Extensive inguinal lymphadenectomy improves overall 5-year survival in penile cancer patients

Results from the Surveillance, Epidemiology, and End Results program

Authors

  • Timothy V. Johnson MD,

    1. Department of Urology, School of Medicine, Emory University, Atlanta, Georgia
    Search for more papers by this author
  • Wayland Hsiao MD,

    1. Department of Urology, School of Medicine, Emory University, Atlanta, Georgia
    Search for more papers by this author
  • Keith A. Delman MD,

    1. Department of Surgery, School of Medicine, Emory University, Atlanta, Georgia
    2. Winship Cancer Institute, School of Medicine, Emory University, Atlanta, Georgia
    Search for more papers by this author
  • Ashesh B. Jani MD, MSEE,

    1. Winship Cancer Institute, School of Medicine, Emory University, Atlanta, Georgia
    2. Department of Radiation Oncology, School of Medicine, Emory University, Atlanta, Georgia
    Search for more papers by this author
  • Otis W. Brawley MD,

    1. American Cancer Society, Atlanta, Georgia
    2. Department of Medical Oncology, School of Medicine, Emory University, Atlanta, Georgia
    Search for more papers by this author
  • Viraj A. Master MD, PhD

    Corresponding author
    1. Department of Urology, School of Medicine, Emory University, Atlanta, Georgia
    2. Winship Cancer Institute, School of Medicine, Emory University, Atlanta, Georgia
    • Department of Urology, School of Medicine, Emory University, 1365 Clifton Road NE, Atlanta, GA 30322
    Search for more papers by this author

Abstract

BACKGROUND:

European Urological Association guidelines recommend potentially curative inguinal lymphadenectomy for certain cases of penile cancer such as grade 3 and pT2-4 lesions, among others. Anecdotally, the authors have noticed that few patients undergo inguinal lymphadenectomy. Therefore, they assessed the frequency of inguinal lymphadenectomy and the impact of dissection extent on survival using the Surveillance, Epidemiology, and End Results (SEER) database.

METHODS:

The authors queried 17 SEER registries from 1988 through 2005 for grade 3 and pT2-4 penile cancer patients without distant metastases. Univariate and multivariate analyses examined predictors of inguinal lymphadenectomy. Kaplan-Meier and Cox regression analyses assessed overall 5-year survival across patient- and disease-related characteristics for patients receiving inguinal lymphadenectomy involving <8 or ≥8 lymph nodes, the latter a surrogate for extent of dissection based on other malignancies.

RESULTS:

Of 593 patients enrolled, only 26.5% received inguinal lymphadenectomy. In addition to grade 3 (P = .031) and pT2-4 disease (P = .004), age <65 years (P < .001) and marital status (P = .002) were significantly associated with receiving lymph node dissection. Increased overall 5-year survival (hazard ratio, 0.54; 95% confidence interval, 0.36-0.79) was observed in patients of all ages who received lymphadenectomy involving ≥8 lymph nodes.

CONCLUSIONS:

A significant number of penile cancer patients at risk for metastases have not received potentially curative inguinal lymphadenectomy. Patients receiving inguinal lymphadenectomy involving ≥8 lymph nodes experienced improved overall 5-year survival. Guidelines should not only be given more emphasis, but possibly be updated to reflect the benefit of extensive lymph node dissection in high-risk penile cancer patients. Cancer 2010. © 2010 American Cancer Society.

The incidence of penile cancer in the United States and Europe is approximately 1 of 100,000 men.1, 2 Lymph node (LN) invasion occurs in 50% of the patients who present with enlarged inguinal LNs and 10% to 15% of the patients who present with nonpalpable LNs.3 Penile cancer exhibits a prolonged locoregional phase before dissemination to distant sites, which differs from other cancers, such as breast cancer, in which LN metastases are presumed to be indicative of systemic disease. As a result, inguinal lymphadenectomy alone can be curative in many cases.1, 4, 5 However, despite recent advances in surgical treatment, the high morbidity associated with open inguinal lymphadenectomy limits its widespread use.1, 2, 5-8

The European Urological Association recommends lymphadenectomy for all patients with clinically palpable inguinal lymphadenopathy, approximately 60% of presenting patients.2 The American Urological Association has not yet issued guidelines addressing this issue. To identify patients with nonpalpable LNs, noninvasive and minimally invasive staging strategies such as sentinel LN sampling may be effective. However, these techniques have not achieved widespread acceptance in the United States.3, 7, 9, 10 Consequently, investigators have shaped guidelines for stratifying these patients.2, 7, 11 According to the European Urological Association's guidelines, patients with low risk of LN metastasis (pTis, pTaG1-2, pT1G1) should be followed, as these patients exhibit a <16.5% incidence of LN metastasis.2, 11, 12 For patients with intermediate risk (T1G2), inguinal lymphadenectomy may be appropriate in the presence of histological factors such as vascular invasion or nodular growth pattern.2, 7, 13 Patients with high risk (pT2-4, G3, or vascular invasion) have a clear indication for inguinal lymphadenectomy, as these patients exhibit a 50% to 70% incidence of LN metastasis.2, 7, 11, 12

Despite this increased risk of metastases, we have anecdotally noticed that relatively few at-risk patients undergo inguinal lymphadenectomy. We have noted that when inguinal lymphadenectomy is performed, LN dissections many times consist of <3 LNs, and sometimes only a needle aspiration is performed. This low LN yield challenges the concept of a curative inguinal lymphadenectomy. Our clinical experience suggests a direct relationship between LNs examined and survival. We propose that a curative inguinal lymphadenectomy requires the dissection of ≥8 LNs, as is recommended by the National Comprehensive Cancer Network (NCCN) for melanoma surgery.14, 15 Some studies suggest this threshold in the treatment of bladder cancer as well, although less of a consensus exists.16 We hypothesize that patients receiving LN dissection of 8 or more LNs will have higher 5-year overall survival. We also hypothesize that a low percentage of patients with grade 3 or T2-4 disease receive lymphadenectomy with a yield of at least 8 LNs. Ours is the first study to test such hypotheses regarding penile cancer using the Surveillance, Epidemiology, and End Results (SEER) database.

MATERIALS AND METHODS

We extracted information on penile cancer patients reported to the 17 registries that comprise the SEER database, which represents 26% of the US population.17 Cases that met the following eligibility criteria were extracted: 1) grade 3 or pT2-4 penile cancer, 2) 1 primary tumor only, 3) age >20 years, and 4) diagnosis between 1988 and 2005. Patients were excluded if they had known metastases at the time of diagnosis or if LN examination data during inguinal lymphadenectomy was unknown or absent. Cases ascertained based on death certificates only or autopsy findings were excluded from the study. Follow-up and vital status information for this study was available through the end of 2005.

All eligible cases were categorized into 2 groups according to whether they received inguinal lymphadenectomy. The 2 groups were compared with respect to age at diagnosis, race, marital status, year of diagnosis, grade, stage, and adjuvant radiation therapy. Frequency analysis and descriptive statistics were performed. Unadjusted associations between various patient characteristics and LN dissection were calculated by calculating crude odds ratios (ORs) with corresponding 95% confidence intervals (CIs). Multivariate logistic regression analyses evaluated the relationship between age, marital status, grade, and stage and adequate LN dissection, adjusting for race, year of diagnosis, and adjuvant radiation therapy. Models were examined for interaction and colinearity.

Kaplan-Meier survival curves were constructed for all patients with grade 3 or pT2-4 disease, with patients stratified by the extent of their LN dissection (<8 LN vs ≥8 LNs) and age at diagnosis (<65 and ≥65 years old). The subgroup with <8 LNs included patients receiving lymphadenectomy involving <8 LNs and patients not receiving lymphadenectomy. Multivariate Cox regression analyses were performed to assess the association between LN dissection involving ≥8 LNs and 5-year overall survival after controlling for age, race, marital status, grade, T classification, and adjuvant radiation. All variables in the model were tested for proportional hazard assumptions, and all models were tested for the presence of interactions between the main independent variable (extent of inguinal lymphadenopathy) and each covariate.

All statistical analyses were performed using SEER*Stat (version 6.4.4; National Cancer Institute, Bethesda, Md) and SPSS 15.0 for Windows (LEAD Technologies, Chicago, Ill).

RESULTS

Patient Demographics

The final study cohort included 593 patients with high-risk penile cancer: 294 (49.7%) patients with grade 3 disease, 234 (39.5%) patients with pT2 disease, 130 (22.0%) patients with pT3 disease, and 39 (6.6%) patients with pT4 disease. It should be noted that although the total cohort consisted of 593 patients, the sum of each category is >593 because categories may overlap. For example, patients may have grade 3 disease and pT2-4. A total of 327 (55.2%) were ≥65 old at the time of diagnosis, and 81.5% of the sampled population was Caucasian. Greater than half the patients (58%) were diagnosed between 2000 and 2005, and over half (54.1%) were married (Table 1). A total of 157 (26.5%) patients had inguinal sampling of at least 1 LN; 24 (4.1%) cases involved only 1 LN, 27 (4.5%) cases involved 2 to 7 LNs, and 107 (18.0%) cases involved ≥8 LNs. Patients were categorized into 2 groups based on whether they received inguinal lymphadenectomy. Patients receiving inguinal lymphadenectomy were more likely to be aged <65 years (P < .001) and currently married (P = .001). There was a trend toward a higher percentage of patients with grade 3 disease (P = .030) and pT2-4 disease (P = .089) in the lymphadenectomy group.

Table 1. Baseline Characteristics in 593 Patients Diagnosed With Grade 3 or T2-T4 Penile Cancer, Stratified by Number of Lymph Nodes Dissected, Using Univariate Analysis
VariablesAll (n=593)Inguinal LymphadenectomyUnivariate Analyses, P
No (n=486)Yes (n=107)
  • ART indicates assisted reproductive technology.

  • a

    County attributes based on 1990 census.

Patient characteristics and county demographicsa
 Age, y   <.001
  <6544.8%38.9%61.1% 
  ≥6555.2%61.1%38.9% 
 Race   .895
  White81.5%81.8%80.3% 
  Black11.1%10.8%12.1% 
  Other7.4%7.4%7.6% 
 Marital status   .004
  Single17.7%15.9%22.9% 
  Married54.1%52.4%58.6% 
  Other28.2%31.7%18.5% 
 Median household income (US$)33,18032,88034,013.538
 % urban80.2%79.5%82.0%.437
 % <high school education24.4%24.4%24.4%.434
 Year of diagnosis   .119
  1988-199317.9%19.5%13.4% 
  1994-199924.1%24.8%22.3% 
  2000-200558.0%55.6%64.3% 
Pathologic features    
 Grade   .100
  Grade 113.7%14.9%10.2% 
  Grade 229.1%26.9%35.0% 
  Grade 349.7%49.9%49.0% 
  Missing/unknown7.6%8.3%5.7% 
 T classification   .089
  T121.1%26.1%15.8% 
  T239.5%42.2%48.2% 
  T322.0%23.1%28.1% 
  T46.6%7.1%7.9% 
  Unknown10.8%1.3%0% 
 % receiving ART11.6%10.3%15.3%.098

Multivariate Logistic Regression Analysis

Patients with grade 3 (OR, 2.18; 95% CI, 1.07-4.42) and pT2-4 (OR, 2.54; 95% CI, 1.35-4.80) disease were more likely to receive LN dissections compared with patients with TxG1 or T1Gx disease, respectively (Table 2). In addition, divorced, widowed, or separated patients (OR, 0.340; 95% CI, 0.17-0.67) and patients aged ≥65 years (OR, 0.39; 95% CI, 0.25-0.61) were less likely to receive inguinal dissections compared with single or married patients or patients <65 years old, respectively.

Table 2. Multivariate Analysis of Predictors of Receiving Inguinal Lymphadenectomy
VariableCrude OR (95% CI)PAdjusted OR (95% CI)P
  1. OR indicates odds ratio; 95% CI, 95% confidence interval.

Age ,y    
 <651 1 
 ≥650.40 (0.28-0.59)<.0010.39 (0.25-0.61)<.001
Marital status    
 Single1 1 
 Currently married0.77 (0.48-1.24).2840.84 (0.49-1.44).528
 Divorced, widowed, separated0.40 (0.23-0.71).0020.34 (0.17-0.67).002
T classification    
 T11 1 
 T2-T41.92 (1.15-3.18).0122.54 (1.35-4.80).004
Grade    
 Grade 11 1 
 Grade 21.91 (1.01-3.60).0451.96 (1.01-3.78).046
 Grade 31.44 (0.79-2.64).2372.18 (1.07-4.42).031
Adjuvant radiotherapy    
 Yes1 1 
 No0.64 (0.38-1.09).1000.89 (0.43-1.57).547

Kaplan-Meier Survival Curves

Kaplan-Meier analysis was performed on patients stratified by ≥8 or <8 LN dissected. Although any inguinal lymphadenectomy improved survival, dissection of ≥8 LNs seemed to confer the greatest survival benefit. The mean (standard deviation) overall 5-year survival for patients with grade 3 penile cancer receiving inguinal lymphadenectomy involving ≥8 LNs and <8 LNs was 66.3% (9.0) and 49.2% (4.6), respectively (P = .010) (Fig. 1). When patients with pT2-4 were analyzed, there was 70.2% (6.3) and 53.6% (3.9) 5-year overall survival, respectively (P = .004) (Fig. 2).

Figure 1.

The 5-year overall survival is shown for all patients with grade 3 penile cancer stratified by lymph node (LN) dissection: ≥8 LNs dissected (8+) or <8 LNs dissected (<8). SE indicates standard error.

Figure 2.

The 5-year overall survival is shown for all patients with pT2-4 penile cancer stratified by lymph node (LN) dissection: ≥8 LNs dissected (≥8) or <8 LNs dissected (<8). SE indicates standard error.

Cox Regression Survival Analyses

Cox regression analysis was conducted to identify predictors of overall 5-year survival in patients with penile cancer (Table 3). As expected, pT2-4 (hazard ratio [HR], 1.60; 95% CI, 1.12-2.30) and grade 3 disease (HR, 3.22; 95% CI, 2.00-5.30) were associated with decreased overall 5-year survival, whereas radiation therapy after lymphadenectomy was associated with increased survival (HR, 0.58; 95% CI, 0.41-0.84). Of the demographic variables, only age ≥65 years was associated with decreased survival (HR, 1.33; 95% CI, 1.02-1.74). After adjusting for grade, stage, adjuvant radiation, age, and other demographic variables, not receiving inguinal lymphadenectomy involving ≥8 LNs significantly reduced overall 5-year survival (HR, 1.86; 95% CI, 1.260-2.76).

Table 3. Multivariate Cox Survival Analyses of the Association Between 5-Year Overall Mortality and Various Patient, Disease, and Treatment Characteristics Among Patients With Penile Cancer, Stratified by Age
VariableHR (95% CI)P
  1. HR indicates hazard ratio; 95% CI, 95% confidence interval.

Age, y  
 <651 
 ≥651.33 (1.02-1.74).037
Marital status  
 Never married1 
 Currently married1.18 (0.81-1.71).386
 Divorced, widowed, separated1.26 (0.84-1.90).663
T classification  
 T11 
 T2-41.60 (1.12-2.30).010
Grade  
 Grade 11 
 Grade 22.80 (1.74-4.50)<.001
 Grade 33.22 (1.96-5.30)<.001
Adjuvant radiotherapy  
 No1 
 Yes0.58 (0.41-0.84).004
Lymph node dissection  
 ≥8 lymph nodes1 
 <8 lymph nodes1.86 (1.26-2.76).002

DISCUSSION

Inguinal Lymphadenectomy Guidelines Are Incompletely Followed

Current penile cancer guidelines from a major urological association recommend inguinal lymphadenectomy for patients with palpable inguinal lymphadenopathy, high-grade disease (grade 3), or pT2-4 disease. No other association has guidelines concerning the treatment of regional LNs for penile cancer, perhaps reflecting the rarity of this disease. Because of the paucity of guidelines and the rather small cohorts used in articles on penile cancer, a large national dataset may be of benefit. By using the SEER-17 database, we identified 593 patients with clinically localized penile cancer (without palpable lymphadenopathy) who met at least 1 of the criteria for high-risk disease (Table 1).

Overall, only 26.5% of these patients received the recommended inguinal lymphadenectomy. However, only 17.9% of all patients received a dissection involving ≥8 LNs. Although inguinal lymphadenectomy is potentially curative in penile cancer patients, our findings appear to demonstrate an underutilization of inguinal lymphadenectomy, a potentially curative procedure. Perhaps, the morbidity associated with inguinofemoral lymphadenectomy limits its widespread use, or perhaps knowledge of the appropriate inguinal lymphadenectomy guidelines are not widely known. The literature has clearly shown the impact of inguinal lymphadenectomy on improved patient survival.1-3, 6, 7 Consequently, measures should be taken to educate urologists regarding the benefits of inguinofemoral lymphadenectomy. Recent advances using minimally invasive approaches to inguinofemoral lymphadenectomy, with consequent reduced morbidity for patients, may help to increase the use of LN dissection.18

Dissection Involving ≥8 Lymph Nodes Increases 5-year Survival

Our data indicate that dissection of any number of LNs improved survival. Furthermore, there appeared to be an almost direct relationship between the extent of lymphadenectomy and survival. However, observing that NCCN guidelines for melanoma historically recommend a dissection of at least 8 LNs (plus sentinel LNs) for inguinal lymphadenectomy, and that a recent publication analyzing a large series of patients supports this number, we hypothesized that a dissection involving at least 8 LNs in penile cancer would improve survival.15 However, we recognize that 8 LNs serve as a marker for an adequate anatomic dissection but should not serve as a minimum number needed to perform an adequate dissection, as there is inherent variability in pathological assessment of LNs as well as anatomic variations in number of LNs.

For high-risk patients with either grade 3 (Fig. 1) or pT2-4 disease (Fig. 2), total LN yield of at least 8 LNs clearly improves overall 5-year survival. These findings suggest that clinical guidelines should not only recommend inguinal lymphadenectomy for high-risk penile cancer patients, but that patients may benefit from an extended lymphadenectomy rather than a limited sampling of the LNs.

As expected, high grade (P < .001), T classification (P = .010), and adjuvant radiotherapy (P = .004) were significant predictors of survival. After controlling for these variables as well as sociodemographics, inguinal lymphadenectomy involving ≥8 LNs was significantly associated with improved survival, nearly doubling overall 5-year survival compared with dissection of <8 LNs (P = .002). Although LN count is a multifactorial variable, the association between increased LNs and improved survival should encourage surgeons to pursue as extensive an anatomic dissection as possible, rather than pursuing a more limited approach under the pretense of avoiding morbidity.

Study Limitations

There are several limitations that are inherent in the type of study we have conducted. SEER captures approximately 26% of the population, but we have not studied the entire US population.17 Importantly, SEER data does not include patient comorbidities, which may be significant confounding factors in the original treatment process, perhaps especially in patients aged >65 years. Also, SEER does not report vascular invasion, which impacts the decision to perform lymphadenectomy on intermediate-risk patients. Therefore, it is unclear whether these results apply to all patients receiving lymphadenectomy. SEER also does not describe laterality of LNs, so it is unclear whether LN yield of ≥8 LNs implies unilateral or bilateral dissection. In addition, “examined LNs” refers to examination by a pathologist, which may not reflect the actual number of LNs resected because of heterogeneity of processing of LNs. Another limitation is that SEER does not disclose risk factors, such as circumcision. SEER also does not report chemotherapy treatment, which can reduce recurrence and improve survival in penile cancer.19 Finally, SEER does not capture cost data, which is especially important when considering the complications and morbidities associated with extensive inguinal lymphadenectomy, or perhaps even the potential costs associated with not performing this procedure.

Conclusions

During the past 2 decades, a significant number of high-risk penile cancer patients have not received inguinal lymphadenectomy with curative intent. Of patients receiving LN dissections, those involving examination of ≥8 LNs experienced significantly higher overall 5-year survival. These findings reinforce the European Urological Association guidelines and may suggest that the guideline should be updated to reflect the benefit of extensive LN dissection in high-risk penile cancer patients.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary