Patients with penile cancer and the risk of (pre)malignant cervical lesions in female partners: a retrospective cohort analysis

Authors


Correspondence: Simon Horenblas, Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, Plesmanlaan 121, 1066CX Amsterdam, the Netherlands.

e-mail: s.horenblas@nki.nl

Abstract

Objective

  • To determine if female partners of patients with penile cancer have more cervical lesions and neoplasia than would be expected from population-based data.

Patients and Methods

  • We included all consecutive patients with primary penile carcinoma in the period 2004–2010.
  • Results of Dutch cervical cancer screening were used to consider (pre)malignant cervical lesions in female partners of patients with penile cancer.

Results

  • In all, 206 women were included.
  • Gynaecological information was available in 195 women: Papanicolaou test (PAP) smears were normal were in 129 partners, 10 smears were abnormal (5.1%, 95% confidence interval 2.5–9.2).
  • PAP2 was found in five, PAP3a in two, PAP3b in two women and PAP4 in one woman.
  • Colposcopy in two women with PAP3b showed cervical intraepithelial neoplasia grade 3 in both.
  • This prevalence was not different from baseline results in the general Dutch population.

Conclusion

  • Female partners of patients with penile cancer did not show more premalignant cervical lesions than in the general population.
Abbreviations
BMD

borderline or mild dyskaryosis

CIN

cervical intraepithelial neoplasia

(hr)HPV

(high-risk) human papillomavirus

PAP

Papanicolaou test

Introduction

Infection with human papillomavirus (HPV) is associated with cervical and penile cancer. Especially infection with high-risk types of HPV (hrHPV) is considered necessary for the development of cancer. HPV can be detected in virtually all cervical carcinoma [1]. The two most common ones, HPV-16 and HPV-18, are found in >70% of all cervical cancers [2, 3]. The overall prevalence of HPV in penile cancer ranges from 22% to 72%. In penile cancer cases 24–65% hrHPV has been detected [4, 5]. HPV-16 and HPV-18 accounted for 60% and 13% of infections [6]. HPV infection is primarily transmitted by genital contact. It is reasonable to assume that sexual behaviour affects rates of HPV infection and disease in sexual partners [7-9].

Previous studies have suggested an association between penile and cervical cancer. Before HPV was considered a common aetiological factor, this suggestion was based on anecdotal reports of couples where the woman had cervical cancer and the man had penile cancer. The most suggestive findings were reported from studies with female partners of men with penile cancer, showing an over-representation of women with cervical cancer in three studies [10-12]. However, another two studies did not confirm the findings of these investigations [13, 14]. Conversely, penile lesions were found in 68% of partners of women with cervical intraepithelial neoplasia [15]. HPV was detected in 59% of penile scrapings in men with penile lesions, containing mainly hrHPV types. Histological features of intraepithelial neoplasia were seen in 32% of lesions [16]. While vaccination programmes against HPV for girls have been implemented in the Netherlands and many other countries, the issue of vaccinating boys is not settled. More information on the exact transmission pattern could be extremely helpful in that matter. This study was initiated with the aim to determine if female partners of a contemporary cohort of patients with penile cancer had more cervical (high-grade) lesions than would expected from population-based data.

Patients and Methods

This study was approved by our Institutional Ethical Committee. Only after written consent, information on gynaecological history and pathology was obtained from partners of patients treated in our institute.

Between January 2004 and October 2010, 418 patients with primary penile carcinoma were treated. Relevant clinical and pathological data were collected from our prospectively kept database. In all, 14 (3.3%) patients could not be contacted. Of these, 10 (2.4%) patients were referred to other hospitals for follow-up or treatment and accurate contact information at our hospital was outdated. Four (0.95%) patients were lost to follow-up because they moved to another country. Another 88 (21.1%) patients had died at the time of registration for this study; their partners could not be contacted due to privacy and ethical issues. Penile cancer was the cause of death in 50 (56.8%) patients. The other 38 (43.2%) patients died of other causes. As a result, 316 patients were available for partner analysis. For the latter, we selected only heterosexual couples with a relationship of >12 months at the time of diagnosis of penile cancer. It was considered not practical and ethically reasonable to try to find and contact all sexual partners of patients that did not have a stable relationship. In all, 75 patients did not have a stable heterosexual relationship, defined as sexual active male–female relation on questioning, including four patients who had a male partner. Several indicated they had sexual intercourse, but due to ethical issues their partners could not be included in this study. The resulting 241 patients were potentially eligible (Fig. 1).

Figure 1.

Total cohort: patients with penile cancer and their partners.

We used Papanicolaou test (PAP) smear data from population-based cervical cancer screening programmes in the Netherlands. Women aged 30–60 years living in the Netherlands and known at the registry of the local civil authorities (Gemeentelijke Basis Administratie) receive an invitation letter to participate in the screening programme for cervical carcinoma once every 5 years [17]. Smear results and additional gynaecological history was collected from GPs. Cervical smears were classified according to the CISOE-A (National Proforma reporting on Composition, Inflammation, Squamous, Other and endometrial, and Endocervical cylindrical epithelium, and Adequacy) classification used in the Netherlands. The results can be translated easily into Bethesda 2001 classification, i.e., low-grade abnormalities such as borderline (PAP 2) or mild (PAP 3a1) dyskaryosis are known as Borderline cytological changes or mild dyskaryosis (BMD), and higher grade abnormalities such as PAP 3a2, PAP3b and PAP4 as >BMD [18]. The prevalence in a Dutch population in the same age group was 2.4% BMD and 0.9% >BMD [19].

Histological examination was done locally and specimens were classified as normal (cervical intraepithelial neoplasia, CIN0), CIN1, 2 or 3 or invasive cancer, according to international criteria.

Results

From the 241 female partners, 206 signed informed consent after questioning. From 11 of these, we received no gynaecological information from their GP. Gynaecological history and PAP-smear results were available for 195 respondents (94.6% response rate). The median (range) age of the women at the time of their partner diagnosis was 61 (32–88) years. From the 195 respondent women, 129 did not have abnormal PAP smears and 10 were abnormal (10/195 [5.1%], 95% CI 2.5–9.2). Five women had PAP2. PAP3a was found in two women, two women had PAP3b and one partner had invasive cervical cancer stage IIA detected with a PAP4 at screening. Of the five PAP2, three women had been treated for myoma and polyps by abdominal hysterectomy 12, 25 and 33 years before diagnosis of their partner's disease. In two women with PAP3a, colposcopy showed no abnormal findings in one woman and CIN1 in the other. Of the two women with PAP3b colposcopy showed CIN3 in both. One partner had invasive cervical cancer stage IIA detected with a PAP4 at screening 8 years before her partner's diagnosis of penile cancer (Fig. 1). The median (range) age of the 10 partners with abnormal PAP smears was 42 (39–53) years. In 56 women no PAP smear was performed. In all, 17 of these women (30.4%) underwent a hysterectomy for various non-oncological reasons except for one case of dysplasia, 27 years before her partner's diagnosis of penile cancer. The remaining women did not participate in the screening for several reasons including age. None of them had a history of cervical cancer. The mean (range) time interval between smear and penile cancer diagnosis was 60.7 (0.2–509.9) months. In sum abnormal PAP smears were found in 10 female partners (7.2%, 10/139) and abnormal histology in four of them (2.9%, 4/139).

Discussion

The present results did not confirm the hypothesis that higher numbers of (pre)malignant findings would occur in current female partners of patients with penile carcinoma than would be expected from the general female population. Overall, in 195 included women, only 10 smears were abnormal (5.1%, 95%CI 2.5–9.2). More specifically, five women were detected with PAP2 and two with PAP3a, resulting in 3.6% BMD (7/195) and two woman were detected with PAP3b, resulting in 1.0% >BMD (2/195). This prevalence was not different from baseline results in a Dutch population in the same age group with 2.4% BMD and 0.9% >BMD [19]. Although we found one (0.05%) invasive cervical cancer in the present study, our cohort was too small to reliably compare this to the general incidence of invasive cervical cancer in the Netherlands (0.03%, 95%CI 0.01–0.07%) [20].

Previous studies [10-12] suggested that cancer of the penis and cervix may share the same aetiological factor(s), because significant numbers of invasive cervical cancers were detected in partners of patients with penile cancer. Possible explanations for the fact that partners of patients with penile cancer in the present study did not have cervical abnormalities are absence of intercourse between the couples and different susceptibility to HPV infection. Furthermore, as it is not yet clear how HPV transmission between humans exactly takes place and as the process of HPV infection to invasive cancer can take decades, it is difficult to interpret the present findings. As the development of HPV induced invasive penile cancer can take more than 15 years, the current partner is not necessarily the source of infection. HPV transmission might have occurred in the past by another female partner other than the current one, resulting in penile disease without cervical disease. Another explanation is different aetiological factors of penile cancer and cervical cancer. In contrast to cervical cancer, lichen sclerosis et atrophicans (balanitis xerotica obliterans), a chronic inflammatory skin disease of unknown aetiology is suggested as a precursor lesion in penile cancer in elderly men, as in vulvar cancer in women [21]. A limitation of the present study is the lack of HPV data in both males and females. This could bias the association between penile cancer and cervical cancer.

Due to the size of the population the present results give only a limited insight into the possible relation between penile cancer and (pre)malignant lesions of the cervix. Another serious limitation of the present study was that partners of deceased patients at the time of inclusion could not participate due to ethical and privacy issues. Another limitation is the absence of data on female partners of patients without a stable sexual relationship. These partners could be more at risk of developing cervical lesions. Understandably, partners could only be contacted after consent of the patients with penile cancer. A significant proportion (57.5%) of these patients died of penile cancer. Furthermore, one could question the completeness of data on PAP-smear results from GPs of partners. GPs in the Netherlands administer first-line care accurately, therefore information on gynaecological history is considered complete.

In conclusion, the present study on current female partners of patients with penile cancer did not show more (pre)malignant cervical lesions than expected from the prevalence in the general female population. As the clinical course of HPV infection in men is to a large extent unknown, and the time of infection of sexual partners is probably associated to premalignant stages of disease, such as flat penile lesion [21], more studies, including partners of >15 years before penile cancer diagnosis and HPV data, should be undertaken to gain more insight in this issue.

Conflict of Interest

None declared.

Ancillary