Three hundred twenty-one (58.7%) patients underwent surgery, while 179 (32.7%) did not have surgery. For 57 (8.6%) participants, it is unknown whether surgery was performed. There was a wide variety of surgical techniques used for treatment. Most patients underwent excisional biopsy (24.5% of all patients), followed by simple or partial removal of the primary site tumor (15.0%), wide local tumor excision, NOS (8.6%), and total surgical removal of the tumor primary site including penectomy (5.3%). Approximately 92 patients received surgery that included lymph node dissection (29% of all patients who received surgery); positive lymph nodes were found in 20 of those cases. The histologies most often associated with positive lymph nodes were melanomas, transitional cell papillomas and carcinomas, and unspecified epithelial tumors; the three most overrepresented histologies regarding positive lymph nodes as a proportion of frequency were transitional cell papillomas and carcinomas (18.5%), nevi and melanomas (14.1%), and soft tissue tumors and sarcomas (11.1%). Full results for histologies and rates at which cancer was found in lymph nodes for each histology are presented in Table 3 .

Patients rarely received adjuvant therapies. Forty-three (7.9%) patients received systemic therapy and 43 received radiation. Thirteen (2.4%) underwent both systemic therapy and radiation. Of the 21 patients (3.8%) that received systemic therapy in addition to surgery, 18 (3.3%) did so after surgery, two (0.37%) did so before surgery, and one (0.18%) did so both before and after surgery. Of the 22 patients (4.0%) that received radiation in addition to surgery, 19 (3.5%) did so after surgery, one (0.18%) did so prior to surgery, and two (0.37%) received both radiation and surgery, but the sequence is unknown. Multivariate analysis revealed a higher concordance between disease Stage and the receipt of adjuvant chemotherapy than disease grade and adjuvant chemotherapy ( Supplementary Table S1 ).

Twenty-three patients received only radiation as treatment, receiving no surgery or chemotherapy. None of the patients in this category had metastases or advanced disease Stage. The distribution of histologies for patients treated solely with radiation is presented as Figure 1 below. The most common treatment modalities and combinations thereof are presented in Figure 2 ; lymphomas were evaluated separately.

Our analysis emphasizes the important role of surgical intervention in treatment. Patients who underwent surgery had a better OS across the entire period observed; however, this survival advantage appears to be greatest in the first 5 years after diagnosis. Although surgery confers a survival benefit, a significant proportion of patients did not undergo surgery. The underutilization of surgical intervention may be explained by the unresectability of the tumor or late disease presentation. However, if this underutilization is due to a lack of guideline-based approach in evaluating these tumors, our findings suggest that it may be advantageous to expand surgical candidacy when feasible. The potential late disease presentation of this uncommon malignancy also emphasizes the necessity of more adequate screening guidelines for this understudied patient population, as surgery is an optimal treatment modality.

Systemic therapy or radiation, used independently or in conjunction with surgery, was utilized in a small proportion of patients. Since there is a non-significant difference in survival during the observed period between patients who did and did not receive radiation or systemic therapy, it cannot be ascertained whether these two treatment pathways are beneficial. Alternatively, it may be the case that radiation or systemic therapy is used in refractory cases where surgery is not advised or is unsuccessful.

This analysis aims to highlight the survival characteristics with respect to a particular histological grouping. This information may be prognostically important for patients with one of these particular tumor subtypes.

Of the most prevalent histological groupings ( Table 2 ), blood vessel tumors, which overwhelmingly consist of KS, had the greatest median OS and 5-year survival. This is in contrast to the previous study by Bhambhvani et al., which found KS to have a significantly worse prognostic outcome when compared to other non-SCC cancers of the penis; it should be noted, however, that this previous work stated the possibility that their results may not represent the present therapeutic landscape (14). As they acknowledged, the severity of KS corresponds with the patient’s level of immunosuppression; therefore, patients with HIV-induced immunosuppression tend to have more severe disease (28, 29). Given that highly active antiretroviral therapy (HAART) has dramatically reduced the severity of immunosuppression caused by HIV, fewer patients diagnosed with HIV are developing KS, an AIDS-defining illness. As a result of these treatments, the KS patient pool consists of those with better immune status, offering this population improved prognostic outcomes compared to other cancers (30, 31). Given that their study includes patients before 2000, when HAART was not widely available, their results do not reflect the current prognostic trend for patients with KS; our present study, thus, not only validates the previous study’s concern regarding HAART changing the prevalence and prognosis of tumors such as KS of the penis but, furthermore, describes and quantifies the current landscape of disease survival, prognosis, and Stage at the time of diagnosis in a time period with greater access to modern antiretroviral treatments that more accurately reflects the present-day clinical context.

Investigating the other common tumor types, we found basal cell neoplasms and melanomas to be two of the most populated groups with comparable median OS and 5-year survival rates of just over 40%, a significant decrease from the average 5-year survival of approximately 80% for SSC of the penis (32). The presence of these UV-dependent tumors on classically unexposed organs such as the penis highlights the importance of thorough dermatologic evaluations by physicians, including examinations of the genitalia.

Unspecified neoplasms and epithelial neoplasms, NOS had the lowest median OS and 5-year survival amongst the most prevalent subtypes. While both of these groups are classified as single categories in the SEER database, it is essential to note that they may represent a wide range of tumor types and may have different clinical characteristics and prognoses. However, epithelial neoplasms, NOS consists of carcinomas, not of squamous cell origin, and unspecified neoplasms consist of any malignant tumor that was not specified and, therefore, put in this category. These classifications may be the result of a lack of identifying microscopic features that could be obtained from the underlying histological sample; a brief analysis investigating grade suggested difficulties in identification were not related to these tumors being relatively undifferentiated (33, 34). This hinders our ability to interpret the data meaningfully and may be cause for physicians to specifically classify neoplasms, when possible, especially when rare.

Most of the patients in our cohort had localized tumors (for lymphomas, localized was defined as Stage I, IE, or IS; for non-lymphomatous cancers, localized was defined as existing solely in the tunica albuginea, an invasive tumor limited to subepithelial connective tissue, but not involving corpus spongiosum or cavernosum, or, if a skin cancer, limited to skin of the penis, prepuce, and/or the glans), which afforded them an increased survival benefit compared to those with tumors at worse stages. The rest of the patients had regional, distant, or unknown/unstaged tumors. Regional and unstaged tumors had similar prognostic indicators and tumors with distant metastases had the worst prognosis, with none of the patients living beyond 5 years. These survival characteristics are generally intuitive and consistent with literature about prognosis with respect to tumor Stage (35–37).

The tumor grade is unknown for 68% of patients; however, there were patients who fell into each category of tumor grade. Of the tumors characterized with grades I–IV, grade I tumors had the best prognosis, followed by grade II tumors. Grades III and IV afforded the worst prognosis. This is consistent with literature that suggests that decreased tumor differentiation is associated with worse outcomes (36, 38).

The results indicate that age was a determining factor in regard to the chance of survival during the observation period. Patients below 70 years had superior survival rates during the entire observation period.

Since Black patients had lower median OS and 5-year survival in comparison to White patients, it was hypothesized that Black patients were diagnosed at a later age. However, 46.7% of Black patients were diagnosed at 70 years or older compared to 51.6% for White patients. Multivariate analysis revealed statistically significant independent contributions of race, stage, marital status, and stage, as well as surgery and positive regional LN on OS ( Supplementary Table S2 ).

This suggests that there are other factors responsible for the worse prognosis for Black patients; although there were trends among survival differences between White and Black patients, such as black patients demonstrating worse survival after receiving beam radiation, and worse survival upon increased disease Stage, these variables were unable to fully explain the survival gap. Further studies are needed to analyze differences in treatment with regard to race for patients with non-SCC of the penis.

The cohort is nearly split in half with respect to marriage status at diagnosis, which is reflective of the U.S. population where 53% of adults are married (39). Given the survival advantage over the 10-year period for married patients compared to unmarried patients, marriage appears to be a protective factor with respect to the survival of a penile non-SCC diagnosis, although this has been observed with penile cancer overall, to our knowledge this represents the first observation of this trend in non-SCC of the penis histologies as well (40). This trend may be attributable to the positive aspects of the marriage dyad such as better health behaviors and social support that are associated with a lower allostatic load (41). In this way, it is consistent with sociological literature that suggests that marriage affords an overall health advantage to men in the form of increased life expectancy and active life expectancy (42, 43). This is particularly relevant given that our entire cohort consists of men.

There is little evidence to support the notion that marriage can decrease the likelihood of being diagnosed with cancer. There is, however, evidence to suggest that social support in the form of marriage can have a significant impact on cancer detection and obtaining adequate treatment (44). This evidence presents a limitation of the argument that marriage itself confers a survival benefit through a decrease in the patient allostatic load, suggesting instead it may confer a benefit through early detection and treatment acquisition and adherence. Due to the rarity of penile non-SCC, it is unknown whether this limitation applies to the results of this cohort.

This study has limitations to note. Due to the rarity of this disease, the relatively small sample size frequently limited the power of statistical analysis. Another limitation involves the potential inclusion of cancers of squamous cell origin, which were intended to be excluded from this study. As some histologies were unable to be pathologically identified, squamous cell tumors may have been inadvertently included in the analysis. The possibility that such tumors may be of very high grade and, thus, dedifferentiation, which may prevent pathological confirmation, was considered; however, no trend between tumor grade and the population of unidentifiable histologies was qualitatively observed.

This study also faced challenges regarding treatment paradigms. Due to the lack of specific data within the SEER database, it cannot be distinguished whether radiotherapy was performed at the primary site or nodal site(s) nor are precise chemotherapeutic regimens recorded. The inability to analyze these treatment nuances limits what we can understand about treatment effectiveness.

The heterogeneity of histological types of non-SCC penile cancer also makes it challenging to develop and analyze specific treatment paradigms. For example, while surgery was a beneficial treatment modality as a whole, this is likely due to the cohort mostly consisting of epithelial neoplasms; other tumor types may not benefit from surgery. As such, while our results may give a broad overview, they may not apply to all histological subtypes.

Age is a potential confounding factor that could have influenced the results. While age was considered in our multivariate analysis, it can influence the rest of the variables not selected for the multivariable analysis. Age can affect the patients’ overall health, their ability to tolerate treatment, the presence of comorbidities, and their access to and utilization of healthcare resources.

Finally, it should be noted that the SEER database encompasses data from American patients only. While the SEER database provides a comprehensive collection of cancer statistics, the demographic makeup of the U.S. differs considerably from other regions in the world. This is especially relevant considering the global importance and varied geographical distribution of penile cancer.