Neurourology and Urodynamics 27:65–70 (2008)

Effect of Lumbar-Epidural Administration of Tramadol onLower Urinary Tract Function

S.K. Singh,1*y M.M. Agarwal,1z Y.K. Batra,2y A.V.K. Kishore,1§ and A.K. Mandal1y1Department of Urology, Postgraduate Institute of Medical Education and Research, Chandigarh, India

2Department of Anaesthesiology, Postgraduate Institute of Medical Education and Research, Chandigarh, India

Aims: Intrathecal and epidural administration of m-agonist opioids is associated with urinary retention, apotentially serious adverse-event. In animal studies tramadol has been found not to affect voiding function. Weevaluated urodynamic effects of epidural tramadol in humans. Methods: Fifteen adults planned for cystoscopyunder local-anesthesia underwent urodynamics (UDS) at baseline and 30 min after administration of 100 mgtramadol in lumbar-epidural space. UDS consisted of filling cystometry, pressure-flow study and pelvic floorelectromyography (EMG). Subsequently, all underwent cystoscopy and were observed for 6 hr. Results: Afterinjection of tramadol, a significant rise was observed in bladder capacity (391.8� 179.6 ml vs. 432.7� 208.8 ml;P¼ 0.019) and compliance (60.1� 51.5 ml/cm H2O vs. 83.0� 63.0 ml/cm H2O; P¼ 0.011) without a significantchange in filling pressure (22.5� 13.2 cm H2O vs. 24.1� 15.1 cm H2O; P¼ 0.576). Filling sensations were delayedsignificantly (P� 0.05). EMG during filling phase showed a significant fall (P¼ 0.027). Peak flow-rate (Qmax),average flow-rate, postvoid residue and detrusor pressure-at-Qmax did not show significant change from baseline(P> 0.05). Three patients had bladder outlet obstruction which did not worsen after the injection. Guarding reflexwas inhibited in seven out of 12 patients who had it at baseline (P¼ 0.016). Conclusions: Epidural tramadolincreases the bladder capacity and compliance and delays filling-sensations, without ill effect on voiding. This seemstrue even for patients with obstructed outflow; however, due to small number of patients a definite conclusioncannot be derived. These results will guide clinician to avoid catheterization in cases where epidural tramadol isused for postoperative pain. The inhibitory effects of tramadol on EMG activity are intriguing and need furtherstudies. Neurourol. Urodynam. 27:65–70, 2008. � 2007 Wiley-Liss, Inc.

Key words: compliance; electromyography; tramadol; urodynamics; voiding dysfunction

INTRODUCTION

Opioid analgesics are one of the cornerstone options forthe treatment of moderate to severe postoperative pain.Behar et al.1 first described the analgesic effects ofepidural morphine in humans. Since then, intrathecal andepidural administration of opioids has become popular in themanagement of both postoperative as well as chronic pain.Use of spinal morphine is mainly limited by side-effectsfor example respiratory depression, itching, nausea, vomitingand urinary retention.2 From a urologist’s perspective,urinary retention is a potentially serious adverse eventwhich is associated with significant discomfort and bladderdysfunction (due to detrusor overstretching). Moreover,catheterization, which may have to be repeated, per se isassociated with significant morbidity in the form of patient-discomfort, urinary tract infections urethral trauma andstricture.3,4

Various studies have proven efficacy of tramadol (systemic,epidural and spinal) in the management of moderate post-operative pain, including that after thoracotomy,5 majorabdominal and pelvic operations,6 and also in neuropathicpain.7 Its side-effect profile is safer than morphine, espe-cially with regards to sedation, respiratory depression anddependence.8 Generally, 50–100 mg has been used forepidural injection. No clinical study is available on its uro-dynamic effects. In certain animal experiments its intrathecalinjection has been found not to cause urinary retention, unlikemorphine.9 The present study was designed to evaluatethe effect of epidural administration of tramadol on bladderurodynamics.

MATERIALS AND METHODS

Fifteen adult patients, seven males and eight females,planned for rigid cystoscopy under local anesthesia for variousindications and not having bothersome lower urinary tractsymptoms were included in the study. Approval to conductthe study was gained from the institutional ethics committee.After explaining the objectives of the study, the proceduresof urodynamics, epidural analgesia and cystoscopy, informedconsent was taken from each patient before inclusion.Indications for cystoscopy included staging procedure incarcinoma cervix in 8, stent removal in 4, evaluation forhematuria in 2 and check-cystoscopy in follow-up case ofbladder tumor in 1. Patients with urinary tract infection,coagulopathy and vertebral deformities precluding lumbarepidural access were excluded from the study.

Tramadol is available for parenteral use on physician’sprescription; the preparation does not contain any preserva-tive. After a baseline urodynamic study (UDS), 100 mgtramadol was injected in lumbar epidural space and a repeaturodynamic study was performed after 30 min. UDS consisted

No conflict of interest reported by the author(s).Karl Erik Andersson led the review process.yProfessor.zAssistant Professor.§Senior Resident.*Correspondence to: S.K. Singh, MS, Mch, House No. 202-A, Sector 24-A,Chandigarh 160 023, India. E-mail: shrawanksingh2002@yahoo.comReceived 23 November 2006; Accepted 25 April 2007Published online 11 July 2007 in Wiley InterScience(www.interscience.wiley.com)DOI 10.1002/nau.20465

� 2007 Wiley-Liss, Inc.

of filling cystometry followed by pressure flow analysis insitting position using ‘Solar Silver’ digital urodynamicmachine (Medical Measurement System, Enschede, TheNetherlands). Pelvic floor EMG was recorded throughout thestudy by perianal surface electrodes using Solar Neuro Module(Medical Measurement System). Initially, the bladder wasemptied using a 12Fr foley catheter. Subsequently, a doublelumen cystometry catheter (6-Fr; Mediplus, Bucks, UK) wasinserted perurethrally for bladder filling and measurement ofvesical pressure using a pressure transducer (Capto SP 844,Memscap AS, Norway). A balloon catheter (5Fr; Medica,Medolia, Italy) was inserted in the rectum to measureabdominal pressure using a similar transducer. Catheterswere ‘zeroed’ to atmospheric pressure keeping the transducersat the level of the superior border of pubic-symphysis. The truedetrusor pressure (Pdet) was computed by software subtractingabdominal pressure (Pabd) from vesical pressure (Pves) on areal-time basis. Sterile normal saline (0.9%, w/v) was used asthe filling medium and infused via the perurethral catheterusing a motor-driven and computer-controlled infusion-pumpat a rate approximately 10% of expected bladder capacity.Volume and pressures at first sensation, first desire, normaldesire, strong desire, maximum cystometric capacity (MCC)and detrusor overactivity (if any) were noted. Detrusorcompliance (DV/DP) was computed in during the tonus limbof the filling phase by linear regression analysis using thesoftware. At the end of the filling phase, the patient was askedto void in a digital gravimetric uroflowmeter (MedicalMeasurement System). Voided volume, residual urine volume,maximum flow rate (Qmax), average flow rate (Qave), Pdet atQmax and maximum Pdet were noted. The Abrams–Griffiths(AG) number defined as (Pdet � Qmax — 2Qmax), detrusor-adjusted mean passive-urethral-resistance-relation factor(DAMPF) and flow-curvature were computed using the soft-ware based on plotting the flow of the InternationalContinence Society nomogram, Shaeffer plot and Chessdetrusor/flow plot, respectively.

Urine production during UDS was calculated both beforeand after the injection of tramadol as follows: Urineproduction ¼ (voided volume þ PVR) � volume filled. ActualMaximum Cystometric capacity was calculated as follows:Actual MCC ¼ {volume filled (or observed MCC) þ urine pro-duction}. Actual MCC was used for statistical analysis inpreference to observed MCC.

Bladder outlet obstruction was defined as an AGnumber > 40 in males and presence of two or more of thefollowing in females PdetQmax > 50 cm H2O, Qmax < 12 ml/sec,urethral resistance (PdetQmax=Q2

max) > 0.2 and significant post-void residue (PVR) in presence of high pressure or resistance.10

A Tuohy graduated epidural needle 18 G � 3.5 in. (PortexLtd., London, UK) was inserted at L3–L4 intervertebral disclevel with the patient in a left lateral position under localinfiltration anesthesia. After confirming the position of theneedle by loss of resistance and negative aspiration for bloodand spinal fluid, 100 mg tramadol (10 mg/ml of 0.9% saline)was injected in the epidural space. The patient’s vitals wererecorded periodically and UDS was repeated after 30 min ofthe epidural injection in a similar manner. Immediately afterthe study, patients were shifted to the endoscopy suite forcystoscopy for respective indication, and were subsequentlyobserved for 6 hr.

STATISTICAL ANALYSIS

All data were fed into a Microsoft Excel spreadsheet andanalyzed using SPSS 10.0 statistical software for Windows.

Normality of data was tested and confirmed using normalquantile plots for all variables. All the variables are presentedas mean � standard deviation (SD). The effect of epiduralTramadol on various urodynamic parameters was analyzedusing the ‘paired t-test’ for continuous variables and the‘McNemar Chi-square test’ for categorical variables. The‘Independent-sample t-test’ was utilized to analyze the effectof presence or absence of guarding reflex on urodynamicparameters. In addition, linear regression analyses were doneto assess the interactions between continuous variables.Correlation-coefficients were tested for significance using thet-test. Correlation coefficients were compared using multipleregression analyses. P values �0.05 were considered statisti-cally significant.

RESULTS

The mean age of the patients was 54.0 � 14.3 years (range,36–80) and the American Urological Association symptomscore 2.34 � 3.41 (range, 0–10). Three patients (two males,one female) had a urodynamically proven bladder outletobstruction (BOO). These males had AG no. 54 and 47,PdetQmax 63 and 60 cm H2O and Qmax of 5 and 7 ml/sec,respectively. One female with BOO had PdetQmax 56 cm H2O,Qmax 5 ml/sec, urethral resistance 2.24 cm H2O/(ml/sec)2 andPVR of 68 ml.

The majority of patients (13/15; 86.6%) tolerated thedrug without significant side effects. One patient developedtransient hypotension and nausea within 20 min of the in-jection of tramadol and was managed with an infusion of500 ml 0.9% saline and i.v. metoclopramide (10 mg),respectively. Another patient complained of nausea 6 hr afterthe injection and was treated with a single dose of oralmetoclopramide.

Urodynamic parameters before and after epidural tramadoladministration are shown in Table I. There was no significantdifference in urine production before and after tramadolinjection (Table I). Typical pressure flow studies beforeand after tramadol injection (of one patient) are shown inFigures 1 and 2, respectively. Maximum cystometric capacityand detrusor-compliance increased significantly (P¼ 0.019and 0.011, respectively) after injection of tramadol withouta significant change in filling pressure (P¼ 0.576). Overall,11 (73.3%) patients showed increased filling volumes and12 (80%) increased compliance. Filling sensations weredelayed significantly as depicted by a significant rise involume at first sensation, normal desire, strong desire andmaximum cystometric capacity (P¼ 0.028, 0.001, 0.003 and0.019, respectively). However, this did not interfere withvoiding; Qmax, Qave and voiding pressures (PdetMax and Pdet atQmax, respectively) did not show a significant change frombaseline study (P¼ 0.861, 0,715, 0.398 and 0.629, respectively).This was true even in obstructed patients (P¼ 0.199 forvoiding Pdet; P¼ 0.383 for Qmax and P¼ 0.427 for PVR).Separate analysis of patients on indwelling ureteral stents(n ¼ 4) also suggested similar findings. Analyzing data of14 patients (excluding data of one patient who requiredmetoclopramide during the Urodynamic study) also showedsimilar results. The AG number, DAMPF and Chess curve alsodid not show any significant change (P¼ 0.649, 0.602 and0.317, respectively). PVR showed a rising trend, thoughstatistically insignificant (P¼ 0.063), from 15.7 ml (range 0–68 ml) to 37.3 ml (range 0–104 ml). No patient complained ofincreased difficulty in voiding or urinary retention eitherimmediately after injection (within 1 hr) or upto discharge(6 hr).

Neurourology and Urodynamics DOI 10.1002/nau

66 Singh et al.

Pelvic floor EMG showed a significant fall after injection(P¼ 0.027) [Figs. 1 and 2]. A progressive rise in EMG activityduring the filling phase was observed in 12 (80%) patientsbefore the injection, whereas only five (33%) showed thisfinding after it. This difference was highly significant(P¼ 0.016). Dividing the patients into two groups on the basisof presence or absence of progressive rise in EMG, we foundfilling capacity and compliance to be lower and EMG activityand maximum filling pressure higher in the former group,though it was statistically insignificant (Table II). This trendwas minimized after the injection (Table II). No effect onvoiding parameters was observed in either group.

When the data were analyzed separately for males andfemales (Table III), presence or absence of ureteral stent andpresence or absence of bladder outlet obstruction, a similartrend was observed in the results.

DISCUSSION

Tramadol is a synthetic 4-phenyl-piperidine analog ofcodeine with a dual mechanism of action of weak m-agonistand inhibitor of serotonin (5-HT) and norepinephrine (NE)reuptake. There is some evidence of direct serotonin-releasingaction as well.8 Its efficacy has been found to be equivalent to

morphine in moderate postoperative pain, including that afterthoracotomy, major abdominal and pelvic surgeries, and infact higher in neuropathic pain (due to the latter mechanism).In severe pain, the efficacy of 100 mg tramadol is comparableto ibuprofen (400 mg) or aspirin (650 mg) with codeine (60 mg)but is less than morphine.11

Delayed sensations and increased bladder capacity andcompliance have been reported earlier with the use of m-agonist opioids other than tramadol, both intravenously aswell as via the spinal-epidural route. The reported change inbladder capacity varies from 20 to 65% depending on theopioid, relative dose, patient-group and route of administra-tion.12,13 The urodynamic effects of tramadol have not beenprospectively investigated in clinical studies; only animalstudies have been conducted. Pandita et al.9 reported asignificant increase in the bladder capacity and compliance(43 � 12%) following intravenous and intrathecal injection oftramadol (5 mg/kg) in female Sprague–Dawley rats. Pehrsonet al.14 reported a 59 � 29% increase in bladder capacity withthe same dose of intravenous tramadol in the same species ofrats with artificially induced cerebral ischemia. The results ofour clinical study corroborate these experimental observa-tions (�12% rise in bladder capacity). The mechanism and siteof action of opioids (including tramadol) affecting bladder

Neurourology and Urodynamics DOI 10.1002/nau

Fig. 1. Pressure-flow study with surface EMG of one patient before injection

of tramadol.

Fig. 2. Pressure-flow study with surface EMG of the same patient after

injection of tramadol showing increased compliance and capacity, decreased

EMG activity, abolished guarding reflex and unchanged voiding parameters.

TABLE I. Urodynamic Parameters Before and After Administration of Tramadol

S. no. Parameter Baseline value 30 min after tramadol P value

1 Volume at first sensation (ml) 78.6 � 88.6 115.4 � 118.0 0.028

2 Volume at first desire (ml) 129.4 � 95.1 170.7 � 137.2 0.069

3 Volume at normal desire (ml) 236.5 � 144.7 311.6 � 177.9 0.001

4 Volume at strong desire (ml) 322.2 � 158.3 373.8 � 190.7 0.003

5 Maximum cystometric capacity (ml) 391.8 � 179.6 432.7 � 208.8 0.019

6 Maximum Pdet filling (cm H2O) 22.5 � 13.2 24.1 � 15.1 0.576

7 Compliance (ml/cm H2O) 60.1 � 51.5 83.0 � 63.0 0.011

8 EMG (mV) during filling phase 26.35 � 21.47 12.96 � 4.47 0.027

9 Voided volume (ml) 376.1 � 183.9 395.3 � 225.6 0.357

10 Post-voided residue (ml) 15.7� 23.0 37.3� 40.0 0.063

11 Urine production during UDS (ml) 65.9� 62.5 39.4� 33.7 0.159

12 Max Pdet during voiding (cm H2O) 49.5 � 21.9 47.6 � 19.2 0.398

13 Pdet at Qmax (cm H2O) 35.9 � 17.2 34.6 � 15.1 0.629

14 Qmax (ml/sec) 12.2 � 6.5 12.0 � 6.4 0.861

15 Qave (ml/sec) 7.5 � 5.5 7.1 � 4.5 0.715

16 AG number 13.5 � 24.5 12.1 � 25.1 0.649

17 DAMPF 40.1 � 18.9 38.3 � 24.6 0.602

18 Chess curve 0.51 � 0.64 0.76 � 1.26 0.317

Effect of Epidural Tramadol on Lower Urinary Tract Function 67

function differs according to the route of administration; theeffect being more rapid and marked when given via anintrathecal or epidural route.15 Moreover, thoracic epiduralhas been reported to be associated with a lower incidence ofurinary retention than lumbar.16 This suggests a primarylumbar-spinal site of action rather than cerebral or peripheral.Coggeshall and Carlton17 found that in rats, opioid m receptorsare concentrated in the dorsal horn of the spinal cord wherethe bladder afferents merge. Delta (d) and k receptors are alsopresent, but in lesser concentration. Both m and d, and not kreceptors are involved in bladder relaxation and impairedsensations by inhibiting the sensory input at the level of thedorsal horn. This is further supported by the absence ofsuch action by non m-agonist opioids, for example nalbuphine,a k-agonist and m-antagonist and pentazocine, a k and s-agonist.12 The accommodation of the bladder to increasingvolumes of urine is predominantly a passive phenomenondependent on its viscoelastic properties and the quiescence ofthe parasympathetic system.18 By inhibiting bladder afferentsat the dorsal horn, m-receptor activation attenuates bladdersensations and may delay initiation of the micturition reflexwhich is normally triggered as the volume exceeds themicturition threshold, therefore causing increased complianceand capacity. Moreover, a direct effect of opioid receptoractivation at the sacral parasympathetic innervation alsosignificantly contributes to compliance.19 Apart from theabove, the 5-HT reuptake inhibitor action of tramadol mightalso contribute to this, as some cat experiments suggest that

bulbospinal projections from serotonergic raphe neurons inthe brainstem mediate inhibitory control of sacral parasym-pathetic outflow and facilitatory control of lumbar sympa-thetic outflow to the urinary bladder.20

Morphine and other opioids with m-agonist action areknown to cause urinary retention and an increase in post-voidresidue with a reported incidence of upto 90%.21 No prospec-tive clinical data is available on evaluation of urodynamiceffects of tramadol. In a clinical report of five cases withtransient impairment of micturition or urinary retention, atemporal relation with the oral intake of tramadol wasobserved. However, causality could not be confirmed in thisreport.22 In another study of 60 patients receiving either 50 mgtramadol or 100 mg tramadol or 10 ml of Bupivacaine 0.25%for postoperative pain relief, only one patient complained ofinability to pass urine in a 24 hr period and he was in theformer group.23 Conversely, animal data suggest that trama-dol does not affect voiding function despite an increase incompliance.9 Our results are in compliance with these animalexperiments. In the present study, there was no significantincrease in PVR (P¼ 0.063); none of the patients developedincreased voiding difficulty or urinary retention. The exactmechanism of favorable effect of tramadol (compared to otherm-agonist opioids) on voiding is not known. Andersson24

suggested that unlike morphine, which has a narrow rangebetween the doses causing inhibition of micturition andincreased bladder capacity, respectively, tramadol has effectsover a considerably wider range. There must be some

Neurourology and Urodynamics DOI 10.1002/nau

TABLE II. Effect of Presence or Absence of Guarding Reflex on Filling Parameters Before and After TramadolInjection

ParameterEMG activity

(mV)Filling capacity

(ml)Filling compliance

(ml/cm H2O)Maximum filling

pressure (cm H2O)

Before tramadol

Guarding reflex present (n ¼ 12) 28.9 � 23.7 352.8 � 156.3 55.2 � 52.6 24.4 � 14.4

Guarding reflex absent (n ¼ 3) 14.0 � 4.4 535.0 � 220.4 78.2 � 53.2 15.6 � 4.1

P value 0.143 0.285 0.546 0.103

After tramadol

Guarding reflex present (n ¼ 5) 15.5 � 5.3 337.0 � 138.2 74.9 � 32.2 22.5 � 10.4

Guarding reflex absent (n ¼ 10) 11.9 � 3.9 472.7 � 224.5 86.2 � 73.2 24.7 � 17.0

P value 0.288 0.203 0.695 0.775

TABLE III. Urodynamic Parameters Before and After Administration of Tramadol Analyzed Separately for Males and Females

No. Parameter

Baseline value,

males (n ¼ 7)

30 min after

tramadol P value

Baseline value,

females (n ¼ 8)

30 min after

tramadol P value

1 Volume at first sensation (ml) 112.60 � 107.28 155.60 � 145.32 0.091 36.00 � 35.02 65.25 � 54.23 0.268

2 Volume at first desire (ml) 169.00 � 166.44 228.00 � 235.08 0.312 110.71 � 40.62 140.28 � 65.16 0.169

3 Volume at normal desire (ml) 238.5714 � 109.47 323.85 � 140.37 0.012 221.71 � 176.01 281.14 � 216.06 0.049

4 Volume at strong desire (ml) 329.14 � 133.31 391.57 � 141.77 0.012 290.71 � 189.91 334.71 � 231.80 0.049

5 Cystometric capacity (ml) 376.71 � 165.34 423.71 � 158.70 0.042 394.28 � 202.23 437.85 � 239.48 0.034

6 Maximum Pdet filling (cm H2O) 21.57 � 14.99 24.42 � 16.19 0.560 23.42 � 12.32 23.71 � 15.12 0.933

7 Compliance (ml/cm H2O) 69.15 � 60.34 86.07 � 77.46 0.061 51.11 � 43.73 79.97 � 50.85 0.084

8 EMG (mV) during filling phase 23.57 � 13.51 13.42 � 4.95 0.048 29.14 � 19.01 12.21 � 3.89 0.040

9 Voided volume (ml) 374.57 � 173.57 375.28 � 174.65 0.979 377.71 � 207.73 415.42 � 280.90 0.275

10 Post-voided residue (ml) 14.85 � 21.03 51.28 � 43.16 0.110 16.57 � 26.60 23.42 � 33.94 0.316

11 Max Pdet during voiding (cm H2O) 58.57 � 26.38 56.71 � 20.83 0.604 40.42 � 12.27 38.5714 � 13.06 0.537

12 Pdet at Qmax (cm H2O) 41.28 � 18.50 39.00 � 15.87 0.747 30.57 � 15.15 30.14 � 14.08 0.939

13 Qmax (ml/sec) 9.71 � 5.49 8.57 � 3.95 0.462 14.71 � 6.79 15.42 � 6.85 0.728

14 Qave (ml/sec) 5.57 � 2.89 4.14 � 1.92 0.158 9.42 � 7.04 10.00 � 4.65 0.807

15 AG number 22.28 � 25.31 21.42 � 22.26 0.821 4.71 � 21.99 2.85 � 25.79 0.715

16 DAMPF 43.85 � 20.86 47.57 � 21.76 0.234 36.28 � 17.58 29.00 � 25.19 0.234

17 Chess curve 0.62 � 0.67 0.72 � 0.72 0.639 0.38 � 0.62 0.80 � 1.78 0.415

68 Singh et al.

additional mechanism by which the m effect of tramadol iscounterbalanced; nor-epinephrine and serotonin reuptakeinhibition by tramadol may be one such although theavailable literature is controversial. Various rat-experimentsindicate that certain spinal and supraspinal a-adrenoceptors(a-1A and a-2) have a facilitatory action on the descendinglimb of the micturition-reflex whereas certain others (a-1D)have inhibitory control over the frequency of thesereflexes.24,25 de Groat20 reviewed the importance of descend-ing serotonergic (5-HT) pathways in modulating lower urinarytract voiding function. Cat experiments have suggested thatthis is an inhibitory pathway, as Duloxetine (a monoaminereuptake inhibitor) increased bladder capacity as well as theEMG activity of the urethral sphincter in ‘irritated bladders’.26

Conversely, Lecci et al.27 reported that this pathway is morecomplicated in rats. They found that intravenous, intrathecaland intracerebroventricular administration of 8-OH-DPAT(a 5-HT-1A agonist) resulted in facilitation of the micturi-tion-reflex in urethane-anesthetized rats and this effect wasblocked by NAN-90 (a 5-HT-1A antagonist). Recently, D’Agos-tino et al.28 demonstrated the presence of prejunctional 5-HTreceptors at parasympathetic nerve terminals in the humanbladder. These were found to be involved in both positive(5-HT4 and 5-HT7) and negative (5-HT1A) feedback mechan-isms regulating cholinergic transmission. Probably, somesystemic absorption following epidural administration oftramadol may influence the detrusor via these mechanisms.

During bladder filling, the sphincter EMG increases pro-gressively, reflecting increased efferent firing in the pudendalnerve which is activated by bladder afferent inputs (guardingreflex).18 In our study this reflex was found to be attenuatedafter administration of tramadol. It can be explained byinhibition of the afferent limb of the guarding reflex at thedorsal horn, as explained earlier. We observed a trend towardshigher compliance and larger capacity in patients who did notshow this guarding reflex on EMG tracing (as compared tothose who did) both before and after administration oftramadol; it again points to the role of sensory inputs fromthe bladder during the filling phase, alteration in which wouldlead to higher compliance as well as the loss of the guardingreflex.

One interesting observation we made was a statisticallysignificant fall in EMG activity after injection of Tramadol. Toour understanding, this has never been reported in theliterature and at present time, it is difficult to explain themechanism involved. The likely explanation may be modula-tion of proprioceptive afferent inputs from the pelvic floor bytramadol at the dorsal horn level, which are important inmaintaining muscle tone, and hence EMG activity. Anotherexplanation may be sought in the peripheral 5-HT releasingeffect of systemically absorbed tramadol, which may haveactivated excitatory 5-HT receptors and reflexly causedrelaxation of the urethral sphincter. Conversely, certain catexperiments suggested that 5-HT receptor activation in thespinal cord by Duloxetine led to increased EMG activity of theurethral sphincter in ‘irritated bladders’ [Thor and Katofiasc,1996]. In human beings, this property of Duloxetine isclinically applicable in the management of stress urinaryincontinence.28 Therefore, molecular studies of greater mag-nitude investigating the neurophysiology of the pelvic floorand urethral pressure profile studies are required to furtherelucidate this conjecture.

Pandita et al.9 suggested a possible diuretic effect oftramadol which could be blocked by desmopressin. Wecalculated urine production both before and after theinjection of tramadol (urine production ¼ voided volume þ

PVR � volume filled) and found no diuretic effect of the drug(65.9 � 62.5 ml before injection and 39.4 � 33.7 ml after theinjection; P¼ 0.159).

CONCLUSIONS

Our study indicates that epidural tramadol increasesbladder capacity and compliance without any significant illeffect on voiding. This is true even in patients with bladderoutflow obstruction; although the number is rather small(n ¼ 3) to draw this conclusion. These results are promisingand will guide the clinician to avoid catheterization in caseswhere epidural tramadol is used for postoperative pain, asagainst morphine and other opioids which do cause urinaryretention. In addition, these patients may require bladderemptying less frequently (due to increased capacity) adding topostoperative comfort. The study also identifies an intriguingand novel effect of tramadol on the pelvic floor EMG which hasa scope for further research.

REFERENCES

1. Behar M, Olshwang D, Magora F, Davidson JT. Epidural morphine intreatment of pain. Lancet 1979;1:527–9.

2. Cousins MJ, Mather LE. Intrathecal and epidural administration of opioids.Anesthesiology 1984;61:276–310.

3. Schaeffer AJ. Infections of the urinary tract. In: Walsh PC, Retik AB, VaughanED, Wein AJ, editors. Campbell’s urology. 8th edition. Philadelphia: Saunders;2002. Vol. I, Chapter 14, p 515–601.

4. Singh SK, Wadhwa P, Bapuraj JR, Jha V. Transcatheter embolization ofinternal pudendal artery pseudoaneurysm following traumatic urethralcatheterization. Int Urol Nephrol 2005;37:93–4.

5. Turker G, Goren S, Bayram S, Sahin S, Korfali G. Comparison of lumbarepidural tramadol and lumbar epidural morphine for pain relief afterthoracotomy: A repeated-dose study. J Cardiothorac Vasc Anesth 2005;19:468–74.

6. Aribogan A, Doruk N, Aridogan A, Akin S, Balcioglu O. Patient-controlledepidural analgesia after major urological surgeries. A comparison oftramadol with or without bupivacaine. Urol Int 2003;71:168–75.

7. Rowlingson JC. Chronic pain. In: Miller RD, editor. Miller’s anesthesia. 6thedition. Philadelphia: Elsevier Churchill Livingstone; 2005. Vol. 1, Chapter 73,p 2763–81.

8. Fukuda K. Intravenous opioid anesthetics. In: Miller RD, editor. Miller’sanesthesia. 6th edition. Philadelphia: Elsevier Churchill Livingstone; 2005.Vol. 1, Chapter 11, p 379–437.

9. Pandita RK, Pehrson R, Christoph T, Friedrichs E, Andersson KE. Actions oftramadol on micturition in awake, freely moving rats. Br J Pharmacol2003;139:741–8.

10. Massey JA, Abrams P. Obstructed voiding in the female. Br J Urol 1988;61:36–9.

11. Moore RA, McQuay HJ. Single-patient data meta-analysis of 3453 post-operative patients: Oral tramadol versus placebo, codeine and combinationanalgesics. Pain 1997;69:287–94.

12. Malinovsky JM, Normand L, Lepage JY, Malinge M, Cozian A, Pinaud M,Buzelin JM. The urodynamics effects of intravenous opioids and ketoprofenin humans. Anesth Analg 1998;87:456–61.

13. Kuipers PW, Kamphuis ET, Venrooji EV, et al. Intrathecal opioids and lowerurinary tract function: A urodynamics evaluation. Anesthesiology 2004;100:1497–1503.

14. Pehrson R, Stenman E, Andersson KE. Effect of tramadol on rat detrusoroveractivity induced by experimental cerebral infarction. Eur Urol 2003;44:495–9.

15. Chaney MA. Side-effects of intrathecal and epidural opioids. Can J Anesth1995;42:891–903.

16. Basse LB, Werner M, Kehlet H. Is urinary drainage necessary duringcontinuous epidural analgesia and after colonic resection. Reg Anesth PainMed 2000;25:498–501.

17. Coggeshall RE, Carlton SM. Receptor localization in the mammalian dorsalhorn and primary afferent neurons. Brain Res Brain Res Rev 1997;24:28–66.

18. Chancellor MB, Yoshimura N. Physiology and pharmacology of bladderand urethra. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ, editors.Campbell’s urology. 8th edition. Philadelphia: Saunders. 2002. Vol. II, Chapter23, p 831–86.

19. Drenger B, Magora F, Evron S, Caine M. The action of intrathecal morphineand methadone on the lower urinary tract in the dog. J Urol Apr 1986;135:852–5.

Neurourology and Urodynamics DOI 10.1002/nau

Effect of Epidural Tramadol on Lower Urinary Tract Function 69

20. de Groat W. Influence of central serotonergic mechanisms on lower urinarytract function. Urology 2002;59:30–6.

21. Bromage PR, Camporesi EM, Durant PAC, Nielson CH. Nonrespiratory side-effects of epidural morphine. Anesth Analg 1982;61:490–5.

22. Meyboom RH, Brodie-Meejer CC, Diemont WL, van Puijenbroek EP. Bladderdysfunction during the use of tramadol. Pharmacoepidemiol Drug Saf 1999;8:S63–4.

23. Delican AE, Vijayan R. Epidural tramadol for postoperative pain relief.Anesthesia 1993;48:326–31.

24. Andersson KE. New pharmacologic targets for the treatment of theoveractive bladder: An update. Urology 2004;63:32–41.

25. de Groat WC, Yoshimura N. Pharmacology of the lower urinary tract. Ann RevPharmacol Toxicol 2001;41:691–721.

26. Thor KB, Katofiasc MA. Effect of Duloxetine, a combined serotonin andnorepinephrine reuptrake inhibitor, on central neural control of lowerurinary tract function in the chloralose anesthetised female cat. J PharmacolExp Ther 1995;274:1014–24.

27. Lecci A, Giuliani S, Santicioli P, et al. Involement of 5-HT-1A receptors in themodulation of micturition reflexes in the anesthetized rats. J Pharmacol ExpTher 1992;262:181–9.

28. D’Agostino G, Condino AM, Gallinari P, Franceschetti GP, Tonini M.Characterization of prejunctional serotonin receptors modulating [3H]acety-lcholine release in human detrusor. J Pharmacol Exp Ther 2006;316:129–35.

29. Mariappan P, Alhasso A, Ballantyne Z, Grant A, N’Dow J. Duloxetine, aserotonin and noradrenaline reuptake inhibitor (SNRI) for the treatment ofstress urinary incontinence: A systematic review. Eur Urol 2007;51:67–74.

Neurourology and Urodynamics DOI 10.1002/nau

70 Singh et al.