The purpose of this study is to determine the effect of a systemic infusion of angiotensin II on haemodynamics and urine output in critically ill patients with severe sepsis/septic shock and acute renal failure.

It will also help determine the feasibility of conducting a definitive and adequately powered randomised controlled trial of angiotensin II in such patients that would assess mortality and need for renal replacement therapy as endpoints.

Sepsis is the most common cause of ARF in the ICU. In last 50 years there have been no significant advances in our understanding of the pathogenesis, prevention or treatment of septic ARF, except for the use of renal replacement therapy (RRT) once it is established.

It has been assumed that hypotension induced by severe sepsis results in organ hypoperfusion and subsequent kidney ischaemia. This ischaemia has been believed to be one of the main factors, if not the principal factor, contributing to development of ARF in severe sepsis. Surprisingly, there is little evidence to support this assumption. Rather, emerging evidence seriously questions this traditional ischaemic-acute tubular necrosis (ATN) paradigm of septic ARF. Patients with severe sepsis have been found to have increased, rather than decreased, renal blood flow, and post mortem examination of kidneys from patients who have died with septic acute renal failure rarely show the appearance of ATN. An animal model of septic ARF found that renal blood flow was increased, while glomerular filtration rate was decreased.

These facts lead us to hypothesise that profound efferent arteriolar vasodilatation may be the cause of the observed decrease in GFR in septic ARF. The only logical explanation for the observation that RBF increases while GFR falls is that both efferent and afferent arterioles dilate, but that efferent vasodilation is greater. A selective efferent arteriolar vasoconstrictor would be expected to restore GFR. Angiotensin II is the most selective known efferent vasoconstrictor.

We hypothesise that early therapeutic intervention with angiotensin II in critically ill patients with severe sepsis/septic shock and kidney dysfunction may improve kidney function such that the need for renal replacement therapy is avoided. This would represent a significant improvement in the care of critically ill patients with severe sepsis/septic shock and ARF, a condition for which no interventions short of RRT have been shown to improve outcome. Novel and successful therapeutic interventions in this patient population would have widespread clinical implications, including improved survival and less need for long-term dialysis, with consequent resource savings.

Inclusion Criteria:

• age ≥ 18 years
• within the first 24 hours of ICU admission
• an expected duration of ICU admission of at least 72 hours
• informed consent by patient or by proxy (i.e. next of kin)
• diagnosis of severe sepsis/septic shock
• diagnosis of kidney dysfunction (minimum RIFLE criteria - 'R'); and
• presence of a central venous catheter.

Exclusion Criteria:

• inability to provide or obtain consent;
• patient is moribund with expected death within 24 hours;
• known chronic kidney disease (CKD) or end-stage renal disease (ESRD) receiving chronic RRT;
• confirmed or suspected acute glomerulonephritis, acute interstitial nephritis, renal vasculitis or post-renal aetiology for kidney dysfunction;
• patient is already receiving (or is about to start) CRRT for acute renal failure at the time of enrolment;
• known or documented allergy to angiotensin II;
• MAP consistently > 100 mmHg with no pressor support and no easily treatable cause (eg. pain); and
• enrolling physician's belief that the study drug could not be administered for the expected study duration.