Endoscopic Ligation of Esophageal Varices

Endoscopic Sclerotherapy

Endoscopic sclerotherapy for variceal bleeding, although described in 1939 [2], was rediscovered in the late 1970s. Soon it became evident that ongoing hemorrhage from a varix could be stopped by the injection of various chemical agents, that this form of intervention decreased the volume of blood transfused, and that additional injection sessions decreased the number of episodes of recurrent bleeding. Sclerotherapy may sometimes be equal or perhaps even superior to portacaval or selective splenorenal shunt surgery in terms of survival and the preservation of hepatic function [3-5]. The results of one trial [6] indicate that endoscopic sclerotherapy decreases the mortality for patients with cirrhosis and variceal bleeding; however, another trial [7] did not find a decrease in mortality. A meta-analysis of seven trials revealed that overall survival for patients with variceal bleeding was improved by sclerotherapy [8]. Thus, it appeared that a definitive treatment for this major complication of portal hypertension had been found. This is not entirely true, but sclerotherapy has become a major technique used in the management of variceal hemorrhage, and it has benefitted patients greatly.

However, endoscopic sclerotherapy has limitations. The injection of a noxious agent into any segment of the human vascular system raises the possibility of untoward effects in organs at a distance from the primary site of injection. The potential risk for such complications would depend to an extent on variceal anatomy. Unfortunately, only a few studies [9-11] exist of normal esophageal venous and variceal anatomy and of the nature of blood flow in these structures. Those studies that are available suggest that an injected sclerosing agent may reach virtually any organ. This has been substantiated by many reports of untoward events after sclerotherapy, some with serious consequences, in organ systems other than the esophagus. In particular, an appreciable risk exists for sepsis after sclerotherapy. Although these systemic complications are relatively rare, they represent an important shortcoming for endoscopic sclerotherapy.

Endoscopic sclerotherapy is less than ideal as a therapeutic measure because tissue injury is fundamental to its mechanisms of action. Once injected, the action of the chemical agent is, to a certain extent, uncontrollable and unpredictable. This accounts for various local complications, including perforation, stricture formation, and ulceration. The destructive basis of sclerotherapy is emphasized by the fact that ulcers at injection sites are expected to occur in all patients. The difference between an ulcer as a desirable consequence of therapy and one that represents a true complication is merely a matter of degree and the clinical behavior of the lesion. Unfortunately, tissue damage probably also occurs with variceal ligation, although to a lesser degree than with sclerotherapy.

Endoscopic Variceal Ligation

Endoscopic variceal ligation of esophageal varices is based on a technique developed in the 1950s for band ligation of hemorrhoids. As originally described by Stiegmann and colleagues [12], use of this technique for esophageal varices involves the mechanical ligation and strangulation of variceal channels by application of small, elastic O rings. Several rings must be applied at various sites; because each ring must be loaded individually on the end of the endoscope, an overtube is used to allow rapid and repeated passage of the endoscope. Actual application of the ring to a varix is by means of an ingenious device attached to the distal end of a standard endoscope. Although a sequence of steps is required, most endoscopists find the technique for band ligation to be less demanding than that for sclerotherapy.

Sclerotherapy Compared with Ligation

The work of Stiegmann and colleagues [13-17] on endoscopic variceal ligation may serve as a model for the development of new technology. Single-arm trials by this group have shown that variceal ligation is comparable to endoscopic sclerotherapy in terms of control of variceal hemorrhage and the prevention of recurrent bleeding, but that variceal ligation has substantially less morbidity.

Although of good quality and credible, this work must be corroborated by other investigators. In the randomized, controlled trial reported by Laine and colleagues [1], there was no difference between variceal ligation and sclerotherapy with respect to recurrent variceal bleeding, volume of blood transfused, length of hospitalization, and survival. Compared with patients who had sclerotherapy, however, those treated with ligation had higher Child-Pugh scores as well as varices of a more advanced endoscopic grade. In the only other randomized, controlled trial of ligation compared with sclerotherapy, Stiegmann and coworkers [18] showed a survival advantage for variceal ligation. In this study, patients were evenly matched for Child-Pugh score and variceal grade at endoscopy.

In the study of Laine and colleagues [1], variceal ligation was superior to sclerotherapy with respect to local esophageal complications, especially stricture formation. However, the percentage of patients (33%) who developed an esophageal stricture as a result of sclerotherapy is remarkably high in this study. This is difficult to explain but may be due to the use of a relatively high concentration of the sclerosing agent. It is probable that the use of a less potent solution would have resulted in fewer strictures, although this might not have prevented recurrent bleeding and might not have decreased the number of treatment sessions required for variceal eradication. The technique of variceal ligation would seem to be inherently safe, but Laine and colleagues [1] did encounter one patient who had an esophageal injury that appeared to be due to placement of the overtube. Although the risk for a systemic complication after variceal ligation should be negligible, bacteremia has been reported [19]. Variceal ligation also required statistically fewer treatment sessions for eradication of varices.

Although the management of cirrhosis, of portal hypertension, and of variceal bleeding was once constrained by a lack of effective methods, the range of options now available has become problematic. Those who do endoscopic examination and therapy must consider whether endoscopic variceal ligation is substantially better than variceal sclerotherapy and consider that perhaps the time as come to abandon the familiar, effective, albeit imperfect treatment method of variceal sclerotherapy.

Conclusion

Experience with variceal sclerotherapy, including data and information derived from many published trials, is substantially greater than that with variceal ligation. Only two randomized, controlled studies of ligation exist, including the one by Laine and colleagues [1]. The follow-up interval for patients having ligation has been comparatively short, and thus the long-term results of ligation are relatively unknown. The technique of variceal ligation is not difficult and should be relatively safe, but the full range of potential and actual complication may not be known. For example, the use of the overtube might be a hazard. However, variceal band ligation must be regarded as a promising technique, one that offers results that are at least equal to sclerotherapy but with fewer serious complications. Endoscopic ligation will probably assume an increasing role in the management of variceal bleeding.