Types of studies

We have included pre-clinical studies regarding the effects of flower pollen extracts as a background and narrative review. Moreover, we have included clinical trials, randomized controlled trials, cohort, and case-control studies for our systematic review and meta-analysis. Editorials, commentaries, and review articles were used only for the background and the narrative review.

Outcome measures

The primary outcome of the study was the improvement of disease-related quality of life in terms of clinical response to the treatment as defined by the investigators. Clinical response to the treatment was generally evaluated in terms of NIH-CPSI and SF-36 questionnaires. Moreover, the improvement of symptoms [urinary and sexual symptoms, in terms of the International Prostatic Symptoms Score (IPSS)] and other questionnaires were also considered as outcome measures, if used by the investigators.

Risk of bias assessment

The risk of bias was performed by using the Newcastle-Ottawa Scale for risk of bias assessment [14].

Search strategy and research methods

We performed a search of literature up to June 2016 using the Medline computerized database of the US National Library of Medicine. The Google Scholar database was used, too. The Medline search was carried-out using Medical Subject Headings and free text terms as follows: ‘pollen extract’, or “flower pollen extract” and ‘prostate’ (exploded) were combined with the terms: ‘treatment’ and ‘therapy’. Abstracts were not considered when full articles focusing on the same studies were available. Due to the limited number of pre-clinical studies published, we also included all non-English language papers as well. In cases of non-English language papers, the paper was included if the abstract was written in English and informative. Overlapping experiments have not been included because they were considered redundant. We considered as background information and as a comparative paper the latest review about the role of flower pollen extract in CP/CPPS patients by Wagenlehner FM published in 2011 [15]. From an initial literature search with pollen extract and prostatitis, a total of 23 extended papers were screened and 15 were selected and included in the present review. Finally, 10 clinical trials and 5 pre-clinical studies were analysed and are discussed in this review (Fig. 1). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and Meta-analyses of Observational studies in Epidemiology (MOOSE) guidelines for the reporting of this present study was used in order to perform an accurate research check-list and report [16, 17]. The meta-analysis was performed using Review Manager 5.3 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) software. The inverse variance technique for the meta-analysis of the hazard ratios has been used. Due to the fact that the studies’ heterogeneity cannot be explained, a random-effects model has been employed which in fact involves an assumption that the effects being estimated in the different studies are not identical.

Review methodology

Two authors performed the study selection independently (TC and PV). All disagreements were resolved by the senior author (VM). Titles and abstracts were used to screen for initial study inclusion. Full-text review was used where abstracts were insufficient to determine if the study met inclusion or exclusion criteria. Two authors (TC and PV) independently performed all data abstraction including evaluation of study characteristics, risk of bias, and outcome measures with independent verification performed by the senior author (VM). The limited number of the studies collected did not require any other authors.

Pre-clinical evidence

Anti-inflammatory effect

It was concluded that pollen extracts are able to inhibit prostaglandin and leukotriene synthesis and this effect is comparable to that of diclofenac and indomethacin and approximately 10 times higher than that of aspirin [20].

Anti-proliferative effects

Several animal models showed that pollen extract has a possible effect on the prostate via the androgen metabolism [21]. Talpur and co-workers demonstrated that pollen extract decreased the size of the prostate in androgen-induced prostatic enlargement in rats [21]. The effect of pollen extract on prostate enlargement is due to the fact that a fraction of this compound is a powerful mitogenic inhibitor of fibroblastic and epithelial proliferation [15]. Moreover, Kamijo and co-workers found that pollen extract protects acinar epithelial cells and inhibits stromal proliferation in association with enhanced apoptosis [12]. Finally, several in vitro studies demonstrated that pollen extract is able to inhibit prostate cancer cell growth, as found by Habib [18]. This effect is even more pronounced in hormone-independent models, suggesting that there might be a place for pollen extract in the control of abnormal growth of hormone-insensitive cells [18].

Clinical evidence and meta-analysis

Non-RCTs

As reported in Table 2, 6 clinical, non randomized trials including 206 patients were selected. The mean response rate in non-controlled studies was 83.6% (62.2%-96.0%). Cai and co-workers in a non-randomized clinical study reported a clinical response rate of 90%, demonstrating that pollen extract in association with vitamins significantly improved total symptoms, pain, and QoL in patients with non-inflammatory CP/CPPS without severe side effects [11]. The same results, in terms of clinical efficacy, were reported by Rugendorff [10] and Buck [23] in two non-randomized trials which reported a clinical response rate of 62.2% and 86.6%, respectively. Moreover, three studies by Japanese researchers demonstrated a high clinical response rate to pollen extract treatment in patients with both class IIIa and class IIIb CP/CPPS [24–26].

RCTs and meta-analysis

The mean response rate in RCTs was 74.4% (70.6%-78.1%). The latest RCT carried out by Iwamura and co-workers demonstrated a response rate of 78.1% in 50 patients affected by CP/CPPS after 8 weeks of treatment [29]. The authors defined the clinical response as a decrease in the NIH-CPSI total score by at least 25% [29]. They did not observe severe adverse events in any patients in their study [29]. On the other hand, Cai and co-workers, in a cohort of patients randomized to pollen extract or ibuprofen, reported a response rate of 75.6% in the flower pollen extract group [27]. Both class IIIa and class IIIb CP/CPPS patients were enrolled and, moreover, it was reported that adverse events were less frequent in the pollen extract group than in the ibuprofen group [27]. In the largest study, Wagenlehner and co-workers demonstrated a clinical response rate of 70.6% [13] in 139 patients affected by inflammatory CP/CPPS and treated for 12 weeks with flower pollen extract. They concluded that the beneficial effect continued to improve after 12 weeks’ treatment showing that pollen extract can be recommended for patients with inflammatory CP-CPPS for long-term treatment [13]. In 2006 Elist, by carrying out a double-blind study which included 60 patients with class IIIa or class IIIb CP/CPPS who were treated with flower pollen extract for 6 months, reported an overall clinical response of 73% [28]. All these 4 RCTs were used for the/included in our meta-analysis. We included 384 patients from 4 studies. The meta-analysis revealed that flower pollen extract could significantly improve patients’ quality of life [OR 0.52 (0.34-.0.81); p = 0.02]. Figure 2 shows the forest plot of the effect of pollen extract on CP/CPPS patients in terms of clinical response rate, as defined by the investigators.

Sub-analysis on the basis of CP/CPPS type (class III a or b)

The analysis of the 4 RCT studies did not permit us to clearly identify which CP/CPPS sub-type was the best candidate to treat with the pollen extract. In this sense, the CP/CPPS class type is not able to predict patients’ clinical response to the treatment. Only one out of four studies enrolled inflammatory CP/CPPS (class A) [13], while the other three studies enrolled both class III a and b [27–29]. Cai and co-workers enrolled 25 patients with inflammatory CP/CPPS (type IIIa) and 62 type IIIb [27]. They found that in the pollen extract group patients affected by type IIIb CP/CPPS showed higher QoL results and a lower pain level following treatment in terms of the NIH-CPSI score (the NIH-CPSI score was 24.8 ± 1.8 at the enrolment versus 11.7 ± 1.7 at the follow-up visit; P < 0.001) when compared with type IIIa CP/CPPS patients [27]. Iwamura and co-workers enrolled 20 participents with class IIIa and 19 with class IIIb, without any reference to the difference between the two groups [29]. Finally, Elist did not report the results stratified by the CP/CPPS class [28]. In the two studies in which in which a data stratification according to class IIIa or b, 84 class A CP/CPPS and 30 class b had been treated with pollen extracts, while 80 class A CP/CPPS and 32 class b were considered as controls. The lack of data did not allow a significant analysis.

Risk of bias assessment

The 4 RCTs included showed few risk of bias. Three studies contained both class IIIa and IIIb CP/CPPS patients, thus introducing the risk of a selection bias. Moreover, the RCT by Elist showed an important risk of a selection bias due to the fact that in this study patients between 20 and 60 years were included.

Main findings

All pre-clinical studies demonstrated that pollen extracts show an important anti-inflammatory effect due to the inhibition of prostaglandin and leukotriene synthesis [20]. Moreover, the dose-dependent, anti-inflammatory action of pollen extract in nonbacterial prostatitis in rats leading to decreased levels of interleukin-1b, interleukin-6 and a tumour necrosis factor, decreases glandular inflammation [15] has been demonstrated. The anti-inflammatory effect of pollen extract is approximately 10 times higher than that of aspirin [20] and did not lead to significant adverse events. This aspect is very important to highlight, due to the fact that the low prevalence of adverse effects correlates with a high patient compliance rate to the treatment. Moreover, several pre-clinical experiences demonstrated that flower pollen extract is able to inhibit 5a-reductase activity in the epithelium and stroma of the prostate in vitro, inhibiting the formation of dihydrotestosterone from testosterone [15]. This could be the reason for the improvements in urinary symptoms reported by the patients. However, the inhibition of 5a-reductase activity requires a long-term treatment as highlighted by several authors [15]. Even if a placebo effect was generally reported in patients treated with phytotherapeutic agents, in the 4 RCTs, clinically significant improvements were only observed in the pollen extract group and not in the placebo group. Finally, while Wagenlehner and co-workers found a decrease in leukocytes in post-prostate massage urine samples in both patients and controls [13], they did not find a significant difference between the two groups in terms of leukocyte number and for this reason leukocytes cannot be correlated with clinical success [13]. This aspect supports the hypothesis that the presence of inflammatory cells in the post-prostate massage urine sample is not a laboratory characteristic that is able to predict treatment response.

Strengths and limitations of the present study

In this review we excluded all studies on the effect of pollen extract on patients affected by benign prostatic hyperplasia or other urological diseases that can determine symptoms. Moreover, we excluded all studies in which the dosage of the compound was indicated in the publication. For this reason, despite the latest review by Wagenlehner we have excluded the paper by Li [30]. On the other hand, the most important limitation of this review is the lack of a pharmacokinetic evaluation of pollen extract. As highlighted by Wagenlehner, pharmacokinetic studies on the absorption, distribution, metabolism or excretion of the active components of flower pollen extracts have not been performed [15]. This is due to the fact that it is not known which compounds are primarily responsible for clinical efficacy [15].

Clinical implications

It is well known that there is no standard treatment or CP/CPPS to date. Amongst all the drugs and therapeutic approaches suggested and used, phytotherapeutic agents are those most widely prescribed in every day clinical practice with variable success. However, their use has only rarely been evaluated in suitable clinical trials. On the other hand, pollen extract has been sufficiently evaluated in preclinical and clinical studies [10, 13, 15]. Herein, we report encouraging results in terms of variations in NIH-CPSI and QoL scores in patients treated with pollen extracts indicating that the use of pollen extract appears to be safe and well tolerated by patients and, for this reason, the compliance to the treatment is high.