For the general population, the risk of adverse outcomes (mortality, progression to end-stage renal disease) increases with higher levels of albuminuria so that it was included in 2012 Kidney Disease Improving Global Outcomes guidelines as key marker for chronic kidney disease (CKD)(16).
Numerous studies in patients with diabetic and non-diabetic renal diseases confirm that marked albuminuria (> 300 mg/day) is associated with a faster rate of CKD progression. On the contrary, moderate-level albuminuria (150-300 mg/day) is not a reliable surrogate marker for CKD progression in intervention clinical trials because reduction of albuminuria can be linked to both improving and worsening CKD progression(17,18).
The deleterious impact of proteinuria, at relatively early post-KT time, on long-term outcome was clearly demonstrated in several previous studies (6,7,9,14). Nevertheless, available studies differed in definition of post-KT harmful proteinuria as well as in post-KT considered time points. That uncertainty was reflected also by available guidelines for transplant management that suggest performing allograft biopsy when there is new onset proteinuria or unexplained proteinuria 3.0 g/mg creatinine or 3.0 g/24 h, with a low level of evidence(19).
Evaluation of risk factors related to post-KT proteinuria was attempted in several studies finding a multitude of different donor-related, recipient related or transplant related factors including, among others, delayed graft function, greater body mass index at transplant, older donor age, greater HLA mismatch, tacrolimus use and antihypertensive use (13,20).
Beyond the causes, however, the intrinsic risk of developing proteinuria at a relatively early stage (first post-KT year) should be addressed as a key risk factor for transplant outcome(21).
Our study demonstrated that occurrence of proteinuria,(≥ 0.5 g/day), at the first post-transplant year, was significantly linked to graft survival and patient survival in the whole population under study. For the first time in literature, to the best of our knowledge, the impact of proteinuria on KTs from different donor age classes was analyzed, demonstrating a synergic effect of proteinuria ≥ 0.5 g/day and donor age ≥ 70 years on DCGS. Very low grade proteinuria (between 0.2 and 0.5 g/day) was not related to outcome. Nevertheless, we demonstrated an association (a trend, not reaching significance) between DCGS and low-grade PTO (≥ 0.2<0.5 g/day) only in kidneys from donors older than 70 years old. (Fig. 3).
Older donor kidneys seemed more sensitive to proteinuric injuries in comparison with younger ones with very relevant differences of DCGS in different donor age classes with same proteinuria. This is also highlighted by the fact that any variation of proteinuria between 6 months and 1 year post-KT portends a worse graft outcome when donor age was ≥ 60 years (Fig 4).
With ageing, kidney undergoes through processes that lead to reduced functional reserve and also to a lower renal reserve response to higher functional requests (e.g. protein load) due to a reduced capacity of renal autoregulation(22–25). These functional changes, that could be also encountered in diabetic patients, were found to be associated with or precede microalbuminuria and glomerular lesions(26,27). Indeed, proteinuria is not considered to be a ‘‘normal’’ physiological process of aging also in cases when the development of a persistent proteinuria increases with age, due to the higher prevalence of diabetes, hypertension and paraproteinemias in the elderly (28). Therefore it could be speculated that the aforementioned process could be amplified in kidneys retrieved from elderly donors, which became more prone to several insults leading to compensatory hyperfiltration of glomeruli that survive reperfusion injury, rejection, and drug toxicity, and,in ultimate analysis, to a faster progression of renal damage. In this context, proteinuria is, at the same time, marker of damage progression and established loss of function, as witnessed by concordance between proteinuria grade and renal function in transplanted patients.
Predisposition to proteinuria development in our population appeared to be certainly related to both donor and recipient characteristics (donor age, pre-transplant diabetes, glomerular Karpinsky score) but, as it is demonstrated by paired kidney analysis, early post-KT events (acute rejection, CMV infections, new onset diabetes after transplantation and urological complications such as ureteral stenosis) contribute to determine kidney fate and prognosis. As shown in table 3, rejections, number of biopsies, NODAT and glomerulonephritis were associated with higher 1-year PTO, surprisingly with similar distribution between donor age groups (data not shown); overall infections and CMV viremia as well as vascular and urological complications were, on the contrary, significantly more frequent in older donor population (data not shown). These data could be explained as a consequence of the indication for decreasing immunosuppressive therapy in case of infection and of the lower quality of older donor tissues in comparison with the younger donor kidneys (29,30). Notably, we found a strong correlation between overall infection and rejection rates, especially when donor age was > 50 years. Therefore, even if rejection risk seems to be similar in all donor ages, susceptibility to external factors (such as infections or urological/vascular complications) plays a major role in older donor populations.
It is well known that kidney allograft function in a stable condition (usually between 3 months and 1 year post KT) is an important predictor of graft failure (31.32). One possible explanation is that, as in chronic kidney disease, lower kidney function is often associated with other cardiovascular risk factors (e.g. hypertension, dyslipidemia and smoke) predisposing to cardiovascular disease and mortality (33). In our study (table 4) we showed that 1-year proteinuria ≥ 0.5 g/day (HR 2.77) is comparable to CKD-EPI < 44ml/min (HR 2.46) in predicting graft failure by multivariate analysis. In this context, donor age ≥ 70 years would make this association even worse.
Among the other clinical variables, CMV viremia post-transplant resulted as an independent predictor of DCGS in Cox multivariate analysis (HR 2.1), as mentioned in previous studies(34,35). As for rejection, when we consider rejection as an event in the entire follow-up, its role is comparable to the one of the other main risk factors (HR respectively 2.5 vs 2.1 and 2.4) (table 4). This is not found for early rejections (1st year rejections) possibly because their role is somehow downsized in a context of a population of older donors and recipients in which other factors are probably more relevant.
Need for surrogate endpoints to enhance trial feasibility has been outlined by a recent review where proteinuria has been defined as a prognostic biomarker(36). In the current scenario, the majority of available organs are represented by “suboptimal” donors (formerly known as ECD or with a Kidney Donor Profile Index greater than 85% according to the recent USA definition). Our study demonstrated prognostic significance of proteinuria, in particular with this kind of donors.
The link among proteinuria, donor age and subsequent higher proteinuria-mediated damage in older donors is an important issue of our study. Several mechanisms were advocated for proteinuria mediated tissue damage such as intratubular complement activation(37,38), intratubular protein overload (39–41), radical oxygen damage induced by tubular reabsorption of iron carrying proteins such as transferrin(42). Older donor kidneys may potentially be more sensitive to such mechanisms, even with a lower grade of proteinuria.
As mentioned in previous studies and confirmed by our data, the evidence suggesting a benefit for ACEi/ARB use in transplant recipients is still lacking. They showed that the use of these agents was often associated with a significant reduction in proteinuria and eGFR without a concurrent improvement of patient or allograft survival as it does in non-transplant settings such as in diabetic nephropathy(43,44). This is also confirmed in a recent randomized controlled trial in which ramipril compared with placebo did not lead to a significant reduction in doubling of serum creatinine, end-stage renal disease, or death in kidney transplant recipients with proteinuria. These data would not support widespread use of this drugs to obtain clinical improvement in transplanted patients (45).
It is well known that Mtor-inhibitors may induce proteinuria by targeting glomerular podocytes(46). This is confirmed also in our cohort. However, excluding patients with mTor maintenance therapy in the first year (16%), the impact of proteinuria on outcome was confirmed.
Our study has some strenghts and some limitations. Strenghts of the study are related to homogeneity of the population, characterized by a wide range of data coming from over a thousand of KTs performed with the same team of surgeons, nephrologists and pathologists. Patients were centrally followed in the long term with all data recorded in patients’ charts.
Another strength, in our opinion, is the validation of prognostic impact of proteinuria in a subset of paired kidneys, thus limiting undetermined donor-derived confounding factors(47).
We acknowledge that a limitation of this study is the absence of protocol graft biopsies for center policy; however, this limitation reduces its importance when we consider that proteinuria impact was shown by some authors to be independent from the underlying renal allograft histology(6).
Other limitations are: absence of routinely donor specific antibody evaluation in the first year, which was available only in a minority of patients, so that we did not evaluate our population under this aspect; moreover lack of qualitative differentiation of urinary protein, considering that tubular or glomerular proteinuria could have different impact on graft outcome, as underlined in previous studies(11,48,49).