Endoscopic methods for the treatment of nonvariceal upper gastrointestinal hemorrhage

In the l970s, diagnostic fibreoptic endoscopy became part of the 
standard practice for evaluation of gastrointestinal disease. In the 1980s, therapeutic 
fibreoptic endoscopy is emerging as standard therapy for many gastrointestinal 
diseases. As the already sophisticated technology continues to blossom, it 
promises to become even more a part of the management of an increasing number 
of gastroenterological problems. Endoscopy can provide both a specific diagnosis 
as well as an identification of the high risk subgroup of patients with either active 
bleeding, or a non bleeding visible vessel that might benefit from endoscopic treatment. 
At endoscopy, patients with active ulcer bleeding have either arterial spurting, 
oozing or oozing beneath an overlying clot. These have poor outcomes: for 
example, when a non bleeding visible vessel is identified, the chances for rebleeding 
are approximately 50% during the period of that hospitalization . With an overlying 
clot without oozing, where dark spots are noted, there is less than a 10% 
chance of rebleeding. There are certain limitations for endoscopic hemostatic 
therapy and there are a few bleeding ulcers with an artery too large to expect 
endoscopic success. The kind of treatment chosen will be dictated by the availability 
of the therapeutic modalities and the skill of the surgeon.

A MAJOR ADVANC'r IN THL MANAtiE, me n t of gastroen terological problems has been in the treatme nt of patients with u p per gastroin teMinc1l tract bleeding.Recen t reviews h ave con sid ered the m a nage m ent of va riceal hemorrhage Va ri o us ap proac hc ~ to n onvariceal bleeding incl ud e tissue contact by rlectrocoagulacion , bot h monopo lar and b ipo lar e lectro des or heater p robt's.A second approach.without tissue contact, consists o f laser p hococoagulation with e ith e r ch e a rgo n or ch e ncod ynium ytt rium al u m in u m garne c( Nd:YAGJ lase r.The thi rd approc1ch is an injccuon therapy of known b leed ing lesion~ with a lco h o l, sclerosa n ts or ad re naline.All of th ese therape utic methods arc rcla• cive ly safe an d effective in co ntrolling b leeding a n d avoidi n g emergencv su rgery.

LASER PHOTOCOAGULATION
In 197 l.Goodale and co-workers (I) reported th e fir st stu d ies using a carbnn d ioxi de laser co con trol b leeding b1ons in anim als.Cu rren tl y th e <l: YAG and argon lasers are used co treat gascromcestlnal b leedi ng.The relative m erit~ of the Nd: YAG and argon lasers for the th erapy of acute up per gastro intestinal b leeding arc as fo llows: when applied end oscopically.both lasers e ffectively produce in itial h emostasis in ma ny pa• mcsure que la tcchnologie d 'avant garde progrcssc, cette prrnique confirme la place qu'elle occupe dans le craitement de certains problcmcs gastrointcstinaux Un progrcs majcur a ninsi etc realise dnns le traitemcnt des pnticnts Souffrant de sa1gnements des voies gastrointestmales Parmi !cs approches divcrses, ii faut inclure Ir contact tissulairc par electrocoagulation.qui comprcnd !'usage des electrodes unipola1res e t bipolaires ou les sondcs thermiqucs.La seconde approche sans mortality for acute and chrome ulcer bleed mg treated with laser as comp.ired to the experience 111 Munich before laser therapy wns available However, this survey did not address the incidence of rebleeding.and this sem•s of pmients wns nor randomized Reports of h igh success rates for laser therapy of u p pergastrrnntestina l b leeding must be evnluated with the perspective that ,1 p prox1matcly 70''., of all episodes are sclf-lim,ted and resolw without specific the rapy.
Of more recent studies cvaluaung Nd:YAG laser therapy for all bleeding from ulcers, ..everal suggest that laser confers a benefit to the patient.However, some reports differ in their results and suggest no benefit.The study of Swain and collengues (3) de::.erves a more deta iled discussion.Of 465 patients with upper gastrointest111al bleedmg, 2 32 had pep1ic ulcers and 147 of the ulcers were e1ther bleeding or showed stigmata of recent hemorrhage.Of 122 patients 111cluded m the study, 62 were treated with Nd:YAG laser w hile 61 served as controls.In the laser treated group b lcedi ng \\'as more effective ly controll ed ( P < 0 .02) and the morta Ii ty was less ( P < L 1.05) than in the control group.In those patients with acnvc bleeding from visible vessels .the laser group fared better Macleod and colleagues (4) reported rhar of 184 patients found at endoscopy w be bkTding from peptic ulcers, 20 were bleeding from arte ries.Eight of these were allocated to placdx) treatment and nll latcr unde rwent eme rgency surge ry for further hemorrhage Of the 12 patients who underwent laser treatment, eight had surgery, but of these only one was still bleeding and actually required surgery.These differences were statlstically significant.The authors concluded that lnser treatment was a snfc and effective method of reducing the incidence of further bleeding and emergency surge ry but the techn ique was difficult and not applicnhle to all paucnts H:i lp ri n and co-authors However.some studies of Nd :YAG laser therapy suggest no benefit.For example, Krejsand co-workers(6)scudied, over 4 3 months, 174 selected patients with either active bleeding (n = 32) or stigmata of recent bleeding (n = 142) due to peptic ulcers who were randomly assigned during endoscopy to either standard treatment with laser phorocoagulation or therapy without photocoagulation.There was no significant difference in outcome between grours.Continued bleedrng or n::bleeding was observed in 22''(, of the l,1ser treated group and in ZO"o of che control group.
Urgent surgery was necessary in 16'\, of the laser treated patients and in 17"/, of the controls.Laser treated patients spenr a mean of 41 h in the intensive care unit, compared with 32 h for control patients.
The mean hospital stay was 12 days in the laser treated group and 11 Jays in the control group.One death occurred in each group When patients with active bleeding were analyzed separately, there was no significant difference 1n outconH:: be-tween laser treated and control groups even though laser photocoagulation initially stopped active bleeding in 88°0 of cases.Among patients with visible vessels, rebleeding occurred in five of 14 ( 36°,~) who received laser rrearment and two of 15 ( 13''.~) controls.Laser treatment rrecip1tated bleeding in four patients and duodenal perforation in one.The authors concluded that Nd:YAG laser photocoagulation did not benefit patients with acute upper gastrointestinal bleeding from peptic ulcers.However. it should be emphasized that Krejs elim111ated many patients who were hemodynamically Lmsrable from the study, thereby drawing a major criticism of the paper.
The results of three argon studies are mixed (7-9) and argon lasers are seldom used at present.

ANGIODYSPLASIA
The routine use of fibreoptic endoscopy in the evalumion of upper gastrointestinal bleeding has demonstrated that angiodysplastic le •ions in the upper tract may hemorrhage.Generally, bleeding from angiodysplastic lesions is selflimited but may be recurrent.If the bleeding lesion is within the reach of the endoscope, it 1s amenable to local therapy.
One of the several treatment modalities for angiodysplastic lesions is laser photocoagulation.There are several reports describing endoscopic obliteration and clinical benefit with both argon and Nd:YAG laser therapy.Waitman et al ( 10) treated 50 patients with argon laser therapy; 3 3 had com piece cessation of bleeding with follow-up of six months co four years.The ocher 17 had markedly reduced bleeding.Bowers and Dixon ( 11) and Jensen and colleagues (9) reported decreased bleeding episodes and reduced transfusion requiremenb with argon laser therapy in patients with angiodysplasia and in a group with classical hereditary tclangiectasia.Fleischer ( l2) reported similar benefits using the Nd:YAG laser.None of these studies incluJed perforation as a complication.
Laser use is nor free from problems.Commercially available lasers arc expensive and the machines arc not portable in the practical sense.Laser therapy is not free from risk, aside from usual risks and complications of endoscopy and anesthesia.Major complications related co laser use itself occu r in approximately 4°{, of patienrs including perforation, fistula to other organs and bleeding.

ENDOSCOPIC ELECTROCOAGULATION
Endoscopic eleca•ocoagulation appears co be an inexpensive.readily available technique for the control of massive bleeding.Endoscopic electrocoagulation may be performed using monopolar or bipolar electrodes, or by fulguration Electrocoagulation results as current flows through tissue near the electrode, hitting and desiccating the tissue ro form a layer broken down and condensed into a necrotic mass.With monopolar clecrrocoagul:uion, current flows through the patient into a ground plate.In bipolar electrocoagulation rhe current dens1t:y 1s very concentratcJ at the bipolar clecrrotip because the tissue contact completes a circuit between rwo wires only a few millimetres apart.This limits the nsk oi injury.which reduces the risk of perforation.
Morcro and colleagues ( 13) report~d efficacy of monopolar clectrocoagulation in the treatment of bleeding gastric ulcers.In a controlled prospective fashion, the efficacy of monopolar clecrrocoagulation in the emergency trearmt'nt of bleeding gastric and stomnl ulcers was studied in 37 patients: 16 underwe nt electrocoagulation while the remaining 21 were treated by convenrional methods (conrrol group).Hemorrhage recurred 111 only one (6.2";,) of the patients in the elecrrocoagulation group, but bleeding recurred in 11 (52.4'~;, ) of the 21 control patients (P <0.05).with no significant difference in mortali ty of the two groups.
Goff I 14) compared the efficacy of hi polar clectrncoagulation with Nd:YAG laser photocoagulation for upper ga,trointestinal bleed mg lesions.The total study group included 33 patients with solitary, actively blecd111g lesions, reblceding lesions or a lesion containmg a visible vessel in the upper gastrointestinal tract.The patients selected for this study were heavily weighted rowarJ high risk i,troups for rebleeding.Altogether.11 patients unJerwent 37 coagulation sessions, 19 of the patients were randomized (eight laser, ll bipolar).In the randomi:cd grour, 4 7.4"., haJ no rebleedmg after therapy (laser 37 5'\,, bipolar 54.5''.,,P<O.l ).The 14 nonranJom1zed racients were primarily treated with bipolnr coagulation because they wt.'rl'felt to be too unstable to transfer to tht.• laser unit ,lt University Hospital.58 6'';, had no further b leed111g.Eleven ( 13".,) of the patient, required surgery but no rnuent dted from bleeding or complications relmed to the study Thu,, there was no significant difference in the frequency  ulator.The Nd: YAG is generally not portable so the patient must be transporred to the laser unit, whereas the bipolar cm1gulator can be taken easily to the patient's bedside Consequently, mosc authors now recommend that the hi polar coagulator he used 111 preference to the Nd.Y AG laser.Smee the heater probe may be more effectiw and cheaper chan the Nd:YAG laser, it is ,1lso potcnually preferable Wlwthcr the heater probe is more efficaciou:, chan the bipolar coagulator re maim co be dccermmed

TH E H EATER PRO BE
The he;itcr probe 1, a device that ,:an s1mulcaneously give heat and pressure lt com1:,ts of ;i hollc)\\ aluminum cylmder with ;in inner hear coil anJ an outer w,1ting of Teflon The alummum ha, high thermal conducuv1ry which prcw1des uniform d1smbut1on of the ht.'at.Storey ( 16) reporced cm 15 patients with ;i gastrn: ulcer, mean age of 60 years and mean pretreatment blood transfusion requirement of (1 6 L. created wirh the hemer probe All but one of the 15 patients avoiJed immedime operation and only two oi the remaming pat1enr, had delayed bleeJ111g, one of which required upcration There were no Jeaths 111 this group Over the ,ame period there were 10 panents wnh duodenal ulcer wnh a mean agl' of 62 years anJ mt•an pretrearment blond transfusion requirement cif 5 L. Operation wns .1vo1Jed111 only t\\'O pai1l'nt:., of whom cmt.• had a smnll delayed hemorrhage which was treated con-,t.'nauvely Of the rt.'st, the technique foiled in three hecause ,1ccess was impns-s1hle owmg to bleeding, concnct \\'as po,sihk but unsucce~sful 111 three patients and there were two patten ts wnh 1111ual success follom•d hy delnycd ht.internal probe ( 1 2 mm in d1amett.•r)wa, u:;ed preferentially with the GIF-lT endoscope (although in two mst,1111:es tlw small probe [2.4 mm] was used) Both probes included water 1mi.:ationIn casl'' of bmk ulcer bleeding, the inactivated heater probe was applteJ with moderntc force, either directly or ci rcumferentially around the bleeding site, to fmd a pren'>e pomt rhat ramponaded hleedmg Upon successful vessel ramponade, water 1 rngauon rrod uced rapid clea rt ng of blond from the ulcer bas<.'At that junc turt.•,several continuous hearer probe.•pulses (each 30 J) were applied to rhermalh seal the compressed vessel If hlcedmg recurred after miual hemostatic success.a second endoscopic heater probl' treatment was performed 1mmediatdy This study indicated that hemer probe treatments were more effective, casin and quicker than laser therapy Johnston l't al f 18) compared YAG la,er, argon laser, monopolar and bipolnr elec-trowagulat1l,n, elcctrofulguration and heater probe 111 coagulation of canll1l' arteries They concluded that the most efft.'ctl\'l'way w coagulate medium s1;:e mcsentertc nrtent.•swas vessel occlusion hv rompres-.1onfollowed hy heat npplication to ~cal 1t Overall hemostatic rank mg ,,•as heater pro he, bipolar, monopolar.YAG.argon.electrofulguracion These data\\ ill he useful to clinicinns plannmg endo,cop1c therapy 0f arterial blced111g future concrolled clm1cal studies com paring heater probe, BICAP and YAG laser \\'di he of 111teresr The maior disadvantage for mono• polar electrocoagulauon mclude, pocenu,11 probe ad herencc to the t1~sut.'The risk l)f perforation has ht•en as high a, 1.6''., 111 some stud 1es.The heatl'r pro ht.• and RICAP were t.k s1gncd so that thl're 1, n(> potennal for acure ussut.•t.'n1s1(111No perforauon~ hnvt.' been reported to datt' with these mstrunwnts Some disadvantages of the RICAP un,r 111clude subopnmal probe stiffness 111 design, lack of prox1m;1l wnter 1rnga11on and a hcmn• rericnced e ndoscopic therapists, severe concomitan t med ical disease that will increase mortality, lack of endoscopic access to the bleeding point.large u nmovable clots and b leeding from th e posterior wall of the cap which will reduce the effectiveness of therapeutic endoscopy.
Par example, quand un vaisseau visible sans saignement est idcnufie, Jes chances de rcsaignements sont de 50' \ , durant l'hospitalisauon Dans le cas d'un ca11lots sus-jacent sans suicemcnt, ou l'on remarque des rnchcs sombres, ii ya moins de IO'\;, de possibilitcs de resaignement.uents; the Nd:YAG laser is tech111c:11l y easier to use because distance from the bleeding lesion is not quite so critic.iiand coaxial gas flow requires carbon dioxide m a nonconc::ict mod e; an d th e risk of perforation is theorenrnlly less wnh the argon laser becaus•e the depth of tissue penetration is le,.,,At present che Nd YAG lnser is used most commonly.were treated at h b cen tre in Munich All paue n ts.ie. an unselected group.were com,id ered .Lesions included variccs.Mallory-Weiss tears.ulcers, erosions.vnscular anomalies and tumours.The overall succe'>s rate of initial hemostasis was 94'\,.Ki e01aber claimed n reduced