- Email: [email protected]
Effects of Repeat Bolus Ethanol Injections on Cardiopulmonary Hemodynamic Changes during Embolotherapy of Arteriovenous Malformations of the Extremities
Effects of Repeat Bolus Ethanol Injections on Cardiopulmonary Hemodynamic Changes during Embolotherapy of Arteriovenous Malformations of the Extremities Byung Seop Shin, MD, Young Soo Do, MD, Hyun Sung Cho, MD, Dong Ik Kim, MD, Tae Soo Hahm, MD, Chung Su Kim, MD, Justin Sangwook Ko, MD, Si Ra Bang, MD, Kwang Bo Park, MD, Sung Ki Cho, MD, Hong Suk Park, MD, and Seonwoo Kim, PhD
PURPOSE: This study was designed to investigate the effects of repeat bolus absolute ethanol injections on cardiopulmonary hemodynamic changes during ethanol embolotherapy of inoperable congenital arteriovenous (AV) malformations in the extremities. MATERIALS AND METHODS: Thirty adult patients (14 male, 16 female; age range, 22–51 years) who underwent ethanol embolotherapy of extremity AV malformations were enrolled in the study. A pulmonary artery catheter was used to measure hemodynamic profiles at baseline (Tbaseline), immediately before (Tpre) and after (Tpost) a bolus injection of absolute ethanol, at the time of the maximum mean pulmonary arterial pressure (PAP) value during a session (Thighest), 10 minutes after the final injection (Tfinal), and after restoration of spontaneous breathing (Tresp). RESULTS: The systolic, mean, and diastolic PAP (P < .01, P < .01, and P < .01, respectively) and the systemic vascular resistance index (P < .05) and pulmonary vascular resistance index (PVRI; P < .05) of Thighest and Tresp were significantly higher than values for Tfinal. The volume of a single bolus injection of absolute ethanol from 0.023 to 0.175 mL/kg of body weight showed that the systolic PAP (P ⴝ .02), pulmonary capillary wedge pressure (P ⴝ .02), and PVRI (P < .01) significantly increased in accordance with the increased single volume of absolute ethanol. A significant increase of the right ventricular end-diastolic volume index and right ventricular end-systolic volume index were observed at a dose of more than 0.14 mL/kg of body weight for a single bolus injection of absolute ethanol. CONCLUSIONS: During ethanol embolotherapy of extremity AV malformations, significant hemodynamic changes can arise during a bolus injection of absolute ethanol. Cardiopulmonary hemodynamic profiles should be monitored closely after a bolus injection of more than 0.14 mL/kg of body weight of absolute ethanol. J Vasc Interv Radiol 2010; 21:81– 89 Abbreviations: AV ⫽ arteriovenous, CI ⫽ cardiac index, CVP ⫽ central venous pressure, HR ⫽ heart rate, LVSWI ⫽ left ventricular stroke work index, PAP ⫽ pulmonary arterial blood pressure, PCWP ⫽ pulmonary capillary wedge pressure, PVRI ⫽ pulmonary vascular resistance index, RVEDVI ⫽ right ventricular end-diastolic volume index, RVEF ⫽ right ventricular ejection fraction, RVESVI ⫽ right ventricular end-systolic volume index, RVSWI ⫽ right ventricular stroke work index, sABP ⫽ systemic arterial blood pressure, SVI ⫽ stroke volume index, SVRI ⫽ systemic vascular resistance index, Tbaseline ⫽ hemodynamic profiles measured 15 minutes after induction, Tfinal ⫽ hemodynamic parameters measured at 10 minutes after the final injection of absolute ethanol, Thighest ⫽ hemodynamic profiles measured at the time of the maximum mean pulmonary arterial blood pressure value during a session, Tpost ⫽ hemodynamic profiles measured within 3 minutes after injection at the time of the maximum mean PAP value, Tpre ⫽ hemodynamic profiles measured immediately before ethanol injection, Tresp ⫽ hemodynamic profiles measured immediately after the patient has resumed spontaneous respiration
Department of Anesthesiology and Pain Medicine (B.S.S., H.S.C., T.S.H., C.S.K., J.S.K., S.R.B.), Surgery (D.I.K.), Department of Radiology and Center for Imaging Science (Y.S.D., S.K.C., K.B.P., H.S.P.), and Biostatistics Unit, Samsung Biomedical Research Institute (S.K.), Samsung Medical Center, Sungkyunkwan University School of Medicine, 50, Ilwon-Dong, Kangnam-Ku, Seoul 135-710, Korea. Received January 8, 2009; final revision received July 22, 2009;
accepted September 29, 2009. Address correspondence to Y.S.D.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2010 DOI: 10.1016/j.jvir.2009.09.026
THE management of an inoperable congenital arteriovenous (AV) malformation in the extremities is still a great challenge for clinicians. There is still difficulty to determine adequate solutions and to understand an unpredictable course to treatment. Incomplete surgical resection of an AV malformation or ligation of a feeding
81
82
•
Effect of Repeat Ethanol Injections in Extremity AVM Embolotherapy
artery often resulted in worsening the AV malformation lesion. It has become apparent that nidus ablation is one of the principal methods of AV malformation management, and the use of this method may allow a hope of a cure for an inoperable AV malformation (1,2). With an accumulation of experience, advances in technology with the use of superselective catheters, and improvement in embolic or sclerosing agents, the role of embolotherapy has expanded as a sole therapeutic mode for AV malformation management. Among the various embolizing or sclerosing agents, absolute ethanol has unique properties of dehydration to induce endothelial damage, blood protein denaturation, and thrombus formation (3). As a result of the eradicative potential of vascular obliteration, the use of absolute ethanol has been demonstrated to result in an decreased rate of recurrence and an increased cure rate in selected AV malformation lesions (2,4). Despite favorable outcomes with the use of absolute ethanol, its use is associated with various complications from localized complications such as skin or mucosa bullae, skin or subcutaneous tissue necrosis, nerve damage, and muscle or cartilage necrosis (5–7) to systemic complications such as acute, severe pulmonary hypertension and serious cardiopulmonary complications including detrimental cardiovascular collapse (8,9). Therefore, although absolute ethanol has became one of the major sclerosing agents for the management of an AV malformation with its potent sclerosing characteristic, issues regarding the risks and benefits of its use for the treatment of an AV malformation are still strongly debated (10 –12). The putative mechanism of cardiovascular collapse associated with absolute ethanol use is attributed to severe pulmonary vasoconstriction, but the exact pathophysiology still requires elucidation. Anesthesiologists have often been involved to administer general anesthesia for the adequate control of severe pain caused by ethanol injection and for close cardiopulmonary monitoring with use of a pulmonary artery catheter during ethanol embolotherapy of a large, inoperable AV malformation in which a large amount of absolute eth-
January 2010
JVIR
Table 1 Demographic Data for Patients with Extremity AV Malformations Characteristic
Value
Patient sex Male Female Age (y) Height (cm) Weight (kg) Mean no. of sclerotherapy sessions before this study Location of lesion Upper extremities Lower extremities
14 16 33.9 ⫾ 9.1 (22–51) 165.0 ⫾ 9.7 65.4 ⫾ 11.4 3.8 ⫾ 3.5 (1–11) 15 15
Note.—Data presented as means ⫾ SD; values in parenthesis are ranges.
anol was used (7). Many investigators have emphasized that ethanol embolotherapy with a large amount of absolute ethanol should be performed with extreme caution to avoid various toxicityrelated cardiopulmonary complications (11,13). Therefore, a better understanding of the effect of repeated ethanol bolus injection on cardiopulmonary function can be crucial to interventional radiologists and anesthesiologists for the adequate management of acute cardiovascular hemodynamic changes and the prevention of detrimental complications during ethanol embolotherapy. Previously, Yakes et al (7) have recommended a limitation of the maximum dose of absolute ethanol as less than 0.5–1.0 mL/kg per session. Mitchell et al (14) have proposed that the volume of each injection of absolute ethanol should be limited to 0.1 mL/kg of body weight per injection with more than a 5-minute interval between each injection. However, details on the acute cardiopulmonary hemodynamic response to repeat bolus injections of ethanol and the determination of the optimal amount of a single bolus ethanol injection to prevent excessive cardiopulmonary hemodynamic changes during ethanol embolotherapy are still unidentified. The aim of this study was to investigate prospectively the cardiopulmonary hemodynamic changes caused by repeat bolus injections of absolute ethanol in patients undergoing ethanol embolotherapy for the treatment of an inoperable soft-tissue AV malformation in the extremities.
MATERIALS AND METHODS Patients The institutional review board at our hospital approved this prospective study, and written informed consent was obtained from each patient. Thirty adult patients who underwent ethanol embolotherapy at our institution between March 2003 and March 2008 were enrolled in the study. All patients (14 men, 16 women; mean age, 33.9 years; age range, 22–51 y) were diagnosed with an inoperable AV malformation in the upper or lower extremities and were referred to the interventional radiologists for ethanol embolotherapy. The AV malformations were located in the upper and lower extremities (n ⫽ 15 each). To investigate the effects of injected absolute ethanol on the cardiopulmonary hemodynamic profiles more clearly, we excluded from the study all patients undergoing embolotherapy with the use of external pneumatic blood pressure cuffs or any vascular occlusion device. We also excluded one male patient who had dilated cardiomyopathy. Among the 30 patients, 17 had preexisting high cardiac output (⬎4.0 L · min⫺1 · m⫺2 of the cardiac index [CI]). The mean number of ethanol embolotherapy sessions per patient was 3.8 (range, 1–11; Table 1). All these patients were monitored with a pulmonary artery catheter because of the radiologists’ plan to use a single bolus injection of absolute ethanol more than 3 mL or a total amount that exceeded 0.25 mL/kg of actual body weight (15). Based on our clinical experience, we started continuous infu-
Volume 21
Number 1
sion of nitroglycerin (0.5–3.0 g/kg/ min) if the mean pulmonary arterial pressure (PAP) increased more than 10 mm Hg from the baseline measurement or if the mean PAP exceeded 25 mm Hg, which is defined as pulmonary hypertension. In addition, we started continuous infusion of nitroglycerin if the amount of subsequent bolus injection of absolute ethanol was more than 5 mL. Procedure Two experienced interventional radiologists (Y.S.D. and K.B.P.) with 14 and 6 years of experience in interventional radiology, respectively, at the time when this study began in 2003 performed all the embolotherapy procedures with absolute ethanol under general anesthesia. Cardiac rhythm and heart rate were monitored continuously with a standard lead II electrocardiogram. To measure continuous arterial blood pressure, a polyvinyl catheter was inserted into the radial artery. A fiberoptic thermodilution pulmonary artery catheter (7.5-F Swan-Ganz Catheter; Baxter, Irvine, California) was introduced through the right internal jugular vein and was connected to a cardiac output-monitoring computer (Explore; Baxter). The routes of vascular access for the ablation of the nidus were then determined. Transarterial and transvenous catheterizations with the use of a coaxial catheter or percutaneous direct puncture were performed and were followed by ethanol embolotherapy. Arteriograms were acquired to determine the flow characteristics of the AV malformations and the injection volume and rate of absolute ethanol used during embolotherapy. The dose of a single bolus injection of absolute ethanol was determined by the amount of contrast material needed to fill the nidus without opacifying the normal vessels. Ethanol embolotherapy was performed at the nidus itself (2,7). The volume of absolute ethanol was administered within the range of 2–10 mL per each single bolus injection (all wholemilliliter volumes except 9 mL were used) (15), and the total amount of absolute ethanol injection during a session was restricted to 1.0 mL/kg of actual body weight (2,7). Each etha-
Shin et al
•
83
nol injection was performed with at least at a 10-minute interval.
RVSWI 共gm-m · m⫺2 · beat⫺1) ⫽ SVI
Data Acquisition and Evaluation
After each bolus injection of absolute ethanol, we obtained hemodynamic profiles during a session as follows:
Fifteen minutes after induction, when the patient’s condition was hemodynamically stabilized, the baseline hemodynamic parameters (Tbaseline) including systemic arterial blood pressure (sABP), heart rate (HR), PAP, central venous pressure (CVP), and pulmonary artery occlusion pressure (PAOP) were recorded. Cardiac output was measured by the thermodilution method with 10-mL bolus injections of cold normal saline solution (4°C) and the average values were calculated from three thermodilution cardiac output measurements within a 10% range. Values obtained from thermodilution cardiac output measurements were used to calculate derived hemodynamic parameters including CI, stroke volume index (SVI), pulmonary vascular resistance index (PVRI), pulmonary capillary wedge pressure (PCWP), systemic vascular resistance index (SVRI), right ventricular ejection fraction (RVEF), right ventricular end-diastolic volume index (RVEDVI), right ventricular end-systolic volume index (RVESVI), left ventricular stroke work index (LVSWI), and right ventricular stroke work index (RVSWI). Calculations of these hemodynamic parameters were performed according to the following standard formulas: CI 共L · min
⫺1
⫺2
· m 兲 ⫽ cardiac output ⁄ body surface area
SVI 共mL · m⫺2 · beat⫺1兲 ⫽ CI/HR ⫻ 1000 PVRI 共dyne · s · cm⫺5 · m⫺2) ⫽ 80 ⫻ 共mean PAP ⫺ PCWP兲/CI SVRI 共dyne · s · cm⫺5 · m⫺2) ⫽ 80 ⫻ 共mean ABP ⫺ CVP兲/CI RVEF 共%兲 ⫽ stroke volume ⁄ end-disstolic volume RVEDVI 共mL · m⫺2兲 ⫽ VI/RVEF RVESVI 共mL · m⫺2兲 ⫽ RVEDVI ⫺ SVI LVSWI 共gm-m · m⫺2 · beat⫺1) ⫽ SVI ⫻ (mean ABP ⫺ PCWP兲 ⫻ 0.0136
⫻ (mean PAP ⫺ CVP兲 ⫻ 0.0136
Tbaseline: Hemodynamic profiles measured 15 minutes after induction; Tpre: Hemodynamic profiles measured immediately before ethanol injection; Tpost: Hemodynamic profiles measured within 3 minutes after injection at the time of the maximum mean PAP value; Thighest: Hemodynamic profiles measured at the time of the maximum mean PAP value during a session; Tfinal: Corresponding parameters measured 10 minutes after the final injection of absolute ethanol; Tresp: Hemodynamic profiles measured immediately after the patient has resumed spontaneous respiration; ⌬Tfinal – Tbaseline: Difference of values between Tfinal and Tbaseline; The largest of ⌬Tpost – Tpre: The largest difference of hemodynamic changes between Tpre and Tpost during a session; and ⌬Tresp – Tfinal: Differences of values between Tresp and Tfinal. We compared the hemodynamic parameters of Tbaseline, Thighest, Tfinal, and Tresp. It was determined if the highest PAPs and the relative hemodynamic profiles were more affected by a single bolus injection of absolute ethanol, an accumulative effect of the total amount of absolute ethanol, or restoration of patient spontaneous respiration that can aggravate the elevation of PAP. Moreover, to determine the point of the most significant hemodynamic changes during a session, we compared the differences of values between ⌬Tfinal – Tbaseline, the largest of ⌬Tpost – Tpre, and ⌬Tresp – Tfinal. We evaluated the hemodynamic parameters of consecutive Tpre and ⌬Tpost – Tpre and the trends in these parameters according to the increasing volume of a single bolus injection of absolute ethanol. As injection cases in which more than seven times in a session were not sufficient for statistical analysis, we performed statistical analy-
84
•
Effect of Repeat Ethanol Injections in Extremity AVM Embolotherapy
sis within a limit of seven times of absolute ethanol injection during a session. Statistical Analysis Data are presented as the median and the interquartile range (ie, 25th to 75th percentile) or as a mean ⫾ SD. Comparisons of Tbaseline, Thighest, Tfinal, and Tresp were performed with the paired t test with Bonferroni correction for the parametric data and Wilcoxon signed-rank test with Bonferroni correction for the nonparametric parameters. The paired t test with Bonferroni correction for the parametric data and the Wilcoxon signed-rank test with Bonferroni correction for the nonparametric data were used for the comparisons of ⌬Tfinal – Tbaseline and ⌬Tresp – Tfinal, the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal, and ⌬Tfinal – Tbaseline and the largest of ⌬Tpost – Tpre. As the number of single bolus ethanol injections and the amount of absolute ethanol injection per patient in a session were different, a mixed model was used for the investigation if Tpre and ⌬Tpost – Tpre were continuously increasing or decreasing along with the increase of the number of single bolus injections of absolute ethanol during a session and if ⌬Tpost – Tpre values were increasing or decreasing during a session as the amount (in mL/kg body weight) of each single bolus injection of absolute ethanol increased. If we assumed that the volume of a single bolus injection of absolute ethanol within the range of 0.023– 0.175 mL/kg of body weight positively correlated with the change of the mean PAP and a correlation coefficient of 0.4 was used in the null hypothesis (ie, upper limit of poor correlation) and 0.75 (ie, upper limit of moderate correlation) in the alternative hypothesis, 30 patients were required by the use of power analysis (␣ ⫽ .05,  ⫽ .8). All statistical analyses were performed with SAS software (version 9.1; SAS, Cary, North Carolina). A P value of less than .05 was considered to indicate a statistically significant difference.
RESULTS The demographic data are presented in Table 1. The total number of single bolus injections of absolute ethanol given to all patients was 181 (mean, 6.0; range, 3–11) and the mean amount of a single bolus injection was
January 2010
JVIR
Table 2 Embolotherapy Data Parameter
Value
Mean procedure duration (min) Mean no. of bolus injections per session Mean volume of absolute ethanol per injection (mL) Mean volume of absolute ethanol of single bolus injection (mL/kg of body weight) Mean volume of total absolute ethanol used (mL) Mean volume of absolute ethanol used (mL/kg of body weight) No. of single bolus injections of absolute ethanol (mL/kg of body weight) 0.02–0.03 0.03–0.04 0.04–0.05 0.05–0.06 0.06–0.07 0.07–0.08 0.08–0.09 0.09–0.10 0.10–0.11 0.11–0.12 0.12–0.13 0.13–0.14 0.14–0.15 0.15–0.16 0.16–0.17 0.17–0.18
217.4 ⫾ 37.5 6.0 ⫾ 2.0 (3–11) 5.1 ⫾ 2.44 (2–8, 10) 0.077 ⫾ 0.033 (0.023–0.157) 37.7 ⫾ 15.8 (15.0–74.0) 0.48 ⫾ 0.21 (0.26–0.89)
9 12 15 18 13 18 16 10 13 10 7 8 9 9 8 6
Note.—Data presented as means ⫾ SD; values in parenthesis are ranges.
5.1 mL ⫾ 2.44 (range, 2–10 mL of absolute ethanol, except for a volume of 9 mL; 0.023– 0.175 mL/kg of body weight). The total amount of absolute ethanol injected during a session ranged between 15 and 74.0 mL (for 0.26 – 0.89 mL/kg of body weight) with a mean amount of 37.7 mL (for 0.48 mL/kg of body weight; Table 2). Twenty-five patients underwent continuous infusion of nitroglycerin (range, 0.5–3.0 g/kg/min) in an attempt to attenuate the elevation of PAPs during the session and the recovery period. Among these 25 patients, nitroglycerin infusion was started at the beginning of embolotherapy in nine patients and in 16 patients it was administered after a few repeat injection of absolute ethanol. Hemodynamic values of Tbaseline, Thighest, Tfinal, and Tresp are presented in Table 3. The comparison of the hemodynamic profiles showed that the systolic, mean, and diastolic sABPs (P ⬍ .05, P ⬍ .05, and P ⬍ .05, respectively), systolic and mean PAPs (P ⬍ .01, and P ⬍ .01, respectively), and RVEF (P ⬍ .05) of Tfinal
were significantly higher than the values for T baseline . The systolic, mean, and diastolic sABPs (P ⬍ .05, P ⬍ .05, and P ⬍ .05, respectively) and PAPs (P ⬍ .01, P ⬍ .01, and P ⬍ .01, respectively), SVRI (P ⬍ .05), and PVRI (P ⬍ .05) of Thighest and Tresp were significantly higher than the corresponding values for Tfinal (Table 3). For the investigation of the largest of ⌬Tpost – Tpre, ⌬Tfinal – Tbaseline, and ⌬Tresp – Tfinal to determine the point of most significant hemodynamic change in a session, the systolic, mean, and diastolic sABPs (P ⬍ .01, P ⬍ .05, and P ⬍ .05, respectively) and the systolic and mean PAPs (P ⬍ .01 and P ⬍ .01, respectively) of ⌬Tfinal – Tbaseline were significantly lower than the corresponding values of the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal. However, the RVEF of ⌬Tfinal – Tbaseline was higher than the values of the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal (P ⬍ .05 and P ⬍ .05, respectively). However, there were no significant differences in the comparison between the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal (Table 4). The mean elevations of the systolic and mean PAPs of the largest of ⌬Tpost – Tpre
Volume 21
Number 1
Shin et al
•
85
Table 3 Cardiopulmonary Hemodynamic Profiles during Ethanol Embolotherapy Hemodynamic Parameter sABP (mm Hg) Systolic Mean Diastolic PAP (mm Hg) Systolic Mean Diastolic HR (beats/min) CI (L · min⫺1 · m⫺2) PCWP (mm Hg) CVP (mm Hg) RVEF (%) SVI (mL · m⫺2 · beat⫺1) RVEDVI (mL · m⫺2) RVESVI (mL · m⫺2) SVRI (dyne · s · cm⫺5 · m⫺2) PVRI (dyne · s · cm⫺5 · m⫺2) LVSWI (gm-m · m⫺2 · beat⫺1) RVSWI (gm-m · m⫺2 · beat⫺1)
Tbaseline
Thighest
Tfinal
Tresp
97.8 ⫾ 11.1 64.0 ⫾ 8.3 52.3 ⫾ 7.7
116.4 ⫾ 18.7*† 84.5 ⫾ 18.6*† 68.5 ⫾ 17.1*†
94.8 ⫾ 11.1* 64.5 ⫾ 7.5* 51.3 ⫾ 8.5*
120.8 ⫾ 19.7*† 78.2 ⫾ 10.4*† 62.8 ⫾ 9.8*†
23.7 ⫾ 4.4 18.6 ⫾ 5.1 12.9 ⫾ 3.3 72.1 ⫾ 10.6 5.10 ⫾ 2.36 14.0 ⫾ 2.7 9.1 ⫾ 2.8 37.7 ⫾ 4.2 72.1 ⫾ 33.9 94.0 ⫾ 28.6 62.2 ⫾ 54.3 1873.6 ⫾ 565.5 255.2 ⫾ 20.2 50.1 ⫾ 20.0 8.7 ⫾ 4.6
38.9 ⫾ 7.3*† 29.2 ⫾ 5.1*† 21.3 ⫾ 4.3*† 87.8 ⫾ 14.6* 6.11 ⫾ 1.67 19.6 ⫾ 5.5* 13.0 ⫾ 3.9* 38.4 ⫾ 7.1 63.9 ⫾ 25.3 111.0 ⫾ 28.5 72.8 ⫾ 36.3 2031.0 ⫾ 584.7† 305.9 ⫾ 54.6*† 52.3 ⫾ 18.9 13.6 ⫾ 5.9*
30.5 ⫾ 4.7* 23.3 ⫾ 4.4* 16.9 ⫾ 5.0 83.3 ⫾ 12.7 5.83 ⫾ 2.24 15.9 ⫾ 4.8 11.6 ⫾ 4.6 42.4 ⫾ 9.9* 72.3 ⫾ 34.4 106.2 ⫾ 30.1 65.4 ⫾ 28.1 1622.8 ⫾ 290.4 260.3 ⫾ 25.0 46.2 ⫾ 18.7 10.8 ⫾ 5.7
39.6 ⫾ 7.3*† 29.7 ⫾ 5.6*† 19.8 ⫾ 5.0*† 91.7 ⫾ 11.2* 6.61⫾ 2.48 18.5 ⫾ 5.3* 13.0 ⫾ 5.4 41.2 ⫾ 8.6* 74.1 ⫾ 27.6 122.1 ⫾ 24.2* 82.1 ⫾ 34.7 1714.9 ⫾ 349.1 293.4 ⫾ 37.3*† 60.8 ⫾ 23.1 15.4 ⫾ 7.6*†
Note.—Data are presented as means ⫾ SD. * P ⬍ .05 vs Tbaseline. † P ⬍ .05 vs Tfinal.
were 14.8 mm Hg ⫾ 7.7 and 10.7 mm Hg ⫾ 6.2, and the mean elevations of ⌬Tresp – Tfinal were 11.7 mm Hg ⫾ 5.3 and 7.4 mm Hg ⫾ 4.7, respectively. The systolic, mean, and diastolic PAPs of Tpre showed a progressive increase from Tbaseline to Tfinal (P ⬍ .01, P ⬍ .01, and P ⬍ .01, respectively) during a session. However, other preinjection hemodynamic parameters did not show a statistically significant increase or decrease as the number of absolute ethanol injections increased during a session. Among the hemodynamic values of ⌬Tpost – Tpre, the systolic and mean PAPs (P ⬍ .01 and P ⬍ .01, respectively) showed a significant decrease along with an increase in the number of bolus injections of absolute ethanol during a session. The elevations of systolic and mean PAPs per 0.1 mL of absolute ethanol per kilogram of body weight of ⌬Tpost – Tpre were 11.1 mm Hg ⫾ 8.7 and 9.2 mm Hg ⫾ 6.6 at the first injection, and 4.2 mm Hg ⫾ 3.1 and 4.9 ⫾ 5.2 at the seventh injection (Fig 1). In 13 patients (43.3%), the highest mean PAPs were reached after the first bolus injection of absolute ethanol, and in nine patients (30.0%) after the final injection. We found that the largest of ⌬Tpost – Tpre of the mean PAPs caused by a single bolus ethanol
injection was evident after the first injection of absolute ethanol in 20 patients (67%) and the lowest single dose of absolute ethanol that produced the maximum elevations of the mean PAPs by the first injection was 0.05 mL/kg body weight. The evaluation of the changes in the hemodynamic profiles according to the increase in volume of a single bolus injection of absolute ethanol from 0.023 to 0.175 mL/kg body weight showed that the systolic PAP (P ⫽ .02), CVP (P ⬍ .01), PCWP (P ⫽ .02), and PVRI (P ⬍ .01) significantly increased in accordance with the increased single volume of absolute ethanol. However, the CI decreased as the amount of bolus injection of absolute ethanol increased (Fig 2). Among the other parameters, we observed a significant increase of the RVEDVI and RVESVI at a dose of more than 0.14 mL/kg body weight of a single bolus injection of absolute ethanol (Fig 3).
DISCUSSION The major findings in this study include that a significant hemodynamic change can arise after the first bolus injection of absolute ethanol and during the period of restoration of spon-
taneous respiration for ethanol embolotherapy of AV malformations of the extremities. In addition, a single bolus injection of absolute ethanol of more than 0.14 mL/kg of body weight can compromise right ventricular function. Based on these results, it can be recommended that the volume of absolute ethanol be determined carefully from the first injection, and the volume should be based on a consideration of PAP in addition to the amount of contrast material needed to fill the nidus. Finally, careful attention is strongly advised to avoid significant hemodynamic changes during the restoration of spontaneous respiration. Although the use of ethanol embolotherapy has demonstrated many promising results, particularly an increased chance for complete remission of some AV malformations, patients are repeatedly exposed to the risks of minor or major complications associated with the chemical toxicity of ethanol as most inoperable soft-tissue AV malformations mandate the need for multistage ethanol embolotherapies. In our previous study (15), we demonstrated that the risk of pulmonary arterial hypertension did not decrease as sessions were repeated, which indicated that the risks of cardiopulmonary complications are sus-
86
•
Effect of Repeat Ethanol Injections in Extremity AVM Embolotherapy
January 2010
JVIR
Table 4 Hemodynamic Changes Caused by a Single Injection and the Total Amount of Absolute Ethanol and after Restoration of Spontaneous Respiration during a Session Hemodynamic Parameter sABP (mm Hg) Systolic Mean Diastolic PAP (mm Hg) Systolic Mean Diastolic HR (beats/min) CI (L · min⫺1 · m⫺2) PCWP (mm Hg) CVP (mm Hg) RVEF (%) SVI (mL · m⫺2 · beat⫺1) RVEDVI (mL · m⫺2) RVESVI (mL · m⫺2) SVRI (dyne · s · cm⫺5 · m⫺2) PVRI (dyne · s · cm⫺5 · m⫺2) LVSWI (gm-m · m⫺2 · beat⫺1) RVSWI (gm-m · m⫺2 · beat⫺1)
Largest ⌬Tpost – Tpre 19.6 ⫾ 22.4 (⫺1.5, 13.5, 35.7) 16.2 ⫾ 18.0 (5.0, 12.0, 23.5) 13.6 ⫾ 16.0 (2.8, 12.0, 17.3)
⌬Tfinal – Tbaseline
⌬Tresp – Tfinal
1.4 ⫾ 9.5 (⫺4.0, 2.0, 6.0)*† 3.3 ⫾ 6.7 (⫺1.0, 5.0, 8.0)*† 1.4 ⫾ 7.2 (⫺4.0, 1.0, 8.0)*†
28.7 ⫾ 14.0 (17.0, 28.0, 32.0) 15.6 ⫾ 9.5 (7.0, 12.5, 21.0) 12.4 ⫾ 8.3 (4.0, 12.0, 18.0)
14.8 ⫾ 7.7 (9.0, 13.5, 22.3) 3.7 ⫾ 3.7 (1.0, 2.0, 7.0)*† 11.7 ⫾ 5.3 (8.0, 10.0, 13.0) 10.7 ⫾ 6.2 (5.0, 11.0, 16.0) 3.4 ⫾ 2.4 (2.0, 4.0, 5.0)*† 7.4 ⫾ 4.7 (5.0, 6.0, 8.0) 7.1 ⫾ 4.4 (2.8, 8.0, 11.3) 2.7 ⫾ 3.0 (0, 3.5, 5.0) 3.9 ⫾ 4.5 (⫺1.0, 3.0, 7.5) 10.2 ⫾ 9.9 (1.0, 10.5, 19.25) 7.8 ⫾ 9.0 (1.0, 7.0, 14.0) 9.4 ⫾ 9.4 (1.0, 4.0, 16.0) 0.27 ⫾ 1.45 (⫺.37, 0.54, 1.11) 0.97 ⫾ 1.30 (.47, 1.21, 1.93) 1.0 ⫾ 0.94 (.05, 1.10, 1.77) 6.0 ⫾ 4.1 (3.0, 5.0, 9.8) 2.1 ⫾ 2.6 (0, 2.0, 3.0)* 3.5 ⫾ 3.5 (1.0, 3.0, 6.0) 3.8 ⫾ 3.7 (.8, 3.0, 7.0) 2.0 ⫾ 2.6 (0, 1.0, 3.0) 2.0 ⫾ 4.2 (0, 1.0, 3.0) ⫺0.2 ⫾ 5.0 (⫺5.0, ⫺1.0, 4.0) 6.1 ⫾ 5.0 (2.8, 8.5, 12.5)*† ⫺0.2 ⫾ 5.8 (⫺4.0, 0, 2.0) ⫺4.2 ⫾ 20.2 (⫺18.8, ⫺1.9, 7.3) 4.4 ⫾ 15.0 (⫺10.0, 7.9, 13.0) 4.7 ⫾ 11.8 (⫺1.8, 2.8, 12.3) 7.8 ⫾ 21.3 (⫺11.5, 5.3, 26.9) 2.1 ⫾ 22.4 (⫺20.8, ⫺1.5, 24.4) ⫺20.7 ⫾ 42.4 (⫺7.3, 6.8, 32.1) 3.3 ⫾ 20.0 (⫺19.5, ⫺.3, 11.6) 4.4 ⫾ 15.6 (⫺56.1, ⫺20.0, 4.1) 17.2 ⫾ 43.9 (⫺16.6, 10.2, 54.7) 78.0 ⫾ 305.7 (⫺86.8, 82.0, 255.8) ⫺156.2 ⫾ 284.4 (⫺470.0, ⫺174.0, 7.0) 94.0 ⫾ 227.9 (⫺74.0, 72.0, 355.0) 39.1 ⫾ 43.3 (16.5, 30.0, 73.0)
14.8 ⫾ 30.8 (⫺21.0, 5.0, 37.0)*
7.3 ⫾ 15.6 (⫺4.1, 8.8, 18.7)
2.5 ⫾ 11.9 (⫺7.3, 1.4, 15.1)
5.7 ⫾ 3.1 (3.5, 5.2, 8.1)
3.4 ⫾ 3.8 (1.8, 4.2, 4.6)
26.5 ⫾ 27.5 (⫺1.0, 30.0, 40.5) 16.2 ⫾ 10.9 (9.1, 13.0, 28.9) 4.2 ⫾ 3.6 (0.3, 5.2, 6.4)
Note.—Data presented as means ⫾ SD where applicable; the numbers in parenthesis are 25th percentile, median, and 75th percentile. There are significant differences of the hemodynamic profiles for the comparison between the largest of ⌬Tpost – Tpre and ⌬Tfinal – Tbaseline, ⌬Tresp – Tfinal, and ⌬Tfinal – Tbaseline. However, there were no statistical differences between the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal. * P ⬍ .05 vs largest of ⌬Tpost – Tpre. † P ⬍ .05 vs ⌬Tresp – Tfinal.
Figure 1. Changes of systolic and mean PAPs per 0.1 mL of absolute ethanol per kilogram of body weight of ⌬Tpost – Tpre (differences among the systolic, mean, and diastolic PAPs recorded when the mean PAP reached a maximum value after injection of absolute ethanol and immediately before the injection of absolute ethanol) decreased according to the increase of the number of injections during a session. Each ethanol injection was performed with at least a 10-minute interval. The changes of PAPs with more than seven injections are not depicted, as the number of cases was not sufficient for statistical analysis. Statistical analysis was performed by use of a mixed model. (*P ⬍ .05.)
tained during multisession ethanol embolotherapies. Ko et al (16) reported that PAP elevations were significantly correlated with the pulmonary arterial ethanol level in patients treated without the use of vascular occlusion techniques. Ethanol injection can result in severe pulmonary vasospasm and acutely increase thin-walled right ventricular afterload and decrease right ventricular cardiac output (17). These effects can be further aggravated by systemic effects of alcohol affecting right atrial and ventricular contractility, including chronotropic and inotropic functions (18 –21). It is important to consider direct myocardial depression caused by the systemic effect of the total amount of absolute ethanol. However, our results have elucidated that significant hemodynamic changes can initiate from the first injection of absolute ethanol, and hemodynamic changes are affected more by the single bolus injection volume rather than by the total amount of absolute
Volume 21
Number 1
Shin et al
10
20
•
87
90 80 PVRI (dyne·s·cm-5·m
8
15
70
*
60 6
*
50
10 *
4
∗
40 30
5
20
2
10
0.17 -0.18
0.15 -0.16
0.13 -0.14
0.11 -0.12
0.09 -0.10
0.07 -0.08
b.
0.05 -0.06
0
0.17-0.18
0.15-0.16
0.13-0.14
0.11-0.12
0.09-0.10
0.07-0.08
0.05-0.06
a.
-2
†
0.03-0.04
0.17- 0.18
0.15- 0.16
0.13- 0.14
0.11- 0.12
0.09- 0.10
0.07- 0.08
0.05- 0.06
0.03- 0.04
-5
0
0.03 -0.04
0
c.
Figure 2. Systolic PAP (a), CVP and PCWP (b), and PVRI (c) of ⌬Tpost – Tpre (differences between hemodynamic parameters measured when the mean PAP reached a maximum value after injection of absolute ethanol and immediately before the injection of absolute ethanol) increased as the dose of a single injection of absolute ethanol increased. However, the CI (b) decreased according to the increase of a single injection dose of absolute ethanol. Statistical analysis was performed with use of a mixed model. (*†P ⬍ .05.)
Figure 3. Sharp increase of RVEDVI and RVESVI from a single dose of 0.14 mL of absolute ethanol per kilogram of body weight of ⌬Tpost – Tpre (differences between hemodynamic parameters measured when the mean PAP reached a maximum value after injection of absolute ethanol and immediately before the injection of absolute ethanol). (*†P ⬍ .05 by use of mixed-model analysis.)
ethanol used during ethanol embolotherapy. First, the systolic, mean, and diastolic PAPs and PVRI of Thighest were significantly higher than the corresponding values of Tfinal, which can reflect the accumulative effect of the total amount of absolute ethanol during a session. Second, even though repeated measurements of hemodynamic param-
eters at Tpre demonstrated that the systolic, mean, and diastolic PAPs increased from the first to the final injection during a session, the systolic and mean PAPs of the largest of ⌬Tpost – Tpre changed more significantly than the corresponding values of ⌬Tfinal – Tbaseline, and there was a lack of significant change of RVEF. In addition, the
systolic and mean PAPs of ⌬Tpost – Tpre showed a significant decrease along with an increase in the number of bolus injections of absolute ethanol during a session. Third, we observed significant increases of RVEDVI and RVESVI at a dose of greater than 0.14 mL/kg of body weight of a single injection of absolute ethanol. The volume of ethanol that resulted in increases of RVEDVI and RVESVI may be considered as an initiating volume that begins to impose a burden on right ventricular contractility. Fourth, the progressive increase in the systolic PAP, CVP, PCWP, and PVRI and decrease in CI of ⌬Tpost – Tpre according to the incremental increase in the single bolus injection volume of absolute ethanol per body weight indicates that a larger single bolus injection dose of absolute ethanol can affect right and left ventricular function. Greyson et al (22) have reported that although right ventricular systolic and diastolic pressures returned to normal values, right ventricular contractile dysfunction persisted for at least 2 hours after 1 hour of pulmonary artery constriction in an animal study. Therefore, we suggest ethanol embolotherapy be performed with a minimal therapeutic volume from the first injection, and the cardiopulmonary hemodynamic profiles be closely monitored when injecting more than 0.14
88
•
Effect of Repeat Ethanol Injections in Extremity AVM Embolotherapy
mL/kg of body weight of absolute ethanol in patients with an AV malformation of the extremities. Our study also showed that significant hemodynamic changes could arise after resuming spontaneous respiration. In addition, ⌬Tresp – Tfinal did not show any significant difference versus the largest value of ⌬Tpost – Tpre. Significant physiologic changes in the cardiopulmonary system can occur during the transition from mechanical ventilation to spontaneous respiration. Three main contributing factors involved in the transition process are the following. First, the shift of intrathoracic pressure from positive to negative increases venous return; second, a sympathetic discharge occurs secondary to tracheal irritation and hypercapnia; and third, there is an afterload increase resulting from the elevated systolic blood pressure and lowered pleural pressure (23–25). A combination of these factors may contribute to the aggravation of already elevated PAPs after repeated injections of absolute ethanol. Therefore, we recommend continuous PAP monitoring and administration of nitroglycerin during the recovery period to ensure the safe recovery of patients. Alternatively, maintenance of mechanical ventilation may be considered as the already elevated PAPs can be acutely increased again (as high as 11.7 mm Hg for the systolic PAP and 7.4 mm Hg for the mean PAPs) during the period of weaning of mechanical ventilation. One of the limitations of this study is that we were unable to measure the blood alcohol level from the pulmonary vascular bed at the point of the maximum mean PAP as we used the thermodilution method to obtain values of the hemodynamic profiles. Therefore, we could not analyze the correlation between the blood alcohol concentration at the pulmonary arterial bed and hemodynamic parameters at the time of the maximum PAP values. Additionally, significant hemodynamic changes after the first injection and the subsequent decrease in systolic and mean PAPs of ⌬Tpost – Tpre along with the increase in the number of bolus injections of absolute ethanol during a session may result from a progressive decrease in sympathetic stimu-
lation arising from pain at the time of injection. However, we could not clearly elucidate or rule out the effect of sympathetic stimulation caused by pain on the elevation of PAPs and the corresponding hemodynamic parameters. Finally, as most of the patients (83.3%) received nitroglycerin during the procedure and recovery period, and as the initiating time of continuous administration of nitroglycerin was different for each patient, we could not evaluate the effect of nitroglycerin on the attenuation of PAP increases during ethanol embolotherapy. In conclusion, significantly greater hemodynamic changes can arise from a single bolus injection of absolute ethanol rather than as a result of the total amount of absolute ethanol used during ethanol embolotherapy of AV malformations of the extremities. Therefore, there should be close collaboration between an anesthesiologist and an interventional radiologist to determine the dose of absolute ethanol and injection intervals, and it is recommended for an interventional radiologist to proceed with the lowest therapeutic dose and use an injection volume of absolute ethanol that is no greater than 0.14 mL/kg body weight. In addition, the cardiopulmonary hemodynamic profiles should be closely monitored when injecting more than 0.14 mL/kg body weight of absolute ethanol in patients with AV malformation of the extremities. Restoration of spontaneous respiration can also aggravate hemodynamic changes. Therefore, it may be advisable to consider maintaining mechanical ventilation during the recovery period as the already elevated PAPs at the end of session can be significantly elevated again at the time of the restoration of spontaneous respiration. References 1. Hyodoh H, Hori M, Akiba H, Tamakawa M, Hyodoh K, Hareyama M. Peripheral vascular malformations: imaging, treatment approaches, and therapeutic issues. Radiographics 2005; 25(Suppl 1):S159 –S171. 2. Do YS, Yakes WF, Shin SW, et al. Ethanol embolization of arteriovenous malformations: interim results. Radiology 2005; 235:674 – 682. 3. Yakes WF, Haas DK, Parker SH, et al. Symptomatic vascular malformations: ethanol embolotherapy. Radiology 1989; 170:1059 –1066.
January 2010
JVIR
4. Yakes WF, Pevsner P, Reed M, Donohue HJ, Ghaed N. Serial embolizations of an extremity arteriovenous malformation with alcohol via direct percutaneous puncture. AJR Am J Roentgenol 1986; 146:1038 –1040. 5. Lee BB, Bergan JJ. Advanced management of congenital vascular malformations: a multidisciplinary approach. Cardiovasc Surg 2002; 10:523–533. 6. Rimon U, Garniek A, Galili Y, Golan G, Bensaid P, Morag B. Ethanol sclerotherapy of peripheral venous malformations. Eur J Radiol 2004; 52:283–287. 7. Yakes WF, Rossi P, Odink H. How I do it. Arteriovenous malformation management. Cardiovasc Intervent Radiol 1996; 19:65–71. 8. Wong GA, Armstrong DC, Robertson JM. Cardiovascular collapse during ethanol sclerotherapy in a pediatric patient. Paediatr Anaesth 2006; 16:343– 346. 9. Garel L, Mareschal JL, Gagnadoux MF, Pariente D, Guilbert M, Sauvegrain J. Fatal outcome after ethanol renal ablation in child with end-stage kidneys. AJR Am J Roentgenol 1986; 146:593– 594. 10. Lee BB, Do YS, Yakes W, Kim DI, Mattassi R, Hyon WS. Management of arteriovenous malformations: a multidisciplinary approach. J Vasc Surg 2004; 39:590 – 600. 11. Villavicencio JL. Primum non nocere: Is it always true? The use of absolute ethanol in the management of congenital vascular malformations. J Vasc Surg 2001; 33:904 –906. 12. Behnia R. Systemic effects of absolute alcohol embolization in a patient with a congenital arteriovenous malformation of the lower extremity. Anesth Analg 1995; 80:415– 417. 13. Lee BB, Kim DI, Huh S, et al. New experiences with absolute ethanol sclerotherapy in the management of a complex form of congenital venous malformation. J Vasc Surg 2001; 33: 764 –772. 14. Mitchell SE, Shah AM, Schwengel D. Pulmonary artery pressure changes during ethanol embolization procedures to treat vascular malformations: can cardiovascular collapse be predicted? J Vasc Interv Radiol 2006; 17: 253–262. 15. Shin BS, Do YS, Lee BB, et al. Multistage ethanol sclerotherapy of softtissue arteriovenous malformations: effect on pulmonary arterial pressure. Radiology 2005; 235:1072–1077. 16. Ko JS, Kim JA, Do YS, et al. Prediction of the effect of injected ethanol on pulmonary arterial pressure during sclerotherapy of arteriovenous malformations: relationship with dose
Volume 21
Number 1
of ethanol. J Vasc Interv Radiol 2009; 20:39 – 45. 17. Yakes WF, Krauth L, Ecklund J, et al. Ethanol endovascular management of brain arteriovenous malformations: initial results. Neurosurgery 1997; 40:1145– 1152. 18. Kelly LF, Goldberg SJ, Donnerstein RL, Cardy MA, Palombo GM. Hemodynamic effects of acute ethanol in young adults. Am J Cardiol 1996; 78: 851– 854. 19. Guarnieri T, Lakatta EG. Mechanism of myocardial contractile depression by clinical concentrations of ethanol: a
Shin et al
study in ferret papillary muscles. J Clin Invest 1990; 85:1462–1467. 20. Delgado CE, Gortuin NJ, Ross RS. Acute effects of low doses of alcohol on left ventricular function by echocardiography. Circulation 1975; 51:535–540. 21. Campbell PH, Barker LA, McDonough KH. The effect of acute ethanol exposure on the chronotropic and inotropic function of the rat right atrium. J Pharm Pharmacol 2000; 52:1001–1010. 22. Greyson C, Xu Y, Cohen J, Schwartz GG. Right ventricular dysfunction persists following brief right ventricular pressure overload. Cardiovasc Res 1997; 34:281–288.
•
89
23. Lemaire F, Teboul JL, Cinotti L, et al. Acute left ventricular dysfunction during unsuccessful weaning from mechanical ventilation. Anesthesiology 1988; 69:171–179. 24. Mancebo J. Weaning from mechanical ventilation. Eur Respir J 1996; 9:1923– 1931. 25. Jardin F, Delorme G, Hardy A, Auvert B, Beauchet A, Bourdarias JP. Reevaluation of hemodynamic consequences of positive pressure ventilation: emphasis on cyclic right ventricular afterloading by mechanical lung inflation. Anesthesiology 1990; 72:966–970.
PURPOSE: This study was designed to investigate the effects of repeat bolus absolute ethanol injections on cardiopulmonary hemodynamic changes during ethanol embolotherapy of inoperable congenital arteriovenous (AV) malformations in the extremities. MATERIALS AND METHODS: Thirty adult patients (14 male, 16 female; age range, 22–51 years) who underwent ethanol embolotherapy of extremity AV malformations were enrolled in the study. A pulmonary artery catheter was used to measure hemodynamic profiles at baseline (Tbaseline), immediately before (Tpre) and after (Tpost) a bolus injection of absolute ethanol, at the time of the maximum mean pulmonary arterial pressure (PAP) value during a session (Thighest), 10 minutes after the final injection (Tfinal), and after restoration of spontaneous breathing (Tresp). RESULTS: The systolic, mean, and diastolic PAP (P < .01, P < .01, and P < .01, respectively) and the systemic vascular resistance index (P < .05) and pulmonary vascular resistance index (PVRI; P < .05) of Thighest and Tresp were significantly higher than values for Tfinal. The volume of a single bolus injection of absolute ethanol from 0.023 to 0.175 mL/kg of body weight showed that the systolic PAP (P ⴝ .02), pulmonary capillary wedge pressure (P ⴝ .02), and PVRI (P < .01) significantly increased in accordance with the increased single volume of absolute ethanol. A significant increase of the right ventricular end-diastolic volume index and right ventricular end-systolic volume index were observed at a dose of more than 0.14 mL/kg of body weight for a single bolus injection of absolute ethanol. CONCLUSIONS: During ethanol embolotherapy of extremity AV malformations, significant hemodynamic changes can arise during a bolus injection of absolute ethanol. Cardiopulmonary hemodynamic profiles should be monitored closely after a bolus injection of more than 0.14 mL/kg of body weight of absolute ethanol. J Vasc Interv Radiol 2010; 21:81– 89 Abbreviations: AV ⫽ arteriovenous, CI ⫽ cardiac index, CVP ⫽ central venous pressure, HR ⫽ heart rate, LVSWI ⫽ left ventricular stroke work index, PAP ⫽ pulmonary arterial blood pressure, PCWP ⫽ pulmonary capillary wedge pressure, PVRI ⫽ pulmonary vascular resistance index, RVEDVI ⫽ right ventricular end-diastolic volume index, RVEF ⫽ right ventricular ejection fraction, RVESVI ⫽ right ventricular end-systolic volume index, RVSWI ⫽ right ventricular stroke work index, sABP ⫽ systemic arterial blood pressure, SVI ⫽ stroke volume index, SVRI ⫽ systemic vascular resistance index, Tbaseline ⫽ hemodynamic profiles measured 15 minutes after induction, Tfinal ⫽ hemodynamic parameters measured at 10 minutes after the final injection of absolute ethanol, Thighest ⫽ hemodynamic profiles measured at the time of the maximum mean pulmonary arterial blood pressure value during a session, Tpost ⫽ hemodynamic profiles measured within 3 minutes after injection at the time of the maximum mean PAP value, Tpre ⫽ hemodynamic profiles measured immediately before ethanol injection, Tresp ⫽ hemodynamic profiles measured immediately after the patient has resumed spontaneous respiration
Department of Anesthesiology and Pain Medicine (B.S.S., H.S.C., T.S.H., C.S.K., J.S.K., S.R.B.), Surgery (D.I.K.), Department of Radiology and Center for Imaging Science (Y.S.D., S.K.C., K.B.P., H.S.P.), and Biostatistics Unit, Samsung Biomedical Research Institute (S.K.), Samsung Medical Center, Sungkyunkwan University School of Medicine, 50, Ilwon-Dong, Kangnam-Ku, Seoul 135-710, Korea. Received January 8, 2009; final revision received July 22, 2009;
accepted September 29, 2009. Address correspondence to Y.S.D.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2010 DOI: 10.1016/j.jvir.2009.09.026
THE management of an inoperable congenital arteriovenous (AV) malformation in the extremities is still a great challenge for clinicians. There is still difficulty to determine adequate solutions and to understand an unpredictable course to treatment. Incomplete surgical resection of an AV malformation or ligation of a feeding
81
82
•
Effect of Repeat Ethanol Injections in Extremity AVM Embolotherapy
artery often resulted in worsening the AV malformation lesion. It has become apparent that nidus ablation is one of the principal methods of AV malformation management, and the use of this method may allow a hope of a cure for an inoperable AV malformation (1,2). With an accumulation of experience, advances in technology with the use of superselective catheters, and improvement in embolic or sclerosing agents, the role of embolotherapy has expanded as a sole therapeutic mode for AV malformation management. Among the various embolizing or sclerosing agents, absolute ethanol has unique properties of dehydration to induce endothelial damage, blood protein denaturation, and thrombus formation (3). As a result of the eradicative potential of vascular obliteration, the use of absolute ethanol has been demonstrated to result in an decreased rate of recurrence and an increased cure rate in selected AV malformation lesions (2,4). Despite favorable outcomes with the use of absolute ethanol, its use is associated with various complications from localized complications such as skin or mucosa bullae, skin or subcutaneous tissue necrosis, nerve damage, and muscle or cartilage necrosis (5–7) to systemic complications such as acute, severe pulmonary hypertension and serious cardiopulmonary complications including detrimental cardiovascular collapse (8,9). Therefore, although absolute ethanol has became one of the major sclerosing agents for the management of an AV malformation with its potent sclerosing characteristic, issues regarding the risks and benefits of its use for the treatment of an AV malformation are still strongly debated (10 –12). The putative mechanism of cardiovascular collapse associated with absolute ethanol use is attributed to severe pulmonary vasoconstriction, but the exact pathophysiology still requires elucidation. Anesthesiologists have often been involved to administer general anesthesia for the adequate control of severe pain caused by ethanol injection and for close cardiopulmonary monitoring with use of a pulmonary artery catheter during ethanol embolotherapy of a large, inoperable AV malformation in which a large amount of absolute eth-
January 2010
JVIR
Table 1 Demographic Data for Patients with Extremity AV Malformations Characteristic
Value
Patient sex Male Female Age (y) Height (cm) Weight (kg) Mean no. of sclerotherapy sessions before this study Location of lesion Upper extremities Lower extremities
14 16 33.9 ⫾ 9.1 (22–51) 165.0 ⫾ 9.7 65.4 ⫾ 11.4 3.8 ⫾ 3.5 (1–11) 15 15
Note.—Data presented as means ⫾ SD; values in parenthesis are ranges.
anol was used (7). Many investigators have emphasized that ethanol embolotherapy with a large amount of absolute ethanol should be performed with extreme caution to avoid various toxicityrelated cardiopulmonary complications (11,13). Therefore, a better understanding of the effect of repeated ethanol bolus injection on cardiopulmonary function can be crucial to interventional radiologists and anesthesiologists for the adequate management of acute cardiovascular hemodynamic changes and the prevention of detrimental complications during ethanol embolotherapy. Previously, Yakes et al (7) have recommended a limitation of the maximum dose of absolute ethanol as less than 0.5–1.0 mL/kg per session. Mitchell et al (14) have proposed that the volume of each injection of absolute ethanol should be limited to 0.1 mL/kg of body weight per injection with more than a 5-minute interval between each injection. However, details on the acute cardiopulmonary hemodynamic response to repeat bolus injections of ethanol and the determination of the optimal amount of a single bolus ethanol injection to prevent excessive cardiopulmonary hemodynamic changes during ethanol embolotherapy are still unidentified. The aim of this study was to investigate prospectively the cardiopulmonary hemodynamic changes caused by repeat bolus injections of absolute ethanol in patients undergoing ethanol embolotherapy for the treatment of an inoperable soft-tissue AV malformation in the extremities.
MATERIALS AND METHODS Patients The institutional review board at our hospital approved this prospective study, and written informed consent was obtained from each patient. Thirty adult patients who underwent ethanol embolotherapy at our institution between March 2003 and March 2008 were enrolled in the study. All patients (14 men, 16 women; mean age, 33.9 years; age range, 22–51 y) were diagnosed with an inoperable AV malformation in the upper or lower extremities and were referred to the interventional radiologists for ethanol embolotherapy. The AV malformations were located in the upper and lower extremities (n ⫽ 15 each). To investigate the effects of injected absolute ethanol on the cardiopulmonary hemodynamic profiles more clearly, we excluded from the study all patients undergoing embolotherapy with the use of external pneumatic blood pressure cuffs or any vascular occlusion device. We also excluded one male patient who had dilated cardiomyopathy. Among the 30 patients, 17 had preexisting high cardiac output (⬎4.0 L · min⫺1 · m⫺2 of the cardiac index [CI]). The mean number of ethanol embolotherapy sessions per patient was 3.8 (range, 1–11; Table 1). All these patients were monitored with a pulmonary artery catheter because of the radiologists’ plan to use a single bolus injection of absolute ethanol more than 3 mL or a total amount that exceeded 0.25 mL/kg of actual body weight (15). Based on our clinical experience, we started continuous infu-
Volume 21
Number 1
sion of nitroglycerin (0.5–3.0 g/kg/ min) if the mean pulmonary arterial pressure (PAP) increased more than 10 mm Hg from the baseline measurement or if the mean PAP exceeded 25 mm Hg, which is defined as pulmonary hypertension. In addition, we started continuous infusion of nitroglycerin if the amount of subsequent bolus injection of absolute ethanol was more than 5 mL. Procedure Two experienced interventional radiologists (Y.S.D. and K.B.P.) with 14 and 6 years of experience in interventional radiology, respectively, at the time when this study began in 2003 performed all the embolotherapy procedures with absolute ethanol under general anesthesia. Cardiac rhythm and heart rate were monitored continuously with a standard lead II electrocardiogram. To measure continuous arterial blood pressure, a polyvinyl catheter was inserted into the radial artery. A fiberoptic thermodilution pulmonary artery catheter (7.5-F Swan-Ganz Catheter; Baxter, Irvine, California) was introduced through the right internal jugular vein and was connected to a cardiac output-monitoring computer (Explore; Baxter). The routes of vascular access for the ablation of the nidus were then determined. Transarterial and transvenous catheterizations with the use of a coaxial catheter or percutaneous direct puncture were performed and were followed by ethanol embolotherapy. Arteriograms were acquired to determine the flow characteristics of the AV malformations and the injection volume and rate of absolute ethanol used during embolotherapy. The dose of a single bolus injection of absolute ethanol was determined by the amount of contrast material needed to fill the nidus without opacifying the normal vessels. Ethanol embolotherapy was performed at the nidus itself (2,7). The volume of absolute ethanol was administered within the range of 2–10 mL per each single bolus injection (all wholemilliliter volumes except 9 mL were used) (15), and the total amount of absolute ethanol injection during a session was restricted to 1.0 mL/kg of actual body weight (2,7). Each etha-
Shin et al
•
83
nol injection was performed with at least at a 10-minute interval.
RVSWI 共gm-m · m⫺2 · beat⫺1) ⫽ SVI
Data Acquisition and Evaluation
After each bolus injection of absolute ethanol, we obtained hemodynamic profiles during a session as follows:
Fifteen minutes after induction, when the patient’s condition was hemodynamically stabilized, the baseline hemodynamic parameters (Tbaseline) including systemic arterial blood pressure (sABP), heart rate (HR), PAP, central venous pressure (CVP), and pulmonary artery occlusion pressure (PAOP) were recorded. Cardiac output was measured by the thermodilution method with 10-mL bolus injections of cold normal saline solution (4°C) and the average values were calculated from three thermodilution cardiac output measurements within a 10% range. Values obtained from thermodilution cardiac output measurements were used to calculate derived hemodynamic parameters including CI, stroke volume index (SVI), pulmonary vascular resistance index (PVRI), pulmonary capillary wedge pressure (PCWP), systemic vascular resistance index (SVRI), right ventricular ejection fraction (RVEF), right ventricular end-diastolic volume index (RVEDVI), right ventricular end-systolic volume index (RVESVI), left ventricular stroke work index (LVSWI), and right ventricular stroke work index (RVSWI). Calculations of these hemodynamic parameters were performed according to the following standard formulas: CI 共L · min
⫺1
⫺2
· m 兲 ⫽ cardiac output ⁄ body surface area
SVI 共mL · m⫺2 · beat⫺1兲 ⫽ CI/HR ⫻ 1000 PVRI 共dyne · s · cm⫺5 · m⫺2) ⫽ 80 ⫻ 共mean PAP ⫺ PCWP兲/CI SVRI 共dyne · s · cm⫺5 · m⫺2) ⫽ 80 ⫻ 共mean ABP ⫺ CVP兲/CI RVEF 共%兲 ⫽ stroke volume ⁄ end-disstolic volume RVEDVI 共mL · m⫺2兲 ⫽ VI/RVEF RVESVI 共mL · m⫺2兲 ⫽ RVEDVI ⫺ SVI LVSWI 共gm-m · m⫺2 · beat⫺1) ⫽ SVI ⫻ (mean ABP ⫺ PCWP兲 ⫻ 0.0136
⫻ (mean PAP ⫺ CVP兲 ⫻ 0.0136
Tbaseline: Hemodynamic profiles measured 15 minutes after induction; Tpre: Hemodynamic profiles measured immediately before ethanol injection; Tpost: Hemodynamic profiles measured within 3 minutes after injection at the time of the maximum mean PAP value; Thighest: Hemodynamic profiles measured at the time of the maximum mean PAP value during a session; Tfinal: Corresponding parameters measured 10 minutes after the final injection of absolute ethanol; Tresp: Hemodynamic profiles measured immediately after the patient has resumed spontaneous respiration; ⌬Tfinal – Tbaseline: Difference of values between Tfinal and Tbaseline; The largest of ⌬Tpost – Tpre: The largest difference of hemodynamic changes between Tpre and Tpost during a session; and ⌬Tresp – Tfinal: Differences of values between Tresp and Tfinal. We compared the hemodynamic parameters of Tbaseline, Thighest, Tfinal, and Tresp. It was determined if the highest PAPs and the relative hemodynamic profiles were more affected by a single bolus injection of absolute ethanol, an accumulative effect of the total amount of absolute ethanol, or restoration of patient spontaneous respiration that can aggravate the elevation of PAP. Moreover, to determine the point of the most significant hemodynamic changes during a session, we compared the differences of values between ⌬Tfinal – Tbaseline, the largest of ⌬Tpost – Tpre, and ⌬Tresp – Tfinal. We evaluated the hemodynamic parameters of consecutive Tpre and ⌬Tpost – Tpre and the trends in these parameters according to the increasing volume of a single bolus injection of absolute ethanol. As injection cases in which more than seven times in a session were not sufficient for statistical analysis, we performed statistical analy-
84
•
Effect of Repeat Ethanol Injections in Extremity AVM Embolotherapy
sis within a limit of seven times of absolute ethanol injection during a session. Statistical Analysis Data are presented as the median and the interquartile range (ie, 25th to 75th percentile) or as a mean ⫾ SD. Comparisons of Tbaseline, Thighest, Tfinal, and Tresp were performed with the paired t test with Bonferroni correction for the parametric data and Wilcoxon signed-rank test with Bonferroni correction for the nonparametric parameters. The paired t test with Bonferroni correction for the parametric data and the Wilcoxon signed-rank test with Bonferroni correction for the nonparametric data were used for the comparisons of ⌬Tfinal – Tbaseline and ⌬Tresp – Tfinal, the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal, and ⌬Tfinal – Tbaseline and the largest of ⌬Tpost – Tpre. As the number of single bolus ethanol injections and the amount of absolute ethanol injection per patient in a session were different, a mixed model was used for the investigation if Tpre and ⌬Tpost – Tpre were continuously increasing or decreasing along with the increase of the number of single bolus injections of absolute ethanol during a session and if ⌬Tpost – Tpre values were increasing or decreasing during a session as the amount (in mL/kg body weight) of each single bolus injection of absolute ethanol increased. If we assumed that the volume of a single bolus injection of absolute ethanol within the range of 0.023– 0.175 mL/kg of body weight positively correlated with the change of the mean PAP and a correlation coefficient of 0.4 was used in the null hypothesis (ie, upper limit of poor correlation) and 0.75 (ie, upper limit of moderate correlation) in the alternative hypothesis, 30 patients were required by the use of power analysis (␣ ⫽ .05,  ⫽ .8). All statistical analyses were performed with SAS software (version 9.1; SAS, Cary, North Carolina). A P value of less than .05 was considered to indicate a statistically significant difference.
RESULTS The demographic data are presented in Table 1. The total number of single bolus injections of absolute ethanol given to all patients was 181 (mean, 6.0; range, 3–11) and the mean amount of a single bolus injection was
January 2010
JVIR
Table 2 Embolotherapy Data Parameter
Value
Mean procedure duration (min) Mean no. of bolus injections per session Mean volume of absolute ethanol per injection (mL) Mean volume of absolute ethanol of single bolus injection (mL/kg of body weight) Mean volume of total absolute ethanol used (mL) Mean volume of absolute ethanol used (mL/kg of body weight) No. of single bolus injections of absolute ethanol (mL/kg of body weight) 0.02–0.03 0.03–0.04 0.04–0.05 0.05–0.06 0.06–0.07 0.07–0.08 0.08–0.09 0.09–0.10 0.10–0.11 0.11–0.12 0.12–0.13 0.13–0.14 0.14–0.15 0.15–0.16 0.16–0.17 0.17–0.18
217.4 ⫾ 37.5 6.0 ⫾ 2.0 (3–11) 5.1 ⫾ 2.44 (2–8, 10) 0.077 ⫾ 0.033 (0.023–0.157) 37.7 ⫾ 15.8 (15.0–74.0) 0.48 ⫾ 0.21 (0.26–0.89)
9 12 15 18 13 18 16 10 13 10 7 8 9 9 8 6
Note.—Data presented as means ⫾ SD; values in parenthesis are ranges.
5.1 mL ⫾ 2.44 (range, 2–10 mL of absolute ethanol, except for a volume of 9 mL; 0.023– 0.175 mL/kg of body weight). The total amount of absolute ethanol injected during a session ranged between 15 and 74.0 mL (for 0.26 – 0.89 mL/kg of body weight) with a mean amount of 37.7 mL (for 0.48 mL/kg of body weight; Table 2). Twenty-five patients underwent continuous infusion of nitroglycerin (range, 0.5–3.0 g/kg/min) in an attempt to attenuate the elevation of PAPs during the session and the recovery period. Among these 25 patients, nitroglycerin infusion was started at the beginning of embolotherapy in nine patients and in 16 patients it was administered after a few repeat injection of absolute ethanol. Hemodynamic values of Tbaseline, Thighest, Tfinal, and Tresp are presented in Table 3. The comparison of the hemodynamic profiles showed that the systolic, mean, and diastolic sABPs (P ⬍ .05, P ⬍ .05, and P ⬍ .05, respectively), systolic and mean PAPs (P ⬍ .01, and P ⬍ .01, respectively), and RVEF (P ⬍ .05) of Tfinal
were significantly higher than the values for T baseline . The systolic, mean, and diastolic sABPs (P ⬍ .05, P ⬍ .05, and P ⬍ .05, respectively) and PAPs (P ⬍ .01, P ⬍ .01, and P ⬍ .01, respectively), SVRI (P ⬍ .05), and PVRI (P ⬍ .05) of Thighest and Tresp were significantly higher than the corresponding values for Tfinal (Table 3). For the investigation of the largest of ⌬Tpost – Tpre, ⌬Tfinal – Tbaseline, and ⌬Tresp – Tfinal to determine the point of most significant hemodynamic change in a session, the systolic, mean, and diastolic sABPs (P ⬍ .01, P ⬍ .05, and P ⬍ .05, respectively) and the systolic and mean PAPs (P ⬍ .01 and P ⬍ .01, respectively) of ⌬Tfinal – Tbaseline were significantly lower than the corresponding values of the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal. However, the RVEF of ⌬Tfinal – Tbaseline was higher than the values of the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal (P ⬍ .05 and P ⬍ .05, respectively). However, there were no significant differences in the comparison between the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal (Table 4). The mean elevations of the systolic and mean PAPs of the largest of ⌬Tpost – Tpre
Volume 21
Number 1
Shin et al
•
85
Table 3 Cardiopulmonary Hemodynamic Profiles during Ethanol Embolotherapy Hemodynamic Parameter sABP (mm Hg) Systolic Mean Diastolic PAP (mm Hg) Systolic Mean Diastolic HR (beats/min) CI (L · min⫺1 · m⫺2) PCWP (mm Hg) CVP (mm Hg) RVEF (%) SVI (mL · m⫺2 · beat⫺1) RVEDVI (mL · m⫺2) RVESVI (mL · m⫺2) SVRI (dyne · s · cm⫺5 · m⫺2) PVRI (dyne · s · cm⫺5 · m⫺2) LVSWI (gm-m · m⫺2 · beat⫺1) RVSWI (gm-m · m⫺2 · beat⫺1)
Tbaseline
Thighest
Tfinal
Tresp
97.8 ⫾ 11.1 64.0 ⫾ 8.3 52.3 ⫾ 7.7
116.4 ⫾ 18.7*† 84.5 ⫾ 18.6*† 68.5 ⫾ 17.1*†
94.8 ⫾ 11.1* 64.5 ⫾ 7.5* 51.3 ⫾ 8.5*
120.8 ⫾ 19.7*† 78.2 ⫾ 10.4*† 62.8 ⫾ 9.8*†
23.7 ⫾ 4.4 18.6 ⫾ 5.1 12.9 ⫾ 3.3 72.1 ⫾ 10.6 5.10 ⫾ 2.36 14.0 ⫾ 2.7 9.1 ⫾ 2.8 37.7 ⫾ 4.2 72.1 ⫾ 33.9 94.0 ⫾ 28.6 62.2 ⫾ 54.3 1873.6 ⫾ 565.5 255.2 ⫾ 20.2 50.1 ⫾ 20.0 8.7 ⫾ 4.6
38.9 ⫾ 7.3*† 29.2 ⫾ 5.1*† 21.3 ⫾ 4.3*† 87.8 ⫾ 14.6* 6.11 ⫾ 1.67 19.6 ⫾ 5.5* 13.0 ⫾ 3.9* 38.4 ⫾ 7.1 63.9 ⫾ 25.3 111.0 ⫾ 28.5 72.8 ⫾ 36.3 2031.0 ⫾ 584.7† 305.9 ⫾ 54.6*† 52.3 ⫾ 18.9 13.6 ⫾ 5.9*
30.5 ⫾ 4.7* 23.3 ⫾ 4.4* 16.9 ⫾ 5.0 83.3 ⫾ 12.7 5.83 ⫾ 2.24 15.9 ⫾ 4.8 11.6 ⫾ 4.6 42.4 ⫾ 9.9* 72.3 ⫾ 34.4 106.2 ⫾ 30.1 65.4 ⫾ 28.1 1622.8 ⫾ 290.4 260.3 ⫾ 25.0 46.2 ⫾ 18.7 10.8 ⫾ 5.7
39.6 ⫾ 7.3*† 29.7 ⫾ 5.6*† 19.8 ⫾ 5.0*† 91.7 ⫾ 11.2* 6.61⫾ 2.48 18.5 ⫾ 5.3* 13.0 ⫾ 5.4 41.2 ⫾ 8.6* 74.1 ⫾ 27.6 122.1 ⫾ 24.2* 82.1 ⫾ 34.7 1714.9 ⫾ 349.1 293.4 ⫾ 37.3*† 60.8 ⫾ 23.1 15.4 ⫾ 7.6*†
Note.—Data are presented as means ⫾ SD. * P ⬍ .05 vs Tbaseline. † P ⬍ .05 vs Tfinal.
were 14.8 mm Hg ⫾ 7.7 and 10.7 mm Hg ⫾ 6.2, and the mean elevations of ⌬Tresp – Tfinal were 11.7 mm Hg ⫾ 5.3 and 7.4 mm Hg ⫾ 4.7, respectively. The systolic, mean, and diastolic PAPs of Tpre showed a progressive increase from Tbaseline to Tfinal (P ⬍ .01, P ⬍ .01, and P ⬍ .01, respectively) during a session. However, other preinjection hemodynamic parameters did not show a statistically significant increase or decrease as the number of absolute ethanol injections increased during a session. Among the hemodynamic values of ⌬Tpost – Tpre, the systolic and mean PAPs (P ⬍ .01 and P ⬍ .01, respectively) showed a significant decrease along with an increase in the number of bolus injections of absolute ethanol during a session. The elevations of systolic and mean PAPs per 0.1 mL of absolute ethanol per kilogram of body weight of ⌬Tpost – Tpre were 11.1 mm Hg ⫾ 8.7 and 9.2 mm Hg ⫾ 6.6 at the first injection, and 4.2 mm Hg ⫾ 3.1 and 4.9 ⫾ 5.2 at the seventh injection (Fig 1). In 13 patients (43.3%), the highest mean PAPs were reached after the first bolus injection of absolute ethanol, and in nine patients (30.0%) after the final injection. We found that the largest of ⌬Tpost – Tpre of the mean PAPs caused by a single bolus ethanol
injection was evident after the first injection of absolute ethanol in 20 patients (67%) and the lowest single dose of absolute ethanol that produced the maximum elevations of the mean PAPs by the first injection was 0.05 mL/kg body weight. The evaluation of the changes in the hemodynamic profiles according to the increase in volume of a single bolus injection of absolute ethanol from 0.023 to 0.175 mL/kg body weight showed that the systolic PAP (P ⫽ .02), CVP (P ⬍ .01), PCWP (P ⫽ .02), and PVRI (P ⬍ .01) significantly increased in accordance with the increased single volume of absolute ethanol. However, the CI decreased as the amount of bolus injection of absolute ethanol increased (Fig 2). Among the other parameters, we observed a significant increase of the RVEDVI and RVESVI at a dose of more than 0.14 mL/kg body weight of a single bolus injection of absolute ethanol (Fig 3).
DISCUSSION The major findings in this study include that a significant hemodynamic change can arise after the first bolus injection of absolute ethanol and during the period of restoration of spon-
taneous respiration for ethanol embolotherapy of AV malformations of the extremities. In addition, a single bolus injection of absolute ethanol of more than 0.14 mL/kg of body weight can compromise right ventricular function. Based on these results, it can be recommended that the volume of absolute ethanol be determined carefully from the first injection, and the volume should be based on a consideration of PAP in addition to the amount of contrast material needed to fill the nidus. Finally, careful attention is strongly advised to avoid significant hemodynamic changes during the restoration of spontaneous respiration. Although the use of ethanol embolotherapy has demonstrated many promising results, particularly an increased chance for complete remission of some AV malformations, patients are repeatedly exposed to the risks of minor or major complications associated with the chemical toxicity of ethanol as most inoperable soft-tissue AV malformations mandate the need for multistage ethanol embolotherapies. In our previous study (15), we demonstrated that the risk of pulmonary arterial hypertension did not decrease as sessions were repeated, which indicated that the risks of cardiopulmonary complications are sus-
86
•
Effect of Repeat Ethanol Injections in Extremity AVM Embolotherapy
January 2010
JVIR
Table 4 Hemodynamic Changes Caused by a Single Injection and the Total Amount of Absolute Ethanol and after Restoration of Spontaneous Respiration during a Session Hemodynamic Parameter sABP (mm Hg) Systolic Mean Diastolic PAP (mm Hg) Systolic Mean Diastolic HR (beats/min) CI (L · min⫺1 · m⫺2) PCWP (mm Hg) CVP (mm Hg) RVEF (%) SVI (mL · m⫺2 · beat⫺1) RVEDVI (mL · m⫺2) RVESVI (mL · m⫺2) SVRI (dyne · s · cm⫺5 · m⫺2) PVRI (dyne · s · cm⫺5 · m⫺2) LVSWI (gm-m · m⫺2 · beat⫺1) RVSWI (gm-m · m⫺2 · beat⫺1)
Largest ⌬Tpost – Tpre 19.6 ⫾ 22.4 (⫺1.5, 13.5, 35.7) 16.2 ⫾ 18.0 (5.0, 12.0, 23.5) 13.6 ⫾ 16.0 (2.8, 12.0, 17.3)
⌬Tfinal – Tbaseline
⌬Tresp – Tfinal
1.4 ⫾ 9.5 (⫺4.0, 2.0, 6.0)*† 3.3 ⫾ 6.7 (⫺1.0, 5.0, 8.0)*† 1.4 ⫾ 7.2 (⫺4.0, 1.0, 8.0)*†
28.7 ⫾ 14.0 (17.0, 28.0, 32.0) 15.6 ⫾ 9.5 (7.0, 12.5, 21.0) 12.4 ⫾ 8.3 (4.0, 12.0, 18.0)
14.8 ⫾ 7.7 (9.0, 13.5, 22.3) 3.7 ⫾ 3.7 (1.0, 2.0, 7.0)*† 11.7 ⫾ 5.3 (8.0, 10.0, 13.0) 10.7 ⫾ 6.2 (5.0, 11.0, 16.0) 3.4 ⫾ 2.4 (2.0, 4.0, 5.0)*† 7.4 ⫾ 4.7 (5.0, 6.0, 8.0) 7.1 ⫾ 4.4 (2.8, 8.0, 11.3) 2.7 ⫾ 3.0 (0, 3.5, 5.0) 3.9 ⫾ 4.5 (⫺1.0, 3.0, 7.5) 10.2 ⫾ 9.9 (1.0, 10.5, 19.25) 7.8 ⫾ 9.0 (1.0, 7.0, 14.0) 9.4 ⫾ 9.4 (1.0, 4.0, 16.0) 0.27 ⫾ 1.45 (⫺.37, 0.54, 1.11) 0.97 ⫾ 1.30 (.47, 1.21, 1.93) 1.0 ⫾ 0.94 (.05, 1.10, 1.77) 6.0 ⫾ 4.1 (3.0, 5.0, 9.8) 2.1 ⫾ 2.6 (0, 2.0, 3.0)* 3.5 ⫾ 3.5 (1.0, 3.0, 6.0) 3.8 ⫾ 3.7 (.8, 3.0, 7.0) 2.0 ⫾ 2.6 (0, 1.0, 3.0) 2.0 ⫾ 4.2 (0, 1.0, 3.0) ⫺0.2 ⫾ 5.0 (⫺5.0, ⫺1.0, 4.0) 6.1 ⫾ 5.0 (2.8, 8.5, 12.5)*† ⫺0.2 ⫾ 5.8 (⫺4.0, 0, 2.0) ⫺4.2 ⫾ 20.2 (⫺18.8, ⫺1.9, 7.3) 4.4 ⫾ 15.0 (⫺10.0, 7.9, 13.0) 4.7 ⫾ 11.8 (⫺1.8, 2.8, 12.3) 7.8 ⫾ 21.3 (⫺11.5, 5.3, 26.9) 2.1 ⫾ 22.4 (⫺20.8, ⫺1.5, 24.4) ⫺20.7 ⫾ 42.4 (⫺7.3, 6.8, 32.1) 3.3 ⫾ 20.0 (⫺19.5, ⫺.3, 11.6) 4.4 ⫾ 15.6 (⫺56.1, ⫺20.0, 4.1) 17.2 ⫾ 43.9 (⫺16.6, 10.2, 54.7) 78.0 ⫾ 305.7 (⫺86.8, 82.0, 255.8) ⫺156.2 ⫾ 284.4 (⫺470.0, ⫺174.0, 7.0) 94.0 ⫾ 227.9 (⫺74.0, 72.0, 355.0) 39.1 ⫾ 43.3 (16.5, 30.0, 73.0)
14.8 ⫾ 30.8 (⫺21.0, 5.0, 37.0)*
7.3 ⫾ 15.6 (⫺4.1, 8.8, 18.7)
2.5 ⫾ 11.9 (⫺7.3, 1.4, 15.1)
5.7 ⫾ 3.1 (3.5, 5.2, 8.1)
3.4 ⫾ 3.8 (1.8, 4.2, 4.6)
26.5 ⫾ 27.5 (⫺1.0, 30.0, 40.5) 16.2 ⫾ 10.9 (9.1, 13.0, 28.9) 4.2 ⫾ 3.6 (0.3, 5.2, 6.4)
Note.—Data presented as means ⫾ SD where applicable; the numbers in parenthesis are 25th percentile, median, and 75th percentile. There are significant differences of the hemodynamic profiles for the comparison between the largest of ⌬Tpost – Tpre and ⌬Tfinal – Tbaseline, ⌬Tresp – Tfinal, and ⌬Tfinal – Tbaseline. However, there were no statistical differences between the largest of ⌬Tpost – Tpre and ⌬Tresp – Tfinal. * P ⬍ .05 vs largest of ⌬Tpost – Tpre. † P ⬍ .05 vs ⌬Tresp – Tfinal.
Figure 1. Changes of systolic and mean PAPs per 0.1 mL of absolute ethanol per kilogram of body weight of ⌬Tpost – Tpre (differences among the systolic, mean, and diastolic PAPs recorded when the mean PAP reached a maximum value after injection of absolute ethanol and immediately before the injection of absolute ethanol) decreased according to the increase of the number of injections during a session. Each ethanol injection was performed with at least a 10-minute interval. The changes of PAPs with more than seven injections are not depicted, as the number of cases was not sufficient for statistical analysis. Statistical analysis was performed by use of a mixed model. (*P ⬍ .05.)
tained during multisession ethanol embolotherapies. Ko et al (16) reported that PAP elevations were significantly correlated with the pulmonary arterial ethanol level in patients treated without the use of vascular occlusion techniques. Ethanol injection can result in severe pulmonary vasospasm and acutely increase thin-walled right ventricular afterload and decrease right ventricular cardiac output (17). These effects can be further aggravated by systemic effects of alcohol affecting right atrial and ventricular contractility, including chronotropic and inotropic functions (18 –21). It is important to consider direct myocardial depression caused by the systemic effect of the total amount of absolute ethanol. However, our results have elucidated that significant hemodynamic changes can initiate from the first injection of absolute ethanol, and hemodynamic changes are affected more by the single bolus injection volume rather than by the total amount of absolute
Volume 21
Number 1
Shin et al
10
20
•
87
90 80 PVRI (dyne·s·cm-5·m
8
15
70
*
60 6
*
50
10 *
4
∗
40 30
5
20
2
10
0.17 -0.18
0.15 -0.16
0.13 -0.14
0.11 -0.12
0.09 -0.10
0.07 -0.08
b.
0.05 -0.06
0
0.17-0.18
0.15-0.16
0.13-0.14
0.11-0.12
0.09-0.10
0.07-0.08
0.05-0.06
a.
-2
†
0.03-0.04
0.17- 0.18
0.15- 0.16
0.13- 0.14
0.11- 0.12
0.09- 0.10
0.07- 0.08
0.05- 0.06
0.03- 0.04
-5
0
0.03 -0.04
0
c.
Figure 2. Systolic PAP (a), CVP and PCWP (b), and PVRI (c) of ⌬Tpost – Tpre (differences between hemodynamic parameters measured when the mean PAP reached a maximum value after injection of absolute ethanol and immediately before the injection of absolute ethanol) increased as the dose of a single injection of absolute ethanol increased. However, the CI (b) decreased according to the increase of a single injection dose of absolute ethanol. Statistical analysis was performed with use of a mixed model. (*†P ⬍ .05.)
Figure 3. Sharp increase of RVEDVI and RVESVI from a single dose of 0.14 mL of absolute ethanol per kilogram of body weight of ⌬Tpost – Tpre (differences between hemodynamic parameters measured when the mean PAP reached a maximum value after injection of absolute ethanol and immediately before the injection of absolute ethanol). (*†P ⬍ .05 by use of mixed-model analysis.)
ethanol used during ethanol embolotherapy. First, the systolic, mean, and diastolic PAPs and PVRI of Thighest were significantly higher than the corresponding values of Tfinal, which can reflect the accumulative effect of the total amount of absolute ethanol during a session. Second, even though repeated measurements of hemodynamic param-
eters at Tpre demonstrated that the systolic, mean, and diastolic PAPs increased from the first to the final injection during a session, the systolic and mean PAPs of the largest of ⌬Tpost – Tpre changed more significantly than the corresponding values of ⌬Tfinal – Tbaseline, and there was a lack of significant change of RVEF. In addition, the
systolic and mean PAPs of ⌬Tpost – Tpre showed a significant decrease along with an increase in the number of bolus injections of absolute ethanol during a session. Third, we observed significant increases of RVEDVI and RVESVI at a dose of greater than 0.14 mL/kg of body weight of a single injection of absolute ethanol. The volume of ethanol that resulted in increases of RVEDVI and RVESVI may be considered as an initiating volume that begins to impose a burden on right ventricular contractility. Fourth, the progressive increase in the systolic PAP, CVP, PCWP, and PVRI and decrease in CI of ⌬Tpost – Tpre according to the incremental increase in the single bolus injection volume of absolute ethanol per body weight indicates that a larger single bolus injection dose of absolute ethanol can affect right and left ventricular function. Greyson et al (22) have reported that although right ventricular systolic and diastolic pressures returned to normal values, right ventricular contractile dysfunction persisted for at least 2 hours after 1 hour of pulmonary artery constriction in an animal study. Therefore, we suggest ethanol embolotherapy be performed with a minimal therapeutic volume from the first injection, and the cardiopulmonary hemodynamic profiles be closely monitored when injecting more than 0.14
88
•
Effect of Repeat Ethanol Injections in Extremity AVM Embolotherapy
mL/kg of body weight of absolute ethanol in patients with an AV malformation of the extremities. Our study also showed that significant hemodynamic changes could arise after resuming spontaneous respiration. In addition, ⌬Tresp – Tfinal did not show any significant difference versus the largest value of ⌬Tpost – Tpre. Significant physiologic changes in the cardiopulmonary system can occur during the transition from mechanical ventilation to spontaneous respiration. Three main contributing factors involved in the transition process are the following. First, the shift of intrathoracic pressure from positive to negative increases venous return; second, a sympathetic discharge occurs secondary to tracheal irritation and hypercapnia; and third, there is an afterload increase resulting from the elevated systolic blood pressure and lowered pleural pressure (23–25). A combination of these factors may contribute to the aggravation of already elevated PAPs after repeated injections of absolute ethanol. Therefore, we recommend continuous PAP monitoring and administration of nitroglycerin during the recovery period to ensure the safe recovery of patients. Alternatively, maintenance of mechanical ventilation may be considered as the already elevated PAPs can be acutely increased again (as high as 11.7 mm Hg for the systolic PAP and 7.4 mm Hg for the mean PAPs) during the period of weaning of mechanical ventilation. One of the limitations of this study is that we were unable to measure the blood alcohol level from the pulmonary vascular bed at the point of the maximum mean PAP as we used the thermodilution method to obtain values of the hemodynamic profiles. Therefore, we could not analyze the correlation between the blood alcohol concentration at the pulmonary arterial bed and hemodynamic parameters at the time of the maximum PAP values. Additionally, significant hemodynamic changes after the first injection and the subsequent decrease in systolic and mean PAPs of ⌬Tpost – Tpre along with the increase in the number of bolus injections of absolute ethanol during a session may result from a progressive decrease in sympathetic stimu-
lation arising from pain at the time of injection. However, we could not clearly elucidate or rule out the effect of sympathetic stimulation caused by pain on the elevation of PAPs and the corresponding hemodynamic parameters. Finally, as most of the patients (83.3%) received nitroglycerin during the procedure and recovery period, and as the initiating time of continuous administration of nitroglycerin was different for each patient, we could not evaluate the effect of nitroglycerin on the attenuation of PAP increases during ethanol embolotherapy. In conclusion, significantly greater hemodynamic changes can arise from a single bolus injection of absolute ethanol rather than as a result of the total amount of absolute ethanol used during ethanol embolotherapy of AV malformations of the extremities. Therefore, there should be close collaboration between an anesthesiologist and an interventional radiologist to determine the dose of absolute ethanol and injection intervals, and it is recommended for an interventional radiologist to proceed with the lowest therapeutic dose and use an injection volume of absolute ethanol that is no greater than 0.14 mL/kg body weight. In addition, the cardiopulmonary hemodynamic profiles should be closely monitored when injecting more than 0.14 mL/kg body weight of absolute ethanol in patients with AV malformation of the extremities. Restoration of spontaneous respiration can also aggravate hemodynamic changes. Therefore, it may be advisable to consider maintaining mechanical ventilation during the recovery period as the already elevated PAPs at the end of session can be significantly elevated again at the time of the restoration of spontaneous respiration. References 1. Hyodoh H, Hori M, Akiba H, Tamakawa M, Hyodoh K, Hareyama M. Peripheral vascular malformations: imaging, treatment approaches, and therapeutic issues. Radiographics 2005; 25(Suppl 1):S159 –S171. 2. Do YS, Yakes WF, Shin SW, et al. Ethanol embolization of arteriovenous malformations: interim results. Radiology 2005; 235:674 – 682. 3. Yakes WF, Haas DK, Parker SH, et al. Symptomatic vascular malformations: ethanol embolotherapy. Radiology 1989; 170:1059 –1066.
January 2010
JVIR
4. Yakes WF, Pevsner P, Reed M, Donohue HJ, Ghaed N. Serial embolizations of an extremity arteriovenous malformation with alcohol via direct percutaneous puncture. AJR Am J Roentgenol 1986; 146:1038 –1040. 5. Lee BB, Bergan JJ. Advanced management of congenital vascular malformations: a multidisciplinary approach. Cardiovasc Surg 2002; 10:523–533. 6. Rimon U, Garniek A, Galili Y, Golan G, Bensaid P, Morag B. Ethanol sclerotherapy of peripheral venous malformations. Eur J Radiol 2004; 52:283–287. 7. Yakes WF, Rossi P, Odink H. How I do it. Arteriovenous malformation management. Cardiovasc Intervent Radiol 1996; 19:65–71. 8. Wong GA, Armstrong DC, Robertson JM. Cardiovascular collapse during ethanol sclerotherapy in a pediatric patient. Paediatr Anaesth 2006; 16:343– 346. 9. Garel L, Mareschal JL, Gagnadoux MF, Pariente D, Guilbert M, Sauvegrain J. Fatal outcome after ethanol renal ablation in child with end-stage kidneys. AJR Am J Roentgenol 1986; 146:593– 594. 10. Lee BB, Do YS, Yakes W, Kim DI, Mattassi R, Hyon WS. Management of arteriovenous malformations: a multidisciplinary approach. J Vasc Surg 2004; 39:590 – 600. 11. Villavicencio JL. Primum non nocere: Is it always true? The use of absolute ethanol in the management of congenital vascular malformations. J Vasc Surg 2001; 33:904 –906. 12. Behnia R. Systemic effects of absolute alcohol embolization in a patient with a congenital arteriovenous malformation of the lower extremity. Anesth Analg 1995; 80:415– 417. 13. Lee BB, Kim DI, Huh S, et al. New experiences with absolute ethanol sclerotherapy in the management of a complex form of congenital venous malformation. J Vasc Surg 2001; 33: 764 –772. 14. Mitchell SE, Shah AM, Schwengel D. Pulmonary artery pressure changes during ethanol embolization procedures to treat vascular malformations: can cardiovascular collapse be predicted? J Vasc Interv Radiol 2006; 17: 253–262. 15. Shin BS, Do YS, Lee BB, et al. Multistage ethanol sclerotherapy of softtissue arteriovenous malformations: effect on pulmonary arterial pressure. Radiology 2005; 235:1072–1077. 16. Ko JS, Kim JA, Do YS, et al. Prediction of the effect of injected ethanol on pulmonary arterial pressure during sclerotherapy of arteriovenous malformations: relationship with dose
Volume 21
Number 1
of ethanol. J Vasc Interv Radiol 2009; 20:39 – 45. 17. Yakes WF, Krauth L, Ecklund J, et al. Ethanol endovascular management of brain arteriovenous malformations: initial results. Neurosurgery 1997; 40:1145– 1152. 18. Kelly LF, Goldberg SJ, Donnerstein RL, Cardy MA, Palombo GM. Hemodynamic effects of acute ethanol in young adults. Am J Cardiol 1996; 78: 851– 854. 19. Guarnieri T, Lakatta EG. Mechanism of myocardial contractile depression by clinical concentrations of ethanol: a
Shin et al
study in ferret papillary muscles. J Clin Invest 1990; 85:1462–1467. 20. Delgado CE, Gortuin NJ, Ross RS. Acute effects of low doses of alcohol on left ventricular function by echocardiography. Circulation 1975; 51:535–540. 21. Campbell PH, Barker LA, McDonough KH. The effect of acute ethanol exposure on the chronotropic and inotropic function of the rat right atrium. J Pharm Pharmacol 2000; 52:1001–1010. 22. Greyson C, Xu Y, Cohen J, Schwartz GG. Right ventricular dysfunction persists following brief right ventricular pressure overload. Cardiovasc Res 1997; 34:281–288.
•
89
23. Lemaire F, Teboul JL, Cinotti L, et al. Acute left ventricular dysfunction during unsuccessful weaning from mechanical ventilation. Anesthesiology 1988; 69:171–179. 24. Mancebo J. Weaning from mechanical ventilation. Eur Respir J 1996; 9:1923– 1931. 25. Jardin F, Delorme G, Hardy A, Auvert B, Beauchet A, Bourdarias JP. Reevaluation of hemodynamic consequences of positive pressure ventilation: emphasis on cyclic right ventricular afterloading by mechanical lung inflation. Anesthesiology 1990; 72:966–970.