The Role of Renal Functional Reserve in Predicting Acute Kidney Injury

Introduction

First described in 1983 by Bosch and colleagues, the concept of renal functional reserve (RFR) refers to the normal kidney’s ability to increase its filtration rate in response to a stimulator such as a protein load.¹ The numerical difference between a baseline and a protein stimulated glomerular filtration rate (GFR) has been termed RFR. In healthy individuals, an increase in GFR after a protein load can range between 6–40% with a mean increase of 26% after a protein meal depending on the experimental conditions of the study.²

The kidney does not operate constantly at its maximum filtration capacity, but rather at about 75% of maximum GFR.³ An individual may have an apparently normal GFR, but a decreased RFR. A lack of GFR increase in response to a protein load has been shown in individuals with reduced nephron mass as a result of a single kidney,^1,4 type I diabetes mellitus,⁵ high grade vesicoureteral reflux,⁶ hemolytic uremic syndrome,^7–9 obesity,¹⁰ and hypertension.¹¹ Using amino acid infusions and plasma clearance of Tc99m diethylene-triamine pentaacetic acid, Barai and colleagues reported a decline in RFR with progression of chronic kidney disease.¹²

Researchers have proposed that a kidney that has a history of injury may be operating at its maximal filtration capacity and, therefore, have reduced, or no available nephrons to increase GFR in response to a stimulus. Acute kidney injury (AKI) can result in a decrease in the number of functional nephrons in the kidney, whereby GFR measurements may not accurately represent the degree of structural injury. This can lead to an adaptive response such as glomerular hyperfiltration in order to maintain homeostasis. An increasing desire by investigators to uncover the silent loss of nephrons that can occur with AKI has led to a renewed interest in finding easily replicated, practical methods for quantifying RFR.

Postulated Mechanisms for RFR

Many mechanisms have been discussed in the literature regarding the changes in GFR that occur in response to a protein load with no single agreed upon process (Figure 1). The most accepted mechanisms for a renal response to a protein load include the involvement of humoral mediators and resetting of tubuloglomerular feedback. It has been proposed that the mechanism may vary based on the disease process.¹³ Animal models support an interplay of hemodynamic and structural changes. Simultaneous changes in renal plasma blood flow with changes in GFR in response to increases in dietary protein have been demonstrated. Interestingly, rats fed high protein diets have shown an increase in kidney size.¹⁴

Figure 1:

Multiple mechanisms are thought to explain the increase in glomerular filtration rate that occurs in response to a protein load. The most commonly cited mechanisms include the involvement of humoral mediators and the resetting of tubuloglomerular feedback resulting in an increase in renal plasma blood flow.

The role of the deactivation of tubuloglomerular feedback in the GFR response to a protein load has been shown in animal models. In dogs using lithium clearance, Wood et al. reported an increase in proximal tubular transport mediated by the sodium amino acid co-transport system after intravenous infusion of amino acids.¹⁵ This decreased distal sodium chloride delivery results in a deactivation of tubuloglomerular feedback and an accompanying increase in GFR. Likely there is a systemic effect of protein consumption, rather than simply a local effect acting on the kidney. Premen and colleagues infused serine, alanine, and proline into the intrarenal artery of dogs that led to a rise in renal blood flow when measured with p-aminohipuric acid, but did not increase GFR.¹⁶ However, an intravenous infusion of these amino acids led to significant elevations in renal blood flow and GFR, supporting the theory that a secondary systemic factor is likely needed before amino acids can cause a change in GFR.

The release of humoral factors at the time of protein consumption has been thought to explain the increase in GFR.^17–19 Several studies’ results in both rats and humans support the role of glucagon in the renal response to a protein load by showing the inhibition of a renal hemodynamic response to amino acid infusion by somatostatin administration.^20,21 However, the increase in GFR with glucagon administration has been found to be moderate, suggesting that glucagon likely plays a facilitative role in the change in renal hemodynamics with amino acid infusion, rather than being the primary factor.²² Insulin-like growth factor-1 has been shown to be higher in the glomeruli and liver of rats fed a high verses low protein diet.²³

Investigators have explored the role of vasoactive mediators such as renal kallikrein and vasoactive kinins in the renal response to protein ingestion.²⁴ In humans, Bolin and colleagues demonstrated that an oral protein load increases GFR concurrently with the urinary excretion of kinin.²⁵ In a cohort of patients with hypertension, Pecly and colleagues reported a lower RFR in obese patients when compared to patients without obesity.¹⁰ They found a lower increase in urinary kallikrein and an inability to elevate serum nitric oxide levels in the patients with obesity.

Ruilope et al. reported evidence of the influence of angiotensin II and prostaglandin in the renal hemodynamic effects of amino acid infusions.²⁶ Healthy human volunteers who received a low sodium diet three days prior had a lack of increase in renal blood flow and GFR when given an amino acid infusion. However, the renal blood flow and GFR response with amino acid infusion was restored with treatment with an angiotensin-converting enzyme (ACE) inhibitor. They also reported in healthy human subjects a decrease in the change in renal blood flow and GFR with amino acid infusion in response to indomethacin administration. Investigations using nonsteroidal anti-inflammatory drugs (NSAIDS) to block prostaglandin synthesis have produced conflicting results regarding their impact on RFR. Some authors have reported a loss of RFR with NSAID administration and others demonstrated a normal RFR after NSAID use.^18,19,27

Methods Used to Quantify RFR

The percent change in GFR after a protein load varies from study to study depending on the type of protein used, dose of protein given or other differences in experimental conditions.² Investigators commonly advise subjects to maintain a diet free of meat, fish, and fowl for 24 hours prior to testing. The majority of studies use an oral red meat protein load.^1,4,10 Researchers have also reported significant changes in GFR after giving subjects dairy products, egg white proteins, or baked goods.²⁸ Compared to other oral methods of protein loading, the use of beef has induced the largest response.

Investigators have employed methods other than an oral protein load to elicit a change in GFR. The intravenous infusion of amino acids has been used;^12,29 however, not all amino acids elicit the same response. In rats the infusion of non-branched chain amino acids has been shown to increase GFR, whereas the branched chain amino acid, leucine, does not modify GFR.³⁰ Amino acid infusions show a faster GFR response when compared to an oral protein load on average (30 to 60 minutes versus 60 to 180 minutes).^1,31 A dopamine infusion has been used in some studies.³¹ The mechanism whereby dopamine increases GFR is thought to be different from a meat meal or the use of amino acids. Experimental results have shown a fall in filtration fraction with a dopamine infusion as a result of a coinciding greater increase in renal plasma blood flow when compared to GFR.³²

The dose of protein given varies from study to study, with the majority of investigators administering 1–2 grams of protein per kilogram of body weight. Rodriguez-Iturbe et al. demonstrated an increased filtration fraction with protein loads of 1.1 and 1.3 grams per kilogram of body weight but not with 0.55 grams per kilogram of body weight.³³ In 18 healthy adult volunteer subjects, Sharma and colleagues tested the effect on GFR of ingesting 1 gram/kilogram versus 2 grams/kilogram of cooked red meat on the GFR response.¹³ They did not report a greater increase in GFR with the larger protein load.

Similar to the type of protein stimulus used, the method used to quantify GFR varies from study to study. Few investigators have reported using inulin or iohexol clearance.^4,34 Comparing creatinine clearance before and after a protein load is the most common method cited in the literature for quantifying RFR.^2,35 An important limitation with the use of creatinine clearance is the secretion of creatinine by renal tubular cells. The clearance of creatinine due to tubular secretion is lower in individuals with a normal GFR when compared to those with a moderately reduced GFR (40–80 mL/min/1.73 m²).³⁶ Given the concern that ingesting food containing preformed creatinine and creatinine precursors could lead to a wide variation in changes in creatinine clearance, some investigators have advocated for the use of milk, cheese and baked goods to stimulate GFR, rather than use of meat products. Hellerstein and colleagues have published on the use of cimetidine to successfully inhibit the tubular secretions of creatinine in RFR studies in children.^37,38 Using cimetidine, they report a very close approximation of inulin clearance and creatinine clearance when they measured the two clearances simultaneously in pediatric patients.

Creatinine clearance studies are time consuming requiring multiple urine collections before and after a protein load, taking on average 6.5–7 hours to complete. Our group has previously shown that changes in cystatin C estimated GFR with cystatin C drawn 125–140 minutes after a protein load may be used to estimate RFR induced by a meat meal in healthy young adults (cystatin C estimated GFR peak vs. baseline: 110.1 vs. 98.1 mL/min/1.73m², p<0.003).³⁹ This method has yet to be validated in any patient groups.

A Renewed Interest in RFR

As a result of renal hyperfiltration, serum creatinine increases only after 50% of the nephrons are lost.⁴⁰ Initially, RFR and baseline GFR may remain intact after an injurious event. However, as a result of multiple renal insults, a patient may become more susceptible to clinical AKI and eventually chronic kidney disease (Figure 2).⁴¹ Patients with chronic medical conditions admitted to the intensive care unit are at a particularly greater risk of kidney function decline with repeat renal insults. Intrigued by the use of RFR to detect the early functional decline in kidney function, investigators recently have focused on the use of RFR as a “stress test” for the kidney to be used to identify those individuals with a normal baseline GFR, but a deficient RFR.⁴¹

Figure 2:

After repeat episodes of acute kidney injury (AKI), a patient may have a deficient renal functional reserve (RFR), yet a normal baseline glomerular filtration rate (GFR). With partial recovery and subsequent renal injury, the patient may become more susceptible to AKI with minor insults, clinically evident AKI and eventually chronic kidney disease. Reprinted from Sharma A, Mucino MJ, Ronco C. Renal functional reserve and renal recovery after acute kidney injury. Nephron Clin Practl. 2014;127(1–4):94–100 by Karger. Reprinted with permission.

Prior to recent years, there have been very few studies describing the association of RFR with outcomes. The use of RFR in children with a history of hemolytic uremic syndrome has been explored, given the uncertainty of future renal health in these patients. Dieguez and colleagues reported that children after hemolytic uremic syndrome with a low response to a protein load (less than a 36% increase) were more likely to develop proteinuria.⁸ Livi et al. showed that in 28 patients with systemic sclerosis without a history of renal disease, lower baseline RFR is associated with a 9.5% decrease in RFR when evaluated five years later, whereas subjects with a higher baseline RFR had a 3.8% decrease in GFR when evaluated at 5 years.⁴²

Importantly, there has been a significant increase in the last 5 years in publications discussing the use of RFR for predicting outcomes. Husain-Syed et al. explored the ability of preoperative RFR to predict AKI within seven days after surgery in 110 patients undergoing elective cardiac surgery requiring cardiopulmonary bypass with a normal resting GFR.⁴³ They stimulated GFR with a 1.2 mg per kilogram of body weight protein load in the form of an oral red meat meal and they calculated RFR using creatinine clearance before and after the protein meal. The investigators report that preoperative RFR was lower in patients who experienced AKI. Additionally, RFR predicted AKI with an area under the receiver operator curve of 0.83 (CI: 0.70–0.96). Patients with a RFR ≤ 15 ml/min/1.73m² were 11.8 times more likely to experience AKI.

The same investigators studied the effect of repeat protein loading in 86 of the original subjects, again using an oral meat protein load and creatinine clearance.⁴⁴ Patients who met Kidney Disease Improving Global Outcomes (KDIGO) AKI criteria or had a rise in cell cycle arrest biomarkers (tissue inhibitor metalloproteinases-2 and insulin-like growth factor-binding protein 7) were more likely to show a decrease in RFR when repeated three months after surgery. No patients without AKI and low postoperative biomarker levels had a decrease in RFR greater than 4.7 ml/min/1.73m².

Limitations to the Routine Use of RFR in Routine Clinical Care

Despite its introduction into the literature over 20 years ago, there is no routinely used method for quantifying RFR in clinical care. The wide variability of study results as a result of differing protocols and patient populations likely is playing a role in the lack of general acceptance of RFR as a routine method to measure kidney function. Importantly, patients of different ages and pathology are frequently grouped together in studies.⁴⁵ A decline in RFR has been shown to occur with age. Studies in humans have demonstrated a decline in GFR of about 0.8 mL/min/1.73 m² per year after the age of 30 years.^46,47 Future investigations should include patients of similar ages and diagnoses.

There is a need for a simple reliable protocol that can be easily replicated. The most commonly cited method for quantifying RFR involves creatinine clearance methods, which are time consuming and have the potential for error due to urine output measurements. In particular, in the pediatric population where patients may not be toilet trained, there is a need for a method that does not rely on urine output determination. The use of other endogenous markers like cystatin C requires waiting for a change in the concentration of the marker to represent a change in GFR after a protein load is administered. Quantifying the disappearance rate of an exogenous marker, such as fluorescent molecules, has been proposed as a method to acquire real-time monitoring of GFR.⁴⁸ The use of real-time monitoring of GFR using fluorescent molecules and transdermal sensors have great promise for evaluating RFR accurately and efficiently.

Skepticism regarding the use of RFR may in part be due to a lack of evidence regarding the impact of RFR on outcomes or potential therapeutic implications of protein administration. Until recent years few researchers have investigated the impact of RFR on outcomes. As discussed above, Husain-Syed reported that in their patient cohort, individuals who met AKI criteria were more likely to a show decrease in RFR when repeated three months after cardiac surgery. Interestingly, no patients in this study met the criteria for AKI using KDIGO urine output criteria.⁴⁴ The authors speculated that there might be a sustained impact of urine output after the protein load. Based on observations supporting the impact of protein on decreasing renal vascular resistance, Pu and colleagues randomized 69 adult patients with estimated GFR values of 20–89 mL/min/1.73m² undergoing cardiac surgery requiring >1 hour of on-pump time to receive a continuous infusion of L-amino acids after anesthetic induction or standard of care.⁴⁹ In the intervention arm, the amino acid infusion was continued until discharge from the intensive care unit. The duration of acute kidney injury defined by the KDIGO criteria was significantly reduced in the patients who received the amino acid infusion. Patients who received the supplementary amino acids demonstrated a significantly greater than baseline estimated GFR when compared to the patients who received standard of care (+10.8% difference, 95% CI, 1.0% to −20%, p=0.033). Additionally urine output was greater in the patient group that received the amino acid infusion.

Future investigations exploring therapeutic considerations if a lower RFR is found are important. Based on studies in animal models, ACE inhibitors and angiotensin II receptor blockers (ARB) medications have the potential to restore the response to a protein load in hyperfiltering states, such as diabetes and hypertension.^26,29 Given the proposed beneficial effects of ACE inhibitors and ARB medications in reducing intraglomerular pressure and proteinuria, an agreed upon method to quantify RFR could lead to future investigations on the impact of these medications in patients with a declining RFR.^50–52

Conclusion

We can be falsely reassured by a normal GFR value when caring for the patients in the intensive care unit, particularly in the setting of decreased muscle mass or hyperfiltration. Future investigations should continue to explore the use of RFR in patients before kidney donation or prior to procedures carrying a higher risk of renal insult. Investigators are beginning to study the use of RFR for assessing renal recovery after AKI. There stands to be a benefit to quantifying RFR in certain patient groups, given the evidence that our patients may have a normal GFR, yet a decline in RFR associated with a greater risk of adverse kidney events.

Synopsis.

Renal functional reserve (RFR) is described as the difference between a glomerular filtration rate (GFR) measured at baseline and after protein stimulation. The percent change in GFR after a protein load varies based on differences in experimental conditions, with the use of an oral meat protein stimulus and a creatinine clearance method to quantify GFR showing the greatest RFR. A decline in RFR has been found in numerous patient groups. Recent investigations have suggested that a lower RFR may be associated with an increased risk of acute kidney injury and eventual chronic kidney disease.

Key Points

• First described over 20 years ago, renal functional reserve is not routinely quantified or applied in clinical practice.

• There are numerous proposed mechanisms explaining the change in glomerular filtration rate that occurs with a protein load, with the most accepted theories include the involvement of humoral mediators and resetting of tubuloglomerular feedback.

• There has been a recent renewed interest in the study of renal function reserve for the purpose of detecting patients more susceptible to clinical acute kidney injury and chronic kidney disease.

• Future investigations should focus on the impact of renal function reserve assessments on outcomes.

Clinical Care Points

• When quantified prior to cardiac surgery, renal functional reserve has been found to be lower in patients that experience acute kidney injury (AKI) postoperatively.

• Patients that experience AKI after cardiac surgery show a decrease in renal functional reserve when repeated 3 months after surgery.

• Providing an amino acid infusion prior to cardiopulmonary bypass has been shown to potentially decrease the duration of AKI after cardiac surgery.