End Points for Clinical Trials in Acute Kidney Injury

Challenges Inherent to the Study of Acute Kidney Injury

Despite advances in modern medicine, acute kidney injury (AKI) is associated with a heavy burden of morbidity and mortality. Even mild forms of AKI are associated with a sevenfold increase in 30-day mortality and increase the risk of multiorgan failure. Severe AKI requiring renal replacement therapy (RRT) is associated independently with a 50% incidence of death among critically ill patients. In spite of the rising incidence of AKI in hospitalized patients no effective therapies other than supportive care currently exist. End point selection is one of the most critical components of clinical trial design. The primary end point must be clinically relevant, important to patients and medical providers, quantifiable, and amenable to therapeutic interventions. The end point should be robust, with negligible confounding factors or bias. In addition, end point selection must be relevant to regulatory agencies (e.g., Food and Drug Administration and European Medicines Agency) if the trial involves a therapeutic agent. The choice of end point is further influenced by the target patient population (e.g., septic AKI vs. cardiac surgery–associated AKI) and treatment objective (e.g., reduction in need for dialysis and/or mortality vs. creatinine improvement) ( Fig. 226.1 ).

Therapeutic clinical trials for AKI have been hampered by numerous barriers such as lack of agreement on the staging and diagnostic criteria for AKI, failure to diagnose AKI in a timely manner, lack of well-defined appropriate study end points, small observed effect size, and inadequate study size population. The development and use of consensus criteria for AKI diagnosis have been an important step in improving the conduct of clinical trials across target patient populations. The Acute Kidney Injury Network (AKIN) group, National Institute of Diabetes and Digestive Kidney Diseases (NIDDK), and the Acute Dialysis Quality Initiative (ADQI) have made ongoing attempts to improve end points for AKI clinical trials. However, some of the challenges inherent to studies of AKI therapeutics persist. First, the event rates for AKI, particularly severe AKI, can be relatively low; to yield acceptable event rates to make clinical trials practical, the target population should be at high risk for AKI. Among 1219 adults who underwent cardiac surgery in the translational research in biomarker end points (TRIBE) study, 426 (34.9%) patients developed AKI after cardiac surgery; however, the incidence of severe AKI (doubling of serum creatinine or need RRT) was only 5%. Therefore with this incidence rate and power of 0.9, and 30% effectiveness of an intervention (i.e., 30% relative reduction in the development of severe AKI from 5% to 3.5%), respectively, 3799 patients per intervention group would be needed to show therapeutic benefit with this event rate. No clinical trial in AKI, either preventative or interventional, has enrolled such high patient numbers to achieve adequate power. Without adequate power, results from clinical trials may be misleading and may misinform the medical provider. In one published summary of preventative AKI trials, only 3 of 30 cardiac surgery trials and only 3 of 28 contrast studies were enrolling 800 or more adult patients total.

Second, although selecting patients at highest risk of AKI increases the event rate, it may reduce the efficacy of the therapeutic intervention in these high-risk patients. Third, other challenges to trial design in the field of critical care nephrology include the identification of study participants early enough such that they may benefit from treatment, the use of therapies that target only a single pathophysiologic process despite the multifactorial causes of some AKI events, and the relative heterogeneity of study populations.

Finally, deciding on the appropriate study end point for phase of AKI clinical trial can depend on whether it is a preventative or intervention trial and can prove challenging. For instance, the use of short-term changes in serum creatinine (or other measures of glomerular filtration rate [GFR]) or changes in biomarker levels as secondary end points may be premature in phase 3 AKI management clinical trials. Other randomized clinical trials in AKI have assessed other hard clinical end points, including death, and have been met with disappointment. Important examples of clinical end points in AKI trials include creatinine- and urine output–based criteria for AKI diagnosis, risk prediction scores, functional and damage biomarkers, requirement for renal replacement therapy, onset of chronic kidney disease (CKD), persistent decline in estimated glomerular filtration rate (eGFR), progression to end-stage renal disease (ESRD), and death. We will discuss individually the different renal and nonrenal end points in AKI trials and discuss methods to improve study outcomes in AKI intervention trials.

Creatinine- and Urine Output–Based Criteria for Acute Kidney Injury Diagnosis

Trial end points that include reduced urinary output or reflect a rise in creatinine typically are applied in earlier phases of drug or device development to demonstrate dose responsiveness, support proof-of-concept, and/or provide preliminary evidence of safety and efficacy. Current consensus definitions of AKI including the Kidney Disease: Improving Global Outcomes (KDIGO) criterion rely on changes in either the level of serum creatinine or urinary output. AKI severity has an important impact on enrollment in AKI prevention trial. A higher stage of AKI diagnostic criteria provides a more appropriate clinically significant AKI end point than less severe AKI diagnostic criteria. Enrollment of patients with early AKI (stage 1 KDIGO) is more sensitive to identification of AKI and will increase statistical power. For example, a therapy that reduces the relative risk of stage 1 AKI by 40% will require approximately 200 study participants (100 in the treatment group and 100 in the control group) if the baseline risk of stage 1 AKI is 40%. A therapy that reduces the relative risk of stage 2 AKI by 50% will require approximately 2200 study participants (1100 in the treatment group and 1100 in the control group) if the baseline risk of stage 2 AKI is 12%. Including participants with early AKI will increase the incidence of false-positive results and, as a result, attenuate the association of AKI with more important subsequent outcomes, such as worsened CKD and death.

Another explanation for the disappointment of interventional trials in AKI and the high mortality associated with AKI is the dependence on serum creatinine for diagnosis of AKI. The diagnosis of AKI may be delayed by the overreliance on serum creatinine as a functional marker, a surrogate for changes in GFR which often occur late in the time course of AKI. A subclinical AKI defined as a renal injury marked by a lack of change in serum creatinine despite changes in structural nephron damage biomarkers is increasingly being recognized as a clinically important type of AKI. Injury biomarkers can detect AKI several hours to days before increases in serum creatinine and is discussed in more detail below.

Although decreased urine output is also incorporated as part of the AKI definitions, limited data exist as to whether early alterations in urine flow are a sensitive marker of AKI. Indeed, there is less certainty of the appropriateness of oliguria alone as a trial entry criterion. Clinical factors such as hydration status, nondiuretic use, cardiovascular hemodynamics, and the extent of early resuscitation measures can affect urine output. Early changes in urinary output are not always true reflections of the decline of GFR and can lead to a misclassification as “mild” AKI. Despite this, some critically ill patients such as those with severe septic shock, consecutive oliguria may provide a valuable measure of AKI risk.