THE POSSIBLE ROLE OF EARLY POST-TRANSPLANT INFLAMMATION IN LATER ANEMIA IN KIDNEY TRANSPLANT RECIPIENTS

Delayed kidney graft function and acute rejection in the early post-transplant period aff ect both short and long-term allograft survival. Allograft rejection, as an infl ammatory state, results in increased erythropoietin resistance, which leads to decreased haemoglobin (Hb) level. We conducted this study to evaluate whether infl ammation in the early post-transplant period could predict later anemia. Th is is a retrospective cohort study based on the analysis of  existing clinical records. Predictor: White blood cells (WBC) count obtained by the end of the fi rst week post-transplant (W). Covariates: Donor's age, recipient's age and sex. Outcome: Anemia identifi ed at  months (M) post engraftment. Median WBC count at W was , x/μL (th – th percentile , x/μL -, x /μL). Mean Hb values at M were , ± , g/L, in males , ± , g/L and in females , ± , g/L. Th e signifi cant correlation was found between WBC at W and Hb at M. Pearson coeffi cient of correlation r was -,, and  confi dence interval (CI) for r was -, to -, (p=,). Univariate logistic regression showed signifi cant association between WBC at W and Hb at M (OR ,;  CI , to ,, p=,). After the adjustment for donor's and recipient's age by transplantation and recipient's sex, multiple regression showed that WBC count remained predictive of anemia at M (OR ,;  CI , to ,, p=,). Early post-transplant infl ammatory response predicts later anemia in kidney transplant recipients. An increase in WBC count in the fi rst week post-transplant by /L increases the risk for anemia after twelve


Introduction
The anemia of chronic kidney disease is, in most patients, normocytic and normochromic, and is due primarily to reduced production of erythropoietin (EPO) by the kidney (a presumed refl ection of the reduction in functioning renal mass), and to shortened red cell survival ().Post-transplant anemia attracts less attention than anemia in chronic kidney disease, it is common and still remains substantial problem.The prevalence of anemia at  years post-transplant was reported to be as high as , ().Anemia can be observed both in early and in late post-transplant period ().In the early post-transplant period anemia assessment that can be attributed to renal reasons is not reliable because the perioperative blood loss, frequent phlebotomies and fl uid shifts interfere with its proper evaluation ().Th erefore, the studies that evaluate anemia usually start with anemia assessment performed at least  months after transplantation (), although there have been recently the indicators that in modern era anemia could be relialably assessed even earlier ().Inflammation is known to contribute to anemia in general population and in chronic kidney disease.Likewise, infection in kidney transplant recipients has been associated with an increased risk of anemia ().It has been suggested that allograft rejection, as an ongoing infl ammatory state, results in increased EPO resistance, which leads to decreased haemoglobin (Hb) level ().Acute rejection in the early post-transplant period ( to  weeks) is known to aff ect kidney function adversely () and kidney dysfunction is a predictor of anemia, due to decreased EPO production ().Th e incidence of subclinical rejections may vary between -  in protocol biopsies taken at various time points between  and  months post-transplant ().In the presence of acute allograft rejection  genes involved in haemoglobin transcription were found down-regulated ().However, no reports have been published about the association between early infl ammatory response and later post-transplant anemia.Th erefore, knowing that delayed graft function within one week post-transplant has a major adverse impact upon both short and long-term allograft survival () and knowing that acute rejection in the early post-transplant period ( to  weeks) is known to aff ect kidney function adversely (), it could be postulated that the early inflammatory response could be associated with later post-transplant anemia.Th erefore, we conducted this study to evaluate whether an infl ammatory response in the early post-transplant period could predict later post-transplant anemia.

Materials and Methods
Design: Retrospective cohort study based on the analysis of  existing clinical records.
Setting: University-based tertiary internal medicine hospital in the town of Tuzla, Bosnia and Herzegovina.

Participants
Inclusion criteria: Patients transplanted and followed up at Tuzla University Medical Center from  -.There were in total  of them.
Exclusion criteria: Patients who did not reach the fi rst year post-engraftment, patients with a renal allograft other than the fi rst allograft and patients with missing data.Th ree patients were excluded because they died before reaching the fi rst year post-engraftment,  has lost his graft in the early post-transplant course,  was excluded because of a renal allograft other than the fi rst allograft and  because of the missing datum on white blood cell count (WBC) in the chart, leaving a total of  patients in the study,  males and  females.

Anamnesis
All patients fulfi lled standard criteria to be eligible to receive their transplants, that is they were free from any untreated current infection, active malignancy with short life expectancy, chronic illness with life expectancy of less than one year, poorly controlled psychosis and active substance abuse.Two patients transplanted in  were receiving EPO treatment by  year post-transplant and both were still anemic by that time.Other anemic patients started receiving EPO later than  year post-transplant because EPO treatment was introduced in our center only in .

Diagnosis
Biopsy-proven acute rejection was diagnosed in  patients,  exhibited signs of delayed graft function,  suffered from postoperative surgical complications,  were hypertensive and  hyperlipemic.There were  diabetics,  with coronary heart disease and  with other cardio-vascular diseases.The study was approved by the local medical ethics committee.
Main predictor: WBC count obtained by the end of the fi rst week post-transplant (W).
Main outcome measure: Anemia identifi ed at  months (M) post-transplant.

Measurements
Anemia was expressed as decreased qualitative variable of Hb.As recommended by the American Society of Transplantation, it was considered present if the Hb concentration was ≤  g/dl in men or ≤  g/ dl in women.Donor and recipient age at transplantation were represented by quantitative variables in years.Induction immunosuppressive protocol included steroids, cyclosporine, antithymocyte globulin (up to ) and humanized anti-IL- receptor antibodies (later).Maintenance immunosuppressive regimen consisted of cyclosporine, azathioprine (up to ), mycophenolate mofetil (later) and steroids.

Statistical analysis
Statistical analyses were performed with Med-Calc software (version ... for Windows, Med-Calc).The degree of association between WBC and Hb was assessed with correlation analysis using Pearson coefficient of correlation after a log-transformation to correct for a lack of normality, where appropriate.Univariate logistic regression analysis was applied to test the relationship between the main predictor and the main outcome, followed by multivariate model adjusted for other covariates.
The significant correlation was found between WBC at W and Hb at M. Pearson coeffi cient of correlation r was -,, and  confi dence interval (CI) for r was -, to -, (p=,) (Figure ).Univariate logistic regression analysis showed signifi cant association between WBC at W and Hb at M (OR ,;  CI , to ,, p=,).After the adjustment for donor's and recipient's age by transplantation and recipient's sex, multiple regression model showed that WBC count remained predictive of anemia at M (OR ,;  CI , to ,, p=,) (Table ).

Discussion
Our results showed signifi cant correlation and association, between WBC count obtained by the end of the fi rst week post-transplant and anemia identifi ed at M.An increase in WBC count at W by   / L increased the risk for anemia at M by  (Table ), which clearly demonstrates the impact of infl ammation in the early post-transplant period on anemia at  months.Th e anemia of chronic disease (infl ammation), was initially thought to be associated primarily with infectious, infl ammatory, or neoplastic disease.However, other observations have shown that anemia of chronic disease can be seen in a variety of conditions, including severe trauma, heart failure, diabetes mellitus, and in those with acute or chronic immune activation (-).Th e anemia is typically normochromic, normocytic, and hypoproliferative.Recurrent or chronic infl ammatory processes are common in individuals with chronic kidney disease, including those with chronic renal failure.Th is is due to many underlying factors, including the uremic milieu, elevated levels of circulating proinfl ammatory cytokines, oxidative stress, carbonyl stress, protein-energy wasting, enhanced incidence of infections and others.It has been suggested that the underlying infl ammatory medical condition causes the release of cytokines such as the interleukins (eg, IL- and IL-) () and tumor necrosis factor (TNF-alpha) by activated monocytes.Th ese cytokines may then unleash a cascade including the secretion of interferon (IFN)-beta and IFN-gamma by T lymphocytes.As an example, IFN-gamma when given to experimental animals can produce the picture of anemia of chronic disease with most of the abnormalities noted above (.Among patients with chronic kidney disease, the presence of an infl ammatory state may also be closely related to accelerated atherogenesis, protein-energy malnutrition, and anemia (-) Likewise, the relationship between infl ammation and anemia in kidney transplant recipients has already been established ().Winkelmayer et al. describe an inflammatory response in the form of rejection and conclude that the rejection could have caused in their study a decrease in Hb level through EPO resistance ().
We have taken into account all risk factors for anemia that could have confounded the association between WBC count at W and Hb at M. Th ose factors were: impaired kidney function, acute rejection episodes, increased donor and decreased recipient age, female recipients and a renal allograft other than the fi rst allograft (, ).Th us, one patient with a renal allograft other than the fi rst allograft was excluded from the study and recipient's age and sex and donor's age were adjusted for in the multivariate analysis.However, we did not adjust for kidney function and acute rejection episodes in the multivariate analysis.Impairment of kidney function was not adjusted for because infl ammation elevates serum creatinine.Winkelmayer et al. in their study of association between anemia and allograft loss did adjust their analysis for creatinine clearance, although they specifi ed that "anemia would be a surrogate of chronic rejection and would refl ect a state of chronic infl ammation due to a failed graft" ().Th us, creatinine clearance in this study should not have been a confounder to be used for the adjustment because chronic rejection and infl ammation deteriorate kidney function, which renders creatinine clearance an intervening variable on a causal pathway to the allograft loss ().Besides, the problem of collinearity between anemia as a surrogate for chronic rejection and creatinine clearance as a marker of deterioration of kidney function seems to be obvious in this analysis, too.
We did not make an adjustment for acute rejection episodes in our study either, even though they are known risk factors for anemia (, ).However, acute rejection episodes did not meet the criteria to be a confounding variable in our analysis.Th e predictor was WBC count which was a surrogate for infl ammation / rejection so that we have avoided the problem of collinearity in a statistical analysis by not adjusting for acute rejections ().Furthermore, acute rejection is also an intervening variable on a causal pathway between the predictor and the outcome, since infl ammation can trigger the rejection in kidney transplant recipients ().Th at is also why WBC count should not have been adjusted for rejection ().

Limitations
Unfortunately we do not have additional parameters of infl ammation, such as cytokines, or CRP, but WBC itself is a valid marker for heightened infl ammatory state.
It could be argued that antirejection drugs including steroids used during this stage might have confounded WBC.Th is, however, is unlikely since all our patients received steroids at the same dose ( mg/kg) in an induction immunosuppressive protocol and dose variation as a possible cause of varying WBC count can be ruled out.Th is raises the likelihood that the more brisk host-response to steroids is perhaps associated with later post-transplant anemia.However, sustained elevation in WBC is more likely to represent a heightened infl ammatory state and less likely due to variable host response to the same dose of steroid.Furthermore, all patients received induction treatment, either with antithymocyte globulin (up to ), or Basilix-imab (later), which are both myelodepressive, so there was no confounding eff ect on WBC count.Th e same stands for Mycophenolate mofetil, or Azathioprine.

Conclusion
Early post-transplant infl ammatory response predicts later anemia in kidney transplant recipients.An increase in WBC count in the fi rst week post-transplant by   / L increases the risk for anemia after twelve months by .
Larger scale studies addressing the limitations discussed above are needed to establish possible association between early post-transplant infl ammatory response and later anemia in kidney graft recipients.

List of Abbreviations
WBC -White blood cells W -fi rst week post-transplant M - months post-transplant EPO -erythropoietin

TABLE 1 .
Association between WBC obtained by the end of the fi rst week post-transplant and anemia after 12 months