18.1 Systemic Circulation

Historically, as early as 1886 De Jager ascribed the rise in carotid and central venous pressures in response to aortic occlusion in anesthetized dogs to the shift of blood from arterial to venous compartments, caused by “elastic energy accumulated in the coats of arteries.” De Jager also noted that occlusion of the inferior vena cava results in a drop of arterial pressure to about half of normal values and of the central venous pressure to zero. He thought it particularly significant that after a temporary release and re-occlusion of the IVC, there was a rise in CVP which he interpreted as venous stasis due to weakening and inadequate diastolic suction of the right ventricle [4].

The findings showed that in spite of profound changes in flows and arterial pressures affected by occlusion of the great vessels, there was relatively little variation in the corresponding RAPs [7]. In order to confirm that the heart indeed “reacts to all changes in the peripheral circulation by maintaining the central venous pressure constant”—as corroborated by Starling and coworkers only a few years earlier on the heart-lung preparation—Barcroft repeated the vessel-occlusion experiments in animals whose heart had been replaced by a mechanical pump [8]. The results indeed showed that, during aortic occlusion, the pump flows had to be increased by up to 60% above baseline to maintain constant RAPs. The pump flows during occlusion of other great vessels, on the other hand, were comparable to baseline flows before occlusion. In view of these findings, Barcroft made the following comment: “The striking thing is that the healthy heart maintained the venous pressure practically constant in the face of flow changes from 990 to 125 c.c. per min., and arterial blood pressure changes of 140 to 32 mm of mercury. Broadly speaking this suggested that, in these experiments, the heart maintained the venous pressure at zero, pumping away what returned to it irrespective of arterial pressure” [8].

Barcroft therefore concluded that the increase in CO following aortic occlusion results from factors related to the peripheral circulation, citing the model proposed by Krogh by which the blood is transferred from a large, highly compliant hepatosplenic reservoir into the low-compliance, upper body compartment with concomitant increase in venous return [9].

These experiments were repeated almost half a century later on anesthetized, atropinized dogs by Stokland and coworkers [10]. In addition to implanting snares around the descending thoracic aorta and IVC, the internal mammary artery and the azygos vein were ligated to prevent major vascular communication between the upper and lower parts of the body. Left ventricular dimension, i.e., the myocardial chord length (MCL), was monitored via a pair of miniaturized ultrasonic probes implanted on the anterior aspect of the LV wall. The results showed that occlusion of the descending aorta results in doubling of SVC flows, while the flow through IVC stabilized at 10% of control. Unlike in Barcroft’s experiments where aortic occlusion resulted in a marked increase in CO, the combined flow through SVC and IVC reached only about 80% of control value. The left ventricular systolic pressure (LVP) increased from 102 ± 7 to 188 ± 11 mmHg and LVEDP rose from 3 to 8 mmHg, the latter corresponding to an increase in MCL by 10%.

Simultaneous occlusion of the descending aorta and IVC resulted in cessation of flow in the IVC and caused MCL to diminish by an average of 38%, while LVP and flows in the SVC remained the same as during the aortic occlusion. Remarkably, after reaching a “steady-state” (usually after 2–3 min), the systolic LVPs and SVC flows during combined aortic and SVC occlusion were virtually the same as during control (pre-occlusion) state.

A heightened “sensitivity to blood volume” during combined aortic and SVC occlusion was demonstrated by infusion of 50 mL aliquots of blood into the jugular vein to the total cumulative volume which restored myocardial dimensions and SVC flows to pre-IVC occlusion levels. Subsequent withdrawal of blood from the upper body compartment to pre-infusion control values suggested a poor correlation between hemodynamics and perturbation in blood volumes. They proposed that about 7–8% of total blood volume was transferred from the lower to upper body vascular compartment, presumably via venous communications in the spinal canal. On the basis of their data Stokland and coworkers concluded that, according to this model, redistribution of blood volume due to “vascular collapse” in the lower body—and subsequent activation of the Frank-Starling mechanism via increase in LV end-diastolic volume—“completely accounts for hemodynamic response” during descending thoracic aortic occlusion [10].

Unlike Barcroft’s and several other reports that demonstrated an increase in CO following the occlusion of the thoracic aorta, the study by Stokland et al. showed an actual decrease in CO. The question was therefore posed as to what factors could account for such a discrepancy in COs between the two studies? Did large LV pressures and distension of the LV cause a weakening of the LV, leading to decreased output, or, was there a significant, but unrecognized backing-up of blood before the right heart, since no values of RAPs were reported in Stokland’s study? To clarify the issue, Stene et al. repeated Barcroft’s aortic occlusion, right heart bypass experiment [11]. In place of the total cardiopulmonary bypass, where the “heart-lung preparation” of another dog served as a source of oxygenated blood, Stene et al. used a mechanical right-heart bypass (RHB). The pump flows were maintained at rates slightly above venous return to keep consistent drainage of the RA (Fig. 18.1). Estimates of the mean systemic pressure (MSP) were obtained by temporarily stopping the pump and measuring the plateau pressures for 7–9 s. The latter served as an index of overall compliance changes in the peripheral circulation during aortic occlusion (OA).