Clinical and technical evidence

A literature search was carried out for this briefing in line with the interim process and methods statement. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.

Published evidence

Five studies are summarised in this briefing with a total of 1,659 critically ill patients admitted to an intensive care unit (ICU). The evidence suggests that increased levels of TIMP-2 and IGFBP7 can predict the risk of acute kidney injury (AKI) and that NephroCheck-guided intervention can reduce the incidence of moderate to severe AKI in critically ill patients. Evidence in 4 studies was generated using the NephroCheck test on a benchtop platform.

Table 1 summarises the clinical evidence as well as its strengths and limitations.

An NIHR HTA-funded systematic review (Hall et al. 2018) assessed a variety of in vitro diagnostic tests, including NephroCheck, for early detection of AKI in ICU patients. This review concluded that using in vitro diagnostic tests in ICU patients could potentially improve patient care. A meta-analysis done as part of the review highlighted limitations arising from between-study heterogeneity. Overall the evidence base was limited because there is no standard definition of AKI and there was a variation in the time of testing.

Overall assessment of the evidence

Diagnostic accuracy and validation

In the included studies, AKI was defined based on a variety of classifications: RIFLE, KDIGO and AKIN. Irrespective of the definition used, the 3 validation studies (Kashani et al. Bihorac et al. and Cuartero et al.) found that an increase in urinary IGFBP 7 and TIMP2 above a specified cut-off point is associated with an increased risk of AKI. However, these studies have not shown that a rise in urinary TIMP-2 and IGFBP7 is specific to the presence of AKI, considering that TIMP-2 and IGFBP7 levels can increase because of a variety of cellular trauma including oxidative stress, ultraviolet radiation, drugs and toxins.

A test specificity of ~50% for the NephroCheck test at a pre-selected cut-off of 0.3 suggests that ~50% of AKI negative patients would have a false-positive result and receive preventive care. This may be an improvement from current standard care, where people who do not have AKI receive unnecessary preventive measures.

Clinical effectiveness and implementation

The Meersch (2017) and Gocze (2017) studies assessed the clinical effectiveness of implementing the KDIGO care bundle in cardiac and non-cardiac surgery patients assessed as being at increased risk of developing AKI based on TIMP-2·IGFBP7 of greater than 0.3. Results of both trials show that NephroCheck-guided care significantly reduced the incidence AKI in the intervention arm. However there was a non-significant difference between the intervention and control arms of both trials for the need for renal replacement therapy and mortality and major adverse kidney events by 30 days. It remains uncertain if reducing AKI incidence translates into any patient or system benefits.

A significant number of further studies are planned or in progress.

Table 1 Summary of selected studies

Kashani et al. (2013)

Study size, design and location

A prospective multicentre observational study across North America and Europe. This study was in 2 phases. Phase 1 was the discovery of novel biomarkers in adults admitted in ICU with at least 1 risk factor for AKI. Phase 2 (Sapphire study) was the validation phase involving 744 adults admitted in ICU with critical illness and no evidence of AKI at the time of enrolment.

Intervention

NephroCheck test on the benchtop instrument at a test cut-off of >0.3, referenced against clinical outcome of confirmed AKI as classified by KDIGO staging.

Key outcomes

In the discovery phase, urinary IGFBP7 and TIMP2 emerged as the best-performing biomarkers from 340 plasma and urinary markers examined (AUC=0.77 and 0.75, respectively, for RIFLE-I/F within 12 to 36 hours).

The AUC for IGFBP7 and TIMP2 was significantly greater (p<0.002) than the AUC of previously known biomarkers (urine and plasma NGAL, plasma cystatin-C, and KIM-1, IL-18, pi-GST, and L-FABP in the urine) of AKI.

Based on an analysis done for 728 patients in the validation cohort, 101 (14%) patients met the primary endpoint of moderate or severe AKI (11%-KDIGO stage 2 and 2.5% KDIGO – stage 3) within 12 hours.

218 (30%) patients developed AKI within 7 days (22% KDIGO stage 2 and 8% KDIGO stage 3) and 49 (6.7%) patients had renal replacement therapy during hospital stay.

A total of 121 (17%) patients died before hospital discharge truncated at 30 days, 161 subjects (22%) met the MAKE30 endpoint.

Strengths and limitations

This large multicentre trial discovered new biomarkers and validated them in separate patient populations. Laboratory technicians were blinded to clinical data and reasons for loss to follow up were clearly stated.

The study was funded by the manufacturer of the NephroCheck Test system (Astute Medical).

Bihorac et al. (2014)

Study size, design and location

A prospective multicentre study of 420 critically ill patients admitted to both surgical and medical ICUs, with an indwelling urinary catheter in the US.

Intervention

NephroCheck test on the benchtop instrument at a test cut-off of >0.3, referenced against clinically confirmed AKI as judged by an independent clinical adjudication committee (CAC) of 3 nephrologists.

Key outcomes

The primary endpoint was the ability of urinary TIMP2 and IGFBP7 to predict moderate to severe AKI within 12 hours of test measurement as judged by the CAC. Of the 408 patients evaluated, 71 (17.4%) reached the key endpoint. The CAC agreed on the diagnosis of AKI in 94% of cases. Inter-rater reliability was excellent, Fleiss' Kappa: 0.86 (95% CI 0.80 to 0.91; p<0.001).

Median urinary TIMP2 IGFBP7 was significantly higher in critically ill patients diagnosed with AKI 1.6 (0.7 to 2.8) than those without 0.3 (0.2 to 0.8), p<0.001.

At a pre-selected cut-off of 2.0, specificity was 95% (95% CI 93 to 97) and sensitivity was 37% (95% CI 26 to 47). At a pre-selected cut-off of 0.3 sensitivity was 92% (95% CI 85 to 98), specificity was 46% (95% CI 41 to 52).

Patients with TIMP2 IGFBP7 between 0.3 and 2.0 had 5 times more risk for AKI (95% CI 3 to 17) than critically ill patients with results below ≤0.3. Patients with test results greater than 2.0 had 17 times the risk for AKI (95% CI 9 to 54).

Strengths and limitations

This study was sufficiently powered for the endpoint and had a complete follow-up. The adjudication committee and laboratory technicians were blinded to the results of the test. Adjudication was based both on the test and a range of other clinical data.

The study was funded by the company.

Cuartero et al. (2017)

Study size, design and location

A prospective observational study of 98 critically ill patients with or without sepsis admitted to ICU with an expected stay of at least 48 hours.

Spain.

Intervention

NephroCheck test on the benchtop instrument at a test cut-off of >0.3, referenced against clinical outcome of confirmed AKI as classified by AKIN staging.

Key outcomes

AKI was classified in this study based on the AKIN classification. Test result scores of TIMP2 IGFBP7 were significantly higher in patients with AKI (1.03, IQR 0.38 to 3.29) compared to those without AKI (0.24, IQR 0.11 to 0.48; p<0.001). These differences were not related to the presence of sepsis.

Patients who developed AKI had higher median test result scores (1.05 [IQR 0.41 to 2.31] for patients without sepsis and 0.98 [IQR 0.36 to 3.94] for patients with sepsis) than those without AKI (0.21 [IQR 0.10 to 0.40] in non-septic patients and 0.32 [IQR 0.15 to 0.63] for those with sepsis) with p<0.001 between subgroups with and without AKI.

In patients with high TIMP2 IGFBP7 test result >2.0, there was a 3.15-fold risk for AKI and a 1.85-fold risk that AKI would be AKIN ≥2.

Strengths and limitations

In this study healthcare providers were blinded to the test results.

This study had a small sample size and was funded by the company.

Meersch et al. (2017)

Study size, design and location

A single-centre unblinded randomised study of 276 adults with an increased AKI risk (as measured by increased TIMP2 IGFBP7 levels) after cardiac surgery in Germany.

Intervention

Comparator

NephroCheck-guided implementation of the KDIGO care bundle.

Standard care for AKI prevention.

Key outcomes

The primary endpoint was the incidence of AKI as defined by KDIGO guidelines within the first 72 hours after surgery.

There was a significantly lower incidence of AKI 72 hours after surgery in the intervention group than in the control (55.1% versus 71.7%, p=0.004).

Within this, the incidence of moderate to severe AKI was significantly lower in the intervention group (29.7% versus 44.9%, p=0.009).

The implementation of the bundle resulted in significantly improved hemodynamic parameters at different time points (p<0.05), less hyperglycaemia (p<0.001) and use of ACEi/ARBs (p<0.001) compared to controls.

At 30, 60 and 90 days of follow-up there was no significant difference in requirement for RRT, all-cause mortality, MAKE, ICU length of stay and hospital length of stay between the intervention and control groups.

Strengths and limitations

Randomisation in this study reduces the risk of bias. The study design is suitable to test the effect of the intervention.

The study was funded by the manufacturer of the NephroCheck test system (Astute Medical) and was not blinded. Standard care may not be generalisable to the UK.

Gocze et al. 2017

Study size, design and location

A single-centre unblinded randomised clinical trial of 121 patients with an increased AKI risk after major non-urgent, non-cardiac surgery in Germany.

Intervention

Comparator

NephroCheck-guided implementation of the KDIGO care bundle.

Standard care for AKI prevention.

Key outcomes

The primary endpoint was the incidence of AKI according to the KDIGO guidelines during the first 7 days after surgery.

The difference in the primary endpoint (all stages of AKI) between the intervention group and the control group was not statistically significant (31.7% versus 47.5% p=0.076). However, the incidence of moderate to severe AKI was significantly lower in the intervention group compared to the control group (6.7% versus 19.7%, p=0.035). The hospital length of stay (p=0.036) and the ICU length of stay (p=0.035) were also significantly shorter in the intervention group compared to the control group.

There was no significant difference in the need for RRT, in-hospital mortality or MAKE at discharge between the intervention and control groups.

Strengths and limitations

The study was ended prematurely after an interim data analysis indicated planned recruitment (n=138) would not result in a statistically significant difference in the primary outcome. This might affect the size of the effect measured in the primary outcome.

There was not direct financial involvement from the company stated in this study, this could minimise any bias towards the outcomes reported.

The standard care in this study may not be generalisable to the UK.

Abbreviations: ACEi/ARBs, angiotensin converting enzyme inhibitors or angiotensin II receptor blockers; AKI, acute kidney injury; AKIN, Acute Kidney Injury Network; AUC, area under the curve; ICU, intensive care unit; IGFBP 7, insulin-like growth factor binding protein 7; IL-18, interleukin-18; KDIGO, Kidney Disease: Improving Global Outcomes; KIM-1, kidney injury molecule-1; L-FABP, liver fatty acid-binding protein; MAKE30, Major Adverse Kidney Events by 30 day; NGAL, neutrophil gelatinase-associated lipocalin; pi-GST, pi-glutathione S-transferase; RIFLE-I/F, Risk, Injury, Failure, Loss, End stage; RRT, renal replacement therapy; TIMP 2 metalloproteinase 2.

Recent and ongoing studies