The impact of pretransplantation serum albumin levels on long-term renal graft outcomes. Transplant Proc. 2013;45:1379-1382. [PMID: 23726577 DOI: 10.1016/j.transproceed.2012.10.063]

Predicting kidney allograft survival with explainable machine learning

INTRODUCTION

Despite significant progress over the last decades in the survival of kidney allografts, several risk factors remain contributing to worsening kidney function or even loss of transplants. We aimed to evaluate a new machine learning method to identify these variables which may predict the early graft loss in kidney transplant patients and to assess their usefulness for improving clinical decisions.

MATERIAL AND METHODS

A retrospective cohort study was carried out with 627 kidney transplant patients followed at least three months. All these data were pre-processed, and their selected features were used to develop an automatically working a machine learning algorithm; this algorithm was then applied for training and parameterization of the model; and finally, the tested model was then used for the analysis of patients' features that were the most impactful for the prediction of clinical outcomes. Our models were evaluated using the Area Under the Curve (AUC), and the SHapley Additive exPlanations (SHAP) algorithm was used to interpret its predictions.

RESULTS

The final selected model achieved a precision of 0.81, a sensitivity of 0.61, a specificity of 0.89, and an AUC value of 0.84. In our model, serum creatinine levels of kidney transplant patients, evaluated at the hospital discharge, proved to be the most important factor in the decision-making for the allograft loss. Patients with a weight equivalent to a BMI closer to the normal range prior to a kidney transplant are less likely to experience graft loss compared to patients with a BMI below the normal range. The age of patients at transplantation and Polyomavirus (BKPyV) infection had significant impact on clinical outcomes in our model.

CONCLUSIONS

Our algorithm suggests that the main characteristics that impacted early allograft loss were serum creatinine levels at the hospital discharge, as well as the pre-transplant values such as body weight, age of patients, and their BKPyV infection. We propose that machine learning tools can be developed to effectively assist medical decision-making in kidney transplantation.

Impact of pretransplantation malnutrition risk on the clinical outcome and graft survival of kidney transplant patients

BACKGROUND

The prevalence of malnourished patients before transplantation and the influence of malnutrition on graft and patient outcomes remain underestimated, despite being associated with higher postoperative morbidity and mortality. This study aimed to develop an easy nutritional screening tool and evaluate the impact of nutritional status on clinical outcome, graft survival (GS) and mortality risk in kidney transplant patients (KTP).

METHODS

In this retrospective cohort study including 451 KTP, we developed a score by using anthropometric, clinical, and laboratory measures performed in the pretransplant evaluation. The patients were stratified into 3 groups according to the final score: G1 (0 or 1 point)=low risk, G2 (2 to 4 points)=moderate risk, and G3 (>5 points)=high risk of malnutrition. The patients were monitored after transplantation at least 1 to 10 years.

RESULTS

Stratifying the 451 patients based on the pretransplant risk score, G1, G2, and G3 were composed of 90, 292, and 69 patients, respectively. Patients from G1 maintained the lowest serum creatinine levels at hospital discharge when compared with others (p = 0.012). The incidence of infection in the patients from G3 was higher than patients from G1 and G2 (p = 0.030). G3 recipients showed worse GS than G1 patients (p = 0.044). G3 patients showed almost threefold higher risk for graft loss (HR 2.94, 95% CI 1.084-7.996).

CONCLUSIONS

KTP with higher malnutrition risk score were associated with worse outcomes and GS. The nutritional screening tool is easy to be used in clinical practice to evaluate the patient in preparation for kidney transplant.

Pre-transplant hypoalbuminemia is not associated with worse short-term outcomes among kidney transplant recipients

INTRODUCTION

Serum albumin is an indicator of overall health status, but it remains unclear how pre-transplant hypoalbuminemia is associated with early post-transplant outcomes.

METHODS

This study included all adult kidney transplant recipients (KTRs) at our center from 01/01/2001-12/31/2017 with serum albumin measured within 30 days before transplantation. KTRs were grouped based on pretransplant albumin level normal (≥4.0 g/dL), mild (≥3.5 - < 4.0g/dL), moderate (≥3.0 - < 3.5g/dL), or severe hypoalbuminemia (<3.0g/dL). Outcomes of interest included: length of hospital stay (LOS), readmission within 30 days, delayed graft function(DGF), and re-operation related to post-transplant surgical complications. We also analyzed rejection, graft failure, and death within 6 months post-transplant.

RESULTS

A total of 2807 KTRs were included 43.6% had normal serum albumin, 35.3% mild, 16.6% moderate, and 4.5% severe hypoalbuminemia. Mild and moderate hypoalbuminemia were associated with a shorter LOS by 1.22 (p < 0.001) and 0.80 days (p = 0.01), respectively, compared to normal albumin. Moderate (HR: 0.58; 95% CI: 0.37-0.91; p = 0.02) and severe hypoalbuminemia (HR: 0.21; 95% CI: 0.07-0.68; p = 0.01) were associated with significantly lower rates of acute rejection within 6 months post-transplant.

CONCLUSION

Patients with pre-transplant hypoalbuminemia have post-transplant outcomes similar to those with normal serum albumin, but with a lower risk of acute rejection based on the degree of hypoalbuminemia.

Serum Albumin Level Before Kidney Transplant Predicts Post-transplant BK and Possibly Cytomegalovirus Infection

Introduction

Opportunistic viral infections cause extensive morbidity and mortality in kidney transplant recipients (KTRs). Low serum albumin levels before and after transplant have been associated with negative outcomes. However, it is uncertain whether serum albumin levels before transplantation are associated with the risk for post-transplantation opportunistic BK polyomavirus (BKV) or cytomegalovirus (CMV).

Methods

We reviewed all KTRs transplanted at our institution between 1 January 2005 and 31 December 2015 with serum albumin measured within 45 days before transplantation in a retrospective observational cohort study. Selected patients were stratified into 3 groups: normal albuminemia (≥3.5 g/dl), moderate hypoalbuminemia (3.49–2.5 g/dl), and severe hypoalbuminemia (<2.5 g/dl). Patients were observed for post-transplantation BKV or CMV according to standard of care.

Results

We included 1717 patients in this study; 72.3% had normal serum albumin, 26.3% had moderate hypoalbuminemia, and 1.5% had severe hypoalbuminemia. Moderate and severe hypoalbuminemia incurred a higher risk for BKV compared with normal serum albumin levels in univariable analysis (moderate hypoalbuminemia: hazard ratio [HR] = 1.5; 95% confidence interval [CI], 1.14–1.90; P = .003); severe hypoalbuminemia: HR = 2.15; 95% CI, 1.01–4.56; P = 0.05). Although not significant after multivariable adjustment, there was still 18% increased risk in moderate hypoalbuminemia and 64% in severe hypoalbuminemia for BKV compared with the normal albumin group. Moderate hypoalbuminemia was associated with a higher risk for CMV infection than normal serum albumin levels in multivariable analysis, although it was not statistically significant (HR = 1.15; 95% CI, 0.36–3.64; P = 0.81).

Conclusions

These findings suggest that pretransplantation hypoalbuminemia is associated with a higher risk for post-transplantation BKV and possibly CMV. More intense screening is warranted for these viruses in recipients with pretransplant hypoalbuminemia.

Bioinformatics analyses on the immune status of renal transplant patients, a systemic research of renal transplantation

BACKGROUND

Kidney transplantation is the most effective treatment for end-stage renal disease. Allograft rejections severely affect survivals of allograft kidneys and recipients.

METHODS

Using bioinformatics approaches, the present study was designed to investigate immune status in renal transplant recipients. Fifteen datasets from Gene Expression Omnibus (GEO) were collected and analysed. Analysis of gene enrichment and protein-protein interactions were also used.

RESULTS

There were 40 differentially expressed genes (DEGs) identified in chronic rejection group when compared with stable recipients, which were enriched in allograft rejection module. There were 135 DEGs identified in acute rejection patients, compared with stable recipients, in which most genes were enriched in allograft rejection and immune deficiency. There were 288 DEGs identified in stable recipients when compared to healthy subjects. Most genes were related to chemokine signalling pathway. In integrated comparisons, expressions of MHC molecules and immunoglobulins were increased in both acute and chronic rejection; expressions of LILRB and MAP 4 K1 were increased in acute rejection patients, but not in stable recipients. There were no overlapping DEGs in blood samples of transplant recipients.

CONCLUSION

By performing bioinformatics analysis on the immune status of kidney transplant patients, the present study reports several DEGs in the renal biopsy of transplant recipients, which are requested to be validated in clinical practice.

Hypoalbuminaemia at time of surgery is associated with an increased risk for overall graft loss after kidney transplantation

AIM

The aim of this retrospective cohort study was to investigate whether pre-operative hypoalbuminaemia (<35 g/L) is associated with adverse outcomes post-kidney transplantation.

METHODS

Our retrospective, single-centre analysis included all patients who received their kidney transplant between 2007 and 2017, with documented admission albumin levels prior to surgery. Survival analyses were undertaken to explore the relationship of pre-transplant hypoalbuminaemia versus other baseline variables upon post-transplant outcomes.

RESULTS

We analysed 1131 kidney allograft recipients transplanted at our centre (2007-2017), with median follow-up 746 days (interquartile range 133-1750 days). Kidney transplant recipients with pre-operative hypoalbuminaemia were more likely older, female, recipients of deceased-donor kidneys and to have longer cold ischaemic times. Recipients with pre-operative hypoalbuminaemia had longer hospital admissions post-operatively but no difference in delayed graft function rates. There was no difference in 1 year creatinine but recipients with hypoalbuminaemia had reduced risk for cellular rejection. We observed significantly worse patient survival (83.2% vs 90.7%, P < 0.001) and overall graft survival (72.5% vs 82.0%, P < 0.001) for recipients with hypoalbuminaemia vs normal albumin levels, respectively, but no difference in death-censored graft survival. In a Cox regression model, adjusted for baseline pre-operative variables, hypoalbuminaemia was independently associated with an increased risk for overall graft loss after kidney transplantation (hazard ratio 1.468, 95% confidence interval 1.087-1.982, P = 0.012).

CONCLUSION

Pre-operative hypoalbuminaemia is an independent risk factor for overall graft loss after kidney transplantation. Further work is warranted to investigate the underlying pathophysiology to determine what supportive measures can be undertaken to attenuate adverse post-transplant outcomes.

Causal relationship between hypoalbuminemia and acute kidney injury

Our meta-analysis published in 2010 provided evidence that low levels of serum albumin (hypoalbuminemia) are a significant independent predictor of acute kidney injury (AKI) and death following AKI. Since then, a large volume of additional data from observational clinical studies has been published further evaluating the relationship between serum albumin and AKI occurrence. This is an updated review of the literature to re-evaluate the hypothesis that hypoalbuminemia is independently associated with increased AKI risk. Eligible studies published from September 2009 to December 2016 were sought in PubMed (MEDLINE) and forty-three were retained, the great majority being retrospective observational cohort studies. These included a total of about 68000 subjects across a diverse range of settings, predominantly cardiac surgery and acute coronary interventions, infectious diseases, transplant surgery, and cancer. Appraisal of this latest data set served to conclusively corroborate and confirm our earlier hypothesis that lower serum albumin is an independent predictor both of AKI and death after AKI, across a range of clinical scenarios. The body of evidence indicates that hypoalbuminemia may causally contribute to development of AKI. Furthermore, administration of human albumin solution has the potential to prevent AKI; a randomized, controlled study provides evidence that correcting hypoalbuminemia may be renal-protective. Therefore, measurement of serum albumin to diagnose hypoalbuminemia may help identify high-risk patients who may benefit from treatment with exogenous human albumin. Multi-center, prospective, randomized, interventional studies are warranted, along with basic research to define the mechanisms through which albumin affords nephroprotection.

State of Harmonization of 24 Serum Albumin Measurement Procedures and Implications for Medical Decisions

BACKGROUND

Measurements of serum and plasma albumin are widely used in medicine, including as indicators of quality of patient care in renal dialysis centers.

METHODS

Pools were prepared from residual patient serum (n = 50) and heparin plasma (n = 48) from patients without renal disease, and serum from patients with kidney failure before hemodialysis (n = 53). Albumin was measured in all samples and in ERM-DA470k/IFCC reference material (RM) by 3 immunochemical, 9 bromcresol green (BCG), and 12 bromcresol purple (BCP) methods.

RESULTS

Two of 3 immunochemical procedures, 5 of 9 BCG, and 10 of 12 BCP methods recovered the RM value within its uncertainty. One immunochemical and 3 BCG methods were biased vs the RM value. Random error components were small for all measurement procedures. The Tina-quant immunochemical method was chosen as the reference measurement procedure based on recovery and results of error analyses. Mean biases for BCG vs Tina-quant were 1.5% to 13.9% and were larger at lower albumin concentrations. BCP methods' mean biases were −5.4% to 1.2% irrespective of albumin concentration. Biases for plasma samples were generally higher than for serum samples for all method types. For most measurement procedures, biases were lower for serum from patients on hemodialysis vs patients without kidney disease.

CONCLUSIONS

Significant differences among immunochemical, BCG, and BCP methods compromise interpretation of serum albumin results. Guidelines and calculations for clinical management of kidney and other diseases must consider the method used for albumin measurement until harmonization can be achieved.

Serum albumin as a significant prognostic factor in patients with malignant pleural mesothelioma

Our aim was to evaluate the prognostic role of the pretreatment serum albumin level in patients with malignant pleural mesothelioma (MPM) receiving platinum-based systemic chemotherapy. From 1995 to 2013, a total of 97 patients receiving platinum-based systemic chemotherapy for newly diagnosed MPM were enrolled. All clinical information and laboratory results were retrospectively collected from the medical records. The Kaplan-Meier method was used to calculate survival. The Cox proportional hazards model was used to identify significant independent prognostic factors for predicting survival. In total, 34 of the 97 patients (35.1 %) had hypoalbuminaemia (albumin ≤ 35 g/l). The 1-year overall survival rate was 44.1 % for patients with hypoalbuminaemia and 72.0 % for patients with a normal albumin level. Multivariate analysis indicated that pretreatment albumin was an independent prognostic factor in MPM. Patients with hypoalbuminaemia had a greater risk of death than those with a normal albumin level [hazard ratio (HR) 1.778; 95 % confidence interval (CI) 1.504-2.998; P = 0.031]. When albumin was entered as a continuous variable in the Cox regression model, the HR of death was significantly decreased by 9.8 % (95 % CI 0.851-0.956) for each 1-g/l increment. The pretreatment serum albumin level is a simple, inexpensive and easily measurable marker with prognostic significance in MPM patients treated with platinum-based systemic chemotherapy.