Comparison of Argon Laser Photocoagulation and Bipolar Electrocoagulation for Endoscopic Hemostasis in the Canine Colon

Transcription

1 GASTROENTEROLOGY 1982;83:830-5 Comparison of Argon Laser Photocoagulation and Bipolar Electrocoagulation for Endoscopic Hemostasis in the Canine Colon DENNIS M. JENSEN, GUSTAVO A. MACHICADO, JORGE TAPIA, and WILLIAM MAUTNER Medical and Research Services, Wadsworth Veterans Administration Hospital and UCLA Center for the Health Sciences, and Center for Ulcer Research and Education (CURE), Los Angeles, California Severe bleeding from the lower gastrointestinal tract commonly occurs from mucosal lesions such as angiodysplasia arid less often from acute ulcerations. Endoscopic coagulation of such lesions is now feasible. Our purpose was to compare the efficacy and histologic injury of bipolar electrocoagulation and argon laser photocoagulation applied at laparotomy in the canine colon. After right colotomy and heparinization in adult mongrel dogs, bleeding standard ulcers were randomly assigned to treatment or control. The colonic ulcers randomized to each group were not significantly different in size or bleeding rate. Efficacy of treatment was evaluated acutely. Adjacent normal mucosa was treated with the same modality as the ulcer to simulate treatment of a mucosal lesion. The same total exposure time (argon laser photocoagulation) or number of pulses (bipolar electrocoagulation) was used for these paired treatments. Gross and histologic injury were determined after 7 days. Both argon laser photocoagulation and bipolar electrocoagulation were uniformly effective in controlling bleeding from standard colonic ulcers and the incidence of full thickness injury was similar, 33%-48%. The histologic damage seen with normal mucosa treatment was consistently less than with standard ulcers for Received November 19,1981. Accepted May 7, Address requests for reprints to: Dennis M. Jensen, M.D., Division of Gastroenterology CHS, UCLA Center for the Health Sciences, Los Angeles, California This study was supported in part by Veterans Administration Research Funds and National Institute of Health Grant AM (to the Center for Ulcer Research and Education). The authors thank Anita Boesman for secretarial assistance; Janet Elashoff for statistical assistance; Esther Braden for histology preparation; Cooper Medical Corporation for use of the argon laser; Medi-Tech Corporation for the bipolar electrodes by the American Gastroenterological Association /82/ $02.50 all treatment groups. The difference is easily explained by the reduction in colonic wall thickness of about 50% with creation of standard ulcers. In our opinion, argon laser photocoagulation and bipolar electrocoagulation are safe enough to be used in clinical trials for control of colonic bleeding from mucosal lesions but not acute colonic ulcers. Severe bleeding from the lower gastrointestinal (ei) tract commonly occui's from mucosal lesions such as angiodysplasia and less often from acute ulcerations (1). Persistent acute or intermittent chronic hemorrhage from the lower GI tract has been traditionally treated by emergency surgery. This treatment may cause significant morbidity and mortality particularly in elderly patients (2). For angiodysplasia, telangiectasia, bleeding sessile lesions, and ulcerated lesions, colonoscopic hemostasis may offer an alternative to emergency surgery for bleeding control or a means of stabilizing patients before elective surgery. Colonic bleeding and nonbleeding lesions have been treated via colonoscopy with argon laser photocoagulation (ALP) (1,3), neodymium-yttriumaluminum-garnet laser (YAG) (4), and monopolar electrocoagulation (MPEC) (5). However, there are no systemic studies of efficacy or safety for any endoscopic hemostatic devices applied to the colon of patients or animals. A major limitation of applying emergency colonoscopic, hemostatic methods has been the inability to carefully and safely examine a colon full of clots, blood, and stool. However, we recently reported that emergency colonoscopy after preparation of patients with saline purge was safe and effective for diagnosis of patients with severe hematochezia (1). Before clinical application of colonoscopic hemostasis, we sought to study two different endoscopically applicable hemostatic methods in the canine

2 October 1982 ENDOSCOPIC HEMOSTASIS IN CANINE COLON 831 colon. In this study, we compared the effectiveness and histologic injury of ALP and bipolar electrocoagulation (BPEC) of bleeding standard colonic ulcers and the tissue damage incurred by treating normal mucosa. Intact mucosa simulated mucosal lesions such as angiodysplasia. Argon laser photocoagulation and BPEC were selected because their margin of safety in the endoscopic treatment of bleeding canine gastric ulcers was significantly greater than Y AG and MPEC as previously reported by our group (6). Materials and Methods A 20-W argon laser (Cooper Medical Devices Corp., San Leandro, Calif.) with a 400-pm flexible lightguide (Quartz Products Corp., Plainfield, N.J.) and full angle of divergence of 10 was used with coaxial carbon dioxide (C0 2 ) gas. The spot size, treatment distance, power density, and CO 2 back pressure (100 II1mH 2 0) at the lesion were controlled as previously described (7). For laser-treated lesions, power density (power output/spot size) and total treatment energy (power x treatment time) were calculated. A Valley Lab SSE2-K electrocoagulation unit (Valley Lab, Boulder, Colo.) was used with a prototype bipolar electrode (Medi-Tech Corp., Watertown, Mass.). The coagulation setting, pulse duration, force of apposition, and infusion rate were controlled as previously detailed (8,9). Argon laser and bipolar electrocoagulation were applied at laparotomy after colotomy. The bipolar electrode was hand held. The laser catheter was also held with a handpiece which controlled treatment distance (7). Statistical comparisons were made using the sign test or the Fisher's exact test (10). The statistical significance level of p < 0.05 was used throughout. Results were not significantly different unless noted. Adult mongrel dogs weighing between 40 and 60 lb were used. Each dog was anesthetized with intravenous sodium pentobarbital and intubated. A sterile midline laparotomy was made. The right colon was mobilized and stool was milked gently out of the ascending colon. An atraumatic clamp was applied on either side of the selected segment, such that blood flow to this segment was not impaired. A 10-cm colotomy was made on the antimesenteric side. The remaining stool was lavaged with sterile normal saline and gentle suction. For each animal, 10 different measurements of the colonic wall thickness at the colotomy were made with a standard micrometer. Intravenous sodium heparin (200 Ulkg initially; then 100 Ulkg-h) was then administered for anticoagulation. Standard ulcers (Quinton Corp., Seattle, Wash.) were made on the mucosal side using a negative suction pressure of 5 in. of Hg. Before randomization, the bleeding rate was immediately measured with a small graduated cylinder for a period of 15 s. Study 1 An extensive study was done to determine effective coagulation (BPEC) and power settings (ALP) for the treatment of standard ulcers in the colon of heparinized mongrel dogs. The lowest effective coagulation setting was 4 for BPEC. A treatment distance of 1.5 cm was chosen for ALP with a corresponding spot size (diameter) of 3.0 mm. The lowest effective power setting for ALP was 6 W. More effective, but higher power settings of 8 Wand 10 W were also chosen for further testing in a histologic study. Study 2 A chronic study was undertaken using 9 dogs. After colotomy and heparinization, standard ulcers were made as described above. Approximately 15 ulcers were made in each dog. Each ulcer was randomly assigned to one of four treatment groups (ALP 6 W, ALP 8 W, ALP 10 W, BPEC) or control. Control ulcers were observed for 3 min and also before surgical closure of the colotomy. The tissue from each standard ulcer ("ulcer plug") was measured and weighed. Efficacy, defined as the ability to stop bleeding completely, was assessed at the time of each treatment. Argon laser photocoagulation treatment times > 120 s or BPEC treatments requiring more than 50 pulses were considered failures. As an indication of full-thickness coagulation or injury, serosal whitening was also assessed acutely for each treatment or control lesion. Immediately adjacent to each ulcer, a similar area of normal mucosa (approximately 1 cm in diameter) was treated with the same modality (ALP 6 W, ALP 8 W, ALP 10 W, or BPEC) and for the same period of time (ALP) or number of pulses (BPEC) as that used for the ulcer. For all control ulcers and control mucosal lesions, CO 2 alone without laser photocoagulation was applied for 20 s. Ulcers and mucosal lesions were labeled with silk sutures. After surgical closure, antibiotics were given for 48 h. The histologic damage for ulcers as well as mucosal areas was assessed from 1-mm tissue sections and from coded histologic sections at 7 days as previously described (8). Colonic wall thickness (from mucosa to the serosa) and depth of injury to normal mucosa treatment were also estimated from these sections using a calibrated microscopic eyepiece. Results Initial Determinations Colonic thickness. The mean micrometer measurement (n = 10) in each animal was considered its colonic wall thickness. The mean colonic wall thickness (±SE) was 1.94 ± 0.07 mm. From histologic sections the mean thickness (± SE) was 2.22 ± 0.11 mm (p > 0.05). The mean standard ulcer depth from histologic sections was 1.00 ± 0.06 mm (n = 20). Colonic ulcer size, weight, and bleeding rate. The mean (±SE) for plug size, plug weight, and 15-s bleeding rate were 8.6 ± 0.2 mm, 27.3 ± 0.8 mg, and 0.53 ± 0.02 cm 3 /15 s, respectively. There were no significant differences among control and treatment groups. Efficacy. None of the control ulcers stopped

3 832 JENSEN ET AL. GASTROENTEROLOGY Vol. 83, No. 4 bleeding during the observation period (Table 1). All the BPEC, ALP 10 W, and ALP 8 W treated lesions stopped bleeding. Ninety-six percent (26/27) of the ALP 6 W lesions stopped bleeding with treatment. Argon laser photocoagulation 6 W treated lesions required more total joules for hemostasis (mean 200 ± 49 J) than either ALP 8 W (mean 175 ± 48 J) or ALP 10 W (mean 173 ± 51 J). Standard colonic ulcer responded to BPEC or ALP thermal coagulation treatments by contracting and shrinking. The shrinkage effect was observed for adjacent mucosa, submucosa, and vessels in the ulcer base. This shrinkage with coagulation was more dramatic with ALP than BPEC. Subjectively, colonic ulcers were easier to treat with ALP than BPEC because contact with the lesion was not required. Sometimes ulcer bleeding increased with BPEC electrode contact or withdrawal from the ulcer base. However, rebleeding of ulcers with manipulation of the colon during serosal inspection was not noted with either ALP or BPEC, as was previously noted in the stomach (6). Histologic Damage Acute serosal whitening was not observed with any of the treatment groups. A coagulation effect of the ulcer or mucosa (whitening of the tissue) was observed after each treatment. None of the treatments re1iulted in tissue vaporization (acute destruction with ulceration). For histologic damage 7 days after randomization, Figure 1 illustrates control (a) and treated (b) standard colonic ulcers and Figure 2 illustrates treated adjacent normal mucosa. These lesions were treated with ALP. Bipolar electrocoagulation treated lesions were indistinguishable on coded sections. At the time of autopsy, there was no evidence of chronic bleeding, perforation, penetration, obstruction, or abscess formation in any animal. The frequency of full-thickness external muscle layer damage (100% EMLD) is tabulated in Table 2 for each treatment group or control for standard ulcers. The tissue damage with ALP 8 W (48% incidence of 100% EMLD) was arithmetically greater than with ALP 10 W (39%), ALP 6 W (33%), or BPEC (33%). All treatment groups had significantly more damage than control standard ulcers of the colon. The frequency of 100% EMLD for normal mucosa treatment and for control groups is shown in Table 2. The tissue damage with BPEC (11% incidence of 100% EMLD) was similar to ALP 10 W (18%) but less than ALP 6 W (26%) or ALP 8 W (26%). The differences were not significant. The mean depth of injury after BPEC or ALP treatment of normal mucosa was mm (Table 2). Discussion The colonic ulcers randomized to each treatment group and control were not significantly different in any of the parameters measured: ulcer size, ulcer plug weight, or bleeding rate. They were standardized. The depth and size of colonic ulcers were similar to standard ulcers of the stomach and duodenum as reported in our previous studies (9). The mean bleeding rate of standard colonic ulcers (0.53 cm 3 /15 s) was similar to duodenal ulcers (0.41 cm 3 / 15 s); but significantly more than esophageal ulcers (0.42 cm 3 /30 s); and significantly less than gastric ulcers (1 cm 3 /15 s). Bleeding rates of colonic ulcers decreased slightly with time. All controls continued to bleed throughout the periods of observation and operation. As the heparin effect wore off, all controls eventually stopped bleeding and none of the dogs exsanguinated. Both ALP and BPEC were uniformly effective in controlling bleeding from standard colonic ulcers. Standard colonic ulcers were similar to ulcers in the stomach, esophagus, and duodenum in that 2-6 bleeding points were present at the ulcer base. In comparison with our previous studies, the shrinkage of ulcers with thermal coagulation in the colon was not as dramatic as in the esophagus but more than in the duodenum (9). This may relate to the relative hemostatic efficacy in the different organs for ALP and BPEC. Colonic ulcers were subjectively harder Table 1. Efficacy of Treatment for Colonic Ulcers Modality Control BPEca Power densityc Tx. distance (cm)d # stop/total e 0/25 27/27 Percent Applicationsf or exposure 8 36 ± 4 Total energy (J)h ALP 6 b ALP 8 ALP /27 27/27 28/ ± 6 22 ± 6 18 ± ± ± ± 51 a BPEG = bipolar electrocoagulation. b ALP 6-10 = argon laser photocoagulation 6-10 W. c Power density is in watts per square centimeter. d Tx. is treatment. e Efficacy of treatment (#stop/total) is number of lesions with complete hemostasis for each treatment group. f Applications are the mean (±SE) number of BPEC applications for hemostasis. 8 Exposure is the mean (±SE) laser exposure time in seconds for hemostasis. h Totlll energy in joules (]) is the mean (±SE) ALP energy required for hemostasis.

4 ENDOSCOPIC HEMOSTASIS IN CANINE COLON October Figure 1. Control (a) and argon laser (ALP) treated (b) standard colonic ulcers 7 days after randomi zation. A is the mucosa; B the ulcer base; C the external muscle layer; D the serosa; E th e depth of external muscle layer damage (EMLD). to treat than esophageal ulcers with ALP or BPEe and required significantly more total energy for hemostasis (e.g., for ALP 10 W: 173 ± 51 J vs. 77 ± 4 However, colonic ulcers were easier to coagulate with ALP or BPEe and required significantly less total ALP energy for hemostasis than duodenal ulcers (e.g., for ALP 10 W: 173 ± 51 J vs. 494 ± 82 The incidence of full-thickness injury to standard colonic ulcers was similar for all ALP powers (33%48%) and BPEe (33%). The tissue injury for ALPtreated colonic ulcers in the present study was significantly less than the esophagus (9) and duodenum (9) but significantly greater than the stomach n. n (7). The mean incidences of 100% EMLD for ALP 10 W were 39% (colon). 70% (esophagus). 60% (duodenum), and 15% (stomach), respectively. For colonic ulcers in this study there was not a good correlation between the extent of histologic damage and animal weight, bleeding rate, or total ALP energy required for hemostasis. The histologic injury with BPEe treatment of colonic ulcers was similar to standard gastric, esophageal, and duodenal ulcers of our previous studies. The mean incidences of 100% EMLD were 33%, 30% (12). 20% (9), and 15% (9). respectively. It is evident from the present study that neither

5 834 JENSEN ET AL. GASTROENTEROLOGY Vol. 83, No.4 Figure 2. Argon laser photocoagulation treated normal mucosa 7 days after randomization. For A-E refer to Figure 1. high-power ALP nor BPEC have as much margin of safety in terms of histologic damage for treatment of standard colonic canine ulcers as with gastric ulcers. Caution is recommended for the treatment of patients with bleeding from similar acute colonic ulcers. Such lesions in patients may reduce the thickness of the already thin colonic wall. The reason for treating normal colonic mucosa was to simulate treatment of more common (1) bleeding colonic lesions (angiodysplasia, telangiectasia, or sessile lesions). These lesions do not usually ulcerate or reduce colonic wall thickness similar to acute standard ulcers. Although the histologic damage seen with normal mucosa treatment was consistently less than with standard ulcers for all treatment groups, the difference was significant only for BPEC (Table 2). The difference is easily explained by the Table 2. Histologic Damage from Treatment of Colonic Ulcers and Mucosa Modality Control BPEC" ALP 6 b ALP 8 ALP 10 Ulcers with 100% 0/25 9/27 e 9/27 13/27 11/28 EMLDc (33%) (33%) (48%) (39%) Mucosa with 100% 0/25 3/27 e 7/27 7/27 5/28 EMLDd (11%) (26%) (26%) (18%) Necrosis depth after mucosa treatment f (74%) (75%) (81%) (73%) Refer to Table 1 for a and b. C Ulcers or d mucosa with 100% EMLD are the fraction of lesions (and percentage) with full-thickness injury to the external muscle layer. All treated ulcers had significantly more damage than control ulcers. e For BPEC there is significantly greater damage to ulcers than mucosa. f The mean depth of necrosis after mucosa treatment is in millimeters (or as a percentage of the total tissue thickness). reduction in colonic wall thickness of about 50% with creation of standard ulcers. Laser photocoagulation and electrocoagulation are potentially useful for destruction of sessile or small colonic polyps by either vaporization or coagulation. Brunetaud et al. have treated rectal polyps with argon and neodymium YAG lasers (13). Low-power density (such as in the present study for ALP) was used for coagulation and high-power densities for vaporization. Similarly Dixon et al. (14) have treated sessile rectal polyps with argon laser in patients with familial polyposis syndrome. Our depth of tissue injury data after BPEC and ALP coagulation of intact colonic mucosa (Table 2) may be useful in planning similar clinical trials. Independent of tissue vaporization, approximately 1. 7 mm depth of tissue coagulation can be anticipated. What is the potential clinical significance of the injury pattern after acute colonic ulcer and mucosa treatment with these modalities? The answer is uncertain at the present time. Careful clinical studies with histologic correlation will be needed for this determination. Although an 11 %-48% incidence of full-thickness muscle injury (Table 2) resulted from treatment of mucosa and colonic ulcers, neither perforations (sealed or overt) nor abscesses were found at autopsy. We would anticipate similar results with clinical application of ALP and BPEC. By contrast, MPEC of acute colonic ulcers and mucosa caused delayed perforations in our subsequent canine colon studies (15). In the present study, Figure lb exemplifies the severe serosa inflammatory reaction 7 days after coagulation for many acute ulcers and a few mucosal lesions with full-thickness exter-

6 October 1982 ENDOSCOPIC HEMOSTASIS IN CANINE COLON 835 nal muscle layer injury. After postoperative recovery, these animals all continued to eat and maintain their activity without overt effects of this injury. With clinical application of these thermal devices, such a severe serosal inflammatory reaction of the colon might result in symptoms similar to the "postpolypectomy syndrome" after mono polar snare electrocautery (16). This syndrome has not been reported for endoscopic thermal modalities except MPEC. Ultimately, careful prospective clinical studies will be needed to determine the safety of promising hemostasic methods such as argon laser and bipolar electrocoagulation in the colon. In our opinion, BPEC and high-power ALP should be considered for control of colonic bleeding from mucosal lesions but not acute colonic ulcers. As a feasibility study, we have initiated a clinical trial to evaluate BPEC and ALP for treatment of bleeding colonic mucosal lesions such as angiodysplasia after saline purge (1). These modalities have not yet been adequately tested clinically to recommend widespread clinical use. References 1. Jensen DM, Machicado GA. Emergent colonoscopy in patients with severe lower gastrointestinal bleeding (abstr). Gastroenterology 1981;80: Eaton AC. Emergency surgery for acute colonic hemorrhagea retrospective study. Br J Surg 1981;68: Brunetaud JM. Enger A. Flament, et al. Utilization d'un laser a argon ionise en endoscopic digestive: photocoagulation des lesions hemorragiques. Rev de Physique Appliques 1979; 14: Kiefhaber P, Nath G, Moritz K. Endoscopic control of massive gastrointestinal hemorrhage by irradiation with a high power neodymium-yag laser. Prog Surg 1977;15: Rogers BGH, Adler F. Hemangiomas of the cecum: colonoscopic diagnosis and treatment. Gastroenterology 1976;71: Johnston JH, Jensen DM, Mautner W. Comparison of endoscopic electrocoagulation and laser photocoagulation of bleeding canine gastric ulcers. Gastroenterology 1982;82: Johnston JH, Jensen DM, Mautner W, et al. Argon laser treatment of bleeding canine ulcers: limitations and guidelines for endoscopic use. Gastroenterology 1981;80: Jensen DM. Endoscopic control of gastrointestinal bleeding. In: Berk JE, ed. Developments in digestive diseases. Vol. 3. Philadelphia: Lea and Febiger, 1980: Machicado GA, Jensen DM, Tapia JI, et al. Treatment of bleeding canine duodenal and esophageal ulcers with argon laser and bipolar electrocoagulation. Gastroenterology 1981;81: Snedcor GW, Cochran WG. In: Statistical methods. 6th ed. Ames, Iowa: Iowa State University Press, 1967:20-6, Jensen DM, Machicado GA, Tapia JI, et al. Comparison of endoscopic multipolar probe and argon laser treatment of bleeding canine gastric ulcers (abstr). Gastrointest Endosc 1981;27: Johnston JH, Jensen DM, Mautner W. A comparison of bipolar electrocoagulation and argon laser photocoagulation with CO 2 in the treatment of bleeding canine gastric ulcers (abstr). Gastrointest Endosc 1978;24: Brunetaud JM, Biserte J, Charlier J, et al. Therapeutic applications of argon ion and Nd YAG lasers. In: Atsumi K, Nimsakul N, eds. Laser Tokyo '81. Sect 5. Tokyo: Japan Society for Laser Medicine. 1981: Dixon JA, Burt RW, Rotering RH, et al. Endoscopic argon laser photocoagulation of sessile colonic polyps (abstr). Gastrointest Endosc 1981;27: Jensen DM, Tapia JI, Machicado GA, et al. Comparison of electrocoagulation and heater probe for hemostasis in the canine colon (abstr). Gastrointest Endosc 1982;28: Waye JD. The postpolypectomy coagulation syndrome. Gastrointest Endosc 1981;27:184.