Content uploaded by Hideki Takizawa
Author content
All content in this area was uploaded by Hideki Takizawa on Feb 15, 2021
Content may be subject to copyright.
Vol.:(0123456789)
1 3
Clinical and Experimental Nephrology
https://doi.org/10.1007/s10157-020-02008-1
ORIGINAL ARTICLE
Incidence andrisk factors ofacute kidney injury, andits effect
onmortality amongJapanese patients receiving immune check point
inhibitors: asingle‑center observational study
YoshinosukeShimamura1 · ShotaWatanabe1· TakutoMaeda1· KokiAbe1· YayoiOgawa2· HidekiTakizawa1
Received: 16 March 2020 / Accepted: 2 December 2020
© Japanese Society of Nephrology 2021
Abstract
Background Immune checkpoint inhibitors (ICPis) are associated with multi-organ immune-related adverse effects. Here,
we examined the incidence rate, recovery rate, and risk factors of acute kidney injury complicated with ICPis (ICPi-AKI)
and evaluted the association between ICPi-AKI and mortality in Japanese patients.
Methods We analyzed 152 consecutive patients receiving ICPis between 2015 and 2019. A logistic regression analysis
was performed to identify risk factors for ICPi-AKI incidence and Cox regression analysis was performed to evaluate the
association between ICPi-AKI and mortality.
Results The mean patient age was 67 ± 10years, with the median baseline serum creatinine level of 0.78mg/dL. Twenty-
seven patients (18%) developed ICPi-AKI, and 19 (73%) of them recovered. Pembrolizumab use and liver diseases were
significant risk factors for the ICPi-AKI incidence. During the follow-up, 85 patients (59%) died, 17 patients (63%) with
ICPi-AKI and 68 (54%) patients without ICPi-AKI, respectively. The ICPi-AKI incidence was not independently associated
with mortality (adjusted hazard ratio, 0.85; 95% confidence intervals, 0.46–1.61).
Conclusions Our finding suggest that pembrolizumab use and liver diseases are associated with a higher risk of ICPi-AKI
development, but ICPi-AKI did not affect mortality. Future multi-center studies are needed to develop optimal management
and prevention strategies for this complication in patients receiving ICPis.
Keywords Acute kidney injury· Immune checkpoint inhibitor· Mortality· Liver disease· Pembrolizumab
Introduction
The advent of immune checkpoint inhibitors (ICPis) has
revolutionized the care of patients with advanced malig-
nancies [1]. Currently, programmed cell death protein 1
inhibitors, nivolumab and pembrolizumab; programmed cell
death ligand 1 inhibitors, atezolizumab, durvalumab, and
avelumab; and cytotoxic T lymphocyte-associated antigen 4
inhibitor, ipilimumab, are used worldwide. Despite their sub-
stantial clinical effect, ICPis are associated with multi-organ
immune-related adverse effects (irAEs) [1]. Dermatologic,
gastrointestinal, hepatic, and endocrine toxicities are com-
mon irAEs; nephrotoxicity with acute kidney injury (AKI)
primarily due to acute interstitial nephritis has been increas-
ingly reported [2–6]. Acute kidney injury complicated with
ICPis (ICPi-AKI) is significant as there is evidence that AKI
is associated with a higher risk of subsequent chronic kidney
disease, healthcare utilization rate, and mortality [7, 8]. The
incidence rate of ICPi-AKI is 1.4–30%, and the wide range is
due to substantial heterogeneity across studies [2–4]. Several
studies reported an association between ICPi-AKI and pro-
ton pump inhibitors (PPIs) [5, 6]. However, the extrapolation
of these results to other populations is limited because these
study populations mostly comprised Caucasians. Although
several ICPi-AKI cases have been reported in Asian patients
[9], to our knowledge, no study has investigated the ICPi-
AKI incidence and its effect on mortality in Asian popula-
tions. Additionally, it has not been validated whether PPI use
is associated with ICPi-AKI in non-Caucasian populations.
Therefore, we conducted a retrospective observational study
* Yoshinosuke Shimamura
yshimamura.tkh@gmail.com
1 Department ofNephrology, Teine Keijinkai Medical Center,
Sapporo, Hokkaido0068555, Japan
2 Hokkaido Renal Pathology Center, Sapporo, Hokkaido,
Japan
Clinical and Experimental Nephrology
1 3
in a Japanese institution to investigate the incidence rate,
recovery rate, and risk factors of ICPi-AKI, and to evaluate
the association between ICPi-AKI and mortality.
Materials andmethods
Study population
This was a single-center retrospective observational study
conducted in Teine Keijinkai Medical Center, which is a
650-bed tertiary referral institution, in Sapporo, Japan. We
retrospectively reviewed medical records of all patients
receiving ICPis and selected 154 consecutive patients who
initiated ICPis for malignancies between March 1, 2015
and October 31, 2019. Baseline laboratory data were avail-
able for all patients. The study was conducted in accord-
ance with the principles of the Declaration of Helsinki and
was approved by the Institutional Review Board of Teine
Keijinkai Medical Center (approval number 2-019096-00).
All study participants provided written informed consent,
and they were presented the right to avoid enrollment in an
opt-out fashion. The study follows the checklist of “Strength-
ening the Reporting of Observational Studies in Epidemiol-
ogy” (STROBE) guidelines [10].
Outcomes anddefinitions
The outcomes were as follows: (1) incidence rate (within
12months of ICPi initiation), recovery rate, and risk fac-
tors of ICPi-AKI and (2) association between ICPi-AKI
and mortality. AKI was defined and staged according to the
Kidney Disease: Improving Global Outcomes criteria [11];
the ICPi-AKI incidence was determined according to the
clinical evaluation of three nephrologists. All AKI cases
were independently reviewed by two board-certified neph-
rologists (Y.S. and K.A.) to determine the most likely etiol-
ogy, and a third board-certified nephrologist (H.T.) resolved
any diagnostic disagreements (n = 2). ICPi-AKI resolution
was defined as recovery of the serum creatinine (sCr) level
to < 0.35mg/dL above the baseline sCr level. Baseline sCr
level was defined as the last available sCr level during the
past 6months since ICPi initiation. The patients were fol-
lowed-up until the study end date (January 31, 2020) or until
the patients were lost to follow-up, had their care withdrawn,
or died.
Covariates
Clinical and pathological data for all participants were
obtained from medical records. The covariates consid-
ered were patient characteristics, including age, sex, and
body mass index [BMI; body weight (kg)/height2 (m2)];
age-adjusted Charlson comorbidity index (ACCI) [12]; the
presence of congestive heart failure, cerebrovascular dis-
ease, peptic ulcers, diabetes mellitus, hypertension, or liver
disease, including inactive hepatitis B virus (HBV) car-
riers, chronic HBV or hepatitis C virus (HCV) infection,
and liver cirrhosis [13]; ICPis used, including nivolumab,
pembrolizumab, atezolizumab, durvalumab, ipilimumab, or
a combination of nivolumab and ipilimumab; concomitant
medications at ICPi initiation, including PPIs, non-steroi-
dal anti-inflammatory drugs (NSAIDs), diuretics, includ-
ing furosemide, thiazides, and spironolactone, and anti-
biotics; kidney biopsy data reviewed by a board-certified
renal pathologist (Y.O.); laboratory parameters, including
baseline sCr level, baseline estimated glomerular filtration
rate (eGFR) [14], chronic kidney disease (CKD) defined as
eGFR < 60mL/min/1.73 m2 at baseline, and dipstick urine
protein at the start of ICPi treatment (negative, ± , 1 + , 2 + ,
3 + , or 4 +); malignancy type, including lymphoma, lung,
gastric, colorectal, genitourinary, breast, and head and neck
cancers; and oral corticosteroid whose initial dose for treat-
ing ICPi-AKI was recorded. Dates of initial ICPi adminis-
tration, discontinuation, and resumption were also collected
from the medical records for each patient. All consecutive
patients receiving ICPis were included to reduce selection
bias.
Statistical analyses
Statistical analyses were performed using STATA version
15.1 software (StataCorp, College Station, TX). Continuous
variables are presented as mean and standard deviation (SD)
or median and associated interquartile range (IQRs). They
were compared using Student’s t-test or Wilcoxon rank-sum
test according to distribution. Categorical variables are pre-
sented as number and percentage, and they were compared
using a χ2 test. Univariate and multivariate logistic regres-
sion analyses were performed to identify risk factors for
ICPi-AKI and factors associated with ICPi-AKI resolution
among patients with ICPi-AKI. Odds ratios (ORs) and the
corresponding 95% CIs were calculated. The independent
variables included in the logistic regression analyses were
selected based on the findings of previous studies [4–8] and
our clinical experience. Age, sex, and variables in the uni-
variate analysis with a p value of < 0.05 were selected as
independent variables for the multivariate analysis.
Cox proportional hazards models were used to investi-
gate possible factors associated with mortality. Hazard ratios
(HRs) and the corresponding 95% confidence intervals (CIs)
were also calculated. Proportional hazards assumptions were
confirmed using a log–log plot of survival. Independent
variables included in the Cox proportional hazards mod-
els were selected based on the findings of previous studies
[4–8] and our clinical experience. ICPi-AKI incidence and
Clinical and Experimental Nephrology
1 3
resolution and variables in the univariate analysis with a
p value of < 0.1 were selected as independent variables for
the multivariate analysis. The Kaplan–Meier method was
used to estimate the incidence of death in patients with and
without ICPi-AKI and the results were compared using the
log-rank test.
Due to the limited number of patients, subgroup analyses
were not performed. Instead, sensitivity analyses were con-
ducted to assess the robustness of our results. The primary
analysis was repeated first, using only patients with lung can-
cer, second, excluding those who received pembrolizumab,
third, excluding those who had liver diseases at baseline,
and fourth, the starting point for the ICPi-AKI group was
set as the day of AKI occurrence to minimize the immortal
time bias. All the reported p values are two-sided. Signifi-
cane levels (α) were set at 0.05. Patients with missing data
were excluded from the analyses and those lost to follow-up
before death were not considered for further analysis.
Results
Study population andcharacteristics
Of the 154 potentially eligible patients who started ICPi
treatment, two were excluded as they had contrast-induced
nephropathy. None of the patients had pre-renal or post-renal
AKI. The remaining 152 patients who completed follow-
up were analyzed. The baseline characteristics of patients
according to the ICPi-AKI incidence are shown in Table1.
There were no differences in the baseline characteristics
between the ICPi-AKI and non-ICPi-AKI groups, except
for a higher proportion of liver diseases and higher num-
ber of patients treated with nivolumab and pembrolizumab
in the ICPi-AKI group. One patient (0.7%) with symptoms
compatible with tubular interstitial nephritis underwent renal
biopsy.
Incidence andrisk factors ofICPi‑AKI
Twenty-seven patients (18%) developed an ICPi-AKI event
at a median of 101 (IQR, 31–158) days after the initiation of
ICPi treatment: 17 patients (63%) with stage 1, eight (30%)
with stage 2, and two (7%) with stage 3. We performed
logistic regression analyses using clinical data, including
age, sex, liver disease, types of ICPis (nivolumab, pembroli-
zumab, atezolizumab, duralumab, and ipilimumab), dipstick
urine protein > 1 + , liver diseases, PPIs, ACCI, baseline sCr,
and CKD to evaluate risk factors for AKI (Table2). The
univariate analyses showed that liver disease was associ-
ated with a higher risk of ICPi-AKI development, but PPIs
were not. The multivariate analysis confirmed that liver dis-
ease was significantly associated with a higher risk of AKI.
Figure1 shows the proportion of patients with AKI stages
by ICPi type. Patients treated with pembrolizumab showed
a significantly higher ICPi-AKI incidence than those treated
with nivolumab (p = 0.009).
Mortality andpredictors ofdeath
Of the enrolled patients, 85 (59%) died during the follow-up.
Of the 27 patients with ICPi-AKI, 17 (63%) died during the
follow-up with death occurring at a median of 18days (IQR
0–194) after ICPi-AKI development. As shown in Fig.2, the
probability of survival in the ICPi-AKI and non-ICPi-AKI
groups was not significantly different. The univariate analy-
ses revealed that the presence of peritoneal metastasis was
associated with a higher risk of death. Sex and baseline sCr
level were associated with a lower risk of death in patients
receiving pembrolizumab than in those receiving nivolumab.
In contrast, the ICPi-AKI incidence, ICPi-AKI resolution,
CKD, and liver diseases were not associated with the risk of
death. The multivariate analysis adjusted for the ICPi-AKI
incidence revealed that ICPi-AKI resolution, sex, baseline
sCr level, CKD, the presence of peritoneal metastasis, and
ICPi type were consistently associated with a lower mortal-
ity risk in patients receiving pembrolizumab than in those
receiving nivolumab (Table2).
ICPi‑AKI resolution
Among the patients with ICPi-AKI, 19/27 (73%) showed
resolution.
Sensitivity analyses
When only patients with lung cancer were considered, the
association between ICPi-AKI and mortality (all causes) was
mostly unchanged, with the analysis only failing to detect a
significant difference (log-rank test, p = 0.486). Addition-
ally, the results did not change when we excluded patients
receiving pembrolizumab from the analysis (log-rank test,
p = 0.415) and patients with liver diseases at baseline (log-
rank test, p = 0.467). Finally, the association did not change
when the starting point was set as the day of AKI occurrence
for the ICPi-AKI group (log-rank test, p = 0.714) (Table3).
Discussion
The key findings of our study are as follows. First, we iden-
tified the presence of liver diseases as an independent risk
factor for ICPi-AKI. To our knowledge, this is the first study
on liver diseases as a predictor of ICPi-AKI incidence in
a cohort of Japanese patients receiving ICPis. The under-
lying mechanisms are likely multifactorial. One possible
Clinical and Experimental Nephrology
1 3
Table 1 Baseline characteristics of the study participants according to ICPi-AKI
ICPi-AKI immune checkpoint inhibitor-associated acute kidney injury, sCr serum creatinine, eGFR estimated glomerular filtration rate, CKD
chronic kidney disease, ACCI age-adjusted Charlson comorbidity index, HBV hepatitis B virus, HCV hepatitis C virus, NSAIDs non-steroidal
anti-inflammatory drugs, PPIs proton pump inhibitors, SD standard deviation
All (n = 152) Non-ICPi-AKI (n = 125) ICPi-AKI (n = 27) p value
Age (years ± SD) 67 ± 10 67 ± 10 67 ± 7 0.638
Women, no. (%) 38 (25) 30 (24) 8 (30) 0.713
Renal biopsy, no. (%) 1 (1) 0 (0) 1 (4) 0.641
Body mass index (kg/m2 ± SD) 22 ± 3 22 ± 3 22 ± 3 0.977
Baseline sCr [mg/dL; median (25%, 75%)] 0.78 (0.67, 0.93 0.78 (0.67, 0.94) 0.77 (0.57, 0.9) 0.321
eGFR [mL/min/1.73 m2; median (25%, 75%)] 72 (55, 87) 71 (54, 84) 76 (58, 100) 0.181
CKD, no. (%) 46 (30) 38 (30) 8 (30) 1.00
Dipstick urine protein, no. (%) 109 (72) 91 (73) 18 (67) 0.521
Negative, no. (%) 55 (51) 45 (50) 10 (55) 0.829
± , no. (%) 28 (26) 26 (29) 2 (11) 0.242
1 + , no. (%) 17 (16) 14 (15) 3 (17) 1.00
2 + , no. (%) 7 (6) 6 (7) 3 (17) 0.157
3 + , no. (%) 1 (1) 1 (1) 0 (0) 1.00
4 + , no. (%) 1 (1) 1 (1) 0 (0) 1.00
ACCI, points [median (25%, 75%)] 6 (6, 7) 6 (6, 7) 6 (6, 7) 0.728
Congestive heart failure, no. (%) 3 (2) 3 (2) 0 (0) 0.960
Cerebrovascular disease, no. (%) 13 (9) 12 (10) 1 (4) 0.539
Peptic ulcer disease, no. (%) 22 (14) 18 (14) 4 (15) 1.00
Liver disease, no. (%)
Inactive HBV carrier, no. (%)
Chronic HBV infection, no. (%)
Chronic HCV infection, no. (%)
Liver cirrhosis, no. (%)
7 (5)
2 (1)
0 (0)
3 (3)
2 (1)
2 (2)
2 (2)
0 (0)
0 (0)
0 (0)
5 (19)
0 (0)
0 (0)
3 (11)
2 (7)
< 0.001
Diabetes mellitus, no. (%) 27 (18) 19 (15) 8 (30) 0.133
Hypertension, no. (%) 79 (52) 61 (49) 18 (67) 0.141
Furosemide, no. (%) 18 (12) 16 (13) 2 (7) 0.647
Thiazides, no. (%) 3 (2) 2 (1.5) 1 (4) 1.00
Spironolactone, no. (%) 8 (5) 6 (5) 2 (7) 0.940
NSAIDs, no. (%) 50 (33) 39 (31) 11 (41) 0.465
PPIs, no. (%) 66 (43) 52 (42) 14 (52) 0.447
Antibiotics, no. (%) 5 (3) 3 (2) 2 (7) 0.467
Nivolumab, no. (%) 79 (52) 71 (57) 8 (30) 0.01
Pembrolizumab, no. (%) 55 (35) 38 (30) 17 (63) 0.001
Atezolizumab, no. (%) 10 (7) 8 (6) 2 (7) 0.848
Durvalumab, no. (%) 8 (5) 8 (7) 0 (0) 0.173
Ipilimumab, no. (%) 2 (1) 2 (1.5) 0 (0) 0.508
Nivolumab and Ipilimumab, no. (%) 2 (1) 1 (1) 1 (4) 0.875
Lung cancer, no. (%) 95 (63) 74 (59) 21 (78) 0.071
Gastric cancer, no. (%) 25 (16) 22 (18) 3 (11) 0.409
Colorectal cancer, no. (%) 1 (1) 1 (1) 0 (0) 0.641
Genitourinary cancers, no. (%) 20 (13) 18 (14) 2 (7) 0.330
Breast cancer, no. (%) 1 (1) 1 (1) 0 (0) 0.641
Lymphoma, no. (%) 1 (1) 1 (1) 0 (0) 0.641
Head and neck cancers, no. (%) 9 (6) 8 (6) 1 (4) 0.590
Oral prednisolone, no. (%) 15 (10) 15 (56)
Oral prednisolone initial dose [mg/day; median
(25%, 75%)]
30 (20, 30) 30 (20, 30)
Clinical and Experimental Nephrology
1 3
mechanism is the concomitant use of ICPis and sofosbu-
vir-based antiviral therapy against chronic HCV infection,
which may have predisposed patients to develop ICPi-AKI.
For instance, several observational studies have reported
that sofosbuvir-based treatment may cause AKI, presum-
ably due to tubulointerstitial nephritis [15, 16]. We specu-
late that blocking the PD1 and CTLA4 pathways with ICPis
may have induced a loss of tolerance of memory T cells and
activated drug-specific T cells, predisposing the patient to
develop ICPi-AKI when concomitantly treated with sofosbu-
vir-based antivirals [5, 6, 17]. Another potential mechanism
was synergistic tubulointerstitial damage caused by ICPis
and cirrhosis-induced ammoniagenesis. Available evi-
dence suggests that untreated metabolic acidosis can cause
Table 2 Clinical factors
associated with the ICPi-AKI
incidence
ICPi-AKI immune checkpoint inhibitor-associated acute kidney injury, PPIs proton pump inhibitors, ACCI
age-adjusted Charlson comorbidity index, sCr serum creatinine, CKD chronic kidney disease, CI confi-
dence interval
Univariate analysis Multivariate analysis
Odds ratio 95% CI p value Odds ratio 95% CI p value
Age ≥ 65 1.36 0.51–3.65 0.541 0.88 0.30–2.62 0.823
Sex 0.75 0.30–1.89 0.541 0.85 0.30–2.46 0.766
Liver disease 13.9 2.55–76.6 0.002 11.1 1.82–67.6 0.009
Nivolumab Reference Reference
Pembrolizumab 3.86 1.52–9.77 0.004 3.80 1.39–10.3 0.009
Atezolizumab 2.16 0.39–11.9 0.384 1.67 0.26–10.7 0.586
Durvalumab
Ipilimumab
Dipstick urine protein > 1 + 1.50 0.54–4.18 0.438
PPIs 1.51 0.66–3.48 0.332
ACCI 0.99 0.63–1.55 0.948
Baseline sCr 0.26 0.04–1.67 0.156
CKD 0.96 0.39–2.39 0.937
Fig. 1 Proportion of patients with different stages of AKI according to the type of immune checkpoint inhibitor. ICPi-AKI immune checkpoint
inhibitor-associated acute kidney injury, AKI acute kidney injury. ⁑p < 0.05
Clinical and Experimental Nephrology
1 3
ammoniagenesis in the kidney, stimulating complement acti-
vation and deposition of complement C and complement
C5b-9, which can induce tubulointerstitial damage [18]. We
speculate that liver cirrhosis may increase systemic ammonia
Fig. 2 Kaplan–Meier curves showing mortality in the ICPi-AKI and non-ICPi-AKI groups of patients. ICPi-AKI, immune checkpoint inhibitor-
associated acute kidney injury
Table 3 Clinical factors
affecting all causes of mortality
ICPi-AKI immune checkpoint inhibitor-associated acute kidney injury, sCr serum creatinine, CKD chronic
kidney disease, ACCI age-adjusted Charlson comorbidity index, CI confidence interval
Univariate analysis Multivariate analysis
Hazard ratio 95% CI p value Hazard ratio 95% CI p value
ICPi-AKI incidence 0.92 0.53–1.59 0.765 0.85 0.45–1.61 0.628
ICPi-AKI resolution 2.73 0.55–13.5 0.217 7.07 0.59–84.8 0.123
Sex 0.58 0.36–0.93 0.023 0.81 0.44–1.49 0.505
Baseline sCr 0.29 0.12–0.70 0.006 0.36 0.07–1.79 0.212
CKD 0.64 0.39–1.05 0.078 0.92 0.42–2.02 0.841
Peritoneal metastasis 2.17 1.13–4.16 0.019 1.41 0.71–2.78 0.325
Dipstick urine protein > 1 + 0.97 0.55–1.73 0.925
Liver disease 0.32 0.08–1.32 0.116
ACCI 1.01 0.81–1.26 0.931
Nivolumab Reference Reference
Pembrolizumab 0.46 0.28–0.77 0.003 0.53 0.31–0.90 0.02
Atezolizumab 0.55 0.20–1.51 0.245 0.65 0.23–1.82 0.413
Durvalumab
Ipilimumab 2.00 0.27–14.7 0.498 1.41 0.71–2.78 0.325
Clinical and Experimental Nephrology
1 3
concentrations and trigger complement activation through
a similar mechanism, resulting in renal tubulointerstitial
damage. Thus, ICPis and liver cirrhosis may synergistically
contribute to the development of ICPi-AKI. However, our
findings could not prove these speculations as we did not
evaluate ammonia concentration in our data set and just one
patient underwent a kidney biopsy. Consistent with the find-
ings of previous studies [5, 6], we advocate the use of kidney
biopsies to accurately determine the cause of AKI in patients
receiving ICPis. Furthermore, we found that pembrolizumab
use was associated with a higher risk for ICPi-AKI. This
was not unexpected as the finding was consistent with that
of previous randomized-controlled trials [19–21]. Although
the exact mechanism of ICPi-induced nephrotoxicity has not
been fully defined, previous studies provide insights into this
issue, including the direct binding of ICPis to the surface
proteins of kidney cells, cross-reactivity of activated T-cells
with the kidney tissue, overproduction of pro-inflammatory
cytokines, and autoantibody formation [1]. Clearly, further
studies are needed to fully understand the underlying mecha-
nisms, which is essential to develop strategies to reduce the
risk of ICPi-AKI.
Second, ICPi-AKI did not affect mortality, which was
confirmed by the results of the sensitivity analyses. This is
not consistent with the findings of previous studies, which
reported that AKI increases mortality risk [7, 8]. A pos-
sible explanation for this discrepancy is our study may not
have had the statistical power to detect differences, as only
18% of our patients developed ICPi-AKI. Close monitoring
and intensive management of patients by experienced medi-
cal professionals may improve the survival of patients with
ICPi-AKI [22]. Regardless, given its rapid onset and high
mortality rate, our results suggest the necessity of timely rec-
ognition and management of this complication. For instance,
machine-learning prediction models and/or electronic AKI
alerts may prove beneficial. In a recent observational study, a
machine-learning prediction model accurately predicted the
risk of AKI [23]. Other studies have also shown that clinical
decision support systems and electronic AKI alerts contrib-
ute to improved outcomes of patients with AKI [24–26].
We anticipate future studies to evaluate whether they can
be successfully applied to in the management of ICPi-AKI.
Third, 70% of patients with ICPi-AKI showed renal func-
tion recovery, although ICPi-AKI resolution was not asso-
ciated with patient survival even after the adjustment for
covariates. Of note, our cohort presented a higher ICPi-AKI
recovery rate than the cohorts in a recent study [6]. This
can be attributed to the fact that 63% of our patients with
ICPi-AKI had stage 1 AKI and may have been more likely
to recover from AKI than those with more severe stages of
AKI. This may also be explained by the fact that ICPis were
withdrawn by all patients with ICPi-AKI and only two of the
patients resumed ICPi treatment after complete resolution
of their ICPi-AKI. While a guideline recommends gluco-
corticoid therapy [27], the results of previous observational
studies are controversial [28, 29]. Investigating whether this
particular patient population may benefit from glucocorti-
coid therapy is a future research interest.
Finally, we could not demonstrate an association between
PPIs and ICPi-AKI, inconsistent with the findings of previ-
ous studies [5, 6]. In our cohort, PPIs were used by over 60%
of the patients at the time of ICPi initiation, comparable
with the proportion in previous studies [5, 6]. The reason
for the inconsistency remains uncertain, but it may be due
to racial differences or related to the duration and/or dosage
of PPIs used.
Our study had some limitations. First, the generalizabil-
ity of our findings may be limited as our study population
comprised only Japanese patients. Second, given its obser-
vational design, residual confounding factors may exist.
Future studies using prospective enrollment and data col-
lection from the initiation of ICPis should be performed to
validate our findings. Third, non-protocolized collection
of laboratory data, such as the lack of controlled intervals,
may cause ascertainment bias, and it is possible that we
missed several AKI cases. Fourth, our multivariable mod-
els may have been underpowered and potential associations
may have been missed owing to the low rates of primary
outcome events. Fifth, it is noteworthy when interpreting
our results that only one patient underwent kidney biopsy
to determine the cause of AKI. Sixth, there is a possibility
that some patients in the ICPi-AKI group might have had
alternative etiologies of AKI because just one patient under-
went renal biopsy. Finally, owing to the small number of
patients, our study was not adequately powered to detect all
potentially significant differences; however, we conducted
this study using the maximum available number of patients
in our medical center.
Even with these limitations, our study had several
strengths. First, we were able to follow-up all patients
throughout the study period. This contributed to the com-
pleteness of the data and overcame weaknesses of other mul-
ticenter studies, where selective laboratory testing may have
led to missing follow-up data. Second, we included patients
with stage 1 AKI because a previous study proved that even
a small increase in the sCr level is associated with higher
mortality [7].
Overall, we determined that ICPi-AKI did not affect mor-
tality in patients receiving ICPis. Furthermore, liver diseases
and pembrolizumab use were independent risk factors for
developing ICPi-AKI in Japanese patients. Even with the
small number of patients, our study is significant because
it is the first to evaluate ICPi-AKI in an Asian population.
Our study has several clinical and research-related implica-
tions that will be beneficial in improving the care of patients
with cancer treated with ICPis. Our findings warrant future
Clinical and Experimental Nephrology
1 3
multi-center investigations on optimal management strate-
gies for ICPi-AKI.
Acknowledgments We would like to thank Editage (www.edita ge.jp)
for English language editing.
Authors’ contributions Conceptualization: YS and HT; Methodology:
YS, SW, KA, and TM; Formal analysis and investigation: YS, YO,
and SW; Writing: original draft preparation: YS; Writing: review and
editing: HT; Supervision: HT.
Compliance with ethical standards
Conflict of interest The authors have declared that no conflict of inter-
est exists.
Ethical approval All procedures performed in studies involving human
participants were in accordance with the ethical standards of the insti-
tutional and/or national research committee at which the studies were
conducted (IRB approval number 2-019096-00) and with the 1964
Helsinki declaration and its later amendments or comparable ethical
standards.
Informed consent Informed consent was obtained from all individual
participants included in the study.
References
1. Sury K, Parezella MA, Shirali AC. Cardiorenal complications of
immune checkpoint inhibitors. Nat Rev Nephrol. 2018;14:571–88.
2. Cortazar FB, Marrone KA, Troxell ML, Ralto KM, Hoenig MP,
Brahmer JR, etal. Clinicopathological features of acute kidney
injury associated with immune checkpoint inhibitors. Kidney Int.
2016;90:638–47.
3. Shirali AC, Perazella MA, Gettinger S. Association of acute inter-
stitial nephritis with programmed cell death 1 inhibitor therapy in
lung cancer patients. Am J Kidney Dis. 2016;68:287–91.
4. Mamlouk O, Selament U, Machado S, Abdelrahim M, Glass WF,
Tchakarov A, etal. Nephrotoxicity of immune checkpoint inhibi-
tors beyond tubulointerstitial nephritis: Single-center experience.
J Immunother Cancer. 2019;7:2.
5. Seethapathy H, Zhao S, Chute DF, Zubiri L, Oppong Y, Strohbehn
I, etal. The incidence, causes, and risk factors of acute kidney
injury in patients receiving immune checkpoint inhibitors. Clin J
Am Soc Nephrol. 2019;14:1692–700.
6. Cortarzar FB, Kibbelaar ZA, Glezerman IG, Budayyeh A, Mam-
louk O, Motwani SS, etal. Clinical features and outcomes of
immune checkpoint inhibitor-associated AKI: A multicenter
study. J Am Soc Nephrol. 2020;31:435–46.
7. Sawhney S, Marks A, Fluck N, Levin A, Prescott G, Black C.
Intermediate and long-term outcomes of survivors of acute kid-
ney injury episodes: a large population-based cohort study. Am J
Kidney Dis. 2017;69:18–28.
8. Fujii T, Uchino S, Doi K, Sato T, Kawamura T, JAKID study
group. Diagnosis, management, and prognosis of patients with
acute kidney injury in Japanese intensive care units: the JAKID
study. J Crit Care. 2018;47:185–91.
9. Tabei A, Watanabe M, Ikeuchi H, Nakasatomi M, Sakairi T,
Kaneko Y, etal. The analysis of renal infiltrating cells in acute
tubulointerstitial nephritis induced by anti-PD-1 antibodies: a case
report and review of literature. Intern Med. 2018;57:3135–9.
10. The STROBE checklist. http://www.plosm edici ne.org/.
Accessed 6 Mar 2020
11. Kellum JA, Lameire N, Aspelin P, Barsoum RS, Burdmann
EA, Goldstein SL, etal. Kidney disease: Improving Global
Outcomes (KDIGO) acute kidney injury work group. KDIGO
clinical practice guideline from acute kidney injury. Kidney Int
Suppl. 2012;2:1–138.
12. Sndararajan V, Henderson T, Perry C, Muggivan A, Quan H,
Ghali WA. New ICD-10 version of the Charlson comorbid-
ity index predicted in-hospital mortality. J Clin Epidemiol.
2004;57:1288–94.
13. Goldman L, Schafer AI. Chronic viral and autoimmune hepati-
tis. In: Goldman-Cecil Medicine. 25th ed. Philadelphia: Elsevier
Saunders; 2014. p. 993–1006.
14. Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, etal.
Collaborators developing the Japanese equation for estimated
GFR Revised equations for estimated GFR from serum creatinine
in Japan. Am J Kidney Dis. 2009;53:982–92.
15. Maan R, Al Marzooqi SH, Klair JS, Karkada J, Cerocchi O, Kow-
gier M, etal. The frequency of acute kidney injury in patients with
chronic hepatitis C virus infection treated with sofosbuvir-based
regimens. Aliment Pharmacol Ther. 2017;46:46–55.
16. Brown PR, Sadiq O, Weick A, Lenhart A, Elbatta M, Fernan-
dez C, etal. Acute kidney injury in patients undergoing chronic
hepatitis C virus treatment with Ledipasvir/Sofosbuvir. Hepatol
Commun. 2018;24:1172–8.
17. Gibson A, Gaulkner L, Lichtenfels M, Ogese M, Park BK, Naisbitt
DJ. The effect of inhibitory signals on the priming of drug hapten-
specific T cells that express distinct Vβ receptors. J Immunol.
2017;199:1223–7.
18. Wesson DE, Buysse JM, Bushinsky DA. Mechanisms of meta-
bolic acidosis-induced kidney injury in chronic kidney disease. J
Am Soc Nephrol. 2020. https ://doi.org/10.1681/ASN.20190 70677
(Accessed March 6, 2020).
19. Herbst RS, Baas P, Kim DW, Felip E, Pérez-Gracia JL, Han
JY, etal. Pembrolizumab versus docetaxel for previously
treated, PD-L1-positive, advanced non-small-cell lung can-
cer (KEYNOTE-010): a randomized controlled trial. Lancet.
2016;387:1540–50.
20. Reck M, Rodríguez-Abreu D, Robinson AG, Hui R, Csőszi
T, Fűlőp A, etal. Pembrolizumab versus chemotherapy for
PD-L1-positive non-small-cell lung cancer. N Engl J Med.
2016;375:1823–33.
21. Bellmunt J, Vaughn WDJ, Lee FJL, Fong L, Vogelzang NJ,
Climent MA, etal. Pembrolizumab as second-line therapy for
advanced urothelial carcinoma. N Engl J Med. 2017;376:1015–26.
22. Rizo-Topete LM, Rosner MH, Ronco C. Acute kidney injury risk
assessment and nephrology rapid response team. Blood Purif.
2017;43:82–8.
23. Koyner JL, Carey KA, Edelson DP, Churpek MM. The develop-
ment of a machine learning inpatient acute kidney injury predic-
tion model. Crit Care Med. 2018;46:1070–7.
24. Mendu ML, Ciociolo GR Jr, Mclaughlin SR, Graham DA, Ghaz-
inouri R, Parmar S, etal. A decision-making algorithm for initia-
tion and discontinuation of RRT in severe AKI. Clin J Am Soc
Nephrol. 2017;12:228–36.
25. Park S, Baek SH, Ahn S, Lee KH, Hwang H, Ryu J, etal. Impact
of electronic acute kidney injury (AKI) alerts with automated
nephrologist consultation on detection and severity of AKI: a
quality improvement study. Am J Kidney Dis. 2018;71:9–19.
26. Al-Jaghbeer M, Dealmeida D, Bilderback A, Ambrosino R, Kel-
lum JA. Clinical decision support for in-hospital AKI. J Am Soc
Nephrol. 2018;29:654–60.
27. Brahmer JR, Lacchetti C, Schneider BJ, Atkins MB, Brassil KJ,
Caterino JM, etal. Management of immune-related adverse events
in patients treated with immune checkpoint inhibitor therapy:
Clinical and Experimental Nephrology
1 3
American Society of Clinical Oncology clinical practice guide-
line. J Clin Oncol. 2018;36:1714–68.
28. González E, Gutiérrez E, Galeano C, Chevia C, de Sequera P,
Bernis C, etal. Early steroid treatment improves the recovery
of renal function in patients with drug-induced acute interstitial
nephritis. Kidney Int. 2008;73:940–6.
29. Fernandez-Juarez G, Perez JV, Caravaca-Fontán F, Quintana
L, Shabaka A, Rodriguez E, etal. Duration of treatment with
corticosteroids and recovery of kidney function in acute interstitial
nephritis. Clin J Am Soc Nephrol. 2018;13:1851–8.
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.