Effect of Prolonged Selective Intramesenteric Arterial Vasodilator Therapy on Intestinal Viability After Acute Segmental Mesenteric Vascular Occlusion

Abstract

Objective

To evaluate the effect of selective intramesenteric artery vasodilator infusion on intestinal viability in a rat model of acute segmental mesenteric vascular occlusion.

Summary Background Data

Although intramesenteric arterial vasodilator infusion may be an effective treatment for nonocclusive mesenteric ischemia, it has also been advocated to increase collateral blood flow after mesenteric vascular occlusion. However, the authors have previously found that intraarterial vasodilators actually reduce collateral blood flow acutely, by preferentially dilating the vasculature of adjacent, nonischemic mesenteric vascular beds, a phenomenon well established in other organs.

Methods

A segment of rat ileum was acutely devascularized, with blood flow provided only by collateral arterial vessels from adjacent, nonischemic bowel. Papaverine (30 or 40 μg/kg/min), isoproterenol (0.06 μg/kg/min), norepinephrine (0.1 or 0.2 μg/kg/min), or vehicle saline was continuously infused into the cranial (superior) mesenteric artery for 48 hours. Viability was then assessed using previously established, objective gross and microscopic criteria.

Results

Although papaverine increased total mesenteric blood flow in normally vascularized rats, it not only failed to improve but actually significantly reduced the length of the devascularized segment maintained viable by collateral blood flow after 48 hours. Isoproterenol had a similar effect. Norepinephrine infusion decreased both normal mesenteric blood flow and viable segment length.

Conclusions

These findings suggest that intraarterial vasodilator therapy fails to improve intestinal viability after segmental mesenteric vascular occlusion.

The intraarterial infusion of a vasodilator may be of benefit in the treatment of nonocclusive mesenteric ischemia, to overcome primary splanchnic vasoconstriction. ^1–3 Such vasodilator therapy has also been advocated for the treatment of mesenteric vascular occlusion, ^4,5 but these clinical reports are both anecdotal and uncontrolled with respect to vasodilator administration. We have previously characterized the physiology of mesenteric collateral blood flow, and its hemodynamic determinants, in a canine model of acute segmental mesenteric vascular occlusion. ⁶ Collateral flow from a nonischemic segment to an adjacent ischemic segment was found to be substantial and was redirected entirely by vasodilatation within the ischemic bed that produced a differential resistance that facilitated collateral nutrient blood flow from the nonischemic segment into the ischemic one. The only vasoactive determinant of this flow was the vasodilatation within the ischemic bed; all other hemodynamic factors, including those in the adjacent, nonischemic segment, played only a passive role. The collateral conduits themselves, primarily the marginal arteries, were not vasoactively responsive to ischemia nor a determinant of collateral flow. These findings indicated that collateral flow after acute segmental mesenteric occlusion was determined primarily by the relative vascular resistances of the ischemic and the adjacent, nonischemic vascular beds.

In a subsequent study, the acute infusion of pharmacologic doses of a vasodilator, either isoproterenol or papaverine, had only a marginal effect on vascular resistance in the ischemic bed but substantially dilated the nonischemic bed, with a consequent increase in blood flow to the nonischemic segment. ⁷ This preferential dilation of the nonischemic bed decreased perfusion pressure to both beds, distal to the site of vascular occlusion, and thereby actually reduced collateral blood flow into the ischemic segment to a small but significant degree. These hemodynamic changes are characteristic of a vascular steal phenomenon.

Findings from these previous studies suggested an apparently paradoxical hemodynamic effect of intraarterial vasodilator infusion on collateral blood flow during acute segmental mesenteric vascular occlusion. However, they do not preclude the possibility that prolonged vasodilator infusion might reverse a superimposed splanchnic vasospasm that has been postulated to appear several hours after the onset of occlusion. ² Because the clinically meaningful end point of mesenteric vascular occlusion is consequent bowel viability, the present study was designed to evaluate directly the effect of prolonged intraarterial infusion of vasoactive (vasodilator and vasoconstrictor) agents on intestinal viability after acute segmental mesenteric vascular occlusion.

METHODS

Experimental Preparation

Male Sprague-Dawley rats (370–410 g) were fasted overnight with access to water and anesthetized with intraperitoneal chloral hydrate (275 mg/kg). Core body temperature, monitored with a rectal probe, was maintained at 36°C to 37.5°C with a heating pad. The root of the cranial (superior) mesenteric artery (SMA) was exposed through a midline celiotomy. Using an operating microscope, the caudal (inferior) pancreaticoduodenal artery was isolated at its origin from the SMA. For the infusion of a vasoactive agent (or a vehicle control), a polyethylene catheter (PE-10) was inserted retrogradely into the cephalic branch of the inferior pancreaticoduodenal artery to its origin from the SMA, such that the tip of the catheter extended to but did not occlude the SMA (Fig. 1). All other branches of the inferior pancreaticoduodenal artery were ligated. The distal end of this catheter was tunneled subcutaneously and brought out through the back of the neck to facilitate the continuous intraarterial infusion of agents by infusion pump. Animals were housed, cared for, and studied in accordance with principles of standard humane procedures for animal experimentation by a protocol approved by the Animal Care and Use Committee of the Johns Hopkins University School of Medicine.

Figure 1. Rat small intestine with segmental arterial devascularization. The proximal end of the 33 ± 1-cm devascularized segment is shown, with the mesentery incised at the proximal point of devascularization. The distal end of the devascularized segment was similarly treated. The stippling in the middle of the segment represents nonviable bowel.

Acute Hemodynamic Effects of Vasoactive Drug Infusion

In preliminary experiments, we first determined the dose–response relationship of SMA blood flow to the intraarterial infusion of the vasoactive agents to be infused in the subsequent experiments (Fig. 2).

Figure 2. Dose–response curve of superior mesenteric artery flow (circle) and systemic mean arterial pressure (triangle) responses to the continuous infusion of papaverine (A), isoproterenol (B), or norepinephrine (C) into the inferior pancreaticoduodenal artery, without segmental devascularization. Arrows indicate doses chosen for subsequent segmental viability experiments.

Papaverine was infused into the root of the SMA at doses varying from 0.4 to 400 μg/kg/min in normal saline (n = 3). SMA blood flow was measured continuously with a pulsed Doppler flowmeter with the flow probe applied to the SMA at its origin from the aorta. By means of a femoral vein, Ringer’s lactate solution was given at 12 mL/kg/h, and a femoral artery was cannulated for the measurement of systemic arterial pressure. Both SMA blood flow and arterial pressure were monitored continuously on a physiologic recorder. In similar fashion, dose–response relationships were obtained for isoproterenol, from 0.01 to 10 μg/kg/min (n = 3), and for norepinephrine, from 0.02 to 4 μg/kg/min (n = 3).

Prolonged Acute Segmental Mesenteric Devascularization

To assess the effect of the long-term selective infusion of vasoactive agents on segmental intestinal viability, a 33 ± 1-cm segment of ileum (measuring retrogradely from the cecum) was devascularized by ligation and division of each of the mesenteric arterial branches that directly supplied this segment. The mesenteric venous branches were preserved to simulate the clinical anatomical situation after embolic occlusion of a mesenteric arterial branch. The mesentery was divided at both the proximal and distal ends of the devascularized segment, but the marginal vessels at each end were preserved to permit collateral blood flow from the adjacent, vascularized intestine to the devascularized segment (see Fig. 1).

A separate polyethylene catheter (PE-10) was tunneled subcutaneously from the peritoneal cavity and also brought out through the back of the neck to facilitate subsequent intraperitoneal administration of antibiotics. After suture closure of the abdomen, the intraarterial infusion was changed from Ringer’s lactate to normal saline (20 mL/kg/h) for 1 hour. The infusion solution was then replaced with a vasodilator (papaverine or isoproterenol), a vasoconstrictor (norepinephrine), or vehicle (saline) control. The infusion was then maintained continuously for 48 hours. During this time, the rats were supported by subcutaneous injections of 30 mL of 5% dextrose-Ringer’s lactate solution containing 1.5 mEq sodium bicarbonate, every 8 hours. Intramuscular ampicillin (50 mg/kg), gentamicin (10 mg/kg), and clindamycin (80 mg/kg), doses previously determined to optimize serum drug levels, ⁸ were given just before the surgical preparation and repeatedly thereafter through the peritoneal catheter every 8 hours. This experimental approach was used here and in similar previous preparations ⁹ to optimize systemic variables (hemodynamics, urine output, and metabolic acid–base balance) and to permit survival of the rats for 48 hours, despite the necessary presence of a segment of nonviable intestine in the peritoneal cavity. This served to mimic the corresponding clinical situation and to allow direct comparison of effects on bowel viability with a minimal influence on rat viability.

Effect of Prolonged Vasoactive Drug Infusion on Intestinal Viability

In the control group (n = 16), normal saline was infused into the root of the SMA at a rate of 1 mL/h for 48 hours. In the papaverine group receiving 30 μg/kg/min (n = 9), papaverine was infused into the root of the SMA at a rate of 1.8 mg/kg in 1 mL/h for 48 hours. In the papaverine group receiving 40 μg/kg/min (n = 8), papaverine was infused into the root of the SMA at a rate of 2.4 mg/kg in 1 mL/h for 48 hours. These doses of papaverine were chosen from the dose–response curve (see Fig. 2) to produce significant vasodilation in the bed of the SMA, but without systemic hemodynamic effects.

In a separate control group of rats (n = 4), to determine possible systemic hemodynamic or toxic effects of a 48-hour papaverine infusion, the inferior pancreaticoduodenal artery was cannulated and papaverine was infused intraarterially at 40 μg/kg/min for 48 hours. In these rats, no intestine was devascularized. At 48 hours, the rats were anesthetized and systemic arterial pressures were measured as described above.

In the isoproterenol group (n = 8), 0.06 μg/kg/min isoproterenol was infused into the root of the SMA at a rate of 3.6 μg/kg in 1 mL/h for 48 hours.

In the norepinephrine group, 0.1 μg/kg/min (n = 5) norepinephrine was infused into the root of the SMA at a rate of 6 μg/kg in 1 mL/h for 48 hours. In the norepinephrine group 0.2 μg/kg/min (n = 5), norepinephrine was infused into the root of the SMA at a rate of 12 μg/kg in 1 mL/h for 48 hours.

Assessment of Intestinal Viability

Forty-eight hours after devascularization, the rats were reanesthetized with intraperitoneal ketamine (60 mg/kg). A femoral artery was cannulated (PE-60) and systemic arterial pressure was measured with a pressure transducer and monitored on a physiologic recorder. In each rat, the devascularized segment was exposed through the midline incision. The correct intravascular placement of the intraarterial catheter was confirmed by observation of the intestine under ultraviolet (360 nm) illumination after the bolus injection of sodium fluorescein (20 mg in 0.2 mL saline) through this catheter. The small bowel was then removed surgically and the rat was euthanized with intravascular pentobarbital. Evaluation of the bowel segment, including the determination of its viable length, was conducted by an observer unaware of the drug infused. The entire devascularized segment, containing both nonviable (central) and viable (at either end) areas, was laid out in a straight line, stretched gently to overcome muscular spasm, and then allowed to return to a tension-free length. At each end of the segment, the point of transition from the viable to the nonviable, necrotic intestine was initially estimated by gross inspection of color and tissue tone and marked with a suture at the antimesenteric border. The overall length of the devascularized segment, the length of the middle, necrotic portion, and the lengths of the viable intestine at each end of the segment were measured to the nearest 0.5 cm (see Fig. 1).

The devascularized segment was then fixed in 10% buffered formalin, and representative samples were taken from the viable and nonviable portions. Longitudinal sections were also taken across the zone of transition from viable to nonviable intestine, along the antimesenteric border, and at each end of the nonviable segment. These specimens were dehydrated in alcohol, embedded in paraffin, sectioned at 8 to 10 μm, and stained with hematoxylin and eosin. Each segment was examined microscopically, again by an independent, masked observer, to verify the point of transition from viable to nonviable intestine. If any difference was found between the point of transition determined by gross inspection and the point determined by microscopic examination, this difference was used to modify the measurement obtained at the initial estimation. The lengths of viable and nonviable intestine were thereby determined grossly but confirmed by histologic examination. Histologic criteria for small intestinal viability at 48 hours are definitive, as previously described. ⁹

Analysis of Data

Lengths of viable intestine within the devascularized segment, beyond the point of devascularization, were expressed in cm (means ± standard error). Total viable length within the devascularized segment was expressed as a percentage of the devascularized length. All comparisons between treatment groups were evaluated for statistical significance by two-tailed Mann-Whitney rank-sum analysis. P ≤ .05 was considered significant.

RESULTS

Acute Effects of Vasoactive Drug Infusion

Over the range of 0.4 to 400 μg/kg/min papaverine, SMA flow increased to 185% of the baseline value with no significant effect on systemic blood pressure. For subsequent prolonged infusion, rates of 30 and 40 μg/kg/min were chosen to increase SMA blood flow 18% and 20%, respectively, over baseline.

Over the range of 0.01 to 1.0 μg/kg/min isoproterenol, SMA flow increased to 160% of the baseline value with no significant effect of systemic blood pressure. Higher infusion rates, to 10 μg/kg/min, resulted in decreases in both SMA flow and systemic pressure. For subsequent prolonged infusion, 0.06 μg/kg/min was chosen to increase SMA blood flow 20% over baseline without systemic hemodynamic effect.

Infusion of 0.02 to 4.0 μg/kg/min norepinephrine decreased SMA blood flow to 57% of the baseline value. Systemic arterial pressure increased over this range to 140% of baseline. For subsequent prolonged infusion, 0.1 and 0.2 μg/kg/min were chosen, which decreased SMA blood flow to 92% and 87% of baseline flow, respectively.

Effects of Prolonged Vasoactive Drug Infusion

All rats survived the 48-hour infusion period with the systemic support provided. Two rats receiving saline (control) were excluded, one because of incidental cecal torsion and necrosis, and the other because of excessive mesenteric devascularization (technical error). One rat receiving isoproterenol was excluded because of infusion catheter breakage. Three rats receiving papaverine were excluded, two because of infusion catheter breakage and the other because a mesenteric arterial branch had been missed and was therefore not divided (consequently there was no necrotic bowel). Decisions about exclusion were made by an observer unaware of the treatment of each rat, before any evaluation of bowel viability. All rats with correctly performed segmental small bowel devascularization (n = 51) survived and were found to have a central segment of necrotic bowel present within the devascularized segment.

Evaluation of Bowel Viability at 48 Hours

At each end of the devascularized segment, a length of devascularized bowel was maintained viable (by collateral blood flow through intramural and patent extramural mesenteric vasculature). Nonviability was readily evident as greenish-black and flaccid, thinned bowel, lacking tissue tone. The transition from viable to nonviable bowel was usually evident by gross examination and palpation and often showed an abrupt change from moderate tone with normal color to flaccidity with necrotic coloration (Fig. 3). In some rats, a short hemorrhagic transition zone was present; in these, the exact point of transition was sometimes modified, albeit only slightly, by the histologic finding. Overall, the length of bowel maintained viable at the distal end of the devascularized segment (adjacent to the ileocecal junction) was greater than the bowel length maintained viable at the proximal end (Table).

Figure 3. Exteriorized small bowel 48 hours after segmental devascularization. (A) The proximal point of devascularization is shown at mark “A,” with division of the mesentery. The bowel between “A” and “B” is grossly viable. Only a part of the nonviable bowel is seen, with the distal part of the devascularized segment still within the abdomen. (B) Transition zone within devascularized segment, with change of viable, intact mucosa to nonviable bowel and with loss of mucosa and muscularis. A cross-section of silk suture (arrow) marks the apparent change of viable to nonviable bowel on gross inspection. (×40)

Table. EFFECT OF VASODILATOR AND VASOCONSTRICTOR INFUSION: VIABLE LENGTHS WITHIN THE DEVASCULARIZED SEGMENT

TDL, total devascularized length; TVL, total viable length within devascularized segment; VLp, viable length, proximal segment; VLd, viable length, distal segment; Pa, systemic mean arterial pressure.

*P < .05 vs. control.

Microscopic examination in all specimens confirmed the presence of viable, intact mucosa at each end of the devascularized segment and necrosis of the mucosa and underlying muscularis propria within the flaccid, necrotic middle portion identified on gross examination. The transition zone was characterized often by an abrupt loss of mucosa, then a short transition within the muscularis from viable muscle to necrosis (Fig. 3B). This microscopic identification of the transition point usually agreed with that determined on gross examination. Adjustments of the transition point were made if needed, but usually were within 0.5 to 1.0 cm of the point determined on gross examination.

Effects of Prolonged Intraarterially Infused Vasoactive Agents

In the 16 control animals, within the standardized devascularized length of 33 ± 1 cm, collateral flow from the adjacent normally vascularized segments was sufficient to maintain 26.6 ± 1.0 cm viable. Despite the presence of necrotic bowel within the peritoneal cavity over the 48-hour period, systemic arterial pressures were not significantly affected. The length of viable bowel at the proximal end of the devascularized segment remained consistently less than that found at the distal end, adjacent to the ileocecal junction.

All four rats in the papaverine (40 μg/kg/min) group without devascularization survived the 48-hour infusion period, with viable small bowel throughout. Systemic arterial pressures at 48 hours were 127 ± 5 mmHg. In the papaverine (30 μg/kg/min) group (n = 9), infusion of this vasodilator failed to generate any increase in total viable length within the devascularized segment. Systemic arterial pressure was not significantly affected. In the papaverine (40 μg/kg/min) group (n = 8), increasing the dose of infused papaverine also failed to improve total viable length; instead, it decreased significantly to 21.8 ± 1.4 cm, with a significant decrease at the proximal end of the devascularized segment (nearest the point of vasodilator infusion). The length of the viable distal end also decreased compared with control. Systemic arterial pressure decreased to 96 ± 14 mmHg.

In the isoproterenol group (0.06 μg/kg/min; n = 8), infusion of this vasodilator not only failed to increase total viable length within the devascularized segment but also significantly decreased the length of the proximal viable end. Systemic arterial pressure was unchanged from that of controls.

In the norepinephrine (0.1 μg/kg/min) group (n = 5), vasoconstrictor infusion decreased the total length of viable bowel within the devascularized segment significantly to 21 ± 1.3 cm. Systemic arterial pressure also decreased to 101 ± 9 mmHg. In the norepinephrine (0.2 μg/kg/min) group (n = 5), an increased dose of vasoconstrictor further decreased the total viable length to 20.4 ± 2.1 cm. Systemic arterial pressure was significantly decreased to 81 ± 14 mmHg. Only these latter rats tended to be lethargic at 48 hours, with a large accumulation of ascites.

DISCUSSION

This model of segmental mesenteric vascular occlusion in rats produced quantitatively reproducible lengths of viable bowel within a standardized length of sequentially devascularized small bowel. This allowed the determination of the effect of intraarterially infused vasoactive agents (both vasodilators and a vasoconstrictor) on the viability of the devascularized bowel. Rats were chosen because the anatomy and physiology of their mesenteric vasculature closely mimics that of the human circulation, with its rich collateral supply to the small intestine, and also because of the ability to withstand the physiologic stress of bowel devascularization and consequent in situ intestinal necrosis. With the systemic support provided by parenteral fluids, glucose, bicarbonate, and antibiotics, all rats survived the 48-hour bowel devascularization period. This period was chosen to ensure that bowel viability or nonviability could be reliably determined by objective histologic criteria, because previous studies had found that although evaluation of bowel subjected to vascular occlusion and examined histologically at periods of less than 48 hours may yield ambiguous results, examination at this time is determinant. ⁹

This model sought to evaluate the changes in bowel viability resulting only from the presence of local collateral blood flow, and the subsequent changes from the addition of selected vasoactive agents. We therefore sought to minimize any additional systemic changes with appropriate systemic support. Indeed, the control rats, despite the presence of 6.5 + 1.0 cm of nonviable, necrotic bowel, were systemically stable at 48 hours, with a mean arterial pressure of 126 + 2 mmHg. Our purpose was to test whether the addition of a vasodilator to this stable preparation improved collateral flow and bowel viability, as has been suggested, ^4,5 or whether its addition failed to increase bowel viability, as we had inferred from earlier, acute hemodynamic studies. ⁷ Within the dose ranges of agents infused, all preparations in the dose–response studies remained stable, as did the rats in which papaverine was infused during a 48-hour period without devascularization, indicating that there was no evident direct toxic effect from these vasoactive agents.

Infusion doses of vasodilators were chosen to increase normal SMA flow approximately 20% over control values. This flow increase was arbitrarily selected to provide a moderate vasodilation to test whether bowel viability was improved, and to avoid possible systemic hemodynamic effects from prolonged high-dose infusions. When the 30-μg/kg/min dose of papaverine failed to show any improvement in bowel survival, we increased the dose to 40 μg/kg/min to preclude the possibility that too low a dose had been chosen. No benefit to bowel survival was shown; indeed, there was a significant reduction in viability. At this higher dose the lower systemic arterial pressure may have contributed to an incremental length of nonviable bowel, but this could not explain the decreased bowel viability at the lower dose. It does, however, suggest that a further increase in the dose would be unlikely to improve viability. Moreover, infusion of the vasodilator isoproterenol also failed to improve bowel survival, even when systemic arterial pressure was the same as that of the controls.

This preparation consistently showed longer viable lengths at the distal end of the devascularized segment compared with those at the proximal end. Collateral blood flow to the proximal part was supplied by a vascular arcade in the midsmall bowel, whereas collateral blood flow to the distal part was supplied by a terminal branch of the cranial mesenteric artery. Because the vasoactive agent was infused closest to the proximal branches of the SMA, one might expect more profound hemodynamic effects at the proximal end of the devascularized segment, and indeed this was reflected in these results.

Boley et al ^2,10 have reported studies of the SMA hemodynamic response to vasodilator infusion for partial SMA occlusion using a canine model. They noted a persistent vasoconstriction in the mesenteric vasculature after partial SMA occlusion and reported increased SMA blood flow from intraarterially infused papaverine. Clinical applicability was advocated for cases of nonocclusive mesenteric ischemia, or after surgical relief of vascular occlusion. A subsequent study in a canine model of 6-hour complete SMA occlusion with concurrent intraarterially infused papaverine purported to show increased bowel survival but has been reported only in abstract form. ¹¹

It is difficult to interpret these studies, as reported, because they represent preliminary reports, which are not traditionally peer-reviewed, and do not present sufficient data to allow evaluation of conclusions. They (apparently) do not represent blinded studies and apparently do not include concurrent (nonischemic) controls (dogs are particularly susceptible to hemodynamic instability, especially splanchnic vasoconstriction and sepsis during extended, open abdominal procedures). ¹²

Based in part on these canine studies, Boley et al ^4,5 have advocated the use of intraarterially infused vasodilators such as papaverine to counteract mesenteric vasoconstriction that they have postulated to characterize not only nonocclusive but also occlusive mesenteric vascular disease. Although management of mesenteric ischemia under this protocol depends in part on the results of angiography, small emboli to the SMA (as would result in segmental occlusive mesenteric ischemia) are also managed initially by papaverine infusion. They have reported apparent improvements in death and complication rates for these patients. These clinical reports, however, are anecdotal, do not cite concurrent controls, and do not examine whether the use of vasodilators is beneficial specifically in cases of occlusive mesenteric ischemia. Moreover, they do not separate the effects of early diagnosis and aggressive management from postulated specific effects of vasodilator infusion.

The phenomenon of nonocclusive mesenteric ischemia is a well-recognized manifestation of a generalized splanchnic vasospastic response to conditions of low cardiac output, which appears to be a hyperresponsiveness of the splanchnic vascular bed to angiotensin II. ^13–16 This vasospasm can lead to bowel ischemia and necrosis, with more profound compromise of the systemic and splanchnic hemodynamic status and resultant end organ damage in the stomach, ¹⁷ small intestine, ¹⁴ colon, ¹⁸ and liver. ¹⁹ Reversal of the splanchnic vasospasm may be seen with improvement of the low-flow cardiac output, interruption of the renin–angiotensin axis with a suitable blocking agent, 13–15,17–19 or selective infusion of a vasodilator into the ischemic vascular bed. ^1,3

We have previously characterized collateral blood flow and its hemodynamic determinants quantitatively in a canine model of acute segmental mesenteric vascular occlusion. ⁶ Collateral flow from a nonischemic segment to an adjacent ischemic segment was found to be substantial, maintaining flow to the ischemic segment at 55% of the preocclusion value. Collateral flow was facilitated entirely by vasodilation within the ischemic bed, which produced differential nutrient blood flow from the nonischemic segment into the ischemic one. The only active determinant of the differential, collateral flow was vasodilation within the ischemic bed; all other hemodynamic factors, including those in the adjacent, nonischemic segment, played only a passive role. These findings indicated that collateral flow after acute segmental mesenteric occlusion was determined almost entirely by the relative vascular resistances of the ischemic and adjacent, nonischemic vascular beds.

In a subsequent study, the acute infusion of pharmacologic doses of the vasodilators isoproterenol and papaverine had only marginal effects on vascular resistance in the (already maximally vasodilated) ischemic bed, but substantially dilated the nonischemic bed with a consequent increase in blood flow to the nonischemic segment. ⁷ This preferential dilation of the nonischemic bed decreased perfusion pressure in both beds, distal to the site of vascular occlusion, and thereby reduced collateral blood flow to the ischemic segment to a small but significant degree. These hemodynamic changes are characteristic of a vascular steal phenomenon. (However, the acute infusion of the vasoconstrictor norepinephrine also reduced nutrient collateral flow, primarily by preferentially constricting the ischemic bed.)

The acute nature of these prior studies precluded the evaluation of possible later hemodynamic sequelae, such as the superimposition of a splanchnic vasospasm resulting from nonocclusive mesenteric ischemia. Although the use of either vasodilators or vasoconstrictors in the setting of segmental vascular occlusion appeared from these studies to be contraindicated, it was important to extend these studies to prolonged infusions to enable assessment of intestinal viability, the clinically meaningful end point of intestinal vascular occlusion.

In a rat model of segmental mesenteric occlusion using a standardized devascularized length of small bowel (similar to the model presented here), Lee et al ²⁰ investigated the effect of systemically administered vasodilator isoproterenol on intestinal viability. A hemodynamically significant dose was administered by means of the femoral vein for 48 hours. This infusion not only failed to improve bowel viability but also decreased it significantly. However, a reduction in systemic perfusion pressure of 38% was noted during dose–response studies, with only a 20% decrease in mesenteric vascular resistance, so that the systemically administered vasodilator appeared to have dilated the systemic circulation more than the splanchnic circulation. As a consequence, SMA blood flow was reduced by 25%. Although it is also likely that the infusion had dilated the nonischemic mesenteric vasculature more than the ischemic, these two factors could not be differentiated. Moreover, this model did not mimic the clinical situation, where a vasodilator would be infused selectively into the SMA.

The current model simulated the clinical counterpart with respect to both segmental vascular occlusion and intraarterially administered vasoactive agents. We sought to test the hypothesis that intraarterially infused vasodilators would result in improved bowel survival (the clinically important end point). We found that the prolonged infusion of two different types of vasodilators not only failed to increase bowel survival but, in the case of papaverine, actually reduced it significantly. These results were consistent with our earlier acute hemodynamic canine models of segmental mesenteric ischemia, which showed that vasodilator infusion did not improve collateral blood flow and implied that prolonged vasodilator infusions would either fail to improve or would actually reduce subsequent bowel viability. ⁷

Vasoconstrictor infusion also significantly reduced bowel viability in segmental mesenteric ischemia. This also is consistent with earlier acute hemodynamic studies that found that preferential vasoconstriction of the already dilated ischemic vascular bed would increase vascular resistance in the ischemic bed and reduce effective collateral blood flow.

One important limitation of the study is that it looked only at a permanent segmental vascular occlusion, not a transient occlusion that would mimic the effects of vasodilator therapy on blood flow after revascularization. Although hyperemia rather than ischemia is the normal sequela of ischemia/reperfusion, ²¹ there are reports of postischemia vasospasm in the mesenteric bed after revascularization. ²² Because this study focused only on a permanent acute segmental vascular occlusion, our finding cannot be considered relevant to a clinical situation involving revascularization.

Although the present study did not address the controversial role of the use of vasodilators for the treatment of primary nonocclusive mesenteric ischemia, there have been no adequate experimental studies or controlled clinical reports documenting benefits from their use in segmental occlusive mesenteric ischemia. The present study indicates that prolonged intraarterial infusion of vasodilators not only failed to increase bowel viability in segmental occlusive mesenteric ischemia but, in the case of papaverine, actually reduced it significantly. These results are in accord with studies of collateral flow in other organs, including the heart, limb, and brain. ^23–25 For patients who show mesenteric vascular occlusion on angiography, these results argue against clinical treatment with intraarterial vasodilators, at least until controlled clinical evidence is present to indicate otherwise.