Abstract

The low incidence of penile cancer in the United States is frequently cited as a reason for not justifying the risk of neonatal circumcision as a prophylactic measure. Although uncircumcised men are uniquely at risk for this malignant neoplasm, previous approaches have used annual incidence data collected without regard to circumcision status, thus tending to underestimate the true risk to this susceptible group. In addition, the concept of lifetime risk has not been addressed. Using data from the Third National Cancer Survey and previously published circumcision prevalence figures in a life-table analysis, we estimated the lifetime risk for cancer of the penis in uncircumcised males. The predicted risk is 166 per 10(5), or one in 600; the estimated median age of occurrence is 67 years. These data deserve to be considered with other morbidity factors in the context of the neonatal circumcision debate.

Circumcision, one of the most cornmon surgical procedures on male subjects in the United States, continues to be controversial in terms of risks and benefits involved. Among the many points of debate is the protective value of the operation against cancer of the penis vs the morbidity of the procedure. The American Academy of Pediatrics states that circumcision is an effective method of preventing cancer of the penis.[1] It has been maintained, however, that in spite of this protective effect, the incidence of cancer of the penis is so low as not to justify the risks of circumcision.[2]

Incidence data for this malignancy are based on the number of cancers per year in 100,000 men, without regard to circumcision status. Since uncircumcised males are essentially the only population at risk, these figures underestimate the incidence of cancer of the penis in this uniquely susceptible group. In addition, annual incidence data do not address the life-time risk for this cancer.

To contribute to the evaluation of neonatal circumcision from the stand-point of penile cancer prevention, we present an analysis of the lifetime risk of cancer of the penis in uncircumcised males.

Subjects and Methods

Age-specific incidence rates for cancer of the penis in US males were obtained for the years 1969 to 1971 from the Third National Cancer Survey.[3] We assumed that virtually all of the reported cancers occurred in uncircumcised males. We therefore adjusted the rates for the estimated fraction of the population that was uncircumcised in each age group. Here we relied on circumcision prevalence values published in 1968 by Stern and Lachenbruch.[4] Their figures are based on physical examination of a nonrandom sample of 330 men. The fraction of uncircumcised men in each age group is given in Table 1, column 3.

Our adjusted age-specific rates were then applied to the US life table of 1971[5] for males of all races. This yielded the cumulative lifetime risk of cancer of the penis in uncircumcised males.

Table 1 summarizes the rate adjustment procedure and shows age-specific risk rates for cancer of the penis in uncircumcised men. These adjusted rates are applied to the 1971 US life table as given in Table 2. This yields a value of 166 per 10⁵, or one in 600, as the cumulative lifetime risk for this malignant neoplasm; the predicted median age of occurrence is 67 years.

Table 1. -- Adjustment of Age-Specific Incidence Rates for Cancer of the Penis

Age Group yr	Crude Incidence for Penile Cancer, 10-5 per Year*	Uncircumcised Fraction of Age Group[4]	Adjusted Risk Rate for Uncircumcised Men, 10-5 per Year
0-30	0.0	...	0.00
30-35	0.1	.39	0.26
35-40	0.4	.39	1.03
40-45	0.5	.50	1.00
45-50	0.9	.50	1.80
50-55	1.3	.54	2.41
55-60	2.6	.54	4.81
60-65	3.4	.56	6.07
65-70	5.4	.45	9.64
70-75	5.8	.45	12.89
75-80	4.9	.45	10.89
80-75	8.3	.45	18.44
85+	7.8	.45	17.33

^* From Third National Cancer Survey.[3]

Comment

Penile Cancer

Cancer of the penis is rare in the United States (annual age-adjusted incidence 1.0 per 10⁵ for 1969 to 1971).[3] The lifetime risk of having this condition was indirectly estimated. This analysis emphasizes incidence and does not specifically predict morbidity or mortality for those contracting the disease. At present, the major morbidity can be ascribed to the psychological and physical trauma of the disease and its therapy (usually surgery).

The median survival time for 67-year-old men (the predicted median age for getting penile cancer) is 7.5 years[6] Life expectancy for unaffected men at this age is 12.1 years.[7] From this, one may roughly estimate 4.6 years of life lost per cancer, or more than 750 years per 10⁵ uncircumcised men. To cause a comparable loss of years of life, circumcision would have to be responsible for approximately ten deaths per 10⁵ neonates, assuming a life expectancy of 76 years at birth.

However, morbidity and mortality issues become at least as much philosophical as medical concerns. They are also subject to change owing to advancements in therapy and its availability to the patient. Therefore, we choose to focus chiefly on the incidence of the disease.

Comparison with the lifetime risks of other malignant neoplasms aids in providing perspective and allowing interpretation of our figures. A similar analysis using the same data sources yielded 225 per 10⁵ as the lifetime risk for cancer of the testes, more than 1,500 per 10⁵ for cervical cancer, and 4,800 per 10⁵ for cancer of the prostate. Thus, for every lifetime incident of cancer of the penis, there are 1.4 cancers of the testes, nine of the cervix, and 29 of the prostate.

Two Basic Assumptions. -- The accuracy of our estimate for lifetime penile cancer risk depends on the validity of two major assumptions. The first is that these cancers occur exclusively in uncircumcised males. This allows one to adjust population-based incidence rates for the fraction of the population that is uncircumcised and, therefore, the fraction at risk. The peculiar susceptibility of uncircumcised males to this malignant neoplasm is well known.[1] Leiter and Lefkovits,[8] in a review of the literature since 1936, found only nine reports of cancer of the penis in neonatally circumcised males.

The actual values of the adjusted age-specific risk rates depend partly on the estimates for frequency of circumcision taken from Stern and Lachenbruch.[4] We are aware of no other age-specific circumcision prevalence data for the United States that are based on physical examination. Although other sources of estimates are available, these data are unreliable[9] in that they are based solely on inquiry.

Table 2. Adjusted Age-Specific Risk Rates for Cancer of the Penis in Uncircumcised Males^*

Age Group	No. Alive at Age x	Estimated No. Alive at Midpoint of Interval	Adjusted Penile Cancer Risk Rate for Uncircumcised Men, 10-5 per Year	Expected No. of Penile Cancers
0-30	100,000	97,036	0.00	0.00
30-35	94,072	93,535	0.26	1.22
35-40	92,997	92,269	1.03	4.75
40-45	91,541	90,455	1.00	4.52
45-50	89,369	87,720	1.80	7.89
50-55	86,070	83,605	2.41	10.07
55-60	81,139	77,549	4.81	18.65
60-65	73,958	69,139	6.07	20.98
65-70	64,318	58,307	9.64	28.10
70-75	52,296	45,547	12.89	29.35
75-80	38,797	31,859	10.89	17.35
80-85	24,921	19,045	18.44	17.56
85+	13,168	6,584	17.33	5.71
Total				166.15

^* Rates applied to the 1971 US life table.[5]

Stern and Lachenbruch's figures are based on a nonrandom sample of men and may not adequately represent the diverse socioeconomic make-up of the US population. Error owing to this sampling bias would probably not have a great effect on the final value for estimated lifetime risk. To illustrate, let us assume the figures from Leiter and Lefkovits[8] systematically overestimate circumcision rates by 50% across all ages. Compensating for this assumed bias reduces the life-time risk for cancer of the penis by 25%, to 124 per 10⁵. The incidence rates taken from the Third National Cancer Survey[3] are also subject to error. The most likely direction of bias is under-estimation of the true incidence owing to incomplete reporting. These neo-plasms in older men may not come to attention, especially if death intervenes from an unrelated cause not requiring autopsy.

The second major assumption on which this analysis is based concerns the validity of applying to the birth cohort of 1971 the age-specific risk rates derived from the experience of much older cohorts. To address this assumption, one must consider the secular trend of penile cancer incidence. The incidence of this malignant neoplasm has remained at a low level and exhibits a decline during the last three decades. In 1947, the age-adjusted risk of penile cancer was 1.4 per 10⁵ per year[10]; in the period from 1969 to 1971, it was 0.9 per 10⁵ per year.[3] Part of this decline may be explained by improved economic and living conditions. To the extent that this is the case, our analysis will over-estimate the lifetime risk for those born in 1971.

The decline may be due in part to a cohort effect, however; ie, increased circumcision rates in younger cohorts would contribute to the overall lowering of penile cancer incidence in the entire population. This latter effect would not affect our analysis since we exclusively address the risk to the uncircumcised male and not to the entire male population. We believe the life-table analysis provides a reasonable estimate of the risk for cancer of the penis in uncircumcised males.

Populations, Predictions, and Hygiene. --—These figures largely reflect the experience of the white majority of the US population. The lifetime risk may be greater for racial groups with higher incidence rates and mortality for this malignant neoplasm. For instance, American blacks suffer a 2.5-fold increased annual incidence; their mortality may be as much as four times higher than that of whites.[11] It is not clear what component of this increased risk may be accounted for by lower circumcision rates.

Since the uncircumcised male is uniquely susceptible, virtually all of these cancers are preventable by neo-natal circumcision. The number of lifetime incident cancers that could be prevented annually by circumcision can be estimated with birth statistics available for 1971. In that year, there were 1,822,910 recorded live male births.[12] If none had been neonatally circumcised, our analysis predicts that one in 600, or more than 3,000 would have penile cancer in their lifetimes.

It has been suggested that careful penile hygiene could provide a level of protection nearly comparable with that provided by circumcision. St John-Hunt[13] and Diddle[14] have indicated, however, that adequate penile hygiene among uncircumcised males in Western countries is in fact not observed on a population basis.

Considerations Other Than Penile Cancer

A decision regarding the value of circumcision may not be made solely on the basis of penile cancer risk. Other points of consideration include the greater incidence of balanitis and phimosis in uncircumcised males as well as the risks of adult circumcisions these conditions may necessitate.[1] In addition, there is a finite complication rate for neonatal circumcisions. Frequency of important nonnatal complications (primarily hemorrhage, infection, and trauma) ranges from 0.06% to 0.25% in the three largest studies published since 1953[15-17]; these studies involved more than 24,000 patients. The rates may be expressed alternately as 60 per 10⁵ and 250 per 10⁵, respectively, and are comparable to the lifetime risk for penile cancer estimated here for uncircumcised males.

Mortality as a complication of circumcision is rare. A review of the literature since 1953 yielded only two reports of circumcision-related deaths.[15,18] Vital Statistics for 1971[7] records no male deaths in categories[19] that include not only circumcision-related deaths but also those owing to complications and misadventures of prophylactic appendectomies and sterilizations, among others. These figures may underestimate the circumcision-related morbidity and mortality owing to reluctance on part of physicians and hospitals to report such incidents. Difficulties in assigning cause of death where multiple medical problems contribute may also be a factor.

Finally, arguments based on religious and social preference continue to play a major role. Such issues are not amenable to the cost vs benefit approaches that can be applied to more medical aspects. We submit that the lifetime risk for penile carcinoma deserves inclusion among these factors figuring in the controversy regarding circumcision. Further investigation is indicated in several areas. Specifically, dependable data regarding circumcision rates and firm definition of related complications and their frequencies are needed. Comparable figures for disease usually attendant to the uncircumcised state (such as complications owing to phimosis and balanitis) would also be helpful. However, given the many fronts on which the subject is being debated, it is likely that circumcision will remain a controversial topic.

References

1. Thompson HC, King LR, Knox E, et al: Report of the Ad Hoc Task Force on Circumcision. Pediatrics 56:610-611, 1975.
2. Morgan WKC: The rape of the phallus. JAMA 193:123-124, 1965.
3. Cutler SJ, Young JL Jr (eds): Third National Cancer Survey: Incidence Data. National Cancer Institute Monograph 41. Bethesda, Md, US Dept of Health, Education, and Welfare, Public Health Service, 1975.
4. Stern E, Lachenbruch PA: Circumcision information in a cancer detection center population. J Chron Dis 21:117-124, 1968.
5. National Life Tables: 1969-1971, publication HRA 75-1150, vol 1, No. 1. Rockville, Md, US Dept of Health, Education, and Welfare, National Institutes of Health, 1976.
6. Axtell LM, Asire AJ, Myers MH (eds): Cancer Patient Survival: Report Number 5, publication 77-992. Bethesda, Md, US Dept of Health, Education, and Welfare, National Institutes of Health, 1976.
7. Vital Statistics of the United States, 1971, Volume II: Mortality, Part A. Rockville, Md, US Dept of Health, Education, and Welfare, 1975.
8. Leiter E, Lefkovits AM: Circumcision and penile cancer. NY State J Med 75:1520-1522, 1975.
9. Lilienfeld AM, Graham S: Validity of deter-mining circumcision status by questionnaire as related to epidemiological studies of cancer of the cervix. J Natl Cancer Inst 21:713-720, 1958.
10. Dorn HF, Cutler SJ: Morbidity from Cancer in the United States, Part 1, monograph No. 56. US Dept of Health, Education, and Welfare, Public Health Service, 1959.
11. Lilienfeld AM, Levin ML, Kessler II: Cancer in the United States. Cambridge, Harvard University Press, 1972, p 76.
12. Vital Statistics of the United States, 1971, Volume I: Natality. Rockville, Md, US Dept of Health, Education, and Welfare, 1975.
13. St John-Hunt D: Circumcision of the newborn: Is it good preventive medicine? Med J Aust 1:1100-1101, 1967.
14. Diddle AW: Should circumcision be routinely? Med Asp Hum Sex 1:32-33, 1967.
15. Speert H: Circumcision of the newborn. Obstet Gynecol 2:164-172, 1953.
16. Shulman J, Ben-Hur N, Neuman Z: Surgical complications of circumcision. Am J Dis Child 107:149-154, 1964.
17. Gee WF, Ansell JS: Neonatal circumcision: a ten-year overview. Pediatrics 58:824-827, 1976.
18. Annunziato D, Goldblum LM: Staphylococcal scalded skin syndrome: A complication of circumcision. Am J Dis Child 132:1187-1188, 1978.
19. Eighth Revision International Classification of Diseases (ICDA), vol 1. US Dept of Health. Education, and Welfare, Public Health Service, 1975, p 405.