Prognostic significance of vascular invasion in patients with bladder cancer who underwent radical cystectomy

Authors


Hideaki Miyake md, Department of Urology, Hyogo Medical Center for Adults, 13-70 Kitaohji-cho, Akashi 673-8558, Japan. Email: hideakimiyake@hotmail.com

Abstract

Abstract  Background:  The objective of this study was to determine whether vascular invasion (i.e. lymphatic and blood vessel invasion) could be a useful prognostic predictor in patients with locally invasive transitional cell carcinoma (TCC) of the bladder who underwent radical cystectomy.

Methods:  This series included 114 consecutive patients undergoing radical cystectomy for primary TCC of the bladder between November 1989 and July 2003. Several clinicopathological characteristics of these patients were analyzed, focusing on the association between vascular invasion and disease recurrence after radical cystectomy.

Results:  Lymphatic and blood vessel invasions were detected in 55 (48.2%) and 33 (29.8%) specimens, respectively. Lymphatic invasion was significantly associated with pathological stage, tumor grade, lymph node metastasis, blood vessel invasion and disease recurrence, whereas blood vessel invasion was significantly related to pathological stage, lymph node metastasis, lymphatic invasion and disease recurrence. Recurrence-free survival in patients with lymphatic invasion was significantly lower than that in those without lymphatic invasion, and a similar significant difference in recurrence-free survival was observed between patients with and without blood vessel invasion. However, multivariate analysis using the Cox proportional hazards model showed that only pathological stage and lymph node metastasis could be used as independent predictors for disease recurrence after radical cystectomy.

Conclusions:  Despite a significant association between several prognostic parameters, vascular invasion was not an independent predictor of disease recurrence; therefore, if there are other conventional parameters available, there might not be any additional advantage to considering the presence of vascular invasion when predicting the prognosis of patients undergoing radical cystectomy for TCC of the bladder.

Introduction

Radical cystectomy has been regarded as the most common therapeutic method for the treatment of locally invasive transitional cell carcinoma (TCC) of the bladder, with an overall 5-year disease-free survival rate of 50–70%.1 However, muscle invasive TCC is an aggressive malignancy that is widely believed to have a high propensity for distant metastasis.2 In fact, recurrence develops in approximately 50% of patients, without evidence of disease after surgery.2,3 Because of the variability in the clinical behavior of TCC of the bladder after radical cystectomy, reliable factors predicting the biological potential of this malignancy are required to select candidates likely to benefit from adjuvant therapies. Therefore, a number of studies have analyzed the clinical outcomes for patients with bladder cancer who underwent radical cystectomy; these studies have identified several useful prognostic factors, such as pathological stage and lymph node metastasis.1,4–8

Although vascular invasion (i.e. lymphatic and/or blood vessel invasion) is an indispensable step in the metastasis of cancer cells, data concerning the prognostic significance of vascular invasion, as detected by histopathological examination in radical cystectomy specimens, remain controversial in the literature.9–16 Accordingly, we retrospectively analyzed data from 114 consecutive patients with bladder cancer who underwent radical cystectomy to determine whether vascular invasion could be a useful predictor of disease recurrence after surgery.

Patients and methods

We included a total of 114 consecutive patients who underwent radical cystectomy with curative intent for primary TCC of the bladder at the Hyogo Medical Center for Adults, Kobe, Japan, between November 1989 and July 2003. Bladder cancer was histologically diagnosed by transurethral resection. Physical examination, laboratory studies, chest radiography, and intravenous pyelogram (IVP) were performed in all patients. Computerized tomography (CT), magnetic resonance imaging, and/or abdominal ultrasonography were used for clinical staging. The tumor stage and grade were examined according to the TNM classification and the World Health Organization system, respectively.17,18 All pathological examinations were performed by a single pathologist.

The indication for radical cystectomy included muscle-invasive bladder cancer, carcinoma in situ of the bladder (refractory to intravesical therapy), and recurrent multifocal, high-grade, superficial bladder cancer (refractory to repeat transurethral resection). The surgical procedures for radical cystectomy remained unchanged during the study period. In men, the bladder, prostate and seminal vesicles were resected, whereas in women, the bladder, anterior vaginal wall and uterus were removed. Urethrectomy was performed in cases diagnosed as having histologically proven TCC of the prostate and/or urethra before radical cystectomy. Standard pelvic lymphadenectomy including the obturator and iliac nodes were performed for all patients. In this series, 30 (26.3%) of the 114 patients received cisplatin-based combination chemotherapy perioperatively, including seven before surgery, 25 after surgery and two before and after surgery. The patients were initially seen 2 months after surgery, and then every 3 months for 2 years and every 6 months until disease progression or death. Follow-up examinations consisted of laboratory studies, urinary cytology, chest radiography, IVP and abdominal and/or pelvic CT, and the interval between examinations was determined by the potential risk of disease progression in each patient.

In the present series, all slides of the bladder were carefully examined with respect to the presence of lymphatic and blood vessel invasion without knowledge of the subsequent postoperative disease course. Presence of cancer cells in vessels with an unequivocal endothelial lining were considered lymphatic invasion. When cancer cells were detected within vessels with a thick vascular wall and red blood cells in the lumen, the finding was considered blood vessel invasion.

All survival data were calculated using the Kaplan–Meier method, and the difference was determined using the log-rank test. The prognostic significance of some factors was assessed using the multivariate Cox proportional-hazards regression model. Differences between the two groups were compared using the Fisher's exact test, chi-square test or unpaired t-test. P-values < 0.05 were considered significant.

Results

Table 1 shows characteristics of the 114 patients included in the present study. Lymphatic and blood vessel invasion were detected in 55 (48.2%) and 33 (29.8%) radical cystectomy specimens, respectively. The relationships between vascular invasion and several clinicopathological factors were then examined. As shown in Table 2, lymphatic invasion was significantly associated with pathological stage, tumor grade, lymph node metastasis, blood vessel invasion and disease recurrence. Similarly, blood vessel invasion was significantly related to pathological stage, lymph node metastasis, lymphatic invasion and disease recurrence.

Table 1.  Patient characteristics
Characteristicn
  • Data are presented as mean/median values.

Patient age at radical cystectomy (year, range)68.2/69.0 (40–84)
Observation period (months, range)47.3/66.5 (6–168)
Gender (%)
 Male91 (79.8)
 Female23 (20.2)
Pathological stage (%)
 pT1 or less34 (29.8)
 pT227 (23.7)
 pT335 (30.7)
 pT418 (15.8)
Lymph node metastasis (%)
 Negative93 (81.6)
 Positive21 (18.4)
Grade
 G10 (0)
 G226 (22.8)
 G388 (77.2)
Lymphatic invasion (%)
 Negative59 (51.8)
 Positive55 (48.2)
Blood vessel invasion (%)
 Negative81 (71.1)
 Positive33 (28.9)
Concomitant carcinoma in situ (%)
 Negative93 (81.6)
 Positive21 (18.4)
Table 2.  Association between vascular invasion and several clinicopathological factors
 Lymphatic invasionBlood vessel invasion
 NegativePositiveP-valueNegativePositiveP-value
Age (years)
 <7028300.4543150.46
 ≥703125 3818 
Gender
 Male46450.6067240.23
 Female1310 14 9 
Pathological stage
 pT1 or less27 7<0.00132 2<0.001
 pT21710 22 5 
 pT3 827 1619 
 pT4 711 11 7 
Lymph node metastasis
 Negative5835<0.0017815<0.001
 Positive 120  318 
Grade
 G1 0 0<0.005 0 00.21
 G220 6 21 5 
 G33949 6028 
Lymphatic invasion
 Negative52 7<0.001
 Positive 2926 
Blood vessel invasion
 Negative5229<0.001
 Positive 726  
Concomitant carcinoma in situ
 Negative50430.3767260.62
 Positive 912 14 7 
Disease recurrence
 Negative5436<0.0016921<0.05
 Positive 519 1212 

In the present series, postoperative recurrence was recognized in 24 (21.1%) of the 114 patients, and 22 of these 24 died of recurrent disease. The 1, 3, 5 and 10-year recurrence-free survival rates of the 114 patients were 84.7%, 72.1%, 70.6% and 70.6%, respectively. We then compared recurrence-free survival according to findings of vascular invasion in the surgical specimen. As shown in Figure 1, recurrence-free survival in patients with lymphatic invasion was significantly lower than that in those without lymphatic invasion. Furthermore, recurrence-free survival in patients with blood vessel invasion was also significantly lower than that in those without blood vessel invasion. We subsequently analyzed the effects of pathological stage on recurrence-free survival according to the presence or absence of vascular invasion. As shown in Table 3, there were no significant differences in recurrence-free survival rates in patients with identical pathological stage, irrespective of the presence or absence of lymphatic and blood vessel invasion.

Figure 1.

Comparison of recurrence-free survival of patients with bladder cancer undergoing radical cystectomy according to vascular invasion using the Kaplan–Meier method. (a) The recurrence-free survival rate in patients with lymphatic invasion in radical cystectomy specimens was significantly lower than that in those without lymphatic invasion (< 0.001, log-rank test). •, without lymphatic invasion; ○, with lymphatic invasion. (b) The recurrence-free survival rate in patients with blood vessel invasion in radical cystectomy specimens was significantly lower than that in those without blood vessel invasion (< 0.001, log rank test). •, without blood vessel invasion; ○, with blood vessel invasion.

Table 3.  Five-year recurrence-free survival rate according to pathological stage, in relation to the presence or absence of vascular invasion
Pathological
stage
Lymphatic invasionBlood vessel invasion
NegativePositiveP-valueNegativePositiveP-value
pT1 or less90.2%74.3%0.3991.1%70.2%0.36
pT288.3%70.5%0.4683.6%68.8%0.31
pT359.2%48.5%0.4258.1%44.5%0.43
pT433.4%18.3%0.2935.1%14.7%0.18

To analyze the degree of contribution to disease recurrence after radical cystectomy, multivariate analysis using the Cox proportional-hazards regression model was performed. As shown in Table 4, pathological stage and lymph node metastasis were independently significant for predicting disease recurrence; however, the remaining factors, including lymphatic and blood vessel invasion could not be considered independent predictors of disease recurrence.

Table 4.  Results of multivariate analysis to determine the association of various clinicopathological factors with recurrence-free survival in patients with bladder cancer who underwent radical cystectomy
VariablesHazard ratio95% CIP-value
  1. CIS, carcinoma in situ.

Age (years) (<70 vs≥70)1.560.59–4.070.37
Gender (male vs female)0.570.24–1.340.20
Pathological stage (pT2 or less vs pT3 or more)0.230.07–0.76<0.05
Lymph node metastasis (negative vs positive)0.180.06–0.51<0.005
Grade (G2 vs G3)0.330.08–1.320.12
Lymphatic invasion (negative vs positive)1.220.32–4.680.76
Blood vessel invasion (negative vs positive)0.690.26–1.790.44
Concomitant CIS (negative vs positive)0.970.23–4.080.97

Discussion

Metastasis, a major cause of mortality in cancer patients, consists of a series of events during which cancer cells detach from the primary tumor by invading the surrounding tissues, enter the circulatory system by penetrating lymphatic and/or blood vessels, and exit the vessels at distant organs to form secondary tumors;19 this suggests that vascular invasion of tumor cells at a primary organ would be an essential step for metastatic formation. However, the prognostic significance of vascular invasion in patients with bladder cancer who have undergone radical cystectomy remains controversial; therefore, we retrospectively analyzed the clinicopathological data of 114 consecutive patients undergoing radical cystectomy for bladder cancer in a single institution, focusing on the prognostic significance of vascular invasion in radical cystectomy specimens.

In the present series, lymphatic and blood vessel invasions were detected in 48.2% and 29.8% of radical cystectomy specimens, respectively, and both of these two factors were significantly related to other prognostic factors. Furthermore, there were significant differences in recurrence-free survival between patients with and without vascular invasion. However, because of the relatively small number of patients in each subgroup, recurrence-free survival in patients with identical pathological stage was not affected by the presence or absence of vascular invasion. Multivariate analysis showed that only pathological stage and lymph node metastasis, but not vascular invasion, could be used as independent predictors of disease recurrence. These findings suggest that pathological stage and lymph node metastasis have a more powerful impact on the prognosis of patients undergoing radical cystectomy than vascular invasion, although these three factors might be closely associated with each other. In fact, the incidence of vascular invasion was frequently observed in accordance with an increase in pathological stage and lymph node metastasis.

As described above, these findings, particularly the prognostic significance of vascular invasion, are not always consistent with previously reported outcomes analyzing the significance of vascular invasion in radical cystectomy specimens.4,9–16,20 Leissner et al. and Honda et al. reported that vascular invasion could be used as an independent prognostic predictor for patients undergoing radical cystectomy,9,20 whereas other investigators could not find or did not address the prognostic significance of vascular invasion using multivariate analysis.4,10–16 These contradictory results might be the result of differences in the background of patients included in each study. For example, the frequency of adjuvant chemotherapy following radical cystectomy varied, although Logothetis et al. demonstrated that patients with vascular invasion were the only subgroup without a survival advantage when treated with adjuvant chemotherapy.15 Considering these findings, to definitively prove the prognostic significance of vascular invasion it will be necessary to analyze the data of a homogeneous group, consisting of a larger number of patients who underwent radical cystectomy.

Accurate diagnosis of tumor infiltration in small vessels is not always easy, particularly in small vessels located in the subepithelial connective tissue. Larsen et al. reported that immunoperoxidase staining of endothelial cells was necessary to differentiate true lymphatic invasion from peritumoral retraction artifacts in bladder cancer,21 whereas Leissner et al. concluded that careful examination of routine staining using hematoxylin and eosin is sufficient to evaluate vascular invasion.9 The present outcome supports the latter view, because of the marked difference in recurrence-free survival between patients with and without vascular invasion. This view has also been supported by experts on urothelial cancer, who suggest that the presence or absence of vascular invasion should be able to be reported without the need for confirmation using specialized staining techniques.22

In conclusion, vascular invasions (i.e. lymphatic and blood vessel invasion) were well correlated to other prognostic factors in patients with TCC of the bladder who were treated with radical cystectomy, and the prognoses of patients with vascular invasion in radical cystectomy specimens were significantly poor compared to those without vascular invasion. However, vascular invasion could not be considered an independent predictor for disease recurrence; therefore, if there are other conventional prognostic parameters available, there might not be any additional advantage to considering the presence of vascular invasion when predicting the prognosis of patients undergoing radical cystectomy for TCC of the bladder.

Ancillary