The pathophysiology of lower urinary tract symptoms in the ageing male population
BPH describes a proliferative process of both the stromal and epithelial elements of the prostate gland; BPH arises in the periurethral and transition zones of the prostate [ 1]. Histological BPH represents an inescapable phenomena for the ageing male population; ≈90% of men will develop histological evidence of BPH by 80 years of age [ 2].
Although nearly all men develop histological BPH, the degree of prostatic enlargement resulting from the hyperplasia is highly variable. The volume of the prostate is most accurately determined using imaging studies such as ultrasonography, CT and MRI. Oesterling et al. [ 3] recently measured the prostate volumes of 464 men aged 40–80 years selected at random from the Olmsted County (USA) population. There was a statistically significant correlation between age and prostate volume (P<0.001; r2=0.185). The overwhelming majority of men in this age range had prostate volumes of 20–60 mL. Although prostate volume was age-dependent, there was a substantial overlap between the categorical age groups.
The severity of LUTS is best quantified using quantitative symptom indices, including the AUA symptom index and the IPSS [ 4]. The development of LUTS is also associated with advancing age [ 5] and often attributed to BPH. Indeed, until recently the constellation of obstructive and irritative symptoms observed in ageing men was termed ‘prostatism’. The differential diagnosis of LUTS in ageing men includes both urological and non-urological conditions. Neurological conditions such as Parkinsonism, cerebrovascular accident, diabetes mellitus, congestive heart failure, bladder cancer, prostate cancer, urinary tract infection, urethral stricture and bladder neck hypertrophy also cause LUTS. Nevertheless, LUTS in the presence of ‘some’ degree of prostatic enlargement often establishes the diagnosis of BPH.
It has often been assumed that the pathophysiology of LUTS in men is the result of the BOO associated with the prostatic enlargement [ 6]. The observations that the development of prostatic enlargement, BOO and LUTS are age-dependent was interpreted to indicate a causal relationship [ 7]. The relationship among these three factors has been represented by three overlapping circles [ 8]; this schematic diagram suggests that in a significant subset of men, enlargement of the prostate causes BOO and LUTS.
The pathophysiology of BOO in men with BPH has been attributed to both static and dynamic factors [ 9]. The static obstruction arises from the bulk enlargement of the prostate encroaching upon the prostatic urethra and bladder outlet, whereas the dynamic obstruction is related to the tension of prostatic smooth muscle. The medical therapies widely used today for the treatment of BPH are targeted at reducing prostate volume and relaxing prostate smooth muscle tension [ 10, 11]. The clinical data showing the effectiveness of androgen suppression and alpha blockade for the treatment of BPH has validated the hypothesis that the pathophysiology of BPH is caused by BOO.
The observations that the development of prostatic enlargement, BOO and LUTS are age-dependent is insufficient evidence that these events are causally related. The relationship among them is optimally defined by measuring these variables in a group of men selected at random from the community. Girman et al. [ 5] measured prostate volume using TRUS, the peak flow rate (Qmax ) and the AUA symptom score in 466 men aged 40–80 years selected at random from the residents of Olmsted County. The P and r2 values for the pairwise relationships of prostate volume and Qmax , prostate volume and symptom score, and Qmax and symptom score are shown in Table 1. These observations suggest that the size of the prostate is a very weak determinant of symptom severity and BOO, and that BOO is only a minor determinant of symptom severity. Barry et al. [ 12] recently reported the relationships between prostate volume, Qmax and symptom scores in a cohort of men participating in a clinical trial examining different treatment options for BPH. The same three relationships were not clinically or statistically significant ( Table 1). The findings of both Girman et al. [ 5] and Barry et al. [ 12] strongly suggest that the idea that the pathophysiology of LUTS in the ageing man is caused by BOO arising from the enlarged prostate is a gross oversimplification.
Relationships between prostatism (LUTS), bladder outlet obstruction and prostatic enlargement using pairwise comparisons. PV, prostate volume; Qmax
, peak flow rate; SS, symptom score
One of the reasons that may explain the poor correlation between BOO and symptom severity is that Qmax is not a reliable proxy for BOO. There is a weak correlation between Qmax and BOO as measured by synchronous pressure-flow studies. As pressure-flow studies are invasive, these measurements have not been performed in a community population. Several investigators have shown no clinically or statistically significant correlation between the severity of BOO and LUTS [ 13].
Studies comparing LUTS, BOO and prostate size amongst different races provides additional evidence that prostate size is not an important factor leading to the development of LUTS and a low Qmax . The AUA symptom score, Qmax and prostate volume have been measured in community-based surveys of men in the USA (Olmsted County) and Japan (Shimomaki-nura) ( Table 2) [ 14]. The Japanese men had smaller prostates, a higher Qmax and more severe symptoms. The discordance between symptom scores, prostate volumes and Qmax in these two population studies is further evidence that the development of LUTS is not explained by BOO or prostate volume.
The IPSS, Qmax
and prostate volume (PV) in Caucasian (American) and Asian (Japanese) men
The overwhelming majority of men with clinical BPH undergo treatment to relieve symptoms and improve their quality of life. Several quantitative questionnaires have been developed to assess the severity of the LUTS associated with clinical BPH. The AUA symptom score has been validated and is the most widely used instrument to measure symptom severity in clinical practice and research [ 12]. Lepor and Machi [ 15] administered the AUA symptom score to a group of 101 men and 99 women aged 55–79 years attending a general health symposium with no emphasis on genitourinary diseases. The mean AUA symptom score was equivalent in the men (6.9) and women (7.7). The AUA score was also grouped into obstructive and irritative symptoms. The obstructive symptom score in the men (2.7) and women (2.9), and the irritative scores (4.2 and 4.8, respectively) were also not significantly different. The development of LUTS characteristic of BPH appears not to be a gender-specific event associated with ageing. It is conceivable that the pathophysiology of LUTS in men and women is different. Nevertheless, the development of LUTS characteristic of clinical BPH in both men and women suggests that there are important nonprostatic mechanisms for the symptoms.
In summary, there is no clinically significant relationship between prostatic enlargement and clinical BPH; there is a weak relationship between LUTS and BOO. Therefore, factors other than prostatic enlargement must account for the severity of clinical BPH. Factors including BOO account for the severity of LUTS.
Prostatic factors contributing to the pathophysiology of clinical BPH
Despite the observation that both men and women develop similar LUTS, the effectiveness of TURP strongly suggests that the prostate and/or prostatic urethra must play an important role in the pathophysiology of clinical BPH. Caine et al. [ 16] reported in 1976 that alpha blockers are effective for the treatment of BPH. Over the last 20 years, over 20 randomized clinical trials have consistently shown the safety and efficacy of various α1-blockers for the treatment of BPH [ 11]. α1-receptors are abundant in the human prostate and mediate the tension of prostate smooth muscle. It has been assumed that the efficacy of α1-blockers is mediated via relaxation of prostatic smooth muscle. The observation by Shapiro et al. [ 17] that 40% of the area density of BPH tissue is smooth muscle provides further evidence that prostatic smooth muscle is an important factor in the development of clinical BPH.
The most appropriate study design to elucidate the pathophysiology of clinical BPH is to compare biochemical and histological measurements in tissue specimens derived from age-matched men with and without clinical BPH. Ideally, the specimens should be derived from men with prostates of equivalent size. Several investigators have compared tissue specimens derived from the outer and inner regions of the prostate, these zones corresponding to the surgical capsule and the hyperplastic tissue, respectively. These comparisons do not provide insights into the pathophysiology of clinical BPH.
Between 1986 and 1989, we investigated the pathophysiology of clinical BPH, using inner gland tissue specimens derived from men undergoing TURP for clinical BPH (symptomatic BPH) and men undergoing cystoprostatectomy for bladder cancer (asymptomatic BPH). Symptom scores, Qmax and prostate volumes were routinely measured before surgery. The α1-receptor density was equivalent in tissue specimens obtained from men with symptomatic and asymptomatic BPH ( Table 3) [ 18]. The contractile response to α1-agonists was also similar between these groups ( Table 3) [ 8]. These studies suggested that the development of clinical BPH was not due to upregulation of the α1-receptor or increased responsiveness of prostatic smooth muscle to α1-agonists. While other investigators have reported that the α1-receptor is upregulated in men with BPH, these studies compared tissues from different regions of the prostate and not inner gland tissue from men with and without clinical BPH. The stromal to epithelial ratio was greater in the men with symptomatic BPH, suggesting that the cellular composition of the inner gland (transition zone) may represent an important factor contributing to the pathophysiology of clinical BPH ( Table 3) [ 19]. As the neurotransmitter for the α1-receptor is noradrenaline, another plausible mechanism contributing to the pathophysiology of clinical BPH is increased adrenergic innervation. Lepor et al. [ 20] reported the relationship between the AUA symptom score and catecholamine level in consecutive men undergoing prostatic biopsy for an elevated PSA or abnormal DRE who had no evidence of prostate cancer. There was an inverse relationship between the AUA symptom score and catecholamine level in the transition zone specimens of the men with no evidence of prostate cancer. The study strongly suggested that the pathophysiology of clinical BPH was not due to increased adrenergic innervation. In summary, the studies from our laboratory identified the cellular composition of the prostate as the only variable contributing to the pathophysiology of clinical BPH.
A comparison of prostatic αl
adrenoceptor (AR) density, phenylephrine responsiveness and cellular composition in men with symptomatic and asymptomatic BPH. Emax
, maximal tension; S/E= stromal to epithelial ratio
To further examine the role of prostatic smooth muscle in the pathophysiology of clinical BPH, 26 men with clinical BPH who were candidates for medical management completed the Boyarsky symptom score and underwent uroflowmetry and TRUS-guided biopsy of the prostate before initiating therapy with the α1-blocker terazosin [ 21]. The mean percentage smooth muscle was quantified from the biopsy specimens. The pairwise relationships between baseline Qmax , percentage smooth muscle, baseline total symptom score, the percentage change in Qmax and percentage change in the symptom score are shown in Table 4. The pairwise relationships show a statistically and clinically significant relationship between the baseline Qmax and the percentage smooth muscle, and no significant relationship between the baseline total symptom score and percentage smooth muscle. These results suggest that the amount of prostatic smooth muscle contributes to BOO and not symptoms; they provide further evidence that LUTS and BOO are not causally related. The relationship between the increase in Qmax and the percentage smooth muscle was highly significant, suggesting that the improvement in BOO secondary to treatment with terazosin is intimately related to the relaxation of prostatic smooth muscle. There was a very weak and statistically insignificant relationship between the percentage change in the total symptom score and the percentage smooth muscle, suggesting that the symptom improvement associated with terazosin may not be mediated via relaxation of prostate smooth muscle. An important implication is that other α1-mediated mechanisms may account for the symptom improvement elicited by terazosin in men with BPH.
The role of prostatic smooth muscle (SM) in the pathophysiology of BPH. SS, symptom score
In summary, prostate smooth muscle density contributes to the severity of BOO and accounts for the α1-blockade-mediated reduction of BOO in men with clinical BPH. Prostatic smooth muscle density does not contribute to the severity of LUTS or the α1-mediated improvement in symptoms in men with clinical BPH.
The pathophysiology of clinical BPH: clinical correlations
If the pathophysiology of LUTS is caused by BOO resulting from enlargement of the prostate, then the decrease in symptom scores in men undergoing treatment for BPH should be directly proportional to the increase in Qmax and the decrease in prostate volume. The overwhelming clinical evidence suggests that the decreases in symptom scores are not proportional to increases in Qmax .
Lepor and Rigaud [ 22] reported the treatment outcome for 30 men with clinical BPH undergoing TURP. Overall, 87% of the patients experienced marked or moderate improvement in their symptoms. In this prospective study, the relationship between changes in Qmax and changes in the obstructive symptom score was not statistically significant (P=0.49; r2=0.25). Similarly, changes in Qmax and changes in the irritative symptom score were not statistically significant (P=0.9; r2=0.085). Schaeffer reported that the symptom improvement after prostatectomy was equivalent in men with and without pressure-flow evidence of BOO.
An analysis of the Veterans Administration (VA) Cooperative Study on the Medical Management of BPH provides insights related to the pathophysiology of clinical BPH [ 23]. In this randomized, placebo-controlled clinical trial, 1229 men with clinical BPH at 31 VA Medical Centers received placebo, terazosin, finasteride, or combination therapy. This represents the first comparison of an α1-blocker and 5-α reductase inhibitor, and the first study to examine the combination of both drugs. Over the entire 52 weeks of the randomized study, the changes in Qmax and AUA symptom score were not significantly different between placebo and finasteride. The changes in Qmax and AUA symptom score were very significantly different between terazosin and placebo, and combination therapy and placebo. The changes in the AUA symptom score and Qmax between terazosin and combination therapy were not statistically or clinically significant. The equivalent effectiveness of placebo and finasteride, and terazosin and combination therapy is strong evidence that 5-α reductase inhibitors at best are of extremely limited value in the medical management of BPH. A subset analysis of the VA study showed a small difference between the changes in Qmax and AUA symptom scores in men receiving finasteride rather than placebo in men with large prostates (personal communication). In men with prostate volumes >50 mL, the treatment-related improvement in AUA symptom score attributed to terazosin was three times that with finasteride.
The proposed mechanism of action of finasteride in men with BPH is a reduction of prostate volume. The P and r2 values for the pairwise relationships between the change in prostate volume and change in Qmax were 0.097 and −0.106, respectively, and for the change in prostate volume and change in AUA symptom score were 0.068 and 0.116 for the 251 men in the VA study on finasteride; both these pairwise relationships approached statistical significance. The mechanism for the minimal efficacy associated with finasteride appears to be weakly determined by a reduction of prostate volume.
If the mechanism for the improvement in symptoms is related to alleviating BOO, then there should be a significant relationship between changes in the AUA symptom score and changes in Qmax . The P andr2 values are shown for the relationship between the changes in AUA symptom score and changes in Qmax for the four treatment groups in the VA study ( Table 5). There is a significant relationship only between the changes in the AUA symptom score and Qmax in the finasteride and combination groups; the r2 values in these groups suggest that this relationship is weak and that factors other than alleviating BOO contribute to symptom improvement.
Pairwise comparisons in the VA study between changes in AUA symptom score and Qmax
Interestingly, there was no significant relationship between changes in the AUA symptom score and Qmax in the terazosin group. If the improvement in symptoms after administering terazosin is not related to Qmax , then excluding men with LUTS and no evidence of BOO may not be justified. A subset analysis of the VA study examined the symptom improvement according to the quartile groups of baseline Qmax . Interestingly, the improvement in the AUA symptom score was equivalent in the lowest (<8.6 mL/s) and the highest quartile baseline Qmax groups (>12.4 mL/s), suggesting that the α1-mediated symptom improvement may not be related to relaxing prostate smooth muscle. If this is the case, men with LUTS and no evidence of BOO should respond to α1-blockers. Lepor et al. [ 24] recently reported that symptom improvement in age-matched men with prostate volumes of equivalent size were equivalent in men with a normal and abnormal baseline Qmax .
In summary, the reduction of prostate volume at best accounts for only a small portion of the minimal effect of finasteride on LUTS. Alleviation of BOO does not account for the improvement in LUTS associated with α blockers. The available clinical data suggests that the improvement in symptoms elicited by α1-blockade and hormonal therapy are not mediated primarily by relax- ation of prostatic smooth muscle and a reduction in prostate volume, respectively.
What is the pathophysiology of clinical BPH? The development of BOO and LUTS is associated with ageing; the overwhelming clinical evidence suggests that these age-dependent variables are not causally related. Undoubtedly, there are some men whose prostatic enlargement causes obstruction and symptoms. Based upon the available data, this subset must be extremely small.
We know that TURP results in highly significant improvements in LUTS. Obviously, the mechanism of symptom improvement in these cases must be related to the prostate, bladder neck or prostatic urethra, as only these tissues are resected. As men without BOO respond equally well to prostatectomy, a reasonable mechanism for the pathophysiology of BPH and the effectiveness of TURP may reside in unobstructive factors involving neurological pathways.
We also know that α1-blockers represent an extremely effective pharmacological strategy for the treatment of BPH. There is increasing evidence that men without BOO respond to α1-blockers. The mechanism for α1-mediated symptom improvement appears to be independent of BOO. It is plausible that the sensory innervation of the prostate may represent a target for the α1-antagonists.
Because there are many urological and non-urological conditions that cause LUTS, and there are age-dependent changes in bladder and neurological function, it is unlikely that there is a single dominant aetiology for LUTS in ageing men. If this is the case, then the optimal management of LUTS will require different and possibly combined therapies. The available results suggest that we should not direct our pharmacological strategies exclusively towards reducing prostate volume or diminishing BOO. Only incremental advances in the medical management of BPH will be derived by developing subtype-selective α1-antagonists or more complete inhibitors of 5α reductase. These advances in the medical management of BPH will require a better understanding of the pathophysiology of LUTS.