Advertisement

Clinical significance of cerebrovascular gas emboli during polidocanol endovenous ultra-low nitrogen microfoam ablation and correlation with magnetic resonance imaging in patients with right-to-left shunt

Open ArchivePublished:September 23, 2010DOI:https://doi.org/10.1016/j.jvs.2010.06.179

      Background

      Foam generated by manual agitation of liquid sclerosant with air or gas is routinely utilized to treat refluxing veins. Although generally well tolerated, serious neurological events have been reported. The composition and properties of the foam, including bubble size and gaseous components, may contribute to the potential for microcirculatory obstruction and cerebral ischemia. We tested an ultra-low nitrogen polidocanol endovenous microfoam with controlled bubble size and density and hypothesized that patients at risk due to the presence of middle cerebral artery (MCA) bubble emboli during microfoam injection would not demonstrate evidence of clinical or subclinical cerebral infarction.

      Methods

      Patients with great saphenous vein incompetence were treated with ultra-low nitrogen (≤0.8%) polidocanol endovenous microfoam injected under ultrasound guidance. Patients with right-to-left shunt were included to evaluate the safety of cerebral arterial bubbles. All patients with MCA emboli detected by transcranial Doppler during endovenous microfoam ablation received intensive surveillance for microinfarction, including brain magnetic resonance imaging and measurement of cardiac troponin-I.

      Results

      MCA bubble emboli were detected in 60 of 82 treated patients; 22 patients had no detectable emboli. Among patients with MCA bubbles detected, 49 (82%) had ≤15 bubbles. No patients developed magnetic resonance imaging abnormalities, neurological signs, or elevated cardiac troponin.

      Conclusions

      Patients treated with foamed liquid sclerosants are commonly exposed to cerebrovascular gas bubbles. In this series of 60 high-risk patients with MCA bubble emboli during or after treatment with ultra-low nitrogen polidocanol endovenous microfoam, there was no evidence of cerebral or cardiac microinfarction. The results of this study cannot be generalized to foams compounded using bedside methodologies, since the composition of these foams is substantially different.
      Saphenous vein incompetence affects 20% to 25% of the adult population.
      • Criqui M.H.
      • Jamosmos M.
      • Fronek A.
      • Denenberg J.O.
      • Langer R.D.
      • Bergan J.
      • et al.
      Chronic venous disease in an ethnically diverse population The San Diego Population Study.
      Nearly half of these patients report symptoms impacting quality of life, and approximately 1% of the adult population develops chronic ulceration due to venous disease, resulting in significant medical costs and loss of productivity.
      • Phillips T.
      • Stanton B.
      • Provan A.
      • Lew R.
      A study of the impact of leg ulcers on quality of life: financial, social and psychologic implications.
      Intravenous sclerosing solutions such as polidocanol and sodium tetradecyl sulfate have been utilized for decades to ablate veins by producing endothelial injury, resulting in venous transformation to a fibrous cord. Saphenous incompetence is commonly treated with endovenous heat ablation (radiofrequency or laser), but venous sclerosants are less invasive. Recent studies have demonstrated that foamed sclerosants
      • Cabrera J.
      • Cabrera J.
      • Garcia-Olmedo M.A.
      Treatment of varicose greater saphenous vein with sclerosant in Microfoam form: long term outcomes.
      are twice as efficacious as liquid sclerosants in the treatment of larger veins including the saphenous trunk veins.
      • Hamel-Desnos C.
      • Desnos P.
      • Wollmann J.C.
      • Ouvry P.
      • Mako S.
      • Allaert F.A.
      Evaluation of the efficacy of polidocanol in the form of foam compared with liquid form in sclerotherapy of the greater saphenous vein: initial results.
      • Rabe E.
      • Otto J.
      • Schliephake D.
      • Pannier F.
      Efficacy and safety of great saphenous vein sclerotherapy using standardised polidocanol foam: a randomised controlled multicentre clinical trial.
      Foamed sclerosants are also effective in treating complex venous malformations
      • Cabrera J.
      • Cabrera J.
      • Garcia-Olmedo M.A.
      • Redondo P.
      Treatment of venous malformations with sclerosant in microfoam form.
      and incompetent veins associated with venous ulcers.
      • Cabrera J.
      • Redondo P.
      • Becerra A.
      • Garrido C.
      • Cabrera J.
      • Garcia-Olmedo M.A.
      • et al.
      Ultrasound-guided injection of polidocanol microfoam in the management of venous leg ulcers.
      Foamed sclerosants are generally created at the bedside by agitating air (78% nitrogen) with liquid sclerosant.
      • Tessari L.
      • Cavezzi A.
      • Frullini A.
      Preliminary experience with a new sclerosing foam in the treatment of varicose veins.
      Variations on the technique include air-to-sclerosant ratios from 2:1 to 8:1, use of different sclerosants in concentrations from 0.5% to 3%, and the use of commercially available devices to assist the process. Consensus conferences in 2003 and 2006 produced guidelines on the creation and clinical use of foamed sclerosants.
      • Breu F.X.
      • Guggenbichler S.
      European Consensus Meeting on Foam Sclerotherapy, April 4-6, 2003, Tegernsee, Germany.
      • Breu F.X.
      • Guggenbichler S.
      • Wollmann J.C.
      Duplex ultrasound and efficacy criteria in foam sclerotherapy from the 2nd European Consensus Meeting on Foam Sclerotherapy 2006, Tegernsee, Germany.
      These diverse foamed sclerosants are widely used and have been relatively safe in a series of 12,713 patients, half of whom were treated with foamed sclerosants.
      • Guex J.J.
      • Allaert F.A.
      • Gillet J.L.
      • Chleir F.
      Immediate and midterm complications of sclerotherapy: report of a prospective multicenter registry of 12,173 sclerotherapy sessions.
      However, there are well-documented reports of significant neurological events including major stroke, seizures, and transient ischemic attacks in patients subsequently demonstrated to have patent foramen ovale.
      • Bush R.G.
      • Derrick M.
      • Manjoney D.
      Major neurological events following foam sclerotherapy.
      • Forlee M.V.
      • Grouden M.
      • Moore D.J.
      • Shanik G.
      Stroke after varicose vein foam injection sclerotherapy.
      Gas was visualized in the vertebral and cerebral arterial systems in two of these cases following injection of modest volumes (4 and 10 mL) of foamed sclerosant. In two recent reports, all patients injected with foamed sclerosants had gas bubbles visualized in the right heart chambers using echocardiography, and some had bubbles in the left heart.
      • Ceulen R.P.M.
      • Sommer A.
      • Vernooy K.
      Microembolism during foam sclerotherapy of varicose veins.
      • Rush J.E.
      • Wright D.D.I.
      More on microembolism and foam sclerotherapy.
      As patent foramen ovale is present in 25% to 30% of adults,
      • Hagen P.T.
      • Scholtz D.G.
      • Edwards W.D.
      Incidence and size of patent foramen ovale during the first 10 decades of life: an autopsy study of 965 normal hearts.
      a significant percentage of patients treated with foamed sclerosants are exposed to gas bubbles in the systemic arterial circulation.
      The composition and properties of gas bubbles, including their size and durability, contribute to their propensity to obstruct the microvasculature
      • Eckmann D.M.
      • Kobayashi S.
      • Min L.
      Microvascular embolization following polidocanol microfoam sclerosant administration.
      and may influence the tendency of certain foams to cause major neurological events. We tested the hypothesis that an investigational ultra-low nitrogen (≤0.8%) polidocanol endovenous microfoam designed to maintain specific microfoam characteristics including bubble size, microfoam density, and rapid gas absorption into blood
      • Wright D.
      • Gobin J.P.
      • Bradbury A.W.
      • Coleridge-Smith P.
      • Spoelstra H.
      • Berridge D.
      • et al.
      Varisolve polidocanol microfoam compared with surgery or sclerotherapy in the kmanagement of varicose veins in the presence of trunk vein incompetence: European randomized controlled trial.
      would not cause clinical or subclinical ischemia or infarction.

      Methods

       Patients

      Patients between 18 and 60 years of age with saphenofemoral junction incompetence and GSV reflux (≥1.0 second) were eligible if they also had clinically important manifestations of venous disease, including edema, skin changes, or healed venous ulcers (CEAP 3, 4, or 5). Excluded were patients with prior evidence of atherosclerotic vascular disease, uncontrolled hypertension, diabetes mellitus requiring insulin, hypercholesterolemia, venous thromboembolism, body mass index >30, other major medical disorders, small saphenous or deep venous incompetence (other than common femoral vein), clinically important echocardiographic abnormalities, medication with oral contraceptives or anticoagulants, abnormal magnetic resonance imaging (MRI) of the brain, and current smokers. Initially each of five participating clinical practices enrolled three to five patients without right-to-left (R-L) shunt to ensure familiarity with study procedures and to receive training from a physician-proctor on the technical aspects of endovenous microfoam ablation; subsequent patients were required to be positive for R-L shunt.
      The protocol was approved by the Institutional Review Boards of participating clinical practices, and all patients provided written informed consent.

       Study design

      The study was conducted under an investigational new drug (IND) application, and the study sponsor developed the protocol in consultation with representatives of the Food and Drug Administration. Because this study was conducted under an IND, it was not possible to include a nonapproved manually generated foam as control.
      Pretreatment evaluation for the presence of R-L shunt was accomplished using transcranial Doppler (TCD) with injection of contrast (8 mL saline agitated with 1 mL air and 1 mL blood).
      • Sastry S.
      • Daly K.
      • Chengodu T.
      • McCollum C.
      Is transcranial Doppler for the detection of venous-to-arterial circulation shunts reproducible?.
      One contrast injection was performed at rest and another while the patient performed the Valsalva maneuver. Unilateral middle cerebral artery (MCA) bubble counts were determined for 15 cardiac cycles to identify patients with shunts that were likely to be intracardiac. Patients testing Spencer grade 2 or greater (≥6 MCA bubbles) at rest or following Valsalva were included.
      • Smith K.H.
      • Spencer M.P.
      Doppler indices of decompression sickness: their evaluation and use.
      TCD surveillance was continued throughout the endovenous microfoam ablation procedure, including during application of compression bandaging and for an additional 10 minutes postprocedure until patients were standing. After each of two periods of postprocedure ambulation, TCD monitoring was resumed such that each patient had a minimum of 30 minutes postprocedure monitoring. All treated patients were hospitalized for 24 hours following the procedure and were monitored with continuous oximetry, electrocardiograms 1 and 24 hours after the procedure, and cardiac isoenzymes and troponin-I were measured every 8 hours.
      Patients with periprocedural MCA emboli underwent additional intensive surveillance including examination by a neurologist, direct ophthalmoscopy and visual field testing within 6 hours, and brain MRI within 24 hours. This testing was repeated at 7 and 28 days, although the 7-day MRI was discontinued from the protocol when it was determined that this time point was capturing no additional information. Any patient reporting visual or neurological symptoms within 24 hours of treatment entered the intensive neurological monitoring schedule regardless of whether MCA emboli had been detected.
      The initial determination of whether MCA bubble emboli were present was made by personnel at the clinical sites. The TCD recordings were subsequently reviewed by an independent expert who counted all MCA emboli. Therefore, it was possible for patients to be initially categorized as negative by the clinical sites but subsequently deemed positive by the TCD expert reviewer. Such patients missed the 24-hour MRI and other early neurological monitoring; in these cases, the clinical sites initiated intensive surveillance procedures at the earliest opportunity.
      Brain MRI was performed using a 1.5 T whole-body system with a dedicated head or neurovascular coil, using a sequence designed to maximize sensitivity to detect small embolic lesions. Baseline MR sequences included sagittal T1 localizer, axial diffusion-weighted image (b-value = 1000) with apparent diffusion coefficient map, axial T2, axial T2*, and axial T2 fluid attenuated inversion recovery.
      Patient recruitment continued until the accumulation of 50 patients with detectable MCA bubbles during or following endovenous microfoam ablation. Clinical follow-up for all patients after 7 and 28 days included duplex ultrasonography of the treated limb to evaluate efficacy and perform a detailed search for deep venous thrombosis.

       Study organization

      Patients were recruited by media advertising. Those who successfully completed a study eligibility questionnaire were referred to participating clinics for informed consent and further screening. Study data were maintained by a contract research organization that also performed the statistical analysis. An independent reviewer adjudicated the MRI images as blinded pairs; each posttreatment sequence matched with the corresponding pretreatment sequence but was blinded to order and date. A TCD expert without knowledge of each patient's R-L shunt status reviewed recordings made during endovenous microfoam ablation and made the final determination regarding presence and number of embolic signals.
      An independent Data and Safety Monitoring Board reviewed progress every 6 months or more frequently and was required to meet in an emergency session within 4 days should any new MRI lesion be discovered by the independent MRI reviewer.

       Treatment

      The investigational microfoam (Varisolve polidocanol endovenous microfoam, BTG International Ltd, W. Conshohocken, Pa) is formulated with a gas mixture designed to maintain specific microfoam characteristics while facilitating rapid gas absorption. A sterile container system dispenses microfoam of reproducibly consistent density, bubble size, and stability. The manufacturing process ensures that gas used to generate the microfoam has an extremely low nitrogen content (≤0.8%).
      After the great saphenous vein (GSV), tributaries, and perforators had been mapped, the GSV was cannulated just above the knee under ultrasound guidance. TCD monitoring began 2 minutes prior to injection of microfoam.
      Microfoam was injected under ultrasound guidance to fill the proximal GSV from the point of cannulation to a point 5 cm distal to the saphenofemoral junction. Additional microfoam was injected to fill the distal GSV and varicose tributaries. The protocol allowed up to 20 mL of microfoam to be delivered in one treatment session (one patient received 24 mL).
      The treated leg was wrapped in a short-stretch bandage with compression pads over the treated venous segments. A thigh-length 30 to 40 mm Hg compression stocking was placed over the dressing. This compression system was worn for 48 hours before removal. The stocking was worn continuously for 2 weeks.

       Statistical analysis

      The study was designed to determine the frequency of new subclinical MRI abnormalities suggestive of emboli or infarction in patients with MCA bubbles detected after microfoam injection. For an observed incidence of zero lesions in 50 patients, the upper 97.5% one-sided confidence limit for the true incidence rate would be 7.11%, calculated using exact binomial probabilities.

      Results

      Between March 2007 and June 2008, 82 patients had endovenous microfoam ablation of the GSV (Fig 1 and Table I), 61 of whom had demonstrated R-L shunt.
      Figure thumbnail gr1
      Fig 1Disposition of patients screened for study enrollment.
      Table ICharacteristics of patients treated with endovenous microfoam ablation (n = 82)
      Age (years)
       Mean ± SD44.5 ± 8.66
       Median45.0
       Range20-59
      Gender, n (%)
       Male16 (19.5)
       Female66 (80.5)
      Race/ethnicity, n (%)
       White78 (95.1)
       Asian2 (2.4)
       Hispanic or Latino2 (2.4)
      CEAP clinical category, n (%)
       CEAP 3 (edema)72 (87.8)
       CEAP 4 (skin changes without ulceration)6 (7.3)
       CEAP 5 (skin changes with healed ulceration)4 (4.9)
      Right-left shunt Spencer grade, n (%)At rest/with Valsalva
       Grade 036 (43.9)/17 (20.7)
       Grade 118 (22.0)/6 (7.3)
       Grade 29 (11.0)/9 (11.0)
       Grade 38 (9.8)/12 (14.6)
       Grade 46 (7.3)/17 (20.7)
       Grade 55 (6.1)/21 (25.6)
      The median volume of injected microfoam was 18 mL (range, 6-24 mL), and following GSV cannulation, the median duration of the procedure was 41 minutes (range, 14-114 minutes). Most MCA emboli were detected within 15 minutes of the microfoam injections (Fig 2). Of 82 treated patients, 60 had one or more MCA bubble emboli detected (Table II). Forty-nine patients (59.8%) had 15 or fewer bubbles detected, and four (4.9%) had more than 50. The maximum number of bubbles counted in any patient was 382 (this latter patient also had a “curtain” of additional emboli, which could not be individually counted). The number of MCA bubbles detected did not correlate with pretreatment Spencer grade (Spearman non-parametric correlation coefficient = 0.3962).
      Figure thumbnail gr2
      Fig 2Transcranial Doppler (TCD) bubbles per minute following initial injection of polidocanol endovenous microfoam (all treated patients combined).
      Table IIMCA bubble emboli during or after endovenous microfoam ablation (expert TCD reviewer)
      Patients with endovenous microfoam ablation82
      Patients with MCA bubble emboli during procedure60 (73%)
      Number of MCA bubble emboli in patients with detectable bubbles (n=60)
       Mean ± SD21.8 ± 62.14
       Median5.0
       Range1-382
      Bubble count category during procedure (categories corresponding to Spencer grading
      Smith KH, Spencer MP. Doppler indices of decompression sickness: Their evaluation and use. Aerosp Med 1970;41:1396-400.
      )
      Number of patients
       1-5 bubbles34
       6-15 bubbles14
       16-50 bubbles8
       51-150 bubbles1
       >150 bubbles3
      MCA, Middle cerebral artery; TCD, transcranial Doppler.
      low asterisk Smith KH, Spencer MP. Doppler indices of decompression sickness: Their evaluation and use. Aerosp Med 1970;41:1396-400.
      MCA bubbles were detected during the microfoam procedure in 54 (89%) of the 61 shunt-positive patients, but 6 of 21 (29%) shunt-negative patients also had MCA bubbles detected. The clinical site and the expert TCD reviewer agreed on the TCD interpretation in 77 of 82 cases (94%).
      All 60 patients with MCA bubble emboli had pretreatment and at least one posttreatment MRI scan. The blinded independent assessor identified 10 patients with minor pretreatment MRI abnormalities and all remained unchanged in subsequent examinations. No lesions were evident in diffusion-weighted MRI sequences at any time point in any patient. Three patients had no 24-hour MRI examinations because MCA bubbles were detected only by the expert TCD reviewer, but all had normal 28-day MRI examinations. Two patients with normal 24-hour MRI examinations declined the 28-day MRI follow-up.
      There were no new neurological symptoms or findings following treatment in any patient, whether or not MCA emboli were detected. All visual field tests were normal or unchanged, and no emboli were detected on fundoscopy. One patient, for whom the expert TCD reviewer counted three MCA emboli, described “twinkling lights” in her peripheral vision appearing approximately 1 hour posttreatment and lasting 20 seconds. Fundoscopy and visual field testing conducted within minutes of the symptom were normal, and the symptom did not recur. Because of the brevity of the visual symptom and the lack of objective signs, the investigator and consulting ophthalmologist did not consider it clinically concerning. Markers of cardiac ischemia were normal for all patients at all time points. Other chemistry and hematology assessments remained within normal limits.
      At the day 28 follow-up visit, duplex ultrasound confirmed complete occlusion of the GSV in 71 of 81 patients (88%) and elimination of saphenous reflux in 73 of 81 (90%).
      Adverse events included pain or discomfort in the treated limb (37 patients), expression of intravascular coagulum (11 patients), and ecchymosis (five patients). One patient who could not be cannulated had extravascular injection of 11 mL of microfoam. This patient experienced no pain or adverse effect. One patient had a symptomatic occlusive thrombus of a short section of popliteal vein, associated with a tense knee effusion, which may have contributed to the thrombosis. She was treated with 6 months of anticoagulation. Five other patients had asymptomatic ultrasound evidence of thrombus involving the deep veins (three with calf vein thrombi and two with nonocclusive thrombus at or near the saphenofemoral junction involving the common femoral vein wall). Sclerosis of the GSV initiates intraluminal thrombosis of the treated superficial vein, and in four additional patients, there was convexity of the GSV luminal thrombus into the lumen of the common femoral vein. In the absence of data that these asymptomatic patients required anticoagulation, most were conservatively managed with serial ultrasound observation (five patients) or short periods of anticoagulation (6-21 days; three patients); one patient was anticoagulated for 90 days. All thrombi resolved, and none of these patients had clinical sequelae or symptoms suggestive of pulmonary embolism.

      Discussion

      Despite widespread use of foamed sclerosants, the neurological safety of physician-compounded foam sclerosant has not been systematically studied. In an intravital microscopic study in the rat, polidocanol foam compounded with air following the Tessari technique
      • Tessari L.
      • Cavezzi A.
      • Frullini A.
      Preliminary experience with a new sclerosing foam in the treatment of varicose veins.
      caused arteriolar obstruction, whereas the investigational microfoam used in the current clinical study did not obstruct the microvasculature at much higher doses.
      • Eckmann D.M.
      • Kobayashi S.
      • Min L.
      Microvascular embolization following polidocanol microfoam sclerosant administration.
      Characteristics of the gas bubbles appear to influence the propensity of foams to cause microvascular occlusion and thus ischemic events such as stroke, as well as more minor events such as transient visual disturbances and chest tightness.
      • Morrison N.
      • Neuhardt D.L.
      • Rogers C.R.
      • McEown J.
      • Morrison T.
      • Johnson E.
      • et al.
      Comparisons of side effects using air and carbon dioxide foam for endovenous chemical ablation.
      The current clinical study of the investigational microfoam was designed to investigate cerebrovascular safety using diffusion-weighted MRI, the most sensitive tool available for the detection of silent cerebral ischemia.
      • Bendszus M.
      • Koltzenburg M.
      • Burger R.
      • Warmuth-Metz M.
      • Hofmann E.
      • Solymosi L.
      Silent embolism in diagnostic cerebral angiography and neurointerventional procedures: a prospective study.
      Diffusion-weighted MRI is highly sensitive to edema formation as an early marker of injury (potentially reversible) while T2 and T2* detect permanent scarring, which, if present, may not become apparent until at least 10 days after the injury. Thus, with the MRI protocol used in this study, minor injury could have been detected, and acute and chronic lesions differentiated. Patients enrolled were specifically selected to have a potential risk of microvascular injury (screened for presence of right-to-left shunt). Each of the 60 patients with detectable MCA emboli during microfoam injection had multiple opportunities to experience an “event” since the bubbles were distributed systemically, and all vascular territories of the brain and other organs were exposed to bubble emboli. The vascular territories most sensitive to ischemic damage were studied intensively, including examination of the entire brain using diffusion-weighted MRI, ophthalmoscopy and visual field examination to detect ischemia of the retina or optic nerve, and troponin-I, which is sensitive to silent myocardial infarction. These methodologies were selected to detect subclinical injury that might not be apparent in a larger series of patients less intensively studied.
      The rate of deep vein thrombosis (one symptomatic deep vein thrombosis, five asymptomatic thrombi) was higher in this trial than in previous trials of endovenous microfoam ablation (11 of 437 [2.5%] in a previous multicenter study).
      • Wright D.
      • Gobin J.P.
      • Bradbury A.W.
      • Coleridge-Smith P.
      • Spoelstra H.
      • Berridge D.
      • et al.
      Varisolve polidocanol microfoam compared with surgery or sclerotherapy in the kmanagement of varicose veins in the presence of trunk vein incompetence: European randomized controlled trial.
      In the current study, deep vein thrombosis may have been increased due to postprocedure immobilization for TCD monitoring and the required 24-hour hospitalization. It is important to emphasize that all treated patients in this study underwent careful posttreatment ultrasound examinations using high-resolution equipment, resulting in detection of small asymptomatic thrombi of uncertain clinical significance. Regardless of treatment modality, detection of deep vein thrombosis is clearly related to whether and how exhaustively thrombi are prospectively sought following the intervention.
      • Keith L.M.
      • Smead W.L.
      Saphenous vein stripping and its complications.
      • Van Rij A.M.
      • Chai J.
      • Hill G.B.
      • Christie R.A.
      Incidence of deep vein thrombosis after varicose vein surgery.
      • Hingorani A.P.
      • Ascher E.
      • Markevich N.
      • Schutzer R.W.
      • Kallakuri S.
      • Hou A.
      • et al.
      Deep venous thrombosis after radiofrequency ablation of greater saphenous vein: a word of caution.
      • Mozes G.
      • Kalra M.
      • Carmo M.
      • Swenson L.
      • Gloviczki P.
      Extension of saphenous thrombus into the femoral vein: a potential complication of new endovenous ablation techniques.
      Manufacturer and User Facility Device Experience (MAUDE) Database
      Because of concern regarding neurological events, consensus guidelines suggest limiting the injected volume of foamed sclerosants to 10 mL or less.
      • Breu F.X.
      • Guggenbichler S.
      • Wollmann J.C.
      Duplex ultrasound and efficacy criteria in foam sclerotherapy from the 2nd European Consensus Meeting on Foam Sclerotherapy 2006, Tegernsee, Germany.
      Despite the injection in this study of relatively large volumes of investigational microfoam (up to 24 mL), in most cases, only very low numbers of MCA emboli were detected. Each MCA carries approximately 3% of cardiac output; therefore, the bubbles detected represent only a fraction of those injected. For example, 30 bubbles counted in one MCA would equate to 1000 bubbles reaching the systemic arterial circulation, which is 0.05% of the bubbles contained in a 20-mL dose of proprietary microfoam (each mL contains approximately 1 million bubbles). Thus, it is clear that most injected bubbles never reach the systemic arterial circulation and are either extinguished in the venous circulation (by absorption of gas into the blood) or filtered by the lung capillaries.
      Most MCA bubbles were detected soon after injection of microfoam; thus, if circulating bubbles posed a risk, this should have been apparent within the first hour postprocedure. Since MCA bubbles were detected in nearly one-third of patients testing negative for R-L shunt, it is evident that in clinical practice, prescreening patients for R-L shunt prior to injection of foamed sclerosants would not be worthwhile. Because periprocedural TCD monitoring continued for up to 60 minutes following microfoam injection, MCA bubbles detected might have represented either intra- or extracardiac shunting.
      Since arterial bubble emboli cannot be avoided during injection of foamed sclerosants, it is important that the injected drug have characteristics consistent with clinical safety. In this series of 60 intensively-studied high-risk patients with cerebral arterial gas bubble embolization following saphenous ablation with a proprietary ultra-low nitrogen endovenous microfoam, there were no clinically important symptoms and no detectable cerebral or cardiac injury. A larger series of unselected patients treated with the proprietary ultra-low-nitrogen microfoam and followed for clinical events will provide additional data to complement the results of this series. The results of this study cannot be extrapolated to other forms of foamed sclerosants since foams compounded using bedside methodologies differ in bubble size and gas composition, properties that may contribute to the potential for microcirculatory obstruction and cerebral ischemia.

      Author contributions

      • Conception and design: JER, DW
      • Analysis and interpretation: JDR, KG, CS, SH, JER, DW
      • Data collection: JDR, KG, CS, SH, JER, DW
      • Writing the article: JER, DW
      • Critical revision of the article: JDR, KG, CS, DW, SH
      • Final approval of the article: JDR, KG, CS, DW, SH, JER
      • Statistical analysis: JER, DW
      • Obtained funding: DW, JER
      • Overall responsibility: JDR

      Appendix

      The following individuals participated in the conduct of the trial:

       Clinical Sites

      Wake Forest University School of Medicine (Winston-Salem NC): J. Regan, M. Bettmann, B. Kouri, L. Patella, P. Tesch; Lake Washington Vascular and Overlake Hospital (Bellevue, Wash): K. Gibson, B. Ferris, D. Pepper, T. Fortney, A. Ebert, C. Leafdale, H. Covert; University of Southern California Medical Center (Los Angeles, Calif): V. Rowe, F. Weaver, D. Hood, C. Pappas, S. Parese; Duke University School of Medicine (Durham, NC): C. Shortell, T. Williams, S. Finley; University of Pittsburgh-Shadyside (Pittsburgh, Pa): S. Hirsch, E. Dillavou, J. Brimmeier, T. Richardson.

       Data and Safety Monitoring Board

      Carlos Kase, MD (chair) and Ravin Davidoff, MD (Boston Medical Center), Howard Rowley, MD (University of Wisconsin-Madison), Mark Espeland, PhD (Wake Forest University School of Medicine).

       Study Proctors

      Mark Isaacs, MD, John Mauriello, MD, David Wright, MB, FRCS.

       Transcranial Doppler Core Laboratory

      (Sentient Neurocare Services, Inc.): Alex Razumovsky, PhD (independent reviewer), David Pilchard.

       MRI Independent Review

      Lyle R. Gesner, MD.

       Perceptive Informatics

      James Paskevitz, MD, Matthew Hayden.

       Parexel Clinical Services

      E. Carter, E. Leip, S. Zandman, T. Zuttermeister.

       BTG International

      D. Wright, J. Rush, G. Suplick, J. Barclay, K. Arcuri, E. Evans, C. Tedesco, P. Mussenden.

      References

        • Criqui M.H.
        • Jamosmos M.
        • Fronek A.
        • Denenberg J.O.
        • Langer R.D.
        • Bergan J.
        • et al.
        Chronic venous disease in an ethnically diverse population.
        Amer J Epidemiol. 2003; 158: 448-456
        • Phillips T.
        • Stanton B.
        • Provan A.
        • Lew R.
        A study of the impact of leg ulcers on quality of life: financial, social and psychologic implications.
        J Am Acad Dermatol. 1994; 31: 49-53
        • Cabrera J.
        • Cabrera J.
        • Garcia-Olmedo M.A.
        Treatment of varicose greater saphenous vein with sclerosant in Microfoam form: long term outcomes.
        Phlebology. 2000; 15: 19-23
        • Hamel-Desnos C.
        • Desnos P.
        • Wollmann J.C.
        • Ouvry P.
        • Mako S.
        • Allaert F.A.
        Evaluation of the efficacy of polidocanol in the form of foam compared with liquid form in sclerotherapy of the greater saphenous vein: initial results.
        Dermatol Surg. 2003; 29: 1170-1175
        • Rabe E.
        • Otto J.
        • Schliephake D.
        • Pannier F.
        Efficacy and safety of great saphenous vein sclerotherapy using standardised polidocanol foam: a randomised controlled multicentre clinical trial.
        Eur J Vasc Endovasc Surg. 2008; 35: 238-245
        • Cabrera J.
        • Cabrera J.
        • Garcia-Olmedo M.A.
        • Redondo P.
        Treatment of venous malformations with sclerosant in microfoam form.
        Arch Dermatol. 2003; 139: 1409-1416
        • Cabrera J.
        • Redondo P.
        • Becerra A.
        • Garrido C.
        • Cabrera J.
        • Garcia-Olmedo M.A.
        • et al.
        Ultrasound-guided injection of polidocanol microfoam in the management of venous leg ulcers.
        Arch Dermatol. 2004; 140: 667-673
        • Tessari L.
        • Cavezzi A.
        • Frullini A.
        Preliminary experience with a new sclerosing foam in the treatment of varicose veins.
        Dermatol Surg. 2001; 27: 58-60
        • Breu F.X.
        • Guggenbichler S.
        European Consensus Meeting on Foam Sclerotherapy, April 4-6, 2003, Tegernsee, Germany.
        Dermatol Surg. 2004; 30 (discussion 717): 709-717
        • Breu F.X.
        • Guggenbichler S.
        • Wollmann J.C.
        Duplex ultrasound and efficacy criteria in foam sclerotherapy from the 2nd European Consensus Meeting on Foam Sclerotherapy 2006, Tegernsee, Germany.
        Vasa. 2008; 37: 90-95
        • Guex J.J.
        • Allaert F.A.
        • Gillet J.L.
        • Chleir F.
        Immediate and midterm complications of sclerotherapy: report of a prospective multicenter registry of 12,173 sclerotherapy sessions.
        Dermatol Surg. 2005; 31: 123-128
        • Bush R.G.
        • Derrick M.
        • Manjoney D.
        Major neurological events following foam sclerotherapy.
        Phlebology. 2008; 23: 189-192
        • Forlee M.V.
        • Grouden M.
        • Moore D.J.
        • Shanik G.
        Stroke after varicose vein foam injection sclerotherapy.
        J Vasc Surg. 2006; 43: 162-164
        • Ceulen R.P.M.
        • Sommer A.
        • Vernooy K.
        Microembolism during foam sclerotherapy of varicose veins.
        N Eng J Med. 2008; 358: 1525-1526
        • Rush J.E.
        • Wright D.D.I.
        More on microembolism and foam sclerotherapy.
        N Eng J Med. 2008; 359: 656-657
        • Hagen P.T.
        • Scholtz D.G.
        • Edwards W.D.
        Incidence and size of patent foramen ovale during the first 10 decades of life: an autopsy study of 965 normal hearts.
        Mayo Clinic Proceedings. 1984; 59: 17-20
        • Eckmann D.M.
        • Kobayashi S.
        • Min L.
        Microvascular embolization following polidocanol microfoam sclerosant administration.
        Dermatol Surg. 2005; 31: 636-643
        • Wright D.
        • Gobin J.P.
        • Bradbury A.W.
        • Coleridge-Smith P.
        • Spoelstra H.
        • Berridge D.
        • et al.
        Varisolve polidocanol microfoam compared with surgery or sclerotherapy in the kmanagement of varicose veins in the presence of trunk vein incompetence: European randomized controlled trial.
        Phlebology. 2006; 21: 180-190
        • Sastry S.
        • Daly K.
        • Chengodu T.
        • McCollum C.
        Is transcranial Doppler for the detection of venous-to-arterial circulation shunts reproducible?.
        Cerebrovasc Dis. 2007; 23: 424-429
        • Smith K.H.
        • Spencer M.P.
        Doppler indices of decompression sickness: their evaluation and use.
        Aerosp Med. 1970; 41: 1396-1400
        • Morrison N.
        • Neuhardt D.L.
        • Rogers C.R.
        • McEown J.
        • Morrison T.
        • Johnson E.
        • et al.
        Comparisons of side effects using air and carbon dioxide foam for endovenous chemical ablation.
        J Vasc Surg. 2008; 47: 830-836
        • Bendszus M.
        • Koltzenburg M.
        • Burger R.
        • Warmuth-Metz M.
        • Hofmann E.
        • Solymosi L.
        Silent embolism in diagnostic cerebral angiography and neurointerventional procedures: a prospective study.
        Lancet. 1999; 354: 1595-1597
        • Keith L.M.
        • Smead W.L.
        Saphenous vein stripping and its complications.
        Surg Clin North Am. 1983; 63: 1303-1312
        • Van Rij A.M.
        • Chai J.
        • Hill G.B.
        • Christie R.A.
        Incidence of deep vein thrombosis after varicose vein surgery.
        Brit J Surg. 2004; 91: 1582-1585
        • Hingorani A.P.
        • Ascher E.
        • Markevich N.
        • Schutzer R.W.
        • Kallakuri S.
        • Hou A.
        • et al.
        Deep venous thrombosis after radiofrequency ablation of greater saphenous vein: a word of caution.
        J Vasc Surg. 2004; 40: 500-504
        • Mozes G.
        • Kalra M.
        • Carmo M.
        • Swenson L.
        • Gloviczki P.
        Extension of saphenous thrombus into the femoral vein: a potential complication of new endovenous ablation techniques.
        J Vasc Surg. 2005; 41: 130-135
        • Manufacturer and User Facility Device Experience (MAUDE) Database