The etiology of penile cancer is multifactorial [12, 25, 26]. In a review of scientific publications on penile cancers between 1996 and 2000, phimosis (inability to fully retract the foreskin), chronic inflammatory conditions (i.e., balanoposthitis, lichen sclerosus et atrophicus), and treatment with psoralen and ultraviolet A photochemotherapy (PUVA) were identified as strong risk factors associated with an odds ratio (OR) greater than 10 [25]. Other risk factors, such as poor hygiene, history of smoking, multiple sexual partners, and history of genital warts, have been identified.

Phimosis is present in up to one half of patients with penile cancer. Although preventive circumcision has been suggested in patients with phimosis living in areas with high incidence of penile cancer [27], adult circumcision failed to show a protective effect since men never circumcised or circumcised after the neonatal period had, respectively, a 3.2 and 3.0 times higher risk compared to men circumcised as neonates [26]. In a population-based case–control study conducted in western Washington State between 1979 and 1998, men not circumcised during childhood resulted at increased risk of invasive (OR = 2.3; 95 % CI: 1.3, 4.1) but not in situ (OR = 1.1; 95 % CI: 0.6, 1.8) penile cancer [28]. Approximately 35 % of men with penile cancer who had not been circumcised in childhood and 7.6 % of controls reported a history of phimosis (OR = 7.4; 95 % CI: 3.7, 15.0). Phimosis was found to be strongly associated with development of invasive penile cancer (OR = 11.4; 95 % CI: 5.0, 25.9) in men not circumcised in childhood. The risk of invasive penile cancer in men with no phimosis and not having been circumcised in childhood was not elevated (OR = 0.5; 95 % CI: 0.1, 2.5). On the contrary, neonatal circumcision is associated with a threefold decreased risk of invasive penile cancer since it protects against phimosis, poor penile hygiene, and retention of desquamated epidermal cells and urinary products resulting in chronic inflammation of the glans and prepuce. Nonetheless, 20 % of penile cancer patients had been circumcised neonatally [25, 28]. Interestingly, Schoen and colleagues determined that the level of protective effect of neonatal circumcision for CIS is not as high as that for invasive penile cancer [29]. Of 89 men with invasive penile cancer whose circumcision status was known, 2 (2.3 %) had been circumcised as newborns and 87 (97.7 %) were not circumcised. Of 118 men with CIS whose circumcision status was known, 16 (15.7 %) had been circumcised as newborns.

Lichen sclerosus (LS), an unusual chronic mucocutaneous condition of the penis preferentially involving the foreskin but also the glans, coronal sulcus, and urethra, was found to cause phimosis and to be associated with the usual, verrucous, papillary, and pseudohyperplastic subtypes of invasive penile carcinoma in 30–50 % of patients [12, 30]. In a retrospective study from Italy, Nasca and colleagues reported penile invasive SCC or premalignant lesions in 9 of 86 men (9.3 %) with a history of penile LS after a mean lag time of 18 (range, 10–34) years. The transition from LS to frank neoplastic foci was histologically evident in all cases of SCC [31]. Powell and coworkers retrospectively found histological or clinical evidence of LS in 11 of 20 patients with SCC of the penis [32]. SCC was well differentiated in seven of the eight cases with histologic evidence of LS in the excision specimen. Only 3 SCCs were well differentiated among the 12 cases with no evidence of LS, although a history of LS sometimes preceding the SCC by 10 years was identified in the records of 7 of these 12 patients. Of the ten deceased patients in this study, seven died from metastatic disease. The authors concluded that the association between SCC of the penis and LS was present even in the patients in whom the clinical presentation of LS or the need for circumcision preceded the SCC by many years. A large retrospective study from Paraguay assessed the anatomic distribution and prevalence of LS in patients with SCC of the penis [30]. The penectomy and circumcision specimens from 207 patients with carcinoma and giant condylomas were examined, and 68 (33 %) patients were found with evidence of LS; however, the true association was felt to be likely underestimated. The preferential anatomic site of LS was the foreskin, although involvement was noted at the level of the glans and coronal sulcus, including the urethra. The gross and microscopic findings suggested that LS may represent preneoplastic condition for at least some types of penile cancers, in particular those not related to HPV. Evidence of LS was found in 28 % of 155 patients with penile carcinoma studied prospectively by Pietrzak and colleagues [33]. Prowse et al. investigated the role of HPV infection and expression of the tumor suppressor protein p16^INK4A in the pathogenesis of penile cancer [34]. In 26 penile SCCs and 20 independent penile LS, HPV DNA was found in 54 % of penile SCCs and 33 % of penile LS patients. Strong immunostaining for p16^INK4A correlated with HPV 16/18 infection in both penile LS and penile SCC. Penile SCC margins were also associated with penile LS in 13 of 26 lesions. HPV was detected in 7 of the 13 SCC cases associated with LS and in 6 of the 11 SCC lesions not involving LS. Barbagli and coworkers evaluated the presence of premalignant or malignant lesions in 130 patients with LS involving the male genitalia. Eleven (8.4 %) men with genital LS showed premalignant or malignant histopathological features including 7 (64 %) with SCC. Based on all these studies, it has been estimated that the risk of malignant transformation of penile LS is similar to vulval lichen sclerosus (4–8 %) [35].

Stern reported on the risk of genital tumors in 892 patient with psoriasis treated with oral methoxsalen (8-methoxypsoralen) and PUVA [36]. In his 12.3-year prospective study, 30 genital neoplasms in 14 (1.6 %) patients were identified between 1976 and 1989. Compared to expected morbidity (based on population incidence data), patients treated with PUVA had a standard morbidity ratio of 95.7 (95 % CI: 43.8, 181.8) for invasive SCC of the penis and scrotum and 58.8 (95 % CI: 26.9, 111.7) for invasive and in situ penile tumors. When compared to the general population or patients exposed to low levels of PUVA, the incidence of invasive SCC in patients exposed to high levels of PUVA was 286 times and 16.3 times higher, respectively (p < 0.001 for both comparisons). After controlling for the level of exposure to PUVA, patients exposed to high levels of ultraviolet B radiation were found to have a risk of genital tumors 4.6 times higher than that in other patients (95 % CI: 1.4, 15.1). Considering the strong dose-dependent increase in the risk of genital tumors associated with exposure to PUVA and ultraviolet B radiation, the author recommended the use of genital protection (e.g., shielding) in men exposed to PUVA or other forms of ultraviolet radiation. In a subsequent update of this experience nearly 10 years later, the development of genital cancer in ten new cohort patients confirmed the persistence of the risk associated with PUVA exposure despite the increased use of genital protection and the decreased use or the discontinuation of PUVA. Men previously unaffected by such tumors developed cancer in the decade starting nearly 15 years after initial exposure to PUVA, with an overall risk of invasive scrotal and penile SCC since enrollment 81.7 times (95 % CI: 52.1, 122.6) higher than that expected in the general population. Multivariate analyses revealed the highest genital tumor risk among men with high-dose exposure to both PUVA and topical tar/ultraviolet B, with an incidence rate ratio of 4.5 (95 % CI: 1.3, 16.1) compared with the low-dose exposure group. Another study by Perkins et al. showed a greater than 300-fold increase in genital tumors in 130 patients with psoriasis treated with PUVA [37].