DETERMINANTS OF MIDWALL CIRCUMFERENTIAL SEGMENTAL LENGTH OF THE CANINE VENTRICULAR SEPTUM AT END DIASTOLE

Citation
Sj. Dong et al., DETERMINANTS OF MIDWALL CIRCUMFERENTIAL SEGMENTAL LENGTH OF THE CANINE VENTRICULAR SEPTUM AT END DIASTOLE, The American journal of physiology, 265(6), 1993, pp. 80002057-80002065
Citations number
23
Categorie Soggetti
Physiology
ISSN journal
00029513
Volume
265
Issue
6
Year of publication
1993
Part
2
Pages
80002057 - 80002065
Database
ISI
SICI code
0002-9513(1993)265:6<80002057:DOMCSL>2.0.ZU;2-A
Abstract
The ventricular septum attaches to the insertion points of the ventric ular free walls, separates, and is shared by both the left (LV) and ri ght (RV) ventricles. Changes in the transseptal pressure gradient (P(T S), the difference between LV and RV pressures) will change the stress in the septum and, therefore, the length of the septal segment (L(s)) . However, since most of the septal myocardial fibers are continuous w ith those of LV free wall and the septum appears to be an integral par t of LV, one might also expect that if the LV transmural pressure (P(l vtm); the difference between LV and pericardial pressure) is raised, L (s) would increase even in the absence of any change in P(TS). Therefo re, we hypothesized that at end diastole, L(s) depends un both P(lvtm) and P(TS). To test this hypothesis, we measured L(s) (sonomicrometry) , LV and RV pressures (micromanometers), and pericardial pressure (fla t liquid-containing balloon) in seven anesthetized open-chest dogs. P( lvtm) was increased through volume loading, whereas P(TS) was maintain ed constant at 10, 5, 0, -5, -10, and -15 mmHg by adjusting the degree of constriction of the pulmonary artery or aorta. These procedures we re performed first with pericardium reapproximated and then after the pericardium had been opened widely. At each controlled P(TS) level, L( s) increased linearly with the increase in P(lvtm). Both the slope and the intercept of this L(s)-P(lvtm) relation were affected by P(TS) in a nonlinear fashion. With use of a model in which L(s) is a linear fu nction of P(lvtm) and a quadratic function of P(TS), there was an exce llent correlation between predicted and observed L(s) (R2 = 0.97 +/- 0 .02, mean +/- SD), providing a clearer picture of the dependence of L( s) on P(lvtm) and P(TS). We conclude that L(s) depends both on P(lvtm) , as indicated by the linear relation between L(s) and P(lvtm), and on P(TS), as indicated by the curvilinear relation between L(s) and P(TS ).