Introduction
Lower urinary tract symptoms (LUTS) suggestive of benign prostatic hyperplasia (BPH) are common in aging men. For decades, the symptoms of this condition have been treated with phytotherapy. Various preparations are available, mainly consisting of extracts from saw palmetto fruit, pumpkin seed, pygeum africanum bark, nettle root, or willow herb [
1,
2]. Meanwhile, selective alpha
1-receptor blockers (ARBs) and 5-alpha-reductase inhibitors (5-ARIs) have increasingly served as first-line medical treatments since they came on the market in the 1990s. However, synthetic drugs have potential side effects that might negatively influence quality of life, such that many patients prefer herbal products due to their excellent safety profile. Thus, plant extracts have been continuously used in clinical practice [
3‐
5].
In recent years, symptom bother and quality of life have become key criteria for therapy decisions [
6,
7]. Considering patient preferences has led to debates about the role of phytotherapy particularly in patients with a low risk of disease progression [
4,
8,
9].
Since 2014, the German S2e guideline suggests the option of treating LUTS/BPH with herbal preparations of proven superiority to placebo in patients who have mild-to-moderate complaints and refuse chemical compounds [
6].
Notably, relevant placebo responses have regularly been observed in randomized controlled trials in patients with LUTS/BPH [
10]. The placebo effect is rapid and may account for 40-60% of the overall symptom relief but tends to diminish over time [
10‐
12]. Consequently, the International Consultation on BPH emphasized the importance of placebo control and follow-up periods of at least 12 months for clinical research in the early 1990s [
13].
Nonetheless, most studies with herbal preparations have shortcomings, such as brief follow-up periods up to only 6 months, lack of placebo control, and small samples with less than 100 participants per treatment group [
6,
14,
15]. To date, no more than five randomized placebo-controlled long-term studies in patients with LUTS/BPH have been reported [
6]. Notably, two of these 12-month studies investigated pumpkin seed soft extract (PSE)
1 [
16,
17].
The first study, published by Bach [
16], showed significant IPSS improvement with PSE versus placebo, and its 12-month follow-up period was recognized as an exception in clinical research on phytotherapy for LUTS/BPH [
6].
The second study (GRANU study) was a three-arm trial testing the efficacy and safety of pumpkin seed (open study arm) and PSE capsules against matching placebo. This study showed no difference between PSE and placebo [
17].
Nonetheless, these two placebo-controlled studies have established the long-term safety of the extract; less than 1% of more than 700 patients treated with PSE reported adverse events with a possible causal relationship to treatment [
16,
17]. The reactions were nonserious, and most of them were gastrointestinal complaints. The types and frequency of all recorded adverse events in the PSE group were similar to those in the placebo group. Serum PSA levels and other laboratory safety parameters showed no relevant changes from baseline after 12 months of treatment [
16,
17]. An observational study confirmed the excellent safety profile of PSE in a large population of 2245 men with LUTS/BPH. After 3 months of treatment, the patients reported symptomatic relief and improved quality of life. Simultaneously, only mild gastrointestinal complaints were reported in no more than 4% of the patients [
18].
In terms of efficacy, we performed a meta-analysis to estimate the benefits of PSE compared to placebo using the data from the original reports of the Bach and GRANU study. Since we used the individual patient listings, the meta-analysis also involved the reanalysis of the initial studies enabling us to provide supplementary information about the individual studies. In addition, we aimed to explore the possible influences of covariates on treatment response using statistical models.
Discussion
In everyday clinical practice, phytotherapy is part of standard therapy for men with uncomplicated but bothersome LUTS suggestive of BPH. The favorable safety profile of herbal extracts is well recognized. However, the efficacy evidence is mainly based on highly heterogeneous studies, and meta-analyses should be interpreted with caution [
4,
6,
7].
The present meta-analysis on the efficacy of soft extract from Uromedic® pumpkin seed (PSE) pooled two studies with basically homogenous designs. Both the Bach and GRANU studies were placebo-controlled and had the same follow-up period of 12 months.
Consistent with the primary studies, an IPSS reduction of ≥5 points from baseline was the threshold of response in the pooled analysis. Thus, responders had an improvement that exceeded the minimum 3-point improvement a patient needs to perceive a clinical benefit [
21‐
23].
The analyzed population consisted of 1389 men aged 64 years on average. At baseline, the patients had mean IPSS and IPSS QoL scores of 16.6 and 3.4 points, respectively. This level of moderate LUTS and related impact on quality of life is typical for patients who seek medical advice and treatment for symptomatic relief [
2].
In these patients, the meta-analysis results suggest beneficial effects of PSE in terms of symptomatic relief. After 12 months, the response rate in the PSE group was 54% (ITT-LOCF) compared to 51% in the placebo group. An advantage of PSE treatment was also observed in the PP set, with response rates of 58% and 53% in the PSE and placebo groups, respectively. Additionally, the mean IPSS change from baseline was greater with PSE (− 5.3 points) than with placebo (− 4.8 points).
The magnitude of symptomatic relief achieved with PSE after 12 months compares favorably with the findings by Debruyne et al. [
25] for hexanic saw palmetto extract and tamsulosin. With these treatments, only 49% of the patients reached an IPSS decrease of ≥5 points, and the mean IPSS change from baseline was − 4.4 points. Notably, based on that study, the European herbal monograph has granted well-established use status for hexanic saw palmetto extract [
26].
In a recent meta-analysis, Russo et al. [
27] found minimal differences in IPSS response between saw palmetto extracts and placebo. However, this analysis was based on studies with immense variability regarding essential features. The small number of patients resulted in significant imbalances in baseline IPSS with up to a 2.9-points difference between the treatment groups in 4 out of the 5 studies.
Another meta-analysis of all available published data for hexanic saw palmetto extract showed a mean IPSS decrease of 5.73 [
2,
15]. However, only 6 of the primary studies were randomized controlled trials, of which only one had a 12-month follow-up period, and none was placebo-controlled [
15,
25].
In contrast, the two PSE studies pooled for the present meta-analysis were equal regarding essential requirements of clinical research in patients with LUTS/BPH, such as placebo control and 12-month follow-up [
13]. As we used patient-level data, our study also repeated the analysis of the individual studies [
28]. All re-evaluation results were consistent with the initially reported study data [
16,
17].
In both studies, the baseline characteristics were well-balanced between treatment groups. However, we noticed possibly important differences between the studies due to the strict selection criteria in the GRANU study conducted approximately a decade after the Bach study [
16]. During this period, ARBs and 5-ARIs had become standard prescription drugs for symptomatic LUTS in Germany [
6,
29]. Thus, the ethics committee tightened the inclusion criteria because treating patients with severe LUTS was no longer accepted in placebo-controlled studies. Consequently, the upper IPSS limit for inclusion was 19 points in the GRANU study [
17].
To some extent, the lower baseline scores in the GRANU study might have reduced the magnitude of IPSS improvement. Notably, the logistic regression and ANCOVA models showed that symptom severity at baseline substantially influenced the IPSS reduction. These findings correspond to other observations, e.g., Roehrborn et al. reported that patients with baseline IPSS above 19 points reached a 3-point reduction more often than patients having lower scores at study start [
23]. Similarly, Debryne et al. [
25] reported a better treatment response in patients with severe symptoms compared to patients with moderate symptoms.
Of the patients studied in the present meta-analysis, approximately two-thirds (924 out of 1389) included in our meta-analysis had been treated in the GRANU study. Thus, a study population with less severe symptoms and no difference between PSE and placebo dominated the present results. Nonetheless, in the pooled analysis of both studies, the estimated response rates in the PSE group exceeded those in the placebo group by 3% and 5% in the ITT and PP sets, respectively. Furthermore, according to the logistic regression model, a patient’s likelihood of achieving a 5-point decrease in the IPSS was 17% higher in the PSE group than in the placebo group. Moreover, regarding the mean IPSS changes from baseline to 12 months the difference between the PSE and placebo groups also favored PSE treatment. In both treatment groups, the decrease in IPSS was accompanied by an improvement in the IPSS-QoL index.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.