Human papillomavirus and risk of prostate cancer: a systematic review and meta-analysis

Abstract

Background: There is growing evidence showing a putative association between high-risk human papillomavirus (HR-HPV) infection and an increased risk of PCa.

Objective: The aim of the current meta-analysis was to evaluate the association between HPV infection and PCa risk.

Methods: This analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis guidelines. We included all studies on HPV DNA or antibodies detected in biopsy tissues or sera. Available data were extracted from the article, including means and standard deviations in all case-control groups.

Results: Thirty studies that investigated the link between HPV-16 and -18 were identified as eligible for this systematic review and meta-analysis, including a total of 6321 participants. The pooled OR showed increased risk of PCa (OR =1.37; p < .01) in men positive for HPV-16. There were seven studies with 2391 PCa cases and 4059 controls investigating the association between HPV-18 infection and PCa risk. Significant heterogeneity between study was found in the pooled analyzes. The pooled OR did not show increased risk of PCa (OR =0.80; p = .49) in men positive for HPV-18.

Conclusions: This meta-analysis suggests that HPV-16 infection could represent a risk factor for PCa, whereas we found no such association for HPV-18. Further well-conducted studies could be useful to confirm this conclusion.

Keywords:

• Human papillomavirus
• HPV16
• HPV18
• prostate cancer
• meta-analysis
• risk factor

Introduction

Prostate cancer (PCa) continues to pose significant health care challenges worldwide. It remains the commonest non-dermatologic cancer in men with 233,000 estimated new cases per year in USA and it is the second most common reason for death due to cancer [1].

Unfortunately, there are few knowledge about the underlying mechanisms this disease. In this regard, both environmental and hereditary components are supposed to play crucial roles in prostate carcinogenesis [2–6]. Moreover, there is growing evidence showing a putative association between high-risk human papillomavirus (HR-HPV) infection and an increased risk of PCa [2,7,8].

Sutcliffe et al. proposed that the sexually transmitted diseases (STDs) may serve an important role in the initiation of PCa [9]. Viral infection may induce chronic and recurrent in ammation of the prostate, and we hypothesize that in ammation of the prostate induced by STDs may increase the risk of PCa [10].

To this regard, it should be point out that inflammation at the tissue level is present in approximately 17% of all cancer cases [11]. Biological studies have provided evidence that active oxygen and nitrogen radicals produced by the inflamed tissue increase the risk of cancer by suppressing antitumor activity and stimulating carcinogenesis [12,13].

It is plausible that in men with pre-existent prostatic inflammation, HR-HPV infection, often localized at the prostate gland level, may contribute to the development of chronic inflammation and regenerative “risk factor” lesions [14]. The immune system plays a key role in preventing tumor formation by recognizing and destroying malignant cells. On the contrary, the selective pressure exerted by the immune system secondary to HPV infection causes tumors to develop multiple mechanisms to escape immune detection, including the induction of immune tolerance, local suppression of the immune response, and systematic disruption of T-cell signaling [15]. Based on these considerations, the aim of the current meta-analysis is to evaluate the association between HR-HPV infection of 16 and 18 genotype and PCa risk.

Methods

Systematic literature search

This analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis guidelines [16]. We performed a systematic literature search of PubMed, EMBASE, Cochrane, and Academic One File databases using Medical Subject Headings (MeSH) indexes, keyword searches, and publication types until December 2016. The search terms combined text words and MeSH terms. For example, the search terms for “cross-sectional”, “prostate cancer”, “human papillomavirus”, “HR-HPV”, “HPV16”, “HPV18”, “risk factors”, and “prostate sample”.

We included all studies on HPV DNA or antibodies detected in biopsy tissues or sera. With respect to HPV detection methods in prostate cancer tissues were used the PCR primers, type-specific PCR primers, and both combined, non-PCR methods as in situ hybridization (ISH), immunohistochemistry and HC2. Exclusion criteria were as follows: studies on immunosuppression patients, case reports, publications not in English, studies without extractable data from the original article, and studies lacking HPV infection. Available data were extracted from the article, including means and standard deviations in all case-control groups.

Study selection

Citation lists of retrieved articles were screened manually to ensure sensitivity of the search strategy. References of the included papers were hand searched to identify other potential relevant studies. Studies were reviewed by two independent reviewers (GIR and AEC); differences in opinion were discussed in consultation with the last author (SLV).

Data were extracted independently by two reviewers. Discrepancies for inclusion between the investigators were resolved by discussion or further consultation with a third author.

Statistical analysis

Meta-analysis was conducted to determine the pooled odds ratio (OR) for HR-HPV and risk of prostate cancer (PCa), based on histological biopsy. We constructed 2 × 2 table including number of events/total number of patients with PCa and number of events/total number of patients with HR-HPV-16 or -18 infection.

Natural logarithm (ln) of the OR [ln(OR)] were calculated to estimate the risk of PCa in patients with HR-HPV. Standard error (SE) of the lnOR was calculated through a first-order Taylor series conversion, where SE[ln(OR)] = (1/OR) * SE[OR]. Begg’s and Egger’s methods were used to assess publication bias, using log transformation of the effect size. The shape of the funnel plots is used to show symmetry and Egger’s methods are applied to indicate significant publication bias for the analyzes exploring association between PC risk and HPV-16 or HPV-18 infection.

Statistical heterogeneity was assessed using the Cochran-Q and I² statistics. Specifically, statistical heterogeneity was tested using the chi-square test. I² ≤ 50%, the variation of the studies was considered to be homogenous, the fixed effect model was adopted. If I² > 50%, there was significant heterogeneity between studies, the random effects model was used. For estimating the association between HPV methods of detection and risk of PCa, a multivariate logistic regression analysis, adjusted for year of study and age has been performed. All p values are two-tailed, α ≤ 0.05 was considered statistically significant (p ≤ .05). The analyzes were performed using RevMan software v.5.1 (Cochrane Collaboration, Oxford, UK) and using SPSS v. 19 software (SPSS Inc, IBM Corp, Somers, NY).

Results

In total, 230 references were identified from the initial online databases and relevant references. After evaluating the title and abstract of each study, 137 studies were excluded as they did not meet the inclusion criteria. Subsequently, we carefully read the full texts of the remaining 96 studies. Thirty studies were identified as eligible for this systematic review, with a total of 6321 participants (Figure 1 and Table 1).

Figure 1. Flowchart of included studies.

Table 1. Characteristics of included studies.

Study ID	Year	Country	Age (years)	Sample	Detection method
Adami et al.	2003	Sweden	Not reported	–	Serum
Aghakhani et al.	2011	Iran	62.77 (±7.40) for PCa 69.45 (±7.96) for controls	Paraffin‐embedded tissue	PCR, general primers.
Anwar et al.	1992	Japan	73 (mean) (ranging from 53 to 88) for all cohort	Paraffin‐embedded tissue	PCR, general primers.
Carozzi et al.	2004	Italy	Not reported	Fresh frozen	PCR
Chen et al.	2011	Japan	63 (median) (range 44–85) for all cohort	Paraffin‐embedded tissue	PCR, type‐specific
Dillner et al.	1998	Sweden	Not reported	Serum	ELISA
Dodd et al.	1993	Canada	79.1 (ranging 69–86) for PCa 71.4 (ranging from 69 to 86) for controls	Fresh frozen	PCR
Ghasemian et al.	2013	Iran	64.20 (±4.37) for PCa 70.08 (±4.20) for controls	Paraffin‐embedded tissue	PCR, general primers.
Hayes et al.	2000	USA	Not reported
Ibrahim et al.	1992	USA	64.0 (ranging 43–84) for PCa 61.0 (ranging 54–79) for control	Paraffin‐embedded tissue	PCR
Korodi et al.	2005	Sweden	Not reported	Serum	ELISA
Kucziyc et al.	2000	Germany	Not reported	Fresh frozen	PCR
Leiros et al.	2005	Argentina	Not reported	Paraffin‐embedded tissue	PCR
May et al.	2008	Germany	65.7(±8.4) for all cohort	Fresh frozen	PCR
McNicol and Dodd	1990	Canada	Not reported	Fresh frozen	PCR, type-specific
McNicol and Dodd	1990	Canada	Not reported	Paraffin‐embedded tissue	Southern blot analysis.
Michopoulou et al.	2014	Greece	65.5 (±5.97) all cohort	Paraffin‐embedded tissue	PCR, general primers
Mokthari et al.	2013	Iran	67.70 (±11.1) for PCa 68.48 (±11.89) for controls	Paraffin‐embedded tissue	Immunostaining
Moyret-Lalle et al.	1995	France	Not reported	Fresh frozen	PCR, type-specific
Noda et al.	1998	Japan	Not reported	Paraffin‐embedded tissue	PCR, consensus primers
Rosenblatt et al.	2003	USA	Range 40–64	Serum	ELISA
Rotola et al.	1992	Italy	Not reported	Fresh frozen	PCR, specific primers
Serth et al.	1999	Germany	67 for PCa (mean) (range, 51–84) 70 for controls (mean) (range, 53–85)	Fresh frozen	PCR
Silverstre et al.	2009	Brasil	62 (±7.3) for PCa 64 (±13.9) for controls	Paraffin‐embedded tissue	PCR, type‐specific
Singh et al.	2015	India	Not reported	Paraffin-embedded tissue	PCR, type-specific primers
Smelov et al.	2011	Netherland	61 (median) for PCa 61 (median) for controls	Paraffin-embedded tissue	PCR
Sutcliffe et al.	2007	USA	Range 40–75	Plasma	ELISA
Suzuki et al.	1996	Japan	Not reported	Fresh frozen	PCR, consensus primers
Tachezy et al.	2012	Czech Republic	63 (mean) for PCa 51 (mean) for controls	Paraffin-embedded tissue	PCR, general primers
Terris and Peehl	1995	USA	66.1 (range: 44–76) for PCa	Paraffin-embedded tissue	PCR
Wideroff et al.	1996	USA	64 and 65 (median) for Black and white PCa 63 and 65 (median) for Black and white controls	Paraffin-embedded tissue	PCR, consensus primers

There were 31 studies with 3050 PCa cases and 3428 controls investigating the association between HPV-16 infection and PCa risk. No significant heterogeneity between studies was found in the pooled analyzes. The pooled OR showed increased risk of PCa (OR =1.38 [95% CI: 1.16–1.64]; p < .01) in men positive for HPV-16 (Figure 2).

Figure 2. Association between HPV 16 and PCa.

There were seven studies with 2393 PCa cases and 4152 controls investigating the association between HPV-18 infection and PCa risk. Significant heterogeneity between study was found in the pooled analyzes. The pooled OR did not show increased risk of PCa (OR =1.05 [95% CI: 0.77–1.44]; p = .75) in men positive for HPV-18 (Figure 3).

Figure 3. Association between HPV 18 and PCa.

Funnel plot was generated to evaluate publication bias using the Egger’s funnel plots (Figures 4 and 5).

Figure 4. Funnel plot of studies investigating HPV 16 as a risk factor.

Figure 5. Funnel plot of studies investigating HPV 18 as a risk factor.

For study investigating HPV-16, we showed significant heterogeneity (bias =1.26 [95% CI: 0.67–1.85]; p < .0) while not for HPV-18 (bias =1.09 [95% CI: −0.67 to 2.85]; p = .17).

The association between HPV methods of detection and risk of PCa has been evaluated by the use of a multivariate logistic regression analysis, adjusted for year of study and age has been performed. We did not find any association between method of detection and risk of PCa (OR =1.12 [95% CI: 0.91–1.39]; p = .21).

Discussion

Prostate cancer is a topic of great interest also for endocrinologists; in particular, the management of androgen therapy [17] and the reflections of the major metabolic diseases on the risk of developing it [18,19]. HPV infections concern the field of STDs and represent another element of cultural linkage.

The HR-HPV infection is the most common infection among young, sexually active individuals. Indeed, it is estimated that about 75–80% of sexually active individuals will become infected in their lifetime [20]. Another important aspect of the clinical management of HR-HPV infection in men compared with women concerns the possible differences in the following parameters: viral load, duration of natural infection, absence of symptoms, lower antibody response, anatomical site differences, and lower incidence [21].

Understanding the factors that influence the progression from latent to clinical cancer is therefore essential to identify patients at high risk of progression at the time of diagnosis and understand the target to prevent and control the passage between infection and clinical disease [22].

Three meta-analyses have been published on the correlation between HR-HPV infections and the risk of prostate cancer [2,7,8]. However, their results are discordant. In the present article, we showed that HPV-16 infection was significantly associated with an increased risk of PCa (OR =1.38), whereas HPV-18 was not (OR =1.05).

In parallel, some studies showed no significant association between HR-HPV and PCa [9,23–25].

In a previous study, we showed that the prevalence of HR-HPV infection in patients with microbial male accessory gland infection was ≈30%, whereas the prevalence decreased to ≈20% among patients with male accessory gland inflammation only [20]. In this regard, it has been shown that HR-HPV could be identified in the prostatic secretions of patients with type III prostatitis and might be associated with the degree of intraprostatic inflammation. This strengthens the evidences about the relationship between HR-HPV and prostatic inflammation [26].

High-risk HPV infections usually last 1–2 years [27], which means that HR-HPV can replicate in host tissues for months in the face of constant cellular turnover and immune surveillance. Its ability to do so depends largely on its potential to regulate both viral and host gene expression [28]. In addition to its differentiation-dependent life cycle organization, HR-HPV actively suppresses both innate and adaptive immune responses [27]. The HPV oncoproteins E6 and E7 are essential for the normal virus life cycle and throughout the cancer development pathway from benign precursor lesions to invasive carcinomas [29,30].

Unfortunately, although HR-HPV has been recognized as an important risk factor of oral, cervical, anal cancer, its role in prostate cancer is far to be understood. However, HR-HPV infection and incidence rate remains high in developing countries due to lack of awareness, followed by mass screening programs, various socioeconomic issues, and low usage of preventive vaccines [31].

Furthermore, a high incidence of HR-HPV DNA has been reported in semen analysis from sexually active men, with and without risk factors for HR-HPV, and from infertile patients [32]. Foresta and colleagues [32] investigated the presence of HPV DNA sequences in semen samples of asymptomatic young adult men who had unprotected intercourse, and reported the presence of HPV semen infection in 10% of these men. Moreover, they reported that a high incidence of HR-HPV infection in sperm was found not only in individuals with risk factors but also in asymptomatic infertile men [32].

Based on these premises, sexual intercourse should be considered an important pathway to the exposure of microorganisms such as HR-HPV infection.

Articles and publications about expressed prostate secretions virus infections have been limited, with only some isolated case reports. CMV, HSV-2 and HR-HPV are the most common virus, all of which can be found in healthy populations, where they will cause latent infection or mild symptoms [33].

Xiao et al. reported that WBC count and lecithin levels were significantly correlated with HR-HPV infection illustrating that there was some relationship between HR-HPV and prostatic inflammation [26].

Before concluding, this meta-analysis exhibited some limitations. In particular, heterogeneity of included studies, as concerning definition of HR-HPV infection and methodology of such studies may represent some drawbacks of the present work.

Conclusion

Results of this meta-analysis of cross-sectional studies showed that HPV-16 infection is associated with PCa. Although the heterogeneity of current literature data concerning techniques of diagnosis of HPV in patients with PCa, HPV infection should be considered as a potential risk factor of PCa.

Disclosure statement

The authors have no conflicting interests to declare.

Additional information

Notes on contributors

Giorgio I. Russo

Giorgio I. Russo is with Department of Urology, University of Catania, Italy.

Aldo E. Calogero

Aldo E. Calogero is a Professor at Department of Andrology and Endocrinology, University of Catania, Italy.

Rosita A. Condorelli

Rosita A. Condorelli is with Department of Andrology and Endocrinology, University of Catania, Italy.

Guido Scalia

Guido Scalia is a Professor at Department of Virology, University of Catania, Italy.

Giuseppe Morgia

Giuseppe Morgia is a Professor at Department of Urology, University of Catania, Italy.

Sandro La Vignera

Sandro La Vignera is a Professor at Department of Andrology and Endocrinology, University of Catania, Italy.