General data
A total of 258 patients were screened, 201 patients were enrolled in the study, and
17 patients were lost to follow up. Therefore, 184 patients with AKI (86 patients
with transient AKI and 98 patients with renal AKI) were included in the study. All
the 184 patients survived to one-year follow up. After regular follow up for one year,
111 of the 184 patients were in the stable renal function group, and 73 patients were
in the renal function deterioration group. Of the 98 patients with renal AKI, 49 were
in the stable renal function group, and 49 were in the renal function deterioration
group (Fig. 1).
Fig. 1. Study flow chart. AKI acute kidney injury
Of the 184 patients with AKI, the ages ranged from 18 to 99 years, and the median
age was 53.00 (31.00, 64.75) years. The male/female ratio was 1.75:1, and the mean
follow-up duration was 12.2?±?1.06 months. The basic serum creatinine (sCr), basic
eGFR and basic 24-h total urine protein levels were 1.12 (0.87, 2.40) mg/dl, 66.68
(28.10, 94.18) ml/min/1.73 m
2
and 3.78 (1.47, 9.08) g, respectively. The baseline creatinine was considered the
lowest value within 3 months prior to diagnosis of AKI.
There were 71 patients with previous chronic renal disease (CKD). CKD is defined according
to Kidney Disease Improving Global Outcomes (KDIGO) Clinical Practice Guidelines for
the Evaluation and Management of Chronic Kidney Disease 12]. The uKIM-1 level at the time of AKI occurrence was 2.37 (1.10, 6.22) ng/mg, and
the peak level of creatinine was 2.43 (1.41, 3.74) mg/dl. There were 123 patients
in AKI stage I, 26 patients in stage II and 35 patients in stage III. Patients with
renal AKI accounted for 53.26 %. The clinical etiology of AKI included insufficient
renal blood perfusion or ischemia (22 %) (insufficient volume, gastrointestinal loss,
cardiac failure, or renal vascular factors), nephrotoxicity (20 %) (nephrotoxic agents
and contrast medium injury, etc.), infection factor (23 %) (sepsis, severe pneumonia
and any kind of infection that caused renal injury), aggravation or activation of
glomerular disease (19 %), and obstruction (11 %). The proportion of patients with
renal function progression was 39.67 % at the one-year follow up (Table 1).
Table 1. General condition of patients
uKIM-1 detection by western blot
Western blot analysis was performed on urine specimens from six patients: two patients
were clinically diagnosed with renal AKI, two patients were clinically diagnosed with
transient AKI, one was a patient with CKD, and one patient was used as a normal control.
uKIM-1 was undetectable in healthy control urine, and a low level was observed in
the urine of the patient with CKD. In sharp contrast, all patients with established
AKI had easily detectable levels of uKIM-1, and higher levels were observed in urine
from patients who were clinically diagnosed with renal AKI (Fig. 2).
Fig. 2. Western blot analysis of urinary kidney injury molecule-1 (uKIM-1) with and without established acute kidney injury (AKI). All patients with established AKI had easily detectable uKIM-1. The uKIM-1 levels
in urine from patients who were clinically diagnosed with renal AKI were higher than
in patients with transient AKI: *P??0.05, **P??0.001. CKD chronic kidney disease
Correlation analysis
Spearman correlation analysis indicated that uKIM-1 positively correlated with the
fractional excretion of sodium (FeNa), renal failure index (RFI) and creatinine elevation
(r?=?0.887, 0.887 and 0.438, respectively, P?=?0.000) (Fig. 3).
Fig. 3. Spearman correlation analyses. a The urinary kidney injury molecule-1 (uKIM-1) content was positively correlated with fractional excretion of sodium (FeNa). b. The uKIM-1 content was positively correlated with the renal failure index (RFI). c. The uKIM-1 content was positively correlated with the times of creatinine elevation
Identification of transient AKI and renal AKI on the basis of uKIM-1 levels
The patients were classified into transient AKI (86 patients) and renal AKI groups
(98 patients) on the basis of renal function recovery during short-term follow up
(?48 h). Comparisons between groups demonstrated that the prognosis of the renal AKI
group was significantly inferior to the transient AKI group. Of the 86 patients with
transient AKI, 44 presented with a FeNa ?1 %, and 48 presented with an RFI ?1 %. Comparisons
between groups indicated that the uKIM-1 level of the renal AKI group was significantly
higher than that of the transient AKI group (Table 2, Fig. 4).
Table 2. Comparison between the transient AKI group and the renal AKI group
Fig. 4. The urinary kidney injury molecule-1 (uKIM-1) levels in the transient acute kidney injury (AKI) group were significantly higher than those of the renal AKI group
ROC curve analysis revealed that the area under the curve (ROC-AUC) was 0.691 (P?=?0.000) for the diagnosis of renal AKI on the basis of uKIM-1. The sensitivity was
66.3 % when the cutoff point of uKIM-1 level was 2.14 ng/mg, and the specificity was
64.7 % (Fig. 5).
Fig. 5. Receiver operating characteristic area under the curve (ROC-AUC) of renal acute kidney injury according to urinary kidney injury molecule-1 (uKIM-1) level
Analysis of long-term prognosis predictions on the basis of uKIM-1 levels
Analysis of long-term prognosis for patients with renal AKI
Of the 98 patients with renal AKI at the regular one-year follow up, 49 were in the
stable renal function group, and 49 were in the renal function deterioration group.
There was no difference by sex in the composition of the groups. However, patients
in the renal function deterioration group were older than those in the renal function
stable group, and their basic renal function was inferior, with a high peak value
of creatinine after AKI. More patients in the renal function deterioration group presented
with FeNa ?1 % and RFI ?1 %, and their uKIM-1 levels were significantly higher than
the stable renal function group (Table 3).
Table 3. Comparison between the renal function deterioration group and the stable renal function
group after one-year follow up in patients with renal acute kidney injury
Gradual Cox regression analysis indicated that patients with renal AKI had higher
uKIM-1 levels with poorer long-term renal prognosis. The risk of renal function deterioration
following AKI increased by 6.4 % for each 1 ng/mg increase in uKIM-1. The prognosis
of patients with renal AKI who presented with FeNa ?1 % was inferior to that of patients
with FeNa 1 % (Table 4).
Table 4. Cox regression shows the independent risk factors for renal function progression in
patients with renal acute kidney injury
The ROC-AUC for the prediction of renal function deterioration in patients with renal
AKI when the uKIM-1 level was 0.680. When the cutoff for uKIM-1 was 2.46 ng/mg, the
sensitivity was 78.6 %, and the specificity 57.9 % (Fig. 6). The K-M curve indicated that uKIM-1??2.46 ng/mg positively correlated with poor
long-term renal prognosis in patients with renal AKI (P?=?0.000) (Fig. 7).
Fig. 6. Receiver operating characteristic area under the curve (ROC-AUC) of renal function progression in patients with renal acute kidney injury
Fig. 7. Urinary kidney injury molecule-1 (uKIM-1) level and kidney prognosis in patients with renal acute kidney injury patients.
uKIM-1??2.46 ng/mg was positively related to poor prognosis. Cum cumulative
Analysis of long-term prognosis for patients with AKI
Of the 184 patients with AKI after the regular one-year follow up, 111 patients were
in the stable renal function group, and 73 patients were in the renal function deterioration
group. There was no difference by sex in the composition of the groups. However, patients
in the renal function deterioration group were older than patients in the stable renal
function group, and their basic renal function was inferior, with a high peak value
of creatinine following AKI. More patients in the renal function deterioration group
presented with FeNa ?1 % and RFI ?1 %, and the uKIM-1 level was significantly higher
than that of the stable renal function group (Table 5).
Table 5. Comparison between the renal function deterioration group and the stable renal function
group after one-year follow up in patients with acute kidney injury
Gradual Cox regression analysis indicated that older patients had lower basic eGFR,
higher uKIM-1, and poorer long-term renal prognosis. The risk of renal function deterioration
after AKI increased by 3.3 % for each one-year increase in age. The risk of renal
function deterioration after AKI increased by 2.4 % for each decrease of 1 ml/min/1.73 m
2
of basic eGFR. The risk of renal function deterioration after AKI increased by 1.8 %
for each 1 ng/mg increase in uKIM-1 (Table 6).
Table 6. Cox regression shows the independent risk factors for renal function progression in
patients with acute kidney injury
The ROC-AUC values for the prediction of renal function deterioration with age, uKIM-1
level and basic eGFR level were 0.687, 0.703 and 0.833, respectively. When the age
cutoff point was 59.5 years, the sensitivity was 54.1 %, and the specificity 78.4 %.
When the cutoff point of the uKIM-1 level was 2.37 ng/mg, the sensitivity was 78.4 %,
and the specificity 60.8 %. When the cutoff point of the eGFR level was 60.35 ml/min/1.73,
the sensitivity and specificity were 71.2 % and 76.3 %, respectively (Fig. 8).
Fig. 8. Receiver operating characteristic area under the curve (ROC-AUC) of renal function progression in patients with acute kidney injury. uKIM-1 urinary kidney injury molecule-1, GFR glomerular filtration rate
The K-M curve revealed that uKIM-1??2.37 ng/mg positively correlated with poor long-term
renal prognosis (P?=?0.000) (Fig. 9).
Fig. 9. Urinary kidney injury molecule-1 (uKIM-1) and kidney prognosis in patients with acute kidney injury. uKIM-1??2.37 ng/mg was
positively related with poor prognosis. Cum cumulative