Dr Cees Korstanje Pharm D PhD, Translational & Development Pharmacology Department, Astellas Pharma Europe BV, Elisabethhof 1, 2350 AC Leiderdorp, The Netherlands. Tel.: +317 1545 5510 Fax: +317 1545 5276 E-mail: firstname.lastname@example.org
WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT
• The efficacy-tolerability profile of tamsulosin in patients with benign prostatic hyperplasia (BPH) is assumed to be associated both with the α1-adrenoceptor selectivity profile of the drug and a small peak : trough ratio in the plasma pharmacokinetic (PK) profile. Tamsulosin is highly bound to plasma proteins, notably α1-acid glycoprotein (AGP). This protein is a high-affinity binding protein and AGP plasma concentration was found to influence the therapeutic (unbound) plasma concentrations for high-AGP-binding drugs.
WHAT THIS STUDY ADDS
• The study actually assessed unbound tamsulosin concentrations in both blood plasma and prostate tissue and reported that the unbound tamsulosin concentrations after multiple dosing in men with BPH, were much higher in prostate than in blood plasma. The assumption is put forward that differential free drug concentrations in prostate and blood plasma may contribute to the relative ‘uroselectivity’ of tamsulosin.
AIM The aim of this small patient study was to investigate tamsulosin concentrations in prostate and plasma samples in order to identify potential differences in the pharmacokinetics (PK) in plasma and prostate contributing to its pharmacodynamic activity profile in patients.
METHODS Forty-one patients with benign prostatic hyperplasia (BPH) scheduled for open prostatectomy were given tamsulosin 0.4 mg for 6–21 days in order to reach steady-state PK. Patients were randomized over four groups to allow collection of plasma and tissue samples at different time points after last dose administration. Samples were collected during surgery and assayed for tamsulosin HCl. The free fraction (fu) of tamsulosin was determined by ultracentrifugation of plasma and prostate tissue spiked with 14C-tamsulosin.
RESULTSCmax in plasma at 4.4 h for total tamsulosin was 15.2 ng ml−1 and AUC(0,24 h) was 282 ng ml−1 h, while for prostate Cmax at 11.4 h post-dose was 5.4 ng ml−1 and AUC(0,24 h) was 120 ng ml−1 h. AUC(0,24 h) for total tamsulosin in prostate was 43% of the plasma AUC(0,24 h). fu was 0.4 % for plasma and 59.1% for prostate. Therefore calculated on unbound tamsulosin, a ratio of 63 resulted for prostate vs. plasma Cmax concentrations.
CONCLUSIONS These data indicate that in patients with confirmed BPH the amount of tamsulosin freely available in the target tissue (prostate) is much higher than in plasma.
Tamsulosin HCl 0.4 mg is the most widely used drug in the treatment of lower urinary tract symptoms associated with benign prostate hyperplasia (BPH). The drug has a relatively low propensity for side effects, which has been well established in randomized placebo-controlled clinical trials . The mechanism for the selectivity of the drug for urinary tract-related symptoms was proposed to be associated with the α1-adrenoceptor selectivity profile of the drug .
Potential selectivity of drugs for prostatic α1-adrenoceptors has been the subject of evaluation in several studies [3–6]. Alfuzosin, 5-methyl-urapidil and Rec 15/2739 were claimed to lower more pronouncedly prostatic tone than blood pressure in vivo, but conclusions have not been consistent between studies. In a few studies an attempt was made to correlate selectivity for the prostate in vivo (based on animal data) with selectivity for any of the subtypes in vitro. The published data indicate that compounds having no selectivity for any of the α1-adrenoceptor subtypes in vitro exhibit a balanced profile in vivo towards prostatic tone and blood pressure in animal models .
It is well established that tamsulosin is relatively selective for α1A and D-adrenoceptors over α1B-adrenoceptors [8, 9]. What this means in terms of mode of action for α1-adrenoceptor blockers in BPH is not completely clear. Discussions and speculations on the relationship between distribution of α1-adrenoceptor subtypes and its consequences for treatment of lower urinary tract symptoms including BPH [10–12] have raised the possibility of prostatic, spinal cord and bladder α1-adrenoceptors as potential contributors to the symptoms of BPH that may have their origin, or association in any of these compartments. Supposing this is relevant for treatment of BPH with α1-adrenoceptor blocking drugs, it is relevant to match the α1-adrenoceptor subsite selectivity of the antagonist drug with the relative expression of α1-adrenoceptor subtypes at organ level.
At the physiological level, it is clear that different organs/physiological functions have a differential α1-adrenoceptor subtype expression and functional dominance . This is well-documented for the human prostate, where α1A-adrenoceptors are functionally dominant . For human and rat bladder it is assumed that α1D- and possibly α1A-adrenoceptors play a functional role , whereas at the ganglionic and central regulatory level α1A and D-adrenoceptors predominate at the mRNA expression level .
The available information regarding subtypes of α1-adrenoceptors involved in blood pressure regulation in humans is very scarce, and the review by Guimarães & Moura  mentions only single studies in this respect. The impression is that there seem to be important species differences, e.g. in the rat α1D-adrenoceptors are most abundant, while in the mouse α1B-adrenoceptors are dominant , and in man α1A and B-adrenoceptors are more abundant than the α1D type, and reported to change to an α1B- type dominance with increasing age .
Whether α1-adrenoceptor subtype selectivity on its own can explain tamsulosin's excellent side effect profile is questionable. In a clinical pharmacology study, tamsulosin (given as 0.4 mg modified release capsules) showed less inhibition of vasoconstriction elicited by the α1-adrenoceptor agonist phenylephrine, than the non-selective α1-adrenoceptor blocker terazosin (given as 5 mg tablets) , while only moderate differences in in situα1A-adrenoceptor site binding in plasma samples were found for tamsulosin and terazosin-treated subjects . This suggests that additional mechanisms might contribute to the clinical selectivity of tamsulosin in BPH.
In animal studies tamsulosin showed a high ‘uroselectivity’ in the dog [6, 7, 22], while this was much lower in the rat . In dogs, for a dose of tamsulosin which caused pharmacological effects (modulation of intra-urethral pressure curves), the area under the concentration–time curve (AUC) values for prostate were about six-fold higher than for plasma, and about two-fold higher in bladder than in plasma . By contrast, in the rat concentrations of tamsulosin in prostate and plasma at steady-state pharmacokinetics were shown to be only marginally different .
This prompted us to investigate tissue concentrations in those compartments mentioned earlier that are potentially open for studies in patients. Prostate, bladder and plasma from patients with BPH were found to be disposable for this investigation. We aimed to determine the free fraction of tamsulosin in plasma and tissue for the following reasons:
1Clear differences in plasma protein binding for tamsulosin have been reported for rats, dogs and humans (about 99% for humans, ±90% for dogs, and ±79% for rats ), and the plasma binding protein for tamsulosin is α1-acid glycoprotein (AGP), which is a high affinity binding protein shown to decrease the bioavailable concentration of bound drugs [26, 27].
2AGP is a ‘heat shock protein’, which can be produced in the liver as well as in the prostate. Thus there is a possibility that AGP concentrations could be variable between plasma and prostate .
Hence, differences in AGP-drug binding can be speculated to impose different free concentrations and thereby possibly different activity for tamsulosin in different body/pharmacokinetic compartments.
We have therefore investigated prostate and bladder tissue obtained from patients who were prepared to take tamsulosin 0.4 mg capsules1 during the time awaiting surgery for open prostatectomy. The study was planned to have patients in steady-state pharmacokinetics for tamsulosin at hospitalization. Patients were allocated to groups with different time intervals from the last capsule taken until the time of surgery. Thereby groups of subjects were obtained with a reasonable spread over the estimated concentration vs. time curves for tamsulosin in plasma, bladder and prostate to construct pharmacokinetic curves for these compartments using population pharmacokinetic techniques and calculate concentration ratios for prostate/plasma and bladder/plasma based on free drug concentrations (fu) in each of these compartments.
The study protocol was approved by the Ethics Committees of each participating hospital (Semmelweis Medical University, Bajcsy Hospital, Ferenc Hospital and Péterfy Hospital, Budapest and Aladár Hospital, Györ, in Hungary). The study was conducted in accordance with the rules of Good Clinical Practise and written informed consent was given by all patients included.
This study had an open label, multicentre, randomized design in a group of 41 ambulatory male subjects with BPH on a waiting list for open prostatectomy.
All eligible subjects were treated with 0.4 mg tamsulosin (Omnic®/Flomax® modified release capsules) once daily, for 6–21 days, prior to open prostatectomy surgery via the suprapubic operation method. At visit 1, eligible subjects were randomized to group 1, 2, 3 or 4, determining the exact date and time of the open prostatectomy 6 to 21 days later. The subjects were hospitalized at visit 2A (day –1), and the final tamsulosin capsule was taken under fasted conditions (not under fed conditions because of anaesthesia) 1 day later at visit 2B (day 0). Prostate surgery was performed 4, 8, 24 or 48 h after last capsule intake, for groups 1–4, respectively.
During the open prostatectomy (visit 3, at day 0, 1, or 2) bladder and prostate tissue samples were taken, and a total of three blood samples were taken at predetermined time points from each subject to determine tamsulosin concentrations in tissue and plasma. The subjects were considered completing the study when they were discharged from the hospital after surgery, approximately 6–14 days after visit 2B.
Thirty-six patients completed the study. They were males, aged 69 ± 8 years (range 52 to 88 years), who had symptomatic BPH with a prostate weight 86 ± 40 g (range 40 to 200 g), diagnosed by digital rectal examination. Their PSA was 12.4 ± 19.7 µg l−1 (range 1.4 to 103.1). For those with 4 µg l−1 < PSA ≤ 10 µg l−1, prostate cancer was excluded by further investigation at the discretion of the investigator; when PSA > 10 µg l−1, prostate cancer was excluded by means of biopsy.
Patients had no neurogenic voiding of any aetiology, urethral stenosis, bladder neck local disease, history of acute or chronic relapsing prostatitis, active urinary tract infection, previous endoscopic prostate resection, pelvis radiation therapy or allergy to previously prescribed α-adrenoceptor blockers. Also patients having received medication for BPH (α-adrenoceptor blockers, finasteride or phytotherapy), within 3 months prior to study initiation, were excluded, as well as patients on other investigational drugs, those having prostate cancer, renal or hepatic insufficiency, significant cardiovascular disease, or mental disease, and those with clinically significant aberrant values for biochemical and haematological laboratory parameters, or those with unacceptable surgical and anaesthetic risk factors.
At hospitalization, patients had been treated with the drug for 6–18 days to obtain a steady state; the mean number of capsules taken per patient was 8.9 ± 3.8 days (range 6–18).
Plasma and tissue samples
Blood samples were taken before, during and after open prostatectomy surgery and bladder and prostate tissue samples during prostatectomy surgery at the times indicated in Table 1. Plasma samples were obtained after centrifugation of about 2 ml Li/heparin blood samples. Tissue samples (prostate specimens of ±5 g; bladder dome full-thickness tissue (±0.7 × 2 cm), were collected without rinsing and frozen immediately at −20°C. The samples were kept at −20°C until shipment, kept on dry ice during shipment to the Bioanalytical Laboratory of Astellas Pharma Europe BV in Leiderdorp, NL and kept at −70°C until analysis. The stability of tamsulosin in these samples was checked and found adequate for these experiments.
Table 1. Group randomization with blood sampling and surgery schedule
Blood samples (sequentially numbered)
Bladder and prostate
t = −0.5 h
t = 4 h
t = 28 h
t = 4 h
t = −0.5 h
t = 8 h
t = 34 h
t = 8 h
t = −0.5 h
t = 6 h
t = 24 h
t = 24 h
t = −0.5 h
t = 24 h
t = 48 h
t = 48 h
Bio analysis and ex vivo drug binding
Bioanalysis of plasma samples for total tamsulosin hydrochloride concentration was done according to methods used earlier , but detection was performed using a triple stage quadrupole mass spectrometer (Thermo Finnigan Surveyor and Thermo Finnigan TSQ 7000). The method was found suitable for quantification of total plasma tamsulosin hydrochloride in the range of 0.1–50 ng ml−1 (LLOQ = 0.1 ng ml−1).
Unbound tamsulosin was determined in pooled samples, after ultrafiltration using filters with a 30 kD cut-off. Further sample treatment and analytical equipment was similar to that for the total plasma tamsulosin assay.
Ex vivo drug binding was investigated in human prostates from those samples that were confirmed to have a tamsulosin concentration below detection limit, as well as in (non-matching) blank human plasma provided by Charterhouse Clinical Research Unit Ltd. (Ravenscourt Park Hospital, London, UK). Prostate tissue was diluted with phosphate buffered saline (pH 7.4) and homogenized using a Potter tube. Plasma was diluted 0, 2, 4, 6, 10, 20 and 40 times and prostate homogenate was diluted 0, 1.5, 2.5, 5 and 10 times to obtain a total dilution of 4, 6, 10, 20 and 40 times. Both plasma and prostate homogenate were spiked with 221 ng ml−114C-labelled tamsulosin (specific activity 4.18 MBq mg−1; radiochemical purity more than 99%, Amersham International plc, Buckinghamshire, UK). To achieve equilibrium between bound and unbound tamsulosin the samples were incubated for 120 min in a 37°C water bath. After incubation 50 µl of the samples were transferred into a scintillation vial for determination of the total concentration of 14C-tamsulosin. After centrifugation for 10 min at 3000 rev min−1, 200–500 µl plasma or prostate tissue was brought onto a Amicon MPS Micropartition device, in which filters with a 10kD cut-off were used (Millipore, Billerica, MA, USA). After the micropartition device was centrifuged for 10 min at 3000 rev min−1, the volume of ultrafiltrate was determined and transferred into a scintillation vial. All scintillation vials were filled up with 4 ml of Ultima Gold scintillation fluid (Perkin Elmer Life and Analytical Sciences; Boston, MA, USA) and subsequently counted in a scintillation counter (Tri Carb 2900TR, Packard Instruments Co., Meride, CT, USA) for 30 min. The relationship between the dilution factor and the data of the prostate or plasma free fraction was fitted using a sigmoid dose–response curve with a variable slope. Values of bottom and top were fixed between 0 and 100. Individual fitting of the prostate data was done with Graphpad Prism version 4 (Graphpad software Inc., San Diego, USA). For the population fit of the plasma and prostate data nonmem version V.1 (University of California, San Francisco, USA) was used.
Plasma and tissue concentration data were subjected to population pharmacokinetic analysis using nonmem version V.1 (University of California, San Francisco). Plasma concentration data were fitted using a one-compartment model with first-order absorption. To model tissue concentrations, it was assumed that a negligible amount (arbitrarily set to 1%) of drug distributed between plasma and tissue by a first-order process. Individual values of the pharmacokinetic parameters were assumed to follow a lognormal distribution. Residual error was modelled as a combination of concentration proportional and additive error. Modelling was performed using the First Order Conditional Estimate with Interaction (FOCEI) method.
Medication was administered once daily after breakfast to patients for a mean of 9.7, 7.8, 7.6 and 10.9 days in groups 1, 2, 3 and 4, respectively. The t1/2 for tamsulosin is approximately 18 h, so a mean time of 7.6–10.9 days of administration is considered sufficient to reach steady-state concentrations. On the day of the prostatectomy the last dose was taken in the fasted state.
During intake of tamsulosin HCl 0.4 mg once daily treatment, five patients in group 1 (62.5%), nine patients in group 2 (90.0%), 10 patients in group 3 (83.3%) and eight patients in group 4 (72.7%) reported at least one treatment emergent adverse event (TEAE). The majority of the TEAEs were of mild to moderate intensity. Three patients experienced an AE of severe intensity: two patients in group 2 (pain and anaemia) and one patient in group 3 (pain). The most common TEAEs were pain and hypotension. All TEAEs were assessed by the investigator as not related to the study drug, except for one patient in group 2 who experienced a peptic ulcer that was considered possibly drug-related by the investigator. The population values of the pharmacokinetic model parameters and their respective inter-subject variability (expressed as CV) are summarized in Table 2.
Table 2. Population values of the pharmacokinetic model parameters and their respective inter-subject variability (expressed as CV)
CL/F (l h−1)
Derived pharmacokinetic parameters were: tmax, Cmax, AUC(0,24 h), t1/2 in plasma and tmax, Cmax and AUC(0,24 h) in prostate tissue. Values are given in Table 3.
Table 3. Pharmacokinetic parameters given as summary of the mean (SD) for tamsulosin in plasma
Cmax (ng ml−1)
AUC(0,24 h) (ng ml−1 h)
Tamsulosin was gradually absorbed reaching a Cmax of 15 ng ml−1 at 4.4 h. After Cmax was reached, the plasma concentration declined slowly with a t1/2 of 18.2 h. (Figure 1A) Distribution to prostate tissue proceeded gradually and Cmax was reached after 11.4 h. (Figure 1B). The AUC(0,24 h) for total tamsulosin HCl in prostate tissue amounted to 43% of the plasma AUC(0,24 h) and, consequently, the average concentration of total tamsulosin HCl in prostate tissue over the day amounted to 43% of the average concentration in plasma.
Tissue binding experiments in prostate showed increases in free fraction with increasing sample dilution (Figure 2). Data could be fitted into sigmoid curves. For both individual and population analysis a sigmoid dose–response relationship with a variable slope that was fitted to the individual data points was used to extrapolate to the undiluted value of the prostate tissue binding. The fu for prostate tissue was found to be 59.1 ± 7.3% (range 51.6–66.6%, n = 4).
Plasma protein binding studies for tamsulosin in pooled samples revealed the protein binding to be 99.6 ± 0.5% (range 98.5–99.91%, n = 8), so consequently fu = 0.4%. This means that protein binding was much stronger in plasma compared with binding to prostate tissue. Pop-PK figures for unbound tamsulosin HCl in both plasma and prostate are depicted in Figure 3A and B. Taking these differences into account the AUC(0,24 h) of unbound tamsulosin in prostate tissue was estimated to be 63-fold higher than the AUC(0,24 h) in plasma. For bladder wall, too few samples were above the limit of quantification to allow meaningful modelling.
The plasma concentration–time curve for total tamsulosin concentration under steady state obtained via population pharmacokinetics was consistent with the pharmacokinetics for tamsulosin HCl 0.4 mg modified release capsules in the elderly reported earlier . In addition, the high plasma protein binding of tamsulosin  was confirmed.
The central finding in this study is that under steady state conditions for tamsulosin dosing in patients, considerably higher unbound drug concentration is found in prostate tissue than in plasma. This is an unexpected finding, since it is assumed that under equilibrium conditions diffusion of unbound drug will lead to equal drug concentrations in all compartments in the body . Appropriate methods have been used to assess the unbound drug concentration in prostate, and corrections have been made for dilution effects that play a role in bioanalysis of tissue samples. The values reported in this small study are therefore the best approximation of the unbound drug concentration in these patients' prostates.
Since the pharmacological activity of a drug directly relates to the unbound concentration, we speculate that this may be why the drug apparently blocks an adequate number of α1A-adrenoceptors in the prostate to display clinical efficacy in BPH, while the free drug concentration in the plasma is thus limited2[30, 31]. The finding of higher free drug ratio at any time after dosing in prostate over blood plasma is a property of the drug tamsulosin HCl and absolute values of free drug in these compartments may be altered by the formulation used. In this respect it is noteworthy that tamsulosin 0.4 mg in an oral controlled absorption system (Omnic/Flomax 0.4 mg OCAS) with lower (total) plasma AUC and Cmax concentrations for tamsulosin HCl has equal clinical efficacy, but has a lower propensity to cause orthostatic symptoms in the elderly than tamsulosin HCl 0.4 mg capsules .
Uroselectivity, as it pertains to α1-adrenoceptor antagonists used in BPH, has been defined in different ways in the literature [33–35]:
Pharmacological uroselectivity may be defined in terms of α1-adrenoceptor subsite selectivity of drugs that preferentially bind to α1A and α1D -adrenergic receptors that are expressed in and regulate the contraction of smooth muscle of the prostate, bladder base and neck, and urethra versus those that regulate vascular smooth muscle contraction and other functions [11, 34].
Clinical uroselectivity has been defined as the relationship of desired clinical effects on obstruction and lower urinary tract symptoms vs. unwanted side effects . Hence, a clinically uroselective α1-adrenoceptor blocker may improve urinary flow rate and BPH symptoms while minimizing unwanted side effects such as dizziness and orthostatic hypotension.
Functional/Physiological uroselectivity of an α1-adrenoceptor blocker refers to a preferential effect on urethral pressure vs. arterial blood pressure, by non-availability of human data established in animal models [34, 35].
Finally, this paper provides arguments to propose a fourth level of selectivity, tissue/drug-partitioning selectivity, that is, how a drug is distributed in target tissues (i.e. prostate, bladder neck and base, prostatic urethra, ganglia) vs. non-target tissues (i.e. blood vessels). An uroselective α1-adrenoceptor blocker may be defined as one that distributes preferentially in the target tissues of the urinary tract vs. other tissues. Our data show that for tamsulosin this fourth mechanism is likely to apply, with differences in local concentration of AGP as the driving force for the higher free drug fraction in prostate than in plasma. For α1-adrenoceptor blockers this is most likely a mechanism that is unique to tamsulosin. To our knowledge, for only one other α1-adrenoceptor blocker, alfuzosin, are published data available on plasma and prostate concentrations of the drug in animals and in man [36, 37]. The study in rats indicated four- and nine-fold higher concentrations of alfuzosin in prostatic tissue compared with plasma at 1 and 6 h after oral drug dosing and the study in humans showed a 2.4-fold difference in plasma and prostate concentations of the drug 12 h after dosing of the last tablet. However, in both studies no information was obtained on the PK profile of the drug in prostate, so it is unclear what this ratio means. Alfuzosin is not highly bound to plasma proteins and not in particular to AGP, making a mechanism as described for tamsulosin in our study unlikely.
Understanding mechanisms of uroselectivity for α1-adrenoceptor blockers in BPH is important to optimize drug therapy. The intrinsic properties of tamsulosin determining its receptor profile, metabolism, and tissue distribution contribute to its uroselectivity. However, a further important factor is the peak : trough ratio imposing on the cardiovascular safety profile for α1-adrenoceptor blockers in BPH. For tamsulosin, reformulation of a capsule dosage form into an oral controlled release matrix formulation (OCAS) further improved the cardiovascular safety profile over that of the capsule [32, 38].
All authors are employees of Astellas Pharma Europe B.V., but have no shares in the company. The study was supported by Astellas Pharma Europe B.V. Omnic®/Flomax® Ocas® is a brand name drug from Astellas Pharma Europe B.V.
The authors wish to acknowledge the urologists, Drs Imre Romics, Tamás Kiss, Lázló Kisbenedek, József Kondás, Ferenc Törzsök, departments of Urology at Semmelweis Medical University, Bajcsy Hospital, Ferenc Hospital and Péterfy Hospital, Budapest, and Aladár Hospital, Györ, Hungary, for their skilful contributions to the study; Dr Maria Milak, Astellas Pharma Hungary, for her excellent organizational contributions, and Dr Piet Swart (now at Novartis Pharma AG, Basel, Switzerland) for his encouraging discussions and guidance on bioanalytical method development.
At the time of study tamsulosin was only available as Omnic/Flomax 0.4 mg modified release capsule.
At the Cmax for tamsulosin in plasma the free drug concentration is about 1.5 × 10−10m, while the dissociation constant in functional studies at the human α1A-adrenoceptor is about 10−10m, and similar at α1D-adrenoceptors, but higher at α1B-adrenoceptors: It is therefore expected that in plasma even at the Cmax the concentration of tamsulosin is not high enough to block an adequate fraction of the α1-adrenoceptors in the blood vessels to blunt appropriate regulation of the blood pressure.