Background: Human papillomavirus (HPV) is causally associated with cervical squamous cell carcinoma (SCC) and its precursor lesions. By analogy, HPV is believed to play a role in penile cancer through progression of HPV-associated penile squamous intraepithelial lesions (SIL). HPV DNA has been reported to be present in 100% of high-grade penile SIL, but the percentage of invasive or infiltrating penile SCC that was positive for HPV DNA has varied from study to study (positivity values ranging from 32% to 82%).
Purpose: To ascertain whether HPV is associated with penile cancer, we used a polymerase chain reaction (PCR)-based assay to test specimens of penile SCC for the presence of HPV DNA.
Methods: A total of 117 formalin-fixed, paraffin-embedded specimens of penile cancer from an equal number of patients who had been diagnosed either at the Memorial Sloan-Kettering Cancer Research Center in New York City between 1964 and 1992 or the Universidad Nacional de Asunción in Paraguay between 1980 and 1992 were analyzed. Specimens were examined without prior knowledge of the histology of the lesions. Methods were used that minimized sample contamination, thus avoiding false-positive results. PCR and Southern blot analyses were used to determine HPV type. The presence of HPV DNA was studied for association with the tumor properties histopathology, growth pattern, tumor grade, regional lymph node status, and anatomic location. Two-sided statistical tests were used to determine P values.
Results: HPV DNA was detected in 26 (22.2%) of 117 specimens. In 23 (88.5%) of the 26 HPV-positive specimens, HPV type 16 (only) was identified. HPV DNA was frequently associated with SCC in areas showing basaloid and/or warty changes (nine [47.4%] of 19 specimens were HPV positive; P = .0125). More highly significant was the association of virus with basaloid SCC (nine [75%] of 12 specimens were HPV positive; P = .0005). However, HPV was not found to be associated with typical SCC of the penis (five [11.1%] of 45 specimens were HPV positive). Virus DNA was more often associated with high-grade tumors (P = .0278) exhibiting aggressive growth (P = .0382) localized to the glans penis (P = .0324). Stepwise logistic regression analysis revealed that only tumor histopathology was a significant predictor of an HPV association.
Conclusions: The presence of HPV DNA was found to be significantly associated only with those penile SCC exhibiting basaloid changes. Furthermore, HPV DNA sequences tended to be associated with higher grade and more aggressive tumor localized to the glans penis. The low frequency of HPV in penile SCC implies that only a small proportion of these cancers arise from HPV-associated penile SIL.