Secondary Logo

Low-Dose Vasopressin in the Treatment of Vasodilatory Septic Shock

Malay, Mary Beth MD; Ashton, Robert C. Jr. MD; Landry, Donald W. MD, PhD; Townsend, Ricard N. MD

The Journal of Trauma: Injury, Infection, and Critical Care: October 1999 - Volume 47 - Issue 4 - p 699

Background Despite appropriate therapy, refractory hypotension often occurs in septic shock. A double-blinded placebo controlled clinical trial was performed to assess the role of low-dose vasopressin (VP) as a pressor agent in septic shock.

Methods Patients admitted to a trauma intensive care unit with vasodilatory septic shock were randomized to receive either VP at 0.04 U/min (n = 5) or placebo (n = 5). Vasodilatory septic shock was defined as a need for catecholamine agents to maintain a mean arterial pressure more than or equal to 70 mm Hg, despite a cardiac index more than 2.5 L/min and a minimal pulmonary artery wedge pressure more than 12 mm Hg. After 1 hour of initiation of the study drug, attempts to discontinue norepinephrine, phenylephrine, and/or dopamine, in respective order, were undertaken provided that the mean arterial pressure remained more than or equal to 70 mm Hg.

Results A vasopressin infusion increased systolic arterial pressure (98 ± 5 to 125 ± 8 mm Hg, p< 0.008) because of peripheral vasoconstriction (systemic vascular resistance increased from 878 ± 218 to 1,190 ± 213 dynes/s per cm−5, p< 0.05). Arterial pressure and systemic vascular resistance were statistically unaffected in the placebo group. Before study termination, measured at 24 hours after drug initiation, two patients in the placebo group died of refractory hypotension. However, all patients receiving VP survived the 24-hour study period and had all other catecholamine pressors withdrawn and blood pressure maintained solely with a low-dose VP infusion.

Conclusion A VP infusion improved arterial pressure and permitted the withdrawal of catecholamine vasopressors. VP is a useful agent in the treatment of refractory septic shock.

From the Allegheny General Hospital, Division of General Surgery (M.B.M., R.C.A., R.N.T.), Pittsburgh, Pennsylvania, and Columbia University, Department of Medicine (D.W.L.), New York, New York.

Address for reprints: Mary Beth Malay, MD, Allegheny General Hospital, Division of General Surgery, 320 East North Avenue, Pittsburgh PA 15212.

Supported by Allegheny-Singer Research Institute, Allegheny General Hospital, Pittsburgh, Pennsylvania.

Septic shock is defined as the circulatory insufficiency that develops in response to overwhelming systemic infection. The most common hemodynamic pattern encountered is characterized by profound peripheral vasodilation resulting in hypotension and end-organ hypoperfusion. 1 Current treatment is supportive and includes appropriate antibiotics, expansion of intravascular volume and, in certain cases, the use of inotropic agents. Despite this therapy, severe hypotension is common and catecholamine vasopressors are often required to maintain arterial pressure. Although norepinephrine has been shown to effectively increase arterial pressure, 2,3 it often exhibits decreased vasopressor activity during septic shock. 4–8 Phenylephrine has also been reported to have a diminished vasoconstrictor effect during sepsis. 9 Given that the mortality rate remains more than 50%, with many of these deaths attributable to refractory hypotension and subsequent multiple organ failure, 10 an alternative pharmacologic agent that is able to restore vascular tone and blood pressure is desirable.

Recently, we reported in an uncontrolled series that patients with septic shock resistant to catecholamine vasopressors may have a defect in the baroreflex-mediated secretion of vasopressin. Administration of a low-dose vasopressin infusion (0.04 U/min) significantly increased arterial blood pressure in these patients and permitted the withdrawal of other catecholamine agents. 11,12 Furthermore, a dose of exogenous vasopressin that provided a plasma concentration expected for the degree of hypotension resulted in a marked pressor response. 12

Although this clinical experience suggested that vasopressin may be a useful agent in the treatment of vasodilatory septic shock, only controlled trials would allow adequate investigation of its role as a potential vasopressor in this complex patient population. By using a double-blinded, randomized clinical design, we studied the hemodynamic effects of low-dose vasopressin in 10 patients with severe septic shock requiring dopamine, phenylephrine, and/or norepinephrine infusions to support arterial blood pressure.

Back to Top | Article Outline


This study was approved by the Institutional Review Board of Allegheny University of the Health Sciences, Allegheny General Hospital. Because all patients in this study were critically ill and unable to give written, informed consent, this consent was obtained from the patient’s surrogate decision-maker.

Back to Top | Article Outline


Eligible patients were older than 18 years of age and had a diagnosis of septic shock defined according to the American College of Chest Physicians/Society of Critical Care Medicine Consensus Committee, 13 including suspected infection or positive blood cultures; mechanical ventilation and at least one or more of the following criteria: heart rate more than 90 beats/min, temperature more than 38°C or less than 36°C, white blood cell count more than 12,000 cells/mm 3 or less than 4,000 cells/mm 3 (or > 10% immature forms), systolic arterial pressure less than 90 mm Hg or a reduction of greater than 40 mm Hg from baseline systolic pressure persisting despite fluid resuscitation and at least one of the signs of organ dysfunction. Organ dysfunction was defined as a PaO2/FIO2 ratio less than 250; creatinine more than 2.0 or the need for dialysis to regulate serum potassium and bicarbonate; a bilirubin more than 2.0 or elevation of transaminases to twice normal; Glasgow Coma Scale score less than 9; platelet count less than 50,000 or an elevation in the prothrombin time to twice normal in the absence of anticoagulation. 14 In addition, all patients had vasodilatory hypotension defined as a cardiac index more than 2.5 L/min and a mean arterial pressure (MAP) less than 70 mm Hg, despite the use of conventional vasopressor agents (dopamine > 3 μg/kg per min and any dose of norepinephrine and/or phenylephrine) initiated after fluid replacement. Because several patients were referred into the study postvasopressor initiation, a minimal pulmonary artery capillary wedge pressure of 12 mm Hg at study entry was chosen to reflect that at least adequate attempts had been made at volume replacement before initiation of catecholamine agents. Pregnant females, patients with electrocardiographic or enzymatic evidence of ongoing myocardial ischemia or infarction, and those with suspected mesenteric ischemia/infarction were excluded.

Back to Top | Article Outline


Randomization and allocation of treatment in eligible subjects was achieved by a computer-generated list created by the hospital pharmacy. In a double-blinded manner, vasopressin (Pitressin injection USP, 8-arginine vasopressin; Parke-Davis) or placebo (normal saline) was administered to subjects through a central vein at a constant infusion rate of 6 mL/h (vasopressin concentration of 0.04 U/min). Intravenous fluids were not changed the hour before or until 1 hour after the study drug infusion began. Attempts to taper and discontinue the dose of conventional vasopressor agents began 1 hour after initiation of the study drug. The protocol specified that norepinephrine, phenylephrine, and/or dopamine infusions were to be decreased in dose decrements of 2 μg/min, 25 μg/min, and 3 μg/kg per min, respectively. These drugs were weaned and discontinued in this sequential order provided that the MAP remained more than 70 mm Hg. Fluid expansion was continued throughout this period and administered according to the pulmonary capillary wedge pressure that was needed to attain the individualized optimal cardiac index. All other concomitant medical treatments were also accepted.

Back to Top | Article Outline

Measurements and Data Collection

All patients had an indwelling arterial line to assess MAP. The electrocardiogram was also monitored throughout the duration of the study. Each patient had a pulmonary artery catheter to guide volume replacement and measure cardiac output. Cardiac output was determined by the thermodilution technique and performed in triplicate. All values were referenced to the level of the right atrium and measured at end-expiration. Hemodynamic parameters and vasopressor doses were recorded at baseline and 1, 4, 8, and 24 hours after inclusion into the study. Systemic vascular resistance and oxygen delivery were calculated according to standard formulas. Serum sodium, base deficit, and plasma creatinine values were measured before study entry and at study completion. The Acute Physiology and Chronic Health Evaluation II score 15 and number of failing organs were used to measure disease severity.

The primary end point of this study was hemodynamic response as defined by an increase in arterial blood pressure. A secondary end point of interest was time to catecholamine pressor-free hemodynamic stability. This secondary end point was defined as a MAP more than 70 mm Hg for more than 30 minutes in the absence of any known vasopressor agent except the study drug and/or a low-dose dopamine infusion (3 μg/kg per min). Untoward events, including bradycardia, arrhythmias, myocardial and/or mesenteric ischemia or infarction, and deaths, were monitored and reported throughout the study.

Back to Top | Article Outline

Statistical Analysis

Hemodynamic variables before and 1 hour after study drug administration were compared by using the Student’s paired t test. Hemodynamic variables and vasopressor dosages between groups were compared by using a two-way analysis of variance for repeated measures. Changes within groups over time were compared with baseline values by using a one-way analysis of variance for repeated measures. Differences in laboratory values attained at 24 hours were compared with values at baseline by using a paired t test.

Back to Top | Article Outline


Ten patients were enrolled in this study and randomized to receive vasopressin (n = 5) or placebo (n = 5). Subject demographics and disease characteristics at entry into the study are displayed in Tables 1 and 2. Groups were similar regarding overall demographics, disease severity, and infectious etiology. Despite adequate volume replacement (pulmonary artery wedge pressure >12 mm Hg) and catecholamine administration, hemodynamic data before study inclusion revealed persistent hypotension attributable to low systemic vascular resistance. Despite randomization, patients allocated to the placebo group had higher baseline pulmonary capillary wedge pressures and significantly greater cardiac indices than the treatment group (Table 3). However, placebo subjects also had a greater base deficit at study entry (see Table 6), thereby eliminating adequacy of fluid resuscitation as a cause for the differences in baseline cardiac indices.

Table 1

Table 1

Table 2

Table 2

Table 3

Table 3

One hour after initiation of the study drug, systolic arterial blood pressure and MAP significantly increased in the treatment group. This increase in blood pressure was attributable to an increase in systemic vasoconstriction (Table 4). Cardiac index, heart rate, and pulmonary artery pressures were relatively unchanged with this increase in arterial pressure. After 1 hour of therapy, patients receiving placebo did not experience any meaningful differences in hemodynamic parameters (Table 4).

Table 4

Table 4

Before attainment of the secondary study end point, measured at 24 hours, two patients in the placebo group died (one patient at 8 hours and one patient at 18 hours). Despite maximum doses of standard catecholamines, both of these deaths were attributable to refractory hypotension (MAP < 50 mm Hg). Therefore, because of the significant reduction in sample size in the placebo group (n = 3), no valid statistical testing could be performed to assess the secondary end point of reduction and withdrawal of conventional catecholamine vasopressors (Table 5).

Table 5

Table 5

Vasopressin administration yielded an observed reduction in conventional vasopressor requirements. At 24 hours, all standard vasopressor agents in the treatment group were able to be withdrawn. However, attempts to discontinue vasopressin in four of the five patients in the treatment group were unsuccessful. The mean arterial pressure in these four patients promptly decreased to less than 70 mm Hg and required reinstitution of vasopressin at 0.04 U/min. Implementation of vasopressin resulted in the expected increase in arterial pressure. At 24 hours, only one patient in the treatment group was able successfully to have all vasopressors, including vasopressin, discontinued and the MAP maintained to more than 70 mm Hg. With vasopressin usage, no significant differences were observed in heart rate, cardiac index, and/or pulmonary artery pressures (Table 5).

Within each group, there were no significant differences noted between baseline laboratory values and values achieved at 24 hours. Because serum sodium was unaffected by vasopressin administration, the known antidiuretic effect of the hormone was not believed to be responsible for the increase in blood pressure. Serum base deficit (reflecting adequacy of fluid resuscitation) and plasma creatinine (depicting an indirect response of the kidneys to vasoconstriction) remained relatively unchanged by vasopressin administration (Table 6).

Table 6

Table 6

Despite continued fluid resuscitation, lower serum creatinine values and higher baseline pulmonary capillary wedge pressures and cardiac indices, the placebo group had greater base deficits both at baseline and at 24 hours when compared with those subjects in the treatment group. Although speculative, the continued base deficit may have been attributable to persistent end-organ hypoperfusion associated with the use of high-dose conventional catecholamine vasopressor agents.

Because of the severity of illness and short duration of this study, patient survival was not an established end point in the initial design of this investigation. However, an important clinical trend was observed. Before termination of the study (24 hours after enrollment), two of five subjects in the placebo group died, whereas all subjects receiving vasopressin survived the 24-hour study period. Both of the deaths in the placebo group were attributable to refractory hypotension (MAP < 50 mm Hg). Throughout the study, there were no observed adverse cardiac events or episodes suggestive of mesenteric ischemia, infarction, or both.

Back to Top | Article Outline


Arginine vasopressin is an endogenous peptide hormone secreted by the neurohypophysis in response to an increase in serum osmolality or a decrease in plasma volume. 16,17 In vitro, it is a more potent vasoconstrictor than angiotensin II or norepinephrine on a molar basis. 18 However, vasopressin is a most intriguing vasopressor hormone in vivo, because it has little pressor activity in normal subjects. 17,19–21 Similarly, patients with the syndrome of inappropriate antidiuretic hormone (SIADH) are not predisposed to hypertension. 16 However, vasopressin antagonists cause marked hypotension in subjects with arterial underfilling; thus, the vasoconstrictor action of vasopressin seems important when intravascular volume or arterial pressure is threatened. 17,18,22,23 Nonetheless, administration of vasopressin to hypotensive or volume-depleted subjects does not result in a marked pressor response, perhaps because V-1 (vasopressin) receptors on vascular smooth muscle are already occupied by endogenous hormone released by baroreflex. 16,17,24–26

Recently, we discovered that patients in vasodilatory septic shock are deficient in vasopressin because of a defect in the baroreflex-mediated secretion of the hormone. 12 Furthermore, administration of low doses of the exogenous hormone yielded a hypersensitive pressor response. 11,12 This study was retrospective and formed the basis for the present controlled clinical trial.

The results of this current prospective study further supports the pressor sensitivity to vasopressin in patients with vasodilatory septic shock. Although knowledge remains limited, sympathetic nerve function seems to be impaired during septic shock. 27 Given that patients administered vasopressin in both our current (Table 4) and previous studies 11,12 did not experience the bradycardia normally observed with an elevation in arterial pressure supports this theory of autonomic insufficiency during septic shock.

Vasopressin is also documented to potentiate the vasoconstrictor actions of conventional catecholamine vasopressors. 28 Because all patients in our study were receiving catecholamine agents, the increase in arterial pressure may have been a synergistic drug effect. However, arterial pressure was still maintained once other vasopressors were withdrawn. Perhaps vasopressin interacted with the elevated endogenous levels of circulating catecholamines that are known to occur during septic shock. 29

An impairment in the release of vasopressin may be yet another possible mechanism to explain this pressor sensitivity to such small doses of the hormone. We have reported that endogenous vasopressin levels are inappropriately low in clinically septic patients. Administration of an exogenous infusion not only increased arterial pressure but appropriately raised serum vasopressin levels in these patients. 12

Finally, desensitization or down-regulation of the catecholamine alpha-1 adrenergic receptors may develop during septic shock 30 and explain the development of refractory hypotension to standard catecholamine vasopressors. Although both vasopressin and the catecholamines use the same phosphatidylinositol second-messenger system, 31 different receptors mediate the vascular vasoconstriction and concomitant increase in arterial blood pressure. Because vasopressin binds to its own V1 vascular receptor, it may assist in restoring peripheral vascular tone if decreased synthesis, increased degradation, or modification of the catecholamine alpha-1 adrenergic receptors yields refractoriness to conventional agents.

Although the mechanism is unknown and remains to be fully elucidated, our present results in conjunction with our previous reports suggest that a continuous low-dose vasopressin infusion (0.04 U/min) increases arterial blood pressure in patients with septic shock. Unfortunately, because of the reduced sample size in the placebo group at 24 hours (n = 3) secondary to death, we were unable to determine the impact of vasopressin in reducing the need for standard catecholamine agents. A larger sample size will be needed to more objectively assess this secondary end point. More importantly, future studies should be undertaken to further investigate the clinical trend of decreased mortality that may be associated with the use of vasopressin in septic shock. Notwithstanding, this current study confirms our previous reports regarding the usefulness of a low-dose vasopressin infusion in the treatment of vasodilatory septic shock.

Back to Top | Article Outline


1. Parillo JE. Septic shock in humans: advances in the understanding of pathogenesis, cardiovascular dysfunction and therapy. Ann Intern Med. 1990; 113: 227–242.
2. Desjars P, Pinaud M, Potel G, Tasseau F, Touze MD. A reappraisal of norepinephrine therapy in human septic shock. Crit Care Med. 1987; 2: 134–137.
3. Meadows D, Edwards JD, Wilkins RG, Nightingale P. Reversal of intractable septic shock with norepinephrine therapy. Crit Care Med. 1988; 16: 663–666.
4. Meadows D, Chernow B, Roth BL. Pharmacologic manipulation of the peripheral vasculature in septic shock: clinical and experimental approaches. Circ Shock. 1986; 18: 141–155.
5. Tsuneyoshi I, Kanmura Y, Yoshimura N. Nitric oxide as a mediator of reduced arterial responsiveness in septic patients. Crit Care Med. 1996; 24: 1083–1086.
6. Houston ML, Thompson WL, Robertson D. Shock: diagnosis and management. Arch Intern Med. 1984; 144: 1433–1439.
7. Chernow B, Rainey TG, Lake CR. Endogenous and exogenous catecholamines in critical care medicine. Crit Care Med. 1982; 10: 409–416.
8. Fink MP, Homer LD, Fletcher JR. Diminished pressor response to exogenous norepinephrine and angiotensin II in septic unanesthetized rats: evidence for a prostaglandin-mediated effect. J Surg Res. 1985; 38: 335–342.
9. Taguchi H, Heistad DD, Chu Y, Rios CD, Oobushi H, Farachi F. Vascular expression of inducible nitric oxide synthase is associated with activation of Ca++ dependent K+ channels. J Pharmacol Exp Ther. 1996; 279: 1514–1519.
10. Parillo JE. Pathogenetic mechanisms of septic shock. N Engl J Med. 1993; 328: 1471–1477.
11. Landry DW, Levin HR, Gallant EM, et al. Vasopressin pressor hypersensitivity in vasodilatory septic shock. Crit Care Med. 1997; 25: 1279–1282.
12. Landry DW, Levin HR, Gallant EM, et al. Vasopressin deficiency contributes to the vasodilation of septic shock. Circulation. 1997; 95: 1122–1125.
13. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med. 1992; 20: 864–873.
14. Marshall JC. Multiorgan failure. In: Wilmore DW, Brennan MF, Harken AH, Holcroft JW, Meakins JL, eds. Scientific American Medicine. New York, NY: Scientific American Inc; 1993: 13–21.
15. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985; 13: 818–829.
16. Jackson EK. Vasopressin and other agents affecting the renal conservation of water. In: Gilman AG, Hardman JG, Limbird LE, eds. Goodman and Gilman’s The Pharmacological Basis of Therapeutics. New York, NY: McGraw Hill; 1996: 715–731.
17. Wagner HW Jr, Braunwald E. The pressor effect of the antidiuretic principle of the posterior pituitary in orthostatic hypotension. J Clin Invest. 1956; 35: 1412–1418.
18. Reid IA, Schwartz J. Role of vasopressin in the control of blood pressure. In: Martini L, Ganong WF, eds. Frontiers in Neuroendocrinology. New York, NY: Raven Press; 1984: 177–197.
19. Ohnishi K, Sato S. Effects of vasopressin on left gastric venous flow in cirrhotic patients with esophageal varices. Am J Gastroenterol. 1990; 85: 294–295.
20. Morton JJ, Padfield PL. Vasopressin and hypertension in man. J Cardiovasc Pharmacol. 1986; 8: S101–S106.
21. Ebert TJ, Crowley AW Jr, Skelton M. Vasopressin reduces cardiac function and augments cardiopulmonary vascular resistance increases in man. J Clin Invest. 1986; 77: 1136–1142.
22. Mohring J, Glanzer K, Maciel JA, et al. Greatly enhanced pressor response to antidiuretic hormone in patients with impaired cardiovascular reflexes due to idiopathic orthostatic hypotension. J Cardiovasc Pharmacol. 1980; 2: 367–376.
23. Huch KM, Runyan KR, Wall BM, Gavras H, Cooke R. Hemodynamic response to vasopressin during V1-receptor antagonism in baroreflex-deficient subjects. Am J Physiol. 1995; 268: R156–R163.
24. Wolfer RS, Lovell NH, Brunner MJ. Exogenous vasopressin does not enhance carotid baroreflex control in the dog. Am J Physiol. 1994; 266: R1510–R1516.
25. Robinson JL. Effect of vasopressin and phenylephrine on arterial pressure and heart rate in conscious dogs. Am J Physiol. 1986; 251: H253–H260.
26. Cowley AW Jr, Monos E, Guyton AC. Interaction of vasopressin and the baroreflex system in the regulation of arterial blood pressure in the dog. Circ Res. 1974; 34: 505–514.
27. Garrad CS, Kontoyannis DA, Piepoli M. Spectral analysis of heart rate variability in the sepsis syndrome. Clin Auton Res. 1993; 3: 5–13.
28. Bartelstone HJ, Nasomyth PA. Vasopressin potentiation of catecholamines in dog, rat, cat, and rat aortic strip. Am J Physiol. 1965; 208: 754–762.
29. Benedict CR, Rose JA. Arterial norepinephrine changes in patients with septic shock. Circ Shock. 1992; 38: L165–L172.
30. Roth BL, Spitzer JA. Altered hepatic vasopressin and α-1 adrenergic receptors after chronic endotoxin infusion. Am J Physiol. 1987; 252: E699–E702.
31. Bendge MJ. Inositol triphosphate and diacylglycerol as second messengers. J Biochem. 1984; 220: 345–360.
Back to Top | Article Outline


Dr. Edward Cornwell (Washington, DC): I would like to congratulate Dr. Malay and colleagues on their work and to thank her for the timely submission of her manuscript.

Down-regulation of adrenergic receptors has long been postulated to contribute to the progressive hemodynamic instability seen in septic shock. In this well-done and well-presented study, Dr. Malay and colleagues assess two parameters in a double-blinded placebo-controlled fashion.

The first was the early hemodynamic response to vasopressin infusion in patients with septic shock and vascular collapse. The second was the augmentation in arterial blood pressure that would allow weaning of other vasoactive drugs. By utilizing hemodynamic data obtained in earlier work as a basis for their statistical power analysis, the authors provide justification for what would otherwise seem to be an inadequate sample size.

I congratulate the authors for the logical progression of their scientific investigation and have several questions regarding their treatment philosophies.

Your eligibility criteria include hypotension requiring adrenergic therapy after “aggressive” fluid resuscitation, defined as a pulmonary capillary wedge pressure greater than 12 mm Hg. Given the well-described myocardial dysfunction associated with septic shock, is it a possibility that the patients’ preloads were inadequate to sustain the cardiac index necessary to avoid vasopressor therapy? Simply put, is fluid resuscitation to a wedge in the low teens before initiating vasoconstrictor therapy truly aggressive?

The second question refers to Table 3 in your manuscript, the slide that was labeled “Baseline Hemodynamic Data,” and for the benefit of the audience that does not have the manuscript, I will walk through this slowly.

Inspection of your baseline hemodynamic data and that attained 1 hour after therapy gives interesting insight. Your pretreatment mean cardiac index was significantly higher in the placebo group, 5.6 versus 4.3 L/min per m2. Now, there was also a trend toward high pulmonary artery occlusion pressures, 20 versus 16, which did not achieve statistical significance, but this may possible be due to a sample size which, while adequate to assess the blood pressure, the magnitude of blood pressure changes that you had previously described may be insufficient to analyze the difference between preload in the two groups.

On inspection of these data and heart rates, which you also provide, again, in Table 4 in your manuscript, we can do the calculation and note that your stroke volume index, your mean stroke volume indices were 40 to 50% higher during the first hour of study in your placebo group. Given that, is it possible that with higher wedges, higher cardiac indices, higher stroke volume indices, that you have identified a group of septic shock patients who are amenable to preload recruitable increases in cardiac index and, therefore, in oxygen delivery?

If this is the case, then your methodology of not changing IV fluids until 1 hour after drug infusion may have the effect of depriving the placebo group of fluid challenges and of the more effective vasoconstrictor. Indeed, by 24 hours, the cardiac index of the three surviving placebo patients decreased to the levels seen in the vasopressor infusion group.

Ohm’s law tells us that pressure is proportional to flow times resistance, and you have nicely shown that vasopressin is superior to placebo in augmenting resistance in this group of patients. It, therefore, follows that without concomitant augmentation of flow as measured by the cardiac index, the placebo group would expect to demonstrate a progressive drop in blood pressure.

So, my final set of questions refers to your practice. Since you may have identified a more effective vasoactive agent in septic shock, how has it affected your current practice in your ICU? Specifically, once you substitute one vasoactive agent for another, are there other end points of resuscitation that you follow?

Is your protocol for ultimately weaning patients from vasopressin substantially different from that of other adrenergic agents?

In summary, while blood pressure is a poor parameter by which to gauge adequacy of resuscitation, we have all cared for patients who are so hemodynamically unstable as to require vasoactive agents.

I congratulate the authors for their continued work in addressing the specific defect in patients with septic shock and look forward to their views on the overall role of this specific therapy. And I would like to thank the Association for the opportunity to discuss the manuscript.

Dr. John H. Siegel (Newark, New Jersey): Thank you very much for presenting an interesting and challenging paper.

There are some questions that I have. Patients with very severe septic shock often have defects in oxygen consumption related to the fact that they have a narrow AV difference that is disproportionate. They also tend to run metabolic insufficiency levels and have frequently had increased levels of lactate. This could be exacerbated by agents like norepinephrine or epinephrine, which increase glucose metabolism.

So, I would like to ask you whether the patients that you had showed evidence that their oxygen consumption levels actually increased, and that they did not have an increase in lactate or the lactate fell?

And, finally, in some studies, vasopressin has been associated with coronary vasoconstriction, and I wonder whether you have any evidence of that or whether you feel there is a class of patients that should be excluded from this type of therapy? Thank you.

Dr. John Porter (Oakland, California): Vasopressin is a potent vasoconstrictor of the splanchnic circulation, and that is why it has its use in GI bleeding with varices. It would be interesting to note what the pHi would do in these patients, because you are probably constricting the splanchnic circulation. And some would suggest that, although the blood pressure was better and SVR was better, if you had splanchnic vasoconstriction, that the patient might do worse.

Dr. Mary Beth Malay (closing): I would like to thank Dr. Cornwell for his kind and insightful comments. In response to his question regarding our requirement that the pulmonary artery wedge pressure be at least 12 mm at study entry, we certainly agree that there is no absolute value of the pulmonary artery wedge pressure that universally determines adequacy of fluid resuscitation. However, given that several of the study patients were referred post-vasopressor initiation, we had to determine a minimum value of the pulmonary artery wedge pressure that suggested attempts at volume replacement had been made before initiation of vasopressor therapy.

Second, your insight into a potential selection bias in the placebo group is appreciated. Although all patients were randomized prior to study entry, it is a remote but real possibility that patients in the placebo group actually may have been in a more hyperdynamic phase of sepsis, as evidenced by their higher pretreatment cardiac indices and lower systemic vascular resistances. This is unlikely, because further analysis reveals that despite higher cardiac indices and higher pulmonary capillary wedge pressures, the base deficit was persistently greater in the placebo group versus the treatment group.

Furthermore, although we withheld intravenous fluids from the patients in the placebo group during the initial hour of study entry and, therefore, may have deprived this group of fluid (the more effective vasopressor), we would have expected a decline in mean arterial pressure if the patients in the placebo group were truly dependent on volume replacement to maintain an adequate cardiac index and mean arterial pressure.

Currently, after septic patients in our ICU are adequately volume-resuscitated (defined as the individualized pulmonary capillary wedge pressure yielding the optimal cardiac index), low-dose vasopressin is initiated if hypotension persists. Simply stated, we only institute conventional catecholamines if volume and/or vasopressin fail to increase arterial pressure. Due to our favorable results, this treatment scheme is now occurring in several ICUs throughout our institution.

Given the low dose of vasopressin used, which by standard drug concentrations translates into an infusion rate of 6 mL/hour, the drug is not weaned but is simply discontinued once the target blood pressure is attained. We have witnessed patients dramatically drop their blood pressure minutes after stopping the infusion. If this occurs, the drug is simply restarted and the same increase in arterial pressure is observed. Eventually, all patients are able to have the infusion discontinued and their blood pressure maintained.

Dr. Siegel, we did not measure serum lactate levels, but we did monitor serum base deficits. The treatment group had both lower baseline and 24-hour base deficits than did patients in the placebo group. It is only speculative that continued catecholamine vasopressors in the placebo group contributed to greater base deficits.

We know that vasopressin inhibits nitric oxide-induced accumulation of cyclic GMP; we also know that in the setting of persistent lactic acidosis, cyclic GMP levels are high and likely contribute to the peripheral vasodilation and subsequent hypotension of septic shock.

It is well documented that pharmacologic doses of vasopressin are associated with significant coronary and mesenteric ischemia. Since only one-tenth of the standard dose of vasopressin was needed to increase arterial pressure, we did not observe EKG evidence of arrhythmias and/or myocardial ischemia.

We share Dr. Porter’s concern regarding splanchnic vasoconstriction and the development of mesenteric ischemia associated with the use of vasopressin. Before implementation of this current study, we examined the dose-dependent effects of vasopressin using a septic porcine animal model. We found that at subpharmacologic doses of vasopressin ranging from 0.04 to 0.12 U/min, mesenteric blood flow was not significantly decreased, despite a marked increase in arterial pressure. Additionally, we did not observe any clinical evidence of mesenteric ischemia in any of the patients in this human study.

I would like to thank the Association for the privilege of presenting our paper at this forum.


Hypotension; Sepsis; Low-Dose vasopressin

© 1999 Lippincott Williams & Wilkins, Inc.