The last decade has witnessed a significant improvement in short-term kidney transplantation outcomes, but long-term graft survival remains suboptimal, with the mean graft survival rate at 10 years ranging from 50% to 70%.1,2 Additionally, the increasing patient lifespan and better access to healthcare have led to a growing population of kidney transplant (KT) recipients who experience graft failure and return to dialysis, which is currently the fourth leading cause of incident dialysis.3,4

Patients with end-stage kidney disease (ESKD) who begin dialysis after graft loss encounter various complications and face an increased burden of morbidity beyond chronic kidney disease (CKD). Previous studies have indicated lower quality of life scores, a higher incidence of depression, increased hospitalizations, and notably higher mortality rates.5–8 Certain characteristics of this patient group may contribute to the poorer outcomes observed, including prolonged CKD duration and ongoing exposure to immunosuppression (IS), which is linked to an increased risk of life-threatening infections,9 endocrine and metabolic dysfunction,10 and specific types of neoplasms.11

Considering the aging global population and the improved survival rates of KT recipients, we are witnessing a growing number of older, frailer patients with more comorbidities returning to dialysis, which might also influence the increased morbidity and mortality rates post-graft failure.12

The Clinical Frailty Scale (CFS) and the Charlson Comorbidity Index (CCI) have been utilized to assess frailty and clinical outcomes in patients with ESKD, including dialysis, KT recipients, and those under conservative care management.13–17 Frailty, a multisystem clinical syndrome resulting from the accumulation of vascular, inflammatory, and age-related challenges, is increasingly prevalent in the CKD hemodialysis (HD) population and significantly impacts clinical outcomes such as hospitalization, institutionalization, decreased quality of life, and mortality.18 The CCI score serves as a valid index with strong discriminatory power in predicting survival at the onset of dialysis.17 For patients facing high morbidity and mortality rates following graft loss, these indices offer valuable tools for multidisciplinary decision-making regarding whether to initiate dialysis or pursue conservative kidney management, engaging both patients and their families.16,19

The community of physicians involved in caring for KT recipients is increasingly aware of the importance of managing post-graft loss.20 Still, coordination with dialysis nephrologists, conservative care teams, and timely referrals for retransplantation remains suboptimal.21 Despite the high morbidity and mortality observed in this patient population, the systematic assessment of frailty and comorbidities is still scarce. Consequently, the potential role of these factors in guiding therapeutic decisions and informing prognosis has been underexplored.22,23

The primary aim of our study was to identify predictors of hospitalization and mortality over a two-year period among ESKD with a failed graft who initiated HD, focusing on comorbidity and frailty scores.

We performed an observational retrospective cohort study of ESKD patients with graft failure who started HD between January 2019 and December 2020 at Coimbra University Hospital, a large tertiary care centre in Portugal. Our study was carried out following the Declaration of Helsinki and approved by the ethics committee of Coimbra University Hospital (#CE-002-2020). Given the study's retrospective nature, sample size, and outcome of our sample population, the same committee waived the need for informed consent.

Patients over 18 years old who had at least two appointments with a transplant nephrologist the year before HD initiation were included. Patients with a previous graft, multiorgan transplant, or graft failure within the first month post-transplant were excluded. Demographic data, comorbidities, and transplant-related variables, including biopsy-proven rejections according to Banff 2017 criteria, were collected from patient records. We have also accessed patients IS withdrawal scheme post-graft failure. Laboratorial values, including serum albumin, hemoglobin and parathormone levels were registered in the last appointment post-graft failure.

The estimated glomerular filtration rate (eGFR) at graft failure was defined by the CKD-EPI formula. Whether planned or unplanned, the decision on start dialysis was based on clinical and laboratory parameters. An urgent HD start was defined as the need to begin HD imminently or within 48h after presentation to the Emergency Room. An unplanned start was defined as the absence of autologous vascular access and/or the requirement for hospitalization to start HD. Residual diuresis was defined as the 24-h urinary output (mL/day) at the last nephrology appointment.

To our knowledge, this study is the first to examine the relationship between frailty, comorbidity, and outcomes in patients with a failing kidney graft who are initiating HD. Over a two-year follow-up period, 37.4% of patients were hospitalized (0.44 events per patient), most commonly due to infection and CV events. Frailty, measured by the CFS, was significantly associated with hospitalization (OR 1.99), highlighting its relevance in predicting post-graft morbidity. In contrast, no significant correlation was found between CCI and hospitalization. Given that the mean time to death was 9.3±5.7 months and that CV-related were the second most common cause, we hypothesize that mortality acted as a competing risk, thereby reducing the likelihood of subsequent hospitalization. Eighteen patients (17%) died during follow-up, most within the first year. Severe infections were the leading cause of death, aligning with previously published data.22 Both CFS and CCI emerged as independent predictors of mortality, as reported in native kidney ESKD.28,29 A cut-off CCI score of ≥8, provided good sensitivity and specificity for two-year mortality risk.

Frailty in ESKD patients with failing grafts is an emerging concern. As graft survival increases and the transplant population ages, comorbidities – particularly CV disease – accumulate.30 In a large study by Gill et al., older age, diabetes, peripheral vascular disease, heart failure, and low serum albumin were the strongest predictors of post-graft failure mortality.22 Other large cohorts have similarly reported high rates of hospitalization and mortality, especially in the first year.8,31,32 However, tools for stratifying patients using frailty and comorbidity scores remain underdeveloped.

Personalizing care for patients with failed allografts has become a key recommendation in recent transplant guidelines.20,21 KT recipients face unique risks due to prolonged IS and complex trajectories of CKD. In a previous work from our group that compared KT recipients with native ESKD patients, we found similar mortality rates despite KT recipients being about 10 years younger – emphasizing the complexity of their risk profile.33

Identifying frail patients at the point of graft failure creates opportunities to intervene early – potentially reducing infection- and CV-related deaths. It also allows tailored discussions around all available post-graft therapies, including conservative kidney management. In native CKD, patients over 80 with high comorbidity and poor functional status may have similar life expectancies, whether managed with dialysis or conservatively.34 We found that older diabetic patients with high CCI scores had a mean survival of only 9 months after starting dialysis, underscoring the importance of considering non-dialysis options. However, none of our patients were referred for conservative care. Managing conservative care post-graft failure presents unique challenges, including ongoing IS.19 While maintaining IS can preserve residual kidney function (RKF) and prevent graft intolerance syndrome – a subclinical inflammatory state that increases morbidity – it also raises risks of infection, malignancy, and metabolic complications.35–37 In our cohort, continued CNI use was not associated with increased hospitalizations or mortality, aligning with recent meta-analyses.38 These findings support individualized IS tapering, with low-dose steroids potentially reducing intolerance syndrome and anuria risk.

Advanced chronic kidney disease units have been established to enhance the quality of life, survival, and morbidity in ESKD.39 The British Transplantation Society recommends dedicated low-clearance transplant clinics offering multidisciplinary tailored care to reduce unplanned dialysis starts and promote early decision-making on dialysis, conservative care, or retransplantation.40,41 More recent studies show better survival rates in patients returning to dialysis, which might reflect the shift of care and the focus on ESKD post-graft failure.42,43 Considering our study results, a low-clearance clinic is being developed for KT patients.

Our study was limited by its retrospective nature and small sample size, including the small number of hospitalization and mortality events that likely affected statistical power. We also lacked data on peritoneal dialysis starts, which may influence outcomes via preserved RKF. Additionally, missing data on induction therapy and donor characteristics further limits the analysis. It is also essential to notice that the study was performed during the SarsCoV-2 pandemic, which had a generally negative impact on the care of chronic patients.44

In conclusion, our findings highlight the importance of frailty and comorbidity in shaping outcomes after kidney graft failure. Patients with a CCI ≥8 face the highest risk of early mortality, particularly in the first-year post-failure. Early identification of at-risk individuals when eGFR declines below 20mL/min/1.73m2 can facilitate smoother transitions, better IS management, tailored symptom control, and proactive retransplantation or conservative care planning.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

We have no conflicts of interest to disclose.