Condom use promotes regression of human papillomavirus–associated penile lesions in male sexual partners of women with cervical intraepithelial neoplasia



Penile HPV-associated lesions are frequently seen in male sexual partners of women with CIN. The natural course and clinical significance of these lesions are unclear. Women with CIN and their male sexual partners were randomized for condom use (condom group n = 68, noncondom group n = 68). Males were screened for the presence of penile lesions, i.e., flat lesions, papular lesions and condylomata acuminata, and of HPV in their penile swabs by PCR testing. Median follow-up time was 13.1 months (range 2.9–57.4). The outcome of our study was clinical regression of penile lesions defined as disappearance of lesions at penoscopy. Potentially prognostic factors, i.e., HPV status, lesion type and age, were studied as well. Outcomes were assessed in 57 men of the condom group and in 43 men of the noncondom group. Condom use shortened the median time to regression of flat penile lesions (7.4 months condom group vs. 13.9 months noncondom group; HR = 2.1, 95% CI 1.2–3.7). This effect was not found for papular lesions (HR = 0.5, 95% CI 0.1–2.8). HPV-negative men showed a significantly shorter median time to regression of flat lesions (3.8 months) compared to men with either HPV-positive status (8.5 months; HR = 0.4, 95% CI 0.2–0.9) or inconsistent HPV status (13.1 months; HR = 0.2, 95% CI 0.1–0.6). Regression of flat penile lesions is HPV-dependent and accelerated by condom use. This effect is probably the result of blocking viral transmission between sexual partners. © 2003 Wiley-Liss, Inc.

Infection with high-risk types of HPV is the main causative agent for cervical cancer and its precursor lesions, i.e., CIN 1–3.1 The relationship between the natural history of CIN lesions and the presence of HPV has been clarified.2 However, the number of studies on the clinical behavior of HPV infection in men is limited.

Of the male sexual partners of women with cervical lesions, 50–65% have HPV-associated penile lesions, i.e., flat lesions, papular lesions and condylomata acuminata.3, 4, 5 The natural history of these lesions and HPV infection in men are incompletely understood because of the use of either insensitive (in situ hybridization) or unreliable (cytology) techniques for HPV detection in previous studies.5, 6, 7 Moreover, penile lesions studied before were often biopsied or treated, resulting in interference with the clinical course.

Studying HPV-associated penile lesions is hampered by a clinical pitfall: most of these lesions are flat, cause no clinical symptoms and are difficult to detect by the naked eye unless a careful visual inspection is performed. These lesions are easily detectable using a penoscope and application of acetic acid. Using the latter approach on male sexual partners of women with CIN lesions in a non-STD setting, we observed penile lesions in 68% of cases, with a predominance of flat penile lesions. A smaller number had papular lesions, and only a few had condylomata acuminata. HPV DNA, mainly of high-risk HPV types, was detected by PCR in 59% of penile swabs.8

Sexual intercourse is the predominant mode of acquiring HPV infection.9 Previous studies in men have shown that failure to use a condom is associated with HPV infection and that consistent condom use reduces the risk of acquiring genital warts.10, 11 Of interest is the work of Hippelainen et al.,6 who studied the clinical course of HPV infections in men. Most of these men were partners of HPV-infected women, and in situ hybridization positivity and condom use were significant prognostic factors for the cure rate. Viral shedding among sexual partners might have consequences for the natural course of genital lesions. If so, condom use is likely to provide an effective blockade of HPV transmission between sexual partners. However, randomized therapeutic intervention with condom use in sexual partners diagnosed with HPV-related genital lesions has not been studied.

We investigated the effect of condom use on the clinical course of HPV-associated lesions in sexual partners in a randomized clinical trial. We describe the influence of condom use and HPV status on the clinical course of penile lesions in male sexual partners of women with CIN.


CI, confidence interval; CIN, cervical intraepithelial neoplasia; EIA, enzyme immunoassay; HPV, human papillomavirus; HR, hazard ratio; PIN, penile intraepithelial neoplasia; STD, sexually transmitted disease.


Study population and design

Sexual couples were enrolled at the colposcopy clinic of the Albert Schweitzer Hospital from July 1997 to June 2002. Women with an abnormal cervical smear (mild dyskaryosis or worse) and/or colposcopic CIN and/or histologic CIN were asked to bring in their male sexual partner. Males were screened for the presence of penile lesions by penoscopy and HPV PCR testing of penile swabs. Women were screened by colposcopy, cytology and HPV testing. Baseline cervical biopsy specimens were taken for histopathologic evaluation. The exclusion criterion was regular condom use at baseline, i.e., using condoms for birth control. Both the period of condom use, i.e., only during the study and for at least 3 months, and the instructions on condom use, i.e., consequently during genital–genital contact, were discussed with the participants. Latex condoms (Durex fetherlite; The Netherlands LRC) with basic lubricant (without spermicidal and/or virus-inactivating substances) were given free, to increase study compliance. The proposal of condom use for at least 3 months was based on a pilot study of 18 couples in which more regression was observed in the condom group after this time interval (data not published). The anticipated power was based on the findings of this 6-month pilot study. Regression rates of penile lesions were 60% in the condom group vs. 25% in the noncondom group. With an HR of 3 and a power of 0.80, the total number of events needed should be at least 31. The pilot study was done between January 1995 and July 1997, and in contrast to the follow-up of women, men were only regularly followed after this period. Couples were randomized for condom use in blocks of 2 at a given time, independently of the presence of penile lesions and the grade of the cervical lesions in the women. Per block of 2 couples, the sequence of condom and noncondom allocation was based on a table of random numbers. Penoscopy, colposcopy and penile and cervical scrapings were repeated after 3, 6 and 12 months and subsequently every 6 months. Both partners were screened usually on the same day (>90%) or, otherwise, within a range of 2 weeks. Condom use was verbally verified at each visit by asking the frequency of failure to use condoms. Independently of each other and in separate rooms, partners answered a questionnaire to obtain information on lifestyle habits, including sexual behavior. This questionnaire was introduced in 1999. The study protocol was approved by the ethics review board of the hospital (protocol number CGE/95/238), and the participants gave informed consent. Of 152 couples asked to join the study, 7 refused to participate and 9 were excluded due to regular condom use, leaving 136 couples randomized. In the present report, males with penile lesions at baseline (n = 113) were selected. Colposcopic assessment of the stage distribution of the cervical lesion in the female partners of these men showed CIN1 in 57 women, CIN2 in 46 women and CIN3 in 10 women.

Penoscopic examination and specimen collection

Penoscopic examination, classification of penile lesions and specimen collection were performed as described earlier.8 Briefly, a Cervex brush (Rovers Medical Devices, Oss, the Netherlands) was used to collect cells from the top of the penis (including the glans, corona, sulcus, frenulum and the inner part of the foreskin) for HPV testing. The brush was placed in 5 ml of PBS solution with 0.001% thimerosal (Merthiolate) (BDH, Poole, UK), sent to the laboratory and stored at 4°C until testing. The colposcope was subsequently used to identify penoscopic abnormalities several minutes after application of 3% acetic acid solution, which is standard in the Netherlands. Penile lesions were categorized into flat lesions (flat or slightly elevated acetowhite lesions in which a punctate pattern could be seen), papular lesions (exophytic lesions with a smooth surface on which a hyperkeratinized layer could be present) and condylomata acuminata.8 Penoscopy was performed blinded of data from previous visits, and findings were recorded on a standardized form. During the study, penile lesions were documented by photography. Photographs were subsequently reviewed and graded by an experienced dermatologist (T.M.S.), unaware of any clinical data. In case of discrepancies (<5%) between the initial finding and the review result, a consensus diagnosis was made with the gynecologist who took the photographs and the dermatologist who reviewed them.

HPV testing

Processing the specimens and testing for HPV were done at the Department of Pathology, VU University Medical Center, as described previously.8, 12 In short, samples were centrifuged and cell pellets resuspended in 100 μl TRIS buffer. Aliquots of 10 μl were used for PCR. DNA quality was tested by β-globin PCR, and testing for HPV was done using the HPV GP5+/6+ PCR EIA. This PCR method enables the identification of 14 oncogenic (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68) and 6 nononcogenic (6, 11, 40, 42, 43 and 44) HPV genotypes by hybridizing PCR products with a high-risk and a low-risk HPV cocktail probe, respectively, in an EIA format. At baseline, 24/100 (24%) penile scrapes were inadequate for HPV testing, i.e., were negative in both the β-globin and the HPV PCR. Of the remaining 76 scrapes, 54 (71%) were positive for HPV: 40 (74%) for high-risk HPV, 5 (9%) for high-risk or low-risk HPV and 9 (17%) for low-risk HPV.

Definitions of regression of penile lesions and HPV status in males

Clinical regression of penile lesions, the primary outcome of interest, was defined as disappearance of all penile lesion types, i.e., no flat and/or papular lesions at penoscopy (Fig. 1). Regression of either flat lesions or papular lesions was defined as disappearance of either all flat lesions or all papular lesions, respectively, irrespective of the presence of the other lesion type. Thus, men with both flat and papular lesions were included in both analyses. The clinical course of condylomata acuminata was not studied since we found only 8 men with this lesion type in our non-STD population. Samples inadequate for HPV testing were not used in the analysis. The remaining were used to define the HPV status in males, which was categorized in 3 groups: (i) HPV-positive group (consistent positive HPV tests), (ii) HPV-negative group (consistent negative HPV tests) and (iii) HPV inconsistent group (HPV tests were not consistent during the study, meaning that alternately positive and negative results were obtained). We did not account for the HPV type because, on the one hand, there was a small number of men with low-risk HPV types and, on the other hand, Kaplan-Meier analyses on the regression of penile lesions showed no statistical difference stratifying for HPV type (HPV-16, other high-risk HPV or only low-risk HPV).

Figure 1.

(Upper panel) Flat penile lesions at penoscopy. (Lower panel) No visible penile lesions after 1 year.

Statistical analysis

Analyses for condom use were by intention to treat. Differences between the condom and noncondom groups were assessed by χ2 or independent t-tests. Kaplan-Meier curves were used to evaluate regression of penile lesions at 24 months. We took the time point of regression of penile lesions as the midpoint between the visits with the last positive and first negative results. Kaplan-Meier curves were compared by means of a 2-sided log-rank test. Cox regression analyses were performed to calculate the HR and 95% CI, adjusted for condom use, HPV status in males (positive, negative or inconsistent), lesion type (flat, papular or both) and age at baseline (tertiles). Within the condom group, we assessed whether the effect of condoms was dependent on the duration of use. Therefore, Kaplan-Meier analyses were performed on men who still had penile lesions present at 6 months and stratified for condom use at this time point. Statistical significance was set at 0.05. For the statistical analyses, SPSS (Chicago, IL) version 9.0 software was used.


Characteristics of the study population

The trial profile is given in Figure 2. One hundred and thirty six couples were randomized for condom use (condom group n = 68, noncondom group n = 68). Penile lesions were present in 113 (83%) men. In 2 men, only condylomata were observed. No follow-up results were available for 9 men. Rates of loss to follow-up were different for both groups: 1/59 in the condom group vs. 8/54 in the noncondom group (p = 0.01). No statistical differences were found between the baseline characteristics of the men under study and those who were lost to follow-up (results not shown). In the noncondom group, penile lesions could not be assessed during follow-up due to recent surgery in one man and partial phimosis in another.

Figure 2.

Trial profile. *Excluded at penoscopy due to previous surgery or phimosis.

Median duration of condom use was 7.0 months (range 2.9–31.2 months). Failure to use condoms was reported by 8 men, with a median of 2 times during the period of condom use (range 1–5). Median follow-up time was 13.1 (range 2.9–57.4) months. Single lesion types were found in 82/100 (82%) and multiple lesion types in 18/100 (18%) men. Flat lesions, papular lesions and condylomata acuminata were observed in 82 (82%), 30 (30%) and 6 (6%) men, respectively. During follow-up, HPV status could not be determined in 3 (3%) men since all samples were inadequate for HPV testing. HPV status was positive in 36 (37%), inconsistent in 43 (44%) and negative in 18 (19%) men. The characteristics of the condom (n = 57) and noncondom (n = 43) groups are listed in Table I. No statistical differences were found between the groups. Questionnaires were obtained from 41 men (41%): 23/57 (40%) of the condom group and 18/43 (42%) of the noncondom group. No statistical differences were found for smoking habits, age at first sexual intercourse, number of sexual partners, number of noncondom sexual partners, history of STDs, duration of the current relation and frequency of sexual intercourse (Table I).

Table I. Characteristics of the Study Population
 Number %Condom group Number (%)Non-condom group Number (%)p
  • 1

    Not determined due to samples being inadequate for HPV testing.

  • 2

    One case of the condom group was excluded because he had 200 sexual partners and 180 noncondom sexual partners. Inclusion resulted in a mean number of 17.1 sexual partners and 13.2 noncondom sexual partners in the condom group, p = 0.3 and 0.3, respectively.

Study subjects100 (100)57 (57)43 (43) 
Age (years; mean, range)38.9 (26.5–57.7)38.7 (26.5–57.7)39.3 (26.8–54.6)0.7
 <34.33320 (35)13 (30)0.5
 34.3–42.43421 (37)13 (30) 
 >42.43316 (28)17 (40) 
Follow-up (months; median, range)13.1 (2.9–57.4)13.5 (2.9–44.7)13.1 (3.0–57.4)0.8
Type of penile lesions   0.7
 Flat65 (65)37 (65)28 (65) 
 Papular17 (17)10 (18)7 (16) 
 Flat/papular12 (12)8 (14)4 (9) 
 Flat/condyloma acuminatum5 (5)2 (4)3 (7) 
 Papular/condyloma acuminatum1 (1)1 (2) 
HPV status in men (baseline)   0.8
 Positive54 (71)31 (72)23 (70) 
 Negative22 (29)12 (28)10 (30) 
HPV status in men (follow-up)   0.7
 Positive36 (37)18 (33)18 (42) 
 Inconsistent43 (44)26 (48)17 (39) 
 Negative18 (19)10 (19)8 (19) 
Characteristics obtained by questionnaires    
 Study subjects41 (41)23 (40)18 (42)0.9
 Smoking   0.2
  No23 (56)15 (65)8 (44) 
  Yes18 (44)8 (35)10 (56) 
  Number of years (mean, range)20.2 (12–30)20.0 (14–28)20.4 (12–30)0.9
  Number of cigarettes per day (mean, range)16.2 (1–25)17.9 (10–20)15.0 (1–25)0.4
 Age at first sexual intercourse (years; mean, range)17.4 (13–25)17.3 (13–25)17.4 (15–22)0.8
 History of STD   0.2
  No38 (93)20 (87)18 (100) 
  Yes3 (7)3 (13)0 (0) 
 Lifetime sexual partners (mean, range)7.62 (1–40)8.82 (1–40)6.1 (2–15)0.3
 Lifetime noncondom sexual partners (mean, range)4.72 (1–15)5.22 (1–15)4.1 (2–15)0.3
 Other sexual partner last year   1.0
  No40 (98)22 (96)18 (100) 
  Yes1 (2)1 (4)0 (0) 
 Type of relationship   1.0
  Married/living together39 (95)22 (96)17 (94) 
  Living apart2 (5)1 (4)1 (6) 
 Duration of relation (years; mean, range)11.5 (0.6–37)12.0 (0.6–32)10.9 (1.3–37)0.6
 Frequency of sexual intercourse    
  Frequency per month (mean, range)7.1 (0–25)6.5 (0–20)7.8 (0.2–25)0.4
  Frequency last month (mean, range)5.1 (0–20)4.4 (0–20)6.0 (0–16)0.3

Influence of condom use on the clinical course of penile lesions

Median time to regression of penile lesions was longer in the noncondom group (n = 43), i.e., 19.2 months, compared to the condom group (n = 57), i.e., 10.4 months (HR = 1.8, 95% CI 1.0–3.3) (Fig. 3, Table II). This effect was significant for flat lesions, i.e., noncondom group (n = 35) 13.9 months and condom group (n = 47) 7.4 months (HR = 2.1, 95% CI 1.2–3.7), but not for papular lesions, i.e., noncondom group (n = 12) >24 months and condom group (n = 18) >22.1 months (HR = 0.5, 95% CI 0.1–2.8). More regression was observed in men with flat lesions compared to men with papular lesions (HR = 0.2, 95% CI 0.05–0.5). In men who still had flat penile lesions at 6 months, median time to regression was similar in those who used condoms <6 months (n = 9) compared to those who used condoms >6 months (n = 12, log rank p = 0.8). Regression of penile lesions was not related to age.

Figure 3.

Clinical course of penile lesions in relation to condom use.

Table II. Prognostic Factors for Clinical Regression of Penile Lesions
 Regression of penile lesions
All typesFlat lesionsPapular lesions
NumberHR (95% CI)pNumberHR (95% CI)pNumberHR (95% CI)p
  1. Cox regression analysis was performed to calculate the adjusted HR and 95% CI. Adjustments were made for all other factors in the table. NA, not applicable.

Condom use         
 No431.0 351.0 121.0 
 Yes541.8 (1.0–3.3)0.05452.1 (1.2–3.7)0.02160.5 (0.1–2.8)0.4
HPV status         
 Negative181.0 141.0 91.0 
 Inconsistent430.4 (0.2–1.1)0.1330.2 (0.1–0.6)0.002120.8 (0.1–9.1)0.8
 Positive360.7 (0.3–1.7)0.4330.4 (0.2–0.9)0.0372.3 (0.2–28.4)0.5
Lesion type         
 Flat691.0 691.0 NA  
 Papular170.2 (0.05–0.5)0.002NA  171.0 
 Flat and papular110.4 (0.1–1.4)0.1111.4 (0.6–3.3)0.4115.5 (0.5–66.1)0.2
Age (years)         
 <34.3331.0 261.0 121.0 
 34.3–42.8321.2 (0.6–2.3)0.7261.3 (0.7–2.7)0.481.8 (0.2–16.0)0.6
 >42.8320.8 (0.4–1.7)0.6281.0 (0.5–2.1)0.981.1 (0.1–11.0)1.0

Influence of HPV status on the clinical course of flat and papular penile lesions

Median time to regression of flat lesions was 3.8 months in HPV-negative men and 8.5 months in HPV-positive men (HR = 0.4, 95% CI 0.2–0.9) (Fig. 4, Table II). No significant difference in median time to regression was found between HPV-positive men and men with inconsistent HPV status (log-rank p = 0.5) (Fig. 4). The majority (83%) of men with flat lesions either were HPV-positive (33/80) or showed inconsistent HPV results (33/80), whereas only 14/80 (18%) were HPV-negative. These findings support the association between flat penile lesions and HPV. In men with papular lesions, HPV status was positive in 9/28 (32%), inconsistent in 12/28 (43%) and negative in 7/28 (25%). A relation between median time to regression of papular lesions and HPV status was not found (Table II).

Figure 4.

Clinical course of flat penile lesions in relation to HPV status.


This randomized clinical trial, performed on a non-STD gynecologic population, showed that in male sexual partners of women with CIN lesions the use of condoms significantly shortened the median time to regression of flat penile lesions. Moreover, flat penile lesions regressed more slowly when HPV was present in the penile swab.

As described by others, we found that flat penile lesions are the most prevalent lesion type in male sexual partners of women with CIN.3, 4, 5, 8 Previous studies have shown an association between flat penile lesions and the presence of HPV by in situ hybridization and PCR.4, 5, 8, 13 Our finding that the regression time of flat lesions was much longer when HPV was present strengthens the value of this association. Similar findings were observed by our group in 2 separate follow-up studies of women with CIN lesions.14, 15 In these prospective studies, the regression time of CIN lesions was significantly shorter if HPV was absent and absence of HPV was associated with cytologic regression of CIN lesions. However, we did not find a relation between the regression time of papular lesions and the presence of HPV, suggesting that the pathogenesis of flat and papular lesions is different.

Interestingly, men with inconsistent HPV results during follow-up showed a similar clinical course of flat penile lesions as men with consistent HPV positivity. Therefore, nondetection of HPV in penile swabs apparently does not necessarily reflect absence of HPV. Instead, inconsistent HPV results are most likely due to relatively low viral load levels in penile swab samples that fall below the detection level. This is supported by the fact that positive signals at HPV testing were generally 5–10 times lower in penile swabs compared to those in cervical scrapes (data not shown). Compared to cervical scrapes, penile swabs often contain fewer cells and/or less DNA due to the keratinized surface of the penis. This low DNA quantity is further reflected by the relatively high number of β-globin PCR-negative penile swabs (i.e., 24%) found in the present and another study.16 As a consequence, defining HPV clearance in males was complicated. Analyses of viral clearance (defined as 2 consecutive negative HPV tests), restricted to men who were HPV-positive at baseline, showed that the HPV clearance rate was higher in the condom group compared to the noncondom group, though significance was not reached due to the low number of men analyzed [2-year rate of viral clearance was 27% in the condom group (n = 26) vs. 7% in the noncondom group (n = 15), p = 0.2].

Different rates of loss to follow-up between the condom and noncondom groups were found. Because men in the intervention group experienced themselves to be directly involved in the treatment of the women due to condom use, they possibly were more motivated to return for examination. Since we found no significant differences concerning the characteristics of the men under study and those who were lost to follow-up, it is unlikely that this difference biased our results.

Since data on sexual characteristics were collected from 1999, our findings on condom use could not be adjusted for such factors. Data were collected in 41% of subjects, which comprises a representative sample of the whole study population, and no differences were found on sex-related factors between the groups. Moreover, given the randomized design of our study, one would expect an even distribution of sex-related factors.

In our study, beneficial effects of condoms were found after condom use for at least 3 months and appeared to be similar when use was extended for a longer period. This observation needs to be interpreted with caution since the exact duration of condom use was not defined at baseline and only a minimal time of 3 months was used. From a rational point of view, the use of condoms should be continued until both sexual partners are negative for HPV. This is supported by data that the protective effect of condom use is related to the presence of HPV in the female sexual partner (Bleeker et al., unpublished data).

A plausible explanation for the observed effect of condom use might be a blockade of viral transmission from the sexual partner, thereby lowering the viral load, accelerating virus elimination and shortening the regression time of the flat penile lesion. The fact that viral load may be a determinant of regression of an HPV-induced lesion became evident from a previous study involving cervical lesions: women with low HPV-16 loads had an increased chance of viral clearance and showed a higher regression rate than women with a higher HPV-16 load.17 This notion is supported by the observation that most penile lesions regress, which is most likely due to much lower viral loads in men than in women. Although most penile lesions run a benign clinical course, a small minority persist and may show histologic features of severe dysplasia (PIN3).18 Long-term follow-up of these men should be considered because of their possible higher risk of developing HPV-associated penile cancer.

Compared to uncircumcised men, Castellsagué et al.16 found a reduced risk of penile HPV infection among circumcised men and, in the case of a history of multiple sexual partners, a reduced risk of cervical cancer in their current female partners. The authors argue that the penile shaft and the outer surface of the foreskin are covered by keratinized, stratified, squamous epithelium, which provides a protective barrier against HPV infection. In analogy, use of condoms in males with penile lesions not only reduces the regression time of their lesions but might also reduce the risk of cervical cancer in their sexual female partners.

Since most penile lesions are flat and difficult to detect with the naked eye, men are not aware of having an HPV infection. A major proportion of the male sexual partners of women with CIN, however, harbor a reservoir of high-risk HPV and form a risk for viral spread to other sexual partners. This risk is considerably reduced by condom use since the number of flat penile lesions is lower.

In conclusion, although this is a relatively small study, our results show that most flat penile lesions regress during follow-up and that this regression is HPV-dependent. Condom use promotes this regression, probably by blocking viral transmission between sexual partners. Therefore, condom use should be advocated by public health systems not only to prevent transmission of STDs like HIV, chlamydia and gonorrhea but also to reduce HPV infections and their possible clinical complications.


We thank Mr. R.E. van Andel and Mrs. M.C.G.T. Verkuyten for technical assistance on HPV testing. We thank Dr. P.D. Bezemer of the Department of Clinical Epidemiology and Biostatistics, VU University Medical Center, for critical reading of the manuscript. We thank Dr. N. Muñoz of the IARC for critical reading of the manuscript and remarks.