THE INCIDENCE OF penis cancer varies widely in different parts of the world. High incidence rates are found in Brazil, Puerto Rico, Jamaica, and certain areas of Africa, where penis cancer is one of the most common types of cancer in males.1 In contrast, the populations of most industrialized countries are at low risk for this cancer.2,3 In Denmark, the incidence of penis cancer decreased significantly from 1.15 per 100,000 person-years (world standardized) in 1943–1947 to 0.82 per 100,000 person-years in 1988–1990.4 Boys circumcised soon after birth are effectively protected against penis cancer, but the beneficial effect diminishes with increased age at circumcision.5–7 Risk factors for penis cancer, other than those resulting from restricted preputial motility, include: venereal infections, measures of sexual promiscuity, poor genital hygiene, and smoking.1,2,8–11 Evidence of an association between syphilis and penis cancer has been reported in several studies.12–16 However, with the detection of certain types of human papillomaviruses (HPV) as a more likely causal agent,17,18 recent large-scale studies assessing the possible importance of prior syphilis infection have been lacking. The present investigation aims to provide a sound epidemiologic background on which to answer the question of whether syphilis should be maintained as a possible factor in the etiology of penis cancer. In our attempt to accomplish this goal, we used national data on incident penis cancers, diagnosed over approximately half a century, and linked these patients with the files of the Danish Syphilis Registry.
Materials and Methods
Since 1943, all incident cancers reported to the Danish Cancer Registry have been classified according to a modified version of the seventh revision of the International Classification of Diseases (ICD-7), and from 1978 by automatic conversion from the International Classification of Diseases for Oncology (ICD-O).19 All cancer notifications under the ICD-7 code 179 were retrieved. Cancers originating in the scrotum or epididymis were excluded and every penis cancer diagnosed before 1978 was manually recoded according to the ICD-O. For some penis cancers, notably from the early part of the study period, histologic information was not available. Such cancers were considered epidermoid, because this is by far the most common type of penis cancer.2,20
All patients diagnosed with penis cancer during the period 1943–1990, and a frequency-matched (by 5-year calendar periods) sample of men with colon or stomach cancer (800 cases each) were linked to the Danish Syphilis Registry at Statens Serum Institut. A search was conducted for recorded information on prior syphilis infection in these patients. The syphilis registry contains records on most cases of syphilis diagnosed in Denmark residents since 1920.21 Serologic screening for syphilis has been widely applied to both in- and outpatients in Denmark since 1910. An average of 267,000 samples have been screened for syphilis each year during the period 1910–1990 in the Danish population, which increased in size from approximately 3 to 5 million inhabitants during the same period. Almost every patient with penis cancer and every patient with either colon or stomach cancer is believed to have been tested for syphilis at least once during adulthood. Consequently, the vast majority of cases of syphilis in these patients have been diagnosed and reported to the registry. Multivariate logistic regression was used to estimate the relative risk, as measured by the odds ratio (OR) of developing penis cancer after a diagnosis of syphilis. In the analyses, we controlled for differences in age (in 10-year age-groups), calendar period (in 5-year periods), place of residence (Copenhagen, including suburbs versus the rest of Denmark) and marital status at the time of cancer diagnosis (never married versus ever married).22
We identified 1,562 cases of invasive penis cancer diagnosed from 1943–1990. In the early period, 1943–1962, 76% (n = 430) had information about the exact histologic type, whereas 96% (n = 467) had such information from 1978–1990. Throughout the study period, most cancer notifications did not contain details about the exact penile localization. The anatomic and histologic distributions of 484 cancers diagnosed during the recent period 1978–1990 are presented in Table 1. Among those with anatomic details available, 59% were localized on the penile glans, 37% on the prepuce, and 3% on the body of the penis.
Of the 1,562 penis cancers, we excluded 39 of nonepidermoid histology: basal cell carcinoma (n = 20), adenocarcinoma (n = 2), transitional cell carcinoma (n = 1), melanoma (n = 9), sarcoma (n = 2), and lymphoma (n = 5). Consequently, our material comprised 1,523 patients with epidermoid penis cancer, including 207 (14%) without specified histology. Median age at diagnosis was 67 years, range 22 to 95 years.
Linkage by means of full name and date of birth (verified identity) resulted in the retrieval of 79 (5.2%) files of the 1,523 patients with penis cancer in the syphilis registry. Syphilis registrations potentially concerning 17 (1.1%) other patients with penis cancer were also identified, but these registrations contained information only on the first letter of the family name and on the date of birth (possible identity). Forty-eight penis-cancer patients (3.2%) had either a record of definite clinical signs of syphilis, or a clinical diagnosis of syphilis without a statement of clinical signs, but with supportive serological findings before their penis cancer diagnosis (verified syphilis). Another 22 patients (1.4%) had a clinical diagnosis of syphilis without a statement of clinical signs and without supportive serological findings (possible syphilis), and 26 (1.7%) were registered because of nontreponemal serological reactivities without anamnestic or clinical signs of syphilis (no syphilis).
Among 800 control patients with colon cancer, 30 patients (3.8%) with verified identity and 17 patients (2.1%) with possible identity were retrieved in the files of the syphilis registry. Nineteen patients (2.4%) had verified syphilis, 15 patients (1.9%) had possible syphilis, and 13 patients (1.6%) had no syphilis. The corresponding figures among 800 control patients with stomach cancer were 20 patients (2.5%) with verified syphilis before the stomach cancer, 21 patients (2.6%) with possible syphilis, and 10 patients (1.3%) with no syphilis among 31 patients (3.9%) with verified identity and 20 patients (2.5%) with possible identity.
The age distribution among cancer patients with past syphilis infection was not markedly different from that of the entire patient populations. The median ages at the time of cancer diagnosis among those patients with penis cancer, colon cancer or stomach cancer who had either a possible or a definite history of syphilis were 69, 71 and 69 years, respectively, as opposed to 67, 70 and 71 years, respectively, in the total patient populations.
Multivariate logistic regression was performed for the 1,516 penis cancer patients and 1,529 control patients (774 colon cancer patients and 755 stomach cancer patients) who had information on both syphilis status and on marital status at the time of cancer diagnosis. With the use of stringent criteria for prior syphilis infection (i.e., the presence of both verified identity and verified syphilis: 43 patients with penis cancer, 13 with colon cancer, and 15 with stomach cancer), the risk of having had syphilis was moderately higher, but not significantly so, among patients with penis cancer than among controls [ORstringent = 1.5 (95% CI:0.9–2.4)] after adjustment for potential confounding by age, calendar period, place of residence and marital status (Table 2). A similar lack of significant association between syphilis and the risk of penis cancer was seen when less stringent criteria for prior syphilis infection were applied (i.e., the presence of either possible or verified syphilis and either possible or verified identity: 70 patients with penis cancer, 34 with colon cancer and 38 with stomach cancer), ORnot stringent = 1.0 (0.7–1.5). This applied also when colon cancer patients, ORstringent = 1.8 (0.9–3.3) and ORnot stringent = 1.2 (0.8–1.8), and stomach cancer patients, ORstringent = 1.3 (0.7–2.3) and ORnot stringent = 0.9 (0.6–1.4) were considered separately as controls.
We subsequently divided the patients into three groups by the year of cancer diagnosis (1943–1962, 1963–1977, and 1978–1990). Except for the period 1963–1977, in which strictly defined anamnestic syphilis was close to significantly more common in patients with penis cancer than controls (ORstringent = 2.2 (0.98–4.8)), no consistent association was present (Table 2).
Almost a century ago, second to phimosis, syphilis was considered the most important risk factor for penis cancer.12 A general carcinogenic effect of syphilis, in which the individual cancer site depended on the location of the primary chancre, was proposed.13 In 1947, a case-control study revealed a fourfold risk of penis cancer in patients with a history of syphilis,14 and later studies supported this finding.15,16 In one study, whose patient material forms part of the present investigation, 9% of 511 patients diagnosed with penis cancer in Denmark during the period 1942–62 were considered syphilis-positive compared with only 4% in an age- and residence-matched control group.16 However, the assessment of syphilis status in penis cancer patients and their controls in that study was unequal. When we restricted our analysis to almost the same period, i.e., 1943–62, we found no support for an association between syphilis and penis cancer (odds ratios of 1.1 using stringent and 0.8 using less stringent criteria). It has also been suggested that syphilis might hasten malignant degeneration and, as a result, cancer may occur at younger ages in patients with prior syphilis.23,24 We found no support for this view either, because the age distributions in those with and without possible or definite prior syphilis were almost identical.
One possible explanation of the previously reported strong association between syphilis and the risk of penis cancer is that early penis cancer in some patients may have been misdiagnosed and initially treated as syphilis, and that some of these incorrect syphilis diagnoses may have persisted even after diagnosis of penis cancer. However, clinical misdiagnosing of syphilis might not entirely explain the association between this venereal disease and penis cancer. Current evidence indicates that certain types of human papillomaviruses play a causal role in some instances of penis cancer.17,18 Because both Treponema pallidum and the relevant HPVs are transmitted sexually, it is reasonable to believe that lacking appreciation and measurement in early studies of the most likely causal venereal factors, i.e., HPVs, have yielded spurious associations with other similarly transmitted agents that are not genuinely involved in the carcinogenesis. This confounding is likely to be partially responsible for the very longstanding, and most likely incorrect, belief that syphilis is a cause of penis cancer. In combination with the findings in two recent case-control studies,8,9 our finding that patients with penis cancer did not have a higher rate of positive syphilis history than other cancer patients suggests that syphilis should be removed from the list of possible causes for penis cancer.
1. Muir CS, Nectoux J. Epidemiology of cancer of the testis and penis. Natl Cancer Inst Monogr No. 53, 1979:157–164.
2. Hall NEL, Schottenfeld D. Penis. In: Schottenfeld D, Fraumeni JF Jr, eds. Cancer Epidemiology and Prevention. Philadelphia, PA: WB Saunders, 1982; 958–967.
3. Parkin DM, Muir CS, Whelan SL, Gao Y-T, Ferlay J, Powell J, eds. Cancer Incidence in Five Continents. Vol. VI. Lyon, France: International Agency for Research on Cancer, 1992. IARC Scientific Publications No. 120.
4. Frisch M, Friis S, Kjaer SK, Melbye M. Falling incidence of penis cancer in an uncircumcised population (Denmark 1943–1990). Br Med J 1995; 311:1471.
5. Dodge OG, Linsell CA. Carcinoma of the penis in Uganda and Kenya Africans. Cancer 1963; 16:1255–1263.
6. Wolbarst AL. Circumcision and penis cancer. Lancet 1932; 1:150–153.
7. Schmautz R, Jain DK. Geographical variation of carcinoma of the penis in Uganda. Br J Cancer 1971; 25:25–32.
8. Brinton LA, Jun-Yao L, Shou-De R, et al. Risk factors for penis cancer: results from a case-control study in China. Int J Cancer 1991; 47:504–509.
9. Maden C, Sherman KJ, Beckmann AM, et al. History of circumcision, medical conditions, and sexual activity and risk of penis cancer. J Natl Cancer Inst 1993; 85:19–24.
10. Daling JR, Sherman KJ, Hislop TG, et al. Cigarette smoking and the risk of anogenital cancer. Am J Epidemiol 1992; 135:180–189.
11. Hellberg D, Valentin J, Eklund T, Nilsson S. penis cancer: Is there an epidemiological role for smoking and sexual behavior? Br Med J 1987; 295:1306–1308.
12. Barney JD. Epithelioma of the penis. An analysis of one hundred cases. Ann Surg 1907; 46:890–914.
13. Touraine A. Cancer et syphilis. Rapports généraux. Rev Med 1933; 50:691–717.
14. Schrek R, Lenowitz H. Etiologic factors in carcinoma of the penis. Cancer Res 1947; 7:180–187.
15. Ekström T, Edsmyr R. Cancer of the penis. A clinical study of 229 cases. Acta Chir Scand 1958; 115:25–45.
16. Jensen MS. Cancer of the penis in Denmark 1942 to 1962 (511 cases). Dan Med Bull 1977; 24:66–72.
17. Krebs HB, Schneider V. Human papillomavirus-associated lesions of the penis: colposcopy, cytology, and histology. Obstet Gynecol 1987; 70:299–304.
18. Malek RS, Goellner JR, Smith TF, Espy MJ, Cupp MR. Human papillomavirus infection and intraepithelial, in situ, and invasive carcinoma of penis. Urology 1993; 42:159–170.
19. Storm HH. Appendix 3(a): The Danish Cancer Registry, a self-reporting national cancer registration system with elements of active data collection. In: Jensen OM, Parkin DM, MacLennan R, Muir CS, Skeet RG, eds. Cancer Registration: Principles and Methods. Lyon, France: International Agency for Research on Cancer, 1991:220–36. IARC Scientific Publications No. 95.
20. Lucia MS, Miller GJ. Histopathology of malignant lesions of the penis. Urol Clin North Am 1992; 19:227–246.
21. Reyn A. The syphilis index at Statens Serum institut, Copenhagen. Dan Med Bull 1954; 1:211–213.
22. SAS Institute. The PROBIT procedure. In: SAS/STAT users guide, version 6. Vol 2, 4th ed. Cary, NC: SAS Institute, 1989:1325–1350.
23. Belote GH, Arbor A. The association of cancer and syphilis as determined by positive serology. Am J Syph 1931; 15:372–375.
24. Michalek AM, Mahoney MC, McLaughlin CC, Murphy D, Metzger BB. Historical and contemporary correlates of syphilis and cancer. Int J Epidemiol 1994; 23:381–385.