Acute Kidney Injury and Chronic Kidney Disease: It’s a Two-Way Street

Abstract

Acute kidney injury (AKI) is frequently seen in hospitalized patients and its incidence increases with the severity of illness. Recent studies have further illuminated the interdependent relationship between AKI and chronic kidney disease (CKD). CKD and proteinuria have been demonstrated to be risk factors for AKI. Moreover, the previous dogma that prognosis is excellent for patients who recover after AKI episodes may not be universally accurate as CKD is emerging as a long-term consequence after AKI. Short-term mortality is lower in CKD patients with AKI.

• AKI
• CKD
• outcomes

INTRODUCTION

Acute renal failure is a frequent occurrence in hospitalized patients and its incidence increases with the severity of illness.^1–3 Until recently there were more than 35 definitions of acute renal dysfunction ranging from percent change in serum creatinine to a need for dialysis. Due to the inconsistencies inherent to the wide variances in AKI definitions, it had been extremely difficult to compare results of different studies in order to draw meaningful and generalizable conclusions.

In 2004 a group of experts proposed the risk, injury, failure, loss, end-stage renal disease (RIFLE) criteria for acute renal failure (based on change in serum creatinine or urine output).⁴ These criteria were further refined by the acute kidney injury network (AKIN) definition in 2007, and the term acute kidney injury (AKI) was coined.⁵ Almost immediately following the establishment of a consensus definition of AKI, intense research efforts pertaining to AKI in various clinical settings commenced.

This outpouring of new literature on AKI resulted in several new observations. Some of the more significant observations are the interrelationships of AKI and chronic kidney disease (CKD). CKD has emerged as not only a risk factor for AKI, but also as a potential durable consequence of AKI. There is also evidence of CKD offering potential protection from short-term mortality in the setting of concurrent AKI.

The following presents the highlights of literature addressing the interdependent relationship between AKI and CKD since the RIFLE and AKIN definitions of AKI have been established.

CKD AS A RISK FACTOR FOR AKI

In a multitude of clinical settings including iodinated radio-contrast exposure, post-cardiac surgery or in general hospitalized patients, there is clear evidence to support CKD as being a distinct risk factor for the development of AKI. Some of the most recent publications are reviewed below.

Using the AKIN definition (rise of 0.3 mg/dL of creatinine over 48 h), Pannu et al. reported the risk of developing inpatient AKI as a function of baseline outpatient serum Cr levels in more than 43,000 hospitalized Canadian patients from 2003 to 2006.⁶ The development of AKI was associated with the presence of CKD and there was a significant stepwise increase in risk as the CKD stage advanced. The odds-ratio for developing AKI with a baseline glomerular filtration rate (GFR) of 45–59, 30–44, and <30 mL/min was 2.9, 6.2, and 18.3, respectively.⁶ A similar finding was reported in the Atherosclerosis Risk in Communities (ARIC) study among a prospective cohort of 11,200 patients aged 45–64 years over an 8-year period. Patients with a baseline GFR of 45–59, 30–44, and 15–29 mL/min had a relative risk for AKI of 3.5, 12.9, and 10.7, respectively.⁷

In a “pre-RIFLE” study, Hsu et al. analyzed the strength of the association of AKI (defined as requiring dialysis) with CKD in more than 600,000 patients from the Kaiser Permanente Database between 1996 and 2003. There was a significant increase in the risk of AKI associated with the advancing stages of CKD with respective odds ratios of 1.95, 6.5, 28.5, and 40 in patients with baseline GFR of 45–59, 30–45, 15–30, and <15.⁸

Collectively, these results suggest that CKD is associated with an increased risk of developing AKI (defined by AKIN or dialysis initiation). These results, however, do not address a causal mechanism. It is plausible that the underlying risk factors that contributed to the presence of CKD in the populations may also have been responsible for the development of AKI. Despite careful adjustment for a multitude of risk factors, the largely observational and retrospective nature of these studies and the imperfect nature of the multivariate regression analyses prevent us from drawing a definitive conclusion.

PROTEINURIA AS AN ADDITIVE FACTOR FOR AKI

The association between AKI and proteinuria has been explored recently. Huang et al. described the association between baseline GFR and presence of proteinuria with AKI (AKIN criteria) in 1235 patients after cardiac surgery.⁹ Baseline CKD was associated with an increased risk for AKI and the presence of any degree of significant proteinuria was itself an independent risk factor.⁹

A large retrospective observational study from Canada looked at records of over 900,000 adults from 2002 to 2007 and stratified the risk for the development of AKI based on levels of GFR and the presence of baseline proteinuria.¹⁰ In each stage of CKD, the presence of any degree of proteinuria had an independent influence on the incidence of AKI. Interestingly, patients with GFR >60 and heavy proteinuria had a higher incidence of AKI than those with GFR of 45–59 and no proteinuria.¹⁰

In interpreting the association of proteinuria and AKI, certain considerations need to be addressed. First, proteinuria (more specifically, albuminuria) can be a marker for endothelial dysfunction, and its presence may identify subgroups of patients with increased microvascular damage. Second, medications that are commonly used to treat hypertension and/or proteinuria (such as angiotensin converting enzyme inhibitors or angiotensin receptor blockers) may impair renal vasomotor autoregulation, thereby increasing the risk of meeting the definition of AKI, although the clinical consequences of immediately reversible AKI in this setting as opposed to other causes of AKI remain unclear. Furthermore, to date, no studies have demonstrated that treatment or reduction of proteinuria affects the incidence or magnitude of AKI.

MORTALITY IN AKI ON CKD

AKI and CKD are independently associated with increased mortality. When occurring together, however, AKI in the setting of CKD paradoxically has been shown to decrease mortality (Table 1).

Table 1.  Characteristics of studies for mortality in AKI on CKD.

				Mortality
Author	Year	Setting	AKI definition	CKD versus without	Comments
Prescott^11	2007	Hospitalized809 ptsUK	RRT	20% versus 30% (10 days)43% versus 50% (90 days)	Patients with first RRT in ICU had higher mortality
Brar^12	2008	Critical care	RRT	31% versus 83% (30 days)	Single center study
		81 pts
Khosla^13	2009	Critical care618 ptsUSA	↑ 0.5 mg/dL (baseline <1.5)Or ↑ 1 mg/dl	31% versus 40% (in hospital)	Earlier nephrology consultation in CKD group (2 days)
Wu^14	2011	Post major surgery in ICU	RIFLE	AKI-F similar to CKD
		17,788 pts
		Taiwan

Notes: pts, patients.

In a large study from the National Inpatient Database, Waikar et al. demonstrated that AKI was associated with lower mortality in patients with CKD, findings that were sustained over the 15 years included in the study.¹⁵ The authors showed that among those who developed AKI, patients with CKD had an overall mortality rate of 22% compared to 30% of patients without CKD.¹⁵ In a similar study of UK patients, Prescott reported similar findings with respect to an apparent mortality benefit conferred by CKD in AKI.¹¹

This association of improved mortality was also exhibited in critically ill patients.¹⁶ In 618 patients with AKI from the PICARD study, the mortality among patients with CKD was 31% compared to 40% without CKD.¹³ It should also be noted that in this study there was a shorter time to nephrology consultation (2 days earlier) in patients with CKD.¹³

The literature, however, is not entirely consistent. Wu et al. studied 17,788 patients admitted to an ICU after major surgery within a 4-year follow-up period. AKI was defined using the RIFLE criteria and mortality was worse in patients with any degree of AKI (including Risk, Injury, and Failure) on CKD than in those without CKD.¹⁴ Yet, a different conclusion may be reached when restricting the analysis only to Stage F (Failure) of the RIFLE classification of AKI, which arguably could be seen as more clinically significant. Among patients with severe AKI (reaching F stage), mortality was similar in patients with or without CKD (47% vs. 45%).

Presently there is no known mechanism explaining the observed lower mortality in patients with AKI on CKD; however, several points might be considered.

First, preconditioning is a physiological state in which organ systems become accustomed to sublethal ischemic events, and putatively develop protective mechanisms. This was demonstrated in animal studies in which a second ischemic event had a lesser impact on the development of AKI in animals that underwent preconditioning.^17,18 Though the mechanism is not known, it is hypothesized that preconditioning probably occurs by inducing a decrease in the inflammation response,¹⁷ an increase in the production of a hypoxia-induced factor¹⁹ and a down-regulation of apoptosis pathways.¹⁸

In humans, a secondary analysis of two randomized controlled trials showed a decrease in the incidence of post-operative AKI in patients that had ischemic preconditioning.²⁰ The CKD status in this study was not reported. Preconditioning has been suggested to be mediated in humans by attenuation of inflammation.²¹

Second, the intensity of injury from comparable degrees of AKI may be attenuated in the CKD population as opposed to the non-CKD population. In a population-based cohort study from Ontario, Canada, Wald et al. identified 3769 patients with acute kidney injury necessitating in-hospital dialysis.²² The risk of end-stage renal disease (ESRD) was greater in those without CKD than those with, suggesting that the disease severity may have been greater due to the absence of some as yet unidentified protective entity present in the CKD population.²²

Lastly, there may be increased attention and focus on renal-specific problems as a whole among the CKD population as compared to the non-CKD population due to a pre-conceived heightened awareness of renal insufficiency. This idea was suggested by Khosla¹³ who found that a nephrology consultation for AKI occurred nearly 2 days sooner in patients with CKD than in those without CKD.

In summary, differences in short-term and long-term outcomes following AKI in patients with or without CKD have been demonstrated but not yet explained. Regarding mortality, the suggestion of a potentially protective role of CKD in AKI is intriguing and warrants further studies.

RISK OF CKD AND ESRD WITH AKI, AND AKI ON CKD

In the traditional view of the natural history of AKI, the renal prognosis was generally considered to be very good, and full recovery to previous levels of kidney function was expected in most patients.

This view has been challenged by recent observational studies that demonstrated that patients with an episode of AKI may be at an increased risk of developing CKD and even ESRD.

A study from the VA system followed 5404 patients with AKI over 5 years (defined by ICD9 coding). Those patients had a 4–6 times risk of developing Stage 4 CKD compared to control subjects without AKI.²³ In a study from the Kaiser Permanente database, severe AKI, as defined by a need for acute dialysis, resulted in an even more impressive 28-fold risk for the development of advanced CKD.²⁴ These findings were replicated in other clinical settings including post-MI, cardiac surgery, HIV, or bone marrow transplantation.²⁵

Other studies support the association of AKI and risk of ESRD ranging from 3- to 40-fold.¹⁰

Furthermore, the multiplicity and duration of AKI episodes may influence outcomes as well. Thakar et al. showed that patients with two, three, or more AKI episodes were more likely to progress to Stage 4 CKD compared to patients with only one episode. The risk for advanced renal disease doubled with each additional AKI episode.²⁶

Duration of the AKI event was directly proportional to mortality in a study by Brown after cardiac surgery.²⁷ Patients with an AKI duration of 1–2 days had an odds ratio for long-term mortality 1.6 compared to controls and the risk went up to 3.4 when AKI was sustained for 7 days or more.²⁸

Coca et al. performed a systematic review and meta-analysis of studies pertaining to the impact of AKI on CKD and ESRD.²⁵ After an episode of AKI, the risk for developing CKD was 8.8 compared to controls and 3.1 for ESRD. When stratified by level of AKI, there was a higher risk for worse outcomes with an increased severity of AKI. It should be noted that there was a high heterogeneity between the meta-analysis studies since the definitions of AKI, CKD, duration of follow up, and clinical settings were markedly different between studies.

The potential mechanisms that could lead an AKI episode to CKD and ESRD are broad and are beyond the scope of this review, but may include nephron loss with subsequent hyperfiltration, tubulo-interstitial fibrosis, and glomerulosclerosis.²⁸

Generally there seems to be strong evidence to support AKI (including its frequency, intensity, and duration) being associated with newly onset CKD and ESRD. The previous dogma that patients who recover from AKI have excellent renal outcomes may not be entirely accurate. This knowledge may inspire the clinician to consider the potential for CKD after an AKI episode and provide appropriate follow up and nephrology referrals as necessary.

CONCLUSION

Recent studies have examined the relationship between AKI and CKD and have suggested that CKD and proteinuria are risk factors for AKI, and the previous tenet proposing recovery after AKI episode portending excellent prognosis may no longer be accurate. CKD is emerging as a long-term consequence of AKI, and short- (but not long-) term mortality from AKI may be lower in patients with CKD as compared to those without.

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