Bladder cancer after managing upper urinary tract transitional cell carcinoma: predictive factors and pathology

Abstract

OBJECTIVE

To evaluate patients with a history of transitional cell carcinoma (TCC) of the upper urinary tract (UUT) to determine the incidence, pathological distribution, and risk factors for developing subsequent bladder tumours.

PATIENTS AND METHODS

Between 1993 and 2003, 103 patients were treated at our institution for UUT-TCC. We reviewed demographic, clinical, surgical, and pathological data from these patients at a median follow-up of 38.7 months, and used univariate and multivariate analyses with logistic regression modelling to determine prognostic variables for bladder recurrences.

RESULTS

In all, 51 (49.5%) patients developed bladder tumours after treatment for UUT-TCC, at a mean interval of 13.2 months. Patient age (P = 0.01), UUT tumour size (P = 0.03), UUT tumour multifocality (P = 0.05), a history of bladder tumours (P = 0.03), and the number of previous bladder tumours (P = 0.05) predicted the development of bladder recurrences on univariate analysis. On multivariate analysis, only a previous history of bladder tumours (odds ratio 2.6, P = 0.05) remained significant. Over 90% of the recurrent bladder tumours were superficial, with two-thirds of these being low to moderate grade. Six patients had muscle-invasive disease, and five had a cystectomy.

CONCLUSION

Bladder tumours occurred in half the patients after treatment for UUT-TCC; >60% of these subsequent bladder tumours were superficial, low- to moderate-grade lesions. Neither the pathology of the UUT tumours nor the method of treatment for the UUT disease was associated with recurrent bladder tumours. Only a history of bladder cancer predicted the development of subsequent bladder tumours.

Abbreviations

UUT

upper urinary tract

OR

odds ratio.

INTRODUCTION

It is estimated that 50 000–60 000 new cases of bladder cancer are diagnosed annually [1]; TCC accounts for >90% of these bladder cancers and 70–80% of TCCs of the bladder present as superficial (not muscle-invasive) tumours (Ta, Tis and T1) that are associated with a high risk of recurrence (70%), but a low risk of progression (10–20%) [2]. TCC of the upper urinary tract (UUT) is much less common, accounting for ≈ 5% of all urothelial malignancies, and <10% of renal tumours [3]. In contrast to TCC of the bladder, urothelial tumours of the UUT tend to present at a more advanced stage [4].

Transitional epithelial cells line the inner surface of the urinary tract; consequently, TCC has been suggested to be a ‘field-change’ disease, with the entire urothelium at risk of developing subsequent tumours [4]. After developing UUT-TCC, subsequent bladder tumours have been reported to occur in 15–50% of cases [5–10]. The prognostic variables for bladder cancer after managing UUT-TCC have been evaluated in only a few studies to date; these studies focused largely on the clinical and pathological characteristics of the UUT tumours, with no analysis of the effect of differences in the method of treatment for the UUT disease. In addition, the most recent studies evaluated East Asian populations, which may represent a different cohort of patients with TCC from those seen in practice in the USA [6,8–10].

Thus we reviewed the demographic, clinical, surgical, and pathological data from >100 consecutive patients treated over a 10-year period for UUT-TCC at our institution, reporting the incidence of subsequent bladder cancer and the predictive factors for developing these tumours. We also studied the pathological grade and stage distribution of subsequent bladder tumours to determine if they are different from de novo bladder cancer.

PATIENTS AND METHODS

Between October 1993 and June 2003, 138 patients were treated at the New York-Presbyterian Hospital for newly diagnosed UUT-TCC; 35 were excluded from the present analysis, including seven with metastatic disease at presentation, 13 with a follow-up of <1 year and 15 who were followed at an outside institution. In all, 103 patients were then available for evaluation, and the demographic information of these patients is summarized in Table 1.

Table 1.  The patients’ demographics

Characteristic	Value
* 71 cases of nephroureterectomy and 16 cases of segmental distal ureterectomy.
Median (range) age, years	70 (38–96)
N (%)
Male	65 (63)
Female	38 (37)
Smoking history:
Yes	51 (49)
No	53 (51)
Side
Right	54 (52)
Left	46 (45)
Bilateral	3 (3)
Presentation
Haematuria	84 (82)
Surveillance/incidental	13 (13)
Flank pain	5 (5)
Flank mass	1 (1)
Location of UUT tumour
Renal pelvis	52 (51)
Upper third ureter	11 (10)
Middle third ureter	10 (10)
Lower third ureter	30 (29)
Multifocal UUT tumours (>1 tumour focus)	24 (23)
Management of UUT tumours
Nephroureterectomy	71 (69)
Open	32 (31)
Hand-assisted laparoscopic	36 (35)
Laparoscopic	3 (3)
Segmental distal ureterectomy	16 (16)
Endoscopic ablation	16 (16)
Management of distal ureter (87)*
Open w/bladder cuff	72 (83)
Cystoscopic	12 (14)
Laparoscopic	3 (3)
History of bladder cancer before UUT-TCC	21 (20)

The diagnosis of UUT-TCC was established by CT, excretory urography, a retrograde ureteropyelogram, and/or ureteroscopy with tissue biopsies. UUT tumours were classified into four locations: renal pelvis, and the upper, middle, and lower ureter. The upper ureter was defined as the site between the PUJ and the iliac crest, the middle ureter as between the iliac crest and the lower brim of the sacrum, and the lower ureter between the lower brim of the sacrum and the vesico-ureteric junction. The management of UUT tumours consisted of ureteroscopic laser tumour ablation (16 patients), segmental distal ureterectomy (16), laparoscopic nephroureterectomy (three), hand-assisted laparoscopic nephroureterectomy (36), or open nephroureterectomy (32). When applicable, the distal ureter was removed either by open bladder-cuff resection (in 72), cystoscopic transurethral unroofing (12), or laparoscopic excision (three) by techniques previously described [11].

Tumours were staged according to the TNM classification [12] and graded using the 1998 WHO classification [13]. The pathological grade and stage distributions of the UUT tumours are listed in Table 2.

Table 2.  The pathology of the UUT tumours

Variable	Patients, n (%)
* Pathological size was not available for 16 cases managed endoscopically.
Stage
pTa	58 (56)
pTis	4 (4)
pT1	15 (15)
pT2	15 (15)
pT3	11 (11)
pT4	0
Superficial (Ta + Tis + T1)	77 (75)
Invasive (T2 + T3 + T4)	26 (25)
Grade
I	20 (19)
II	44 (43)
III	39 (38)
Size (n = 87)*, cm
Median,	3
<2	30 (34)
2–5	47 (54)
>5	10 (12)

The median (range) follow-up after surgery was 38.7 (13–135) months; the follow-up consisted of interval history and physical examination, urinary cytology, chest X-ray, and abdominopelvic CT. Surveillance cystoscopy was performed every 3 months for the first 2 years, every 6 months for the next 2 years, and yearly thereafter.

In all, 19 potential risk factors for recurrent bladder tumours were evaluated; 13 of these variables are listed in Tables 1 and 2 and the remaining six include: the time delay from initial diagnosis to definitive therapy (nephroureterectomy or segmental ureterectomy); surgical margin status; and the stage, grade, multifocality and number of previous bladder tumours. Univariate analysis using the chi-square log-rank test and multivariate analysis using Cox's proportional hazards regression model were used. Only variables significant by univariate analysis were considered for the multivariate analysis. Fisher's exact test was used to assess differences in the grade and stage distribution between UUT and bladder tumours, and the grade and stage distribution of bladder tumours depending on a previous history of bladder cancer. The unpaired Student's t-test was used to compare the number of recurrent bladder tumours depending on the previous history of bladder tumours.

RESULTS

In all, 51 of the 103 (49.5%) patients treated for UUT-TCC developed subsequent bladder tumours, at a mean (range) interval of 13.2 (1–45) months after therapy for UUT. Of these 51 patients, 42 (82%) developed an intravesical recurrence <2 years after managing their UUT tumour. Twenty-one patients had a history of superficial bladder cancer before developing UUT-TCC, and 15 of these (71%) developed subsequent bladder tumours after managing their UUT lesion. In considering the 82 patients with no history of bladder cancer, 36 (44%) developed subsequent bladder tumours after managing their UUT-TCC.

On univariate analysis, patient age (P = 0.01), UUT tumour size (P = 0.03), and UUT tumour multifocality (P = 0.05) predicted subsequent bladder tumours, while the pathological grade, stage and location of the UUT tumours did not. A history of bladder tumours (P = 0.03) and the number of previous bladder tumours (P = 0.05) also predicted the development of intravesical recurrences. Conversely, the pathological stage, grade and multifocality of the previous bladder cancer did not. Similarly, the method of treatment of the UUT tumour (endoscopic ablation vs laparoscopic nephroureterectomy vs open nephroureterectomy), the surgical management of the distal ureter (open vs cystoscopic vs laparoscopic), and the surgical margin status of patients treated with ureteric resection were not associated with intravesical recurrences. (Table 3).

Table 3.  Univariate analysis of risk factors for developing bladder tumours after managing UUT-TCC

Variable	Chi-square	P
Patient:
Age (<70 vs >70 years)	6.54	0.01
Gender (M vs F)	0.31	0.58
Smoking history (Yes vs No)	0.31	0.58
Presentation (Haematuria vs other)	0.46	0.50
UUT tumour
Size (<3 vs >3 cm)	4.75	0.03
Foci (multifocal vs unifocal)	3.78	0.05
Side (L vs R)	0.55	0.46
Location:
Renal pelvis vs upper third ureter	0.03	0.88
Renal pelvis vs middle third ureter	0.24	0.63
Renal pelvis vs lower third ureter	1.47	0.23
Renal pelvis vs ureter (all)	0.69	0.41
Stage (T):
Ta vs Tis	0.01	0.94
Ta vs T1	0.02	0.92
Ta vs T2	0.01	0.94
Ta vs T3	0.01	0.94
Ta vs T4	0.01	0.94
Ta/Tis/T1 (superficial)   vs T2/T3/T4 (invasive)	0.01	0.94
Grade:
I vs II	1.9	0.17
I vs III	3.68	0.06
II vs III	0.50	0.48
Management:
Surgical technique
Open vs endoscopic ablation	0.68	0.41
Open vs laparoscopic (including   hand-assisted)	0.59	0.44
Management of distal ureter
Open bladder cuff vs cysto/lap	2.1	0.15
Delay from initial diagnosis to  nephroureterectomy, months	1.50	0.22
Margin status	0.10	0.75
Previous bladder tumours:
Yes vs No	4.96	0.03
N previous (single vs multiple)	3.80	0.05
Stage (T)
Ta vs Tis/T1	0.03	0.97
Grade
I vs II	0.09	0.77
I vs III	0.32	0.57
II vs III	0.17	0.68
Tumour foci (multifocal vs unifocal)	1.71	0.19

On multivariate analysis, only a history of bladder cancer (odds ratio (OR) 2.6, P = 0.05) remained significantly associated with developing bladder tumours after treatment for UUT-TCC. Multifocal UUT tumours almost doubled the risk (OR 1.9) of bladder recurrences, and the increase in risk was almost significant (P = 0.10; Table 4).

Table 4.  The multivariate analysis of risk factors for developing bladder tumours after managing UUT-TCC

Variable	OR	P
Previous history of bladder cancer	2.6	0.05
UUT tumour multifocality (>1 focus)	1.9	0.10
N previous bladder tumour occurrences	1.7	0.14
UUT tumour size	1.4	0.15
Age	1.1	0.35

There were 117 bladder tumours in the 51 patients with intravesical recurrences after treatment for UUT-TCC, at a median follow-up of 38.7 months, giving a mean (range) number of recurrent bladder tumours per patient of 2.3 (1–7). Patients with a history of bladder cancer tended to have more recurrences (2.9 vs 2.1, P = 0.08). Of the 117 tumours, 111 (95%) were superficial, with 73 (66%) being low- to moderate-grade (I and II) and 38 (34%) being high-grade (III). Six patients had muscle-invasive (T2) intravesical recurrences, and five had a cystectomy during the follow-up. Three of these six patients had T2 bladder cancer at their initial follow-up cystoscopy, while the remaining three had superficial disease that progressed to muscle invasion. The pathology of bladder tumours in the 15 patients with a history of bladder cancer did not differ from that in the remaining 36 with bladder recurrences in stage, grade or multifocality (Table 5).

Table 5.  Pathological stage and grade distribution of the 117 recurrent bladder tumours diagnosed in 51 patients after management of UUT-TCC

Variable	Tumours, n (%)
Overall (51 patients, 117 bladder tumours)
Stage
pTa	87 (74)
pTis	7 (6)
pT1	17 (15)
pT2	6 (5)
pT3	0
pT4	0
Superficial (Ta + Tis + T1)	111 (95)
Invasive (T2 + T3 + T4)	6 (5)
Grade
I	26 (22)
II	47 (40)
III	44 (38)
No previous history of bladder tumours (36 patients, 74  tumours)
Stage
pTa	52 (70)
pTis	5 (7)
pT1	14 (19)
pT2	3 (4)
pT3	0
pT4	0
Superficial (Ta + Tis + T1)	71 (96)
Invasive (T2 + T3 + T4)	3 (4)
Grade
I	12 (28)
II	12 (28)
III	19 (44)
History of bladder tumours (15 patients, 43 tumours)
Stage
pTa	35 (81)
pTis	2 (5)
pT1	3 (7)
pT2	3 (7)
pT3	0
pT4	0
Superficial (Ta + Tis + T1)	40 (93)
Invasive (T2 + T3 + T4)	3 (7)
Grade
I	12 (28)
II	12 (28)
III	19 (44)

Of 103 UUT tumours, 77 (75%) were superficial while 26 (25%) were stage ≥ T2; conversely, 95% of the bladder recurrences were superficial and only 5% of the bladder tumours were invasive. There was a statistically greater distribution of invasive UUT tumours than invasive bladder tumours (chi-square 24.6, P < 0.001). There was no correlation between the stage of UUT tumours and subsequent intravesical recurrences in individual patients. However, patients with high-grade UUT tumours had more than five times the risk of developing high-grade bladder recurrences than patients with low-grade UUT-TCC (OR 5.3, P < 0.01).

DISCUSSION

Bladder tumours reportedly occur after UUT-TCC in 15–50% of cases [5,7,14,15]; nevertheless, only a few studies to date have focused on potential risk factors for the development of these subsequent bladder tumours. A review of these studies shows a marked heterogeneity in identifiable risk factors. Hisataki et al.[9] reported a 35% incidence of bladder tumours at 5 years in 69 patients previously treated for UUT-TCC. In that study the pathological stage of the UUT tumour was a significant prognostic factor for subsequent intravesical recurrences. Koga et al.[8] reported a 34% incidence of subsequent bladder tumours at a median follow-up of 35 months among 85 patients surgically treated for UUT-TCC. However, in that study female gender, postoperative systemic chemotherapy, and incomplete distal ureterectomy were all identified as risk factors for developing subsequent bladder tumours. In 2003, Kang et al.[10] reported a series of 189 patients in Taiwan with UUT-TCC, and noted a 31% intravesical recurrence rate at a median follow-up of 91 months. They found that the multifocality of the UUT tumour was the only predictive variable for bladder recurrences. Most recently, Matsui et al.[6] reported a 42% intravesical tumour recurrence rate after UUT-TCC at a median follow-up of 39.7 months, and identified tumour multiplicity, pathological stage, tumour size, and surgical method as risk factors for subsequent bladder tumours.

In the present study we reviewed a contemporary series of 103 consecutive patients treated for UUT-TCC, and report a 49.5% intravesical tumour recurrence rate at a median follow-up of 38.7 months. Over 80% of patients with subsequent bladder tumours developed these lesions <2 years after managing their UUT tumour. When excluding those patients with a history of bladder cancer, the incidence of bladder tumours after UUT-TCC (44%) remain among the highest reported.

We showed that after managing UUT-TCC, patient age, UUT tumour size, UUT tumour multifocality, and a history of bladder cancer were significant predictors of intravesical recurrences by a univariate log-rank test. However, on multivariate analysis, only a history of bladder tumour was a significant independent risk factor (OR 2.6) for future intravesical recurrences. Multifocality of the UUT tumour almost doubled the incidence (OR 1.9) of future bladder recurrences, and this was nearly significant (P = 0.10). Interestingly, the method of surgical therapy used to treated the UUT tumour did not affect subsequent bladder recurrences, and nor did the management of the distal ureter.

In considering the pathology of intravesical recurrences after UUT-TCC, 95% of these tumours were superficial lesions, with two-thirds of these being low to moderate grade. This stage distribution of bladder tumours after UUT-TCC is similar to the pathological distribution of bladder tumours presenting de novo, and parallels the reported distribution of intravesical recurrences after UUT-TCC [10,16]. There was no difference in the pathology of bladder tumours between patients with a history of bladder cancer and those with newly diagnosed bladder tumours after managing UUT-TCC. Further, there was no correlation between the pathological stage of UUT tumours and the subsequent bladder tumours that developed in these patients. Interestingly, high-grade UUT tumours were five times more likely to develop high-grade intravesical recurrences than were low-grade UUT tumours.

Overall, half of the patients with UUT-TCC developed subsequent bladder tumours, at a mean interval of 13.2 months after managing their UUT lesion. Indeed, if these patients were followed for longer the recurrence rate would probably be even higher. This high rate of early bladder recurrence underscores the need for rigorous follow-up cystoscopy at regular intervals. Almost all of these bladder tumours were superficial, low- to moderate-grade lesions, and thus were amenable to endoscopic resection, as opposed to more invasive therapies. This high percentage of superficial tumours may partly reflect the aggressive cystoscopic screening programme we used in the follow-up of these patients. Our analysis failed to identify specific characteristics of UUT tumours that would allow risk stratification in surveillance for future bladder recurrences. Clearly, patients with a history of bladder cancer require close monitoring for future recurrences. It also appears that multifocal UUT tumours double the risk of intravesical recurrences, and that high-grade UUT tumours have five times the risk of developing high-grade bladder tumours. Given these results we advocate that close cystoscopic surveillance is essential in patients after managing UUT-TCC.

In conclusion, bladder tumours occurred in half of patients after treating them for UUT-TCC; >60% of these subsequent bladder tumours were superficial, low- to moderate-grade lesions. Neither the pathology of the UUT tumours nor the method of treatment for the UUT disease was associated with subsequent intravesical recurrences. Only a history of bladder cancer predicted the development of subsequent bladder tumours.

CONFLICT OF INTEREST

None declared.