The circulation in portal hypertension.

While increased resistance to portal blood flow is a central feature of portal hypertension, the varied clinical presentations of this syndrome and their relation to the underlying divergent splanchnic hemodynamic patterns remain puzzling. Some patients present with massive ascites, repeated hemorrhage from ruptured esopha-geal varcies, and severe hypersplenism, whereas others with equivalent or even higher portal pressure readings are symptom-free. Nonetheless, each of these serious complications is generally conceded to be somehow a direct consequence of high portal pressure. Portal pressure, however, depends not only upon resistance to portal flow through the liver and through portasystemic collaterals bypassing the liver, but also upon the volume of splanchnic blood feeding into the portal system from the gastrointestinal tract, pancreas, and spleen. Thus, at any given moment, the level of portal pressure is the resultant of a whole set of rapidly changing interrelated flows and vascular resistances in the splenic, mesenteric, hepatic, and systemic circulations (1). Less than 30 yr ago, the "syndrome" of pulmonary hypertension appeared equally perplexing. Pulmonary artery pressure likewise depends upon a complex set of rapidly changing interrelated flows and vascular resistances albeit in the systemic and pulmonary circulations. But then the cardiac catheter was introduced into the heart and great vessels, and blood flow and vascular pressure measurements were made at key points in these two circulations. Today, a variety of forms of pulmonary hypertension are recognized which have their own distinctive clinical presentations and are defined by specific hemodynamic criteria (2). Unfortunately, the events in portal hypertension take place in the relatively inaccessible splanchnic bed, and methodology continues to be a major stumbling block in obtaining complete hemodynamic data. Direct measurement of individual organ blood flow usually must be made under artificial or less than ideal conditions (e.g., at laparotomy), and indirect quantitation of portasystemic shunted flow is at best a rough approximation (3). Thus, "catheterization data" (simultaneous measurements of flow and pressure) at key points in the splanchnic circuit (splenic, mesenteric, hepatic, and portal) in patients with portal hypertension comparable to "catheterization data" in patients with pulmonary hypertension are simply not available. The influence of splanchnic blood flow on portal dynamics is well illustrated in experimental right-sided congestive heart failure in the dog. Banding of the pulmonary artery alone or in combination with disruption of the tricuspid valve regularly produces marked systemic venous hypertension, massive hepatomegaly, engorge-'Supported in part by grants from the USPHS (HL …