SEARCH

SEARCH BY CITATION

Keywords:

  • penile carcinoma;
  • squamous cell carcinoma;
  • partial penectomy;
  • total penectomy;
  • inguinal lymph node involvement;
  • vascular embolization;
  • lymphatic embolization

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

The objective of the current study was to identify independent clinical and pathologic variables that were predictive of lymph node involvement in patients with squamous cell carcinoma of the penis in a multicenter series with the intent to select patients who were suitable to undergo immediate inguinal lymphadenectomy.

METHODS

Data were analyzed from 175 patients who underwent surgery for penile carcinoma in 11 urologic centers participating in the Gruppo Uro-Oncologico del Nord-Est (Northeast Uro-Oncological Group) Penile Cancer Data Base. Pathologically positive lymph nodes were defined as the presence of histologically confirmed lymph node metastasis in patients who underwent either immediate or delayed inguinal and/or pelvic lymphadenectomy. Patients who had clinically positive lymph nodes with cytologically positive fine-needle aspiration results and who had not undergone lymphadenectomy were censored.

RESULTS

Overall, lymph-node involvement was observed in 71 of 175 patients (40.6%) included in the analyses. After analyzing the whole group of patients, the following variables were identified as independent predictors of pathologic lymph node metastasis: clinical lymph node status, pathologic stage of the primary tumor, venous and lymphatic embolizations, and histologic grade. In the subgroup of patients with clinically negative lymph nodes, tumor thickness, histologic grade, lymphatic and venous embolizations, infiltration of both corpus spongiosum and urethra, and pathologic stage of the primary tumor (according to the 1997 TNM classification system) were predictive of lymph node involvement on univariate analysis. The generated logistic regression model showed that venous and/or lymphatic embolizations and infiltration of the corpus spongiosum and/or urethra were independent predictors of pathologic lymph node metastasis in patients with clinically negative lymph nodes.

CONCLUSIONS

Venous and/or lymphatic embolizations played relevant roles as predictors of pathologic lymph node involvement in patients with penile neoplasia and should be considered important parameters in determining which patients with clinically negative lymph nodes should undergo immediate lymphadenectomy. Cancer 2005. © 2005 American Cancer Society.

The presence and extension of inguinal and pelvic lymph node metastases are the most relevant prognostic factors in squamous cell carcinoma of the penis. At initial presentation, clinically palpable inguinal lymph nodes are present in 28–64% of patients with penile carcinoma.1 In 47–85% of these patients, metastatic inguinal lymph node involvement is verified.2–4 In the remaining patients, groin lymph node enlargement is due to overlapping inflammatory reactions caused by infection of the primary tumor. Pelvic lymph node extension is found in 22–56% of patients who have positive inguinal lymph nodes.5, 6

In patients with clinically unpalpable inguinallymph nodes, occult micrometastases are present in 12–24%.7 Early lymphadenectomy has yielded 5-year disease-specific survival rates of 58–88%.1, 8, 9 The routine prophylactic inguinal lymphadenectomy is limited by relevant postoperative complications experienced by 50% of patients (range, 24–87%) and by mortality related to the procedure, which is estimated as > 3%.5

The removal of the sentinel lymph node, as labeled by Cabanas10 through penis lymphangiography, and the limited lymph node dissection, as proposed by Catalona,11 to reduce surgical complications, are not free of false-negative findings, and their routine clinical use has been questioned.4, 12 In 2001, Horenblas et al. proposed the use of dynamic lymphoscintigraphy with isosulphan blue and technetium 99m-nanocolloid sulphur to identify the sentinel lymph node in patients with clinically negative inguinal lymph nodes.5 This technique provided an opportunity to limit extended lymph node dissection only to patients who had positive sentinel lymph nodes. The currently available results are promising, although they come from a single center.13, 14

With the objective of reducing the number of prophylactic lymphadenectomies in patients with clinically negative inguinal lymph nodes, the pathologic findings of the primary tumor could be used to define the prognostic risk of lymph node metastasis, permitting the identification of high-risk patients. Pathologic extension of the primary tumor, histologic grade, vertical or superficial growth pattern, and both lymphatic and vascular embolization are the histopathologic parameters assessed most frequently to predict inguinal lymph node involvement.15–20 The retrospective nature of the series, the lack of data on a large cohort of patients, and the single-center setting of the studies limit the reliability of published data.

The Gruppo Uro-Oncologico del Nord Est (i.e., Northeast Uro-Oncological Group) (GUONE) is a multidisciplinary group that includes urologists, oncologists, radiotherapists, and pathologists working in northeastern Italy. In March 2004, GUONE created a Penile Cancer Data Base with the purpose of collecting data retrospectively on patients with penile squamous cell carcinoma observed from 1980 to 2002 in the participating urologic and oncologic centers. The objective of the current study was to identify clinical and pathologic variables that may be independent predictors of metastatic inguinal lymph node involvement with the intent to identify patients with squamous cell carcinoma of the penis who may be suitable for immediate inguinal lymphadenectomy.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

The GUONE Penile Cancer Data Base collected clinical and pathologic data from 265 patients who were treated for squamous cell carcinoma of the penis between 1980 and 2002 in 11 urologic and oncologic centers of northeastern Italy (in the cities of Belluno, Brescia, Bolzano, Padova, Pordenone, Trento, Treviso, Trieste, Udine, Verona, and Vicenza). Thirty-nine patients (14.7%) were excluded from this study as they were lost during follow-up. Moreover, 51 patients (19.2%) who had undergone either penis-preserving treatments or radiotherapy for primary tumors were not included in the analysis because of the lack of several assessed pathologic parameters. One hundred seventy-five patients who had undergone either partial or total amputation or emasculation became the object of the study.

In each patient, the following clinical variables were evaluated: age, execution of circumcision at a pediatric age (age < 18 years); history of sexually transmitted disease; time elapsed between the onset of penile lesions and histologic tumor diagnosis; clinical size, site, and tumor morphology; clinical stage of both the primary tumor and inguinal lymph nodes; and the presence of distant metastasis. Clinical stage was assigned according to 1978 TNM classification system (T0, no evidence of primary tumor; Tis, carcinoma in situ; T1, superficial exophytic lesions measuring ≤ 2 cm in greatest dimension; T2, lesions measuring > 2 cm but < 5 cm or with minimal depth of invasion; T3, lesions measuring > 5 cm or with deep invasion; and T4, tumor invading adjacent structures).21 We elected to use this older edition of the TNM classification system because of its reliability in clinical practice compared with subsequent versions based on clinical definition of the local extension of the primary tumor.22 Clinical lymph node status was assigned by inguinal lymph node palpation and either ultrasound or computed tomography scan (N0, no evidence of regional lymph node metastasis; N1, metastasis in a single, superficial inguinal lymph node; N2, metastases in multiple or bilateral, superficial lymph nodes; N3, metastasis in deep inguinal and/or pelvic lymph node).

The following pathologic variables of the primary tumor were assessed: positive resection margins; tumor thickness; growth pattern; histologic grade; lymphatic and vascular embolization; and infiltration to the corpora cavernosa, corpus spongiosum, or urethra. The pathologic findings were assigned by pathologists from each center involved in the study. The pathologic stage was reported according to the 1997 TNM classification system as follows: pT1, tumor invading subepithelial connective tissue; pT2, tumor invading corpora cavernosa or corpus spongiosum; pT3, tumor invading urethra; pT4, tumor invading other adjacent structures.23 Histologic grade, based on the Broders classification,24 was classified as follows: Grade 1, well differentiated; Grade 2, moderately differentiated; and Grade 3, poorly differentiated. The tumor growth pattern was classified as superficial or vertical, as proposed by Cubilla et al.25 Tumor thickness was measured from the top of the tumor to the deepest tumor cell.18 Lymphatic embolization was defined as the presence of tumor emboli within endothelium-lined spaces bound by a thin wall without either smooth muscle fiber or red blood cells.26 The same morphological aspect but with the presence of both parietal smooth muscle fibers and red blood cells, was defined as venous embolization.26

To assess the risk of metastatic lymph node involvement, patients were classified according to both the European Association of Urology (EAU) risk group (low risk: pTis, pTa/Grade 1–2, or pT1/Grade 1; intermediate risk: pT1/Grade 2; and high risk: ≥ pT2 or Grade 3)20 and the Solsona risk group (low risk: pTa–pT1/Grade 1; intermediate risk: pT1/Grade 2–3 and pT2/Grade 1; high risk: pT2/Grade 2–3 or ≥ pT3).27 Pathologically positive lymph nodes were defined as the presence of histologically confirmed lymph node metastasis in patients who underwent either immediate or delayed inguinal and/or pelvic lymphadenectomy. Moreover, those patients with clinically positive lymph nodes who had cytologically positive fine-needle aspiration results and had not undergone lymphadenectomy were censored.

Statistical Analysis

The Pearson chi-square test was used to compare categorical variables and to perform univariate analyses. Logistic regression was used to identify the independent predictive value of the prognostic variables that showed significant results on univariate analysis. The Kaplan–Meier method was used to estimate disease-free survival, which was defined as the time between radical excision of the primary tumor to the pathologic diagnosis of lymph node recurrence. All analyses showed minimal statistical significance, with 2-sided P values < 0.05. All clinical and pathologic data on the patients were collected in a single electronic data base and were analyzed by one of the authors (V.F.) using the Statistical Package for Social Science (version 12.0; SPSS Inc., Chicago, IL).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

The median (± standard deviation) age of the 175 analyzed patients was 62.0 years ± 12.2 years (range, 34–91 years). Only 17 patients (9.7%) had undergone circumcision at a pediatric age. Sexually transmitted diseases were present in 19 patients (10.9%). The mean time elapsed between the onset of the penile lesion and the histologic diagnosis of penile carcinoma was 4 ± 12.8 months (range, 1–96 months). At initial presentation, phimosis was present in 61 patients (34.9%). The median clinical tumor size was 30.0 ± 14.3 mm (range, 3–80 mm). Macroscopically, the lesion was exophytic in 66 patients (37.7%), ulcerated in 91 patients (52%), and nodular in 18 patients (10.3%). The primary tumor was located at the prepuce in 16 patients (9.1%), the glans in 88 patients (50.3%), the coronal sulcus in 59 patients (33.7%), and shaft in 12 patients (6.9%).

According to the 1978 TNM classification system, the clinical stage of the primary tumor was Ta/T1 in 32 patients (18.3%), T2 in 104 patients (59.4%), T3 in 33 patients (18.9%), and T4 in 6 patients (3.4%). The inguinal lymph nodes were clinically negative (cN0) in 115 patients (65.7%). In the remaining patients, inguinal lymph nodes were palpable monolaterally in 24 patients (13.7%), palpable bilaterally in 32 patients (18.3%), and fixed to adjacent tissue in 4 patients (2.3%). Clinically positive pelvic lymph nodes were identified in 15 patients (8.6%). The clinical lymph node status (according to the 1997 TNM classification system) was negative (cN0) in 115 patients (65.7%), cN1 in 17 patients (9.7%), cN2 in 27 patients (15.4%), and cN3 in 16 patients (9.1%). At presentation, distant metastases were found in 5 patients (2,9%).

Partial and total penectomies were undergone by 149 patients (85.1%) and 22 patients (12.6%), respectively; only 4 patients underwent emasculation (2,3%). The pathologic stage of the primary tumor was pTa–pT1 in 59 patients (33.7%), pT2 in 86 patients (49.1%), pT3 in 29 patients (16.6%), and pT4 in 1 patient (0.6%).

In patients with clinically positive lymph nodes, pelvic and/or inguinal lymphadenectomy was undergone by 42 patients (70%). In this subset, the removed lymph nodes were confirmed as histologically positive in 29 patients (69%). Eighteen patients (30%) with clinically positive lymph nodes who had cytologically positive fine-needle aspiration results did not undergo lymphadenectomy, because they were either unwilling or unfit to undergo surgery, and they received radiotherapy and/or chemotherapy.

Among the patients with cN0 status, a surveillance program was followed in 85 patients (73.9%). At a median follow-up of 26 months (range, 2–292 months), lymph node recurrence was observed in 18 patients: Nine of those patients underwent delayed lymphadenectomy, whereas the other 9 patients underwent only fine-needle aspiration biopsy of palpable inguinal lymph nodes. Lymph node recurrences were observed within 6 months after treatment of the primary tumor in 9 of 18 patients (50%), within 1 year in 14 of 18 patients (77.7%), and before 24 months in all patients. Figure 1 shows the disease-free survival curve in the patients with clinically negative lymph node status who underwent a surveillance program.

thumbnail image

Figure 1. Disease-free survival probability in patients who had penile carcinoma with clinically negative lymph node status and who underwent surveillance. All 18 events occurred within 24 months after treatment of the primary tumor.

Download figure to PowerPoint

The remaining 30 patients (26.1%) with cN0 status underwent prophylactic inguinal lymphadenectomy. The removed lymph nodes were pathologically negative (pN0) in 24 patients (80%) and positive in 6 patients (20%). Table 1 summarizes the pathologic features of the patients according to their pathologic lymph node status.

Table 1. Pathologic Features of the 30 Patients with Clinically Negative Lymph Nodes who Underwent Immediate Inguinal Lymphadenectomya
VariableNo. of patients (%)P value
pN0 (n = 24)pN+ (n = 6)
  • pN0: negative pathologic lymph node status; pN+: positive pathologic lymph node status.

  • a

    Few data were missing for some patients.

Tumor thickness  0.21
 ≤ 5 mm8 (100)0 (0) 
 > 5 mm14 (70)6 (30) 
Growth pattern  0.35
 Vertical11 (69)5 (31) 
 Superficial10 (91)1 (9) 
Invasion depth  0.11
 ≤ 3 mm10 (100)0 (0) 
 > 3 mm12 (67)6 (33) 
Histologic grade  0.95
 Grade 19 (82)2 (18) 
 Grade 2–312 (75)4 (25) 
Lymphatic embolization  0.002
 Absent18 (95)1 (5) 
 Present2 (29)5 (71) 
Vascular embolization  0.008
 Absent20 (85)3 (15) 
 Present0 (0)3 (100) 
Corpora cavernosa infiltration  0.12
 Absent15 (94)1 (6) 
 Present9 (64)5 (36) 
Corpus spongiosum infiltration  0.02
 Absent15 (100)0 (0) 
 Present9 (60)6 (40) 
Urethra infiltration  0.06
 Absent22 (88)3 (12) 
 Present2 (40)3 (60) 
Solsona risk groups  0.17
 Low2 (100)0 (0) 
 Intermediate11 (92)1 (8) 
 High9 (64)5 (36) 

Overall, lymph node involvement was observed in 71 of 175 patients (40.6%) who were included in the analysis: Thirty-five patients had histologically positive lymph nodes at early lymphadenectomy, 9 patients had histologically positive lymph nodes at delayed lymphadenectomy, and 27 patients had histologically positive lymph nodes at fine-needle aspiration biopsy (18 patients at the time of the initial diagnosis, who received radiotherapy and/or chemotherapy, and 9 patients during follow-up) (Table 2).

Table 2. Clinical and Pathologic Lymph Node Status in the 175 Analyzed Patients
Clinical lymph node statusTreatment (no. of patients)No. with positive lymph node status (n = 71)
  1. cN0: negative clinical lymph node status; pN+: positive pathologic lymph node status; LAD: lymphadenectomy; N+: positive lymph node status; CHT: chemotherapy; RT: radiotherapy; cN+: positive clinical lymph node status; FNA: fine-needle aspiration.

cN0 (n = 115)Surveillance (85)9 pN+ (delayed LAD)
  9 N+ (CHT and RT)
 Immediate LAD (30)6 pN+
cN+ (n = 60)Delayed LAD (42)29 pN+
 RT and CHT (18)18 N+ (FNA)

Among the evaluated clinical variables, those that were predictive of lymph node involvement on univariate analysis were tumor site (P = 0.007), macroscopic tumor morphology (P = 0.01), and clinical stage both of the primary tumor (P = 0.000) and of locoregional lymph nodes (P = 0.000) (Table 3). On multivariate analysis, only the clinical stage of lymph nodes was an independent predictor of pathologic lymph node metastasis (Table 4).

Table 3. Clinical Variables Predictive of Lymph Node Involvement: Univariate Analysis (Pearson Chi-Square Test) in the Entire Cohort of Patients
VariableNo. of patientsNo. of patients with pN+ status (%)Univariate P value
  • pN+: positive pathologic lymph node status; N0: negative lymph node status; N+: positive lymph node status.

  • a

    According to the 1978 TNM classification system.

  • b

    According to the 1997 TNM classification system.

Age  0.66
 ≤ 60 yrs7431 (41.9) 
 > 60 yrs10139 (38.6) 
Diagnosis delay  0.21
 ≤ 4 mos9238 (41.3) 
 > 4 mos7223 (31.9) 
Clinical tumor size  0.30
 ≤ 30 mm9433 (35.1) 
 > 30 mm6026 (43.3) 
Phimosis at presentation  0.65
 Present6123 (37.7) 
 Absent11447 (41.2) 
Tumor site  0.007
 Prepuce/glans10433 (31.7) 
 Sulcus/shaft7137 (52.1) 
Macroscopic growth pattern  0.01
 Exophytic6617 (25.8) 
 Ulcerated9144 (48.4) 
 Nodular189 (50.0) 
Clinical tumor classificationa  0.000
 Ta/T1328 (25.0) 
 T210436 (34.6) 
 T3–T43926 (66.7) 
Clinical lymph node statusb  0.000
 N011525 (21.7) 
 N+6045 (75.0) 
Table 4. Independent Predictive Clinical Variables of Inguinal Lymph Node Involvement in the Entire Cohort of Patients: Multivariate Logistic Regression Analysis
VariableHR95% CIP value
  • HR: hazard ratio; 95% CI: 95% confidence interval; cN0: negative clinical lymph node status; cN+: positive clinical lymph node status.

  • a

    According to the 1978 TNM classification system.

  • b

    According to the 1997 TNM classification system.

Tumor site (distal vs. proximal)1.7360.825–3.6550.14
Macroscopic growth pattern (exophytic, ulcerated, nodular)1.4950.831–2.6910.18
Clinical tumor classification (Ta/T1 vs. T2 vs. > T2)a1.5170.825–2.7890.18
Clinical lymph node status (cN0 vs. cN+)b8.5133.969–18.256< 0.001

Among the pathologic variables that were analyzed, the following were predictors of pathologic lymph node involvement on univariate analysis: tumor thickness (P < 0.001), histologic grade (P < 0.001), lymphatic embolization (P < 0.001), venous embolization (P < 0.001), corpora cavernosa infiltration (P = 0.004), corpus spongiosum infiltration (P = 0.005), urethral infiltration (P < 0.001), and pathologic stage (pT) of the primary tumor (P < 0.001). In addition, both the EAU (P = 0.000) and the Solsona (P = 0.000) risk-group assessments proved predictive (Table 5). On multivariate analysis, the pathologic stage of the primary tumor, venous embolization, lymphatic embolization, and histologic grade turned out to be independent predictors of pathologic lymph node metastasis (Table 6).

Table 5. Pathologic Variables that Were Predictive of Lymph Node Involvement: Univariate Analysis in the Entire Cohort of Patients (Pearson Chi-Square Test)
VariableNo. of patientsNo. of patients with pN+ status (%)Univariate analysis (P value)
  • pN+: positive pathologic lymph node status; pT: pathologic tumor status.

  • a

    According to the 1997 TNM classification system.

Surgical margins  0.08
 Negative16060 (37.5) 
 Positive117 (63.6) 
Tumor thickness  0.000
 ≤ 5 mm6214 (22.2) 
 > 5 mm9146 (50.5) 
Growth pattern  0.06
 Superficial6319 (30.2) 
 Vertical9141 (45.1) 
Histologic grade  0.000
 Grade 1497 (14.3) 
 Grade 2–311756 (47.0) 
Lymphatic embolization  0.000
 Present3630 (83.3) 
 Absent11125 (22.5) 
Venous embolization  0.000
 Present2926 (89.7) 
 Absent11829 (24.6) 
Corpora cavernosa infiltration 0.004
 Present7941 (51.9) 
 Absent9629 (30.2) 
Corpus spongiosum infiltration 0.005
 Present6133 (54.1) 
 Absent11437 (32.5) 
Urethra infiltration  0.000
 Present3021 (70.0) 
 Absent14549 (33.8) 
Pathologic tumor classificationa 0.000
 pTa/pT15912 (20.3) 
 pT28637 (43.0) 
 pT3–pT43021 (70.0) 
European Association of Urology risk group (pT and grade)0.000
 Low251 (4.0) 
 Intermediate238 (34.8) 
 High11854 (45.8) 
Solsona risk group (pT and grade) 0.000
 Low251 (4.0) 
 Intermediate5516 (29.1) 
 High8646 (53.5) 
Table 6. Independent Predictive Pathologic Variables of Inguinal Lymph Node Involvement in the Entire Group of Patients: Multivariate Logistic Regression Analysis
VariableHR95% CIP value
  • HR: hazard ratio; 95% CI: 95% confidence interval.

  • a

    According to the 1997 TNM classification system.

Tumor thickness (≤ 5 mm vs. > 5 mm)1.4350.538–3.8330.47
Pathologic tumor classification (pTa/pT1 vs. pT2 vs. > pT2)a2.2881.118–4.6840.02
Histologic grade (Grade 1 vs. Grade 2–3)4.2681.278–14.3640.01
Venous embolization (absent vs. present)5.2401.139–24.1010.03
Lymphatic embolization (absent vs. present)6.9411.967–24.4980.003

When patients were stratified according to their clinical lymph node status, the variables that were significant on univariate analysis among patients with cN0 status were tumor thickness (P = 0.01), histologic grade (P = 0.01), lymphatic embolization (P < 0.001), venous embolization (P < 0.001), corpus spongiosum infiltration (P = 0.01), urethral infiltration (P < 0.001), and pathologic stage of the primary tumor (pT) (P < 0.001). Moreover, the Solsona risk-group assessment also proved predictive of lymph node involvement (P = 0.007) (Table 7). To avoid colinearity in generating the multivariate model, we clustered the presence of either venous or lymphatic embolization in a single variable. Moreover, according to the data from the univariate analysis, local extension of the primary tumor was categorized as follows, by splitting the group of patients with pT2 tumors into pT1 and pT2 tumors (corpora cavernosa invasion) versus pT2 (corpus spongiosum invasion) and pT3 tumors. The generated logistic regression model showed that venous and/or lymphatic embolizations and corpus spongiosum and/or urethral infiltrations were independent predictors of pathologic lymph node metastasis in patients with cN0 status (Table 8).

Table 7. Pathologic Variables Predictive of Lymph Node Involvement in Patients with Negative Clinical Lymph Node Status: Univariate Analysis (Pearson Chi-Square Test)
VariableNo. of patientsNo. of patients with pN+ status (%)Univariate P value
  • pN+: positive pathologic lymph node status.

  • a

    According to the 1997 TNM classification system.

Surgical margins  0.31
 Negative10923 (21) 
 Positive52 (40) 
Tumor thickness  0.01
 ≤ 5 mm516 (12) 
 > 5 mm527 (33) 
Growth pattern  0.45
 Superficial499 (18) 
 Vertical5313 (25) 
Invasion depth  0.001
 ≤ 3 mm504 (8) 
 > 3 mm5419 (35) 
Histologic grade  0.01
 Grade 1434 (9) 
 Grade 2–36920 (29) 
Lymphatic embolization  < 0.001
 Present117 (64) 
 Absent8814 (16) 
Venous embolization  < 0.001
 Present108 (80) 
 Absent8913 (15) 
Corpora cavernosa infiltration  0.44
 Present4311 (26) 
 Absent7214 (19) 
Corpus spongiosum infiltration  0.01
 Present3613 (36) 
 Absent7912 (15) 
Urethra infiltration  < 0.001
 Present149 (64) 
 Absent10116 (16) 
Pathologic tumor classificationa  < 0.001
 pTa/pT1455 (11) 
 pT25611 (20) 
 pT3-pT4149 (64) 
European Association of Urology risk group (pT and grade)0.09
 Low221 (4) 
 Intermediate174 (23) 
 High7319 (26) 
Solsona risk group (pT and grade)  0.007
 Low221 (4) 
 Intermediate416 (15) 
 High4917 (35) 
Table 8. Independent Predictive Pathologic Variables of Inguinal Lymph Node Involvement in Patients with Clinically Nonpalpable Lymph-Nodes: Multivariate Logistic Regression Analysis
VariableHR95% CIP value
  1. HR: hazard ratio; 95% CI: 95% confidence interval.

Tumor thickness (≤ 5 mm vs. > 5 mm)1.3680.371–5.0430.63
Histologic grade (Grade 1 vs. Grade 2–3)3.5640.930–13.6560.06
Spongiosum/urethra infiltrations (absent vs. present)4.6321.337–16.0510.01
Venous/lymphatic embolizations (absent vs. present)5.3811.370–21.1290.01

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Squamous cell carcinoma of the penis spreads mainly through the lymphatic system to inguinal and pelvic lymph nodes. The presence of metastasis at inguinal and pelvic lymph nodes is the most relevant, unfavorable prognostic factor in these patients. The 5-year disease-specific survival rate in patients with pathologically positive lymph nodes after they undergo inguinal lymphadenectomy varies in relation to the extension of lymph node involvement: 50–80% in patients with pN1 status, 9–30% in patients with pN2 status, and 0–15% in patients with pN3 status.28 Only 17–45% of clinically palpable lymph nodes proved to be positive pathologically. Our data showed that the presence of clinically positive inguinal lymph nodes was related significantly to the presence of metastases. Therefore, in patients who have palpable, mobile inguinal lymph nodes, bilateral radical inguinal lymphadenectomy is recommended strongly. Patients who have fixed inguinal masses should undergo neoadjuvant chemotherapy, if necessary, followed by radical inguinal and pelvic lymphadenectomy.20

Moreover, approximately 20% of patients with clinically negative lymph nodes may present with pathologic lymph node involvement after they undergo inguinal lymphadenectomy.18–20 The opportunity to identify patients with cN0 status who may have a higher risk of harboring occult micrometastasis may allow a decrease in the number of patients who undergo early inguinal lymphadenectomy, with the clear advantage of limiting the risk of surgical complications. Inguinal lymphadenectomy has surgical morbidity rates ranging from 24% to 87% and has a perioperative mortality rate > 3%.1, 5

This multicenter study highlighted that vascular and lymphatic embolizations, in addition to pathologic stage and histologic grade of the primary tumor, play relevant roles as predictors of pathologic locoregional lymph node involvement in patients with squamous carcinoma of the penis. Specifically, in the patients with clinically negative lymph nodes, which definitely comprised the most interesting group of patients, our multivariate analysis underscored that the histologic findings of vascular and/or lymphatic embolizations and the presence of corpus spongiosum and/or urethral infiltration were independent predictors of inguinal lymph node metastasis.

The relevance of venous and lymphatic embolizations as variables predictive of metastatic lymph node involvement was reported in 1996 by Lopes et al. in a multivariate analysis of 145 patients who had undergone partial penectomy.17 Those authors showed rates of pathologically positive inguinal lymph nodes as high as 63.5% and 30.6% in patients with and without lymphatic embolization, respectively. Similarly, 77.8% and 44% of patients with and without venous embolization had inguinal metastases, respectively.17 In 2001, Slaton et al. underscored that the presence of vascular embolization could be used to select patients with infiltration to the corpora cavernosa or corpus spongiosum (pT2) to undergo early lymphadenectomy.18 Also in that series, only 25% of patients with pT2 tumors without vascular embolization had lymph node metastasis; conversely, 71% of patients with pT2 tumors who had vascular embolization had pathologically positive lymph nodes.18 In the same report, the pathologic stage of primary tumors and the percentage of poorly differentiated tumors were variables that were predictive of lymph node involvement.18 More recently, Ficarra et al. analyzed 30 patients who underwent either partial or total penectomy in a multivariate data set and found that the histologic grade of the primary tumor and lymphatic embolization were independent variables that were predictive of lymph node involvement. In that article, lymph node metastases were reported in 100% and 22% of patients with and without lymphatic embolization, respectively.29

The correlation between local excision of the primary tumor and lymph node involvement has been shown clearly in the literature.2, 16, 18, 22 The rate of lymphatic progression in patients who had penile carcinoma without corpora cavernosa or corpus spongiosum infiltration (pTa–T1) ranged between 5% and 11%.12, 30, 31 Conversely, in patients who had penile carcinoma with invasion to the corpora cavernosa, corpus spongiosum, or urethra (pT2–pT3), the rate increased up to 60–75%.16, 30, 32

In patients with clinically negative lymph nodes, we observed a significant increase in the risk of pathologic lymph node involvement among those with infiltration to the corpus spongiosum. These data suggest the possibility that spongiosum infiltration should be considered closer to urethral invasion (pT3) than to corpora cavernosa involvement (pT2).

In most studies, the histologic grade proved to be a more powerful prognostic factor than the pathologic stage.16, 32, 33 The rates of pathologically positive lymph nodes were 4–24% in well differentiated tumors (Grade 1), 46–79% in moderately differentiated tumors (Grade 2), and 82–100% in poorly differentiated tumors (Grade 3).27, 32, 34 The 2004 EAU guidelines on penile carcinoma reconfirmed that the pathologic stage and histologic grade of the primary tumor were the most relevant variables for selecting patients who should undergo immediate inguinal lymphadenectomy.20

A more accurate definition of the risk categories can be obtained by combining the pathologic stage and the histologic grade of the primary tumor. In 1992, Solsona et al. identified three categories with different risks of lymph node involvement. The “low-risk” group included patients with pT1/Grade 1 tumors, patients with pT1/Grade 2–3 and pT2/Grade 1 tumors formed the “intermediate-risk” category, and patients with pT2/Grade 2–3 or ≥ T3 tumors formed the “high-risk” category.27 In a recent prospective study, the same authors reported lymph node involvement rates of 0%, 33%, and 83% in the low-risk, moderate-risk, and high-risk groups, respectively.15 The roles of venous and lymphatic involvement were not assessed in that report. Our data confirmed that the patient risk-stratification schemes proposed both by the EAU guidelines20 and by Solsona et al.27 were related significantly to the percentage of pathologic locoregional lymph node involvement.

Villavicencio et al. observed a good correlation between the tumor growth pattern and groin lymphatic involvement.16 Lymph node metastases were detected in 70% and 14% of patients with vertical and superficial growth patterns (P = 0.0009), respectively.16 In our series, those data did not achieve statistical significance.

The results from this study confirmed and supported the 2004 EAU guideline recommendations for early inguinal lymphadenectomy in all patients who have poorly differentiated (Grade 3) or ≥ pT2 tumors; however, we propose surveillance for patients with low-stage (Ta–T1) and low-grade tumors (Grade 1–2).20 However, the current data highlighted that lymphatic and/or venous embolizations resulted in a clear indication for inguinal lymphadenectomy, independent of the primary tumor grade or stage.

Dynamic sentinel node biopsy, as described by Horenblas, may be useful for selecting patients with clinically negative lymph nodes who are candidates for early lymphadenectomy.5 The administration of both a blue dye and a radioactive tracer intradermally around the tumor before penectomy allow a more accurate identification of the sentinel lymph nodes compared with traditional penile lymphangiography reported by Cabanas in 1977.10 The dynamic lymphoscintigraphy has shown a specificity of 100% and a sensitivity of 78–80%, and its use was validated recently in a prospective study.13, 14 If these data are reconfirmed in other prospective studies, then, in the near future, this technique may become the optimal way to select patients who have lymph node micrometastases as candidates for early lymphadenectomy. Possible limitations to the increased use of this procedure may include the low availability of necessary equipment and the cost.

In conclusion, the clinical lymph node status was found to be the only independent clinical predictor of pathologic lymph node involvement. Bilateral radical lymphadenectomy should be considered in all patients who have palpable inguinal lymph nodes. In the current study, all patients with nonpalpable lymph nodes who had undergone a surveillance program had lymph node progression within 24 months after surgery for the primary tumor.

In the cN0 subgroup, the presence of venous and/or lymphatic embolizations and tumor extension to the corpus spongiosum with or without urethral invasion turned out to be the strongest predictors of pathologic lymph node involvement. Early bilateral inguinal lymphadenectomy should be recommended in all patients with cN0 disease who have those features. These findings should be included in all pathologic reports.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES