Use of Selective Mesenteric Vasodilator Peptides in Experimental Nonocclusive Mesenteric Ischemia in the Dog

GASTROENTEROLOGY 1986;90:669-76

Use of Selective Mesenteric Vasodilator Peptides in Experimental Nonocclusive Mesenteric Ischemia in the Dog KEITH L. MAcCANNELL, CHRISTINE A. NEWTON, KARL LEDERIS, JEAN RIVIER, and MICHAEL TIFF ANY

Departments of Medicine and Pharmacology and Therapeutics, University of Calgary Faculty of Medicine, Calgary, Alberta, Canada; and Salk Institute, La Jolla, California

Three structurally related peptides, ovine corticotropin-releasing factor, sauvagine, and urotensin I [4-41], are selective mesenteric vasodilators in dogs. To assess the possible benefit of these peptides in nonocclusive mesenteric ischemia, they were compared with a nonselective vasodilator, sodium nitroprusside, in the anesthetized dog. Mesenteric blood flow was reduced by ~30%, without lowering of systemic arterial pressure, by either digoxin or pericardial tamponade. In the digoxin model, i.v. infusions of corticotropin-releasing factor, sauvagine, and urotensin I restored intestinal vascular resistance and mesenteric blood flow to control values, without causing a fall in systemic arterial blood pressure. In the tamponade model, only urotensin I was assessed, and it produced the same restoration of hemodynamic variables. On the other hand, in both models, i.v. infusions of nitroprusside, which were effective in correcting intestinal vascular resistance, produced a fall in arterial blood pressure (presumably because of systemic dilatation), which prevented restoration of mesenteric blood flow. Intestinal oxygen uptake was not altered by tamponade, but was reduced by 23% in the digoxin model, where it was restored to control values by both the peptides and nitroprusside. The increased oxygen extraction seen in both models was corrected by the peptides but not by nitroprusside, suggesting that nitroprusside may have a direct and offsetting metabolic effect on the gut. Nonocclusive mesenteric ischemia (NOMI) occurs in heart failure or hypovolemia as a consequence of Received December 14, 1983. Accepted September 18, 1985. Address requests for reprints to: Dr. K. 1. MacCannell, Departments of Medicine (Gastroenterology) and Pharmacology and Therapeutics, Health Sciences Center, University of Calgary, 3330 Hospital Drive N.W., Calgary, Alberta T2N 4NI, Canada. © 1986 by the American Gastroenterological Association 0016-5085/86/$3.50

depressed cardiac output (1-4). Many patients with this condition are being treated with cardiac glycosides, which are known to cause decreased mesenteric perfusion (5). The mortality with NOMI is ~80%, although improved survival may be possible with early diagnosis and intervention (6,7). Boley (6) has popularized aggressive roentgenographic confirmation of the diagnosis, coupled with close-arterial infusion of papaverine, a vasodilator alkaloid, directly into the mesenteric artery. Because it has been assumed that vasoactive agents administered intravenously would produce systemic cardiovascular effects that would preclude their use, experimental approaches to the management of NOMI have focused on the search for agents that would be active in low doses when administered intraarterially into the mesenteric artery and that would be cleared rapidly from the circulation (8). Using models of experimental NOMI produced by hemorrhage-induced hypotension or by digitalisinduced mesenteric vasoconstriction, a number of vasoactive substances have been examined for possible benefit when administered intraarterially. These include dopamine (9); isoproterenol (9); glucagon (9,10); prostaglandins El (9,11), D2 (12), and 12 (12); histamine (10); perhexilene (10); and adenosine analogues (12). None has been ideal. An agent that selectively dilated the mesenteric circulation and that had no other cardiovascular action would offer a distinct advantage over such drugs, as one could give this latter agent intravenously, and yet target its actions for the mesenteric circulation. We have reported in this journal (13) that three chemically related peptides, urotensin I [4-41] (U I), Abbreviations used in this paper: CRF, corticotropin-releasing factor; NOMI, nonocclusive mesenteric ischemia; NP, sodium nitroprusside; S, sauvagine; U I, urotensin 1.

670

MAcCANNELL ET AL.

sauvagine (S), and ovine corticotropin-releasing factor (CRF), all of which are available in synthetic form, are selective dilators of the mesenteric circulation (superior and inferior mesenteric outflows) in the dog. Selectivity is so marked that these agents do not dilate the celiac outflow. They have no direct actions on the heart, and the slight hypotension that they produce (at its maximum, about a 10%-15% fall in systemic arterial blood pressure) is solely a function of the mesenteric vasodilatation: once maximal dilatation is achieved, the peptides produce no additional decrease in blood pressure. No new cardiovascular actions emerge, even at very high doses (14). The mesenteric vasodilatation is not secondary to a "metabolic" action on the intestinal mucosa or muscle. Although all three peptides release corticotropin, the mesenteric vasodilator effect does not appear to be related to adrenocorticotrophic hormone or cortisol release (13). The effect of all three agents on intestinal oxygen kinetics is identical to the effect produced by passive increases in intestinal blood flow (13) and quite unlike the effect produced by drugs such as acetylcholine or isoproterenol, which have a direct metabolic action on the intestinal wall. This communication reports on the experimental use of these peptides in two models of mesenteric ischemia, both produced in the dog, one by the use of pericardial tamponade, and the other by the use of cardiac glycosides.

Materials and Methods Reduction in Mesenteric Blood Flow by Pericardial Tamponade Twelve mongrel dogs of either sex, weighing from 22 to 33 kg (mean ± SEM 26.7 ± 1.0 kg) and fasted 24-48 h, were anesthetized with 30 mg/kg of sodium pentobarbital, intubated, and ventilated with room air using a Harvard respiration pump (Harvard Apparatus Co., South Natick, Mass.). The right femoral artery and vein were catheterized for measurement of systemic arterial and caval pressures, respectively, and a Swan Ganz catheter (Electronics for Medicine, Pleasantville, N.Y.) was inserted via the right internal jugular vein for measurement of pulmonary artery pressure and cardiac output by thermal dilution. All pressures were sensed with Statham transducers (Statham, Oxnard, Calif.), adjusted to the level of the right atrium, and calibrated against mercury. The left femoral artery was cannulated to permit sampling of arterial blood, and the left femoral vein was used for drug infusions. A midline laparotomy was done and an electromagnetic flow probe (Micron Instruments, Los Angeles, Calif.) was placed around the root of the cephalic (superior) mesenteric artery. Mesenteric venous pressure was measured with a catheter inserted antegrade by a small vein in the mesenteric arcade; this catheter was also used to sample mesenteric venous blood. The chest was then

GASTROENTEROLOGY Vol. 90, No.3

opened in the fifth left intercostal space, and a perforated soft polyethylene tube was introduced into the pericardial space; the pericardial opening was sealed around the catheter with cyanoacrylate adhesive (commonly known as "super glue"). Both the abdominal and thoracic incisions were approximated, but no attempt was made to evacuate the pneumothorax. The heart rate was recorded with lead II percutaneous electrodes. All parameters were recorded on either a Beckman RM (Beckman Instruments Inc., Fullerton, Calif.) or a Gould recorder (Gould Inc., Cleveland, Ohio). After a 1-h stabilization period, 1-ml blood samples were withdrawn in triplicate from the femoral artery and mesenteric vein, using the technique of Lautt (15), for determination of oxygen kinetics. Nonocclusive mesenteric ischemia was then produced by injecting -120 ml of normal saline into the pericardial space so as to reduce mesenteric blood flow by 25%-30%. This technique, described by Starr et al. (16), induces reproducible, stable mesenteric ischemia. At the end of 30 min of tamponade, additional blood samples were taken and, in the presence of continuing tamponade, the first drug was given for 30 min in a dose adequate to restore mesenteric blood flow to control values, or, if flow could not be restored, for at least 45 min so as to achieve the maximal flow increase possible. Blood samples were now taken, and the fluid was removed from the pericardial space, allowing hemodynamic variables to return to normal. At the end of a second 3D-min control period, additional blood samples were taken, and the tamponade was reestablished so as to produce the same reduction in mesenteric blood flow as had been achieved during the first period of tamponade. Blood samples were taken after 30 min, and, in the presence of tamponade, the second drug was now administered as before--blood samples again being taken at the end of the infusion. The tamponade was now removed, and blood samples were again taken after a 3D-min control period. Hemodynamic measurements were continued throughout. The order of presentation of synthetic U I and sodium nitroprusside (NP) was alternated, each being given first to 6 dogs. Both agents were given intravenously. The mean dose of U I was 8.25 ± 1.23 ng/kg. min and that of NP was 7.69 ± 1.10 /Lg/kg· min; the rates of infusion were 0.2-0.4 mllmin. Although this protocol of alternating the order of administration of the agents should have removed any concern about order artifacts, we studied the effects of tamponade alone in an additional 3 control dogs. In these animals, the same protocol was used except that saline was administered during all periods-the control periods and the "drug" periods.

Reduction of Mesenteric Blood Flow Using Digoxin All animals were prepared as described above, with the following exceptions: the thoracotomy and the insertion of the Swan Ganz catheters were not done, and the common colic artery was cannulated in those instances where agents were administered intraarterially into the mesenteric circulation. Twenty-four mongrel dogs of ei-

March 1986

ther sex were used; they weighed from 16.0 to 32.1 kg (mean ± SEM 24.1 ± 3.5 kg). Following a l-h stabilization period, l-ml blood samples were withdrawn in triplicate as above. Nonocclusive mesenteric ischemia was then induced with digoxin, following the method of Schwaiger et al. (10): a bolus of 75 J.Lg/kg (4.8-9.6 ml) was given intraveneously, followed by 2.5-10 J.Lg/kg. min at 30-min intervals until mesenteric blood flow was reduced 25%-30%, at which time additional blood samples were taken. No additional digoxin was administered after this time. The mean dose of digoxin was 103.6 ± 5.2 J.Lg/kg. Preliminary experiments indicated that flow remained depressed after digoxin administration for at least 135 min in the absence of additional intervention. Synthetic U I, S, CRF, or NP was now administered (each to 6 dogs) to restore mesenteric blood flow as in the first set of experiments. The infusion, at a rate of 0.2-0.4 ml/min, was continued at least 15 min after flow had been restored to control levels, or for 45 min if flow could not be restored. Blood samples were taken at this time and again 30 min after the drug infusion had stopped. Sauvagine, CRF, and NP were each given intravenously to a series of 6 dogs in doses of 1.5 ± 0.3 ng/kg· min, 78 ± 11 ng/kg. min, and 8.9 ± 2.4 J.Lg/kg. min, respectively. Urotensin I was administered intraarterially into the common colic artery of an additional 6 dogs in a dose of 9 ± 2 ng/kg . min. In 3 of these latter dogs, a second challenge with U I, this time intravenously, was given after the second control period. At the time of the second administration, the effects of the first challenge had disappeared and the vasoconstrictive effects of the digoxin were again dominant. Digoxin and NP (Sigma Chemical Co., St. Louis, Mo.) were prepared from powder, the former being dissolved in 20% ethanol/distilled water and the latter dissolved in distilled water. The peptides used in these experiments were synthesized using solid phase methods, with purification by preparative and semi preparative high-performance liquid chromatography. Each synthetic peptide had the expected amino acid composition, and each behaved as a single peak using several high-performance liquid chromatography systems. Mesenteric resistance was calculated by dividing the pressure drop across the intestine (arterial blood flow minus mesenteric venous pressure) by the mesenteric blood flow. Blood samples were analyzed for total hemoglobin (THb), percent oxygen saturation (% 02Hb), percent carboxyhemoglobin (% metHb), and oxygen content (vol % O2) using an Instrumentation Laboratory 282 co-oximeter (Instrumentation Laboratory, Inc., Lexington, Mass.), which was calibrated using dog blood. Volume % O 2 was calculated using the following formula: vol % O 2 = THb x 1.39 x (% O 2 Hb/l00). These data were then used to determine intestinal oxygen uptake (02U = A-V O 2 difference x SMA flow) and percent oxygen extraction [(% 02E = A-V O 2 difference/vol % O 2 arterial) x 100]. Data were analyzed using analysis of variance and the Student's Hest for paired data. The t-test was used only when the analysis of variance showed statistically significant treatment effects.

NONOCCLUSIVE MESENTERIC ISCHEMIA 671

Results Effect of the Selective Mesenteric Vasodilator Peptide, Urotensin I, in Mesenteric Ischemia Produced by Pericardial Tamponade

Preliminary experiments done on 3 dogs established that there were no observable differences between the first and second period of tamponade in otherwise untreated dogs. Table 1 illustrates the effect of tamponade on hemodynamics and oxygen kinetics. These data were derived from the six periods of tamponade in these 3 control animals and the 24 periods of tamponade in the 12 experimental animals reported below: Cardiac output was measured in 10 of the 15 dogs, and on average, fell 21%. Mean pulmonary artery and inferior caval pressures rose slightly during tamponade, but mean arterial blood pressure did not change. Tamponade produced a fall in mesenteric blood flow averaging 26.5%. The indices of intestinal O2 extraction (A-V O2 difference and % O 2 extraction) were increased, without change in O2 uptake. Figure 1 shows the response to Lv. infusions of U I and of NP during tamponade-induced mesenteric ischemia. Urotensin I, although given intravenously, corrected the tamponade-induced alterations in mesenteric flow and resistance, and did so without altering systemic arterial blood pressure. It also corrected the tamponade-induced changes in A-V O2 difference and O 2 extraction. Indeed, the values for all these variables during tamponade + U I periods were very similar to values obtained during control periods in the absence of tamponade. In contrast, although Lv. NP reduced the tamponade-induced Table 1. Effect of Pericardial Tamponade on Hemodynamics and Intestinal Oxygen Kinetics Cardiac output (mllmin) Mesenteric blood flow (mllmin) Mean systemic arterial pressure (mmHg) Intestinal vascular resistance (PRU a ) Mesenteric venous pressure (mmHg) Mean pulmonary artery pressure (mmHg) Inferior caval pressure (mmHg) Intestinal O2 extraction A-V O2 difference

Control

Tamponade

2710 ± 150 202.5 ± 11.9 131.4 ± 3.1

2140 ± 110b 148.8 ± 8.9" 130.1 ± 3.3

67.9±4.0

91.4 ± 6.4 b

5.9 ± 0.3

7.5 ± 0.3"

18.2 ± 1.2

19.3 ± 1.0

4.8 ± 0.1

6.9 ± 0.2"

3.6 ± 0.3

5.2 ± O.4 b

19.3 ± 1.4 6.8 ± 0.4

27.1 ± 1.8" 7.1 ± 0.4

(m1l100 ml of blood)

Percent O2 extraction Intestinal O 2 uptake (mllmin) a

PRU, peripheral resistance units = mmHg/ml . min. "p < 0.01.

b

p < 0.05.

672

GASTROENTEROLOGY Vol. 90, No, 3

MAcCANNELL ET AL.

.CONTROL

~ TAMPONADE

**p < 0 .01 *** p < 0 .001

UROTENSIN I I.V. (N = 12) * * /\/\

** /\/\

150

* p < 0 .05

~TAMPONADE + DRUG

6 .0

100

4.0

4 .0

50

2 .0

2 .0,..O\J" _

SODIUM NITROPRUSSIDE I. V. (N = 12)

* /\

* /\ 10

20

2.

10

6 .0

~~ ~ ~

4 . .

2 .0

~ SYSTEMIC ARTERIAL PRESSURE (mm Hg)

SMA FLOW (ml I min)

MESENTERIC RESISTANCE (PRU)

A-V OXYGEN DIFFERENCE (m1 11 00mll

OXYGEN EXTRACTION (% )

OXYGEN UPTAKE (ml I min)

Figure 1. Response of anesthetized dogs to pericardial tamponade and to the i.v. administration of either sodium nitroprusside, a nonselective vasodilator, or urotensin I, a selective mesenteric vasodilator peptide. The first and fourth blocks represent control periods during which there was no tamponade. Tamponade was present during the second and third blocks. The second block shows the effect of the tamponade only, and the third block shows the effect of tamponade and drug. If analysis of variance showed a significant treatment effect for a variable, differences between adjacent blocks were calculated by paired t-test. Note that the doses of nitroprusside, which produce similar changes in mesenteric vascular resistance, drop arterial pressure and cannot restore mesenteric blood flow. This may be responsible for the restoration in intestinal oxygen extraction being less complete with the nitroprusside. PRU, peripheral resistance units; SMA, cephalic (superior) mesenteric artery.

elevation in mesenteric vascular resistance, it produced a marked fall in systemic arterial blood pressure, presumably because generalized vasodilation made it impossible to restore mesenteric blood flow fully. Sodium nitroprusside tended to correct the tamponade-induced elevations in A-V O 2 difference and O 2 extraction, but, unlike U I, the correction was not statistically significant.

Effect of Selective Mesenteric Vasodilator Peptides on Mesenteric Ischemia Produced by Digoxin The administration of digoxin to 24 dogs produced an average lowering of mesenteric blood flow of 30.5%, without a significant change in arterial blood pressure (Table 2). The increase in intestinal vascular resistance was accompanied by a

small but significant increase in mesenteric venous pressure, without a corresponding change in caval pressure. Digoxin produced small but significant increases in the indices of intestinal O2 extraction, but these changes were offset by the decrease in mesenteric blood flow, such that a mean fall in intestinal O 2 uptake (A-V O 2 difference x blood flow) of 23.3% occurred. Although only a small number of animals was used in each subset of digoxin-treated dogs, the same pattern of response was seen in each group (compare Table 2 with Figure 2); however, because of the small number the responses were frequently not statistically significant. In doses that were effective in restoring the digoxin-induced increase in intestinal vascular resistance to control levels, i.v. NP caused a significant fall in systemic arterial blood pressure (Figure 2). Small doses did not cause a fall in blood pressure,

March 1986

NONOCCLUSIVE MESENTERIC ISCHEMIA 673

Table 2. Effect of Digoxin on Hemodynamics and Intestinal Oxygen Kinetics Control Systemic arterial pressure (mmHg) Mesenteric venous pressure (mmHg) Interior caval pressure (mmHg) Mesenteric blood flow (mllmin) Intestinal vascular resistance

128.8 ± 2.3

Digoxin 127.8 ± 4.1

5.5 ± 0.3

6.9 ± O.4 b

4.7 ± 0.3

5.0 ± 0.3

156.5 ± 9.8

108.7 ± 7.9 b

85.4 ± 5.4

123.0 ± 9.0 b

4.0 ± 0.3

4.7 ± 0.4c

17.7 ± 1.1 6.0 ± 0.4

20.4 ± 1.9 C 4.6 ± 0.4c

by expressing them as percentage change from control (p < 0.001 and p < 0.01, respectively). In the case of U I, the changes in O2 extraction were not significant even when the data were normalized. All of the peptides tended to raise the digoxin-depressed O2 uptake toward normal, but the changes were not significant for any of the peptides, whether expressed in absolute terms or as percentile change. The actions of all peptides were short-lived, and the effects of the digoxin were again apparent within 30 min of stopping the infusion.

(PRUO)

Intestinal O 2 extraction A-V O2 difference (mlll00 ml blood)

Percent O2 extraction Intestinal O2 uptake (mllmin)

°PRU, peripheral resistance units c

P

< 0.05.

= mmHg/ml .

min. b p < 0.001.

but they were also ineffective in correcting the digoxin-induced disturbance in intestinal vascular resistance. As in the tamponade experiments, the decrease in perfusion pressure again prevented restoration of mesenteric blood flow. Indices of O2 extraction, already increased by digoxin, were additionally increased by NP relative to control values. Relative to the depressed postdigoxin value, intestinal O2 uptake also was increased by NP in spite of the fall in mesenteric blood flow. Because O 2 uptake is the product of blood flow and A-V O 2 difference, this increase in O2 uptake reflected the increased O 2 extraction. The effect of NP on hemodynamics and on O2 kinetics was sustained, with effects still being marked 30 min after discontinuing the agent. Intravenous CRF, S, and U I also lowered the digoxin-induced elevation in intestinal vascular resistance but, in contrast to NP, all three peptides restored the depressed mesenteric blood flow, and none caused a fall in systemic arterial blood pressure, with the exception of CRF, which caused a small, but statistically significant, decrease. A comparison between i.v. and i.a. infusions was made only in the case of U I: the hemodynamic responses to i.v. U I were not distinguishable from those of i.a. U lor i.v. CRF or S. Each of the peptides decreased the digoxin-elevated indices of O2 extraction. In the case of CRF, the fall in A-V O2 difference and % O 2 extraction was significant (p < 0.01 and p < 0.01, respectively). The changes in these indices with S were not significant when expressed in absolute values but became significant when intra animal variability was removed

Discussion In these experiments, the imposition of pericardial tamponade or the administration of digoxin decreased mesenteric blood flow and increased mesenteric vascular resistance. These hemodynamic effects were fully reversed by the three structurally related peptides, ovine CRF, S, and U I. All three peptides were effective when administered intravenously, and achieved the desired hemodynamic effect by this route of administration without lowering arterial blood pressure and, hence, mesenteric perfusion pressure. We have demonstrated the same response to these peptides with two different models of mesenteric ischemia, and this provides added assurance that our results can be generalized. Digoxin or tamponade reduced mesenteric blood flow by ~30%, without lowering systemic arterial perfusion pressure. This is not a minor insult: for example, in the tamponade experiments, cardiac output was reduced by 21%. Preliminary experiments indicate that the results given here can be replicated under more severe conditions. We have previously shown that although CRF is less potent than U I or S, these three peptides cannot be distinguished one from the other with respect to their hemodynamic actions (14). All three are selective dilators of the canine mesenteric circulation (13,14,17) and have no other hemodynamic actions even when large doses are given intravenously (14). It is not surprising, therefore, that the effects of i.v. U I closely resembled those of U I delivered directly into the mesenteric artery, or that the effects observed were very similar to i.v. CRF or S on the one hand and i.a. U I on the other. These results suggest possible beneficial results with i.v. administration of these peptides in NOM!. There are, at present, two caveats to be applied to this suggestion. First, we have reported previously (17) that native U I also produces selective vasodilatation in the subhuman primate, but a systematic examination of possible species differences has not been done. Second, one has no assurance that the animal models of NOMI

674

MAcCANNELL ET AL.

GASTROENTEROLOGY Vol. 90. No.3

SAUVAGINE I.V. (N=6)

"lii ' 'm "'ru "hl "Ill * * *** ** *

* * * **

/\

* *

;/'

150

* *

6.0

8.0 6.0

100

100

100

4.0

20

4.0

50

50

50

2.0

10

2.0

*

OVINE CRF I.V. (N =6)

"'In ' 'jfi ' 'lei * * *

***

*

/'-..

* **

/\

;'/\

150

8.0

6.0

6.0

100

100

4.0

20

4.0

50

50

50

2.0

10

2.0

' 'Ill

tfi ***

150

100

100

50

50

*

*

m"m m * ***

8.0 /\ /\

100

200

/\ . 111 *

UROTENSIN I I.A. (N = 6) * * * /\*/\

"'m ' krt "hi /\

6.0

100

4.0

20

50

2.0

10

*

4.0

2.0

*

UAOTENSIN I I.V. (N =3)

"IiI

6.0

*

'''b ' '1Jj ' 'm

"m "IIi

,.om "'Ill ' 'In

'1[1 "Ill "tj]

**

/\/\

* * *

/\/\

150

100

100

100

50

50

50

*

4.0

20

4.0

2.0

10

2.0

*

*

150

100

100

100

50

50

50

SYSTEMIC ARTERIAL PRESSURE (mm Hg)

6.0

*

SODIUM NITROPRUSSIDE I.V. (N=6) * * /'-.. * ** /".

* /\

C OEC

8.0

6.0

C OEC SMA FLOW (ml/min)

*

6.0

CO EC* MESENTERIC RESISTANCE (PRU)

6.0

4.0

20

4.0

2.0

10

2.0

C DEC A-V OXYGEN DIFFERENCE (mI/100ml)

Figure 2. Response of anesthetized dogs to digoxin and to the administration of either sodium nitroprusside or one of three selective mesenteric vasodilator peptides. C. control; D, digoxin; E, experimental (therapeutic intervention). Note that, because of the long action of digoxin, its effects were present during the experimental period and the second control period (last block in each panel). Thus, the last control block in this figure is not comparable to the last control block in Figure 1. The second and third blocks are entirely comparable: the second block shows the effect of digoxin only and the third block shows the effect of digoxin plus medication.

C DEC

CD EC

OXYGEN EXTRACTION

OXYGEN UPTAKE (ml/min)

(%)

* NO SIGNIFICANT TREATMENT EFFECT BY ANOVA

* p< 0.05

**

p
***

p< 0.001

are ideal. However, reduced cardiac output (such as is produced in the tamponade model) and digitalis glycosides (such as are used in the digoxin model) are certainly contributory to the clinical syndrome in many, if not most, patients with NOM!. Intestinal O 2 uptake is defined as the amount of O 2 taken up by the intestine per minute, while intestinal O 2 extraction is defined as the amount of O2 removed by the tissues from each milliliter of blood. In the present experiments, digoxin, but not tamponade, reduced O 2 uptake (Tables 1 and 2), in spite of similar effects on mesenteric blood flow and intestinal vascular resistance. Similar results have been reported previously with digoxin (10) and with another cardiac glycoside, ouabain (18,19). In our hands, digoxin produced a small increase in O 2

extraction (Table 2), an unexpected finding, but one supported by the data of Schwaiger et al. (10). This profile for digoxin differs from that reported for vasoconstrictor sympathomimetic amines [reduction in both O2 extraction and O 2 uptake (20,21)] or that reported for mechanically produced reduction in intestinal blood flow in the absence of drugs [increase in O2 extraction with no change in O2 uptake (22,23; see also our data for tamponade, Table 1)]. The effects of digoxin on O 2 kinetics are therefore complex, and probably reflect not only decreased blood flow but also a direct metabolic action on the intestinal wall. The present experiments indicate that the digoxininduced fall in intestinal O2 uptake, which parallels the decrease in intestinal blood flow, can be restored

March 1986

by both Lv. peptide or by Lv. NP (Figure 2). Nitroprusside, however, did not correct the hemodynamic alterations produced by digoxin or by pericardial tamponade: we could not achieve effects with Lv. NP that were comparable to those produced by Lv. peptide, without producing unacceptable systemic effects. We did not examine the effects of NP delivered intraarterially into the mesenteric artery, and We concede that effects comparable to those obtained with Lv. peptides probably could have been achieved with NP delivered directly into the mesenteric artery in a dose small enough to prevent systemic effects. The point of possible clinical relevance is that La. administration of these mesentericselective vasodilator peptides was not necessary. If these peptides should prove to be useful in the treatment of clinical NOMI, aggressive interventions such as superior mesenteric artery infusions would not be necessary. Moreover, because the peptides have no other cardiovascular actions, even at high doses (14), the dose schedule need not be critical. It should be noted, however, that higher doses of CRF are required than are required of the other two peptides and that these doses of CRF are in the range where one might expect adrenocorticotrophic hormone release. It is also possible that intravenously administered NP might have corrected the digoxin- or tamponadeinduced decrease in mesenteric blood flow, had supplemental fluids been administered together with the NP, but the necessity of having to expand the vascular volume in such a circumstance would also designate stich a nonselective vasodilator as an inferior agent. In the anesthetized but otherwise untreated dog, the effect of the three synthetic peptides on intestinal O 2 kinetics is to produce a decrease in O 2 extraction with no change in O2 uptake (13). This establishes that these peptides have no direct metabolic effect on the intestine because exactly the same pattern is observed when intestinal blood flow is increased mechanically in the absence of drugs (22,23). When the peptides or NP were administered in the presence of digoxin (see Figure 2), the depressed O2 uptake was restored to control values. The peptides, but not NP, corrected the digoxin- and tampoIiadeinduced increments in O 2 extraction. Indeed, in the case of the digoxin-induced increment in O 2 extraction, NP produced a small additional increase, which may represent a metabolic effect of NP. At the present time, there is no completely satisfactory treatment of nonocclusive mesenteric ischemia, but the most common approach is to infuse a vasodilator, such as NP or papaverine, directly into the mesenteric circulation after selective catheterization of the superior or inferior mesenteric artery.

NONOCCLuSIVE MESENTERIC ISCHEMIA

675

Intravenous administration is avoided because of systemic cardiovascular effects. The intraarterial infusion is a major invasive procedure in a critically ill and frequently elderly patient, and would be rendered unnecessary with the availability of an effective agent that could be given intravenously. This appears to be the promise of the CRF-like peptides. Release of adrenocorticotrophic hormone and cortisol does not appear to be a problem at the doses of U I and S used. Moreover, a synthetic fragment of U I, U I [4-28], appears to have little adrenocorticotrophic hormone-releasing activity (Rivier J, personal communication), while retaining selective mesenteric vasodilatory activity (MacCannell K, Kobayashi Y, and Lederis K, unpublished observations). This fragment could be the optimal agent for use in humans, if humans respond in a similar fashion to the dog.

References 1. Fogerty TJ, Fletcher WS. Genesis of non-occlusive mesenteric ischemia. Am J Surg 1960;111:130-7. 2. Larsen A. Non-occlusive intestinal gangrene. Acta Chir Scand 1970;136:227-34.

3. Price WE. Rohrer GV. Jacobson ED. Mesenteric vascular diseases. Gastroenterology 1970;57:599-604. 4. Sullivan JE. Vascular diseases of the intestine. Med Clin North Am 1974;58:1473-85. 5. Harrison LA. Blaschke J, Phillips BS. Price WE. Cotton M deV. Jacobson ED. Effect of ouabain on the splanchnic circulation. J Pharmacol Exp Ther 1969;169:321-7. 6. Boley SJ. Sprayregen S. Siegelman SS. Veith FJ. Initial results from an aggressive roentgenographic and surgical approach to acute mesenteric ischemia. Surgery 1977;82:848-55. 7. Habboushe F. Wallace HW. Nusbaum M. Bauni S. Dratch P. Blakemore WS. Non-occlusive mesenteric vascular insufficiency. Ann Surg 1974;180:819-22. 8. Lanciault G. Fang WF. Jacobson ED. Bowen JC. Evalt.lation of potential agents for treatment of non-occlusive mesenteric ischemia in the dog. Circ Shock 1976;3:239--46. 9. Ulano HB. Treat E. Shanbour LL. Jacobson ED. Selective dilation of the constricted superior mesenteric artery. Gastroenterology 1972;62:39-47. 10. Schwaiger M. Fondacaro JD. Jacobson ED. Effects of glucagon. histamine and perhexilene on the ischemic canine mesenteric circulation. Gastroenterology 1979;77:730-5. 11. Davis LJ, Anderson JI Wallace S. Jacobson ED. Experimental use of prostaglandin E, in nonocclusive mesenteric ischemia. Am J Roentgenol Radium Ther Nucl Med 1975;125:99-110. 12. Fondacaro JD. Walus KM. Schwaiger M. Jacobson ED. Vasodilation of the normal and ischemic canine mesenteric circulation. Gastroenterology 1981;80:1542-9. 13. MacCannell KL. Hamilton PL. Lederis K. Newton CA. CRFlike peptides produce selective dilatation of the dog mesenteriC circulation. Gastroenterology 1984;87:94-102. 14. MacCannell KL. Lederis K. Hamilton PL. Rivier J. Amunine (ovine CRF). ur6tensin I and sauvagine. three structurally related peptides. produce selective dilatation of the mesenteric circulation. Pharmacology 1982;25:116-20.

GASTROENTEROLOGY Vol. 90, No.3

676 MAcCANNELL ET AL.

15. Lautt WW. Method for measuring hepatic uptake of oxygen or other blood borne substances in situ. J Appl PhysioI1976;40: 269-74. 16. Starr FL, Samphilipo MA, White RI, Anderson JH. A reproducible canine model of non-occlusive mesenteric ischemia. Invest RadioI1982;17:34-6. 17. MacCannell KL, Lederis K. Dilatation of the mesenteric vascular bed of the dog produced by a peptide, urotensin I. J Pharmacol Exp Ther 1977;203:38-48. 18. Frizzell RA, Markscheid-Kaspi L, Schultz SG. Oxidative metabolism of the rabbit ileal mucosa. Am J Physiol 1974; 226:1142-8. 19. Pawlik W, Shepherd AP, Mailman D, Jacobson ED. The effects

20.

21. 22. 23.

of ouabain on intestinal oxygen consumption. Gastroenterology 1974;67:100-6. Shepherd AP, Mailman D, Burks TF, Granger HJ. Effects of norepinephrine and sympathetic stimulation on extraction of oxygen and B6Rb in perfused canine small bowel. Circ Res 1973;33:166-74. Shepherd AP, Pawlik W, Mailman D, Burks TF, Jacobson ED. Effects of vasoconstrictors on intestinal vascular resistance and oxygen extraction. Am J PhysioI1976;230:298-305. Lutz J, Henrich H, Bauereisen E. Oxygen supply and uptake in the liver and intestine. Pflugers Arch 1975;360:7-15. Ohman U. Blood flow and oxygen consumption in the feline small intestine: responses to artificial distension and intestinal obstruction. Acta Chir Scand 1976;142:329-33.