Acute kidney injury following hepatectomy for hepatocellular carcinoma: incidence, risk factors and prognostic value

Abstract

Background

Acute kidney injury (AKI) following hepatectomy remains understudied in terms of diagnosis, severity, recovery and prognostic value. The aim of this study was to assess the risk factors and prognostic value of AKI on short- and long-term outcomes following hepatectomy for hepatocellular carcinoma (HCC).

Method

This is a retrospective analysis of a single-center cohort of 457 consecutive patients who underwent hepatectomy for HCC. The KDIGO criteria were used for AKI diagnosis. The incidence, risk factors, and prognostic value of AKI were investigated.

Results

AKI occurred in 67 patients (15%). The mortality and major morbidity rates were significantly higher in patients with AKI (37% and 69%) than in those without (6% and 22%; p < 0.001). Renal recovery was complete in 35 (52%), partial in 25 (37%), and absent in 7 (11%) patients. Advanced age, an increased MELD score, major hepatectomy and prolonged duration of operation were identified as independent predictors of AKI. AKI was identified as the strongest independent predictor of postoperative mortality but did not impact survival.

Conclusion

AKI is a common complication after hepatectomy for HCC. Although its development is associated with poor short-term outcomes, it does not appear to be predictive of impaired long-term survival.

Introduction

Approximately 30–40% of patients of acute kidney injury (AKI) occur in postoperative settings.^{1, 2, 3} AKI, which has mostly been studied in cardiothoracic surgery,¹ has been associated with increased morbidity and mortality,^{4, 5, 6} higher costs of care,⁷ and worse long-term survival.^{8, 9}

To date, only 6 studies have assessed the frequency and impact of AKI on outcomes following liver resection.^{10, 11, 12, 13, 14, 15, 16} Using various diagnostic criteria, the incidence of AKI ranges from 3% to 17%, and 3 recent studies have confirmed the association of AKI development with short-term mortality.^{12, 14, 16} Whether AKI following hepatectomy also worsens the long-term outcomes remains unknown because of the exclusive focus of past studies on short-term mortality. Of particular note is the lack of well-established risk factors for AKI in the context of hepatic resection; such information is of the utmost importance to help patients and surgeons during the decision process leading to surgery. The latter question, the aforementioned caveats of previous reports (which are scarce), and the opportunity for improved standardization resulting from the recent international definition of AKI endorsed by Kidney Disease Improving Global Outcomes (KDIGO) in 2012 inspired the present study.¹⁷

The objective of this retrospective study was to apply the KDIGO criteria to diagnosis, severity, and recovery to scrutinize the incidence, clinical course, risk factors, and prognostic value of AKI regarding short- and long-term outcomes in a large single-center cohort of consecutive patients who underwent liver resection for a homogenous indication, namely, hepatocellular carcinoma (HCC).

Methods

Study population

All patients who underwent liver resection for HCC between 1989 and 2014 were identified from a prospectively maintained data file. This study was approved by the Institutional Review Board of the center and conformed to the ethical guidelines of the 1975 Declaration of Helsinki.

Selection criteria for liver resection

Patients were selected for surgery provided the following criteria were fulfilled: (1) Child class A liver function; and (2) a sufficient volume of the future remnant liver. As previously reported,^{21, 22} portal hypertension was not considered an absolute contraindication for hepatectomy. The final decision to proceed to surgery relied on multidisciplinary meetings.

Intraoperative anesthesia and surgical technique

The anesthetic approach to liver resection has been previously described.^{18, 19} Anesthesia was specifically adapted to the risks of massive hemorrhage, gas emboli, general hypothermia, abrupt hemodynamic changes, and coagulation disorders following ischemia-reperfusion injury.²⁰ Normovolemia was restored upon completion of the resection by fluid expansion using crystalloid solutions. Blood products were transfused to maintain a hemoglobin level >9 g/dl and a platelet count >50 × 10⁹/l.

Hepatectomies were performed using an open or laparoscopic approach, as previously described.^{18, 19, 21} Major hepatectomy was defined as the resection of 3 or more Couinaud segments.²² When needed, major hepatectomy was prepared by portal vein embolization to achieve a future remnant liver ≥40% of the whole non-tumorous liver volume.²³ The type and duration of vascular clamping and the transfusion needs were recorded. Abdominal drains were placed near the transection surface in all patients.

AKI diagnostic criteria, severity staging and recovery definition

The criteria for AKI diagnosis, severity, and recovery provided by the KDIGO guidelines were used.²⁴ AKI was defined as either 26 μmol/l (0.3 mg/dl) within 2 days after surgery or a 50% increase above the baseline serum creatinine (sCr) within 7 days after surgery. The urinary output was not included because of missing data. The severity of AKI was stratified according to the ratio of the highest sCr value reached after surgery to the baseline sCr^{17, 24}: stage 1 = 50%–99% increase; stage 2 = 100%–199% increase; and stage 3 = 200% or greater increase, an increase in sCr to 354 μmol/l (or 4 mg/dl) or greater, or any requirement for renal replacement therapy (RRT).

Renal recovery was classified as complete (sCr returned to a level less than 50% above baseline at discharge), partial (increase in sCr > 50% above baseline sCr at discharge but without the need for RRT), and absent (RRT at the time of hospital discharge when not already needed before surgery). Baseline sCr was defined as the minimum value of sCr available within 3 months before surgery, including the value on the admission day.²⁵ At baseline, patients were classified as having chronic kidney injury (CKI) whenever the baseline estimated glomerula filtration rate (eGFR) was <60 ml/min/1.73 m².²⁶ The estimated GFR was calculated according to the Modification of Diet in Renal Disease Equation²⁷:

eGFR = 186.3 × [(serum creatinine at the time of measurement (μmol/l)/88.4)] ^−1.154 × (age ^−0.203) × 1.212 (if patient is West African) × 0.742 (if patient is female).

Indications for RRT included AKI onset with oliguria (<500 ml per day urine output), hyperkaliema, metabolic acidosis, uremia and/or fluid overload. Intermittent or continuous hemodialysis was performed at the discretion of the intensivist. Hemodialysis was performed the night before surgery for patients treated with chronic RRT.

Definitions of outcomes

Postoperative mortality was defined by death occurring within 90 days of surgery or at any time during hospitalization. A root-cause analysis of all postoperative deaths was performed by conducting a thorough review of all clinical data in the electronic medical records and charts.

Complications were assessed during the same period. The following complications were retrospectively classified according to consensus or acknowledged definitions: postoperative liver failure,²⁸ ascites,²⁹ biliary fistula,³⁰ and infection.²⁹ Postoperative complications were retrospectively classified according to the Dindo-Clavien classification and further classified as moderate (grade I or II) or severe (grade III or IV).³¹ The durations of stay in the intensive care unit (ICU) and in the hospital were recorded. No patient was lost to follow-up.

Statistical analysis

Continuous variables are presented as the means ± standard deviations or medians (interquartile range) and were compared using Student's t-test or the non-parametric Mann–Whitney test for 2 groups or using the ANOVA or Kruskal–Wallis tests for > 2 groups, depending on the (non-)normality of the variable distributions, as assessed with the Shapiro–Wilk test for normality. Categorical variables are presented as numbers (percentages) and were compared across groups using the χ² or Fischer's exact test, as appropriate.

Multivariate Cox analyses (for criteria for inclusion, see below) were performed to identify the independent predictors of binary outcomes, including postoperative mortality and development of AKI. Pre- and intraoperative potential risk factors were considered for AKI risk analysis, and the mortality risk analysis also included postoperative variables. Model discrimination was assessed by the area under the receiver operating curve (AUROC).

Overall survival (OS, assessed from the date of hepatectomy to the date of last follow-up or death) was analyzed after excluding patients who died within 90 days after surgery. The latter was studied because the inclusion of survivors from the postoperative period not only induced bias (as patients may not have had enough time to recover their renal function before they died) but also because being dead with recovered kidneys has little significance from a patient's perspective.³² Patients who underwent liver transplantation during the follow-up were censored at the transplantation date. Survival curves were computed using the Kaplan–Meier method and p values for comparison were computed with the log-rank test. Model discrimination was evaluated by Harrell's concordance index. Potential predictors for entry in the multivariate analyses (Cox and logistic models) were selected based on a univariate screening using a liberal p value of ≤0.1; these variables were then considered for multivariate analysis. All tests were two-tailed, and a p value < 0.05 was considered to be statistically significant. Statistical analyses were performed using Stata v14.1 statistical software (StataCorp, College Station, TX).

Results

From 1989 to 2014, 485 patients who underwent liver resection for HCC were identified from a prospectively maintained data file. The study population consisted of 457 HCC patients after excluding 28 (6%) patients without available baseline sCr levels.

Study population

Pre-, intra-, and postoperative characteristics of the study population are summarized in Table 1, Table 2, Table 3. During the postoperative period, AKI occurred in 67 (15%) patients. The incidence of AKI over time was 27% (15/56), 14% (21/151), and 12% (31/250) during the periods 1989–1997, 1998–2005, and 2006–2014, respectively (global p = 0.026; p for trend = 0.019). The AKI severity was stage 1 in 43 (64%), stage 2 in 12 (18%), and stage 3 in 12 (18%) patients. Postoperative mortality was significantly correlated with increased severity (26%, 33%, and 83% for stages 1, 2, and 3, respectively, p < 0.001). The root causes of mortality included liver failure (n = 24, 50%), sepsis (n = 6, 13%), pneumonia (n = 3, 7%), hemorrhage (n = 2, 4%), tumor recurrence (n = 2, 4%), cardiogenic shock (n = 1, 2%), gas embolism (n = 1, 2%), and pulmonary embolism (n = 1, 2%). In the remaining 8 patients (17%), a single root cause could not be identified, and death resulted from a mixture of liver insufficiency, sepsis, and multiple organ failure.

Table 1.

Preoperative patient characteristics of the whole series according to the occurrence of postoperative AKI

Variables	Whole series N = 457	No AKI N = 390 (85%)	AKI patients N = 67 (15%)	P Value
Patient characteristics
Age (years)	60.5 ± 13.4	60 ± 14	64 ± 10	0.027
Gender, male	379 (83)	319 (82)	60 (90)	0.119
Body mass index (kg/m2)	26 ± 11	25.4 ± 4.5	29.6 ± 28.4	0.011
ASA class ≥ III	125 (27)	101 (26)	24 (36)	0.092
Comorbidities
Chronic pulmonary disease	35 (8)	26 (7)	9 (13)	0.054
Cardiovascular disease	31 (7)	22 (6)	9 (13)	0.019
Diabetes mellitus	75 (16)	65 (17)	10 (15)	0.722
Chronic kidney injury	70 (15)	52 (13)	18 (27)	0.005
Body mass index ≥ 30 kg/m2	57 (13)	49 (13)	8 (12)	0.887
Underlying liver disease
Viral hepatitis	219 (48)	185 (47)	34 (51)	0.440
Alcohol	71 (16)	59 (15)	12 (18)
Hemochromatosis	23 (5)	18 (5)	5 (7.5)
Metabolic syndrome	34 (7)	32 (8)	2 (3)
Autoimmune	3 (0.5)	2 (0.5)	1 (1.5)
Healthy liver	107 (23)	94 (24)	13 (19)
Cirrhosis	206 (45)	171 (44)	35 (52)	0.202
Portal hypertension	114 (25)	98 (25)	16 (24)	0.827
MELD score	8.5 ± 2.6	8.4 ± 2.5	9.3 ± 3.2	0.013
BCLC stage A vs B or C	274 (60)/183 (40)	240 (60)/150 (40)	34 (51)/33 (49)	0.160
Liver and kidney function tests
Platelet count (×103/mm3)	209 ± 117	207 ± 113	219 ± 143	0.447
Bilirubin (μmol/l)	14 ± 10	13 ± 8	19 ± 18	<0.001
AST (IU/l)	63 ± 61	61 ± 59	72 ± 73	0.198
Prothrombin time (% of normal)	87 ± 13	87 ± 13	87 ± 14	0.772
Albumin (g/l)	40.6 ± 5.8	40.6 ± 6.1	40.7 ± 4.6	0.941
Baseline serum creatinine (μmol/l)	89 ± 29	87 ± 25	96 ± 45	0.018
Baseline eGFR (ml/min/1.73 m2)	85 ± 31	86 ± 29	83 ± 39	0.502

Footnotes: AKI, acute kidney injury; SD, standard deviation; ASA, American Society of Anesthesiologists; eGFR, estimated glomerular filtration rate; MELD, Model for End-stage Liver Disease; BCLC, Barcelona Clinic Liver Cancer; AST, aspartate aminotransferase.

No AKI includes patients without any occurrence of KDIGO criteria.

Categorical variables are presented as numbers (percentages) and continuous variables are expressed as the means ± standard deviation.

Table 2.

Intraoperative patient characteristics of the whole series and according to the occurrence of postoperative AKI

Variables	Whole series N = 457	No AKI N = 390 (85%)	AKI patients N = 67 (15%)	P Value
Surgical approach				0.046
Laparoscopic	128 (28)	116 (30)	12 (18)
Open	329 (72)	274 (70)	55 (82)
Major hepatectomy	241 (53)	194 (50)	47 (70)	0.002
Any type of vascular clamping	296 (65)	249 (64)	47 (70)	0.318
Type of clamping				0.535
Pedicle intermittent	188 (41)	160 (41)	28 (42)
Pedicle continuous	63 (14)	53 (14)	10 (15)
Standard total vascular exclusion	34 (7)	26 (7)	8 (12)
Vascular exclusion preserving the caval flow	11 (2)	10 (3)	1 (1.5)
No clamping	161 (35)	141 (36)	20 (30)
Clamping time (minutes)	22 ± 23	22 ± 22	26 ± 27	0.163
Red blood cell transfusion
Yes	82 (18)	61 (16)	21 (31)	0.002
Number of red blood cells units transfused	0.8 ± 2.8	0.7 ± 2.0	1.8 ± 5.5	0.003
Duration of operation (minutes)	208 ± 81	202 ± 76	249 ± 97	<0.001
Duration of operation (minutes) > 300 min	33 (7)	10 (3)	23 (34)	0.008

Footnotes: AKI, acute kidney injury; KDIGO, Kidney Disease Improving Global Outcomes; SD, standard deviation.

No AKI includes patients without any occurrence of KDIGO criteria.

Categorical variables are presented as numbers (percentages) and continuous variables are expressed as the means ± standard deviation.

Table 3.

Short-term outcomes of the whole series according to the occurrence of postoperative AKI

Outcomes	Whole series N = 457	No AKI N = 390 (85%)	AKI patients N = 67 (15%)	P Value
90-Day mortality	48 (11)	23 (6)	25 (37)	<0.001
Overall morbidity	184 (40)	138 (35)	46 (69)	<0.001
Major morbidity (≥grade III)b	60 (13)	45 (12)	15 (22)	0.015
Liver failure	69 (15)	47 (12)	22 (33)	<0.001
Ascites	70 (15)	53 (14)	17 (25)	0.013
Biliary fistula	9 (2)	5 (1)	4 (6)	0.011
Pulmonary complications	66 (14)	49 (13)	17 (25)	0.006
Infectious complications	73 (16)	50 (13)	23 (34)	<0.001
Need for RRTa	5 (1)	0	5 (8)	<0.001
Renal function recovery				<0.001
Complete recovery	35 (8)	–	35 (52)
Partial recovery	25 (6)	–	25 (37)
No recovery	7 (2)	–	7 (11)
Duration of stay, days
In ICU	6 ± 8	5 ± 7	12 ± 14	<0.001
In Hospital	13 ± 11	12 ± 9	20 ± 18	<0.001

Footnotes: AKI, acute kidney injury; SD, standard deviation; RRT, renal replacement therapy; ICU, intensive care unit.

No AKI includes patients without any occurrence of KDIGO criteria.

Categorical variables are presented as numbers (percentages) and continuous variables are expressed as the means ± standard deviation.

^aRRT included dialysis or hemodiafiltration.

^bDindo-Clavien classification.

Independent predictive factors for AKI are shown in Table 4. The model discrimination analysis yielded an AUROC of 0.72. Independent predictors of postoperative mortality are shown in Table 5.

Table 4.

Univariate and multivariate analyses of predictive factors for AKI

Variable	Univariate analysis	Multivariate analysis
P Value	Odds ratio (95% Confidence interval)	P Value
Demographics
Age (years)a	0.027	1.030 (1.001–1.050)	0.040
Female sex (yes vs no)	0.119
Clinical variables
Cirrhosis (yes vs no)	0.202
Portal hypertension (yes vs no)	0.827
ASA ≥ III (yes vs no)	0.092		0.411
Chronic pulmonary disease (yes vs no)	0.054		0.190
Cardiovascular disease (yes vs no)	0.019		0.210
Diabetes mellitus (yes vs no)	0.722
Chronic kidney injury (yes vs no)	0.005		0.07
Body mass index ≥ 30 kg/m2 (yes vs no)	0.887
eGFR (ml/min/1.73 m2)a	0.502
MELD scorea	0.013	1.000 (1.001–1.050)	0.040
BCLC stage (A vs B or C)	0.160
Intraoperative variables
Extent of liver resection (minor vs major)	0.002	2.7 (1.5–5.0)	0.001
Vascular clamping (yes vs no)	0.318
Clamping time (min)a	0.163
Packed red blood cell transfusion (yes vs no)	0.002		0.30
Duration of operation > 300 min (yes vs no)	0.008	2.4 (1.0–5.9)	0.05

Footnotes: ASA, American Society of Anesthesiologists; eGFR, estimated glomerular filtration rate; MELD, Model for End-stage Liver Disease; BCLC, Barcelona Clinic Liver Cancer.

^aContinuous variables.

Table 5.

Univariate and multivariate analyses of risk factors for postoperative mortality

Variable	Univariate analysis	Multivariate analysis
P Value	Odds ratio (95% Confidence interval)	P Value
Demographics
Age (years)a	0.085		0.410
Female sex (yes vs no)	0.006		0.471
Clinical variables
Cirrhosis (yes vs no)	0.024		0.252
Portal hypertension (yes vs no)	0.475
ASA ≥ III (yes vs no)	0.007		0.880
Chronic pulmonary disease (yes vs no)	0.698
Cardiovascular disease (yes vs no)	0.004		0.190
Diabetes mellitus (yes vs no)	0.090	3.2 (1.3–8.2)	0.015
Chronic kidney injury (yes vs no)	0.017		0.500
Body mass index ≥ 30 kg/m2 (yes vs no)	0.359
eGFR (ml/min/1.73 m2)a	0.778
MELD scorea	<0.002	1.2 (1.0–1.3)	0.018
BCLC stage (A vs B or C)	0.006		0.459
Intra-operative variables
Extent of liver resection (minor vs major)	0.082		0.768
Vascular clamping (yes vs no)	0.027		0.633
Clamping time (min)a	0.053		0.995
Packed red blood cell transfusion (yes vs no)	<0.001	2.9 (1.2–6.7)	0.016
Duration of operation > 300 min (yes vs no)	0.37
Complications
AKI (yes vs no)	<0.001	6.6 (2.8–15.5)	<0.001
Liver failure (yes vs no)	<0.001	5.7 (1.8–17.6)	0.003
Ascites (yes vs no)	<0.001		0.785
Biliary fistula (yes vs no)	0.952
Infectious complications (yes vs no)	<0.001		0.696
Pulmonary complications (yes vs no)	<0.005		0.593

Footnotes: ASA, American Society of Anesthesiologists; eGFR, estimated glomerular filtration rate; APRI, aminoaspartate transferase to platelet ratio index; MELD, Model for End-stage Liver Disease; BCLC, Barcelona Clinic Liver Cancer; AKI, acute kidney injury; KDIGO, Kidney Disease Improving Global Outcomes.

^aContinuous variables.

Prognostic value of AKI for long-term survival

At the end of the study period, 205 (43%) patients had died. The prognostic value of AKI development on the long-term outcomes was measured in 409 survivors of the postoperative period. For the survival calculation using the Kaplan–Meier method, the 30 patients who underwent salvage liver transplantation for tumor recurrence were censored at the time of transplant as alive with recurrence.

After a median follow-up at 27 months (range, 3–252 months) post-hepatectomy, the OS rates at 1, 3, and 5 years were 85%, 67%, and 60%, respectively. The OS at 5 years was similar in patients who developed AKI (n = 367) and those who did not (n = 42) (54% vs 61%, respectively, p = 0.208) and in the subset of AKI patients (n = 42), in those who completely recovered at discharge (n = 29) and those who did not (n = 13) (33% vs 61%, respectively, p = 0.943).

Discussion

In the present series, AKI developed frequently following hepatectomy for HCC and was a strong independent predictor of postoperative mortality. More interestingly, the OS rates were not affected by the development of AKI or the absence of complete recovery after AKI developed, as shown using the most recent AKI diagnostic criteria¹² in a large, single-center cohort of consecutive patients with various demographic, tumor, and operative characteristics.

Similar to the results from a previous report,³³ the postoperative mortality rate in the present study was 11% among the whole study population. Except for diabetes mellitus and AKI (see below), the 3 other independent predictors of mortality identified here (Table 5)—the MELD score,³⁴ intraoperative blood cell transfusion,^{35, 36, 37, 38} and liver failure^{28, 37}—have already been extensively validated and do not require further comment. The postoperative mortality rate rose to 37% in patients who developed AKI, which is much higher than reported elsewhere (range: 14%–26%) in the same hepatectomy setting.^{10, 11, 12, 13, 14, 15, 16} This discrepancy may have resulted from the following: (i) the KDIGO criteria, which capture not only severe AKI (i.e., >50% increase above baseline sCr) but also increases in sCr as small as 26 μmol/l; and (ii) the 90-day delay (to capture surgery-related events and exclude disease-related events) used to measure postoperative mortality compared to the 30-day delay used in some series,^{13, 14, 16} which underestimates the true mortality by approximately 50%.³⁹

Similarly to other reports, older age^{12, 14} and the need for major hepatectomy¹² were identified as independent predictors of AKI (Table 4). Logically, the MELD score emerged as the third independent predictor. Indeed, 2 of its 3 components—serum bilirubin and creatinine—have already been identified as risk factors.^{12, 14} Yet, duration of operation was identified as an independent risk factor for AKI. This may be explained by the fact that the long period of hypotension-related anesthesia might impair renal function. The risk model for predicting AKI achieved a similar performance (AUC = 0.72) as those models used by Kambakamba et al. (AUC = 0.765)¹² and Slankamenac et al (AUC = 0.8).¹³ Most importantly, the current model identified mostly the variables that were available before operation, increasing its clinical value. The strong independent correlation between the development of AKI and the 90-day mortality underlines the clinical importance of the identification of patients at high-risk of this complication to improve the following: (i) the inclusion of advice from a nephrologist in selecting potential candidates for surgery; (ii) the efficacy-efficiency balance of biomarker measurements for earlier diagnosis of AKI²⁴ in the postoperative period; and (iii) adequate volume expansion, use of diuretics, or administration of vasoactive drugs⁴⁰ and even earlier initiation of RRT postoperatively before fulfillment of the usual indication criteria.⁴¹

The present study also identified diabetes and intraoperative transfusion as independent predictors of postoperative mortality. Whereas the impact of intraoperative blood transfusion on short-term outcome has been extensively reported,⁴² the current finding concerning diabetes mellitus remains debated^{43, 44} and must be confirmed by further studies.

To the best of the authors' knowledge, the present study is the first to assess the impact of AKI on long-term survival in the setting of liver surgery. In the present analysis, OS were similar regardless of whether patients developed AKI or not or if they had (at discharge) completely recovered from AKI or not after AKI developed. This absence of impact on the long-term outcomes compensates for the short term, poor prognostic value of AKI and should be considered during the decision process to achieve the patient's informed consent for high risk surgery.^{45, 46}

The shortcomings of the current study include its retrospective nature; the long period covered, and a probable patient selection bias inherent to a tertiary liver center. In addition, one third of the AKI detected over the study period were impacted on by pre-existing chronic kidney injury (prevalence in the previous series: 1%–19%),^{14, 16, 17, 18, 19, 20} which may reduce the number of AKI patients (approximately two third) that had their renal function impaired by the liver resection. This number may be judged not powered to show a relative impact of AKI on survival. Despite this, although not significant there was a 7% difference of OS between the two groups. In addition, it is intuitively clear that AKI influenced the other complications, and vice versa. However, as the chronology of complications could not be retrieved from the data file, this point deserves further evaluation. Finally, the manner and timing of intraoperative volume resuscitation after the hepatectomy is crucial to the prevention of postoperative AKI. However these data are difficult to capture in a retrospective study and practices regarding the management of intraoperative volume resuscitation may vary widely from one institution to another. This issue deserves further specific evaluation.

Conclusions

AKI following liver resection for HCC, as defined by the KDIGO criteria, is a frequent complication and a strong predictor of 90-day mortality. The development of AKI or even the absence of complete recovery does not impact long-term outcomes. In addition to the current knowledge, the present results suggest that AKI should be viewed as a dynamic process and a distinct therapeutic target. The short-term outcomes of liver resection for HCC might be improved by preventing AKI through a more stringent selection of HCC candidates for surgery and the early diagnosis and rapid treatment of AKI. The multidisciplinary meetings used to decide whether to proceed to surgery and postoperative management might be improved by involving nephrologists when discussing patients at risk of developing AKI.

Conflict of interest

None to declare.

Funding or grants

None to declare.