Xi YANG ,Chen HUA ,Hui CHEN,Lan LUO,Gang MA,Xiaojie HU,YaJing QIU,Wenxin YU,Lei CHANG,Yunbo JIN ,Xiaoxi LIN

ABSTRACT Background Ethanol embolotherapy is considered an optimal choice for the treatment of arteriovenous malformations (AVMs);however,there are some complications associated with this treatment.This study aimed to prospectively investigate systemic hemodynamic changes in high-flow AVMs using ethanol embolotherapy.Methods From September 2012 to September 2014,34 male patients and 26 female patients with AVMs who underwent embolotherapy (100 sessions in total) with absolute ethanol were included in this study.Invasive systolic blood pressure (SBP) and heart rate (HR) were recorded before and after each injection and throughout the procedure.Differences between the initial and highest SBP (ΔmaxSP) and HR values (ΔmaxHR),as well as the initial and final SBP (ΔSP) and HR (ΔHR) values,were analyzed.We aimed to explore the potential association between these values and the amount of ethanol that was used.Results The total ethanol used was variable (0.01-0.40 mL/kg;mean,0.20 mL/kg).SBP and HR increased after ethanol injection in most sessions (91 in 100 sessions).SBP decreased in 9 sessions (9 in 100 sessions),while HR,oxygen saturation,and end-tidal CO2decreased in one of the 9 sessions.ΔmaxSP and ΔmaxHR averaged 38.4 mmHg and 27.8 bpm,respectively (both P＜0.05),while ΔSP and ΔHR averaged 3.4 mmHg and 4.0 bpm,respectively (both P＜0.05).ΔmaxSP and ΔmaxHR were positively correlated with the total dose of ethanol injected.Conclusions Elevations in SBP and HR during ethanol embolotherapy are common,temporary,and most likely pain-mediated;these increases tend to be positively correlated with ethanol dose.Hypotension may be regarded as an acute complication of ethanol embolotherapy.Hypotension combined with bradycardia,oxygen desaturation,and decreased end-tidal CO2may be a potential predictor of cardiovascular collapse.

KEY WORDS Systemic hemodynamic change;Ethanol embolotherapy;Arteriovenous malformations;Hypotension;Blood pressure;Heart rate

INTRODUCTION

Arteriovenous malformations (AVMs) are congenital vascular lesions characterized by hypertrophied inflow arteries and shunting through a primitive vascular nidus into tortuous dilated outflow veins,with no intervening capillary bed.AVMs are refractory due to their resistance to treatment and have a high recurrence rate,which presents a major challenge to clinicians worldwide.Clinical management focuses on multidisciplinary approaches,including surgery and embolotherapy with various agents[1].Ethanol has been used for the embolization of AVMs due to its advantage in reducing recurrence by endothelial cell destruction in the nidus[2].Ethanol embolotherapy has been a primary therapeutic choice for AVMs in our center since 2006 and has achieved an excellent outcomes[2].However,ethanol embolotherapy has various acute and chronic complications associated with serious morbidity.Complications include skin and/or mucous membrane bullae,skin and/or subcutaneous tissue necrosis,nerve damage,deep venous thrombosis and/or pulmonary embolism,massive muscle or cartilage necrosis,transient hemoglobinuria,pulmonary hypertension,and cardiovascular collapse (CVC)[3].

CVC is a rare and dangerous complication that has been reported 11 times in the literature[4-14].The exact cause of CVC remains unknown;currently,there are two predominant hypotheses.One hypothesis proposes that the hemodynamic changes caused by ethanol may induce precapillary pulmonary artery vascular spasm or microthromboembolism in the pulmonary circulation and result in increased pulmonary artery pressure (PAP) that leads to elevated right ventricular (RV) afterload,decreased RV contractility,and decreased RV cardiac output.This sequence progresses to decreased left ventricular filling,decreased left ventricular cardiac output,systemic hypotension,and decreased coronary artery perfusion[2].Another hypothesis attributes the systemic dynamic change to a direct toxic effect of ethanol on the patient’s biventricular myocardium conducting system,leading to bradyarrhythmia,hypotension,and,eventually,cardiac arrest[6].

After reviewing the reported cases (14 patients) of CVC resulting from ethanol embolotherapy,hypotension (7 of 14) was the most common initial clinical manifestation of CVC,followed by bradycardia (4 of 14),desaturation(3 of 14),and tachypnea (2 of 14) (Table 1)[4-14].Current studies have mostly focused on pulmonary artery hemodynamic changes and have elucidated the correlation between ethanol injection and pulmonary hypertension[12-19].The aim of our study was to prospectively investigate systemic hemodynamic changes in high-flow soft-tissue AVMs in low-dose ethanol embolotherapy.

MATERIALS AND METHODS

Patients

Approval from the Institutional Review Board of the Shanghai Ninth People’s Hospital was obtained for the prospective study,and written consent was obtained from all patients after a discussion regarding the advantages and risks of ethanol embolotherapy.A total of 60 patients(34 men and 26 women) with AVMs who underwent ethanol embolization under general anesthesia between September 2012 and September 2014 were included in the study.SBP and HR were monitored before,during,and after ethanol embolotherapy in all patients.SBP and BP were recorded before and after each injection,as well as before and after the entire procedure.Oxygen saturation,end-tidal CO2,and three-lead ECG were also monitored.All 60 patients underwent 100 sessions of ethanol embolotherapy under general anesthesia.SBP was measured for systolic,mean,and diastolic values using a pressure monitoring catheter (Edwards Lifesciences,Irvine,CA,USA).Fifty-three (88.3%) of the soft-tissue AVMs were located in the head and neck region,while 7 (11.7%) were found in the extremities (Table 2).

Procedures

All embolotherapy procedures involving the use of absolute ethanol were performed under general anesthesia.Invasive SBP monitoring was implemented by introducing a pressure-monitoring catheter into the left dorsal pedal artery.Superselective arteriography was performed to determine the angioarchitecture and hemodynamic features of the major compartments of the AVMs.Ethanol embolization was directed against the nidus instead of the vascular feeders.A 21-gauge needle with butterfly grips (Surflo Winged Infusion Set,21-gauge,0.75 inch;Terumo,Belgium) and a microcatheter (Prowler 14 Infusion Catheter;Codman &Shurtleff,Inc.,Miami Lakes,Florida) was advanced based on the preceding angiography.Angiography was performed to determine the exact flow characteristics of the AVMs,volume of ethanol used during embolization,and rate of injection.Test injections of contrast medium (Ultravist 300;Schering,Berlin,Germany) were performed under fluoroscopic monitoring.The volume of ethanol used was determined based on the amount of contrast medium required to fill the AVM nidus without opacifying normal vessels.In most cases,absolute ethanol was used;however,diluted ethanol (50%-70%) with a contrast medium (Ultravist 300;Schering) was chosen in the case of microfistulas.

After ethanol injections,arteriography was performed via the arterial approach after an interval of 5-10 min to determine whether the AVMs were completely embolized.This technique was meticulously repeated until complete embolization of at least one compartment of the AVMs was achieved.The volume of ethanol used ranged from 2 mL (most used) to 4 mL in a single injection,and the total dose never exceeded 0.4 mL/kg in one session.The SBP and HR values were recorded at the start and end of each procedure,and the highest/lowest SBP and HR val-ues were recorded after the injection of absolute ethanol.

Table 1 Review of published literature on CVC as a complication of ethanol injection [4-14]

Table 2 Demographic and ethanol embolotherapy data for patients

Anesthesia Techniques

Anesthesia was induced with propofol (2 mg/kg),fentanyl(0.02 mg/kg),and cisatracurium (2 mg/kg).Patients were maintained with mechanical ventilation on pure oxygen with sevoflurane at an end-tidal concentration of 2% and propofol at 4 mg/kg/h intravenously.Oxygen saturation was maintained between 97% and 100%,with an endtidal CO2level of 35-45 mmHg.Patients were fully hydrated with equilibrium liquid before the procedure and at approximately 8 mL/kg/h during the procedure based on physiological needs and fluid loss.When hypotension or bradycardia occurred,ephedrine or norepinephrine was administered intravenously.We excluded SBP and HR data whenever extra anesthetics or medications were used that would affect these parameters.

Data Acquisition

Hemodynamic parameters,including systemic SBP (systemic systolic,mean,and diastolic BP) and HR were recorded at the following points:hemodynamic profiles measured 15 min before ethanol injection (SP1,MP1,DP1,and HR1);hemodynamic profiles measured within 3 min after injections at the time of maximum/minimum mean BP value and HR (SP2,MP2,DP2,and HR2);hemodynamic profiles measured 15 min after the final injection of absolute ethanol (SP3,MP3,DP3,and HR3);the difference of values between SP3 and SP1 (ΔSP),MP3 and MP1 (ΔMP),DP3 and DP1 (ΔDP),and HR3 and HR1 (ΔHR);the largest difference of hemodynamic changes between SP2 and SP1 (ΔmaxSP),MP2 and MP1(ΔmaxMP),DP2 and DP1 (ΔmaxDP),and HR2 and HR1(ΔmaxHR) during a session (Table 3).

Table 3 Date acquisition

Statistical Analysis

The initial,maximum/minimum,and final SBP values(systolic,mean,and diastolic BP) and HR were investigated and compared with the total amount of absolute ethanol used in each session.Data are presented as mean ± standard deviation (SD).All the sessions were assumed to be independent.Comparisons of SP,MP,DP,and HR were performed using pairedt-tests (IBM SPSS Statistics 19.0,IBM Corporation,Armonk,NY,USA).Bivariate correlations were assessed to determine whether ethanol dose correlated with ΔmaxSP and ΔmaxHR.If the correlations were significant,simple linear regression analyses were performed to explore whether the cumulative ethanol dose had a significant effect on ΔmaxSP and ΔmaxHR.Differences were considered statistically significant atP＜0.05.

RESULTS

The total amount of absolute ethanol used to control the AVM nidus varied;0.3-24.0 mL was used per session(range,0.01-0.40 mL/kg),with a mean of 10.0 mL.The mean number of therapy sessions per patient was 1.7.Among the 100 total sessions of ethanol embolotherapy,95 sessions were performed using the direct puncture method,and the remaining 5 sessions were performed using both methods (Table 2).The mean values of initial systolic,mean,and diastolic blood pressures were 90.8 ± 12.2 mmHg (SP1),70.0 ± 9.6 mmHg (MP1),and 49.2 ± 9.2 mmHg (DP1),respectively.The mean values of the highest systolic,mean,and diastolic blood pressures were 129.2 ± 23.6 mmHg (SP2),103.0 ± 18.6 mmHg(MP2),and 76.8 ± 15.7 mmHg (DP2),respectively.The mean SBP values at the end of the sessions were 94.2 ± 12.8 mmHg (SP3),72.7 ± 10.2 mmHg (MP3),and 51.3 ± 9.2 mmHg (DP3),respectively.The mean initial,highest,and final HR values were 69.3 ± 13.8 bpm(HR1),97.0 ± 21.1 bpm (HR2),and 73.3 ± 12.7 bpm(HR3),respectively.The SP2,MP2,DP2,and HR2 were significantly greater than SP1,MP1,DP1,and HR1,respectively (ΔmaxSP=38.4 ± 23.3 mmHg;ΔmaxMP =33.0 ± 18.0 mmHg;ΔmaxDP=27.7 ± 15.1 mmHg;ΔmaxHR=27.8 ± 15.7 bpm;P＜0.05 for each group).These increases were mainly due to the pain sensed by the vasa nervorum in the arteries and veins.The SP3,MP3,DP3,and HR3 were also significantly greater than SP1,MP1,DP1,and HR1,respectively (ΔSP=3.4 ± 13.0 mmHg;ΔMP=2.7 ± 10.1 mmHg;ΔDP=2.1 ± 9.4 mmHg;ΔHR=4.0 ± 8.3 bpm;P＜0.05 for each group) (Fig.1).The ΔmaxSP and ΔmaxHR were positively correlated with the total dose of ethanol injected(Pearsonr=0.30,R2=0.11,P＜0.05;Pearsonr=0.34,R2=0.09,P＜0.05) (Fig.2).

Fig.1 Changes in systemic systolic,mean,and diastolic blood pressure (SP,MP,and DP) and HR during ethanol embolotherapy. Data are presented as mean ± SD.The mean value of the initial SBPs was 90.8 ± 12.2 mmHg (SP1).The mean value of the highest SBPs was 129.2 ± 23.6 mmHg (SP2).The mean SBP value at the end of the sessions was 94.2 ± 12.8 mmHg.The mean initial,highest,and final HR values were 69.3 ± 13.8 bpm,97.0 ± 21.1 bpm,and 73.3 ± 12.7 bpm,respectively.The SP2 and HR2 were significantly greater than SP1 and HR1,respectively (ΔmaxSP=38.4 ± 23.3 mmHg,ΔmaxHR=27.8 ± 15.7 bpm;P＜0.05).The SP3 and HR3 were also significantly greater than SP1 and HR1,respectively (ΔSP=3.4 ± 13.0 mmHg,ΔHR=4.0 ± 8.3 bpm;P＜0.05).

Fig.2 Changes in systemic SBP (ΔmaxSP) and HR (ΔmaxHR) on the Y-axis plotted against the X-axis ethanol dose (mL/kg).The relationship was statistically significant,with both P-values less than 0.05;however,the correlation was not strong.The Pearson correlation values were 0.30 and 0.34,and R2 values equal to 0.11 and 0.09,respectively.

Decreases in SBP immediately after ethanol injection occurred in 9 of the 100 sessions.The noninvasive oxygen saturation decreased from 97% to 93%,the end-tidal CO2decreased from 40 mmHg to 30 mmHg,and HR decreased from 75 bpm to 53 bpm in one of the 9 sessions(Table 4).Ethanol injection was stopped immediately after a decrease in SBP was observed.Five of these cases were transient and were restored within two min after stopping ethanol injection;the other 4 were persistent and needed intravenous vasopressor (ephedrine) to restore SBP.The HR,end-tidal CO2,and oxygen saturation were also restored 5 min after intravenous ephedrine injection.SBP and HR were normal after the procedure in all 9 sessions.

Table 4 Patients with an SBP decrease immediately after injection

DISCUSSION

SBP and HR Increase after Ethanol Injection

We observed that SBP and HR increased after ethanol injection during ethanol embolization procedures in most sessions (92%),with an average of 38.4 mmHg (SBP) and 27.8 bpm (HR).After the procedure,SBP and HR were slightly higher than before the procedure:an average SBP of 3.4 mmHg and HR of 4.0 bpm.Three previous studies reported similar results for SBP and HR increase during ethanol embolotherapy.Mitchell et al.found an average increase of 11.6 mmHg in noninvasive cuff systemic SBP after ethanol injection in the treatment of vascular malformations,which was associated with a rise in systolic PAP[12].Ko et al.and Behnia also observed an increase in SBP and HR during ethanol embolotherapy for arteriovenous malformations[17,20].Increases in SBP and HR are believed to be related to pain resulting from ethanol injection causing sympathetic stimulation,even when patients are under general anesthesia[12].We found that the effect of ethanol on hemodynamics was temporary since SBP and HR decreased in less than 5 min after the increase caused by ethanol injection and remained slightly higher after the procedure compared to values obtained before the procedure.We suggest that increases in SBP and HR are common,temporary,and do not pose a risk during ethanol injection.

Hypotension and Cardiovascular Collapse

In our study,hypotension occurred in 9% of the sessions after ethanol injection.Five of the hypotension cases were transient,and four were persistent until intravenous ephedrine was administered.The HR,end-tidal CO2,and oxygen saturation decreased in one of the 9 sessions.Similar results were found in a retrospective study of 67 patients who received percutaneous ethanol injections for the treatment of hepatocellular carcinoma.Nimmaanrat et al.reported 10 patients (14.9%) with hypotension during the procedure,in which one of the patients also experienced bradycardia and oxygen desaturation leading to cardiovascular collapse[13].

Cardiovascular collapse is characterized by acute and severe hypotension that can decrease coronary perfusion and lead to pulseless electrical activity that necessitates cardiopulmonary resuscitation[13].Acute hypotension can be caused by a decrease in acute left ventricular filling and/or a decrease in acute left ventricular contractility.An acute decreased left ventricular filling can be caused by pulmonary hypertension and peripheral vasodilatation.

Shin et al.performed elaborate clinical research on the effect of ethanol on the hemodynamics of the pulmonary artery during ethanol embolotherapy.They demonstrated a consistent elevation of PAP with a single ethanol dose of 0.05 mL/kg[20]and a high incidence of acute pulmonary hypertension in each sclerotherapy session without a lasting effect on PAP[16].They also found that injections limited to 0.14 mL/kg every 10 minutes obviated the chance of CVC occurrence[19].Furthermore,Ko et al.demonstrated that PAP reflects the pulmonary arterial ethanol level and is positively related to the dose of ethanol[14].Pulmonary hypertension caused by ethanolinduced precapillary pulmonary artery vascular spasm led to elevated right ventricular afterload,decreased RV contractility,decreased RV cardiac output,and decreased left ventricular filling.However,some authors have doubts regarding PAP reaction to the injected ethanol.Mitchell et al.[12]found that the minimal increase in PAP during these procedures did not elucidate the cause of the rare complication of cardiovascular collapse during ethanol embolization for vascular malformation.

Ethanol has been widely recommended as a peripheral vasodilator agent since 1967[21].Many studies have confirmed the acute effects of ethanol on blood pressure as a vasodilator[22-27].Several studies have found a transient increase in blood pressure approximately 10 min after alcohol ingestion,followed by a decrease in blood pressure that lasted for several hours[28-29].The possible mechanism involves the relationship between ethanol and cellular electrolyte metabolism.Acetaldehyde acetate,a metabolite of ethanol,can also induce peripheral vasodilatation,especially in patients with congenital acetaldehyde hydrogenase deficiency[30].

Ethanol can also induce an acute decrease in left ventricular contractility.In a population of 64 young healthy subjects,Cameli et al.[31]found that low blood concentrations of ethanol can acutely impair left ventricular and right ventricular function.Similar results were observed in 10 normal subjects in a study by Delgado,along with the finding that myocardial contractility decreases following the ingestion of alcohol[32].Acute ethanol exposure can directly impair myocardial contractility through the following cellular mechanisms:by directly inhibiting the calcium myofilament interaction,by inhibiting the intracellular contractile apparatus,by diminishing the Ca2+transients,and by direct toxicity on cardiac myocytes caused by ethanol and its metabolites[33-35].These changes lead to decreased left ventricular stroke volume,systemic hypotension,and impaired coronary blood flow,which further impair left and right ventricular contractility[21].

In previous hemodynamic studies of ethanol embolotherapy for vascular malformation,hypotension was rarely reported and was not regarded as a complication of ethanol embolotherapy.We believe that hypotension may be a potential indicator of CVC during the procedure,especially when it is persistent and/or combined with bradycardia,oxygen desaturation,and tachypnea.Adrenaline and its analogs are very effective in reversing hypotension by increasing coronary perfusion pressure,improving coronary blood flow,and reversing ventricular free wall ischemia[36].We believe that hypotension should be regarded as an acute complication of ethanol embolotherapy for AVMs.When hypotension is persistent and presents with bradycardia,oxygen desaturation,and endtidal CO2decrease,ethanol injection should be stopped immediately,and measures should be taken in time to prevent CVC.

Hemodynamic Change and Ethanol Dose

In our study,increases in SBP and HR were positively correlated with ethanol dose.This is in line with the results of Mitchell et al.,in which systemic SBP was strongly associated with the amount of ethanol in ethanol embolotherapy for vascular malformations[12].The systemic absorption of ethanol in embolotherapy is well known,and the relationship between the injected ethanol dose and serum ethanol level is well documented[37-38].Although a safe ethanol dose has not been confirmed,the accepted practice over the past several years has been to keep the total volume of ethanol at less than 1 mL/ kg,based on the experience of Yakes[1].Mason et al.[39]demonstrated that when keeping the ethanol dose less than 1 mL/kg,the serum ethanol level after the procedure will be below the intoxication level in most patients,with a few cases at a slightly intoxicated level and none at a toxic level.However,this does not protect the patient from potential cardiovascular collapse.Most of the reported cases of CVC occurred when using a“safe”dose of ethanol (Table 1).In Mason’s study,the serum ethanol level was measured 30 min after the last injection of ethanol,which may result in underestimating the peak serum ethanol level and the effect of bolus ethanol injected at one time.Moreover,Yata et al.[39]found that ethanolinduced sludge can cause transient pulmonary hypertension or cardiopulmonary collapse without ethanol in pigs.In our study,the ethanol dose never exceeded 0.4 mL/kg in all procedures of ethanol embolotherapy for vascular malformations.In our single-center experience,a low ethanol dose given in one session was associated with a low incidence of CVC.

Limitations

One of the limitations of this study was that we were unable to measure the serum ethanol level and PAP at the point of maximum SBP.Therefore,we could not analyze the correlation between hypotension and serum ethanol levels and were unable to determine whether hypotension was caused by pulmonary hypertension.Additionally,we could not confirm whether persistent hypotension would lead to cardiovascular collapse or could be reverted without using vasopressors,which may result in serious sequelae when persistent hypotension occurs and no measures are taken to avoid it.We based our hypothesis of CVC occurrence on previously reported cases.

CONCLUSIONS

Increases in SBP and HR during ethanol embolotherapy are common,temporary,and most likely pain-mediated;the increase in these parameters was positively correlated with ethanol dose.Hypotension may be regarded as an acute complication of ethanol embolotherapy procedures.Hypotension combined with bradycardia,oxygen desaturation,and end-tidal CO2decrease may be potential predictors of cardiovascular collapse.

ETHICS DECLARATIONS

Ethics Approval and Consent to Participate

This study received ethical approval from the Ethics Committee of the Shanghai Ninth People’s Hospital,Shanghai Jiao Tong University School of Medicine.All participants provided written informed consent before study enrollment.

Consent for Publication

All of the authors have consented to the publication of this article.

Competing Interests

The authors declare no conflicts of interest.The authors state that the views expressed in the article are their own and not the official position of the institution or funder.