Acute complications of excimer laser coronary angioplasty: A detailed analysis of multicenter results

OBJECTIVES
The aim of this study was to document and analyze the incidence and consequences of complications of excimer laser coronary angioplasty.


BACKGROUND
Excimer laser coronary angioplasty has been reported to be a safe and feasible alternative or adjunct to conventional balloon angioplasty, but serious and unique complications have been observed.


METHODS
Data on 1,595 interventions of excimer laser coronary angioplasty in 1,521 patients were analyzed, using a merged data base from the U.S. and European Percutaneous Excimer Laser Coronary Angioplasty (PELCA) registries.


RESULTS
Procedural success was achieved in 89.3% of interventions. Stand-alone laser angioplasty was performed in 17.8% of interventions. Complications included dissection (22.0%), vasospasm (6.1%), filling defects (4.8%), abrupt reclosure (6.1%), embolization (2.3%), perforation (2.4%), arrhythmia (0.7%) and aneurysm formation (0.3%). Major complications were non-Q wave myocardial infarction (2.3%), Q wave myocardial infarction (1.0%), coronary artery bypass grafting (3.1%) and death (0.7%). Logistic regression analysis revealed correlation between dissections and the use of larger catheter size (p = 0.0005), high energy per pulse levels (p = 0.0001 for native vessels), lesion length > 10 mm (p = 0.001) and presence of a side branch (p = 0.01). The incidence of perforations was higher in women (p = 0.004), in treatment of total occlusions (p = 0.02) and in the presence of a side branch (p = 0.03). Fatal complications were correlated with patients with multivessel disease (p < 0.0001), patients with acute myocardial infarction (p = 0.0009) and older patients (> 70 years old, p = 0.004). The incidence of major complications decreased after performance of 50 laser angioplasty procedures at one institution (p = 0.02).


CONCLUSIONS
This analysis defines both the learning curve and the profile of complications for excimer laser angioplasty and provides insight into the selection of appropriate patients and proper performance of the procedure.

Analysis of energy per poke.The incidence of minor and severe dissections and perforations of the vessel wall were analyzed according to the maximal energy per pulse used in the procedure.Energy per pulse was calculated from the Ruence (mJ/mmf) and the area of the fibers at the catheter tip.The mean energy per pulse was 21 f 7 mJ in 340 interventions with dissections and 20 + 8 mJ in 1,132    the use of catheters with a diameter >1.4 mm and e tlttence 250 ml/mm'.Additional analysis showed an association of higher energy per pulse levels with an increased incidetw of dissections, especially in native coronary arteries.These findings are consistent with the results of other studies (37).
However. the results might also reflect a subset of lesions where a second pars with higher energy densities was required after an initially failed laser approach.The information in the data set does not allow further analysis of this potentially confounding factor.The influence of arterial wall morphology on the occurrence of dissections is substaotiated by the finding that severe dissections were more Irequently observed in younger patients, women and patients without diabetes.The overall incidence of vasospasm was low and is comparable to conventional coronary angioplasty.Differences in the reported incidence of vasospasm after laser angioplasty are probably the result of diUerent procedural protocols.If additional balloon angiaplarty is performed immediately after laser ablation, the interpretation of complications is restricted to the combined procedure.This limitation accounts not only for the observation of vow spasm but for the total scope of laser-related complications.Protocols that allow additional procedures like coronary angioplasty only alter a period of 20 % 30 min detect higher incidences of complications as a direct consequence of laser irradiation and laser-induced injury (7.13).
The mechanism of thrombus formation after laser irradiation is largely unknown and may include laser-iadependent consequences of deep vessel wall injury (38,391 and laserspecific processes.In the current clinical experience.the finding of intraluminal filling defects was rare but was frequently associated with vessel reclosure.Therefore.sufficient anticoagulant and antiaggregant treatment is mandatory in excimer laser angioplasty. Abrupt vessel closure at the intervention site complicating the coup of angioplasty is a common phenomenon in conventional balloon diiation (35,40-42), with an incidence of 4.6% co 6.8% (35.40).In this report, vessel closure zfter laser angioplasty was seen in 6.1%.and in 4.7% the operators classified the vessel closure as laser related.There was no baseline variable associated with abrup: vea! r~slosure.However, filling defects, vasospasm and dissections were associated with vessel closure in the multivariate analysis, which corresponds to the mechanisms of vessel closure atler conventional coronary aogioplasty (41-W Distal embolization was associated with myocxdial infarction in I6 (44%) of 36 patients.Embolization was more frequently observed in the treatment of saphenous vein grafts.The iocidence of distal embolization has also been found to be increased in target lesions containing thrombus (22).These data suggest that embolizalion of plaque material i! ~VNJ totarget l&m morphology, not to laser irradiation.The passage of the laser catheter is associated with shear stress at the cuter surface.Thus.embolization might occur in lesions susceptible to detachment of material.Advancement of the catheter tie into the lesion without efficient ablation might increase ihe risk of mechanical alteration of the plaque becao~ tissue ablation is dependent on contact of the laser probe with the target.the etch rate of laser ablation is low (1.45).and the steerability of the catheters is limited (7.12).As a consequence, the aepoach to friable lesic~w.especially in sapheooos vein grafts.shotdd be cautious.and advaoccment of the catheter should be strictly limited to the experimentally defined etch rates.
Perforaliw was awxiated with a maior clinical even1 in IS patients and led lo emergency 0peiGiin in 13.Baseline variables with ao indcpcodenl conelat&oo to a higher ineidense otthis comdiilioo were female ES&~.lrcatmcnt of total occlusioos a& lesions at a veaoel branch point.This is consistent with e recently published study oo the tirsl 764 patients of the U.S. PELCA regislry (23).which fowl a higher overall incidence ofcompiicatioos correlated with the treatment of lesions at anarterial biforcntion.Analysis of the individual complications was not performed in this study.The higher incidence of pcrforaiions in occlusions and branch points may be due to an increased risk of laser irradiation in the deeper vessel wall layers because the potential for misalignment is higher under these conditions.Although acoustic side effects (14) and vapor bubble expansion (IS) in the deeper wall layers remain to be investigated.tissue disruption is likely to occur.Thus.whether cxlmvasation of contrast media aller laser irradiation represents an oblique direct channel or a complex dissection with s~at1~ring of medial and dventitial tissue remains open to question.However, what is important is that balloon angioplasty CBR resolve the extmvasation in many cases (i3,46,47l, probably by simply compressing the vessel wall.A recently published study (4s) shows a relation between the catheter/vessel diameter ratio and perforations and suggests a safety margin of I mm between catheter size and vessel diameter.There-fore, attempts to reduce the risk of perforation shoud include I) carecu!taraet lesior selection.2) a cautiou approach to total occlusions and lesions at vessel branch points, 3) choice of laser catheter size in relation to vessel diameter.and 4) restriction of laser irradiation to wuations with coaxial catheter alignment.
The incidence of major complications was low and is comparable to that of conventional coronary angioplasty (29,35,49,50).In the small number of patients treated for acute myocasdial infarction in the European PELCA reg~stry, the incidence of Q wave infarction was increased, which is concordant with the experience in the treatment of acute ischemic episodes with balloon dilation (43).Excluding these patients treated for acute myocardial infarction from the analysis results in an incidence of 0 wave infarction of0.9A and a reduction of incidence of fatal complications to 0.6%.Fatal complications were associated with patients who had a high risk for percutaneous intervsntioas.No target lesion and no interventional variables were associnted with a hiiher incidence of fatal complications.These findings suggest that fatal outcome is more related to the individual clinical situation than to the laser angioplasty procedun.However.unless there is evidence for benefit of laser angioplasty compared with conventional ballwn angioplasty for specific target lesions.patients at high risk should be excluded fmm excimcr laser angioplasty trials.
As previously reported (35.49.Sl.52).oar data suggest a learning curve for excimer laser angioplasty.Whereas there was no change in the incidence of minor complications.a significant reduction in major complications with growing operator experience was observed.This reduction is predominantly due to a decrease in the incidence of myocardial infarction and coronary artery bypass gmtling.The low incidence of fatul complications does not allow further interpretation.The success rates showed a significant iacrease a&r performance of 20 and SO eacimer laser angioplarty procedures per ic?!itulion.Thus, probably as a result of a variation in patient se!ection and operator skill.the clinical results of excimer laser angioplasty impmved with growing operator experience.
Study iimitatiam.The study population was heterogenow in terms of patient and lesion selection as well as interventional strategies.The data are the combined results of two registries.However, although detailed analysis of failwe and procedural sacces~ revealed differences for specific target lesions, there was no substantial difference in the overall results and complication rates of both registries (23.24).This was not a randomized trial.and the data presented are purely observational.In addition, because of the nature of multicenter trials.the report of complications is based on the interpretation of different investigators.However, the analysis represents the results of excimer laser angioplasty performed in multiple centers in Europe and the United States.The analysis of the influence of multivessel disease is limited by the late inclusion of this variable in the U.S. registry, leaving missing values in 44% of patients.A major drawback of both registries is the almost routine use of additional balloon angioplasty Peter laser irradiation.Because of this pmcedwal strategy, ldcntification of the acuie and 'ubawte consequences of stand-atone laser an&opiasty on thz angotographic appearance of the target vessel was not possible.
Summary.This study documented the relative safety of excimer laser coronary angioplasty in selected patients and a learning carve resulting in increasing success rates and reduction of major complications with growing operator experience.Risk groups for individual complicadons were identified in this analysis that have direct implications for pehem ssieiiioe and interventional technique.The influence of catheter size and energy per pulse on the incidence of disseclions eras documented, saages:iagclinicai relevance of acoustic Gde effects and vapor babble formation a&r laser irradiation of tissue and blood.Fatal corn&cations were more frequent in patients at hii risk for percotaneous interventions.Including patients with acute ischemic syndromes.
,I[ KARL R. K 4RSCH.MD, FACC, AND THE COINVE~TIGATORS OF THE U.S. AND EUROPEAN PERCUTANEOUS EXCIMER LASER CORON.ARV ANW~PLASTY (PELCA) REGISTRIES iiibingen ond Berlin, Germany: Bosron.Mussachaserrs: Paris, frm~e: New York, ~Vew York: Dwhan!.North Carolina nary ahlg&pMy In 1,521 @it$ were aMly"zed.using B merged data b&se from tm U.S. and Eumtxan Percutanmus Excboer Lrrrr Comwy Ansioecrrtg (PELCA) re&lrics.laser an@aplasty WIS performed in 17.8% of btlerwatlom.complbxuols iadudfd d&wctiml(z2.o%),YBSO-lp~l (6.1%).till@ d&b (4.8%), abrupt reclosurr (6.1%).embW&n (2X6), $&Ion 12.4%), arrhyIbmin (0.7%) and mewyam fwm&, (0.3%).Mf&u ccmplk&?tts were tton.Q R.C yorudw inlwttcm (2.3%).Q mw myowdisl infarction When excimer laser irradiation is used in the setting of percutaneous transluminal coronary angioplasty.the major advantage of this technique is thought to be ablation of the atherosclerotic plaque.with only minimal thermal and mechanical injury to the adjacenl vessel wall wuctures (1,2i.(1.0%).mroosry artery bypass .@tiig(3.1%) nnd death (0.7%).Lo&tie rcwessimt a&sh ewe&d curreLation between hgth ;I0 ,&I (p = O.&l) and prrvnn of B side braorh (p = 0.01).The iwideoce of pwfomtions I*Bs bigher in WXWI ip = 0.001).in treatment of total wclosiom @ = 0.02) and in tbe presence of B side braneb (p = 0.03).Fatal wmp~tiws Hem ro&ated with p&tents with multiressel d pstiwts witb acute myornrdbd inlrrrlion (p r pstipne (>70 yean old.p = 0.&?4).The tideme of mBjor compttcntions decnssed SRer prfonnnnre of 50 War angiopla~$ pnwdorps St Or& bistibdioo (p = 0.02).Conclusions.This aw&sis de&~ both the learning EWVC nnd the pm& of compltcsticms for exciwr lsrrr lnt?foplasty =d provides insight Into the selfftioo of appmpriat@ patkttts aad proper performwe Of Ibe procedure.(, Am Colt C,,,+io11594;23:1305_13j As opposed to the mandatory fissuring and cracking of the plaque and overdistension of the arterial wdl using conventional coronary angioplasty (3-S), the new method lhus migtt have the theoretic potentml of reducing complicntions and prolxbly as a result of reduced injury of dcreasing lhe incidence of restenosis (6).The feasibility and relative safety of excimer laser co~owry angioplasty were demonstrated in early clinical trials in a limited number of patients (7-9).and retinemenrs in laser catheter technology resulted in improvement of acute interventional success, comparable to the results of conventional coronary anglopktsty (IO-!?!.However, the nte of successful stand-alone orocedures was limited by suboptimal reduction of steoosIs'9everity in larger diameter vessels.nccessi-Wine additional balloon dilation in most cases (ll,l2).Furthermore.preliminary results did not show B reduction in the incidence of restenosis (7.12.13).Recent experimental studies suggest major side effects of excimer lase: inadiation.Acoustic damage and rapid vapor buhhie expansiotl we likely to alter adjacent tissue signifi-cantly (14-16).In addition to the debate abcu: indications for excimer laser angioplarty (I I .&IF19).ongoing studies of the undesirable effects of laser irradiation (20-22) might elucidate the mechanisms of clinically siguilicantCompkations.The primary aim of this analysis was to docttmenl the incidence and consequences of acute complicalions after excimer laser angioplasty using the curreatly available technology.An attempt was made to identify associations between morpholqjc and interventional variables and the incidencr of complications that might have clinical implications for patient selection and perfomwtce of the procedure.Furthermore, the data for clinical results and complications were analyzed according to the number of procedures ?erformed per clinical site to identify th': learning curve of excimer laser angioplasty.Methods Patient anr&nen~.Analysis of complications was performed using a single data base after merging the data sets of the European and the U.S. Percutaneous Excimer Laser Coronary Angioplasty (PELCA) registries.Both registies were designed to evaluate the immediate and long-term success and complications of coronary excimer laser aagie plasty.The individual local ethical committees gave ap proval for the participation in the excimer laser angioplasly trial.All patients were candidates for coronary aagiopiasty uy clinical and angiographic findings, and all patients signed infonaed consent for the laser procedure.The selection of patients for laser angioplasly was based on the preference and experience of the individual investigator.Patient selection criteria did not change throughout the cottrs~ of the study.The laser procedure was perforated as previously described (7.9.12).Additional interventions were determined by !he operator's discretion.After the procedure the patients were monitored according to local practice.Data bsx.The data collection forms were completed by the participating investigators and then mailed to the data collection centers, where they were reviewed for complete.ness, accuracy and consistency.The European PELCA registry contributed data on 247 patients treated at nine clirtical sites between September 1990 and July 199' (see Appendix).The number of patients per center varied from 2 to 123.Quantitative analysis of the intervention angiogmms was performed at a central core laboratory (Deutsches Herzzentrum Berlin).The data collection forms and intervention angiogram results were sent to the University of Ttibiagen and entered into a data base for statistical analysis using the SAS program (SAS 6.04, SAS Inc.).The U.S. PELCA registry contributed the data sets of 1,274 patients treated at 30 clinical sates (see Appendix) between May 1989 and December 1991.The number of patients per site varied from 1 to 142.The intervention a&gram was analyzed by the investigators at the individual sites, aad the results were included in the data collection form.The data collection forms were sent to Spctrawtics Inc. and entered into a data base for statistical analysis by SAS (SAS 6.64).la addition to the original data of both data sets, a new variab!e was created to obtain information about the "karning cttwe*' of the procedure.Au interventions were numbered chmnologically for each investigational site.In the final analysis three groups were classitied: I) the fi:st 20 interventions at one site., 2) interventions 21 to 50, and 3) all interventions after 5i? excimer laser aagioplasties at the individual hospital.The definitions of morphologic.procedural aad outcoiie variables used have been described clsewherr (23.24).Tlte data bases were merged, aad a total of 97 variables were accumulated.The analysis was performed at the University dTi!hingen.La%r m.All interventions were perforated with a xenomchloride excimer laser system (CVX 300.Spectraaetits).as previourly described (12.23.24).&t&&al a&++.Statistical analysis W&E p&orated in cooperation with the Witate for Medti lnfommtioa Rocessing.University of Ttlhingen.AU tests were carried out with the SAS software Version 6.04.A srepwiae model-building strategy was appSed (25).To determine the wociation of patknl-.lesion-and laserspecific factors with each complii, alI uiahks (mabk I) were examined by univariate analysis (chi-square test or Fisher exact test).Ia a second step.all variahla showing a univatiate aasorlatioa with the compIk&a were entered in a backward logistic npnssion aaalysir for each iadividual complicalion.Additionally.varinhks that were probably eratioas or clinical e&ience were cnimd into the mdel.This sacoral step was petfmmed acpamtely for patient, ksioa aad iaterventional vtics.In the third step.all wiabks that showed a muitivariate association with the complicatiwt were catered into a model of stepwise logistic regression for each Wmplieatiott.This Rnd modal was optimized according to the lest of wncordance, Thus.Ihe fiaal modal showed tha variables with independent influence on the incidence of the cmttplkalii.observationa with miaaktg data for the speeipc analysis were treated as "missing values" awotdii to the SAS software algorithms; a p value < 0.05 was considered atatiatitally signifieaot.Data are presented as man value k SD. uesult.5 htknt ppnlakn, The data base conraimd data on I.595 interventions of excimer laser aagioplaaty petfumted ia 1,521 patients (mean age 60 2 12 years).The haselim characteristics are preaen!edia Table 2, showing the in-idence of the variable in tht available data sets.Symptoms of angina pectorio were classihed according lo the Caaadian Cardiovasfslar Society fttnctionrd score (26).Data w the number of diseasrd vessels were available for 854 patients only, of whom 3% (46%) had multivessel disease.Maltives-sei disease was considered present if there was significant stenosis (MO%) in at least two major coronary arteries.Praalural ontmme.Laser success was achieved in 81.1% ofall interventions.Procedural SUCCESS was repORed for 89.3% of the interventions.Stand-alone laser aogioplasty was performed in 285 interventions (17.8%).Additional procedures Ferfotmed to resolve complications were coronary angioplasty in 158 cases (9.9%).atherectomy in I (0.1%).thrombolysis in 14 (0.9%).bypass operation in 49 (3.1%).intmcorowy stenting in 3 (0.2%) and other pmcedweo in 22 (1%).
two with atrial hbrillalian and six with prriprwxduml ventricular Ushycardia or ventricular hbrillalion.CAEG = coronary tiey bypur gmfling: MI = myccardial infarction.interventions without dissections (p = 0.03).No significant differences were found for severe dissections and petforations, A second analysis was performed for interv~mtions in native target vessels, excluding saphenous vein grafts.Mean energy per pulse was 21 k 7 mJ in 307 interventions with dissections ard 19 + 7 ml in 854 interventions without dissections (p = O.OWl).Laser-related dissections were associated with a mean energy per puke of 20 f 7 ml in contrast to fP f 8 mJ for interventions without laser-related dissections (p = 0.03).No significant differences in mean energy per pulse were found for severe dissections and perforations.MuRlvarialc analyrls.The results of the multivariate analysis are presented in Table 4.The first three columns show the results of the separate analysis for patient, lesion and intervention variables together with the results of the final analysis, showing the variables with an association to a higher incidence of the specific complication in the last column.Dissections were seen more frequently in women (risk ratio [RR] 1.5, confidence interval [Cl] 1.1 to 1.8.p = 0.03), with the use of catheters >I.4 mm (RR 1branch points (RR 2.4.CI t.5 to 3.3, p = 0.05) and in the treatment of total occlusions (RR 3.1.Cl 2.3 to 3.8, p = 0.04).Filling defects were correlated tc the treatment of ulcerated lesions (RR 3.2, Cl 2.5 to 3.8, p = 0.0Xt61.Embolizations were more frequent in the treatment of saphenous vein grafts (RR 11.9, Cl 10.9 to 12.9, p = 0.0001) and ulcerated lesions (RR 4.0, Cl 3.1 to 5.0, p = 0.003).A multivariate analysis of the coincidence of complieations showed a correlation of abrupt reclosure with the observation of dissections @ = O.tW), vasospasm (p = 0.001) and tilting defects (p = 0.001).Lcrrning carve.The rate of interventions classified as laser success increased from 77% in the first 20 interventions per clinical site to 81% in interventions 21 to 50 and 86% after 50 interventions (chi-square test, p = 0.001).Procedutal success increased from 86% to 90% and 92%.respectively (chi-square test, p = 0.004).

Figure
Figure I shows the incidence of minor complications (occurrence of at least one minor complication, not including severe dissections and perforations), severe dissections.petforations and major complications (infarction.coronary artery bypass grafting, death) in the three groups of the learning curve analysis, There was no signiticaot difference in the occurrence of minor comolications and t&orations.Severe dissections were more f&tent in the grst 20 interventioris per site (p = 0.03).The incidence of major cumplications decreasedirom 8.5% in the tirst 20 interventions per site to 5.;X in interventions 21 IO SO and 4.5% in the subs-gent interventions (p = 0.02).The rosptive iocidelner of the individual major compiications in the three chmnologk patient groopr were as follows: myocardll infarction occurred in 5%, 1.9% and 2.3%: corcmary artery bypass grafting was necessary in 3.6%.3.7% and 13%: and fatal complications were reported in 09%.0.5% and 0.6%.