SEARCH

SEARCH BY CITATION

Keywords:

  • relative survival;
  • incidence;
  • neoplasm survival;
  • mortality;
  • penile cancer;
  • penile carcinoma

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References

We examined trends in the incidence and mortality, and described the survival of patients with penile squamous cell carcinoma in the Netherlands between 1989 and 2006. On the basis of nationwide population-based data, 3-year moving average European age-standardized incidence and 10-year relative survival estimates were calculated. Penile squamous cell carcinomas were categorized according to stage grouping based on the TNM classification. In the 17-year study period, 2000 primary penile cancers were diagnosed in the Netherlands of which 1883 (94%) were squamous cell carcinomas. Median age at diagnosis was 68 years. The majority of patients (57%) were diagnosed with localized tumors (Stage 0 or I). The percentage of missing disease characteristics increased with increasing age. The 3-year moving average incidence rate of patients with penile squamous cell carcinoma increased significantly from 1.4 per 100,000 person-years in 1989 to 1.5 in 2006 with an estimated annual percentage of change of 1.3%. Ten-year relative survival of patients according to the different stage groups was 93% for Stage 0, 89% for Stage I, 81% for Stage II, the 9-year survival was 50% for patients with Stage III disease and a 2-year survival of 21% for patients was found for Stage IV disease. Our study shows that the incidence rate of penile squamous cell carcinoma in the Netherlands has increased slightly, especially the incidence of carcinomas in situ. Patients with Stage III and IV tumors have poor survival.

Primary penile cancer is a relatively rare neoplasm in the Western world with an age-adjusted incidence rate of around 1 per 100,000 men.1 More than 95% of these tumors are squamous cell carcinomas.2 Causative risk factors for penile cancer include phimosis,3 tobacco use4 and infection with human papillomavirus (HPV),5–7 while neonatal circumcision is suggested to have a protective effect.8 There is a worldwide geographic variation in incidence that could be caused by differences in socioeconomic status, hygiene, religious and cultural conditions.1, 9 For example, the incidence of penile cancers ranges from 0.04 per 100,000 men in Jewish populations in Israel (high incidence of circumcision) to 3–4 per 100,000 men in non-Western countries, such as Brazil, India and Southern Africa.1 Although the exact pathogenesis is largely unknown,7 inflammation may represent a critical component in tumor development or progression as many penile cancers arise at sites of infection, chronic irritation or injury.3 Prognosis is affected by tumor stage, grade and especially presence of lymph node involvement at diagnosis.10, 11

In recent years, the overall incidence of penile cancer has decreased in the United States.12, 13 However, reported incidence rates varied by race/ethnicity being higher in Hispanic and Afro-American men.1, 12–14 Trends of incidence of penile carcinoma in the Netherlands (with a large majority of Caucasian inhabitants) could therefore differ from those reported in the United States. We conducted a study of the trends in incidence and survival in the Netherlands in the period 1989–2006. Although a short overview is given of all penile malignancies, the main focus of our study is on penile squamous cell carcinomas.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References

Data on primary malignant penile cancers diagnosed in the Netherlands from 1989 to 2006 were obtained from the nationwide population-based Netherlands Cancer Registry (NCR) covering 14.8 million inhabitants in 1989 to 16.3 million inhabitants in 2006. The NCR combines the data of the 8 Dutch regional cancer registries since 1989. These 8 regional cancer registries receive lists of newly diagnosed cancer patients on a regular basis from the pathology departments of the hospitals, all participating in a nationwide network (PALGA). In addition, the medical records departments of hospitals provide lists of diagnoses of outpatients and hospitalized cancer patients. Following these notifications, trained registrars extract patient and tumor characteristics (topography, histology, stage and date of diagnosis) data from the medical records.

All penile malignancies were classified according to the International Classification of Diseases for Oncology.15 If patients had 2 or more invasive penile squamous cell carcinomas, only the first tumor was included in the survival analyses. A new primary tumor was defined as an invasive tumor >3 months after an in situ tumor diagnosis (n = 10), or a difference in subtype compared to the former penile tumor (n = 1). For patients with 2 invasive tumors (n = 3), the second tumor was excluded for all analyses of this study due to the high chance that the second tumor was a recurrence. For patients with a noninvasive tumor following a previous invasive tumor diagnosis, only the invasive tumor was included in the survival analyses. To characterize the stage of the disease all the available pathological data of the first 3 months of diagnosis were used. If (parts of) the pathological stage was unknown or missing, (parts of) the clinical stage was used to determine the stage of the penile neoplasm. To analyze changes in stage distribution the chi-square test was used. Penile squamous cell carcinomas were categorized according to stage grouping based on the 6th TNM classification, that is Stage 0 (Tis/Ta N0/Nx M0/Mx or Tx Nx Mx and registered as tumor in situ), Stage I (T1 N0/Nx M0/Mx), Stage II (T1 N1 M0/Mx, T2 N0-1 M0/Mx), Stage III (T1-2 N2 M0/Mx, T3 N0-2 M0/Mx) and Stage IV (T4, any N, any M, or any T, N3, any M, or any T, any N, M1), respectively.16 Stage was categorized as missing if the T-stage was categorized as 0 or X and the tumor was registered as invasive. Tumors were graded as well differentiated (G1), intermediately differentiated (G2), poorly differentiated (G3), undifferentiated (G4) or unknown (Gx).

Mortality data were retrieved from the website of the NCR.17 Three-year moving average age-standardized incidence and mortality rates were calculated per 100,000 person-years. Standardization was performed according to the European standard population. Evaluation of the trend in incidence and mortality rates were performed by calculating the estimated annual percentage changes (EAPC) with the joinpoint program.18

Data on vital status (available until January 1, 2008) were obtained from the hospital records and the mortality register of the Central Office for Genealogy (an institution that registers all deaths in the Netherlands via the municipal population registries). Because the cause of death is not supplied by the Central Office for Genealogy, it is unknown in the NCR database. Hence, we calculated relative survival estimates for patients with penile squamous cell carcinomas. Relative survival is an estimation of the disease-specific survival. It is calculated as the absolute survival amongst cancer patients divided by the expected survival for the general population with the same sex and age structure.19 For patients with an in situ tumor before occurrence of an invasive tumor follow-up was calculated from the incidence date of the invasive tumor. Follow-up on vital status was available for patients diagnosed since 1995 only. Hence, patients diagnosed with penile carcinoma before 1995 were excluded from the survival analyses (n = 502). Vital status is missing for 7% (n = 96) of the 1368 patients with squamous cell carcinomas diagnosed since 1995. Patient older than 95 years (n = 10) were excluded from the survival analyses resulting in the inclusion of 1262 patients with squamous cell carcinoma of the penis for the relative survival estimates, that is 67% of the total number of patients diagnoses with squamous cell carcinoma of the penis between 1989 and 2006. For evaluation of possible independent prognostic factors, a multivariate survival analysis was performed according to the Cox proportional hazards model. Because the cause of death was unknown in the NCR database, we used overall survival for this analysis.

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References

Between 1989 and 2006, 2000 primary penile cancers were diagnosed in the Netherlands in 1986 men. Squamous cell carcinomas were the most common histological subtype comprising of 94% of the reported malignancies. Other penile tumors were melanoma (2%), Pagets disease (<1%), basal cell carcinoma (<1%), adenocarcinoma (<1%) and a group of other histologies (3%). There was little variation in the histological distribution over time.

Focusing on the squamous cell carcinomas, the median age at diagnosis was 68 years. The incidence of these penile tumors, however, increased with age. The highest age-specific incidence rates of noninvasive en invasive penile carcinoma were found in men between 80 and 84 years old (1.7 per 100,000) and men of 85 years and older (17 per 100,000), respectively.

An overview of the tumor characteristics at diagnosis according to age is shown in Table 1. Nineteen percent of the squamous cell carcinomas were carcinoma in situ (n = 353) and 81% (n = 1530) were invasive. The majority of tumors were staged T1 (42%), well differentiated (29%) and diagnosed at a localized stage (58%, Stages 0 and I). The percentage of invasive tumors was lower in the age category 20–39 years (49%). The percentage of patients with missing primary tumor stage, regional lymph node involvement and disease stage tend to increase with higher age of the patients. Twenty-one percent of the patients aged 80 years had a missing primary tumor stage and 31% of these patients had missing regional nodal involvement status, while these percentages were only 6% and 11%, respectively, in patients aged 20–39 years. Of the patients with primary invasive penile tumors and known regional involvement (n = 1162), staging of the lymph nodes was clinically in 63% (n = 728) and surgically in 37% (n = 434), respectively.

Table 1. Overview of penile squamous cell carcinoma characteristics at diagnosis in the Netherlands between 1989 and 2006
inline image

The 3-year moving average overall European age-standardized incidence rate of penile squamous cell carcinoma increased from 1.4 per 100,000 person-years in 1989 to 1.5 in 2006 (Fig. 1), with EAPC of 1.3% (95% confidence interval (95% CI): 0.1–2.6%) over the whole study period. The 3-year moving average European age-standardized incidence rate of noninvasive carcinomas increased significantly from 0.1 in 1989 to 0.3 per 100,000 person-years, with an EAPC of 4.5%, (95% CI: 2.0–6.9%). The incidence rate of the invasive tumors was relatively stable and varied between the 0.9 and 1.3 per 100,000 person-years (EAPC = 0.9% 95%CI: −0.6 to 2.4%).The 3-year moving average of the mortality rate of all penile cancers is also shown in Figure 1 and varied between the 0.2 and 0.4 per 100,000 person-years in the Netherlands. The decreasing trend of the mortality was near significant (EAPC = −2.4%, 95%CI: −4.8 to 0.1%).

thumbnail image

Figure 1. Three-year moving average European age-standardized penile squamous cell carcinoma incidence rates and 3-year moving average European age-standardized of all penile cancer mortality rate per 100,000 person-years in the Netherlands.

Download figure to PowerPoint

Focusing on disease stage at diagnosis, the percentage of men with Stage 0 tumors increased from 16% in the period 1989–1994 to 20% in the period 2000–2006 (p = 0.13) (Fig. 2). There was little variation over time in the percentage of Stage I, II and III tumors. Stage IV tumors decreased from 5% in the first period to 2% in the second period and increased again to 3% in the last period (p = 0.07). The percentage of missing stage decreased significantly from 15% in 1989–1994 to 9% in 2001–2006 (p < 0.001).

thumbnail image

Figure 2. Disease stage trends by year of diagnosis of patients diagnosed with penile squamous cell carcinoma from 1989 to 2006.

Download figure to PowerPoint

Although evaluation of the different treatment modalities was not the primary focus of this article, 91% of the patients (n = 1708) were treated with surgical resection including 7% (n = 124) with adjuvant treatment (i.e., chemotherapy, radiotherapy or a combination). Relatively more patients were treated with adjuvant treatment with increase of stage (data not shown).

Patients with poorly differentiated tumors (G3) had worse 10-year survival compared to those with better differentiated tumors (66% (95% CI: 49–82%) for G3 vs. 73% (95% CI: 63–83%) for G2 and 77% (95% CI: 65–88%) for G1-tumors, respectively). As expected, patients with regional lymph node involvement (N+) had considerable worse survival compared to those without nodal involvement (N0) (Fig. 3, 38% (95% CI: 26–51%) vs. 90% (95% CI: 82–97%). Ten-year relative survival of patients with Stage 0 (noninvasive) tumors at diagnosis was 93% (95% CI: 81–103%), this was 89% for patients with Stage I tumors (95% CI: 79–98%) and 81% for patients with Stage II tumors (95% CI: 65–96%, Fig. 4). A 9-year relative survival of 50% (95% CI: 35–64%) was found for patients with Stage III tumors and the 2-year relative survival of patients with Stage IV tumors was 21% (95% CI: 10–36%). No survival estimates were possible after 9 and 2 years for patients with Stage III and IV tumors, respectively, because <10 patients were alive at that follow-up time.

thumbnail image

Figure 3. Ten-year annual relative survival of patients with penile squamous cell carcinoma according to regional lymph node involvement (N0 = no regional lymph node involvement, N+ = presence of regional lymph node involvement, Nx = missing regional lymph node status). No survival estimate was possible after 9 for patients with N+ tumors, because <10 patients were alive at that follow-up time).

Download figure to PowerPoint

thumbnail image

Figure 4. Ten-year annual relative survival of patients with penile squamous cell carcinoma according to disease stage at diagnosis (no survival estimates were possible after 9 and 2 years for patients with Stage III and IV tumors, respectively, because <10 patients were alive at that follow-up time).

Download figure to PowerPoint

The multivariate analysis showed that age, primary tumor stage, presence of regional lymph node involvement and presence of distant metastasis were all prognostic factors for overall mortality (Table 2).

Table 2. Cox multivariate regression analysis of variables associated with survival for patients diagnosed with invasive penile squamous cell carcinoma
inline image

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References

This study shows that the overall European age-standardized incidence rate of penile squamous cell carcinoma in the Netherlands has increased slightly from 1.4 to 1.5 per 100,000 person-years between 1989 and 2006 (with an EAPC of 1.3%). This finding is in contrast to recent studies from Finland and the United States, where the incidence rates decreased over time.9, 12, 13 For example, Goodman et al. reported an EAPC of −1.2% between 1973 and 2003 in the United states, and showed that the incidence declined more markedly in Afro-American men compared to Whites.13 However, they only investigated invasive penile cancers possibly explaining parts of the differences in results. Our figures indicate that the significant increase in incidence of carcinoma in situ contributes to the overall increasing incidence rates of penile squamous cell carcinomas. Rippentrop et al. reported an increasing incidence trend of carcinomas in situ in the United States between 1973 and 1998.14 A possible explanation for the higher rate of carcinomas in situ could be better awareness and less hesitation to seek treatment leading to earlier diagnosis and treatment.

In Denmark, a statistically significant decline in the overall rate of penile cancer was found between 1943 and 1990 despite a low and stable national circumcision rate of ∼1.6%.20 In that 50-year period, the proportion of Danish dwellings having a bath and routine access to clean water gradually increased over the study period.20 It is likely that such improvements in sanitary conditions have found place in the Netherlands far before the current study period (1989–2006) explaining why no decreased incidence was found in the current series. The incidence of penile carcinoma was already very low in 1989, the first year of the NCR. In an extra analysis using only the incidence data between 1970 and 1989 of the Comprehensive Cancer Centre South, located in the Southern region of the Netherlands and covering 1.0 million inhabitants in those years, the incidence rates did not change over time.

A likely explanation for the contrary results in invasive penile cancer incidence rates is not easily made. The incidence rates in the United States were evaluated using the Surveillance Epidemiology and End Results (SEER) data covering ∼26% of the US population. The incidence of penile carcinoma has shown disparities in the United States being higher in Hispanic and Afro-American men.12–14 Although these data are considered representative of the greater US populations, they might not be representative for certain cancer sites.21 In the current study, we used data of the NCR fully covering the Netherlands. Demographic differences between the populations might explain the different outcomes.

This study was consistent with previous studies that showed that ∼95% of primary penile malignancies were squamous cell carcinomas, the majority of the tumors were well and intermediate graded (29 and 35%, respectively),1, 12, 13 and survival is dependent on the stage at diagnosis with more deaths in higher staged tumors.10, 11 Our multivariate analysis indicate that more advanced primary tumors (T3 and T4), presence of regional lymph node involvement and presence of distant metastasis are in particularly poor prognosticators for worse survival. Of interest, although the median age of patients was 68 years and the age-adjusted incidence of penile carcinoma peaks after 80 years, 1% of men with penile carcinoma were younger than 30 years at diagnosis and 5% younger than 40 years. While the exact pathogenesis of penile squamous cell carcinoma is largely unknown, the HPV virus is an established etiologic factor in at least 40% of penile tumors.6 Unfortunately, HPV status was not known in the patients in our study. Previous studies have shown that some subtypes are typically HPV related.22 However, subgroup analysis of histologic subtypes was also not possible because the histological subtype was often not classified by the pathology laboratories.

Most of the penile tumors (58%) were diagnosed at a localized stage, that is Stages 0 (19%) and I (39%). Primary noninvasive carcinomas (Stage 0) do not metastasize. Hence, invasive nodal staging is not needed in this subgroup. On the other hand, patients with clinical Stage I tumors (cT1N0M0) could have occult metastases. These patients have potentially curable disease when treated adequately after diagnosis, emphasizing the need for optimal staging and treatment. The single most important prognostic factor for cancer-specific survival is presence of inguinal nodal involvement.10, 11 Surgical removal of occult nodal metastases offers a survival benefit compared to lymphadenectomy when occult disease becomes clinically apparent during close surveillance.23, 24 Unfortunately, elective lymphadenectomy is associated with significant morbidity,25 and is unnecessary in ∼80% of clinically node-negative patients,26 precluding its prophylactic use. Dynamic sentinel node biopsy is considered a more suitable staging method that only removes the lymph nodes on a direct lymphatic drainage pathway of the tumor.27 Only groins with tumor-positive sentinel nodes undergo a completion ipsilateral inguinal lymphadenectomy avoiding unnecessary morbidity in tumor-negative groins.

During the years, less patients were registered with a missing stage, indicating that staging by the clinician in the medical reports and/or the quality of registration by our trained registrars has improved. The percentage of missing disease characteristics seems to increase with increasing age. This suggests that older patients are staged less accurately and are potentially undertreated. An explanation could be that this subgroup of patients has more comorbidity interfering with (invasive) staging.

Our data show that relative survival rates in patients with Stages 0–II tumors is fairly good (81–93%) suggesting contemporary treatment is effective in the majority of these men. Patients with Stages III and IV tumors have poor survival (Fig. 4). Relative survival rates of these patients appear to be at least similar as survival of patients in the United States as reported by SEER.28 Several previous studies have shown that survival in node-positive patients is negatively influenced by the extent of nodal disease,10, 11 especially pelvic metastases (Stage IV) being a particularly poor prognosticator.29, 30 Patients in previous series were treated by surgery with or without radiotherapy suggesting this management is suboptimal in those with prognostic unfavorable features. Recently, TPF chemotherapy treatment [taxanes (T), cisplatin (P) and fluorouracil (F)] has shown potential in downstaging malignant disease in patients with unresectable or recurrent nodal penile carcinoma.31 Induction treatment followed by surgical resection in responding patients may lead to improved outcome in those with prognostic unfavorable stage groups.32

Using data from population-based data registries is not without limitations. The primary focus of the NCR is collecting data on all incident cancer cases, including stage at diagnosis and patient survival data. Although the scope and magnitude of these data make it excellent for studying rare malignancies, several variables that could be of interest were missing, for example HPV status, tobacco use, circumcision status and socioeconomic factors. Furthermore, ethnicity was not included in our study, which has recently shown to be of importance in the incidence rates in the United States.1, 12, 13 Finally, although the NCR also collects some information about treatment, full details were unavailable. Hence, we were not able to analyze the survival of patients with advanced penile tumors (Stage III or IV) stratified by treatment modality.

In conclusion, our data indicate that the incidence of penile squamous cell carcinoma has increased slightly between 1989 and 2006 in the Netherlands caused largely by the increased incidence of carcinomas in situ. Survival is dependent on disease-stage at diagnosis. Especially, patients with Stage III and IV tumors have poor survival.

References

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References