Introduction

Penile cancer is a locally aggressive disease which leads to mutilating treatments. It is more common in developing countries and shows progressive reduction in incidence in developed regions—as a consequence, it can be considered a neglected disease [1–4]. Northeast and North of Brazil, the two less developed regions of the country, concentrates most cases [5]. The Brazilian Society of Urology estimates that 42 and 40 % of all cases in Brazil occur in Northeast and North regions, respectively (unpublished data). The Brazilian Ministry of Health estimates a thousand penile amputations per year with 363 annual deaths due to penile cancer (year 2010). The state of Bahia, in the Northeast region of Brazil, has one of the highest incidences of penile cancer in Brazil. It is a general assumption between physicians that the disease is heterogeneous on clinical evolution and outcome. In order to describe epidemiological and pathological features of penile cancer in a high-risk area of Brazil, we reviewed cases from Hospital Aristides Maltez at Salvador, Bahia—the main institution in the state which provides oncologic treatment for penile cancer in the public health system.

Materials and methods

Patients and study design

This is a cross-sectional study. It was based in Hospital Aristides Maltez: the main institution that practice oncologic assistance for penile cancer for 351 municipalities of the state of Bahia. All consecutive patients who sought medical assistance on the division of Urology from January 1997 to December 2007 with a diagnosis of penile squamous cell carcinoma were eligible to be included in the study. Exclusion criteria were as follows: primary tumors surgically resected in other institutions, no tumor staging information on medical records, no available histological slides for review, and available histological slides that could not be used to confirm diagnosis due to technical issues. Clinical follow-up was recorded, and patients were grouped in the following groups: alive free of disease, alive with disease, dead in consequence of the disease or complications of treatment, dead due to other cases, and lost from follow-up. The study was approved by the Research Ethics Committee of Hospital Aristides Maltez located in Salvador, Brazil (224/08).

Tumor subtyping and grading

Tumor slides were reviewed and subtypes were classified or reclassified based on the World Health Classification of Tumors (2004) [6]. Tumor grading was performed as follows: Grade 1 is an extremely well-differentiated carcinoma, with a minimal deviation from the morphology of normal/hyperplastic squamous epithelium; Grade 2 tumors show disorganized growth as compared to grade 1 lesions, higher nuclear-to-cytoplasmic ratio, evident mitoses, and, although present, less prominent keratinization; Grade 3 are tumors showing any proportion of anaplastic cells, identified as solid sheets or irregular small aggregates, cords or nests of cells with little or no keratinization, high nuclear-to-cytoplasmic ratio, thick nuclear membranes, nuclear pleomorphism, clumped chromatin, prominent nucleoli, and numerous mitosis [7, 8]. Cytological features (koilocytes) suggestive of HPV infection were evaluated. No method based on direct detection of viral DNA was employed.

Statistical analyses

Statistical analyses were performed using the SPSS 17.0 software package. Categorical data were compared by a Chi-square test for trend, and the numerical data were compared using a Student’s t test. Kaplan–Meier curves were established to compare survival from different groups. Curves were compared using log-rank (Mantel–Cox) test. A p value <0.05 was considered significant.

Results

From January 1997 to December 2007, 378 patients with diagnosis of penile cancer were referred to the Division of Urology of the Hospital Aristides Maltez. Lack of available data leads to exclusion of 30 cases. Among 348 cases of penile cancer, 32 were only biopsied in the referral center. Local excision/postectomy was performed in 28 (8 %), partial penectomy in 262 (75 %), total penectomy in 8 (2 %), and emasculation in 18 (5 %) cases. Seventy-seven patients underwent lymphadenectomy.

Thirty-four patients were excluded for different reasons (e.g., only biopsies available), remaining 314 patients for further clinicopathological analysis.

The mean age was 57 ± 16 years (range 25–91), and 53 (17 %) patients were less than 40 years by the time of diagnosis. There was a low rate of formal learning: 42 % had incomplete primary/elementary education, and 45 % were illiterate. Poor quality of medical records precluded analysis of sexual behavior. Cigarette smoking was reported by 137 (44 %) patients and alcohol consumption by 127 (40 %). Forty-one patients (13 %) had phimosis, and 29 (9 %) reported previous circumcision. These characteristics of patients enrolled in the study are summarized in Table 1. The mean interval between the onset of symptoms and seek for medical assistance was 9.8 ± 11 months.

Table 1 Sociodemographic and clinical history of 314 patients with penile carcinoma assisted in Hospital Aristides Maltez, Salvador, Brazil, from 1997 to 2007
Full size table

Tumors had a mean size of 4.2 cm (±2.0 cm) ranging from 0.5 to 13.0 cm. The glans was the most common primary site of involvement (250 cases or 80 %). Most cases (183 or 58 %) showed intermediate histological grade, and 210 (67 %) had invasion deeper than 5 mm. There was a decreasing frequency of involvement from corpus spongiosum (243 or 77 %), corpus cavernosum (121 or 39 %) to distal urethra (47 or 15 %). Evaluation of tumor front revealed that in 168 (54 %), the growth pattern was infiltrative. Nodal and distant metastasis were present in 98 (31 %) and 7 (2 %) of all patients, respectively. Cytological features suggestive of HPV infection were observed in 239 (76 %) of all cases. Based on WHO classification, the frequency of squamous cell carcinomas subtypes was 241 (77 %) for usual, 26 (8 %) for warty, 11 (4 %) basaloid, 4 (1 %) papillary, 2 (1 %) verruciform, 1 (0.3 %) sarcomatoid, and 29 (9 %) of mixed subtypes. Morphologic features of this series are detailed in Table 2.

Table 2 Histopathological and staging features of 314 patients with penile carcinoma assisted in Hospital Aristides Maltez, Salvador, Brazil, from 1997 to 2007
Full size table

Clinical follow-up detected enlargement of inguinal lymph nodes in 109 patients (35 %): 98 cases (31 %) had clinical suspicious of metastatic disease and 77 (25 %) were submitted to lymphadenectomy in the institution. Metastases were confirmed in 66 cases (86 %), while in 11 cases (14 %), removed lymph nodes showed only reactive features.

Twenty-seven patients underwent chemotherapy, and 43 were treated with radiotherapy. Forty patients experienced local recurrence. The mean follow-up for the 314 patients was 22.4 ± 26.5 months. A total of 42 cases (13 %) were lost during follow-up. At the end of the study, 66 patients (21 %) were alive with persistent disease, while 42 (13 %) had died due to penile carcinoma. Clinical follow-up of 38 patients with more than 5 years of follow-up gave the following results: alive free of disease (n = 28), alive with disease (n = 5), dead in consequence of the disease or complications of treatment (n = 4), and dead due to other cases (n = 1).

We were interested on how disease’s presentation and progression differed between younger (<40 years old) and older (≥40 years old) patients. Younger patients had higher frequency of some morphological features of worse prognosis such as infiltrative growth pattern (70 vs 50 %) and perineural invasion (26 vs 13 %) (p < 0.05). Younger patients experienced disease recurrence in 45 % of all cases with only 25 % of older patients showing recurrent disease at follow-up (p < 0.05). The frequency of death due to penile carcinoma was 19 % for patients <40 years old, 11 % for the age group between 40 and 60 years, and 13 % for older (>60 years old patients) (p < 0.05). Comparison of survival curves between patients less than 40 years and the age group between 40 and 60 years is shown in Fig. 1 (Mantel–Cox: 0.025). For the same age groups, the frequency of lymph node metastasis was 49, 34, and 21 %, respectively.

Fig. 1
figure 1

Survival curves of patients from the <40 (blue line) and 40–60 (green line) year age groups

Full size image

Survival curves showed no statistical difference between different carcinoma subtypes. At the 60th month of follow-up, disease-specific survival rate was 86 % for the usual subtype, 88 % for warty, 88 % for mixed subtypes, and 66 % for basaloid.

Discussion

A previous study, based on data from the same reference center for treatment of penile carcinoma in the state of Bahia, described 811 cases (803 cases of squamous cell carcinoma) from the period of 1952 to 1983 [9]. Importantly, 170 patients (21 %) were 40-year-old or younger, a frequency which is similar to the described herein (17 %). Of all patients, 20 % had already inguinal lymph node enlargement at the time of diagnosis. In this previous report, 65 % of affected patients had concurrent phimosis and none reported circumcision in childhood (125 or 15 % reported the procedure during adolescence and adulthood). In the present study, 41 (13 %) patients had phimosis and 29 (9 %) reported previous circumcision. Among the 811 cases from 1952 to 1983, the intervals between onset of symptoms and seeking medical assistance were less than 3 months, between 3 and 9 months, and longer than 9 months were 16, 43, and 41 %, respectively. In the present series, the mean interval between disease onset and medical assistance was 9.8 ± 11 months. No follow-up was reported in that previous study.

Despite economic growth, social development during the last decades in Brazil, particularly in Northeast region of Brazil and state of Bahia, the sociodemographic characteristics of patients with penile carcinoma seem to be unchanged. The illiteracy rate in Brazil dropped from 51 % in 1950 to 10 % in 2010 [10]; however, 45 % of all patients in this study were illiterate, and other 45 % had incomplete elementary education. Northeastern Brazil also experienced a sixfold rise in gross domestic product per capita from 1950 to 2010 [11]. In contrast, the present study covering the period from 1997 to 2007, in addition to the previous one covering the period from 1952 to 1983, did not detect a trend for reduction in diagnosis of new cases of penile cancer in this reference center. Poor formal education is a well-known factor associated with penile cancer, and recent studies from different settings show that the sociodemographic characteristics of affected patients seem to persist [2, 12] albeit the incidence of disease decreases in developed countries [1–4].

In the present series, the presence of cytological features suggestive of HPV infection had no impact on prognosis. Some previous studies are discordant on the possible impact of HPV infection on prognosis [13–15]. Survival curves also did not differ between different tumor subtypes.

The mean age in this case series, 57 years, is comparable with most series in the literature. However, the reported series from Brazil show a higher frequency of younger individuals (<45 years at the moment of diagnosis) [5, 16, 17]. The present series showed that 17 % of patients were less than 40 years at the time of diagnosis, and cancer-specific death rate in this age group was 19 % (in contrast to 11 and 13 % in the 41–60 and >60 age groups). Squamous cell carcinomas in younger patients were more likely to exhibit infiltrative growth pattern, perineural invasion, and recurrence. Distinct age groups did not differ on the frequency of specific subtypes of squamous cell carcinomas. The current literature and proposed prognostic indexes do not include age at cancer diagnosis as important information of prognostic value [18–21]. Recently, a multicenter study reported that zoophilia (sex with animals) was an independent factor for penile cancer, even though it did not impact on prognosis [22]. Although it is believed that zoophilia is common in Brazil, our medical records had no information on this behavior and precluded the evaluation on how patients from distinct age groups differed on this issue.

Conclusion

Regardless of squamous cell carcinoma subtypes, penile cancer in Northeastern Brazil seems to show more aggressive features and behavior when presented at younger age. If confirmed in other large series from endemic areas of penile cancer, younger age at diagnosis should be considered a prognostic factor in penile cancer.