In the world of transplants, the introduction of new immunosuppressants has led to an improvement in short- and medium-term graft survival1–3; nevertheless, long-term graft loss remains a problem.4 The many factors associated with these long-term losses have been analysed by a large number of studies.5 Among them, recent studies6 show that proteinuria is an independent risk factor predictive of graft failure for recipients of all ages.7 At the same time, continued renal graft function decline is a strong predictor of mortality in renal transplant patients.6

The prevalence of proteinuria one year after transplantation ranges from 11% to 45%,8 or even higher in patients treated with proliferation signal inhibitors (PSI). Two mechanisms for the development of proteinuria after transplantation are described: tubular origin, due to inadequate protein reabsorption in the proximal tubule cells damaged by ischaemia/reperfusion phenomena, rejection or tubular toxicity, or glomerular origin due to increased passage of higher molecular weight proteins such as albumin through the glomerular barrier due to de novo or transplant glomerular disease, chronic rejection or drug toxicity.

Proteinuria in transplantation is usually indicative of some type of graft disease and generally tends to occur in relation to chronic graft nephropathy (histological substrate with interstitial fibrosis and tubular atrophy), acute rejection, transplant glomerular disease or recurrence of primary kidney disease.8–10 It can be associated with immunological or non-immunological factors11 and factors related to both the donor and the recipient. Degree of HLA sensitisation, age, obesity, hypertension and other cardiovascular risk factors can all contribute to the development of post-transplant proteinuria. Immunosuppressive therapy has also been associated with proteinuria,12 with reports that PSI in particular can induce it or make it worse if it already exists.13,14

Nephrotic-range proteinuria is associated with a worse renal graft outcome, as occurs in other diseases in the native kidney. However, urinary protein excretion usually considered as mild (<500mg/day) from the first months after transplantation15 is also associated with worse transplantation outcomes and leads to much poorer graft and patient survival rates, and may even provide information on the graft.16–17 However there are no studies that focus on the evolution of urinary protein excretion and its repercussion on renal graft survival.

To analyse the prevalence of proteinuria in the renal transplant population in a particular regional area (Andalusia in southern Spain). To identify contributory factors and the consequences of proteinuria, particularly for graft and patient outcome. To evaluate the impact of the evolution of urinary protein excretion over time on graft survival.

We present data from the Renal Transplant Registry of Andalusia, which forms part of a general registry of replacement treatment for chronic kidney disease in the region that has been in operation since 1984.

For this review, we have analysed the data from 1 January 2000 to 31 December 2009, continuing the observation up to 25 May 2012 (12.25 years or 147 months). Data were collected retrospectively for the years 2000–2003 and prospectively from 2004 to 2009. We performed the analysis on the urinary protein excretion records available for 1815 patients after a median follow-up of 77.89±32.94 months (12–147) during this period.

The urinary protein excretion, or simply proteinuria, was quantified as mg/day, categorising the patients for the analysis into three levels according to severity:

• I.

  Urinary protein excretion below 300mg a day

• II.

  Urinary protein excretion 300–1000mg a day

• III.

  Urinary protein excretion over 1000mg a day

The urinary protein excretion data were collected at two patient follow-up points, three months and one year after transplantation, leading to the creation of 6 groups of patients according to severity of urinary protein excretion and time of determination: I3m, II3m and III3m, according to the classification at 3 months and I1y, II1y and III1y, for one year after transplant.

Statistical analysis: The SPSS v21 statistical package was used.

An overall descriptive analysis was performed on the entire study population. Comparative analysis was carried out according to proteinuria groups, for which we performed a bivariate analysis using ANOVA, applying Welch's test for those groups in which there was a lack of homoscedasticity in the variances of the groups. For analysis of multiple comparisons in quantitative variables that showed statistically significant differences with respect to the proteinuria groups, we used multiple comparison tests indicated for groups with unequal variances and Bonferroni for groups with equal variances. Pearson's Chi-square test was used to compare qualitative variables.

The level of significance was considered to be p<0.05.

Patient and graft survival censored for death was studied by a Kaplan–Meier analysis and a multivariate analysis was performed using a Cox regression model by means of manual stepwise regression.

The following variables were introduced into this model: donor age; donor gender, cause of donor death; recipient age; recipient gender; pre-transplant cytotoxic antibodies; maximum cytotoxic antibodies; HLA-A, -B and -DR incompatibility; ACE inhibitors (3 months); ARA2 (3 months); creatinine (3 months); creatinine (1 year); proteinuria (3 months); and proteinuria (1 year).

We studied a population of 1815 kidney transplants performed since 2000 with a follow-up period of up to 12.25 years to analyse the impact of proteinuria on renal transplantation outcomes.

Analysing factors that can influence the development of proteinuria, we found donor-related factors, such as cause of death (CVA vs non-CVA), to be determinants. In our case, the groups with the highest levels of proteinuria were recipients from donors who had died from cerebrovascular causes. It might be assumed that grafts from donors who have died from a CVA would have a worse cardiovascular profile. Since we believed that this relationship might be influenced by donor age, we analysed the cause of death, whether CVA or not, according to quartiles of donor age and found significant differences: as expected, there were more deaths from CVA in older age groups. We also found differences in proteinuria according to quartiles of donor age, especially at three months after transplantation, where donor grafts of the extreme age quartiles (1st and 4th) were those with most proteinuria. This leads us to believe that the age of the donor also influences the development of proteinuria, although we only found differences in the analysis of donor age at three months and not at one year by proteinuria groups. Recent studies18–21 show that older donor age has a negative effect on renal graft, both in deceased-donor and living-donor donation, but none of these studies analyse proteinuria. Many authors have found a higher degree of delayed renal function in older donors.22 In our case, delayed renal function was associated with higher levels of proteinuria and a greater risk of graft loss.

Moreover, in addition to our group, other authors23 have found that higher pre-transplant levels of cytotoxic antibodies negatively affect graft outcome; in our case, we found that patients with higher antibody levels had more proteinuria and, indirectly, worse graft outcome.

Our patients who had more post-transplant hypertension had more proteinuria and worse graft outcome, with post-transplant hypertension acting as an independent factor for graft loss. This could be explained by the fact that hypertension is an expression of multiple factors, some graft-related, some recipient-related and others related to immunosuppressive therapy or chronic renal disease,24 which continue after the transplant as cardiovascular disease; in this clinical setting, we know that the renin–angiotensin–aldosterone axis (RAAS) is activated and also, that this activation favours the onset of proteinuria.25 Improvement of proteinuria and its consequences have also been reported after treatment with RAAS blockers.26 In our case, the percentage of patients treated with angiotensin converting enzyme (ACE) inhibitors and angiotensin II receptor antagonists (AIIRAs) at three months after transplantation was small and no differences were, therefore, found between proteinuria groups.

Immunosuppressive induction therapy affects the renal graft.27 We analysed induction therapy and, although we did not find it to be a risk factor for graft loss, we did find that patients who had received thymoglobulin had significantly more proteinuria one year after transplantation than those treated with anti-CD25 or those who had received no induction. Although this finding may seem paradoxical, one explanation could be that in our region, use of thymoglobulin induction is rare and confined to patients with high levels of anti-HLA sensitisation pre-transplant.

We found that post-transplantation proteinuria (at three months and one year) is predictive of graft outcome and is associated with poorer renal function, a lower graft survival rate and even with higher patient mortality rates. In the literature, there is evidence that even for the same degree of renal dysfunction, increased proteinuria can increase the risk of cardiovascular death.28 Oblak et al. recently reported that patients who have had an episode of acute rejection followed by increased proteinuria have worse outcomes than those who have suffered rejection but no increase in proteinuria,29 with both worse graft survival and increased risk of death. Some authors, such as Roodnat,30 believe that proteinuria affects not only graft survival, as we found, but also that of the patient. In our multivariate analysis, we failed to demonstrate that severity of proteinuria was associated with worse recipient survival. In the Kaplan–Meier study, higher levels of proteinuria at one year correlated with worse patient survival, but this may only reflect their association with worse renal function, which, as in the majority of published studies, was identified as a predictor of mortality in the adjusted analysis.

However, we did find that the changes in proteinuria over time seemed to significantly affect transplant outcome, worsening patient and graft survival in those in whom proteinuria worsened between three months and one year post-transplantation. Proteinuria remained stable in more than two thirds of our patients between three months and one year post-transplantation, while it worsened in 12% and improved in 18%.

Compared to cases in which proteinuria improved or remained stable between three months and one year, recipient survival was significantly worse in the group of patients with worsening proteinuria, regardless of the level they started with (<300mg or 300–1000mg/day). Nevertheless, improvement in their proteinuria had no effect on patients survival.

We think that in these patients with improvement in proteinuria (18%), this could be related to the problems after transplant and disappears in the first months, but not with graft pathology. This group, which improves proteinuria has not improved survival, possibly because the proteinuria was not a problem relative to the graft.

When analysing graft survival overall for all patients, we found that the factor most likely to worsen survival is progression of proteinuria. However, when survival for each proteinuria group at three months and one year were analysed separately, here we did find that progression of proteinuria was associated with worse survival rates, while improving proteinuria also improved graft survival, in both cases compared to patients who remained stable. It might be expected that the difference in survival would be related to a parallel worsening in renal function, but we found no differences in serum creatinine at one year in terms of whether the proteinuria progressed, remained stable or improved. We found that there were differences at three months, with renal function being worse in the group in which proteinuria subsequently improved. This may be the result of immediate post-transplant factors that could be resolved, allowing both proteinuria and renal function to improve.

We are therefore able to state that the progression of proteinuria in the first year after transplantation, on its own, will predict long-term graft survival. The mechanism underlying this association is beyond the scope of this work. Proteinuria, especially its progression, may indicate one or more ongoing damaging processes, such as chronic rejection, recurrence of underlying disease or overload of a kidney which was inadequate due to donor characteristics. Moreover, high-level proteinuria may itself exacerbate the progress of the organ.

Based on these results, we believe that in addition to analysing the degree of proteinuria at a particular time, we should monitor changes over time.

In another vein, our analysis found that factors that appear to increase the risk of proteinuria at three months, such as donor age, disappear at one year, while the cause of death of the donor remains among the factors that influence the development of proteinuria both at three months and one year.

In our case, as had previously been found by authors of our group,31 the Cox regression showed the age of the recipient as mortality factors. In addition, renal function, as measured by serum creatinine, at one year post-transplant was also a determinant of patient survival.

In the Cox analysis for graft survival, proteinuria, renal function at three months and one year post-transplant were predictive factors for graft loss.

In conclusion, we found that proteinuria is a marker of poor renal graft outcome, especially when it progresses. Although its cause may be multifactorial, it is clear that we can consider it as an epiphenomenon of a series of events that will ultimately lead to a worse outcome for the transplanted patient.

The authors declare no conflict of interest.