rectal bleeding, therefore, in a patient who has under­gone EIS may require that the sclerotherapy be per­formed by means of flexible sigmoidoscopy.

Those patients whose livers permit survival beyonda course of EIS appear to be at risk for complicationsat sites other than the esophagus. We will have tonote carefully the number of local and remote compli­cations that can follow EIS and weigh them in thebalance against the putative benefits. Based on theinformation to date, it would appear that EIS re­stricted to the distal esophagus and done with a shortneedle and, perhaps, more paravariceal than intravar­iceal injections of the sclerosant may reduce the num­ber of remote complications that plague the scleroth­erapist.

Bernard M. Schuman, MD

REFERENCES1. Gottfried EB, Goldberg HJ. Mucosal bridge of the distal esoph­

agus after esophageal variceal sclerotherapy. Gastrointest En­dose 1985;31:267-9.

2. Agha FP. The esophagus after endoscopic injection sclerother­apy: acute and chronic changes. Radiology 1984;153:37-42.

3. Evans DMD, Jones DB, Cleary BK, et al. Oesophageal varicestreated by sclerotherapy: a histopathological study. Gut1982;23:615-20.

4. Saks BJ, Kilby AE, Dietrich PA, et al. Pleural and mediastinalchanges following endoscopic injection sclerotherapy of esoph­ageal varices. Radiology 1983;149:639-42.

5. McGrew W, Goodin J, Stuck W. Fatal complication of endo­scopic sclerotherapy: Serratia marcescens bacteremia with de­layed esophageal perforation. Gastrointest Endosc 1985;31:329­31.

6. Ayres SJ, Goff JS, Warren GH. Endoscopic sclerotherapy forbleeding esophageal varices: effects and complications. AnnIntern Med 1983;98:900-3.

7. Cohen FL, Koerner RS, Taub SJ. Solitary brain abscess follow­ing endoscopic injection sclerosis of esophageal varices. Gas­trointest Endosc 1985;31:331-33.

8. Cohen LB, Korsten MA, Scherl EJ, et al. Bacteremia afterendoscopic injection sclerosis. Gastrointest Endosc 1983;29:198-200.

9. Brayko eM, Kozarek RA, Sanowski RA, et al. Bacteremiaduring esophageal variceal sclerotherapy: its cause and preven­tion. Gastrointest Endosc 1985;31:10-2.

10. Snady H, Korsten MA, Waye JD. The relationship of bacter­emia to the length of injection needle in endoscopic varicealsclerotherapy. Gastrointest Endosc 1985;31:243-6.

11. Grobe JL, Kozarek RA, Sanowski RA, et al. Venography duringendoscopic injection sclerotherapy of esophageal varices. Gas­trointest Endosc 1984;30:6-8.

12. Monroe P, Morrow CF, Millen JE, et al. Acute respiratoryfailure after sodium morrhuate esophageal sclerotherapy. Gas­troenterology 1983;85:693-9.

13. Sukigara M, Omoto R, Miyamae T. Systemic dissemination ofethanolamine oleate after injection sclerotherapy for esophagealvarices. Arch Surg 1985;120:833-6.

14. Seidman E, Weber AM, Morin CL, et al. Spinal cord paralysisfollowing sclerotherapy for esophageal varices. Hepatology1984;4:950-4.

15. Foutch PG, Sivak MV. Colonic variceal hemorrhage after en­doscopic injection sclerosis of esophageal varices: a report ofthree cases. Am J GastroenteroI1984;79:756-60.

VOLUME 31, NO.5, 1985

Letters to the Ed itor

Laser therapy is cost-effective

To the Editor:

The ultimate realization of "therapeutic endoscopy" is inthe application of laser technology via the endoscope for thetreatment of gastrointestinal disease. Critics of laser appli­cations in gastroenterology have had three major concernsabout its use: (1) the technology may not offer an improve­ment over more conventional methods of treatment; (2) theequipment is not mobile; and (3) it is too expensive.

In the field of gastroenterology the laser was originallyused for the control of upper gastrointestinal (UGI) hemor­rhage, and initial reportsl

-4 (uncontrolled studies) suggested

that the laser could affect hemostasis in approximately 90%of unselected cases. Most patients with UGI hemorrhagestop bleeding spontaneously without any intervention, andtherefore, it has been argued that the laser in these uncon­trolled series may not have been of any significant benefit.The early controlled trials,5.6 that were conducted suggestedthat laser therapy was of no benefit over conventional meth­ods. The more recent controlled trials, however, report amore favorable result.7

-9 Bown lO has recently concluded that

the controlled trials that included only endoscopically ac­cessible targets and categorized the patients by specificbleeding points demonstrated the use of the laser to beadvantageous. When the results of the controlled studieswere combined, Bown found that for a high risk subset ofUGI bleeders (bleeding and nonbleeding visible vessels)accessible to the endoscope, the control group faired lesswell than the laser-treated group (64% of controls rebled vs.34% in the argon trials vs. 16% in Nd:YAG trials). Prelim­inary data from a controlled study by Swain et al. l1 suggestthat the Nd:YAG laser-treated group had a decreased bleed­ing rate and need for emergency surgery with a resultantdecrease in mortality. Thus, data are now emerging whichconfirm the impression that many investigators have had,namely, that laser therapy in the treatment of gastrointes­tinal hemorrhage is efficacious.

Many of the patients with upper gastrointestinal hemor­rhage who would benefit from laser therapy are frequentlycared for in the intensive care unit. Critics of laser therapyargue that any endoscopic method of treatment for UGIbleeding must be portable so it can be taken to the patient'sbedside in the intensive care unit rather than moving thepatient to a designated treatment area. Many potential lasercandidates have multisystem disease, are medically unstable,and are rightfully located in the intensive care unit. Therisk of aspiration in this group of patients during an emer­gency endoscopic procedure is high, and an anesthesiologistmay be needed to intubate and monitor such a patient. Infact, the endoscopic therapy of a UGI bleeder should beadministered in a designated area, such as a laser treatmentroom, where all endoscopic tools are readily obtainable andwhere trained personnel are clustered.

Although the current cost to purchase and install aNd:YAG laser is between 95 and 100 thousand dollars, wenow have an effective endoscopic method of controlling UGIhemorrhage, and the wider indications for its use both within

349

the field of gastroenterology and without will make its use acost-effective endeavor.

Fred Kessler, MDDivision of Gastroenterology

Mt. Sinai Medical CenterCleveland, Ohio

REFERENCES1. Dwyer R, Yellin A, Craig J, Cherlow J, Bass M. Gastric hemo­

stasis by laser phototherapy in man. J Am Med Assoc1976;236:1383-4.

2. Nath G, Gurish W, Kiefhaber P. First laser endoscopy viafiberoptic transmission system. Endoscopy 1973;5:208-13.

3. Fruhmorgen P, Reidenbach F, Bodem R, Kaduk B, Demling L.Experimental examinations on laser endoscopy. Endoscopy1974;6:116-22.

4. Kiefhaber P. International experience with lasers for gastroin­testinal bleeding. Proceedings of the International Laser Con­gress, 1979.

5. Ihre T, Johnson C, Seligsson 0, Torngren S. Endoscopic YAGlaser treatment in massive UGI bleeding. Scand J Gastroenterol1981;16:633-40.

6. Escourrou J. Nd:YAG laser therapy for acute gastrointestinalhemorrhage. International Society for Laser Surgery, 4th Con­gress, Tokyo, 1981.

7. Rutgeerts P, Vantrappen G, Broeckaert L, et al. Controlledtrial of YAG laser treatment of upper digestive hemorrhage.Gastroenterology 1982;83:410-6.

8. Swain C, Storey D, Northfield T, Bown S, Kirkham J, SalmonP. Controlled trial of argon laser photocoagulation in bleedingpeptic ulcers. Lancet 1981;2:1313-6.

9. Jensen D, Machicado G, Tapia J. Berlin D, Mautner W. En­doscopic argon laser photocoagulation of patients with severegastrointestinal bleeding. Gastrointest Endosc 1982;28:151.

10. Bown S. Review of controlled trials of laser therapy for hem­orrhage from peptic ulcers. Lasers Surg Med 1983;3:121.

11. Swain C, Bown S, Salmon P, Kirkham J, Northfield T. Con­trolled trial of Nd:YAG laser photocoagulation for bleedingpeptic ulcers. Lasers Surg Moo 1983;3:111.

A method for maintaining patency of anintracholedochal stent

To the Editor:

With the increased use of endoprosthesis for bile ductobstructions, the occlusion of the stents and the necessityto replace them or pass a second stent become problems.1

-5

We were presented with a problem of an obstructed stentin a patient who had had a transhepatically placed stent. Byutilizing the available guide wire via a side-viewing duoden­oscope, we manipulated it through the existing holes of thestent and were able to flush out the sediment. This wasapparently successful since the patient had resolution of hisjaundice.

I suggest this as a simple and inexpensive method formaintaining patency of an occluded stent.

J. Rattan, MDIchilov HospitalTel Aviv, Israel

REFERENCES1. Huibregtse K, Haverkamp RJ, Tytgat GN. Transpapillary po­

sitioning of a large 3 mm biliary endoprosthesis. Endoscopy1981;13:217-9.

2. Zinman DS, Clement AR. Endoscopic stents and drains in the

350

management of pancreatic and bile duct obstruction. Surg ClinNorth Am 1982;62:837-44.

3. Kerlan RK, Pogany AC, Goldberg HI, Ring EJ. Percutaneousbiliary drainage in the management of cholangiocarcinoma. AmJ Roentgenol 1983;141:1295-8.

4. Isoley JS, Dick R, George P, Kirk RM, Hobbs KEF, SherlockS. Percutaneous transhepatic endoprosthesis for bile duct ab­straction. Gastroenterology 1984;86:905-9.

5. Camener JL, Gagler BW, Zvidemo GR. The use of Silastictranshepatic stents in benign and malignant biliary stricture.Ann Surg 1978;188:552-61.

Intraperitoneal hemorrhage followingfiberoptic gastroscopy

To the Editor:

Splenic injury may complicate fiberoptic colonoscopyl buthas not been reported to follow upper gastrointestinal (UGI)endoscopy. We report two cases separated by 8 years inwhich intraperitoneal hemorrhage occurred as a result ofUGI endoscopy. Both cases were women in their eighthdecade undergoing emergency endoscopy for severe UGIhemorrhage; wide-bore endoscopes were used on both occa­sions. In both patients the bleeding lesions were high on thelesser curvature of the stomach, necessitating inversion ma­neuvers to identify the source of bleeding.

In both instances laparotomy was undertaken to controlthe gastrointestinal hemorrhage. On opening the abdomen,fresh blood and clot were observed overlying the gastricfundus and gastric surface of the spleen, the bleeding arisingfrom a torn, short gastric vein. In one case hemorrhage wasarrested by ligature; in the other, splenectomy was per­formed for continuing peroperative bleeding.

We draw attention to this unreported complication whichmay occur during UGI endoscopy when performing inversionmaneuvers with a wide-bore endoscope. Elderly patientswith frail tissues may be particularly prone to sustain theseinjuries which may result in unexplained and continuinghemorrhage.

T.C. B. Dehn, MS, FRCSE.C.G. Lee, MCh, FRCS

John Radcliffe HospitalHeadington, Oxford, England

REFERENCE1. Meyers MA. Complications of gastrointestinal endoscopy. Mt

Sinai J Moo 1981;48:191-200.

Pancreas divisum may be onlycoincidental

To the Editor:

Pancreas divisum (PD) is defined as failure of the dorsaland ventral buds of the pancreas to fuse. As a result, theadult pancreatic ductal system remains as it was before theeighth week of fetal life. In PD the duct of Wirsung drainsonly the ventral pancreas and a small segment of the pos­terior and inferior part of the head of the gland through themajor papilla. The duct of Santorini becomes the main

GASTROINTESTINAL ENDOSCOPY