Acute kidney injury after major hepatectomy: association with postoperative complications and post-hepatectomy liver failure

Abstract

Background

Patients undergoing liver resections are at increased risk of acute kidney injury (AKI). The objective of this study was to examine AKI after major hepatectomy and the associated risk of adverse postoperative outcomes, including post-hepatectomy liver failure (PHLF) and mortality.

Methods

All adult patients undergoing major hepatectomy between 2010 and 2021 at three tertiary referral centres in Sweden and Germany were included. Postoperative AKI was based on standardized criteria using serum creatinine and urine output. Logistic regression models were used to explore associations with adverse events.

Results

In all, 1561 patients were included in the study, of whom 400 (25.6%) developed AKI of any grade. Risk factors for AKI included male sex, higher age, and a low preoperative estimated glomerular filtration rate. Patients with AKI were at increased risk of intensive care unit admission (adjusted odds ratio (aOR) 6.06; 95% confidence interval (c.i.) 2.83 to 12.95), Clavien–Dindo grade ≥ IIIa (aOR 2.08; 95% c.i. 1.61 to 2.67) and PHLF grade B or C (aOR 2.87; 95% c.i. 2.11 to 3.90). Patients with only AKI and patients with both AKI and PHLF grade B or C had an increased risk of 90-day mortality than patients with neither. Postoperative AKI was associated with an increased risk of 90-day mortality even after adjustment for PHLF.

Conclusion

AKI after major hepatectomy is a common complication and is associated with considerable morbidity and mortality. Postoperative AKI may be both an early risk factor for PHLF after major hepatectomy and independently contribute to 90-day mortality.

This study explored early acute kidney injury after major hepatectomy and its association with post-hepatectomy liver failure. One-quarter of all included patients developed early acute kidney injury, and these patients were also more likely to later experience post-hepatectomy liver failure.

Introduction

Acute kidney injury (AKI) is a common complication after gastrointestinal surgery with a prevalence of around 13%, with variations depending on the type of surgery and whether the surgery is elective or acute¹. A sudden, temporary, or sustained decrease in kidney function during the perioperative period is associated with both co-morbidity and morbidity after surgery², where also milder episodes can confer an increased risk of death³. Because a low estimated glomerular filtration rate (eGFR) increases the risk of postoperative AKI, and given the increase of chronic kidney disease (CKD) among the general population⁴, more patients undergoing surgery may be at risk of postoperative AKI. The most commonly used definition of AKI is the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines⁵, defining AKI by changes in both serum creatinine and urine volume output, because oliguria with normal creatinine levels can be a presentation of AKI⁶. Patients undergoing major hepatectomy, including associating liver partition and portal vein ligation for staged hepatectomy (ALPPS), are at increased risk of postoperative AKI than patients undergoing minor liver resections^7–9. It is assumed that, apart from general surgical risk factors such as high blood loss and long duration of surgery, there are also hepatectomy-specific risk factors for the development of AKI, including the hepatorenal syndrome¹⁰. A feared adverse event following major hepatectomy is post-hepatectomy liver failure (PHLF). AKI can be one feature of its presentation, although the extent to which immediate postoperative AKI contributes to the risk of developing PHLF or is a secondary effect of PHLF is unclear¹¹. The aim of this study was to investigate the impact of postoperative AKI after major hepatectomy on patient outcome using standardized criteria for AKI based on both creatinine measurements and urine volume output. Furthermore, a possible association of AKI with other postoperative complications, including PHLF, was explored.

Methods

Study population

All patients aged ≥ 18 years who underwent major hepatectomy (defined as three or more liver segments) between 2010 and 2021 at three tertiary referral centres for hepatopancreatobiliary surgery (Karolinska University Hospital, Linköping University Hospital, both in Sweden, and Asklepios Hospital Hamburg, Germany) were included. In Asklepios Hospital, ALPPS remained standard, whereas in the remaining institutions it was only performed in research studies or as a rescue treatment after failed hypertrophy. Patient-related data were retrospectively collected from digital patient record systems. Patients without any postoperative measurements of creatinine or urine volume output were excluded.

The study was approved by regional ethics committees and the requirement for informed consent was waived due to the retrospective nature of the study.

Exposures and measures

Postoperative AKI was defined using the KDIGO guidelines⁵. The guidelines define AKI as an increase in serum creatinine by ≥ 0.3 mg/dl (26.5 µmol/l) within 48 hours (h), or a ≥ 1.5-fold increase in serum creatinine from baseline within 7 days, or urine volume ≤ 0.5 ml/kg/h for 6 h⁵. In this study, to explore early AKI after major hepatectomy, only the first two postoperative days were used to define AKI. Further, because data on hourly urine output were not available, 24-h urine volume was used and AKI was defined as a urine volume of ≤ 0.5 ml/kg/h on either the first or second postoperative day. The grading of AKI (stage 1–3) was performed according to the KDIGO guidelines⁵, but stage 2 was defined as a urine output ≤ 0.5 ml/kg/h for 24 h on either the first or second postoperative day, and stage 3 was defined as a urine output ≤ 0.3 ml/kg/h for 24 h on either the first or second postoperative day. Table S1 provides details on missing data for serum creatinine and urine output, as well as the timing of AKI diagnosis. Preoperative eGFR was based on the last recorded serum creatinine before surgery, which was routinely collected the day before surgery. Standardized future liver remnant (sFLR; Table S2) was calculated using preoperative information on the future liver remnant based on computed tomography, and total liver volume was estimated using the Vauthey formula¹² based on body surface area according to the Mosteller formula¹³.

Outcomes

The outcomes of interest included postoperative AKI, as defined above, postoperative complications, and mortality. Postoperative complications included Clavien–Dindo grade ≥ IIIa¹⁴, a hospital stay > 14 days, and readmission within 90 days. Biliary leakage and postoperative haemorrhage were defined according to the International Study Group of Liver Surgery (ISGLS), and grade B and C were considered clinically significant^15,16. Intensive care unit (ICU) admission was standard for all operated patients at Asklepios Hospital and was therefore only evaluated in patients from centres without routine ICU admissions. PHLF was defined using the ISGLS criteria, with grades B and C considered clinically significant¹⁷. Time-to-event data for 90-day mortality are reported according to the presence of postoperative AKI and postoperative PHLF with or without postoperative AKI.

Statistical analysis

Descriptive statistics for continuous variables are presented as either the mean and standard deviation (s.d.) or median and interquartile range (i.q.r.), with data analysed using Student’s t test or the Mann–Whitney U test. Categorical variables are presented as counts and percentages and were analysed using the χ² test or Fischer’s exact test. To investigate intra- and postoperative risk factors for AKI after surgery, an analysis was performed using unadjusted and adjusted logistic regression models presenting odds ratios (ORs) with their 95% confidence interval, (c.i.). Multivariable models were adjusted for the predefined preoperative variables. AKI and postoperative complications were also analysed using unadjusted and adjusted logistic regression models with the adjusted models including the same variables. Univariable and multivariable Cox proportional hazards models, assessed with the Schoenfeld residuals test, were used for time-to-event data to assess 90-day mortality. Additional adjustment was performed for PHLF when the impact of AKI alone was assessed. Kaplan–Meier estimates of survival were used to visualize 90-day mortality. Analyses were performed using Stata^® version 16.1 software (StataCorp, College Station, TX, USA). Two-sided P < 0.05 was considered statistically significant.

Results

Study population

A total of 1573 patients underwent hepatectomy during the study period; 12 did not have postoperative measurements of creatinine or urine output allowing for AKI categorization and so were excluded. The remaining 1561 patients were included in the study. The median age at the time of surgery was 65 (i.q.r. 56–73) years and most (882 patients, 56.5%) were male (Table 1). As indicated in Table 2, the most common indication for surgery was colorectal liver metastases (742, 47.5%), followed by cholangiocarcinoma (295, 18.9%), hepatocellular carcinoma (188, 12.0%), and gallbladder cancer (31, 2%), whereas the most common procedure was right hemihepatectomy (736, 47.1%), followed by extended right hemihepatectomy (307, 19.7%), left hemihepatectomy (309, 19.8%), extended left hemihepatectomy (111, 7.1%), and ALPPS step 2 (98, 6%). All included procedures were open major hepatectomies. In total, 617 of all included patients (39.5%) received neoadjuvant oncological treatment (Table 1). The median follow-up time was 1.9 (i.q.r. 0.7–4.1) years.

Table 1.

Patient characteristics overall and according to the presence of postoperative AKI

	Postoperative AKI			P*
	No AKI (n = 1161)	AKI (n = 400)	Overall (n = 1561)
Baseline characteristics
Age (years), median (i.q.r.)	65 (55–72)	67 (57–73)	65 (56–73)	0.043
Missing	6 (0.5%)	0 (0%)	6 (0.4%)
BMI (kg/m2), mean(s.d.)	25.1(4.3)	27.6(4.8)	25.7(4.5)	< 0.001
Missing	5 (0.4%)	2 (0.5%)	7 (0.5%)
Sex				< 0.001
Female	558 (48.1%)	121 (30.3%)	679 (44.5%)
Male	603 (51.9%)	279 (69.7%)	882 (56.5%)
Preoperative creatinine† (µmol/l), mean(s.d.)	74.1(21.1)	81.9(27.7)	76.1(23.3)	< 0.001
Missing	44 (4%)	8 (2%)	52 (3%)
Preoperative eGFR (ml/min/1.73 m2), mean(s.d.)	86.5(18.9)	82.6(22.2)	85.5(19.9)	< 0.001
Missing	50 (4%)	8 (2%)	58 (4%)
Preoperative eGFR categories				< 0.001
≥90 ml/min/1.73 m2	506 (45.3%)	157 (40.0%)	663 (43.9%)
≥60 to < 90 ml/min/1.73 m2	509 (45.6%)	172 (44.9%)	681 (45.1%)
≥45 to < 60 ml/min/1.73 m2	75 (7%)	38 (10%)	113 (7.5%)
< 45 ml/min/1.73 m2	27 (2%)	25 (6%)	52 (3%)
Co-morbidities
Cardiovascular disease	440 (37.9%)	233 (58.3%)	673 (43.1%)	< 0.001
Missing	1 (0.1%)	0 (0%)	1 (0.06%)
Diabetes	136 (11.7%)	79 (20%)	215 (13.8%)	< 0.001
Missing	1 (0.1%)	0 (0%)	1 (0.06%)
Pulmonary disease	94 (8%)	34 (9%)	128 (8.2%)	0.807
Missing	2 (0.2%)	0 (0%)	2 (0.1%)
Disease characteristics
Liver disease
Fibrosis	546 (65.9%)	152 (67.9%)	698 (66.3%)	0.002
Cirrhosis	22 (2%)	14 (4%)	36 (2%)	0.065
Missing	333 (28.7%)	176 (44.0%)	509 (32.6%)
Diagnosis				< 0.001
CRLM	580 (50.0%)	162 (40.5%)	742 (47.6%)
CCC	193 (16.7%)	102 (25.6%)	295 (18.9%)
HCC	130 (11.2%)	58 (15%)	188 (12.1%)
Gallbladder cancer	21 (2%)	10 (3%)	31 (2%)
Other	235 (20.3%)	68 (17%)	303 (19.4%)
Missing	2 (0.2%)	0 (0%)	2 (0.1%)
Treatment or interventions
Neoadjuvant treatment	489 (42.4%)	128 (32.3%)	617 (39.8%)	< 0.001
Missing	7 (0.6%)	4 (1%)	11 (0.7%)
Portal vein embolization or ligation	197 (17.2%)	103 (26.3%)	300 (19.5%)	< 0.001
Missing	16 (1.4%)	9 (2.3%)	25 (1.6%)
Pringle manoeuvre	214 (18.4%)	88 (22%)	302 (19.4%)	0.114
Time of Pringle manoeuvre (min), mean(s.d.)	24.4(12.3)	26.4(13.9)	24.9(12.7)	0.338

Values are n (%) unless otherwise stated. The presence of postoperative AKI is based on definitions suggested by KDIGO. †Creatinine (mg/dL) = creatinine (µmol/L) ÷ 88.4. AKI, acute kidney injury; i.q.r., interquartile range; BMI, body mass index; s.d., standard deviation; eGFR, estimated glomerular filtration rate; CRLM, colorectal liver metastases; CCC, cholangiocarcinoma; HCC, hepatocellular carcinoma. *Analysed using Student‘s t-test or Mann-Whitney U-test for continous variables and chi-square or Fischer‘s exact test for categorical variables.

Table 2.

ORs for AKI according to perioperative characteristics

Perioperative characteristic	No. of individuals (%)	No. of patients with AKI (%)	ORs for developing AKI*
			Unadjusted	P	Adjusted†	P
Type of resection
Right hemihepatectomy	736 (47.2%)	182 (24.7%)	Reference		Reference
Left hemihepatectomy	309 (19.8%)	68 (22%)	0.86 (0.63, 1.18)	0.347	0.84 (0.60, 1.18)	0.324
Extended right	307 (19.7%)	88 (29%)	1.22 (0.91, 1.65)	0.186	1.34 (0.97, 1.85)	0.072
Extended left	111 (7.1%)	21 (19%)	0.71 (0.43, 1.18)	0.183	0.68 (0.40, 1.15)	0.148
ALPPS Step 2	98 (6%)	41 (42%)	2.19 (1.42, 3.38)	< 0.001	2.55 (1.55, 4.20)	< 0.001
Operation duration ≥ 4 h	997 (63.9%)	262 (26.3%)	1.10 (0.87, 1.40)	0.431	0.95 (0.74, 1.23)	0.716
Pringle manoeuvre	302 (19.4%)	88 (29%)	1.25 (0.95, 1.66)	0.115	1.16 (0.86, 1.56)	0.337
Blood loss (≥ 1000 ml)	1047 (67.1%)	314 (29.9%)	2.13 (1.63, 2.78)	< 0.001	1.90 (1.44, 2.53)	< 0.001
sFLR	271 (17.3%)	64 (24%)
< 30%	98 (6%)	26 (10%)	Reference		Reference
30 to < 40%	99 (6%)	19 (7%)	0.66 (0.34, 1.29)	0.222	0.74 (0.36, 1.51)	0.406
≥ 40%	74 (5%)	19 (7%)	0.96 (0.48, 1.90)	0.899	1.00 (0.47, 2.12)	1.000

*Values in parentheses are 95% confidence intervals. †Adjusted for sex, age at the time of surgery, preoperative estimated glomerular filtration rate, diabetes, cardiovascular disease, pulmonary disease, and body mass index. ORs, odds ratios; AKI, acute kidney injury; ALPPS, associated liver partition with portal vein ligation for staged hepatectomy; sFLR, standardized future liver remnant.

Postoperative AKI after major hepatectomy

When applying the KDIGO criteria for AKI using creatinine and urine volume measurements on the first two postoperative days, 400 of all patients who underwent major hepatectomy (25.6%) developed AKI of any stage. AKI stage 1 was diagnosed in 130 (8.3%) patients, 199 (12.7%) patients had AKI stage 2, and 71 (5%) patients had AKI stage 3. The majority of all patients who developed AKI after surgery (352, 88.0%) fulfilled the KDIGO criteria within the first postoperative day. Patients with higher age, male sex, higher body mass index, higher preoperative creatinine levels, cardiovascular disease, and diabetes were more likely to develop postoperative AKI (Table 1). Among patients who received neoadjuvant oncological treatment, which is only administered to patients with colorectal liver metastases, 20.7% (128) developed AKI. The surgical risk factors for postoperative AKI included undergoing an ALPPS step 2 procedure and intraoperative blood loss ≥ 1000 ml (adjusted OR 2.55 (95% c.i. 1.55 to 4.20) and 1.90 (95% c.i. 1.44 to 2.53), respectively; Table 2). There were no differences in sFLR and the odds of AKI (for example, for sFLR ≥ 40 versus < 30%: adjusted OR 1.00; 95% c.i. 0.47 to 2.12).

AKI and the risk of postoperative adverse outcomes

Patients with AKI after major hepatectomy also had an increased risk of other postoperative complications. Among the 1044 patients not admitted to the ICU routinely after surgery, patients with AKI had an increased odds of ICU admission within the first day after surgery (adjusted OR 6.06; 95% c.i. 2.83–12.95; Table 3). In addition, patients with AKI had an increased odds of readmission within 90 days (adjusted OR 1.67; 95% c.i. 1.11 to 2.52), a hospital stay ≥ 14 days (adjusted OR 2.47; 95% c.i. 1.91 to 3.18), and Clavien–Dindo grade ≥ IIIa complications (adjusted OR 2.08; 95% c.i. 1.61 to 2.67). Postoperative AKI was associated with postoperative haemorrhage grade ≥ B (adjusted OR 2.07; 95% c.i. 1.14 to 3.77). Of all included patients, 245 developed PHLF grade ≥ B, and 108 of these patients also developed postoperative AKI (Table 3). Patients with AKI also had an increased odds of PHLF grade B or C compared with patients with normal postoperative kidney function (adjusted OR 2.87; 95% c.i. 2.11 to 3.90).

Table 3.

Postoperative AKI and ORs for postoperative complications after major hepatectomy

Postoperative complication	No. of individuals (%)	No. of patients with AKI (%)	Postoperative AKI versus no AKI*
			Unadjusted OR	P	Adjusted OR†	P
ICU admission within 24 h‡	37 (4%)	23 (62%)	6.92 (3.50, 13.70)	< 0.001	6.06 (2.83, 12.95)	< 0.001
Readmission within 90 days	176 (19.1%)	46 (26%)	1.56 (1.06, 2.28)	0.024	1.67 (1.11, 2.52)	0.014
Hospital stay ≥ 14 days	503 (32.5%)	187 (37.2%)	2.40 (1.89, 3.04)	< 0.001	2.47 (1.91, 3.18)	< 0.001
Clavien–Dindo grade ≥ IIIa	519 (33.9%)	180 (34.7%)	2.01 (1.58, 2.54)	< 0.001	2.08 (1.61, 2.67)	< 0.001
PHLF grade B or C	245 (15.8%)	108 (44.1%)	2.77 (2.09, 3.68)	< 0.001	2.87 (2.11, 3.90)	< 0.001
Biliary leakage grade ≥ B	169 (11.3%)	53 (31%)	1.37 (0.97, 1.94)	0.078	1.37 (0.94, 1.99)	0.096
Postoperative haemorrhage grade ≥ B	56 (4%)	23 (41%)	2.08 (1.20, 3.58)	0.009	2.07 (1.14, 3.77)	0.017

*Values in parentheses are 95% confidence intervals. Unadjusted and adjusted ORs from logistic regression analysis are shown for postoperative complications after major hepatectomy comparing postoperative AKI according to Kidney Disease: Improving Global Outcomes to normal postoperative kidney function. †Adjusted for sex, age at the time of surgery, preoperative estimated glomerular filtration rate, diabetes, cardiovascular disease, pulmonary disease, and body mass index. ‡Among patients not routinely admitted to the ICU after surgery (n = 1044). AKI, acute kidney injury; ORs, odds ratios; ICU, intensive care unit; h, hours; PHLF, post-hepatectomy liver failure.

Postoperative mortality

In all, 89 patients died before the 90-day follow-up. Compared with patients with normal postoperative kidney function, patients with postoperative AKI had an increased risk of 90-day mortality (adjusted hazard ratio (HR) 3.70; 95% c.i. 2.33 to 5.88; Table 4 and Fig. S1). This risk remained increased for all AKI stages (stage 1: adjusted HR 4.08 (95% c.i. 2.26 to 7.37); stage 2: adjusted HR 2.76 (95% c.i. 1.50 to 5.07); stage 3: adjusted HR 5.53 (95% c.i. 2.81 to 10.87). After additionally adjusting for PHLF grade B or C, patients with postoperative AKI still had an increased risk of 90-day mortality (HR 2.37; 95% c.i. 1.49 to 3.79; Table 4). The HRs comparing patients with postoperative AKI, PHLF, or the combination of AKI and PHLF are presented in Table 5. Compared with patients without AKI or PHLF, patients with AKI alone had an increased risk of 90-day mortality (adjusted HR 2.28; 95% c.i. 1.13 to 4.60), but patients with PHLF alone had a higher risk of 90-day mortality (adjusted HR 6.33; 95% c.i. 3.27 to 12.26; Fig. S2). Patients with AKI and PHLF after surgery had an even higher risk of 90-day mortality (adjusted HR 16.64; 95% c.i. 9.33 to 29.67). The risk for adverse postoperative outcomes among patients who developed AKI, including PHLF and 90-day mortality, were similar after excluding people who underwent the ALPPS step 2 procedure (Table S3 and Fig. S3). Among patients who developed PHLF, a more severe AKI stage was associated with increased mortality (Table S4).

Table 4.

Cox’s proportional hazards regression analysis for 90-day mortality after major hepatectomy according to postoperative AKI and AKI stage

	Unadjusted HR	P	Adjusted HR*	P	Adjusted HR†	P
Postoperative AKI	3.42 (2.25, 5.19)	< 0.001	3.70 (2.33, 5.88)	< 0.001	2.37 (1.49, 3.79)	< 0.001
AKI stage
No AKI	Reference		Reference		Reference
Stage 1	4.09 (2.36, 7.10)	< 0.001	4.08 (2.26, 7.37)	< 0.001	2.81 (1.55, 5.11)	0.001
Stage 2	2.37 (1.33, 4.21)	< 0.001	2.76 (1.50, 5.07)	0.001	1.67 (0.91, 3.10)	0.101
Stage 3	5.21 (2.74, 9.98)	< 0.001	5.53 (2.81, 10.87)	< 0.001	3.49 (1.77, 6.90)	< 0.001

Values in parentheses are 95% confidence intervals. Unadjusted and adjusted HRs are shown for postoperative mortality according to AKI and AKI stage. *Adjusted for sex, age at the time of surgery, preoperative estimated glomerular filtration rate, diabetes, cardiovascular disease, pulmonary disease, and body mass index. †Additionally adjusted for post-hepatectomy liver failure grade B or C. AKI, acute kidney injury; HR, hazard ratio.

Table 5.

Cox’s proportional hazards regression analysis for 90-day mortality after major hepatectomy according to postoperative AKI and PHLF grade B or C

	Unadjusted HR	P	Adjusted HR*	P
No AKI or PHLF	Reference		Reference
AKI	2.27 (1.17, 4.41)	0.016	2.28 (1.13, 4.60)	0.022
PHLF	6.40 (3.49, 11.76)	< 0.001	6.33 (3.27, 12.26)	< 0.001
AKI and PHLF	15.07 (8.81, 25.77)	< 0.001	16.64 (9.33, 29.67)	< 0.001

Values in parentheses are 95% confidence intervals. Unadjusted and adjusted HRs are shown for postoperative mortality according to the presence of AKI. *Adjusted for sex, age at the time of surgery, preoperative estimated glomerular filtration rate, diabetes, cardiovascular disease, pulmonary disease, and body mass index. AKI, acute kidney injury; PHLF, post-hepatectomy liver failure; HR, hazard ratio.

Discussion

In this study of 1561 patients undergoing major hepatectomy at three tertiary referral centres, postoperative AKI of any grade, based on serum creatinine and urine output, was prevalent (26% of all operated patients). Pre- and intraoperative risk factors for AKI included male sex, higher age, lower preoperative eGFR, ALPPS step 2 procedure, and higher intraoperative blood loss. Patients with postoperative AKI were more likely to develop PHLF, whereas patients with both postoperative AKI alone and in combination with PHLF had a significantly higher 90-day mortality.

It has previously been shown that patients undergoing any type of liver resection are at risk of developing AKI after surgery, with an incidence ranging from 9 to 25% depending on the degree of liver resection^8,9,18–20. The reasons for a comparatively high risk of kidney injury after liver resection remain uncertain, although factors such as intraoperative hypotension, acute tubular necrosis, or haemodynamic changes in portal blood flow may increase the susceptibility of the kidneys during the immediate operative period^10,21. In addition, the definition of AKI in general has varied and only recently been harmonized following the introduction of the KDIGO criteria^5,22,23. In the present study on major hepatectomies, a direct relationship between sFLR and the risk of developing AKI was not found, suggesting that other factors contribute to postoperative AKI rather than a reduction in liver volume alone. It should be noted that only a minority of included patients were evaluated for sFLR. Current treatment practices at the three institutions in the present study include the measurement of central venous pressure, aimed at low pressure (< 5 mmHg) during surgery, which is thought to decrease perioperative blood loss^24,25 PHLF is a major cause of short-term mortality following liver surgery^26–28, with an incidence of around 10% after major hepatectomy²⁹. In the present study, postoperative AKI was an independent risk factor for PHLF and, in combination with AKI, significantly contributed to postoperative mortality. It has previously been shown that an increase in portal vein pressure to the remnant liver after hepatectomy can increase the risk of AKI³⁰. The mechanism is hypothesized to involve a hepatorenal reflex that causes kidney hypoperfusion^10,31,32. At the same time, liver failure in itself can cause AKI through the hepatorenal syndrome. Based on the current ISGLS definition of PHLF, which does not allow a diagnosis of PHLF earlier than 5 days after surgery, these results suggest that postoperative AKI within the first 2 days after hepatectomy may be an early sign of adverse postoperative events in general, and PHLF in particular. However, it is not possible by this study alone to clearly define the cause or the sequence of postoperative AKI and PHLF, and pathophysiological changes probably interact in the immediate postoperative period. AKI is likely to be caused by a number of interrelated factors associated with the patient, the procedure, or the postoperative course that are not directly due to the to the future liver remnant. However, the results of this study suggest that meeting the definition of AKI early after surgery is associated with a higher risk of developing PHLF, by the current definitons of AKI and PHLF. Therefore early AKI in the postoperative phase could be considered as a high-risk sign of an adverse outcome³³.

This study has limitations. It was impossible to distinguish between AKI due to the actual liver resection or postoperative kidney failure caused by emerging PHLF. A decrease in urine output after major surgery could also be affected by factors other than direct kidney injury. In addition, the results may have been influenced by the practices at the three institutions, and sFLR was only measured in a minority of patients included in the study, which could cause selection bias. Another limitation is that information on changes in eGFR after ALPPS step 1 was not available. The 2-day cut-off for AKI after surgery could miss some cases of AKI, although this cut-off was chosen to better reflect direct postoperative causes of AKI. The study did not include information on how AKI was treated during the postoperative course with regard to, for example, initiation of dialysis, fluid resuscitation, and the duration of abnormal creatinine levels, although this may not invalidate the mere presence of AKI being associated with PHLF.

This study of patients undergoing major hepatectomy found that the prevalence of postoperative AKI was unexpectedly high and that it was associated with other postoperative complications. Postoperative AKI within 2 days after surgery was as independent risk factor for PHLF grade B and C. Patients with AKI alone or in combination with PHLF had a significantly increased risk of 90-day mortality, and AKI was an independent risk factor for mortality also after adjusting for PHLF.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Author contributions

OS: Conceptualization, methodology, formal analysis, writing—original draft. TR: Methodology, investigation, writing—review & editing. KH: Methodology, investigation, writing—review & editing. JE: Methodology, investigation, writing—review & editing. AEK: Methodology, investigation, writing—review & editing. RB: Methodology, investigation, writing—review & editing. PG: Methodology, writing—review & editing. PS: Methodology, writing—review & editing. ES: Methodology, writing—review & editing. KJO: Concepualtization, methodology, writing—review & editing. BB: Concepualtization, methodology, writing—review & editing. SG: Concepualtization, methodology, writing—review & editing.

Disclosure

RB has received funding from the Cancer Research Funds of Radiumhemmet outside of the present study. The remaining authors declare no conflict of interest.

Supplementary material

Supplementary material is available at BJS Open online.

Data availability

Data sets generated during and/or analysed during the present study are not publicly available but are available from the corresponding author upon reasonable request.