The Relationship between Characteristics of Inguinal Lymph Nodes and Pelvic Lymph Node Involvement in Penile Squamous Cell Carcinoma: A Single Institution Experience

The Relationship between Characteristics of Inguinal Lymph Nodes and Pelvic Lymph Node Involvement in Penile Squamous Cell Carcinoma: A Single Institution Experience Giovanni Lughezzani, Mario Catanzaro, Tullio Torelli, Luigi Piva, Davide Biasoni, Silvia Stagni, Alessandro Crestani, Andrea Guttilla, Daniele Raggi, Patrizia Giannatempo, Andrea Necchi, Giorgio Pizzocaro, Maurizio Colecchia, Roberto Salvioni and Nicola Nicolai* From the Departments of Urology (GL, MC, TT, LP, DB, SS, AC, AG, RS, NN, GP), Medical Oncology (DR, PG, AN) and Pathology (MC), Fondazione Istituto di Ricovero e Cura a Carattere Scientifico Istituto Nazionale dei Tumori, Milano, Italy

Purpose: We determined predictors of pelvic lymph node metastases in patients with penile cancer. Materials and Methods: We retrieved a total of 188 node positive inguinal groins from 142 patients treated for penile cancer. Logistic regression models were fitted to test for predictors of pelvic lymph node metastases. The minimum p value method was used to determine the most significant cutoff values of each predictor. Results: Pelvic lymph node metastases were observed in 45 cases (31.7%). The 5-year cancer specific survival rate was 71.0% vs 33.2% in patients with inguinal vs pelvic lymph node metastases. The most significant cutoff values were 3 inguinal lymph node metastases and a metastasis diameter of 30 mm. According to univariable logistic regression models the number of inguinal metastases (OR 1.92, p <0.001), the diameter of the metastases (OR 1.03, p ¼ 0.001) and extranodal extension (OR 8.01, p <0.001) were significant predictors of pelvic lymph node metastases. These variables were also independent predictors of metastases in multivariable logistic regression models (p 0.012). Patients with 3 or more inguinal lymph node metastases and those with a metastasis diameter of 30 mm or greater were at 4.77 and 2.53-fold higher risk, respectively, of harboring pelvic lymph node metastases (p 0.006). The proportion of metastases increased significantly from 0% in cases with no risk factors to 57.1% when all 3 risk factors were observed (p <0.001). Conclusions: The number and diameter of inguinal lymph node metastases as well as extranodal extension are significantly associated with pelvic lymph node metastases. These variables should be considered to determine the need for pelvic lymph node dissection. Patients with no risk factors may be spared this dissection.

Abbreviations and Acronyms CSS ¼ cancer specific survival DSNB ¼ dynamic sentinel lymph node biopsy EAU ¼ European Association of Urology ENE ¼ extracapsular extension LND ¼ lymph node dissection LNM ¼ lymph node metastases RSS ¼ risk stratification score SCC ¼ squamous cell carcinoma Accepted for publication October 21, 2013. * Correspondence: Department of Urology, Fondazione Istituto di Ricovero e Cura a Carattere Scientifico Istituto Nazionale dei Tumori, Via Venezian 1, 20133, Milano, Italy (telephone: þ39 02/23902163; e-mail: nicola.nicolai@ istitutotumori.mi.it).

Key Words: penis; carcinoma, squamous cell; lymph node excision; neoplasm metastasis; prognosis PELVIC LNM have a tremendously detrimental effect on the survival of patients with penile SCC.1e3 Those with pathologically determined

inguinal LNM are at increased risk for further lymphatic spread to the pelvic lymph nodes.3,4 However, the proportion of patients with pelvic

0022-5347/14/1914-0977/0 THE JOURNAL OF UROLOGY® © 2014 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.

http://dx.doi.org/10.1016/j.juro.2013.10.140 Vol. 191, 977-982, April 2014 Printed in U.S.A.

www.jurology.com

j

977

978

INGUINAL LYMPH NODES AND PELVIC LYMPH NODE INVOLVEMENT IN PENILE CANCER

LNM among those with inguinal LNM is extremely variable. It is difficult to assess pelvic lymph nodes with conventional imaging.5,6 Pelvic LND is frequently performed as an adjunct to inguinal LND but it may represent overtreatment in a nonnegligible proportion of patients, resulting in additional morbidities and hospitalization costs.7,8 According to the 2010 EAU penile cancer guidelines pelvic LND should be done when 2 or more metastatic inguinal lymph nodes are found at inguinal LND or ENE is observed in at least 1 inguinal lymph node metastasis.9 We evaluated predictive factors associated with pelvic LNM in a contemporary population of patients with pathologically determined inguinal LNM.

MATERIALS AND METHODS Study Population At our institution 142 consecutive patients with penile SCC and pathologically ascertained inguinal LNM were treated between 1985 and 2012. A total of 188 node positive inguinal groins were evaluated and ipsilateral pelvic lymph node involvement was determined. A total of 28 patients with no clinical evidence of pelvic LNM underwent inguinal LND only. Since they did not experience recurrent disease at a minimum 2-year followup postoperatively, they were included in study and grouped together with individuals without evidence of pelvic LNM at pathological evaluation.

Indications and Surgical Technique At our institution the indications for and extension of LND for penile cancer slightly changed with time. Until the 2000s only patients with clinically evident nodal disease and those with clinical T2 or greater disease were selected for bilateral inguinopelvic LND. Patients with T1 and grade 1 or 2 SCC of the penis were initially spared from surgery.10 Since 2001, only patients with a T1 G1 tumor have been spared LND (fig. 1). Inguinal LND was performed in patients with 1) clinical N0 disease in whom DSNB revealed metastases in biopsied node(s) or who had no tracer accumulation in the groin, 2) clinical N0 disease who were at intermediate or high risk for lymph node involvement according to institutional guidelines (all G2 or greater and all pT2 or greater) and who did not undergo DSNB and 3) clinically resectable cNþ disease. Synchronous pelvic LND was performed in patients with cN2 or greater inguinal metastases or pelvic lymph nodes suspicious on preoperative cross-sectional imaging. According to EAU guidelines, delayed pelvic LND was planned when 2 or more positive inguinal lymph nodes were found at pathological examination or ENE was observed in the ipsilateral groin. Inguinal LND consisted of the removal of superficial and deep inguinal lymph nodes.11 Pelvic LND consisted of the removal of external iliac and obturator lymph nodes.

Node Count and Staging Any lymph node containing viable cancer was counted as positive as well as lymph nodes showing complete

Figure 1. Decisional algorithm for inguinal and pelvic LND in patients with penile cancer treated at our institution since 2001. ILND, inguinal LND. þve, positive. SN, sentinel node. lnm, LNM. PLND, pelvic LND.

pathological remission, considered fibrosis after chemotherapy.12e14 A fixed or gross nodal mass was counted as 1 lymph node regardless of size, and as ENE.12 We used the UICC TNM, 7th edition to determine clinical and pathological nodal category. Pathology reports included the number of lymph nodes removed from each groin, number of positive lymph nodes, ENE and lymph node diameter (maximum diameter determined by the pathologist).

Statistical Analysis We used the chi-square and Mann-Whitney tests to compare categorical and continuous variables, respectively. The Kaplan-Meier method was used to explore CSS rates and the log rank test was used to compare survival curves. The minimum p value approach according to Mazumdar and Glassman15 was applied to determine the most significant cutoff value for each variable. Univariable and multivariable logistic regression models were fitted to test the predictors of the primary outcome (pelvic LNM), defined as LNM on pathological evaluation on the same side where positive inguinal lymph nodes were detected. Univariable and multivariable predictive accuracy values were quantified according to the ROC AUC. The method of DeLong et al16 was used to test the statistical significance of different accuracy values. All statistical tests were performed using S-PLUS Professional, version 1 (MathSoftÒ) or SPSSÒ, version 17.0. All tests were 2-sided with statistical significance considered at 0.05.

RESULTS Descriptive Analysis Median patient age was 63 years (IQR 53e72). The median number of inguinal lymph nodes removed was 9 (IQR 6e12) and the median number of positive inguinal nodes was 2 (IQR 1e3). The median diameter of positive inguinal lymph nodes was

INGUINAL LYMPH NODES AND PELVIC LYMPH NODE INVOLVEMENT IN PENILE CANCER

25 mm (IQR 15e40). ENE was observed in 98 inguinal groins (52.1%). Overall we performed 160 inguinopelvic (85.2%) and 28 inguinal (14.8%) LNDs. Pelvic LNM were pathologically detected in 45 of 142 patients (31.7%). Chemotherapy was administered preoperatively and/or postoperatively in 102 patients (71.8%). Of the patients 44.4% received adjuvant chemotherapy only and 24.6% received neoadjuvant chemotherapy (supplementary table 1, http:// jurology.com/). Chemotherapy regimens were heterogeneous, including a combination of cisplatin and 5-fluorouracil (plus taxanes in more recent cases) with vincristine, bleomycin and oral methotrexate used more frequently.17 Adjuvant chemotherapy was more often administered to patients with pelvic LNM than those with inguinal LNM

979

only (60.0% vs 37.1%). In contrast, the proportion of patients who received neoadjuvant chemotherapy did not differ between the 2 groups. Mean followup was 51 months. CSS statistically differed according to the extent of nodal involvement (log rank test p <0.001, fig. 2). Specifically, the 5-year CSS rate was 71.0% in patients with inguinal LNM only vs 33.2% in those with pelvic LNM. Patients with pelvic LNM had a significantly higher number of inguinal positive lymph nodes than patients without pelvic LNM (median 2 vs 1, p <0.001). Similarly, the median diameter of metastatic inguinal nodes was significantly greater in patients with pelvic LNM (34 vs 20 mm, p ¼ 0.004). The incidence of ENE was significantly higher in individuals with pelvic LNM than those with inguinal LNM only (84.3% vs 40.1%, p <0.001).

Figure 2. Kaplan-Meier curves show CSS rates by pelvic lymph node involvement

980

INGUINAL LYMPH NODES AND PELVIC LYMPH NODE INVOLVEMENT IN PENILE CANCER

According to the minimum p value approach the most significant cutoffs were 3 or greater for the number of positive inguinal lymph nodes and 30 mm or greater for lymph node diameter. Univariable and Multivariable Logistic Regression Analyses Predicting Pelvic LNM On univariable logistic regression analyses ENE (OR 8.01, 95% CI 3.50e18.35, p <0.001), the number of positive inguinal lymph nodes (OR 1.92, 95% CI 1.45e2.55, p <0.001) and positive inguinal lymph node diameter (OR 1.03, 95% CI 1.01e1.05, p ¼ 0.001) were significant predictors of pelvic LNM (table 1). Similar results were obtained when considering categorical parameters according to their most significant cutoffs. Patients with 3 or more positive inguinal lymph nodes were at 4.77-fold higher risk (95% CI 2.35e9.68) for harboring pelvic LNM than those with 1 or 2 metastatic lymph nodes (p <0.001). A positive inguinal lymph node diameter of 30 mm or greater was associated with a 2.53-fold higher risk of pelvic LNM (95% CI 1.310e4.900, p ¼ 0.006). In multivariable logistic regression models all 3 variables (ENE, number of positive lymph nodes and lymph node diameter) emerged as independent predictors of pelvic LNM independently of whether variables were coded continuously or categorically (each p 0.012, table 1). On accuracy analysis ENE was the most informative predictor of pelvic LNM (AUC 72.2%), followed by the continuously coded number of positive inguinal lymph nodes (70.2%) and lymph node diameter (65.9%). Combining the 3 parameters led to significant improvement in the discriminative ability of each single predictor (AUC 81.1% to 81.9%, p <0.001). Risk Stratification Score Risk categories were defined according to the presence of the 3 risk factors, including ENE, 3 or more positive nodes and a positive node diameter of

30 mm or greater. The 3 parameters were combined to create a RSS to determine the probability of pelvic LNM based on inguinal LNM characteristics, including 0dno risk factor, 1d1 factor, 2d2 factors and 3d3 factors (table 2). The proportion of pelvic LNM significantly increased from 0% (0 of 49 cases) when there were no risk factors to 57.1% (8 of 14) when all 3 risk factors were observed (p <0.001). Logistic regression models confirmed these findings. Specifically, the presence of 2 or 3 risk factors was associated with a 10.2-fold higher risk (95% CI 4.73e21.99) of harboring LNM relative to that with 0 or 1 risk factor (p <0.001). Interestingly, on univariable logistic regression analysis the RSS showed the highest discriminative ability for predicting pelvic LNM (AUC 76.0%).

DISCUSSION Indications for pelvic LND in patients with penile cancer remain controversial. This topic is based on 2 main premises. 1) About 20% to 25% of patients with inguinal LNM from penile SCC have pelvic LNM. This proportion is not negligible but it indicates that most patients with inguinal LNM may be spared pelvic LND. 2) Pelvic lymph node involvement is a well-known prognostic factor in patients with penile SCC. Historical reports of sparse data indicate only 10% 5-year survival.18 Two more recent series described no survivors of 21 patients at 5 years and a 16% rate of long-term survival in 25 patients.8,19 On the other hand, we noted 33% 5-year CSS, substantially better than previous findings. Therefore, we assume that a nonnegligible proportion of patients with pelvic LNM may be cured after lymph node excision that includes pelvic lymph nodes. Individuating patients at different levels of risk for harboring pelvic LNM represents an important clinical need. Preoperative computerized tomography has high specificity but limited sensitivity

Table 1. Univariable and multivariable logistic regression analyses predicting pelvic LNM based on inguinal lymph node characteristics Univariable % Predictive Accuracy

ENE (yes vs no) Pos inguinal lymph nodes: No. (continuous) Diameter (continuous) No. (categorical 1-2 vs 3 or greater) Diameter (categorical less than 30 mm vs 30 or greater) Risk category (0-1 vs 2-3)

72.2

8.014 (3.500e18.348) <0.001 6.288 (2.564e15.417) <0.001 6.610 (2.771e15.766) <0.001

70.2 65.9 66.2 61.3

1.921 1.028 4.766 2.534

* Predictive accuracy 81.9%. † Predictive accuracy 81.1%.

p Value

OR (95% CI)

p Value

Multivariable, Categorical†

Predictors

76.0

OR (95% CI)

Multivariable, Continuous*

OR (95% CI)

p Value

(1.445e2.554) <0.001 1.939 (1.413e2.661) <0.001 e e (1.011e1.045) 0.001 1.026 (1.006e1.046) 0.012 e e (2.346e9.680) <0.001 e e 4.726 (2.115e10.563) <0.001 (1.310e4.900) 0.006 e e 2.681 (1.240e5.797) 0.012

10.202 (4.733e21.992) <0.001

e

e

e

e

INGUINAL LYMPH NODES AND PELVIC LYMPH NODE INVOLVEMENT IN PENILE CANCER

Table 2. Patients with vs without pelvic LNM by EAU criteria vs proposed criteria for pelvic LND No. Inguinal No. Inguinal þ p Value Only (%) Pelvic (%) (chi-square test) EAU No. pos inguinal nodes: Less than 2 2 or Greater ENE: Absent Present

<0.001

78 (84.8) 59 (61.5)

14 (15.2) 37 (38.5)

82 (91.1) 55 (56.1) Proposed

8 (8.9) 43 (43.9)

No. pos inguinal nodes: Less than 3 114 (81.4) 3 or Greater 23 (47.9) Pos inguinal node diameter (mm): Less than 30 85 (81.0) 30 or Greater 52 (62.7) ENE: Absent 82 (91.1) Present 55 (56.1) Risk category: 0 49 (100) 1 52 (82.5) 2 30 (48.4) 3 6 (42.9)

<0.001

<0.001 26 (18.6) 25 (52.1) 0.005 20 (19.0) 31 (37.3) 8 (8.9) 43 (43.9)

<0.001 <0.001

0 11 (17.5) 32 (51.6) 8 (57.1)

(20% to 37.5%) for identifying pelvic LNM.5,6 Positron emission tomography/computerized tomography showed high accuracy for detecting pelvic LNM20 but these findings must be validated. Clinical and pathological information may also be associated with pelvic LNM. Lont et al reported that the tumor grade of involved lymph nodes and 2 or more inguinal lymph nodes were independent predictors of pelvic LNM in 102 patients.8 Similarly, Zhu et al noted that the number of positive inguinal lymph nodes (3 or more), the lymph node ratio (30% or greater), ENE and p53 mutation were associated with pelvic LNM in a small series of 33 cases.5 In another study of 27 positive inguinal groins Zhu et al found that ENE was associated with pelvic LNM.21 Our study shows that 3 parameters (the number of positive lymph nodes, maximum diameter of positive lymph nodes and presence of ENE) are independently associated with the presence of pelvic LNM. Specifically, 3 or more positive inguinal lymph nodes represented the optimal cutoff value to predict pelvic LNM. Interestingly, the proportion of patients with pelvic LNM increased from 38.5% of those with 2 or more positive inguinal lymph nodes to 52.1% of those with 3 or more positive inguinal nodes. The proportion of patients with pelvic LNM did not differ among those with fewer than 2 vs fewer than 3 positive inguinal nodes (15.2% vs 18.6%). Also, inguinal lymph node diameter was associated with pelvic LNM and 30 mm was the most significant cutoff. In addition, we confirmed the important relationship between ENE and pelvic LNM. Based on

981

our findings we developed a RSS that divides patients with inguinal LNM into 4 risk categories of harboring pelvic LNM. Specifically, the proportion of pelvic LNM progressively increased from 0% for no risk factor to 17.5% for 1 to 51.6% for 2 to 57.1% for 3. Based on historical data the 2010 EAU penile cancer guidelines recommend pelvic LND when 2 or more inguinal nodes were involved by disease or ENE was observed in at least 1 inguinal node.4,9 Based on our RSS pelvic LND could be safely omitted in patients with no risk factors, who represented 35.8% of those in our series. Pelvic LND should be performed in all patients with 2 or more risk factors (26.3% of our series) since more than half will harbor pelvic LNM. Unfortunately, patients with only 1 risk factor (37.9% of our series) have a probability of pelvic LNM that does not differ from the pretest probability of the population as a whole. They should be prudently considered for pelvic LND. From the clinical perspective future directions may include attempts to determine the number and largest diameter of metastatic inguinal nodes on preoperative imaging.5,6 When obtained preoperatively, this information may help plan the indication for and extension of lymph node dissection, and may help avoid repeat and delayed surgical interventions.22,23 The retrospective nature of our series represents a potential limitation. However, we included only patients treated after 1985 because treatment has not significantly changed since then. Preoperative chemotherapy could be considered a limitation due to the potential disappearance of some LNM after treatment and to the shrinkage of responsive LNM. Less than 10% of our cases showed complete pathological remission, which mirrors other experiences.18 These cases counted as having positive nodes. Although our rate of partial response was approximately 50%, in line with other data,22 the diameter of metastatic lymph nodes remained significant on multivariable analysis. This finding did not change after excluding patients undergoing neoadjuvant therapy (supplementary tables 2 and 3, and supplementary figure, http://jurology.com/). Since our pathologists did not routinely record lymphovascular invasion in the primary tumor, we were limited in accounting for this parameter in our analyses. In a previous study a conglomerated nodal mass counted as 1 lymph node.12 Despite the logical criticism that the true number of metastatic nodes is unknown in such cases, the number of lymph nodes (including counting a nodal mass as 1) remained statistically significant on multivariable analysis.

982

INGUINAL LYMPH NODES AND PELVIC LYMPH NODE INVOLVEMENT IN PENILE CANCER

The inclusion of 28 patients who underwent inguinal LND may also be considered a limitation. However, we only included those with no relapse at a minimum 2-year followup after inguinal LND. As such, it is highly unlikely that they had disease that metastasized to the pelvic lymph nodes. Moreover, results of an analysis that included patients who underwent inguinopelvic dissection were virtually the same (data not shown). Finally, the clinical staging of pelvic lymph nodes is not routinely recorded in our database. Nonetheless, this information is controversial and time dependent since we cannot assume that the accuracy of radiological imaging did not change with time. On the other hand, the pathological characteristics of inguinal lymph nodes remain the most important parameter to be evaluated to determine the risk of pelvic lymph node involvement.

CONCLUSIONS Inguinal metastases in 3 or more inguinal nodes, a pathological lymph node diameter of 30 mm or greater and the presence of ENE could represent strong, independent predictors of pelvic LNM. According to our findings patients with no risk factors may avoid pelvic LND while this should be considered in those with 1 or more risk factors. Although our results await further validation, identifying these factors predictive of inguinal LNM by noninvasive diagnostic techniques may allow for the planning of concomitant inguinal and pelvic dissection in most patients, avoiding additional surgery and leading to a decreased duration of cure.

ACKNOWLEDGMENTS Ilaria Lofaro assisted with database maintenance and data retrieval.

REFERENCES 1. Leijte JA, Gallee MP, Antonini N et al: Evaluation of current TNM classification of penile carcinoma. J Urol 2008; 180: 933.

9. Pizzocaro G, Algaba F, Horenblas S et al: EAU penile cancer guidelines 2009. Eur Urol 2010; 57: 1002.

2. Ornellas AA: Management of penile cancer. J Surg Oncol 2008; 97: 199.

10. Pizzocaro G, Piva L and Nicolai N: Lymphadenectomy for cancer of the penis. In: Lymph Node Surgery in Urology. Edited by JP Donohue. Oxford, United Kingdom: Isis Medical Media 1995; chapt 14, pp 118e130.

3. Protzel C, Alcaraz A, Horenblas S et al: Lymphadenectomy in the surgical management of penile cancer. Eur Urol 2009; 55: 1075. 4. Ornellas AA, Seixas AL, Marota A et al: Surgical treatment of invasive squamous cell carcinoma of the penis: retrospective analysis of 350 cases. J Urol 1994; 151: 1244. 5. Zhu Y, Zhang SL, Ye DW et al: Predicting pelvic lymph-node metastases in penile cancer patients: a comparison of computed tomography, Cloquet’s node, and disease burden of inguinal lymph-nodes. Onkologie 2008; 31: 37. 6. Graafland NM, Teertstra HJ, Besnard AP et al: Identification of high risk pathological node positive penile carcinoma: value of preoperative computerized tomography imaging. J Urol 2011; 185: 881.

11. Daseler EH, Anson BJ and Reimann AF: Radical excision of the inguinal and iliac lymph node glands; a study based upon 450 anatomical dissection and upon supportive clinical observations. Surg Gynecol Obstet 1948; 87: 679. 12. Carver BS, Cronin AM, Eggener S et al: The total number of retroperitoneal lymph-nodes resected impacts clinical outcome after chemotherapy for metastatic testicular cancer. Urology 2010; 75: 1431. 13. Bermejo C, Busby JE, Spiess PE et al: Neoadjuvant chemotherapy followed by aggressive surgical consolidation for metastatic penile squamous cell carcinoma. J Urol 2007; 177: 1335.

7. Nelson BA, Cookson MS, Smith JA Jr et al: Complications of inguinal and pelvic lymphadenectomy for squamous cell carcinoma of the penis: a contemporary series. J Urol 2004; 172: 494.

14. Joerger M, Warzinek T, Klaeser B et al: Major tumor regression after paclitaxel and carboplatin polychemotherapy in a patient with advanced penile cancer. Urology 2004; 63: 778.

8. Lont AP, Kroon BK, Gallee MP et al: Pelvic lymph node dissection for penile carcinoma: extent of inguinal lymph-node involvement as an indicator for pelvic lymph-node involvement and survival. J Urol 2007; 177: 947.

15. Mazumdar M and Glassman JR: Categorizing a prognostic variable: review of methods, code for easy implementation and applications to decision-making about cancer treatments. Stat Med 2000; 19: 113.

16. DeLong ER, DeLong DM and Clarke-Pearson DL: Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988; 44: 837. 17. Pizzocaro G, Nicolai N, Piva L et al: Chemotherapy for cancer of the penis. In: Principles and Practice of Genitourinary Oncology. Edited by D Raghavan, HI Scher, SA Leibel et al. Philadelphia: LippincottRaven 1997; chapt 95, pp 973e977. 18. Sanchez-Ortiz RF and Pettaway CA: The role of lymphadenectomy in penile cancer. Urol Oncol 2004; 22: 236. 19. Pandey D, Mahajan V and Kannan RR: Prognostic factors in node-positive carcinoma of the penis. J Surg Oncol 2006; 93: 133. 20. Graafland NM, Leijte JA, Valdes Olmos RA et al: Scanning with 18F-FDG-PET/CT for detection of pelvic nodal involvement in inguinal node positive penile carcinoma. Eur Urol 2009; 56: 339. 21. Zhu Y, Zhang SL, Ye DW et al: Prospectively packaged ilioinguinal lymphadenectomy for penile cancer: the disseminative pattern of lymph-node metastasis. J Urol 2009; 181: 2103. 22. Pagliaro LC, Williams DL, Daliani D et al: Neoadjuvant paclitaxel, ifosfamide, and cisplatin chemotherapy for metastatic penile cancer: a phase II study. J Clin Oncol 2010; 28: 3851. 23. Pizzocaro G, Nicolai N and Milani A: Taxanes in combination with cisplatin and fluorouracil for advanced penile cancer: preliminary results. Eur Urol 2009; 55: 546.