Abstract
It remains unclear whether lymphadenectomy alters regional node recurrence after nephroureterectomy in patients with urothelial carcinoma of the renal pelvis. The predictive factors for regional node recurrence are still unclear. In this study, we retrospectively examined how the extent of lymphadenectomy influences regional node recurrence in patients with urothelial carcinoma of the renal pelvis.
From January 1988 through July 2013, we performed nephroureterectomy in 180 patients with non-metastatic (cN0M0) urothelial carcinoma of the renal pelvis at two Japanese institutes. Regional nodes were determined according to our previous mapping study: complete lymphadenectomy designates that all regional sites were dissected; incomplete lymphadenectomy that all sites were not dissected. A third group included those without lymphadenectomy.
The 5-year cancer-specific and recurrence-free survival was significantly higher in the complete lymphadenectomy group than in the incomplete lymphadenectomy or without lymphadenectomy groups (P = 0.03). The incidence of regional node recurrence was significantly lower in the complete lymphadenectomy group at 2.9% (2/67) than in the incomplete lymphadenectomy at 18.1% (4/22) or without lymphadenectomy at 10.9% (10/91) groups (P = 0.03). In patients with incomplete lymphadenectomy, 75% of regional node recurrence occurred outside of the dissected sites. Complete lymphadenectomy is shown to be a likely predictive factor of reduced risk of recurrence at the regional nodes by multivariate analysis, after adjusting for patient age, pathological T stage, and pathological nodal metastases.
This study shows that template-based lymphadenectomy reduced the risk of regional node recurrence in patients with urothelial carcinoma of the renal pelvis and appears to result in improved survival.
INTRODUCTION
The therapeutic benefit of lymphadenectomy in upper tract urothelial carcinoma has been examined by a variety of investigators. We have reported significant reduction in risk of cancer death by using template-based lymphadenectomy in patients with pathological Stage 3 or higher (1). Since then, there has been an accumulation of reports about improvement of patient survival by lymphadenectomy (2–5). We have recently reported results from a non-randomized prospective study showing that lymphadenectomy significantly improved cancer-specific, overall survival in patients with renal pelvic cancer (6). These results further support evidence of the therapeutic benefit of lymphadenectomy in patients with urothelial carcinoma of the renal pelvis (UCRP).
However, the reason for this improvement in patient survival after lymphadenectomy remains unclear. The lymph nodes are a major site of lymphatic metastases from urothelial carcinoma of the upper urinary tract (7,8). Incidence of local recurrence including regional node metastases has been high after radical nephroureterectomy according to previous reports (9–12). Thus, lymphadenectomy may reduce the risk of regional node metastases, which results in higher patient survival. Our prospective study also showed a lower incidence of regional node recurrence in patients with UCRP after template-based lymphadenectomy than without lymphadenectomy (6). These results, however, need to be supported by an increased number of patients. In addition, the influence of the extent of lymphadenectomy on the recurrence behaviors remains undetermined.
To further investigate the influence of lymphadenectomy on the pattern of lymphatic recurrence, we examined in detail the recurrent patterns in the regional nodes of our retrospective cohort of patients with UCRP from two institutes. We also analyzed the influence of the extent of lymphadenectomy on regional node recurrence in these patients.
PATIENTS AND METHODS
Patients
From January 1988 through July 2013, we performed radical nephroureterectomy in 210 patients with UCRP at two Japanese institutes. Of these, 180 patients with no lymph node metastasis and no visceral metastases at pre-operative radiological examination (cN0M0) were the subjects of this study.
Staging of the tumor was performed according to the UICC TNM classification (13). The pre-operative stage was determined by computed tomography (CT) scan. For multifocal tumors, patients were considered to have renal pelvic cancer when tumors of the renal pelvis were of a higher pathologic stage compared with heir ureteral tumors, or were larger in size if the tumor stage was equivalent among them. Table 1 shows patient backgrounds.
Patient characteristics according to type of lymphadenectomy in 180 non-metastatic patients with urothelial carcinoma of the renal pelvis
| CompLND | IncompLND | No-LND | P value | |
|---|---|---|---|---|
| Number of patients | 68 | 20 | 92 | |
| Mean patient age, year. (median; range) | 67.2 (67; 51–83) | 66.6 (68; 51–76) | 73.5 (75; 36–91) | <0.001 |
| Gender M/F, n | 51/17 | 15/5 | 62/30 | 0.52 |
| Mean follow-up, months (median; range) | 66.8 (47.1; 2.6–213) | 68.0 (38.1; 6.9–208) | 47.0 (26.3; 1.0–225) | 0.01 |
| No. cT stage (≤cT1/cT2/cT3/cT4) | 10/27/28/3 | 2/10/7/1 | 23/39/27/3 | 0.48 |
| No. cN stage (cN0/cN1, 2) | 68/0 | 20/0 | 92/0 | – |
| No. pT stage (≤pT1/pT2/pT3/pT4) | 22/11/32/3 | 2/3/13/2 | 30/11/44/7 | 0.20 |
| No. pN stage (pN0/pN1/pN2/pNx) | 63/2/3/0 | 13/2/5/0 | 0/3/1/88 | <0.001 |
| No. tumor grade (low/high) | 41/27 | 10/10 | 44/48 | 0.29 |
| LVI, (%) | 30 (44%) | 15 (75%) | 45 (49%) | 0.04 |
| Mean LN removed (median; range) | 11.5 (11; 4–32) | 4.6 (3; 2–16) | – | <0.001 |
| Adjuvant chemotherapy, n (%) | 7 (10%) | 1 (4%) | 1 (1%) | 0.02 |
| Recurrence, n (%) | 9 (13%) | 6 (30%) | 26 (28%) | 0.04 |
| Regional lymph node only, n (%) | 2 (3%) | 1 (5%) | 3 (3%) | 0.91 |
| Lymph node and distant organ | 2 (3%) | 3 (15%) | 9 (10%) | 0.10 |
| Distant organs only, n (%) | 5 (7%) | 2 (10%) | 14 (15%) | 0.28 |
| Cancer death, n (%) | 8 (12%) | 6 (30%) | 23 (25%) | 0.05 |
| CompLND | IncompLND | No-LND | P value | |
|---|---|---|---|---|
| Number of patients | 68 | 20 | 92 | |
| Mean patient age, year. (median; range) | 67.2 (67; 51–83) | 66.6 (68; 51–76) | 73.5 (75; 36–91) | <0.001 |
| Gender M/F, n | 51/17 | 15/5 | 62/30 | 0.52 |
| Mean follow-up, months (median; range) | 66.8 (47.1; 2.6–213) | 68.0 (38.1; 6.9–208) | 47.0 (26.3; 1.0–225) | 0.01 |
| No. cT stage (≤cT1/cT2/cT3/cT4) | 10/27/28/3 | 2/10/7/1 | 23/39/27/3 | 0.48 |
| No. cN stage (cN0/cN1, 2) | 68/0 | 20/0 | 92/0 | – |
| No. pT stage (≤pT1/pT2/pT3/pT4) | 22/11/32/3 | 2/3/13/2 | 30/11/44/7 | 0.20 |
| No. pN stage (pN0/pN1/pN2/pNx) | 63/2/3/0 | 13/2/5/0 | 0/3/1/88 | <0.001 |
| No. tumor grade (low/high) | 41/27 | 10/10 | 44/48 | 0.29 |
| LVI, (%) | 30 (44%) | 15 (75%) | 45 (49%) | 0.04 |
| Mean LN removed (median; range) | 11.5 (11; 4–32) | 4.6 (3; 2–16) | – | <0.001 |
| Adjuvant chemotherapy, n (%) | 7 (10%) | 1 (4%) | 1 (1%) | 0.02 |
| Recurrence, n (%) | 9 (13%) | 6 (30%) | 26 (28%) | 0.04 |
| Regional lymph node only, n (%) | 2 (3%) | 1 (5%) | 3 (3%) | 0.91 |
| Lymph node and distant organ | 2 (3%) | 3 (15%) | 9 (10%) | 0.10 |
| Distant organs only, n (%) | 5 (7%) | 2 (10%) | 14 (15%) | 0.28 |
| Cancer death, n (%) | 8 (12%) | 6 (30%) | 23 (25%) | 0.05 |
No., number; cT, clinical T stage; cN, clinical N stage; pT stage, pathological T stage; pN stage, pathological N stage; CompLND, complete lymphadenectomy; IncompLND, incomplete lymphadenectomy; LND, lymphadenectomy; LN, lymph nodes; LVI, lymphovascular invasion.
Patient characteristics according to type of lymphadenectomy in 180 non-metastatic patients with urothelial carcinoma of the renal pelvis
| CompLND | IncompLND | No-LND | P value | |
|---|---|---|---|---|
| Number of patients | 68 | 20 | 92 | |
| Mean patient age, year. (median; range) | 67.2 (67; 51–83) | 66.6 (68; 51–76) | 73.5 (75; 36–91) | <0.001 |
| Gender M/F, n | 51/17 | 15/5 | 62/30 | 0.52 |
| Mean follow-up, months (median; range) | 66.8 (47.1; 2.6–213) | 68.0 (38.1; 6.9–208) | 47.0 (26.3; 1.0–225) | 0.01 |
| No. cT stage (≤cT1/cT2/cT3/cT4) | 10/27/28/3 | 2/10/7/1 | 23/39/27/3 | 0.48 |
| No. cN stage (cN0/cN1, 2) | 68/0 | 20/0 | 92/0 | – |
| No. pT stage (≤pT1/pT2/pT3/pT4) | 22/11/32/3 | 2/3/13/2 | 30/11/44/7 | 0.20 |
| No. pN stage (pN0/pN1/pN2/pNx) | 63/2/3/0 | 13/2/5/0 | 0/3/1/88 | <0.001 |
| No. tumor grade (low/high) | 41/27 | 10/10 | 44/48 | 0.29 |
| LVI, (%) | 30 (44%) | 15 (75%) | 45 (49%) | 0.04 |
| Mean LN removed (median; range) | 11.5 (11; 4–32) | 4.6 (3; 2–16) | – | <0.001 |
| Adjuvant chemotherapy, n (%) | 7 (10%) | 1 (4%) | 1 (1%) | 0.02 |
| Recurrence, n (%) | 9 (13%) | 6 (30%) | 26 (28%) | 0.04 |
| Regional lymph node only, n (%) | 2 (3%) | 1 (5%) | 3 (3%) | 0.91 |
| Lymph node and distant organ | 2 (3%) | 3 (15%) | 9 (10%) | 0.10 |
| Distant organs only, n (%) | 5 (7%) | 2 (10%) | 14 (15%) | 0.28 |
| Cancer death, n (%) | 8 (12%) | 6 (30%) | 23 (25%) | 0.05 |
| CompLND | IncompLND | No-LND | P value | |
|---|---|---|---|---|
| Number of patients | 68 | 20 | 92 | |
| Mean patient age, year. (median; range) | 67.2 (67; 51–83) | 66.6 (68; 51–76) | 73.5 (75; 36–91) | <0.001 |
| Gender M/F, n | 51/17 | 15/5 | 62/30 | 0.52 |
| Mean follow-up, months (median; range) | 66.8 (47.1; 2.6–213) | 68.0 (38.1; 6.9–208) | 47.0 (26.3; 1.0–225) | 0.01 |
| No. cT stage (≤cT1/cT2/cT3/cT4) | 10/27/28/3 | 2/10/7/1 | 23/39/27/3 | 0.48 |
| No. cN stage (cN0/cN1, 2) | 68/0 | 20/0 | 92/0 | – |
| No. pT stage (≤pT1/pT2/pT3/pT4) | 22/11/32/3 | 2/3/13/2 | 30/11/44/7 | 0.20 |
| No. pN stage (pN0/pN1/pN2/pNx) | 63/2/3/0 | 13/2/5/0 | 0/3/1/88 | <0.001 |
| No. tumor grade (low/high) | 41/27 | 10/10 | 44/48 | 0.29 |
| LVI, (%) | 30 (44%) | 15 (75%) | 45 (49%) | 0.04 |
| Mean LN removed (median; range) | 11.5 (11; 4–32) | 4.6 (3; 2–16) | – | <0.001 |
| Adjuvant chemotherapy, n (%) | 7 (10%) | 1 (4%) | 1 (1%) | 0.02 |
| Recurrence, n (%) | 9 (13%) | 6 (30%) | 26 (28%) | 0.04 |
| Regional lymph node only, n (%) | 2 (3%) | 1 (5%) | 3 (3%) | 0.91 |
| Lymph node and distant organ | 2 (3%) | 3 (15%) | 9 (10%) | 0.10 |
| Distant organs only, n (%) | 5 (7%) | 2 (10%) | 14 (15%) | 0.28 |
| Cancer death, n (%) | 8 (12%) | 6 (30%) | 23 (25%) | 0.05 |
No., number; cT, clinical T stage; cN, clinical N stage; pT stage, pathological T stage; pN stage, pathological N stage; CompLND, complete lymphadenectomy; IncompLND, incomplete lymphadenectomy; LND, lymphadenectomy; LN, lymph nodes; LVI, lymphovascular invasion.
The Institutional Review Board of both institutes approved this prospective study protocol.
Surgery, Regional Lymph Nodes and Lymphadenectomy
Open nephroureterectomy was performed based on a procedure previously described with a retroperitoneal approach (14). Some patients underwent laparoscopic nephroureterectomy, but the excision of the bladder cuff was performed by open procedure with a small incision.
Figure 1 shows the anatomical template of lymphadenectomy for renal pelvic cancer, which covers all regional sites, proposed by our institutes (Tokyo Women's and Wakayama Medical: TWWM-template) (15,16). For tumors of the right renal pelvis: the right renal hilar, paracaval, retrocaval nodes and interaortocaval nodes were dissected. For tumors of the left renal pelvis: the left renal hilar and para-aortic nodes were dissected. The lower boundary of the template was at the level of the inferior mesenteric artery. All lymphadenectomies were performed by open procedure. Indication and extent of lymphadenectomy was based on the surgeons' preference until December 2004, since we did not have clear criteria for selecting patients. But after January 2005, complete lymphadenectomy (CompLND) was performed for all patients except for those with severe comorbidities or advanced age. The reason for omission of lymphadenectomy for those patients was to avoid excessive invasion and to minimize operation time. Thus, all of our incomplete lymphadenectomy (IncompLND) cases were performed before 2005.
Schema of anatomical template for lymphadenectomy in tumors of the renal pelvis proposed by our institutes (Tokyo Women's and Wakayama Medical: TWWM-Template)
We divided the subjects into three categories according to the status of lymphadenectomy. Complete lymphadenectomy was designated when the regional nodes were all dissected according to the TWWM-template (CompLND). When the extent of lymphadenectomy did not include all regional sites, it was categorized as IncompLND. All patients with IncompLND underwent surgery before 2005 in the retrospective cohort. The final group is comprised of patients who did not undergo lymphadenectomy (No-LND).
Patient Follow-up and Determination of Lymph Node Recurrence Sites
After surgery, patients were evaluated for recurrence with a CT scan every 6 months and for intravesical recurrence with cystoscopy and urine cytology every 3 months for the first 2 years. Subsequently, a follow-up examination was carried out every 6–12 months.
Recurrent sites in the lymph nodes were determined by CT scans. A lymph node increasing sequentially and exceeding 1 cm was considered a metastasized node.
Chemotherapy
Adjuvant chemotherapy was considered when nodal involvement and/or disease infiltrating the surrounding adipose tissue was observed. However, we made our final decision after considering patient comorbidity, performance status and willingness to receive therapy. Chemotherapy consisted of one to three courses of the M-VAC (consisting of methotrexate, vinblastine, doxorubicin and cisplatin) or the GC regimen (gemcitabine and cisplatin). No patients received neoadjuvant chemotherapy in this study cohort.
Statistical Analysis
Univariate analysis was performed using the Mann–Whitney U test to compare patient factors with continuous variables among groups and the χ2test to compare patient factors with categorical variables among groups. Intravesical recurrence was not included as the endpoint to determined recurrence-free survival (RFS). Cancer-free or RFS was calculated by the Kaplan–Meier method, and statistical significance was analyzed by the log-rank test. Univariate and multivariate analyses for prognostic factors influencing regional node recurrence were performed using the Cox proportional hazards model with stepwise selection. A difference was considered significant when P < 0.05.
RESULTS
Patient Characteristics
Table 1 shows patient characteristics according to type of lymphadenectomy in 180 non-metastatic patients with UCRP. Sixty-eight patients were categorized as CompLND, 20 as IncompLND and 92 as No-LND. Age of patients was significantly higher in the No-LND. Mean follow-up was over 5 years in the CompLND and IncompLND groups. There was no statistical difference in stage distribution among groups. The subjects of this study were clinically non-metastatic patients, thus all patients belonged to cN0. Pathological T stage did not differ between groups. Pathologically, lymph node metastases (pN1/2) were found in five patients (7.4%) in CompLND and in seven patients (35%) in IncompLND. Four patients (4.3%) were found to have node positive disease in the No-LND group by sampling a few swollen nodes. Tumor grade was similar among groups, but lymphovascular invasion was more frequently observed in IncompLND than in the two other groups. Adjuvant chemotherapy was given to more patients with CompLND than those with IncompLND or No-LND.
Recurrence-free and Cancer-specific Survival
We first examined and compared RFS and cancer-specific survival (CSS) among groups (Fig. 2). The 2 and 5-year RFS was 87.4 and 84.3% in the CompLND, 80.0 and 66.0% in the IncompLND, and 71.3 and 66.3% in the No-LND groups. RFS was significantly higher in patients with CompLND than those with IncompLND or No-LND (P = 0.03). The 2- and 5-year CSS was 95.1 and 90.7% in the CompLND, 83.7 and 63.7% in the IncompLND and 82.0 and 67.6% in the No-LND groups. CSS was also significantly higher in CompLND than in the IncompLND or No-LND groups (P = 0.03). Thus, higher RFS is likely to be associated with higher CSS in patients with CompLND.
Patient survival compared among three groups. Survival rate was calculated by the Kaplan–Meier method, and statistical significance was compared using the log-rank test. Results of recurrence-free survival are shown on the left and those of cancer-specific survival on the right. Thick black line indicates CompLND. Thick light gray line indicates IncompLND. Thin black line indicates No-LND.
Influence of the Extent of Lymphadenectomy on Recurrence Sites after Surgery
We next analyzed the influence of the extent of lymphadenectomy on the site of recurrence after surgery (Fig. 3a). Some patients showed metastases in multiple sites. Thus, we set up our data concerning the incidence of metastasis into three groups: regional nodes recurrence, distant nodes alone or distant organs. The incidence of regional node recurrence was significantly lower in patients with CompLND (2/68, 3%) than with IncompLND (4/20, 20%) or with No-LND (10/92, 11%) (P = 0.03). In the IncompLND group, 75% of regional node recurrence occurred outside the dissected sites. However, the incidence of metastases to distant nodes alone was very low and did not differ among groups. There was no statistical significance in the incidence of distant organ metastases among the groups (P = 0.28). These results show that template-based lymphadenectomy significantly decreases regional node recurrence but had no effect on distant node or organ metastases.
Incidence of recurrence by site compared among three groups. Black bar indicates regional nodes; dark gray bar, distant lymph nodes alone; intermediate gray bar, distant organs alone and light gray bar, no recurrence.
Predictive Factors for Regional Node Recurrence
We further examined the prognostic factors predicting regional node recurrence with univariate and multivariate analyses (Table 2). The univariate analysis showed that age, CompLND, advanced pathological T stage (>pT3), pathological node metastases (pN+) and high grade were significant factors. The multivariate analysis with stepwise selection showed that age, ≥pT3 and pN+ were significant independent factors for regional node recurrence. CompLND did not reach statistical significance (P = 0.06), but can be considered as a marginally significant factor.
Prognostic factors influencing regional node recurrence in patients with urothelial carcinoma of the renal pelvis
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Age | 1.08 | 1.02–1.15 | 0.002 | 1.08 | 1.02–1.15 | 0.002 |
| Gender (male) | 1.18 | 0.41–4.23 | 0.76 | |||
| CompLND | 0.20 | 0.03–0.72 | 0.01 | 0.17 | 0.04–1.07 | 0.06 |
| LVI | 4.93 | 1.58–21.5 | 0.004 | |||
| pT stage (≥pT3) | 3.5 × 106 | 6.79–∞ | <0.001 | 3.6 × 106 | 10.1–∞ | 0.01 |
| pN+ | 18.5 | 6.58–49.4 | <0.001 | 17.8 | 6.04–54.2 | <0.001 |
| High grade | 6.03 | 1.98–26.3 | 0.001 | |||
| Adjuvant chemotherapy | 2.97 | 0.46–10.6 | 0.20 | |||
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Age | 1.08 | 1.02–1.15 | 0.002 | 1.08 | 1.02–1.15 | 0.002 |
| Gender (male) | 1.18 | 0.41–4.23 | 0.76 | |||
| CompLND | 0.20 | 0.03–0.72 | 0.01 | 0.17 | 0.04–1.07 | 0.06 |
| LVI | 4.93 | 1.58–21.5 | 0.004 | |||
| pT stage (≥pT3) | 3.5 × 106 | 6.79–∞ | <0.001 | 3.6 × 106 | 10.1–∞ | 0.01 |
| pN+ | 18.5 | 6.58–49.4 | <0.001 | 17.8 | 6.04–54.2 | <0.001 |
| High grade | 6.03 | 1.98–26.3 | 0.001 | |||
| Adjuvant chemotherapy | 2.97 | 0.46–10.6 | 0.20 | |||
HR, hazard ratio; CI, confidential interval.
Prognostic factors influencing regional node recurrence in patients with urothelial carcinoma of the renal pelvis
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Age | 1.08 | 1.02–1.15 | 0.002 | 1.08 | 1.02–1.15 | 0.002 |
| Gender (male) | 1.18 | 0.41–4.23 | 0.76 | |||
| CompLND | 0.20 | 0.03–0.72 | 0.01 | 0.17 | 0.04–1.07 | 0.06 |
| LVI | 4.93 | 1.58–21.5 | 0.004 | |||
| pT stage (≥pT3) | 3.5 × 106 | 6.79–∞ | <0.001 | 3.6 × 106 | 10.1–∞ | 0.01 |
| pN+ | 18.5 | 6.58–49.4 | <0.001 | 17.8 | 6.04–54.2 | <0.001 |
| High grade | 6.03 | 1.98–26.3 | 0.001 | |||
| Adjuvant chemotherapy | 2.97 | 0.46–10.6 | 0.20 | |||
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Age | 1.08 | 1.02–1.15 | 0.002 | 1.08 | 1.02–1.15 | 0.002 |
| Gender (male) | 1.18 | 0.41–4.23 | 0.76 | |||
| CompLND | 0.20 | 0.03–0.72 | 0.01 | 0.17 | 0.04–1.07 | 0.06 |
| LVI | 4.93 | 1.58–21.5 | 0.004 | |||
| pT stage (≥pT3) | 3.5 × 106 | 6.79–∞ | <0.001 | 3.6 × 106 | 10.1–∞ | 0.01 |
| pN+ | 18.5 | 6.58–49.4 | <0.001 | 17.8 | 6.04–54.2 | <0.001 |
| High grade | 6.03 | 1.98–26.3 | 0.001 | |||
| Adjuvant chemotherapy | 2.97 | 0.46–10.6 | 0.20 | |||
HR, hazard ratio; CI, confidential interval.
DISCUSSION
In the present study, we demonstrate that in patients with UCRP, CompLND is a marginally significant factor for reducing the risk of regional node recurrence. Regional node recurrence in patients with IncompLND was observed at sites where lymphadenectomy was not performed (Extra-LND extent). This study further confirmed the therapeutic role of lymphadenectomy performed on the basis of the anatomical template (TWWM-template) proposed by our institutes.
Our recent prospective study showed the reduction of risk for regional node recurrence by preforming lymphadenectomy (6), which is supported by the present study with an increased number of patients. As a mechanism for this type of risk reduction, Roscigno et al. (4) hypothesizes that lymphadenectomy may remove microcancer deposits in the lymph nodes that cannot be detected by routine pathological examination with hematoxylin and eosin staining. This hypothesis is confirmed in a study by Abe et al. (5) in which they examined nodal specimens with immunohistochemistry using anticytokeratin antibodies. Of 51 patients with pN0 from routine pathological examinations, 7 (13.7%) showed micrometastases in the lymph nodes detected by immunohistochemical staining. Five of seven patients survived without recurrence. These results may also account for the reduced incidence of regional node recurrence observed in our study. We are currently examining micrometastases of the lymph nodes at the genetic level to support this hypothesis.
To the best of our knowledge, this is the first report determining prognostic factors to predict recurrence at the regional nodes. Advanced age, advanced pathological T stage and pathological node metastases are significant, independent factors that predict regional node metastases. These factors are reported also to be predictive of cancer-specific survival and/or overall survival (17–19). In other words, these conventional prognostic factors associated with unfavorable survival are also predictive of regional node recurrence. Regional node recurrence after surgery has been reported to be high, (and may be reported as recurrence to the soft tissues in some studies), ranging from 8.1 to 20% in patients with upper tract urothelial carcinoma (9–12). This suggests that prevention of regional node recurrence results in the improvement of patient survival. The present study shows that lymphadenectomy based on the TWWM-template is also an independent factor that reduces the risk of regional node recurrence. In fact, lymphadenectomy is the only modifiable factor with which urologists can intervene to reduce the risk of regional node recurrence. We also should note that adjuvant chemotherapy shows no influence on lowering regional node recurrence and is reported to have no impact on patient survival (20,21).
We should emphasize that lymphadenectomy has to be template based. The present study showed that regional node recurrence at the Extra-LND extent was 75% in the IncompLND group in which dissection did not include all regional sites. In addition, IncompLND did not improve recurrence-free survival. This suggests that all regional sites should be dissected to reduce the risk of regional node recurrence. We have previously shown that the extent of lymphadenectomy should be determined based on anatomical template, not on the number of lymph nodes (22). Some multi-institutional studies failed to show a therapeutic benefit of lymphadenectomy because their survival rate was similar among patients who underwent lymphadenectomy with no pathological node metastases (pN0) and those without lymphadenectomy (pNx) (19,23–25). However, the extent of lymphadenectomy was not clearly determined because of the retrospective nature of these studies. This may have compromised the therapeutic benefit of lymphadenectomy.
The unfavorable outcome of the IncompLND group may be due to election bias due to the higher incidence of patients with positive LVI or pathological node metastases in this group. But, this study showed a significant difference between CompLND and No-LND for both RFS and CSS. We have already confirmed the therapeutic role of lymphadenectomy using the TWWM-template in a prospective study of patients with UCRP. (6) Thus, we suggest that the role of IncompLND is questionable since the studies which also support the therapeutic benefit of lymphadenectomy utilized a similar extent of lymphadenectomy to the TWWM-template.
Finally, we should emphasize the limitations of this study. First, although this is a multi-institutional study, it includes only two institutes. A larger number of institutes and patients will be needed to further support our results. Second, this is a retrospective study. Thus, the favorable outcome of the CompLND group may be attributed to patient selection bias. We also need a prospective, randomized trial to support our results with a higher level of evidence.
CONCLUSIONS
In the present study, we demonstrate that lymphadenectomy needs to be based on the TWWM-template to reduce the risk of regional node recurrence. Since this is the only modifiable factor that can influence patient survival, we believe that CompLND should be considered whenever possible.
Conflict of interest statement
None declared.
Acknowledgements
The authors thank Elizabeth Kiritani for English language assistance, and Noriko Hata for secretarial work.
