Reversible anuric acute kidney injury secondary to acute renal autoregulatory dysfunction

Abstract

Autoregulation of glomerular capillary pressure via regulation of the resistances at the afferent and efferent arterioles plays a critical role in maintaining the glomerular filtration rate over a wide range of mean arterial pressure. Angiotensin II and prostaglandins are among the agents which contribute to autoregulation and drugs which interfere with these agents may have a substantial impact on afferent and efferent arteriolar resistance. We describe a patient who suffered an episode of anuric acute kidney injury following exposure to a nonsteroidal anti-inflammatory agent while on two diuretics, an angiotensin-converting enzyme inhibitor, and an angiotensin receptor blocker. The episode completely resolved and we review some of the mechanisms by which these events may have taken place and suggest the term “acute renal autoregulatory dysfunction” to describe this syndrome.

• Acute kidney injury
• autoregulation
• glomerular hemodynamics

Introduction

The approach to the patient with acute kidney injury (AKI) involves a process with the aim of characterizing it as being due to prerenal, intrinsic renal, or postrenal factors. Autoregulation of glomerular capillary pressure (Pgc) acts to attenuate the impact of decreased renal perfusion pressure which would otherwise lead to a reduction in the glomerular filtration rate (GFR). While a prerenal state can cause a decrease in GFR, complete anuria would be extremely unlikely to occur. Various medications, both over the counter and prescription, can interfere with autoregulation and potentially reduce Pgc and the GFR. The combination of a diuretic, a nonsteroidal anti-inflammatory drug (NSAID), and angiotensin-converting enzyme inhibitor (ACEI) or angiotensin receptor blocker (ARB) may be particularly disruptive to autoregulation. Here we describe a patient who developed anuric AKI on this drug combination with complete recovery and review mechanisms for autoregulation and how this may become disrupted.

Case

A 53-year-old woman with type 2 diabetes mellitus and hypertension was admitted to the ambulatory surgery department at 0700 h for elective total knee replacement at 1200 h. She took her usual medications up to the day prior to surgery and at 0400 h on the day of surgery, which included lisinopril (20 mg daily), olmesartan (40 mg daily), hydrochlorothiazide (25 mg daily), and furosemide (40 mg daily). Her serum creatinine was 1.1 mg/dL two weeks prior and also on the morning of surgery. On admission her blood pressure (BP) was 132/76 mmHg, heart rate 84/min, the lungs were clear, and there was no edema. At 1000 h, she was administered 200 mg celecoxib. Surgery was initiated at 1215 h with intrathecal hydromorphone anesthesia and a urinary catheter was inserted. Left total knee replacement proceeded uneventfully with estimated blood loss of 50 cc and a stable mean arterial pressure of approximately 75 mmHg throughout surgery. However, her urine output was noted to be <10 cc/h, and 1500 cc of isotonic saline was administered with no increase in urine output. At 1500 h, she was brought to the recovery room and was now noted to be anuric. Her creatinine was again 1.1 mg/dL at this time. Renal ultrasound revealed no hydronephrosis and normal renal arterial and venous blood flow. Isotonic saline was continued at 150 cc/h and 80 mg of furosemide was administered intravenously followed by an infusion at 10 mg/h but her anuria persisted for 16 h postoperatively, at which time the creatinine had increased to 2.4 mg/dL. Her urine output then increased to >100 cc/h from the 16th to 28th h, and then to >250 cc/h after the 28th h, while the creatinine improved to 1.5 mg/dL at 48 h postoperatively, and then to baseline at 1.1 mg/dL at 60 h postoperatively. Urinalysis was negative for protein, blood, white blood cells or casts though markers of tubular injury such as neutrophil gelatinase-associated lipocalin (NGAL) or kidney injury molecule-1 (KIM-1) were not measured. Forty-eight hours postoperatively her urine osmolality was 629 mOsm/L with a fractional excretion of sodium of 0.15%, implying intact tubulointerstitial function and making acute tubular necrosis (ATN) and interstitial nephritis unlikely. Sixty-eight hours postoperatively her creatinine was 0.9 mg/dL. These events are summarized in Table 1.

Table 1. Summary of events, urine output and serum creatinine levels.

Time	Events	Serum creatinine	Urine output
−0600	Lisinopril 20 mg
	Olmesartan 40 mg
	Hydrochlorothiazide 25 mg
	Furosemide 40 mg
−0200	Celecoxib
0	Surgery		<10 mL/h
0300	Saline	1.1 mg/dL	0 mL/h
	Furosemide
1600		2.1 mg/dL
2000		2.4 mg/dL	>100 mL/h
2800			>250 mL/h
4800		1.5 mg/dL
6000		1.1 mg/dL
6800		0.9 mg/dL

Discussion

AKI may result from prerenal conditions, renal parenchymal disease, or obstruction. Severe, prolonged prerenal states and parenchymal disorders may lead to ATN. These two disorders account for >75% of all AKI in hospitalized patients.1 In our patient, despite almost 24 hours of anuria, she completely recovered normal renal function and did not manifest any signs of renal parenchymal damage. There was also no evidence of any systemic hemodynamic disturbance sufficient to cause such an event. Given the combination of medications she was on, we hypothesize that exposure to the NSAID celecoxib on a background of two diuretics and both an ACEI and ARB resulted in combined afferent arteriolar constriction and efferent arteriolar dilatation causing a profound decrease in Pgc to cause a reversible episode of anuria.

In the absence of such drugs, prerenal conditions and profound hypoxia are countered by compensatory renal autoregulatory mechanisms to preserve renal blood flow (RBF) and maintain pO₂. These protective mechanisms include prostaglandins, nitric oxide, adenosine, and tubuloglomerular feedback, and the fall in GFR which results in the setting of ATN has been termed “acute renal success” given that the decrease in GFR prevents uncontrolled volume losses that would otherwise occur with a sudden impairment of tubular reabsorption of solute and water.2 A normal GFR is the product of several Starling forces present in the glomerular microcirculation, namely, hydrostatic and oncotic pressures within the glomerular capillary and within Bowman’s space. The GFR is kept constant over a broad range of BPs by the effects of angiotensin II on the afferent and efferent arteriole, the effects of prostaglandin E₂, nitric oxide, and adenosine on the afferent arteriole, and other mechanisms. A reduction in blood flow and oxygen to the afferent arteriole is compensated by constriction of the efferent arteriole and dilatation of the afferent arteriole which helps to maintain Pgc and GFR. ACEIs and ARBs impair efferent arteriolar constriction, lowering Pgc and GFR. NSAIDs block afferent arteriole dilatation, preventing inflow to the glomerulus. Fortunately, most individuals with normal cardiac output and renal microcirculation do not require prostaglandin-mediated afferent vasodilatation in the basal state and will maintain a normal GFR while taking NSAIDs. In contrast, the GFR of a patient with low effective arterial volume is dependent upon angiotensin II-mediated efferent arteriolar constriction and prostaglandin-mediated afferent arteriolar dilatation, hence sensitizing them to drugs that disturb these mechanisms. It should be emphasized, however, that these relationships are complex and there are other factors which could also interfere with autoregulation. For example, while nitric oxide plays a physiological role in the regulation of vascular tone, excessive nitric oxide production can interfere with autoregulation and could be a contributing factor in scenarios such as that described here.3

Hypotension is a relative term when describing AKI in a susceptible individual, as described by Abuelo.4 He described AKI occurring in the absence of frank hypotension, in susceptible individuals with preexisting renal microvascular disease, or in patients with an altered microcirculation due to NSAIDs, cyclosporine, tacrolimus, radiocontrast agents, ACEIs, ARBs, hypercalcemia, hepatorenal syndrome, or renal artery stenosis. Abuelo called this “normotensive ischemic acute renal failure”, as renal perfusion pressure drops below the autoregulatory range. Normal renal autoregulation is sustained down to a mean arterial pressure (MAP) of 80 mmHg. However, in patients with impaired autoregulation, the GFR falls even while the MAP remains in the normal range. In contrast to prerenal azotemia, where the reduction in renal function results from the consequences of low effective arterial volume, the present case is an example of a scenario in which the problem appears to be a hemodynamic one at the level of the glomerulus itself, with a reduction in Pgc sufficient to cause anuria with apparently preserved effective arterial volume.

Our patient was hemodynamically stable throughout her uneventful surgery but had several variables present which increased her susceptibility to a decrease in Pgc including diabetic microvascular disease, hypertensive vascular disease, and the concomitant use of an ACEI, ARB, diuretics, and the NSAID that she was given shortly before the procedure. The concurrent use of diuretics, NSAIDs, and ACEIs or ARBs has been referred to as “the nephrotoxic triple whammy”.5,6 Lapi et al. reported a 30% increased incidence of AKI when patients were treated with this combination, especially at the start of treatment.7 Our patient likely had a severe enough perturbation of the renal microcirculation to result in anuria, with blood flow, and pO₂ sufficient to prevent necrosis, but not enough to maintain the GFR8 — a syndrome which perhaps could be described as “acute renal autoregulatory dysfunction” (“ARAD”). Despite the apparent decrease in her Pgc, flow through the efferent arteriole and the vasa recta presumably remained sufficient to maintain tubular integrity and avoid ATN, attested to by the increased urine osmolality and volume, and to the rapid recovery of renal function after the withdrawal of the offending medications and increasing her intravascular volume.

There are a number of limitations which must be kept in mind in attempting to draw conclusions from this case. First, despite the patient starting out with a normal serum creatinine this does not exclude the possibility that her renal function had declined sooner than was detected with serum creatinine levels. Cystatin C or urinary markers of tubular injury might have been able to assess this. It should also be noted that a surgical procedure, even if overtly uneventful, may induce numerous changes in levels of various cytokines and inflammatory mediators that can also impact upon renal function aside from the effect of medications. Hence, it needs to be emphasized that this episode of AKI with these three classes of drugs took place in the context of surgery and might not occur in this manner in another clinical scenario.

In conclusion, the patient we have discussed is an example of a scenario that could best be described as a profound, pharmacologically induced marked decrease in Pgc due to disrupted autoregulation sufficient to cause sustained but completely reversible anuria in the setting of surgery. We suggest using the term “ARAD” to describe rapidly reversible AKI which appears to be due to factors which derange autoregulation of Pgc at the level of the glomerulus and without systemic hemodynamic abnormalities altering renal perfusion. Greater awareness of this syndrome and the role of medications in inducing it will hopefully reduce the incidence of this event, perhaps by lowering the dose or temporarily withholding such agents during procedures or intercurrent illnesses.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.