https://orcid.org/0000-0002-0629-325X
Figures
Abstract
Background
Acute kidney injury (AKI) after lung transplantation (LuTx) is associated with increased long-term mortality. In this prospective observational study, commonly used AKI-definitions were examined regarding prediction of long-term mortality and compared to simple use of the serum creatinine value at day 7 for patients who did not receive hemodialysis, and serum creatinine value immediately before initiation of hemodialysis (d7/preHD-sCr).
Methods
185 patients with LuTx were prospectively enrolled from 2013–2014 at our center. Kidney injury was assessed within 7 days by: (1) the Kidney Disease Improving Global Outcomes criteria (KDIGO-AKI), (2) the Acute Disease Quality Initiative 16 Workgroup classification (ADQI-AKI) and (3) d7/preHD-sCr. Prediction of all-cause mortality was examined by Cox regression analysis, and clinical as well as laboratory factors for impaired kidney function post-LuTx were analyzed.
Results
AKI according to KDIGO and ADQI-AKI occurred in 115 patients (62.2%) within 7 days after LuTx. Persistent ADQI-AKI, KDIGO-AKI stage 3 and higher d7/preHD-sCr were associated with higher mortality in the univariable analysis. In the multivariable analysis, d7/preHD-sCr in combination with body weight and intra- and postoperative platelet transfusions predicted mortality after LuTx with similar performance as models using KDIGO-AKI and ADQI-AKI (concordance index of 0.75 for d7/preHD-sCr vs., 0.74 and 0.73, respectively). Pre-transplant reduced renal function, diabetes, higher BMI, and intraoperative ECMO predicted higher d7/preHD-sCr (r2 = 0.354, p < 0.001).
Citation: Doricic J, Greite R, Vijayan V, Immenschuh S, Leffler A, Ius F, et al. (2022) Kidney injury after lung transplantation: Long-term mortality predicted by post-operative day-7 serum creatinine and few clinical factors. PLoS ONE 17(3): e0265002. https://doi.org/10.1371/journal.pone.0265002
Editor: Justyna Gołębiewska, Medical University of Gdansk, POLAND
Received: September 20, 2021; Accepted: February 20, 2022; Published: March 4, 2022
Copyright: © 2022 Doricic et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The work was supported by a grant from the German Federal Ministry of Education and Research (BMBF-No: 03INT502AB) to HH and RG. Funders had no influence on the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: ADQI, Acute Disease Quality Initiative; AKI, acute kidney injury; CI-95%, 95% confidence interval; CKD, chronic kidney disease; CKD-EPI, Chronic kidney disease epidemiology collaboration; d7, day 7; d7/preHD-sCr, serum creatinine value at day 7 for patients who did not receive hemodialysis, and serum creatinine value immediately before initiation of hemodialysis; ECMO, extracorporeal membrane oxygenation; eGFR, estimated glomerular filtration rate; FEV1, forced expiratory volume in one second; FFP, fresh frozen plasma; FVC, forced vital capacity; HR, hazard ratio; ICU, intensive care unit; IQR, interquartile ranges; KDIGO, Kidney Disease Improving Global Outcomes; LAS, Lung Allocation Score; LuTx, lung transplantation; Max, maximum; Min, minimum; PGD, primary graft dysfunction; pRBC, packed red blood cells; RRT, renal replacement therapy; sCr, serum creatinine; SD, standard deviation; UO, urine output
Introduction
Acute kidney injury (AKI) is a common complication after lung transplantation (LuTx), with incidence rates of 39–69% [1–5] and is strongly associated with an increased risk for death [5–7]. However, reliable individual assessment of this risk is impeded by heterogenous and rather complex definitions of AKI. Applying common AKI definitions such as the Kidney Disease Improving Global Outcomes (KDIGO) definition [8] and the Acute Disease Quality Initiative 16 Workgroup definition (ADQI) [9] requires time and effort. Specifically, serum creatinine (sCr) has to be monitored prior to LuTx and on seven consecutive postoperative days to apply the KDIGO criteria. Based on the dynamic sCr changes, patients are categorized into three different AKI stages [8]. Moreover, the ADQI definition uses the KDIGO criteria but adds persistence of AKI as an attribute, which is defined as AKI presence for more than 48 hours [9]. An alternative AKI criterion proposed by KDIGO and ADQI is urine output (UO). UO needs to be measured for 24 hours and AKI can then be staged according to specific cut-offs for UO reduction [8]. However, due to unavailable data on UO, this alternative criterion has not been used in many studies [2, 5, 10–12].
Based on these considerations and on the fact that categorization of a continuous parameter like serum creatinine may decrease its informational value [13], this study examines whether a simple assessment of kidney function by the serum creatinine value at day 7 for patients who did not receive hemodialysis, and serum creatinine value immediately before initiation of hemodialysis (d7/preHD-sCr) predicts mortality equally reliable as the aforementioned measures of AKI. Furthermore, clinical and laboratory predictors of impaired post-transplant renal function are explored.
Methods
Study population
Patients were consecutively recruited for this study. In the period from June 2013 to December 2014, 193 adult patients received a LuTx, of which 185 patients consented to participate in the study (95.9%). Exclusion criteria were combined organ transplantations. The study was approved by the local ethics committee (no. 6895) and initiated in 2013 to examine the primary endpoint "AKI" in terms of clinical factors and biomarkers. The following results now refer to the secondary endpoint "long-term outcome". Clinical, laboratory and surgery-related factors were documented pre-, peri- and post-transplantation. Patients were followed up for 5 years post-transplant, with an annual routine consisting of clinical and laboratory parameters including renal function and immunosuppressive drug levels. The mean follow-up was 57 ± 20 months.
Minimally invasive sternum-sparing anterolateral thoracotomy performed in the transplant procedure whenever possible. The majority of patients received bilateral LuTx (n = 182, 98.4%); only 3 patients had single LuTx. Initially, all patients received a triple maintenance immunosuppression with tacrolimus, mycophenolate mofetil and steroids. Within the first year after transplantation, tacrolimus target levels were 8–12 μg/liter. Induction therapy was not used. In all patients, primary graft function (PGD) was scored at different postoperative time points after transplantation according to the International Society for Heart and Lung Transplantation guidelines and PGD grade 3 was used for analysis [14, 15].
Renal function and AKI definitions
Renal function was determined by calculating the estimated glomerular filtration rate (eGFR) with the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)–formula [16]. Baseline serum creatinine values were collected within 7 days before transplantation. No patient was on dialysis prior to LuTx. All patients who required renal replacement therapy (RRT) after LuTx underwent hemodialysis. Serum creatinine value at day 7 for patients who did not receive hemodialysis, and serum creatinine value immediately before initiation of hemodialysis was used for analysis (d7/preHD-sCr). The KDIGO criteria [8] were used to grade AKI as stage I (sCr increase by 1.5 to 1.9-fold from baseline or absolute sCr increase by ≥0.3 mg/dl or ≥26.5 μmol/l), stage II (sCr increase by 2.0 to 2.9-fold from baseline), or stage III (sCr increase by 3.0-fold from baseline or sCr increase by ≥4.0 mg/dl or ≥353.6 μmol/l or initiation of RRT within the first 7 days after LuTx). The ADQI definition [9] was used to differentiate between transient AKI (return of sCr below KDIGO AKI stage I within 48 hours of AKI onset) and persistent AKI (sustained sCr elevation with at least KDIGO stage I beyond 48 hours after AKI). Urine output, which is an alternative criterion in both definitions (KDIGO, ADQI), was excluded as an AKI criterion because the data was not available for all patients.
Statistical analysis
Statistical analyses were performed with IBM SPSS statistical software version 26.0 and the rms [17] package from R Studio software 3.6.0 (R Core Team, 2020). Continuous data with normal distribution are reported as mean value ± standard deviation. Data with non-normal distribution are reported with median and interquartiles. The Kolmogorov-Smirnov test was used to identify variables with and without normal distribution. Group comparisons were made with the student’s t-test and the Mann–Whitney U-test accordingly. To examine cut-points of continuous parameters in the context of survival analyses, we used R Studio package ’CutpointsOEHR’ [18]. Categorical variables are presented as numbers and percentages and comparisons were made using Fisher’s exact test and chi-square test. Kaplan-Meier analysis and log-rank test were used to describe patient survival for all causes of mortality.
For the Cox regression analyses, linearity of continuous variables was verified by categorizing the variables and comparing the ß-coefficients from the univariate Cox regression. Variables with a p-value <0.05 in the univariate analysis and further variables that were deemed biologically relevant were included in the multivariable modeling, which was performed by stepwise backward selection (p-value threshold <0.15). The bootstrapping procedures in R software were used to validate the models. The Harrell’s concordance index was used to report the performance of the models [19]. Kaplan-Meier curves were constructed from multivariate Cox model for 4 separate risk groups ("very low" to "high") using cut points on the prognostic index determined by the Cox method (cut points: 16th, 50th, and 84th percentiles of the prognostic index) [20, 21]. In the absence of any significant difference and with identical Kaplan-Meier curves, the two risk groups “very low” and “low” were summarized into one group "low" for better clarity. Multivariable linear regression analysis was applied to assess the effect of pre-and perioperative factors on d7/preHD-sCr after LuTx, using backward selection and a cut-off p value of <0.2. Unstandardized predicted serum creatinine values were calculated using the regression equation based on the unstandardized regression coefficients. Differences with p <0.05 were considered as significant.
Results
1. Study population
Patient characteristics, pre- and postoperative factors of the entire study population and the subgroups with and without AKI stage I-III according to KDIGO are shown in Table 1. Leading indications for LuTx were idiopathic pulmonary fibrosis (33.0%), chronic obstructive pulmonary disease (23.8%) and cystic fibrosis (20.0%). Mean age at transplantation was 48y (± 12yrs). Sex distribution between patients with and without AKI was comparable. Patients with AKI post LuTx were younger and more frequently had diabetes and better kidney function before LuTx. Surgery time and post-operative intensive care unit treatment was longer and more packed red blood cells (pRBC) were given during surgery to the patients, that later developed AKI. Primary graft dysfunction of grade 3, which has been shown predictive for survival after lung transplantation, was only numerically more prevalent in patients with AKI.
2. Incidence and characteristics of AKI
Of the 185 patients, 115 (62.2%) developed KDIGO-AKI, mostly stage 1 (40.0%) and stage 2 (44.3%). Sixty patients with AKI (52.2%) did not recover renal function within 48 hours thus presenting with persistent ADQI-AKI (Table 1). Fig 1 shows the serum creatinine levels in the different AKI stages. Median d7/preHD-sCr was comparable between no AKI and KDIGO-AKI stage 1 (Fig 1A) and between no AKI and transient ADQI-AKI (Fig 1B). Patients with transient AKI had KDIGO-AKI stage 1 in 69% and the remaining patients had KDIGO-AKI stage 2. Patients with persistent ADQI-AKI had mostly KDIGO-AKI stage 2 (56.7%) and stage 3 (30.0%), with 4 patients requiring dialysis (Fig 1C).
A-C Serum creatinine values at day 7 after LuTx in patients without AKI and with AKI. According to the KDIGO-AKI stages (A) and the ADQI-AKI grading (B) and distribution of the AKI stages (C). Boxes and whiskers represent medians, lower and upper quartiles, and the extreme values. *p < = 0.05; **p < = 0.01; ***p < = 0.001. (D) Long-term mortality after LuTx. Kaplan-Meier curves were stratified by stages of AKI according to the KDIGO classification. No AKI vs. KDIGO-AKI stage 1, log rank p = 0.311; No AKI vs. KDIGO-AKI stage 2, log rank p = 0.573; No AKI vs. KDIGO-AKI stage 3, log rank p = 0.002; KDIGO-AKI 1 vs. KDIGO-AKI stage 2, log rank p = 0.176; KDIGO-AKI stage 1 vs. KDIGO-AKI stage 3, log rank p = 0.043; KDIGO-AKI stage 2 vs. KDIGO-AKI stage 3, log rank p = 0.017.
3. AKI and patient long-term survival
In the post-transplant follow-up period of 57 ±20 months, one patient was lost in the -to-follow-up and 36 (19.5%) patients died. Leading causes of death were infections (n = 16; 44.4%), chronic lung allograft dysfunction (n = 12; 33.3%) such as bronchiolitis obliterans syndrome and restrictive allograft syndrome, and malignancies in 4 patients. One patient each died from sudden cardiac death, acute graft failure, hemoptysis and ileus. Five of the 16 infectious deaths occurred in the early post-transplant course within the first 90 days. All infectious deaths except one were in patients with AKI. Overall, in the patients without AKI 9 deaths occurred, whereas the remaining 27 deaths were observed in patients with AKI, with the worst survival in KDIGO-AKI stage 3 (Fig 1D). Notably, three of the four patients who required dialysis within the first 7 days after transplantation died.
In univariable Cox analyses (Table 2), KDIGO-AKI stage 3, persistent ADQI-AKI and higher d7/preHD-sCr were associated with reduced survival. Regarding the d7/preHD-sCr, there was a linear relation with mortality, without discernible cut-off value. Further factors associated with reduced survival were higher body weight and several variables relating to a more complex course of surgery and postoperative treatment such as operating time, re-operation, ECMO treatment, length of stay in the ICU and ward, substitution of blood components and primary graft dysfunction at 72 hours after transplantation. The four patients with dialysis requirement until day 7 had the longest operation times (median 374.5 min), and more transfusions of pRBC and FFP (median 12.5 and 9.5) intraoperatively.
Based on these univariable results (Table 2), multivariable models with the different AKI definitions and d7/preHD-sCr were created. These results are summarized in Table 3A–3C. The predictive performance was similar for the two AKI definitions and d7/preHD-sCr, with a concordance index of 0.75 for d7/preHD-sCr, 0.74 for KDIGO-AKI and 0.73 for ADQI-AKI and, after 200-fold bootstrapping, 0.74 for d7/preHD-sCr, and 0.70 for the other two models. The variables used in the three models were similar.
The calibration of the three models was comparable, with similar proportions of patients assigned to low, moderate and high mortality risk according to the model’s prognostic index (Fig 2). We also tested the d7/preHD-sCr model including dialysis as an additional variable and the results are shown in S1 Table. The performance of this model and the risk factors remained unchanged. An additional model using the eGFR calculated from d7/preHD-sCr showed a concordance index of 0.74/0.71 (without/with bootstrapping) and thus had no advantage. PGD grade 3 was not retained as a significant factor in the models.
A-C. Calibration of the multivariable Cox models with KDIGO-AKI criteria (A), ADQI-AKI grading (B) and d7/preHD-sCr (C). Kaplan-Meier curves are shown for three separate risk groups ("low" to "high") as described in Methods. The table below shows the events observed in the three risk groups.
In a post-hoc analysis of the long-term course after discharge from the hospital, renal function was examined in the patients assigned to low, moderate and high mortality risk. The four patients who required dialysis within the first 7 days were in the high-risk group. Fig 3 shows the last known sCr value during routine follow-up over 57 ±20 months in the three risk groups defined by the d7/preHD-sCr prediction model. sCr values at last follow-up were higher in all three groups, but highest in patients with moderate and high mortality risk.
For comparison, the d7/preHD-sCr value is also depicted for the three risk groups. For patients with dialysis, the serum creatinine level at initiation of renal replacement therapy was used. Boxes and whiskers represent medians, lower and upper quartiles, and the extreme values. *p < = 0.05; **p < = 0.01; ***p < = 0.001.
4. Prediction of d7/preHD-sCr
To estimate the relationship between d7/preHD-sCr and several pre- and perioperative factors, univariable and multivariable linear regression analyses were performed (Table 4). In the multivariable analysis, pretransplant renal function, diabetes, higher BMI and need of intraoperative ECMO were retained as significant predictors of d7/preHD-sCr. Fig 4 illustrates the relation between individually observed d7/preHD-sCr levels and the sCr values predicted by the multivariable model, explaining more than a third of variation in d7/preHD-sCr (r2 = 0.354).
Results were obtained by the multivariable regression model with pretransplant serum creatinine, diabetes, BMI and need of intraoperative ECMO as factors. Individual patient values and the regression line with the 95% confidence interval boundaries are shown.
Discussion
This study underlines the importance of kidney injury in patient outcomes after lung transplantation. Similar to previous studies we show that higher degree of AKI early after LuTx is associated with death [4, 5]. In addition, end stage renal failure and chronic renal function impairment in the long-term were more common in patients with severe AKI early after LuTx.
Unlike previous studies which used more complex definitions of AKI in the prediction of mortality risk, this study suggests that simple use of d7/preHD-sCr is equally reliable. Regular laboratory monitoring after major surgery is essential and use of this information should be as simple as possible. This includes close monitoring of sCr to detect renal dysfunction and initiate further diagnostics and therapy. Mortality risk prediction based on KDIGO and ADQI criteria is established, but any reliable simplification is an improvement. There is good rationale to use the serum creatinine. First, categorization of variables that are continuous is rarely sensible because information is lost, particularly with broad categories [13]. Second, calculation of AKI stages is more cumbersome and may be prone to error, requiring daily measurement and on-going comparison of the actual value with baseline serum creatinine before LuTx and urine output monitoring. Notably, although the established definitions suggest urine output as an alternative criterion for AKI, many studies did not consider urine output for AKI assessment, mainly for practical reasons of availability [2, 5, 10–12]. This suggests that parameters for assessment of AKI should be kept as simple as possible. A further, principal aspect is that the d7/preHD-sCr reflects pre-existing chronic renal function impairment and acute renal function decline, as patients may not fulfill the KDIGO or ADQI criteria but nevertheless have impaired renal function and thus an increased risk of mortality.
Fairly accurate estimation of the mortality risk was shown with all parameters of kidney injury, d7/preHD-sCr, the KDIGO-AKI and the ADQI grading, together with a few additional clinical factors. The multivariable models all had similar concordance indices and differentiation of mortality risks as illustrated by the calibration studies. The model with d7/preHD-sCr included body weight and need of intra-operative and post-operative platelet infusions. These risk factors were also included in the other two models. Using the eGFR at day 7 instead of d7/preHD-sCr was not advantageous in the model, probably because of the inaccuracy of eGFR calculations in critically ill patients as reported in liver transplant patients [23, 24].
The incidence of AKI is high early after LuTx, affecting one to two thirds of patients [1–5] and in our study 61% of patients. This broad range perhaps reflects different populations at risk, varying surgical and medical practice and use of different definitions of AKI. Likewise, the association of AKI with death was reported highly variable among different studies [1–6, 25, 26]. Severity of AKI was important for the mortality risk in our study, regardless of the AKI definition used. A strong link between severe AKI and death was also reported by others, particularly in patients with AKI who require temporary dialysis or who are on incident chronic dialysis treatment [6, 27–29]. This is in line with a previous study that identified severe AKI with renal replacement therapy after non-cardiac surgery as an independent risk factor for severe infections [30]. In our study, four patients required dialysis before day 7 and further six patients developed dialysis dependency within three months post-transplantation. Of these patients, all remained on dialysis and seven of them died. Infections were the leading cause of death in the early post-transplant period and even more so in the long-term and were almost exclusively prevalent in patients with AKI. In this regard, AKI has been described as an impaired immune state associated with dysfunctional monocyte cytokine production, predisposing to infections [31]. In the AKI group, sCr levels before transplantation were lower compared with the group without AKI. A similar observation was reported by others and appears to be related to the subgroup of patients with cystic fibrosis [1, 3, 11]. In our study, CF patients had a lower median pre-transplant sCr of 47.0 compared with patients without AKI. Hyperfiltration due to diabetes may be one explanation for the lower sCr in these patients, as 19 out of 37 had diabetes.
Apart from AKI, particularly factors that indicate a more complicated transplantation and post-transplantation course such as operation time, a greater need of substituting blood components, and post-surgical complications like re-operation showed associations with mortality. Platelet and plasma infusions were retained in the multivariable analyses, suggesting an independent effect on patient survival. Platelet transfusions were also a risk factor for mortality in patients with liver transplantation [32] and with lung transplantation [33].
More complex surgical courses and postoperative complications in turn can precipitate AKI, which would be in this case a surrogate of a more severe transplant course. On the other hand, an independent effect of AKI on long-term survival is indicated by various studies. AKI was linked to the development of chronic kidney disease in several study populations [34, 35] In turn, chronic kidney disease is an independent factor for mortality, besides classical risk factors like cardiovascular disease and diabetes [36]. Our post-hoc analysis emphasizes that patients after LuTx are at high risk of developing or aggravating clinically relevant chronic renal function impairment. Declining renal function was observed in all three risk groups in the long-term, most notably in the moderate- and high-risk group. This is probably due, in part, to the extensive exposure to calcineurin inhibitors [37–39].
Given the clinical relevance of AKI, identifying patients at risk is important. Pre-existing diabetes mellitus, need of intraoperative ECMO, higher BMI and impaired renal function prior to LuTx allowed prediction of higher d7/preHD-sCr. These factors were risk factors for AKI in previous studies [4, 6]. As there are no specific therapies to treat AKI, identification of patients at higher risk is crucial to modify treatment strategy, e.g. avoidance of nephrotoxic drugs and radiocontrast media, maintaining adequate cardiovascular function and treatment of relevant infections [40].
Our study has several limitations. First, this is a single-center study with a moderately-sized study population. Second, like in many other studies [2, 5, 10, 11], urine output was too difficult to obtain in a systematic and reliable fashion and was therefore, not used for the staging of AKI by the KDIGO and ADQI. Third, external validation of our models needs to be performed in the future. Lastly, additional biomarkers are not included in our current models and might help improve identification of patients at risk for AKI and increased mortality [41–43].
Conclusion
This study demonstrates the importance of reduced renal function and AKI for the long-term survival after LuTx. Findings should help identify patients at risk for AKI and to adjust pre- and post-operative management accordingly. Together with few clinical factors, the continuous parameter d7/preHD-sCr predicts mortality equally well as the commonly used AKI definitions by KDIGO and ADQI and is easier to apply.
Supporting information
S1 File. Anonymized data set with variable description.
https://doi.org/10.1371/journal.pone.0265002.s001
(XLSX)
S1 Table. Multivariable Cox models including postoperative dialysis.
https://doi.org/10.1371/journal.pone.0265002.s002
(PDF)
Acknowledgments
This work is dedicated to Prof. Dr. Faikah Güler, who initiated this study and unfortunately passed away far too early in her life.
We thank Dr. Hannah Lang very much for the linguistic revision.
References
- 1. Rocha PN, Rocha AT, Palmer SM, Davis RD, Smith SR. Acute Renal Failure after Lung Transplantation: Incidence, Predictors and Impact on Perioperative Morbidity and Mortality. American Journal of Transplantation. 2005;5(6):1469–76. pmid:15888056
- 2. Arnaoutakis GJ, George TJ, Robinson CW, Gibbs KW, Orens JB, Merlo CA, et al. Severe acute kidney injury according to the RIFLE (risk, injury, failure, loss, end stage) criteria affects mortality in lung transplantation. J Heart Lung Transplant. 2011;30(10):1161–8. pmid:21620737; PubMed Central PMCID: PMC3185168.
- 3. Jacques F, El-Hamamsy I, Fortier A, Maltais S, Perrault LP, Liberman M, et al. Acute renal failure following lung transplantation: risk factors, mortality, and long-term consequences. Eur J Cardiothorac Surg. 2012;41(1):193–9. pmid:21665487; PubMed Central PMCID: PMC3241081.
- 4. Wehbe E, Brock R, Budev M, Xu M, Demirjian S, Schreiber MJ Jr., et al. Short-term and long-term outcomes of acute kidney injury after lung transplantation. J Heart Lung Transplant. 2012;31(3):244–51. pmid:21996350.
- 5. Fidalgo P, Ahmed M, Meyer SR, Lien D, Weinkauf J, Cardoso FS, et al. Incidence and outcomes of acute kidney injury following orthotopic lung transplantation: a population-based cohort study. Nephrol Dial Transplant. 2014;29(9):1702–9. pmid:24957809.
- 6. George TJ, Arnaoutakis GJ, Beaty CA, Pipeling MR, Merlo CA, Conte JV, et al. Acute kidney injury increases mortality after lung transplantation. Ann Thorac Surg. 2012;94(1):185–92. pmid:22325467; PubMed Central PMCID: PMC3601658.
- 7. Wehbe E, Duncan AE, Dar G, Budev M, Stephany B. Recovery from AKI and short- and long-term outcomes after lung transplantation. Clin J Am Soc Nephrol. 2013;8(1):19–25. pmid:23037982; PubMed Central PMCID: PMC3531657.
- 8. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl. 2012.
- 9. Chawla LS, Bellomo R, Bihorac A, Goldstein SL, Siew ED, Bagshaw SM, et al. Acute kidney disease and renal recovery: consensus report of the Acute Disease Quality Initiative (ADQI) 16 Workgroup. Nature reviews Nephrology. 2017;13(4):241–57. Epub 2017/02/28. pmid:28239173.
- 10. Fidalgo P, Ahmed M, Meyer SR, Lien D, Weinkauf J, Kapasi A, et al. Association between transient acute kidney injury and morbidity and mortality after lung transplantation: a retrospective cohort study. J Crit Care. 2014;29(6):1028–34. pmid:25138690.
- 11. Ishikawa S, Griesdale DE, Lohser J. Acute kidney injury within 72 hours after lung transplantation: incidence and perioperative risk factors. J Cardiothorac Vasc Anesth. 2014;28(4):931–5. pmid:24360152.
- 12. Sikma MA, Hunault CC, van de Graaf EA, Verhaar MC, Kesecioglu J, de Lange DW, et al. High tacrolimus blood concentrations early after lung transplantation and the risk of kidney injury. Eur J Clin Pharmacol. 2017;73(5):573–80. Epub 2017/01/31. pmid:28132082; PubMed Central PMCID: PMC5384949.
- 13. Royston P, Altman DG, Sauerbrei W. Dichotomizing continuous predictors in multiple regression: a bad idea. Stat Med. 2006;25(1):127–41. Epub 2005/10/12. pmid:16217841.
- 14. Christie JD, Carby M, Bag R, Corris P, Hertz M, Weill D, et al. Report of the ISHLT Working Group on Primary Lung Graft Dysfunction part II: definition. A consensus statement of the International Society for Heart and Lung Transplantation. J Heart Lung Transplant. 2005;24(10):1454–9. Epub 2005/10/08. pmid:16210116.
- 15. Christie JD, Bellamy S, Ware LB, Lederer D, Hadjiliadis D, Lee J, et al. Construct validity of the definition of primary graft dysfunction after lung transplantation. J Heart Lung Transplant. 2010;29(11):1231–9. Epub 2010/07/27. pmid:20655249; PubMed Central PMCID: PMC2963709.
- 16. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604–12. Epub 2009/05/06. pmid:19414839; PubMed Central PMCID: PMC2763564.
- 17. Harrell Frank E. Jr. (2020). rms: Regression Modeling Strategies. R package version 6.0–1. https://CRAN.R-project.org/package=rms.
- 18. Chen Y, Huang J, He X, Gao Y, Mahara G, Lin Z, et al. A novel approach to determine two optimal cut-points of a continuous predictor with a U-shaped relationship to hazard ratio in survival data: simulation and application. BMC Med Res Methodol. 2019;19(1):96. Epub 2019/05/11. pmid:31072334; PubMed Central PMCID: PMC6507062.
- 19. Bansal A, Heagerty PJ. A Tutorial on Evaluating the Time-Varying Discrimination Accuracy of Survival Models Used in Dynamic Decision Making. Med Decis Making. 2018;38(8):904–16. Epub 2018/10/16. pmid:30319014; PubMed Central PMCID: PMC6584037.
- 20. Cox DR. Note on Grouping. Journal of the American Statistical Association. 1957;52(280):543–7.
- 21. Scheffner I, Gietzelt M, Abeling T, Marschollek M, Gwinner W. Patient Survival After Kidney Transplantation: Important Role of Graft-sustaining Factors as Determined by Predictive Modeling Using Random Survival Forest Analysis. Transplantation. 2020;104(5):1095–107. Epub 2019/08/14. pmid:31403555.
- 22. Egan TM, Murray S, Bustami RT, Shearon TH, McCullough KP, Edwards LB, et al. Development of the new lung allocation system in the United States. Am J Transplant. 2006;6(5 Pt 2):1212–27. Epub 2006/04/15. pmid:16613597.
- 23. Francoz C, Prie D, Abdelrazek W, Moreau R, Mandot A, Belghiti J, et al. Inaccuracies of creatinine and creatinine-based equations in candidates for liver transplantation with low creatinine: impact on the model for end-stage liver disease score. Liver Transpl. 2010;16(10):1169–77. Epub 2010/09/30. pmid:20879015.
- 24. Kalafateli M, Wickham F, Burniston M, Cholongitas E, Theocharidou E, Garcovich M, et al. Development and validation of a mathematical equation to estimate glomerular filtration rate in cirrhosis: The royal free hospital cirrhosis glomerular filtration rate. Hepatology. 2017;65(2):582–91. Epub 2016/10/26. pmid:27779785.
- 25. Bennett D, Fossi A, Marchetti L, Lanzarone N, Sisi S, Refini RM, et al. Postoperative acute kidney injury in lung transplant recipients. Interact Cardiovasc Thorac Surg. 2019. pmid:30649317.
- 26. Hennessy SA, Gillen JR, Hranjec T, Kozower BD, Jones DR, Kron IL, et al. Influence of hemodialysis on clinical outcomes after lung transplantation. J Surg Res. 2013;183(2):916–21. Epub 2013/03/14. pmid:23481566; PubMed Central PMCID: PMC4217044.
- 27. Brown JR, Kramer RS, Coca SG, Parikh CR. Duration of acute kidney injury impacts long-term survival after cardiac surgery. Ann Thorac Surg. 2010;90(4):1142–8. Epub 2010/09/28. pmid:20868804; PubMed Central PMCID: PMC3819730.
- 28. Coca SG, King JT Jr., Rosenthal RA, Perkal MF, Parikh CR. The duration of postoperative acute kidney injury is an additional parameter predicting long-term survival in diabetic veterans. Kidney Int. 2010;78(9):926–33. Epub 2010/08/06. pmid:20686452; PubMed Central PMCID: PMC3082138.
- 29. Duff S, Murray PT. Defining Early Recovery of Acute Kidney Injury. Clin J Am Soc Nephrol. 2020;15(9):1358–60. Epub 2020/04/03. pmid:32238366; PubMed Central PMCID: PMC7480548.
- 30. Tagawa M, Nishimoto M, Kokubu M, Matsui M, Eriguchi M, Samejima KI, et al. Acute kidney injury as an independent predictor of infection and malignancy: the NARA-AKI cohort study. J Nephrol. 2019;32(6):967–75. Epub 2019/10/17. pmid:31617159.
- 31. Himmelfarb J, Le P, Klenzak J, Freedman S, McMenamin ME, Ikizler TA, et al. Impaired monocyte cytokine production in critically ill patients with acute renal failure. Kidney Int. 2004;66(6):2354–60. Epub 2004/12/01. pmid:15569326.
- 32. Pereboom IT, de Boer MT, Haagsma EB, Hendriks HG, Lisman T, Porte RJ. Platelet transfusion during liver transplantation is associated with increased postoperative mortality due to acute lung injury. Anesth Analg. 2009;108(4):1083–91. Epub 2009/03/21. pmid:19299765.
- 33. Ong LP, Thompson E, Sachdeva A, Ramesh BC, Muse H, Wallace K, et al. Allogeneic blood transfusion in bilateral lung transplantation: impact on early function and mortality. Eur J Cardiothorac Surg. 2016;49(2):668–74; discussion 74. Epub 2015/04/29. pmid:25913825.
- 34. Chawla LS, Amdur RL, Amodeo S, Kimmel PL, Palant CE. The severity of acute kidney injury predicts progression to chronic kidney disease. Kidney Int. 2011;79(12):1361–9. pmid:21430640; PubMed Central PMCID: PMC3257034.
- 35. Chawla LS, Kimmel PL. Acute kidney injury and chronic kidney disease: an integrated clinical syndrome. Kidney Int. 2012;82(5):516–24. pmid:22673882.
- 36. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med. 2004;351(13):1296–305. Epub 2004/09/24. pmid:15385656.
- 37. Barraclough K, Menahem SA, Bailey M, Thomson NM. Predictors of decline in renal function after lung transplantation. J Heart Lung Transplant. 2006;25(12):1431–5. Epub 2006/12/21. pmid:17178337.
- 38. Bloom RD, Reese PP. Chronic kidney disease after nonrenal solid-organ transplantation. J Am Soc Nephrol. 2007;18(12):3031–41. Epub 2007/11/28. pmid:18039925.
- 39. Esposito C, De Mauri A, Vitulo P, Oggionni T, Cornacchia F, Valentino R, et al. Risk factors for chronic renal dysfunction in lung transplant recipients. Transplantation. 2007;84(12):1701–3. Epub 2008/01/01. pmid:18165784.
- 40. Ronco C, Bellomo R, Kellum JA. Acute kidney injury. Lancet. 2019;394(10212):1949–64. Epub 2019/11/30. pmid:31777389.
- 41. Ledeganck KJ, Gielis EM, Abramowicz D, Stenvinkel P, Shiels PG, Van Craenenbroeck AH. MicroRNAs in AKI and Kidney Transplantation. Clin J Am Soc Nephrol. 2019;14(3):454–68. Epub 2019/01/04. pmid:30602462; PubMed Central PMCID: PMC6419285.
- 42. Albert C, Zapf A, Haase M, Rover C, Pickering JW, Albert A, et al. Neutrophil Gelatinase-Associated Lipocalin Measured on Clinical Laboratory Platforms for the Prediction of Acute Kidney Injury and the Associated Need for Dialysis Therapy: A Systematic Review and Meta-analysis. Am J Kidney Dis. 2020;76(6):826–41 e1. Epub 2020/07/18. pmid:32679151.
- 43. Kane-Gill SL, Peerapornratana S, Wong A, Murugan R, Groetzinger LM, Kim C, et al. Use of tissue inhibitor of metalloproteinase 2 and insulin-like growth factor binding protein 7 [TIMP2]*[IGFBP7] as an AKI risk screening tool to manage patients in the real-world setting. J Crit Care. 2020;57:97–101. Epub 2020/02/23. pmid:32086072.