contrast medium is contraindicated [6,19]; moreover, when CT shows suspicious findings which are not corroborated by positive urinary cytology, second-line investigations (RP-guided biopsy as well as upper urinary tract endoscopy) are useful for characterization of the lesions detected by CT [10].

In spite of high CTU and RP NPV, it should be noted that Carcinoma In Situ (CIS) – neoplastic flat lesion of the urothelium and the basal membrane [1,20] - generally may not be identified with radiographic tecniques [3]; however, CIS is often associated with positive urinary cytology, which is highly sensitive (> 90%) [21,22]. If urine cytology is positive or suspicious and CTU is negative, there is need for upper urinary tract endoscopy, performing selective cytology and biopsy [5,23]. This event happened in one our case – negative at both CTU and RP (Figure 3) - in which RP-guided biopsy was not performed, since absence of abnormal findings did not provide indications for sampling site. Other two patients – with negative imaging studies (CTU and RP) and positive urinary cytology – underwent ureteroscopy that was also negative; they were both tumor-free at follow-up. Nevertheless, if CTU is negative and urinary cytology is negative or atypical, there is a very low UUT-TCC probability [7], so the use of second-line investigations is not justified and patients can undergo the clinical and instrumental follow-up.

To sum up, second line investigations should be used only in selective cases after CTU, to improve its diagnostic performances: in our study sensitivity increased from 90.5% to 95.2%, specificity from 84.4% to 100%; diagnostic accuracy from 86.4% to 98.5%; PPV from 73.1% to 100%; NPV from 95% to 97,8%.

Accurate preoperative staging is important for the proper management of the patient: CTU, compared to conventional methods (IVU and RP), allows a contemporary study of the urinary tract and of the whole abdomen, to assess the extension of disease with a single examination.

However, CTU is not reliable to differentiate between superficial tumor, in which the lamina propria of the mucosa is not invaded (Ta), and tumor extended to subepithelial connective tissue (T1) or to muscularis (T2). Moreover, there are some limits in distinction between early stage (Ta-T2) and advanced one (T3-T4) [11,15,24]: the distinction between T2 (tumor confined to the muscularis) and T3 (tumor extended beyond the muscularis) is not always feasible, since the microscopic infiltration of extra-parietal fat is not detectable with currently available imaging techniques; in addition, tumor may sometimes cause an extra-parietal inflammatory reaction, which mimics the neoplastic infiltration and causes an error of CT overstaging [15].

We retrospectively compared CT images with T parameter on the surgical specimen, and evaluated correctly tumour stage in 80% of the cases (12/15 observations): we did not recognize peri-tumoral fat tissue micro invasion in one pT3 lesion, and we overestimated two pT2 lesions because of the presence of inflammatory desmoplasia induced by the tumor; our results are slightly lower than those reported by Fritz et al. (87.8%) [15], who assessed a larger number of cases (41 UUT-TCCs in 39 patients).

Several limitations of our study have to be considered: our series is numerically limited and not homogeneous, because it includes three patient kinds (as reported in Table 1) and not every case was managed according to the same diagnostic work flow; however it is the result of a mono-centric experience, acquired through a 2 year long period. The study, likewise others reported in literature [2,3], is a retrospective data analysis, proposing to discriminate patients with- from those without neoplastic disease. Moreover, the 24 month follow-up may not be sufficient, but other published works [2,3,25] have used a shorter follow-up (12 months).

  

    
 
    
Haematuria 
    
Cystoscopy 
    
Urine    cytology 
  
  

    
+
    
-
    
+
    
-
    
C5
    
C4
    
C3
    
<C3
  
  

    
I group    (32 patients)
      
(1st event of haematuria)
    
32
    
--
    
--
    
32
    
8
    
13
    
11
    
--
  
  

    
II    group (7 patients)
      
(staging    in new diagnosis of bladder TCC)
    
7
    
--
    
7
    
--
    
5
    
2
    
--
    
--
  
  

    
III    group (27 patients)
      
(Follow-up    in prior TCC)
    
14
    
13
    
5
    
22
    
11
    
6
    
--
    
10
  
  

    
TCC:    Tumor Cell Carcinoma. 
      
 
  

Table 1:  Study population (66 patients).