remifentanil for fetal immobilization and maternal sedation

Remifentanil for Fetal Immobilization and Maternal SedationDuring Fetoscopic Surgery: A Randomized, Double-BlindComparison with DiazepamMarc Van de Velde, MD, PhD, Dominique Van Schoubroeck, MD, Liesbeth E. Lewi, MD,Marco A. E. Marcus, MD, PhD, Jacques C. Jani, MD, Carlo Missant, MD, An Teunkens, MD, andJan A. Deprest, MD, PhD

Departments of Anaesthesiology and Obstetrics and Gynaecology, University Hospital Gasthuisberg, KatholiekeUniversiteit Leuven, Leuven, Belgium, and the Department of Anaesthesiology, University of Maastricht, The Netherlands

Obstetric endoscopy procedures are routinely performedat our institution to treat selected complications of mono-chorionic twin gestation. We perform these proceduresunder combined spinal epidural anesthesia plus maternalsedation. In the absence of general anesthesia, fetal immo-bilization is not achieved. We hypothesized that remifen-tanil would induce adequate maternal sedation and pro-vide fetal immobilization, which is equal or superior tothat induced by diazepam. Fifty-four second trimesterpregnant women were included in this randomized,double-blind trial. After combined spinal epidural anes-thesia, maternal sedation was initiated using either incre-mental doses of diazepam or a continuous infusion ofremifentanil. Maternal sedation, hemodynamics, side ef-fects, and fetal hemodynamics and immobilizationwere evaluated before, during, and for 60 min aftersurgery. Remifentanil produced adequate maternalsedation with mild but clinically irrelevant respira-tory depression (respiratory rate 13 � 4 breaths/min

and Pco2 38.6 � 4 mm Hg at 40 min of surgery),whereas diazepam resulted in a more pronouncedmaternal sedation but no respiratory depression (re-spiratory rate 18 � 3 breaths/min and Pco2 32.7 �3 mm Hg at 40 min of surgery). Compared with diaz-epam, fetal immobilization with remifentanil oc-curred faster and was more pronounced, resulting inimproved surgical conditions; the number of grossbody and limb movements was 12 � 4 (diazepam)versus 2 � 1 (remifentanil) at 40 min of surgery. Be-cause of this, the mean (range) duration of surgerywas significantly shorter in the remifentanil-treatedpatients, 60 (54 –71) min versus 80 (60 –90) min in thediazepam group. We conclude that remifentanil pro-duces improved fetal immobilization with good ma-ternal sedation and only minimal effects on maternalrespiration.

(Anesth Analg 2005;101:251–8)

A s a result of advances in high-resolution ultra-sound, an increasing number of fetal conditionsare diagnosed early in gestation. Some of these

conditions are life threatening or may cause irrevers-ible organ damage but may benefit from a prenatalintervention (1–3). Miniaturization of endoscopes has

revived the interest in fetoscopy and today it has adistinct place in modern fetal medicine (1,3). Fetalsurgery includes all types of surgery in which directinterventions on the fetus are performed, but few ofthem are amenable to treatment by endoscopy. Theterm “obstetric endoscopy” was proposed for feto-scopic procedures on the placenta, the umbilicalcord, and fetal membranes. The technique of lasercoagulation of the vascular anastomoses on themonochorionic placenta for twin-to-twin transfu-sion syndrome is well established and has recentlybeen shown to be superior to amniodrainage in arandomized controlled trial (4). Therefore, the num-ber of these procedures will undoubtedly increase(2– 4). Another application of obstetric endoscopy isselective feticide by cord occlusion, used in mono-chorionic twin pregnancies complicated by severediscordant anomalies (5,6).

Supported, in part, by a 2002 Society of Anesthesia and Reanima-tion of Belgium (SARB) grant for experimental research, by “Kredietaan Navorsers” (nr. 1.5.080.03) granted by the Fund for ScientificResearch Flanders (Fonds voor Wetenschappelijk Onderzoek Vlaan-deren), and by grants financed by the European Commission (EuroTwin 2 Twin, QLG1-CT-2002–01632, to Drs Jani and Lewi).

Accepted for publication January 7, 2005.Address correspondence to: Marc Van de Velde, MD, PhD.,

Director Obstetric Anesthesia and Extra Muros Anesthesia, De-partment of Anaesthesiology, University Hospitals Gasthuisberg,Herestraat 49, B - 3000 Leuven, Belgium. Address e-mail [email protected].

DOI: 10.1213/01.ANE.0000156566.62182.AB

©2005 by the International Anesthesia Research Society0003-2999/05 Anesth Analg 2005;101:251–8 251

These procedures are often performed under localanesthesia (4). However, we liberally use combinedspinal epidural anesthesia as a means of maternalanesthesia. Combined spinal epidural and local anes-thesia provide neither fetal immobilization nor anes-thesia. Based on clinical experience, we believe thatfetal movements may lead to fetal trauma and mayhamper surgery, leading to incomplete coagulation ofvessels, failure of surgery and an increase in the du-ration of the intervention (7). Increasing the durationof endoscopic surgery may increase the risk of iatro-genic preterm, prelabor rupture of membranes (8,9).

Fetal immobilization has been traditionally ob-tained by maternal administration of diazepam (DZP),which is associated with maternal sedative effects.Although it provides maternal sedation, in our expe-rience DZP produces insufficient fetal immobility.Remifentanil (REMI) is a novel, short-acting opioid,which is rapidly hydrolyzed by nonspecific plasmaand tissue esterases. It has been used for intraopera-tive sedation in patients undergoing regional or localanesthesia (10–14). In term pregnant women under-going elective Cesarean delivery under epidural anes-thesia, it produces excellent maternal sedation withoutadverse maternal effects (15). Kan et al. (15) demon-strated that IV REMI, in a dose of 0.1 �g · kg�1 · min�1

and part of a regional anesthetic technique, rapidlyand extensively crosses the placenta (umbilical vein/maternal artery ratio, 0.88). An initial dose responsestudy determined that a dose of 0.1 �g · kg�1 · min�1

of REMI produced excellent fetal immobilization insecond trimester pregnant patients (16). Based on thisdose-response study, we hypothesized that REMI in adose of 0.1 �g · kg�1 · min�1 would induce superiorfetal immobilization during obstetric endoscopic sur-gery as compared with DZP and at the same timeprovide appropriate maternal sedation. Therefore, weinitiated a randomized, double-blind trial comparingthe effects of IV DZP versus IV REMI in pregnantwomen undergoing obstetric endoscopic surgery un-der neuraxial block. We postulated that fetal immobi-lization and maternal sedation provided by REMIwould be at least as good as, if not superior to, DZP.

MethodsAfter Institutional Ethics Committee approval, 54healthy (ASA I–II) women in the second trimester ofpregnancy (gestational age, 16–25 wk), carrying amultiple pregnancy and scheduled for either feto-scopic laser coagulation or cord occlusion, providedwritten and informed consent to this randomized,double-blind trial.

Before anesthesia and surgery, all patients receivedprophylaxis for acid aspiration using 30 mL oral so-dium citrate 0.3 M, metoclopramide 10 mg IV, and

ranitidine 50 mg IV 60 min before initiation of anes-thesia. Mothers were prehydrated using 1000 mL oflactated Ringer’s solution through an IV catheter inthe right forearm. A second IV cannula was positionedin the left antecubital vein to infuse maternal sedativedrugs. Under local anesthesia, the left radial arterywas cannulated to allow continuous arterial bloodpressure measurements and repetitive blood sam-pling. Combined spinal epidural anesthesia was per-formed at the L3-4 or L4-5 interspace with the patientsitting. The epidural space was identified using an18-gauge Tuohy needle using the loss of resistance tosaline technique. The dura was entered using a 27-gauge pencil point spinal needle and 8 mg of hyper-baric bupivacaine 0.5% was injected into the spinalspace, after which a 20-gauge epidural catheter wasadvanced 4 cm into the epidural space. Anesthesiawas maintained by additional epidural top-ups ofropivacaine 0.75% at the discretion of the attendinganesthesiologist. If hypotension (defined as a decreasein mean arterial blood pressure of �10% from baselinevalues recorded immediately before anesthesia) oc-curred, ephedrine or phenylephrine was administeredat the discretion of the attending anesthesiologist.

The patient was then positioned in the supine posi-tion with 15 degrees left lateral tilt to prevent aorto-caval compression (17). Supplemental oxygen (5L/min) was routinely administered by face mask. Af-ter baseline recordings, maternal IV sedation wasstarted. Patients were randomized to 2 groups of 27patients by a computer-generated list. Study drugswere prepared and administered by an anesthesiolo-gist not involved in further management of the pa-tients. Patients, surgeon, and attending anesthesiolo-gist were blinded as to the sedative drugs used. In theDZP group a continuous infusion of saline mimickedthe REMI infusion. DZP was initiated using a dose of5 mg IV, followed 10 min later by an additional 5 mg.Additional 2.5-mg increments of DZP were givenwhen maternal sedation was judged inadequate by anobserver assessment of alertness/sedation scale(OAA/S) score of 5 or when fetal immobility wasjudged inadequate by the surgeon. In case a top-updose of DZP was required, an increase in the “sham”saline infusion rate was performed simultaneously. Asto the maximum total dose of DZP, no additionaltop-ups were given if maternal sedation was profound(OAA/S score of 3 or less), maternal arterial blood gasanalysis revealed a pH �7.35 or a Pco2 of �45 mm Hg,or maternal respiratory rate decreased to �8 breathsper minute.

In the REMI group a continuous infusion of REMIwas started at an initial flow rate of 0.1�g · kg�1 · min�1 (dilution of REMI 50 �g/mL) and at0 and 10 min a bolus of normal saline was given tomimic the DZP administration. The initial dose ofREMI was based on a previous dose-response study at

252 OBSTETRIC ANESTHESIA VAN DE VELDE ET AL. ANESTH ANALGFETAL IMMOBILIZATION WITH REMIFENTANIL DURING FETOSCOPIC SURGERY 2005;101:251–8

our institution (16). Sham saline boluses and an in-crease of the REMI infusion rate with 0.025�g · kg�1 · min�1 were performed if maternal seda-tion was inadequate (OAA/S score of 5) or if fetalimmobility was judged to be insufficient by the sur-geon. The REMI or saline infusion was decreased by0.025 �g · kg�1 · min�1 if maternal sedation was pro-found (OAA/S of 3 or less), maternal blood gas pHdecreased to less than 7.35, the arterial Pco2 increasedto more than 45 mm Hg, or maternal respiratory ratedecreased to �8 breaths per minute. If maternal apneaoccurred, cricoid pressure was applied and mask ven-tilation was initiated until spontaneous respiration re-sumed and the REMI infusion would be stoppedimmediately.

At the end of surgery the REMI infusion wasstopped. The observation period started at the mo-ment of first administration of REMI until 60 min afterthe end of surgery. In both groups, all necessarychanges in infusion rate and additional boluses ofDZP were made by an anesthesiologist not involved indata recording.

Before the study, demographic data, medical his-tory, relevant obstetrical data, maternal arterial bloodpressure as measured invasively, maternal heart andrespiratory rate were recorded. Maternal side effectswere noted. Maternal sedation was evaluated by theattending anesthesiologist using the OAA/S (18). Wetargeted the sedation to aim at an ideal OAA/S scoreof 4; a score of �4 was considered profound sedationand a score �4 was considered insufficient sedation.Sedation was evaluated at baseline, at 20, 40, and60 min during surgery, and at 10, 20, 30, and 60 minafter completion of surgery. Maternal arterial bloodgas analysis was performed before the start of seda-tion, every 20 min during surgery, and at 10, 20, 30,and 60 min after the end of surgery.

Fetal heart rate was recorded every 15 min usingultrasound. Fetal mobility was assessed before, dur-ing, and after surgery by taping 5 min ultrasoundsequences of fetal movement every 20 min throughoutsurgery and 10, 20, 30, and 60 min after the end ofsurgery. These taped sequences were evaluated off-line by an experienced ultrasonographer. For that pur-pose the video sequences were randomly presentedwith patient identification blinded. The baseline re-cording was presented first for each patient. Twotypes of evaluation were performed: a visual analoguescale score for mobility (0 � immobile fetus and 100 �baseline mobility) and the number of gross bodymovements and limb movements per 5-min period. Iffetuses were immobile before the start of sedation,patients were excluded from further analysis. Only thefetal movements of the non-stuck twin were recordedin case of twin-to-twin transfusion syndrome, andonly movements of the normal fetus were recorded incase of selective feticide.

At the end of the intervention, the surgeon assessedoverall fetal immobility and operating conditions us-ing a four-point scale: 1 � excellent, 2 � good, 3 �moderate, 4 � inadequate or no immobilization. Thissubjective score represented an overall subjective im-pression and is further referred to as the surgicalassessment score.

Perinatal variables included the number of surviv-ing fetuses, gestational age at delivery, and neonatalsurvival and complications.

Data were analyzed using two-way repeated meas-ures analysis of variance followed by Scheffe’s post hoctesting as required. Categorical data were analyzedusing �2 analysis and Fisher’s exact test. Data arepresented as a mean � sd, median and interquartilerange, or as percentage of the group total. P � 0.05was considered as statistically significant. Our prelim-inary experience with DZP sedation demonstrated ad-equate fetal immobilization in approximately 30% ofpatients; adequate maternal sedation was achieved inmost mothers. In a dose finding study for REMI weachieved fetal immobilization in more then 80% ofpatients using 0.1 �g · kg�1 · min�1; maternal seda-tion was adequate. For sample size calculations, weexpected a 50% increase in adequate fetal immobilityfrom 30% to 80% of fetuses when using REMI. Wecalculated the number of patients required in eachgroup to demonstrate a statistically significant differ-ence to be 23 subjects (� � 0.05, � � 0.05).

ResultsIn two patients in each group, fetuses were immobilebefore the start of sedation and surgery and thereforethese were excluded. This left 50 patients for analysis,25 in each group. Gestational age at intervention, thenumber of laser coagulations, and cord occlusionswere comparable in the two groups. There was nosignificant difference in gestational age at delivery andsurvival rates between the treatment groups, both forlaser cases and cord occlusions (Table 1). The inci-dence of preterm labor and delivery was not signifi-cantly different between the two groups.

Results related to maternal sedation and fetal im-mobilization are summarized in Table 2 and Figures 1through 6. REMI produced excellent levels of maternalsedation in all patients. Only one patient (4%) hadan OAA/S score �4 and was therefore consideredto be profoundly sedated during surgery (Fig. 4).The mean REMI infusion rate was 0.115 � 0.020�g · kg�1 · min�1. The most rapid REMI infusionrate was 0.150 �g · kg�1 · min�1. In the DZP group,11 women (44%) were profoundly sedated (OAA/Sscore �4). The mean total DZP dose was 14.5 �4.8 mg. Maternal respiratory rate in the REMI groupdecreased during surgery; it remained stable in the

ANESTH ANALG OBSTETRIC ANESTHESIA VAN DE VELDE ET AL. 2532005;101:251–8 FETAL IMMOBILIZATION WITH REMIFENTANIL DURING FETOSCOPIC SURGERY

DZP group. As a result of maternal hypoventilation,an increase in Pco2 and a decrease in pH was notedin the REMI group (Figures 1 through 3). The lowestrespiratory rate and pH and highest Pco2 in anypatient at any stage occurred in one patient treatedwith REMI after 40 min of treatment. Her respira-tory rate was 7 breaths/min, Pco2 was 48 mm Hg,and pH was 7.31. REMI infusion was stopped andthe respiratory depression spontaneously resolvedafter several minutes. Maternal hemodynamics re-mained stable throughout the procedure. Similardoses and number of top-ups of ephedrine andphenylephrine were needed in both groups. Dura-tion of surgery was significantly longer in the DZPgroup, 80 (60 –90) minutes versus 60 (54 –71) min-utes in the REMI group.

REMI induced a significantly higher degree of fetalimmobilization, whereas DZP had little effect on fetalmobility as evaluated by subjective surgical and ob-jective ultrasound scores (Figs. 5 and 6). The numberof fetal gross body and limb movements decreasedfrom 18 � 3 to 2 � 1 at 40 min of surgery in the REMIgroup; this decrease was much less in the DZP group,from 17 � 4 to 12 � 4 at 40 min of surgery. Thesubjective appreciation of fetal immobilization by thesurgeon, who was blinded as to the medication, wasgood to excellent in 23 of 25 patients (92%) in the

REMI group, whereas this was good to excellent inonly 8 of 25 (32%) in the DZP group. No significantchanges in fetal heart were noted in either group. Noearly or late decelerations or fetal bradycardia wererecorded.

DiscussionThis randomized double-blind study in patients un-dergoing obstetrical endoscopic surgery demonstratesthat REMI induces excellent fetal immobilization andmaternal sedation during surgery, while DZP pro-vides less fetal immobilization and deeper maternalsedation.

Some in utero conditions are amenable to surgicalinterventions (1–4,19). At our institution, obstetric en-doscopy procedures are performed regularly. Mostcases are for treatment of twin-to-twin transfusionsyndrome because laser therapy has been proven to bebetter then amniodrainage (1,3,4). In addition, selec-tive feticide procedures in selected monochorionictwin pregnancies may require in utero endoscopic cordocclusion. These procedures usually do not requirematernal general anesthesia (19,20). General anesthe-sia in pregnancy is associated with a more frequentincidence of maternal mortality and morbidity (21),

Table 1. Demographic and Obstetrical Data in the Study Population: None of the Differences were Significant

Remifentanil group Diazepam group

Age (yr) 30.2 � 5.1 29.6 � 4.8Weight (kg) 70 � 11 71 � 14Height (cm) 167 � 5 167 � 8Gestational age at procedure (wk) 19.9 � 2.5 19.6 � 2.6Gestational age at delivery (wk) 34 (33–36) 33 (29.5–35)Coagulation of chorionic vessels for TTTs (%) 84 76Cord coagulation (%) 16 24Neonatal survival for TTTS patients (%) 71 63Survival of non-target fetus in case of cord coagulation (%) 75 67

Data are presented as a mean � sd; median and interquartile range or percentage of group total.TTTS � twin to twin transfusion syndrome.

Table 2. Data on maternal sedation, surgeon satisfaction with fetal immobility and surgical conditions, duration ofsurgery and the need for ephedrine and phenylephrine in the study population

Remifentanil group Diazepam group P value

Mean remifentanil infusion (�g/kg/min) 0.115 � 0.020* 0 �0.00001Diazepam dose (mg) 0 14.5 � 4.8* �0.00001OAA-score � 4 (number of patients) 1 11* �0.001Satisfaction score 1 or 2 as evaluated by surgeon (n) 23/25 8/25* �0.0001Duration of surgery (min) 60 (54–71) 80 (60–90)* 0.024Ephedrine (mg) 21 � 8 25 � 11 NSPhenylephrine (�g) 50 (0–275) 250 (50–500) NSEpedrine (number of patients) 22 21 NSPhenylephrine (number of patients) 10 12 NS

Data are presented as a mean � sd, median and interquartile range and number of patients.OAA/S � observer assessment of alertness scale; NS � not significant.* P � 0.05 versus remifentanil treated patients.


mainly as the result of airway problems. Most Euro-pean centers prefer local or regional anesthesia tech-niques for these cases. However, regional anesthetictechniques do not provide fetal immobilization or fetalanalgesia. Fetal movements may lead to fetal trauma,may hamper or prolong surgery, or may even result infailure to complete the planned surgery. Prolongation

of surgery may increase the risk of iatrogenic preterm,prelabor rupture of membranes (8,9).

To obviate these problems, we initially used IV DZPto obtain fetal immobilization. However, the effects onfetal mobility of IV maternal DZP were unpredictableand often disappointing, and maternal sedation wasprofound. In the present trial we confirmed this ob-servation, with only a small percentage of fetuses be-ing adequately immobilized. It has been shown that

Figure 1. Respiratory rate (breaths per minute) in patients receivingeither diazepam (DZP) or remifentanil (REMI) IV sedation. X-axis:BL � baseline measurement; 10–60 and 10A–60A � measurement10–60 min after start and, respectively, end (A) of surgery. In theREMI group, the REMI infusion was stopped at the end of surgery.In the DZP group the last bolus was given on clinical indication andno top-ups were administered after the end of surgery. * P � 0.05REMI versus DZP at each time point; ** P � 0.05 versus baselinewithin one group.

Figure 2. Arterial Pco2 in patients receiving either diazepam (DZP)or remifentanil (REMI) IV sedation. X-axis: BL � baseline meas-urement; 10–60 and 10A–60A � measurement 10–60 min after startand, respectively, end (A) of surgery. In the REMI group, the REMIinfusion was stopped at the end of surgery. In the DZP group thelast bolus was given on clinical indication and no top-ups wereadministered after the end of surgery. * P � 0.05 REMI versus DZPat each time point; ** P � 0.05 versus baseline within one group.

Figure 3. Arterial pH in patients receiving either diazepam (DZP) orremifentanil (REMI) IV sedation. X-axis: BL � baseline measure-ment; 10–60 and 10A–60A � measurement 10–60 min after startand, respectively, end (A) of surgery. In the REMI group, the REMIinfusion was stopped at the end of surgery. In the DZP group thelast bolus was given on clinical indication and no top-ups wereadministered after the end of surgery. * P � 0.05 REMI versus DZPat each time point; ** P � 0.05 versus baseline within one group.

Figure 4. Observer assessment of alertness (OAA/S) score of �4 inpatients undergoing endoscopic, intrauterine surgery with eitherdiazepam or remifentanil sedation. BL � baseline measurement;10A � measurement 10 min after the end of surgery; 20A � meas-urement 20 min after the end of surgery; 30A � measurement30 min after the end of surgery; 60A � measurement 60 min afterthe end of surgery. In the remifentanil group, the remifentanilinfusion was stopped at the end of surgery. In the diazepam groupno additional boluses of diazepam were administered after the endof surgery.


DZP crosses the placenta rapidly but that the fetalcapillary blood concentration varies considerably, atleast in term infants (22), and that neonatal effects arelargely unpredictable. It was also demonstrated thatthe transfer of DZP across the human placenta isslower in early pregnancy than during labor (23). Inaddition, there are concerns of DZP being associatedwith neurodevelopmental changes in neonates andcongenital abnormalities when used chronically (24–26). Administration of DZP outside the period of or-ganogenesis using a single bolus has never been asso-ciated with teratogenic effects. Furthermore, DZP doesnot provide fetal analgesia and fetal and maternalrecovery is slow after DZP administration.

We decided to use DZP as the control group in thepresent trial despite the possibility of using other moreshort-acting benzodiazepines. Theoretically, othermore short-acting benzodiazepines, such as midazo-lam, may yield more consistent and more controllablematernal sedation. However, placental passage andthus fetal immobilization remains unpredictable aswell (27,28). Placental passage of midazolam in preg-nant ewes is small, with a fetal/maternal plasma con-centration ratio of 0.15 (27). Also, in term pregnanciesthe placental transfer of midazolam is considerablyless than that of thiopental and REMI (15,28).

Remifentanil is a novel ultra-short-acting opioid forIV use that is clinically proposed for sedation duringsurgical interventions in the nonpregnant and preg-nant population (10–15). In general, opioids have alarge transplacental passage (29,30) and as a conse-quence produce fetal “sleep.” We therefore speculated

that REMI would provide excellent fetal immobiliza-tion. REMI rapidly and extensively crosses the pla-centa (umbilical vein/maternal artery ratio, 0.88) interm pregnancies (15). Other opioids have also beenshown to have a rapid and large transplacental pas-sage in early human gestation (29–31). Although nopharmacokinetic data on REMI are available at mid-gestation and our study similarly does not providesuch information, our observations clearly show thatREMI effectively crosses the placenta and causes fetalimmobilization.

In contrast to DZP, REMI has the potential, as doother opioids, to provide effective fetal analgesia afteraccidental direct stimulation (e.g., touching with en-doscopes). Therefore, it has been suggested that painrelief has to be provided during in utero interventionson the fetus from mid-gestation (20 weeks) on (32–34).Direct administration of fentanyl to the human fetushas been shown to block the fetal stress responseduring mid-gestational in utero interventions (35). Inour trial inadvertent touching of an immobilized fetusresulted in fetal “awakening.” Therefore, when fetalanalgesia or blunting of the fetal stress response isrequired, additional drugs (opioids and nondepolar-izing muscle relaxants) must be administered directlyto the fetus. It must be stressed, however, that fetalanalgesia is not generally required during in uteroprocedures on the placenta and cord (the proceduresperformed in the present trial), as direct fetal traumashould not occur.

Maternal sedation during lengthy or stressful inutero interventions is useful to relieve anxiety and

Figure 5. Visual Analog Scale (VAS) for fetal mobility (0 � fetus iscompletely immobile; 100 � baseline fetal mobility) in patientsreceiving either diazepam (DZP) or remifentanil (REMI) IV seda-tion. X-axis: BL � baseline measurement; 10–60 and 10A–60A �measurement 10–60 min after start and, respectively, end (A) ofsurgery. In the REMI group, the remifentanil infusion was stoppedat the end of surgery. In the DZP group the last bolus was given onclinical indication and no top-ups were administered after the endof surgery. * P � 0.05 REMI versus DZP at each time point; ** P �0.05 versus baseline within one group.

Figure 6. Number of gross body and limb movements during a5-min registration period at various time points throughout theprocedure. X-axis: BL � baseline measurement; 10–60 and 10A–60A � measurement 10–60 min after start and, respectively, end (A)of surgery. In the remifentanil (REMI) group, the remifentanil infu-sion was stopped at the end of surgery. In the diazepam (DZP)group the last bolus was given on clinical indication and no top-upswere administered after the end of surgery. * P � 0.05 REMI versusDZP at each time point.


improve patient cooperation. Especially in emotion-ally stressful situations, such as selective feticide,effective maternal sedation can be useful from apsychological viewpoint. In twin-to-twin transfu-sion syndrome, the mother usually has seriousdiscomfort from polyhydramnios, which is only re-lieved at the end of the endoscopic procedure. In thepresent trial, REMI produced adequate maternal se-dation, whereas DZP often resulted in sedation thatwas considered too deep. Unfortunately, as withany opioid, REMI was associated with mild respi-ratory depression. In our series, this never becameclinically relevant, as none of the patients experi-enced respiratory arrest or signs of severe respira-tory acidosis. The sedative and respiratory depres-sant effects of REMI were extremely short-lived.This is in line with previous investigations in vol-unteers after bolus or continuous IV infusions orREMI (36,37). When respiratory depression occurs,reduction of the REMI infusion or brief cessationrapidly restores maternal respiration.

REMI may be used to induce fetal immobilization inother diagnostic or interventional procedures. For exam-ple intrauterine transfusion through the umbilical cordmay benefit from IV maternal REMI administration tosedate the mother and immobilize the fetus. In thosecases when perforation of the fetal abdominal wall isrequired for intrahepatic vein transfusion, REMI wouldbe insufficient to provide adequate fetal analgesia andimmobilization. Direct fetal administration of opi-oids and muscle relaxants could be required.

Another application is for fetal magnetic resonanceimaging studies, when some degree of immobilizationmay be helpful. Despite advances in magnetic reso-nance imaging technology, fetal movements induceartifacts hampering diagnostic accuracy (38,39). Ben-zodiazepines have been used for these indications but,based on the present study, they may produce unre-liable fetal immobilization. In addition, because theywill result in lengthy maternal sedation, REMI may bea better alternative.

An alternative to REMI may be propofol, as it is aneffective, controllable maternal sedative. We decidednot to study propofol because it lacks analgesic prop-erties. Whether it provides similar fetal immobilizingproperties as REMI needs to be established.

We conclude that maternally administered REMI isa superior alternative to maternal DZP to induce ma-ternal sedation and fetal immobilization. Further stud-ies must be conducted to establish long-term effects ofREMI on the fetus and to establish its place in otherfetal diagnostic and therapeutic interventions.

The authors wish to express their sincere gratitude to the midwiferystaff of the labor and delivery ward of the UZ Leuven, where theseprocedures are routinely performed.

References1. Deprest JA, Gratacos E. Obstetrical endoscopy. Curr Opin Ob-

stet Gynecol 1999;11:195–203.2. De Lia JE, Cruikshak DP, Keye WR. Fetoscopic neodynium:YAG

laser occlusion of placental vessels in severe twin-twin transfu-sion syndrome. Obstet Gynecol 1990;75:1046–53.

3. Lewi L, Van Schoubroeck D, Gratacos E, et al. Monochorionicdiamniotic twins: complications and management options. CurrOpin Obstet Gynecol 2003;15:177–194.

4. Senat MV, Deprest J, Boulvain M, et al. A randomized trial ofendoscopic laser surgery versus serial amnioreduction for se-vere twin-to-twin transfusion syndrome at midgestation.N Engl J Med 2004;351:136–44.

5. Challis D, Gratacos E, Deprest J. Selective termination in mono-chorionic twins. J Perinat Med 1999;27:327–8.

6. Deprest J, Evrard V, Van Schoubroeck D. Fetoscopic cord liga-tion. Lancet 1996;384:890–1.

7. Rosen MA. Anesthesia for fetal surgery and other intrauterineprocedures. In: Chestnut DH, ed. Obstetric anesthesia, 3rd ed.Philadelphia: Elsevier–Mosby, 2004:96–109.

8. De Lia JE, Kuhlmann RS, Lopez KP. Treating previable twin-twin transfusion syndrome with fetoscopic laser surgery: out-comes following the learning curve. J Perinat Med 1999;27:61–7.

9. Deprest JA, Van Ballaer PP, Evrard VA, et al. Experience withfetoscopic cord ligation. Eur J Obstet Gynecol Reprod Biol 1998;81:157–64.

10. Machata AM, Gonano C, Holzer A, et al. Awake nasotrachealfiberoptic intubation: patient comfort, intubating conditions,and hemodynamic stability during conscious sedation withremifentanil. Anesth Analg 2003;97:904–8.

11. Sa Rego MM, Inagaki Y, White PF. Remifentanil administrationduring monitored anesthesia care: are intermittent boluses aneffective alternative to a continuous infusion? Anesth Analg1999;88:518–22.

12. Volmanen P, Akural EJ, Raudaskoski T, Alahuhta S. Remifen-tanil in obstetric analgesia: a dose-finding study. Anesth Analg2002;94:913–7.

13. Joo HS, Perks WJ, Kataoka MT, et al. A comparison of patientcontrolled sedation using either remifentanil or remifentanil-propofol for shock wave lithotripsy. Anesth Analg 2001;93:1227–32.

14. Lauwers M, Camu F, Breivik H, et al. The safety and effective-ness of remifentanil as an adjunct sedative for regional anesthe-sia. Anesth Analg 1999;88:134–40.

15. Kan RE, Hughes SC, Rosen MA, et al. Intravenous remifentanil:placental transfer, maternal and neonatal effects. Anesthesiol-ogy 1998;88:1467–74.

16. Missant C, Van Schoubroeck D, Deprest JA, et al. Remifental forfoetal immobilisation and maternal sedation during endoscopictreatment of twin-to-twin transfusion syndrome: a preliminarydose finding study. Acta Anaesth Belg 2004;55:239–44.

17. McLennan CE. Antecubital and femoral venous pressure innormal and toxemic pregnancy. Am J Obstet Gynecol 1943;45:568–91.

18. Chernik DA, Gillings D, Laine H, et al. Validity and reliability ofthe observer’s assessment of alertness/sedation scale: studywith intravenous midazolam. J Clin Psychopharmacol 1990;10:244–51.

19. Sebire NJ, Souka A, Skentou H, et al. Early prediction of severetwin-to-twin transfusion syndrome. Hum Reprod 2000;15:2008–10.

20. De Lia JE, Kuhlmann RS, Harstad TW, Cruikshank DP. Feto-scopic laser ablation of placental vessels in severe previabletwin-twin transfusion syndrome. Am J Obstet Gynecol 1995;172:1202–11.

21. Hawkins JL, Koonin LM, Palmer SK, Gibbs CP. Anesthesiarelated deaths during obstetric delivery in the United States,1979–1990. Anesthesiology 1997;86:277–84.

22. Bakke OM, Haram K. Time course of transplacental passage ofdiazepam. Clin Pharmacokinetics 1982;7:353–62.


23. Kanto J, Erkkola R. The feto-maternal distribution of diazepamin early human pregnancy. Ann Chir Gynaecol Fenniae 1974;63:489–91.

24. Saxen I, Saxen L. Association between maternal intake of diaz-epam and oral clefts. Lancet 1975;2:498.

25. Mehanny SZ, Abdel-Rahman MS, Ahmed YY. Teratogenic effectof cocaine and diazepam in CF1 mice. Teratology 1991;43:11–7.

26. Kellog CK. Benzodiazepines: influence on the developing brain.Prog Brain Res 1988;73:207–28.

27. Vree TB, Reekers-Keeting JJ, Fragen RJ, Arts TH. Placental transferof midazolam and its metabolite 1-hydroxymethylmidazolam inthe pregnant ewe. Anesth Analg 1984;63:31–4.

28. Bach V, Carl P, Ravlo O, et al. A randomised comparisonbetween midazolam and thiopental for elective cesarean sectionanesthesia: III. Placental transfer and elimination in neonates.Anesth Analg 1989;68:238–42.

29. Shannon C, Jauniaux E, Gulbis B, et al. Placental transfer offentanyl in early human pregnancy. Hum Reprod 1998;13:2317–20.

30. Cooper J, Jauniaux E, Gulbis B, et al. Placental transfer offentanyl in early human pregnancy and its detection in fetalbrain. Br J Anaesth 1999;82:929–31.

31. Krishna BR, Zakowski MI, Grant GJ. Sufentanil transfer in thehuman placenta during in vitro perfusion. Can J Anaesth 1997;44:996–1001.

32. Giannakoulopoulos X, Sepulveda W, Kourtis P, et al. Fetalplasma cortisol and beta endorphin response to intrauterineneedling. Lancet 1994;344:77–81.

33. Giannakopoulos X, Teixeira J, Fisk N, Glover V. Human fetaland maternal noradrenaline responses to invasive procedures.Ped Res 1999;45:494–9.

34. Anand KJS, Maze M. Fetuses, fentanyl, and the stress response.Anesthesiology 2001;95:823–5.

35. Fisk NM, Gitau R, Teixeira JM, et al. Effect of direct fetal opioidanalgesia on fetal hormonal and hemodynamic stress responseto intrauterine needling. Anesthesiology 2001;95:828–35.

36. Babenco HD, Conard PF, Gross JB. The pharmacodynamic effectof a remifentanil bolus on ventilatory control. Anesthesiology2000;92:393–8.

37. Nieuwenhuijs DJF, Olofsen E, Romberg RR, et al. Responsesurface modeling of remifentanil-propofol interaction on cardio-respiratory control and bispectral index. Anesthesiology 2003;98:312–22.

38. Levine D, Stroustrup Smith A, McKenzie C. Tips and tricks offetal MR imaging. Radiol Clin N Am 2003;41:729–45.

39. Luks FI, Carr SR, Ponte B, et al. Preoperative planning withmagnetic resonance imaging and computerized volume render-ing in twin-to-twin transfusion syndrome. Am J Obstet Gynecol2001;185:216–9.


remifentanil for fetal immobilization and maternal sedation

Education