Rapid Decline in 51Cr-EDTA Measured Renal Function During the First Weeks Following Lung Transplantation
Abstract
We previously described a 54% decline in renal function at 6 months after lung transplantation (LTx). We hypothesized that this decline is a very early event following LTx.
Thirty-one consecutive patients (16 females/15 males), mean age 49 (±13) years, with emphysema, cystic fibrosis/bronchiectasis or idiopathic pulmonary fibrosis were included in an analysis of renal function before and after LTx. The glomerular filtration rate (GFR) was measured using the 51Cr-ethylenediaminetetra acetic acid plasma clearance single injection technique (mGFR) at baseline before transplantation and at 1, 2, 3 and 12 weeks postoperatively.
Mean mGFR declined from 103 ± 18 to 65 ± 22, 53 ± 16 and 57 ± 18 mL/min/1.73m2 at 1-, 3- and 12-weeks post-LTx (p < 0.0001), respectively. In a time-dependent repeated measures ANOVA, risk factors for a decline in mGFR posttransplant included: time (p < 0.0001), acute renal failure within 2 weeks post-LTx (p = 0.0003), use of heart and lung machine (p = 0.04), and the use of ephedrine (p = 0.048), as well as increasing age, older than 18 years at LTx (p = 0.006). These data demonstrate that renal function, measured with an isotope method, decreases dramatically during the first week after LTx.
Introduction
Kidney function, as monitored by plasma creatinine (p-creatinine) and estimated clearance equations, has been found to decline after lung transplantation (LTx) (1). Using direct measured clearance methods, we previously found that there is an early decline in measured glomerular filtration rate (mGFR), and this decline is greater than that estimated by indirect methods such as the Cockcroft Gault method and by p-creatinine (2). Our group and others found an overall cumulative incidence of renal disease (K/DOQI Stage 5) of 22% among 390 adult lung transplant patients at 10 years in our National Center of Lung Transplantation, and a 54% and 62% decline in mGFR at 6 and 24 months after transplantation, respectively (2,3).
It is unknown how fast the decline in the early postoperative period is, how primary graft dysfunction (PGD) and perioperative factors (such as the use of vasoactive substances, cardiopulmonal bypass, hypotension and blood loss) influence renal function, and how the high level of blood cyclosporine concentration needed with these patients (4) affects immediate postoperative renal function.
We hypothesize that renal function declines immediately after LTx. To demonstrate this, we measured renal function, using a plasma clearance method, before LTx and in the first weeks following LTx.
Materials and Methods
Patients
We performed an observational prospective study of immediate posttransplant renal function. All the patients on the waiting list for LTx in the Danish National Lung Transplant Programme were asked to participate and all these patients gave informed consent. From December 2006 to December 2007, all consecutive (35) single- (SLTx) and double-lung transplant (DLTx) recipients were included in the study.
Three patients withdrew informed consent after transplantation and one had a prolonged stay in the intensive care unit (ICU) with no 51Cr-EDTA clearance measurements taken; these four patients were excluded from the study, leaving 31 patients who were included in the analyses. The transplantation procedure and the selection of recipients and donors have been described previously in detail (5).
All patients had a normal urine analysis assessed by dipstick measurement and urine cultures and none of the patients had any signs of overt renal disease (assessed by abdominal ultrasound) prior to transplantation. All patients had glomerular filtration rates (GFR) within the normal range before transplantation.
The GFR was determined using the 51Cr-ethylenediaminetetra acetic acid (EDTA) plasma clearance single-injection technique (6) at baseline before transplantation, and then at 1-, 2-, 3, and 12 weeks postoperatively.
All routine p-creatinine samples collected during the study period were included in the analysis.
Primary graft dysfunction (PGD) was assessed according to ISHLT standards (7) and radiological analyses were performed by a specialist in respiratory medicine, who was blind to clinical information about the patients.
GFR measurements collected at our institution using the single-injection 51Cr-EDTA method have a variation coefficient of 4.5% and p-creatinine measurements have a variation coefficient of 4%.
Trough levels of cyclosporine were determined using Cyclosporine Monoclonal Whole Blood assay (TDx-FLx, Abbott) and AUC-cyclosporine/week (AUC-CyA) 1, 2, 3 and 12 weeks postoperatively were estimated using the trapezoid rule.
The local scientific ethics committee for Copenhagen and Frederiksberg, Denmark, approved the study (#KF 01279825). Collection and data registration were also licensed by the Danish Data Protection Agency (#2006–41-5640).
Immunosuppressive treatment and control
All patients received the standard immunosuppressive treatment used at our institution, consisting of preoperative induction with 500 mg of methylprednisolone and treatment with antithymocyte-globulin (ATG) at a dose of 1.5 mg/kg for the first 3 days. A maintenance immunosuppressive treatment regimen, consisting of cyclosporine, azathioprine and prednisolone, was started on the first postoperative day. Trough levels of cyclosporine were intended to be around 200 μg/L during the whole study period. The standard dose of azathioprine was 1.5 mg/kg and prednisolone was tapered within 2 weeks to a dose of 5 mg daily.
Bronchoscopy with transbronchial biopsies (TBB) were performed at weeks 2, 4, 6 and 12, and acute rejection episodes equal to or higher than grade A2 according to the ISHLT Classification (8) were treated with iv methylprednisolone 1g daily for 3 days, followed by prednisolone tapered to maintenance dose over a 3-week period. Recurrent rejection in one patient was treated with a change from cyclosporine to tacrolimus.
Antimicrobial prophylaxis and treatment
All lung transplant procedures were covered by broad-spectrum antibiotic prophylaxis with meropenem and ciprofloxacin until the removal of all intrathoracic lines.
All patients received 3 months of CMV prophylaxis, irrespective of donor and recipient CMV serotype, with oral valganciclovir at a standard dose of 450 mg twice daily and fungal prophylaxis with oral voriconazole at a standard dose of 200 mg twice daily.
In addition, all patients received life-long prophylaxis against pneumocystis carinii and Toxoplasma gondii with a combination of sulfamethoxazole (400 mg) and trimethoprim (80 mg) treatment, once daily. All prophylactic treatments were initiated on the first postoperative day.
Antihypertensive treatment
At baseline before transplantation, nine patients received a loop diuretic; two were given a beta-receptor blocker, one a calcium channel antagonist and one an ace inhibitor.
Operative techniques and perioperative monitoring
The procedures used in the Danish National Lung Transplant Programme have been described in detail previously (5). Briefly, SLTx procedures were preferred for patients with emphysema (n = 19). SLTx generally involved a posterolateral thoracotomy resection of the recipient's native lung, and implantation of the donor lung without the need for cardiopulmonary bypass.
Sequential DLTx was performed in patients with cystic fibrosis (CF), bronchiectasis and primary pulmonary hypertension (PPH) (n = 3) or via sternothomy and the use of cardiopulmonal bypass (n = 9).
Mean arterial pressure (MAP), diuresis, intubation time and eventually extubation in the operation room (OR), use and length of cardiopulmonary bypass, blood loss and fluid administration and arterial blood gas were registered by careful review of the anesthesia charts. Furthermore, the use of vasoactive substances, ephedrine, noradrenalin and dopamine were registered.
Data analysis
Data pertaining to age, gender, height, weight, pretransplant lung disease and type of transplant were ascertained for all patients. Measured GFR values were corrected for surface area and standardized to a body surface area of 1.73 m2.
Data analyses were performed using Statistical Analysis Software (SAS) version 9.1. Unless specified otherwise, continuous data are described as mean ± standard deviation (± SD) or median and interquartile range (IQR) for normal and skewed distributions, respectively. Group comparisons of continuous data were performed using t-tests for normally distributed data and nonparametric methods for nonnormally distributed data. Chi-square or Fisher's exact tests were used for group comparisons between categorical data, where appropriate. A p-value of <0.05 was used to determine significance. Multivariate analyses were performed using repeated measures analyses of variance (ANOVA) with mGFR as dependent outcome. Patients were entered into the model as a random effect. All other variables were entered as possible time-dependent fixed effects. A general Satterthwaite approximation was chosen for the determination of denominator degrees of freedom for fixed effects. The model was reduced sequentially by backward elimination (p < 0.05). Subsequently, we conducted an individual forward selection with the remaining nonsignificant parameters to ensure that these were not significant in the final model. The following patient parameters were evaluated: mean- and AUC-cyclosporine (CyA) at 1, 2, 3 and 12 weeks and interactions with time, recipient age older than 18 years (10 year increments), gender, body mass index, pretransplant mGFR, acute renal failure (ARF). ARF was defined as rapid rising urea and doubling of p-creatinine, oliguria or GFR <15 mL/min during the first 2 weeks posttransplantation. The time at which ARF occurred was recorded and ARF was only used as independent variable for outcomes occurring later in time (i.e. ARF was included as a time-dependent independent variable). Other variables that were analyzed in the model included: PGD, use and length of cardiopulmonary bypass, use of trasylol and interaction with cardiopulmonary bypass, use of epidural blockade during intubation and operation, ephedrine, dopamine, treated rejections and presence of arterial hypertension or diabetes mellitus requiring treatment at baseline, type of transplantation (SLTx, DLTx) and transplant indication (emphysematous, CF, IPF and PPH).
Results
A total of 155 B-CyA, 155 p-creatinine, and 143 mGFR values for the 31 patients were included in the subsequent analyses. The mean mGFR before transplantation was 103 ± 19 mL/min/1.73m2 (Table 1).
| Patient group | N | P-creatinine0 (μmol/L) | p-Value | mGFR0 (mL/min/1.73m2) | p-Value |
|---|---|---|---|---|---|
| All | 31 | 60 ± 16 (34–107) | 103 ± 19 (63–141) | ||
| Gender: | |||||
| Male | 15 | 67 ± 16 (48–107) | <0.008 | 106 ± 19 (63–141) | 0.3 |
| Female | 16 | 53 ± 13 (34–85) | 101 ± 18 (81–140) | ||
| Disease category: | |||||
| COPD/A1AT def. | 21 | 60 ± 16 (42–107) | 0.61 | 101 ± 18 (63–141) | 0.17 |
| CF/bronchiectasis | 7 | 58 ± 16 (34–85) | 115 ± 19 (87–140) | ||
| PPH/IPF | 3 | 68 ± 19 (53–89) | 95 ± 12 (81–104) | ||
| Transplant type: | |||||
| SLTx | 19 | 77 ± 15 (34–131) | 0.45 | 93 ± 18 (41–135) | 0.17 |
| DLTx | 12 | 81 ± 16 (48–160) | 96 ± 22 (40–141) | ||
- Data are presented as mean ± SD (range). Only p-creatinine were significantly different at the p < 0.05 level using Holm's correction for multiple pair-wise comparisons.
- COPD = chronic obstructive pulmonal disease; CF = cystic fibrosis; PPH = primary pulmonal hypertension; IPF = idiopathic pulmonal fibrosis; SLTx = single lung transplantation; DLTx = double lung transplantation.
After the transplantation, the mean mGFR declined to 67 ± 21, 61 ± 25 and 56 ± 17 mL/min/1.73m2 at 1, 3 and 12 weeks (Figure 1). The acute decline was seen in all patients. This corresponds to a 35% and 47% reduction (p = 0.0001) in mean mGFR at 1 and 12 weeks after transplantation compared to baseline values. Mean p- creatinine before the transplantation and 12 weeks after increased from 0.060 ± 16 μmol/L to 0.098 ± 29 μmol/L (p = 0.0001, Figure 2). Five patients developed ARF within the first 2 weeks, not biopsied but with oliguria, marked increase in urea and a doubling of p-creatinine or an mGFR below 15 mL/min. All were treated conservatively, with some recovery of renal function.
51Cr-EDTA measured GFR (mL/min) before and in the weeks after LTx.
Development in p-creatinine (μmol/L) following LTx in 31 patients.
The mean B-CyA-concentration at 1, 2, 3 and 12 weeks were 168 ± 65, 196 ± 61, 240 ± 52 and 206 ± 96 μg/L, and the decline in mean mGFR coincided with a rise in B-CyA concentration, which is in accordance with data previously described by our research group (2).
SLTx was performed in 19 patients and DLTx in 12 patients (Table 1). Cardiopulmonal bypass by heart-and-lung machine (HLM) was used with 9 patients, including 5 of the 7 CF patients. The mean HLM-perfusion time was 169 minutes (range 79–253 minutes). Trasylol anticoagulant was used in 7 of 9 patients supported by HLM and in 2 patients not supported by HLM. Single lung ventilation was performed in 25 patients during transplantation.
Eight patients were extubated in the operating room before transfer to an ICU; the rest of the patients were transferred while intubated.
Perioperative readings of MAP were registered every 15 minutes and on average were below 60 mmHg for 15 min (range 15–255 minutes) in 20 patients during intubation and the operation. The mean MAP of all patients was 67 ± 8 mmHg.
Blood loss was on average 1 L and was replaced by red blood cells (RBC) and titrated to an average hematocrit of 30–35% supplemented by fresh frozen plasma (FFP), crystalloids and colloids, targeting a neutral fluid balance at the end of the operation. Mean diuresis per hour was 0.2 L (range 0.01–0.63 L/h). Four patients were operated on again within 24 h due to bleeding (Table 2).
| Preoperative | |
|---|---|
| Age (years, ± SD) | 49 (13) |
| COPD (n) | 21 |
| CF (n) | 7 |
| PPH/IPF (n) | 3 |
| P-hemoglobin (mmol/L) | 7.5 ± 1 |
| Systolic blood pressure (mmHg) | 134 ± 27 |
| Diastolic blood pressure (mmHg) | 81 ± 15 |
| Antihypertensive (n,%) | |
| 0 | 18 (58) |
| 1 | 9 (29) |
| 2 | 4 (13) |
| Use of insulin (n,%) | 2 (6) |
| Perioperative | |
| Mean-MAP (mmHg) | 67 ± 8 |
| MAP <60mmHg/15 min (n, range(min)) | 20/31 (15–255) |
| Diuresis (l/h) | 0.2 (0.01–0.6) |
| Blood loss (liter, range) | 1 (0.3–2.7) |
| P-hemoglobin (μmol/L) | 6.4 ± 1 |
| Use of epidural blockade (n) | 18/31 |
| Use of vasopressors (n) | |
| Ephedrine | 24/31 |
| Dopamine | 23/31 |
| Noradrenalin | 14/31 |
| Postoperative (n,%) | |
| PGD = 0 or bleeding | 6 (19) |
| PGD = 1 | 9 (29) |
| PGD = 2 | 10 (32) |
| PGD = 3 | 3 (10) |
| 0 rejections <12 weeks | 9 (29) |
| 1 rejections <12 weeks | 11 (36) |
| 2 rejections <12 weeks | 9 (29) |
| 3 rejections <12 weeks | 3 (6) |
| Acute renal failure <2 weeks | 5 (16) |
| Antihypertensive at 12 weeks | |
| 0 | 4 (13) |
| 1 | 22 (71) |
| 2 | 3 (10) |
| 3 | 2 (6) |
- COPD = chronic obstructive pulmonal disease; CF = cystic fibrosis; PPH = primary pulmonal hypertension; IPF = idiopatic pulmonal fibrosis; MAP = mean arterial pressure; PGD = Pprimary graft dysfunction.
The most frequent PGD scores were 1 and 2 (29% and 32%, respectively) and repeated TBB revealed one or more rejections, graded L-A2 or more and requiring treatment in 22 patients within the first 12 weeks (Table 2).
Fourteen (45%) previously nondiabetic patients received insulin at some time postoperatively, but only two (6%) continued on this therapy at 12 weeks. Antihypertensive treatment, primarily a loop diuretic, was given to 13 (42%) patients preoperatively and to 27 patients (87%) at 12 weeks (Table 2).
There was no mortality within the study period.
Time-dependent analysis of repeated measurements
A time-dependent repeated measures ANOVA model described a 35% decline in mGFR at 1 week posttransplant and a 47% decline in mGFR at 12 weeks posttransplant. The results of backward elimination multivariate analysis of the recipient, transplant and early posttransplant variables with mGFR as the dependent outcome are shown in Table 3. Time (p < 0.0001), ARF postoperative (p = 0.0003), use of cardiopulmonal bypass (p = 0.04), older than 18 years (per 10 year increments) at the time of LTx (p = 0.006) and use of ephedrine (p = 0.048) were independent risk factors for a decline in mean mGFR posttransplantation. However, AUC-CyA, PGD, use of trasylol, dopamine, antihypertensive therapy, insulin and treated acute rejections were not significant risk factors for the decline in mGFR. Furthermore, gender, BMI, pretransplant mGFR, transplant indication and type of transplantation were not significant factors in the final model describing the time-dependent acute decline in renal function.
| Parameter | Estimate | 95% Confidence lntervals | p-Value |
|---|---|---|---|
| Intercept | 136 | 117, 155 | <0.0001 |
| One week after LTx | − 34 | − 41, − 27 | <0.0001 |
| Two weeks after LTx | − 38 | − 45, − 31 | <0.0001 |
| Three weeks after LTx | − 46 | − 54, − 39 | <0.0001 |
| Twelve weeks after LTx | − 43 | − 51, − 36 | <0.0001 |
| ARF posttransplant within 2 weeks | − 20 | − 31, − 9 | 0.0003 |
| Use of heart and lung machine | − 14 | − 26, −1 | 0.04 |
| Ephedrine | − 12 | − 24, − 0.1 | 0.048 |
| Recipient age older than18 year (per 10 year increments) | − 6 | −10, − 2 | 0.006 |
- The intercept is the estimated mean GFR for an 18-year-old patient at time 0 before transplantation. All other parameters are regression coefficients, estimating the time-dependent changes in mean mGFR associated with each variable.
Discussion
An irreversible loss of kidney function of 35% and 47% at 1 and 12 weeks, respectively, after LTx has been demonstrated. Several risk factors for decline in renal function were identified. Time after transplantation (p = 0.0001), ARF within 2 weeks posttransplantation (p = 0.0003), older than 18 years (per 10 year increments) at the time of LTx (p = 0.006), use of cardiopulmonal bypass (p = 0.04) and the use of ephedrine (p = 0.048) were independent risk factors for the decline in mGFR posttransplantation. P-creatinine was significantly increased at 12 weeks postoperative (p = 0.0001) and the means of B-CyA and AUC-CyA increased, coinciding with the decline in mGFR, but did not seem to be independent risk factors according to this analysis, nor were there significant interactions between mean B-CyA, AUC-CyA and time.
Clinical, biochemical and physiological events during the perioperative and immediate postoperative period therefore seem to be even more important to the development of impaired renal function than previously expected. This is the first study to show this using a reliable isotope GFR method.
Previous retrospective studies by Rocha et al. of 296 LTx patients (9) demonstrated that 166 patients (56%) experienced a doubling of p-creatinine within approximately 1 week (ARF) and that risk factors for ARF requiring dialysis (ARFD) were a diagnosis other than COPD, lower baseline GFR, posttransplant use of amphotericin and mechanical ventilation beyond 1 day. However, when dialysis was necessary, then all clinical outcomes, including mortality, were greatly affected.
Antihypertensive agents, primarily loop diuretics, were used by 13 patients in the current study before the transplantation and blood pressure measurements just before the operation revealed normal blood pressure in the majority of patients. During the operation, 20 patients were peroperatively registered as having a MAP below 60 mmHg, despite the persistent pre- and peroperative fluid therapy and the use of vasoactive substances performed by the anesthesiologists, and the use of ephedrine was a significant risk factor for the decline in mGFR in this context.
Use of cardiopulmonal bypass by HLM was a significant risk factor for the decline in mGFR. The use of HLM in 5 out of 7 CF patients also underlines how vulnerable to kidney damage these patients are (2). Rinder et al. demonstrated that cardiopulmonal bypass increases circulating neutrophil responses and this is associated with deteriorating postoperative renal function (10). Studies of complications with cardiac surgery and liver transplantation clearly describe how perioperative hypotension and the following use of vasopressors, and the use of cardiopulmonal bypass beyond 94 minutes are risk factors for postoperative renal failure (11–13).
Four patients were reoperated on within 24 hours due to bleeding. However, blood loss, use of trasylol during operation and rebleeding were not significant risk factors for the decline in renal function.
Of the factors that play a role in kidney failure after LTx, especially the long-term pathophysiology of CyA-induced kidney damage is well described and cyclosporine is considered to be the main cause of decline in renal function in transplant patients. The mean B-CyA concentrations or AUC-CyA at times 1, 2, 3 and 12 weeks were not documented to be independent risk factors for the acute decline in mGFR, but doses of cyclosporine given to these patients by far exceed nephrotoxic levels documented to induce fibrosis and irreversible damage as seen in biopsies after months of CyA treatment (14–16). These changes could explain in part the late development and sustained renal failure and hypertension seen in our cohort, judged by the fact that at 12 weeks, 25 patients were receiving antihypertensive therapy (Table 2). Strategies to avert or slow the progression of posttransplant renal disease include calcineurin inhibitor dose reduction or withdrawal, with or without the substitution of other immunosuppressive agents and the use of renoprotective agents like, ACE-I inhibitors, or Ang-II receptor blockers (17,18).
In conclusion, 51Cr-EDTA measured mean GFR declined 35% and 47% within 1 and 12 weeks after LTx. The decline coincided with a rise in cyclosporine concentration but was not associated with level of cyclosporine. This study suggests that current concentrations of cyclosporine used in lung transplantation are far above the level where renal damage occurs and that strategies for renoprotective regimens are urgently needed, especially in the vulnerable peri- and postoperative period.
Acknowledgments
The study was supported by an unrestricted grant from the Danish Kidney Foundation, the Helen Bjørnows Foundation and the Novartis Research Foundation. M Hornum was supported by a fellowship from Rigshospitalet, Denmark.
The authors would like to thank the laboratory technicians Annette Vinding, Helle Christensen, Andreas Haltorp and Torquil Watt, MD, for their skilful contribution to the data collection and analysis and Karl Kristensen, Department of Biostatistics, University of Copenhagen, for statistical revision.
