EXPERIMENTAL Vasospasm of the Flap Pedicle: The Effect of 11 of the Most Often Used Vasodilating Drugs. Comparative Study in a Rat Model Petr Hyza, M.D., Ph.D. Libor Streit, M.D. Daniel Schwarz, Ing., Ph.D. Tomas Kubek, M.D. Jiri Veselý, M.D., Ph.D. Brno, Czech Republic Background: There has been no review study published yet comparing the effects of the vasodilating drugs that are most often used in clinical practice empirically. The aim of the authors' study was to perform this comparison and to select the drugs that are able to release vasospasm and the drugs that reduce vasospasm duration most effectively in an experimental model in vivo. Methods: Pedicled groin flaps were dissected in 300 male Wistar rats. Vasospasm was induced by tension applied on the pedicle in the axial direction using a 15-g weight. The blood perfusion of the flap was monitored using a laser Doppler device. The duration of vasospasm was defined as the time from the release of tension until blood flow began to rise. These times were detected using automated computerized detection. The effects of 11 different drugs were studied in 14 groups. The drugs were applied locally; some of them were tested in different concentrations or applied parenterally. Results: Ten percent magnesium sulfate reduced the duration of vasospasm most effectively (p < 0.01). Verapamil applied locally and also pentoxifylline applied parenterally were also very effective. In contrast, the duration of vasospasm was extended after local application of 2% lidocaine (p < 0.01). Conclusions: The authors concluded that 10% magnesium sulfate applied locally has the best ability to relieve surgically induced vasospasm because of the highest level of significance and reliability. The finding that local application of 2% lidocaine prolongs vasospasm may be surprising. (Plast. Reconstr. Surg. 134: 574e, 2014.) Microsurgical free tissue transfer is currently an important method in reconstructive surgery, and vasospasm is a common complication associated with flap dissection. Vasospasm is a local and persistent spastic contraction of smooth muscles in the vessels that is caused by pathologic stimulus. Vasospasm is a functional problem that usually resolves spontaneously. In case vasospasm is prolonged, the blood supply of tissues is significantly reduced and flap failure may occur.1-6 The pathogenesis of vasospasm is complex, and it is often associated with other pathologic entities, such as the mechanism of thrombosis, ischemia-reperfusion injury of the flap, or the no-reflow phenomenon.7-12 Thus, From the Department of Plastic and Aesthetic Surgery, St. Anne University Hospital; and the Institute of Biostatistics and Analysis, Masaryk University. Received for publication August 27, 2013; accepted March 13, 2014. Copyright © 2014 by the American Society of Plastic Surgeons DOI: 10.1097/PRS. 0000000000000570 pharmacologic interventions for vascular complications in microsurgery may focus on thrombosis (antiaggregants and anticoagulants), ischemia-reperfusion injury or the no-reflow phenomenon (drugs that modulate flap tolerance to ischemia), and vasospasm. We believe that no study comparing the effects of the most often used vasodilating drugs on mechanically induced vasospasm of the flap pedicle has been published yet. The main objective of our previous experiments was to develop an experimental model that would simulate reliably possible surgical causes of vasospasm induction during flap dissection. This model should also enable easy and standardized application of the drug while being well adapted for testing the drug within the study.13 The main aim of this study was (1) to compare the effects of different vasodilating drugs with the control group and also to compare their effect Disclosure: The authors have neither conflict of interest nor funding support for publishing this article. 574e www.PRSJournal.com Volume 134, Number 4 • Comparison of Vasodilating Drugs Table 1. Medications Used in the Experiment Route of Median Median Group Drug Name Name of Active Substance Administration Manufacturer t (sec) ř*» (sec) 1 Agapurin injection Pentoxifyllinum, 20 mg/ml Parenteral Zentiva a.s., Hlohovec, 138 936.67 Slovak Republic 2 Agapurin injection Pentoxifyllinum, 20 mg/ml Local Zentiva a.s., Hlohovec, 90 298 Slovak Republic 3 Prostavasin injection Alprostadilum, 4 jig/ml Parenteral Schwarz Pharma AG, Mon- 183 1161 heim, Germany 4 Perlinganit Glyceroli trinitras, 1 mg/ml Local Schwarz Pharma AG, Mon- 261 773 heim, Germany 5 Lidocaine 2% Lidocaini hydrochloridum, Local Egis Pharmaceuticals PLC, 1184 2217 20 mg/ml Budapest, Hungary 6 Magnesium Magnesii sulfas heptahydricus, Parenteral Hoechst-Biotika s.r.o., Mar- 158 577 sulfuricum 10% 0.1 g/ml tin, Slovak Republic 7 Magnesium Magnesii sulfas heptahydricus, Local Hoechst-Biotika s.r.o., Mar- 119 315 sulfuricum 10% 0.1 g/ml tin, Slovak Republic 8 Magnesium Magnesii sulfas heptahydricus, Local Hoechst-Biotika s.r.o., Mar- 180 617 sulfuricum 20% 0.2 g/ml tin, Slovak Republic 9 Marcaine 0.5% Bupivacaini hydrochloridum, Local AstraZeneca, UK, Ltd., 209 1105 5 mg/ml Macclesfield, Cheshire, United Kingdom 10 Mesocaine 1% Trimecaini hydrochloridum, Local Zentiva a.s., Praha, Czech 174 701 10 mg/ml Republic 11 Nipruss Sodium nitroprusside, 1 mg/ Local Schwarz Pharma AG, Mon- 205 602 ml heim, Germany 12 Papaverin Spofa Papaverini hydrochloridum, Local VUAB Pharma a.s., 187 433 injection 30 mg/ml Roztoky, Czech Republic 13 Lekoptin injection Verapamili hydrochloridum, Local Lek Pharmaceuticals d.d., 44 275 2.5 mg/ml Ljubljana, Slovenia 14 Prostavasin injection Alprostadilum, 4 jig/ml Local Schwarz Pharma AG, Mon- 251 900 heim, Germany 15 Control group 379 883 vasospasm duration; t time to hyperperfusion. between the groups, (2) to select the drugs that are able to reduce vasospasm duration, and (3) to select the drugs that reduce vasospasm duration most effectively. Our hypothesis was that application of the studied drugs will reduce the duration of mechanically induced vasospasm in the experimental model. MATERIAL AND METHODS Three hundred male Wistar rats were used. The median weight of the rats was 358.2 ± 51.5 g. The experiment was approved by the Ethics Committee at St. Anne University hospital in Brno and performed under standard conditions (i.e., temperature, 24° to 25°C; light conditions; sterility). The rats were anesthetized using ketamine (100 mg/kg) and xylazine (10 mg/kg). Ether was used at the beginning of anesthesia. All dissections were performed by the same surgeon. The effects of 11 different drugs were studied in 14 groups. The drugs were mostly applied locally, and the effects of the most potent drugs were tested also in different concentrations or applied parenterally/intraperitoneally (Table 1). No drug was applied in the control group. Dissection Technique The holder for the laser Doppler probe (small straight probe 407-1; PeriMed AB, Jarfalla, Sweden) was placed on the skin at a defined point (2 cm above and laterally from the genital tubercle). A skin paddle of 1.5 x 1.5 cm of the groin flap based on the superficial inferior epigastric artery was outlined. The flap pedicle was designed at the medial part of the flap. The pedicle was dissected sharply, leaving a thick fat and connective tissue layer surrounding the pedicle (Fig. 1). Two strips of perivascular connective tissue ("perivascular flaps") with a length of 0.5 cm that remained distally connected with the pedicle were raised (Fig. 2). One of them comprised the inferior epigastric nerve. A weight of 15 g was attached to the above-described perivascular flaps by the atraumatic monofilament nylon suture (Premilene 6-0; B. Braun Melsungen AG, Melsungen, Germany). All dissections were performed by the same surgeon, as in the control group. Timing of the Experiment, Induction of Vasospasm, and Application of Drugs After dissection, the rat was moved to the measuring area and the flap was left resting for 575e Plastic and Reconstructive Surgery • October 2014 Fig. 1. Flap dissection. The flap was dissected out sharply. The pedicle was neither touched by instruments nor pulled or traumatized in any other way, to prevent induction of vasospasm. Fig. 2. Pedicle dissection. Two perivascular flaps (stripes) were created on both sides of the pedicle using sharp dissection by scissors. 25 minutes. Then, the laser Doppler probe was attached to the holder, the recording of flap perfusion was turned on, and the flap was left for another 5 minutes resting. Next, the preprogrammed script was begun; the basal level of the flap perfusion was measured for 5 minutes. The 15-g weight was gently hung on the block and the pedicle was pulled for the next 5 minutes and then released. Administration of drugs began immediately after application of the weight. The drugs were applied locally by continuous manual application of drops on the flap pedicle; the rate was 0.5 ml/minute, and dropping continued for 2 minutes after releasing the weight or the drugs were administered systemically using a 2-ml bolus of drug solution given intraperitoneally. Following the stimulus, the perfusion curve was recorded for 30 minutes. The time of stimulation termination was set as 0 (Table 2). Each step of the experiment was defined by the appropriate time period using PeriSoft for Windows software script (PeriSoft for Windows 2.5; PeriMed) and an appropriate voice signal allowed us to control the timing of the experiment. Laser Doppler Perfusion Measurements A laser Doppler flowmeter (PeriFlux system 5000; PeriFlux 5010 Laser Doppler Perfusion Monitor unit, small straight probe 407-1 and mounting ring PH 07-4; PeriMed) was used to measure blood flow through the pedicle in all groups. The Periflux 5010 Laser Doppler Perfusion Monitor unit was attached to the probe just before the start of the measurement through a simple holding frame to prevent signal errors. A laptop personal computer with PeriSoft for Windows software was used as a signal recorder. All of the values from the laser Doppler flowmeter were exported from the PeriSoft software into an ASCII text formatted file for each measurement. The sampling period was set up to T = 0.03 second. Each measurement resulted in a blood perfusion signal, which showed flap perfusion dynamics for the typical shape of the signal (Fig. 3). Blood perfusion signals were described by two time intervals, which were extracted with the use of an automated signal delineation method. The first and the most clinically important parameter was vasospasm duration, the time interval during which the flap was not perfused because of persistent vasospasm. The second, rather auxiliary parameter was time to hyperperfusion, the time interval between the termination of stimulus and the time point when the highest hyperperfusion level was reached. Both parameters, vasospasm duration and time to hyperperfusion, were statistically compared between particular groups. Signal Processing The Savitzky-Golay polynomial filter was used to smooth the raw measured signals that were 576e Volume 134, Number 4 • Comparison of Vasodilating Drugs Table 2. Timing of the Measurements Time Step of the Experiment -45 min Attachment of the probe fixation ring to the flap, beginning of dissection -40 min Flap left resting for next 25 minutes -15 min Attachment of the probe to the fixation ring and beginning of measurement of flap perfusion -10 min Beginning of script, measurement of basal level of flap perfusion -5 min Beginning of stimulation, hanging of 15-g weight, beginning of local application of drug by dropping or bolus intraperitoneal injection 0 min Termination of the stimulation, continu- ous dropping, continuous measurement of perfusion 2 min Termination of the dropping, continu- ous measurement of perfusion 30 min Termination of the measurement, disconnection of the probe from the flap, and termination of the experiment corrupted by impulse noise. This preprocessing step resulted in enhanced signals on which our algorithm for automated delineation of the signals was applied, to extract parameters relevant to the focus of our study. The delineation algorithm was designed and implemented in a form of a set of functions and scripts in Matlab (MathWorks, Inc., Natick, Mass.), so that it simulated analytical thinking of an expert providing manual delineation. Because of the remarkable high shape variability in the signals—mainly during the vasospasm and reperfusion phases (Fig. 4)—we were describing each signal by two time intervals as mentioned above. Whereas the time interval vasospasm duration represented duration of the vasospasm, the time interval time to hyperperfusion represented how rapid or slow flap reperfusion was. The other parameters described in Figures 3 and 4 represent various blood perfusion levels during each measurement. The most important one was the threshold blood perfusion level h, as it served for extracting the time interval vasospasm duration. The threshold L was determined as the sum of the blood perfusion level after stimulus termination (L ) and the tenth of the reperfusion wave amplitude given by the difference between levels L, „ and L: hp s L, ■L0 + hp s 10 (1) where L was determined as the tenth percentile of signal values during the stimulation of vasospasm, LQ was determined as the tenth percentile of signal values after the end of stimulation and before the reperfusion wave reached half of its amplitude, and L was determined as the median of signal values during 3 seconds around the time point in which the maximum blood hyperperfusion value was found. The median filtration was preferred before simple averaging, as the blood perfusion signal values were not normally distributed. RESULTS The vasodilatory effects of the drugs were evaluated in 14 groups and compared with the control group. Some of the 280 measurements had to be removed from further statistical processing: six measurements because of an excessive number of artifacts in the signals; three measurements because of errors in the stimulus application (operating personnel mistake); and three measurements because of partial absence of data (spontaneous shutdown of the recorder or the computer). No reperfusion was observed in the interval of 30 minutes after releasing the weight in 13 measurements, and the signals had atypical shapes as well. In such cases, time intervals were determined to be as follows for subsequent data processing: vasospasm duration, 30 minutes; and time to hyperperfusion, 60 minutes. The existence of differences between the groups was tested by nonparametric analysis of variance (Kruskal-Wallis) tests. Significant differences in the time intervals were detected (p = 1.034 x 10"10 for vasospasm duration and p = 1.772 x 10-7 for time to hyperperfusion). It means that at least two of the group medians of the time intervals vasospasm duration and time to hyperperfusion differed significantly. Distributions of vasospasm duration and time to hyperperfusion values across the groups are shown in the box and whisker plots (Figs. 5 and 6). Additional exploration of the differences between groups was needed to provide information, whether the medians were significantly different from each other. Post hoc Wilcoxon rank sum tests were used together with the Bonferroni method of correction for multiple comparisons. The results of the post hoc analysis are shown in Tables 3 and 4. The reported p values were determined according to the Bonferroni correction method multiplied by the total number of tests performed (n = 105), to prevent false-positive results. Evaluation of Vasospasm Duration (Bonferroni Correction, n = 105) Vasospasm duration was significantly shorter compared with the control group in groups 1 577e Plastic and Reconstructive Surgery • October 2014 Fig. 3. Typical shape of the perfusion curve and extraction of the time intervals vasospasm duration and time to hyperperfusion. The following time intervals could be differentiated on the curve: 7, resting perfusion in the flap after its dissection; 2, segment of significantly reduced levels up to an arrest of perfusion of the flap was defined from the application of the stimulus to the vasospasm, applying the weight until it was released at the end of stimulus (point ES); 3, the segment of significantly reduced perfusion, which represents persisting vasospasm of the vascular pedicle. At the beginning of this segment was regularly observed peak artifact (A) that developed by the movement of the probe until disconnection of the weight. At the end of this segment occurs restoration of flap perfusion and the curve changes at the end of vasospasm (point EV) into (4) a segment with a further increase of flap perfusion, which continues to maximal perfusion values at the point of flap hyperperfusion (F/-/P).The flap turned pink already at the lower third of the increasing part of the curve, whereas in the area of maximal values there was already compensatory hyperemia of the flap that manifested with redness (point FHP). 5, Subsequently, slower stabilization of flap perfusion to the level of normal perfusion and pink color of the flap occurred. Gray solid lines, raw measured signals; black solid lines, signals after preprocessing; gray dashed lines, delineated signal levels (horizontal) and time intervals (vertical). A signal delineation algorithm was used to extract automatically the following parameters relevant to the study: Ls, blood perfusion level during vasospasm stimulation;L0, blood perfusion level after stimulus termination;/.^, the highest blood hyperperfusion level; Lss, steady-state blood reperfusion level; t0, termination of the vasospasm stimulus (recorded manually); time interval between stimulus termination and reperfusion wave onset, defined as point ES to point EV;thp, time interval between stimulus termination and the time point when the highest hyperperfusion level was reached, defined as point ES to point FHP. (pentoxifylline administered parenterally), 7 (10% magnesium sulfate applied locally), and 13 (verapamil applied locally). Vasospasm was significantly longer in group 5 (2% lidocaine applied locally). Evaluation of Time to Hyperperfusion (Bonferroni Correction, n= 105) Time to hyperperfusion was significantly shorter only in group 7 (10% magnesium sulfate applied locally). The results mentioned 578e Volume 134, Number 4 • Comparison of Vasodilating Drugs 1 5. -a - L, - k ?o fv fep Time -1- •-1-1- liss I, - L i i — ! 1 1 k t, thp Time - —i—<-1—-1-1- r L, to t,, 4 Fig. 4. Blood perfusion signal examples demonstrating the variability of perfusion parameters values (Ls, L0, Lhp, Lss, and Lt) in different subjects. It is also obvious that very similar duration of vasospasm time could be associated with very different times to hyperperfusion (cenferand below). above were calculated using Bonferroni correction for all post hoc tests (n = 105) performed after nonparametric analysis of variance. If the groups with drug were compared only with the control group using the post hoc analysis performed with multiple Wilcoxon tests, more positive results were obtained, because Bonferroni correction was calculated for a lower number of tests performed (n = 14). The results of post hoc analysis, in which all drug groups are compared with one control group only, are reported in Table 5. Evaluation of Vasospasm Duration (Bonferroni Correction, n = 14) Vasospasm duration differed significantly from the control group in groups 1, 2, 5, 6, 7, 10, and 13 at the 5 percent significance level 5 and in groups 1, 5, 7, and 13 at the 1 percent significance level. Vasospasm duration was shortened in groups 1 and 2 (pentoxifylline applied parenter-ally and locally), 6 and 7 (10% magnesium sulfate applied parenterally and locally), 10 (1% trime-caine applied locally), and 13 (verapamil applied locally), and prolonged in group 5 (2% lidocaine applied locally). Evaluation of Time to Hyperperfusion (Bonferroni Correction, n = 14) The time to hyperperfusion in groups 5 and 7 differed significantly from the control group at the 5 percent significance level. Time to hyperperfusion was longer in group 5 and shorter in group 7. DISCUSSION The pathogenesis of vascular complications in microsurgery is often a complex process in which three basic pathologic entities may play different roles: thrombosis, ischemia reperfusion injury or the no-reflow phenomenon, and vasospasm. One pathologic entity is frequently manifested more than the other. The personal experience of the microsurgeon seems to be the most important factor in management of the vascular complication and in specific therapy selection. The most often used drugs in the prevention and treatment of vascular complications include antiaggregants and anticoagulants. Others are used less frequently. In certain cases, the selection of a proper pharmacologic management that is specifically focused on vasospasm may save flap vitality. 579e Plastic and Reconstructive Surgery • October 2014 1300 1600 1400 1200 1000 800 600 400 200 0 ; i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Fig. 5. Distributions of the time interval vasospasm duration (tv) values across the groups. On each box, the central mark is the median, the edges of the box are the 25th and 75th percentiles, the whiskers extend to the most extreme data points not considered outliers, and outliers are plotted individually. 4500 4000 3500 3000 2500 1*2000 1500 1000 500 0 - t- „. T T 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Fig. 6. Distributions of the time interval time to hyperperfusion (th) values across the groups. On each box, the central mark is the median, the edges of the box are the 25th and 75th percentiles, the whiskers extend to the most extreme data points not considered outliers, and outliers are plotted individually. An ideal pharmacologic agent should act in the pedicle and in the microcirculation of a flap, even after topical application.14 A considerable number of medicaments are routinely used in clinical practise. Their efficacy on surgically induced vasospasm has been disputable in all of the cases. Differentiation of vasospasm that subsides spontaneously from that resulting from a specific therapy seems to be difficult in clinical practice. This was the main reason for creating this study, with a primary aim of comparing the effect of different commonly available vasodilating drugs on surgically induced vasospasm. The goal of subsequent analysis was to select those that are effective and also to choose the most effective one. To achieve this, we compared the distributions of vasospasm duration and time to hyperperfusion between all groups and to the distributions of the time parameters in the control group. Pentoxifylline applied parenterally and 10% magnesium sulfate and verapamil applied locally were evaluated as the most potent vasodilators. 580e Volume 134, Number 4 • Comparison of Vasodilating Drugs Table 3. Results of Post Hoc Wilcoxon Tests for Vasospasm Duration Showing Statistically Significant Differences between Specific Groups Group Group 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 1 1 0.382 0.005f 1 1 1 1 1 1 1 1 1 0.037f 2 1 0.194 0.003f 1 1 1 1 1 0.260 1 1 1 0.163 3 1 0.020-j- 1 1 1 1 1 1 1 0.116 1 1 4 0.016f 1 0.032t 1 1 1 1 1 0.006f 1 0.785 5 0.006f 0.003f 0.005f 0.018f 0.008f 0.047f 0.004f O.OOlf 0.091 0.045f 6 1 1 1 1 1 1 0.914 1 0.360 7 1 0.804 1 0.142 1 1 0.950 0.023f 8 1 1 1 1 0.566 1 1 9 1 1 1 0.184 1 1 10 1 1 0.836 1 0.237 11 1 0.009f 1 1 12 1 1 0.381 13 0.107 0.041f 14 1 *The reported Rvalues were corrected with the Bonferroni method (i.e., multiplied by the total number of tests performed; n = 105). fSignificant differences. Table 4. Results of Post Hoc Wilcoxon Tests for the Time to Hyperperf usion Showing Statistically Significant Differences between Specific Groups Group Group 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 1 1 1 0.017f 1 1 1 1 1 1 1 1 1 1 2 1 1 O.Olf 1 1 1 1 1 1 1 1 1 1 3 1 0.02t 1 0.026f 1 1 1 1 1 0.960 1 1 4 O.Olf 1 0.008-j- 1 1 1 1 1 1 1 1 5 O.Olf 0.002t 0.009f 1 0.030f 0.080 0.008f 0.002t 0.182 0.207 6 0.785 1 1 1 1 1 1.0 1 1 7 1 0.111 1 0.1172 1 1 0.382 0.026f 8 1 1 1 1 1 1 9 1 1 1 1 1 1 10 1 1 1 1 1 11 1 1 1 1 12 1 1 1 13 1 1 14 1 *The reported Rvalues were corrected with the Bonferroni method (i.e., multiplied by the total number of tests performed; n = 105). fSignificant differences. Although medians for vasospasm duration and time to hyperperfusion for verapamil and pentoxifylline were shorter than medians for 10% magnesium sulfate, differences expressed in p values were more significant for 10% magnesium sulfate than for verapamil and pentoxifylline and were also the most significant of all. This is true for both parameters (vasospasm duration and time to hyperperfusion) and for both Bonferroni corrections. Therefore, the efficiency of 10% magnesium sulfate applied locally can be considered as the most reliable. Magnesium sulfate is used especially by neurologists and neurosurgeons in the treatment of cerebral vasospasm and other diseases.15"18 However, the positive effect of magnesium sulfate has not yet been reliably demonstrated in microsurgery. The mechanism of action of magnesium sulfate has not been clarified, although its vasodilating features have been well described in vitro19-22 and in vivo.23-26 Pentoxifylline is a xanthine derivative that activates smooth muscle relaxation directly or by inhibition of phosphodiesterase. Improvement of skin-flap survival was observed after using pentoxifylline in rats.2728 We used pentoxifylline empirically several times parenterally after free flap transfer, with a promising effect on relieving vasospasm; this treatment was well tolerated by the patients. Nimodipine (L-type calcium channels) improves flap survival in randomized experimental 581e Plastic and Reconstructive Surgery • October 2014 study in rats.29 Weinzweig et al. compared the effect of verapamil and nifedipine on vasospasm induced by cold. Verapamil was evaluated as more effective.30 Verapamil seems to be less suitable for wider clinical use because of its cardiac side effects. Lidocaine and other local anaesthetics have often been used as vasospasm-relieving therapy. Several studies confirmed local and dose-dependent vasodilating effect of lidocaine.31-34 Beek-man et al. examined the effect of lidocaine in different concentrations on the vasospasm of rat-tail artery induced by local administration of ergotamine. The efficacy of lidocaine increased proportionally up to the concentration of 12%. Clinically most often used 2% lidocaine had just a slight vasodilating effect, and tissue toxicity was observed after the use of 20% lidocaine.33 Ohta et al. compared the vasodilating effect of lidocaine in concentrations of 2%, 4%, 10%, 20%, and 40%. Vasospasm of the rat femoral artery was produced by administration of fresh blood from another rat. They did not find any effect of 2% lidocaine; its efficiency gradually increased up to the concentration of 20%.34 Kim et al. demonstrated the vasodilating potential of 2% lidocaine, which is a concentration used clinically, in the porcine gastroepiploic artery.35 In contrast, the article by Evans et al.14 confirmed that 2% lidocaine acts as a spasmolytic agent in a rabbit carotid model. The safety-related dose of lidocaine is disputable, although Johnstone et al. demonstrated the safety of topical use of Table 5. Results of Post Hoc Wilcoxon Tests for Vasospasm Duration and Time to Hyperperf usion Showing Statistically Significant Differences between Control Group and Particular Drug Groups P Group tv 1 0.0050f l 2 0.0218f 0.2579 3 0.5501 1 4 0.1046 3.1504 5 0.0060f 0.0276f 6 0.0480-j- 0.8987 7 0.0030f 0.0034f 8 0.1474 0.8875 9 1 1 10 0.0316f 1 11 2.1870 1 12 0.0508 0.2222 13 0.0054f 0.3487 14 4.0031 1 vasospasm duration; £ time to hyperperfusion. *The reported p values were corrected with the Bonferroni method (i.e., multiplied by the total number of tests performed; n= 14). fSignificant differences. 4% lidocaine in microsurgical procedures.36 The local anaesthetics tested in our study, except for trimecaine, did not cause significant shortening of vasospasm duration or time to hyperperfusion. In our study, 2% lidocaine even significantly prolonged vasospasm duration. We can conclude that local anesthetics with the exception of trimecaine have only a minor vasodilating effect on surgically induced vasospasm. We assume that the use of 2% lidocaine as a spasmolytic agent should be reevaluated in microsurgery. Wang et al. studied the action of sodium nitroprusside on vasospasm induced by ischemia and reperfusion within the circulation of the cremaster muscle in a rat. Sodium nitroprusside was evaluated as a potent spasmolytic agent.37'38 Price and Pearl studied the effect of locally applied nitroglycerin cream and intravenous allopurinol on an excessively long skin flap in a rat. Both medicaments enhanced flap survival separately, but their combination did not lead to higher parameters of flap survival.39 The effect of local application of nitroglycerin had been disputable in other experimental studies.4041 The effectiveness of papaverine was examined and confirmed in several comparative experimental studies.14'32 42 Alprostadil (prostaglandin E ) is a potent vasodilator and antiaggregant. Chen et al. confirmed its ability to relieve vasospasm and compared it with 1 % lidocaine in the rabbit.43 The use of the automated delineation algorithm prevented us from another source of measurement uncertainty and thus ensured maximum objectivity and accuracy in obtaining the values of the two parameters vasospasm duration and time to hyperperfusion selected to describe the signals. An alternative use of only one time interval to extract the useful information from the signals would result in the inability to distinguish between the duration of vasospasm and the flap reperfusion speed. The nonredundancy of our signal description is also demonstrated in the example signals (Fig. 4). Nearly the same short duration of vasospasm in the examples shown in Figure 4, center and below do not necessarily imply the same reperfusion speed (time to hyperperfusion) of the flaps. However, the time interval vasospasm duration was considered in our study to have a higher importance in comparison of lengths of vasospasm. Perfusion of the flap near level L on the perfusion curve could be considered sufficient for survival of the flap with regard to the clinical correlation with the color of the flap (the 582e Volume 134, Number 4 • Comparison of Vasodilating Drugs flap was turning pink). At level L , the flap was red and compensatory hyperemia was observed. At level L , perfusion of the flap was already optimal. The time when the flap achieves this hyperperfusion (time to hyperperfusion) is influenced not only by the duration of vasospasm but also by metabolic and other factors. Therefore, two different flaps with the same duration of vasospasm could have different values of time to hyperperfusion (Fig. 4). This corresponds also to a higher variability of the parameter time to hyperperfusion compared with vasospasm duration (Figs. 5 and 6). The time interval vasospasm duration therefore corresponds to the duration of vasospasm, whereas the parameter time to hyperperfusion shows the time required to achieve the peak of compensatory mechanisms. CONCLUSIONS Drugs that reduce the duration of vasospasm in experimental model conditions after its induction by axial tension most effectively (p < 0.05) include the following (Table 5): 10% magnesium sulfate applied locally, pentoxifylline applied parenterally, and verapamil applied locally. The following drugs are also effective: pentoxifylline applied locally, 10% magnesium sulfate applied parenterally, and 1% trimecaine applied locally. Significant prolongation of vasospasm was observed after administration of 2% lidocaine. The most effective drug was 10% magnesium sulfate applied locally, which shortened the vasospasm duration most significantly (p = 0.0030) and the time to hyperperfusion (p = 0.0034) (Table 5). Therefore, we used 10% magnesium sulfate for subsequent experiments in an analogic porcine model to verify its effectiveness. Although we cannot draw any general conclusions for human medicine, this study may help microsurgeons to review their opinion regarding pharmacologic treatment of vasospasm and perhaps incorporate magnesium sulfate into the protocols for therapy of vasospasm in the future. In contrast, we assume that the use of 2% lidocaine as a spasmolytic agent should be reevaluated in microsurgery. LiborStreit, M.D. Department of Plastic and Aesthetic Surgery St. Anne University Hospital Berkova 34 612 00 Brno, Czech Republic streit@fhusa.cz ACKNOWLEDGMENT This study was supported by a grant from Internal Grant Agency of the Ministry of Health, Czech Republic (IGA-NR 8368-5). REFERENCES 1. Acland RD. Factors that influence success in microvascular surgery. In: Serafin D, Buncke HJ, eds. Microsurgical Composite Tissue Transplantation. St. Louis: Mosby; 1979:76-82. 2. Puckett CL, Winters RR, Geter RK, Goebel D. Studies of pathologic vasoconstriction (vasospasm) in microvascular surgery. J Hand Surg Am. 1985;10:343-349. 3. Veselý J, Samohýl J, Barinka L, Němec A. Tissue shock in free flaps in the experiment on the rat: Significance, classification and effect (in German). Handchir Mikrochir Plast Chir. 1987;19:269-272. 4. Veselý J, Samohyl, J, Barinka L, Němec A, Smrčka V. Spastic complications in free flap transfers. Rozhl Chir. 1990;69:682-688. 5. Weinzweig N, Gonzalez M. Free tissue failure is not an all-or-none phenomenon. Plast Reconstr Surg. 1995;96:648-660. 6. Seaber AV. Experimental vasospasm. Microsurgery 1987;8:234-241. 7. Saetzler RK, Lehr HA, Barker JH, Kamler M, Galla TJ, Messmer K. Visualization of nutritive perfusion following tourniquet ischemia in arterial pattern skin flaps: Effect of vasoactive medication. Plast Reconstr Surg. 1994;94:652-660. 8. Hjortdal VE, Sinclair T, Kerrigan CL, Solymoss S. Venous ischemia in skin flaps: Microcirculatory intravascular thrombosis. Plast Reconstr Surg. 1994;93:366-374. 9. Hjortdal VE, Sinclair T, Kerrigan CL, Solymoss S. Arterial ischemia in skin flaps: Microcirculatory intravascular thrombosis. Plast Reconstr Surg. 1994;93:375-385. 10. Khalil AA, Aziz FA, HallJC. Reperfusion injury. Plast Reconstr Surg. 2006;117:1024-1033. 11. May JWJr, Chait LA, O'Brien BM, Hurley JV. The no-reflow phenomenon in experimental free flaps. Plast Reconstr Surg. 1978;61:256-267. 12. Feng LJ, Berger BE, Lysz TW, Shaw WW. Vasoactive prostaglandins in the impending no-reflow state: Evidence for a primary disturbance in microvascular tone. Plast Reconstr Surg. 1988;81:755-767. 13. Hyza P, Streit L, VeselyJ, Schwarz D. New experimental model for studying vasospasm and implications for atraumatic handling of the vessels. In: Book of Abstracts, International Congress of Plastic Surgery, 2007, Brno. Brno: TA-Service; 2007:40. 14. Evans GR, Gherardini G, Gurlek A, et al. Drug-induced vasodilation in an in vitro and in vivo study: The effects of nicardipine, papaverine and lidocaine on the rabbit carotid artery. Plast Reconstr Surg. 1997;100:1475-1481. 15. Boet R, Mee E. Magnesium sulfate in the management of patients with Fisher Grade 3 subarachnoid hemorrhage: A pilot study. Neurosurgery 2000;47:602-606; discussion 606-607. 16. Zausinger S, Westermaier T, Baethmann A, Steiger HJ, Schmid-Elsaesser R. Neuroprotective treatment paradigms in neurovascular surgery: Efficacy in a rat model of focal cerebral ischemia. Acta Neurochir Suppl. 2001;77:259-265. 17. Veyna RS, Seyfried D, Burke DG, et al. Magnesium sulfate therapy after aneurysmal subarachnoid hemorrhage. JNeurosurg. 2002;96:510-514. 18. Schmid-Elaeser R, Kunz M, Zausinger S, Prueckner S, Briegel J, Steiger HJ. Intravenous magnesium versus 583e Plastic and Reconstructive Surgery • October 2014 nimodipine in the treatment of patients with aneurysmal subarachnoid hemorrhage: A randomized study. Neurosurgery 2006;58:1054-1056. 19. Altura BT, Altura BM. Withdrawal of magnesium causes vasospasm while elevated magnesium produces relaxation of tone in cerebral arteries. Neurosci Lett. 1980;20:323-327. 20. Altura BM, Altura BT. Magnesium ions and contraction of vascular smooth muscle. Fed Proc. 1981;40:2672-2679. 21. Kimura T, Yasue H, Sakaino N, Rokutanda M, Jouga-Saki M, Araki H. Effects of magnesium on the tone of isolated human coronary arteries. Circulation 1989;79:1118-1124. 22. Noguera MA, D'Ocon MP. Modulatory role of magnesium on the contractile response of rat aorta to several agonists in normal and calcium-free medium. J Pharm Pharmacol. 1993;45:697-700. 23. Chi OZ, Pollak P, Weiss HR. Effects of magnesium sulfate and nifedipine on regional cerebral blood flow during middle cerebral artery ligation in the rat. Arch Int Pharmacodyn Ther. 1990;304:196-205. 24. DiPette DJ, Simpson K, Guntupalli J. Systemic and regional hemodynamic effect of acute magnesium administration in the normotensive and hypertensive state. Magnesium 1987;6:136-149. 25. Kemp PA, Gardiner SM, Bennett T, Rubin PC. Magnesium sulphate reverses the carotid vasoconstriction caused by endothelin-I, angiotensin II and neuropeptide-Y, but not that caused by NG-nitro-L-arginine methyl ester, in conscious rats. Clin Sei (Land.) 1993;85:175-181. 26. Kemp PA, Gardiner SM, March JE, Bennett T, Rubin PC. Effects of NG-nitro-L-arginine methyl ester on regional haemodynamic responses to MgS04 in conscious rats. Br J Pharmacol. 1994;111:325-331. 27. Monteiro DT, Santamore WP, Nemir P Jr. The influence of pentoxifylline on skin-flap survival. Plast Reconstr Surg. 1986;77:277-281. 28. Nieto CS, Garcia MJS, Garcia PB. A comparative study on the effect of various pharmacological agents on the survival of skin flaps in the rat. Br JPlastSurg. 1992;45:113-116. 29. Stark GB, Dorer A, Jaeger K, Narayanan K. The influence of the calcium channel blocker nimodipine on flap survival. Ann Plast Surg. 1989;23:306-309. 30. Weinzweig N, Lukash F, Weinzweig J. Topical and systemic calcium channel blockers in prevention and treatment of microsurgical spasm in a rat epigastric island flap model. Ann Plast Surg. 1999;42:320-326. 31. Puckett CL, Winters RR, Geter RK, Goebel D. Studies of pathologic vasoconstriction (vasospasm) in microvascular surgery. J Hand Surg Am. 1985;10:343-349. 32. Hou SM, Seaber AV, UrbaniakJR. Relief of blood-induced arterial vasospasm by pharmacologic solutions. J Reconstr Microsurg. 1987;3:147-151. 33. Beekman WH, Sluimers JE, Kort WJ, van der Meulen JC. Resolution of experimental microvascular vasoconstriction in rats by topical application of lidocaine hydrochloride in varying concentrations. Ann Plast Surg. 1988;21:570-575. 34. Ohta I, Kawai H, Kawabata H, Masada K, Ono K. Topical use of Xylocaine for relieving vasospasm: Effect of concentration./ Reconstr Microsurg. 1991;7:205-208. 35. Kim DC, Chen D, Perrotta VJ, et al. Porcine gastroepiploic artery as an in vitro experimental model to study vasodilators in microsurgery. Ann Plast Surg. 1996;36:502-507. 36. Johnstone RE, Wax MK, Bishop DJ, Chafin JB. Large doses of topical lidocaine during microvascular surgery are not associated with toxic blood concentrations. Anesthesiology 1995;82:593-596. 37. Wang WZ, Anderson G, Firrell JC. Arteriole constriction following ischemia in denervated skeletal muscle. J Reconstr Microsurg. 1995;11:99-106. 38. Wang WZ, Anderson G, Fleming JT, et al. Lack of nitric oxide contributes to vasospasm during ischemia/reperfu-sion injury. Plast Reconstr Surg. 1997;99:1099-1108. 39. Price MA, Pearl RM. Multiagent pharmacotherapy to enhance skin flap survival: Lack of additive effect of nitroglycerin and allopurinol. Ann Plast Surg. 1994;33:52-56. 40. Rohrich RJ, Cherry GW, Spira M. Enhancement of skin-flap survival using nitroglycerin ointment. Plast Reconstr Surg. 1984;73:943-948. 41. Nichter LS, Sobieski MW, Edgerton MT. Efficacy of topical nitroglycerin for random-pattern skin-flap salvage. Plast Reconstr Surg. 1985;75:847-852. 42. Ueda K, Harii K. Comparative study of topical use of vasodilating solutions. Scand J Plast Reconstr Surg Hand Surg. 2003;37:201-207. 43. Chen LE, Seaber AV, UrbaniakJR. Vasodilator action of prostaglandin El on microcirculation of rat cremaster muscle. Microsurgery 1990;11:204-208. 584e