Clinical Kidney Journal, 2022, vol. 15, no. 11, 2089–2096 https:/doi.org/10.1093/ckj/sfac159 Advance Access Publication Date: 23 June 2022 Original Article O R I G I NA L A RT I C L E Increased mortality after kidney transplantation in mildly frail recipients María José Pérez-Sáez 1,∗ , Carlos E. Arias-Cabrales1,∗ , Dolores Redondo-Pachón1 , Carla Burballa1 , Anna Buxeda1 , Anna Bach1 , Anna Faura1 , Ernestina Junyent1 , Ester Marco2 , Leocadio Rodríguez-Mañas 3 , Marta Crespo1 and Julio Pascual1,4 , for the FRAIL-MAR Study Group 1 Nephrology Department, Hospital del Mar, Barcelona, Spain, 2 Physical Medicine & Rehabilitation Department, Parc de Salut Mar (Hospital del Mar-Hospital de l’Esperança), Rehabilitation Research Group, Hospital del Mar Research Institute, Universitat Autònoma de Barcelona, Barcelona, Spain, 3 Geriatrics Department, Hospital Universitario de Getafe, Madrid, Spain and 4 Nephrology Department, Hospital Universitario 12 de Octubre, Madrid, Spain ∗ These authors contributed equally to this work. Correspondence to: Julio Pascual; E-mail: julpascual@gmail.com ABSTRACT Background. Physical Frailty Phenotype (PFP) is the most used frailty instrument among kidney transplant recipients, classifying patients as pre-frail if they have 1–2 criteria and as frail if they have ≥3. However, different definitions of robustness have been used among renal patients, including only those who have 0 criteria, or those with 0–1 criteria. Our aim was to determine the impact of one PFP criterion on transplant outcomes. Methods. We undertook a retrospective study of 296 kidney transplant recipients who had been evaluated for frailty by PFP at the time of evaluating for transplantation. Results. Only 30.4% of patients had 0 criteria, and an additional 42.9% showed one PFP criterion. As PFP score increased, a higher percentage of women and cerebrovascular disease were found. Recipients with 0–1 criteria had lower 1-year mortality after transplant than those with ≥2 (1.8% vs 10.1%), but this difference was already present when we only considered those who scored 0 (mortality 1.1%) and 1 (mortality 2.4%) separately. The multivariable analysis confirmed that one PFP criterion was associated to a higher risk of patient death after kidney transplantation [hazard ratio 3.52 (95% confidence interval 1.03–15.9)]. Conclusions. Listed kidney transplant candidates frequently show only one PFP frailty criterion. This has an independent impact on patient survival after transplantation. Received: 31.3.2022; Editorial decision: 1.6.2022 © The Author(s) 2022. Published by Oxford University Press on behalf of the ERA. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com 2089 2090 M.J. Pérez-Sáez et al. GRAPHICAL ABSTRACT Keywords: frailty phenotype, Fried, survival, transplant INTRODUCTION The problem may arise when robustness or pre-frailty con- cepts are modified from the original PFP definition. Several studies carried out both in dialysis patients and KT recipients Among kidney transplant (KT) recipients, the overall pooled have considered as non-frail patients those who had not only prevalence of frailty before transplantation has been recently 0 but also 1 PFP criteria [4, 5, 14, 15] (instead of only those estimated at 17.1% [1]. This condition implies poorer results af- with 0 criteria). This responded to the claim that too few end- ter transplantation, not only regarding clinical outcomes such stage renal disease patients would have 0 criteria. Therefore, as longer delayed graft function or inpatient stay [1], higher rate two misclassifications may potentially be present: (i) when an- of early readmissions to the hospital [2], or poorer tolerance to alyzing intermediate frail patients, these studies only consider immunosuppressants [3], but also in terms of higher mortality those recipients with 2 PFP criteria, and (ii) for non-frail pa- after transplantation [4–6]. tients, they are considering those recipients with not only 0 but Although many scales have been used to evaluate frailty in also 1 PFP criterion. In other studies involving renal patients, KT recipients [7, 8], the Physical Frailty Phenotype (PFP) [9] has the original PFP score to define different categories of robust- been the most utilized [1]. Its original version comprises five ness, pre-frailty, and frailty has been used [16]. Therefore, while domains and classifies patients as frail if they present with ≥3 there is consensus in considering as frail patients those who criteria. Non-frail patients are considered with 0 criteria and fulfill ≥3 PFP criteria and this is relevant for outcomes anal- intermediate or pre-frail patients those with 1 or 2 criteria. In yses [1], confusion might be present when analyzing results the general population, not only frailty but also pre-frailty has of pre-frail patients, a much more common condition among been associated with poorer health outcomes [9–12]. In the KT candidates and recipients, especially if we only consider 1 setting of KT, pre-frailty has also been associated with worse PFP criterion to define pre-frailty. To our knowledge, no stud- outcomes after transplantation, although it has not been found ies have analyzed the relevance of the presence of 1 PFP crite- to increase the risk of mortality after transplantation when rion, and no studies have differentiated results after transplan- adjusted models were performed [4]. In other studies, pre-frail tation accordingly to the original PFP definition (0, 1–2, and ≥3 and frail transplant recipients have been merged and outcomes criteria) vs the modified PFP classification performed in renal have been considered together for both categories [3, 5, 13]. population. Mortality in mildly frail KT recipients 2091 Table 1. Baseline characteristics and main outcomes of 296 KT recipients according to their PFP score All KT (n = 296) PFP 0 (n = 90) PFP 1 (n = 127) PFP ≥2 (n = 79) P-value Baseline characteristics Age at transplantation (years, mean ± SD) 62.6 ± 12.3 60.9 ± 11.5 61.1 ± 14.1 63.2 ± 11.3 .412 Sex (female, n (%)) 87 (29.4) 17 (18.9) 30 (23.6) 40 (50.6) <.001 BMI (kg/m2 , mean ± SD) 27.2 ± 4.8 27.2 ± 4.5 26.8 ± 4.8 27.8 ± 5.2 .288 Hypertension (n, %) 285 (96.3) 86 (95.6) 124 (97.6) 74 (94.9) .553 Diabetes mellitus (n, %) 102 (34.5) 28 (31.1) 41 (32.3) 33 (41.8) .275 Coronary artery disease (n, %) 41 (13.9) 12 (13.3) 18 (14.2) 11 (13.9) .984 Peripheral vascular disease (n, %) 25 (8.4) 4 (4.4) 13 (10.2) 8 (10.1) .262 Cerebrovascular disease (n, %) 22 (7.4) 8 (8.9) 2 (1.6) 12 (15.2) .001 Renal replacement therapy modality (n, %) Hemodialysis 164 (65.5) 51 (56.7) 81 (63.8) 62 (78.5) Peritoneal dialysis 66 (22.3) 23 (25.6) 31 (24.4) 12 (15.2) .037 Preemptive transplant 36 (12.2) 16 (17.8) 15 (11.8) 5 (6.3) Time on dialysis prior to KT (months, mean ± SD) 24.7 ± 21.1 21.0 ± 19.9 25.2 ± 22.0 27.5 ± 20.7 .162 Retransplant (n, %) 21 (7.1) 7 (7.8) 6 (4.7) 8 (10.1) .325 Transplant characteristics Type of donor (n, %) Living donor 28 (9.5) 12 (13.3) 11 (8.6) 5 (6.3) Brain-death donor 146 (49.3) 37 (41.2) 69 (54.4) 40 (50.6) .094 Circulatory-death donor 122 (41.2) 41 (45.5) 47 (37.0) 34 (43.1) Donor age (years, mean ± SD) 64.1 ± 15.3 64.1 ± 13.0 63.6 ± 16.3 65.1 ± 16.2 .794 Thymoglobulin induction (n, %) 37 (12.6) 10 (11.2) 11 (8.7) 16 (20.5) .043 Calcineurin inhibitor + mycophenolic acid (n, %) 166 (57.8) 50 (57.5) 74 (60.2) 42 (54.5) .734 Calcineurin inhibitor + mTOR inhibitors (n, %) 119 (41.5) 37 (42.5) 48 (39.0) 34 (44.2) .751 Tacrolimus C0 /D during first year after transplant 1.44 ± 0.56 1.44 ± 0.56 1.49 ± 0.57 1.37 ± 0.52 .409 (ng/mL/mg, mean ± SD)a Follow-up (months, median [IQR]) 22 [10–38] 19 [11–38] 22 [10–39] 23 [6–38] .850 KT, kidney transplant; PFP, Physical Frailty Phenotype; SD, standard deviation; mTOR, mammalian target of rapamycin; C0 /D, concentration/dose; IQR, interquartile range. a C0 /D ratio = tacrolimus trough concentration (ng/mL)/daily dose (mg). Comparisons were made among three categories: PFP 0, PFP 1, and PFP ≥ 2. Chi-square test was used to compare categorical variables, ANOVA test to compare quantitative variables with normal distribution and Kruskal–Wallis test for quantitative variables without normal distribution. We have previously reported a high prevalence of pre-frailty ing was 12.4 [84.2–184.7] days. PFP comprises five components: between KT candidates and its negative impact on outcomes shrinking (self-report of unintentional weight loss of 4.5 kg dur- while waiting for transplantation [17, 18]. In the present study, ing the past year), weakness [grip strength below an established we aim to analyze results after transplant according to the cut-off based on sex and body mass index (BMI)], exhaustion patient’s frailty status, focusing on the specific weight of having (self-report), low activity (kilocalories per week below an estab- only 1 PFP criterion on transplant outcomes. lished cut-off) and slowed walking speed (walking time of 4.5 m below an established cut-off by sex and height). Each component or question scores 0 or 1 depending on its absence or presence, MATERIALS AND METHODS respectively. In the original version, robust patients were defined Study design as a score 0, pre-frail as those who ranked 1–2, and frail patients were defined by a score ≥3. In previous studies involving dial- This is a retrospective cohort study carried out at Hospital del ysis and KT population, 0 and 1 scores were merged as a sin- Mar, Barcelona, Spain. Between June 2016 and February 2021, 296 gle category of robust patients, claiming that few patients with KT candidates who had been evaluated for frailty received a KT. end-stage renal disease score 0 [4, 5, 14, 15]. We explored the po- Median time to follow-up after KT was 22 months [interquartile tential misclassification of labeling as robust patients those who range (IQR) 10–38]. Clinical and epidemiological variables were scored 1 and, therefore, both classifications were analyzed sepa- collected from our local database. rately, including patients who scored 1 as robust and as pre-frail patients. Ethics The Institutional Review Board of Hospital del Mar approved the Study variables study, and all enrolled participants provided written informed consent at the time of frailty evaluation. The study followed the Baseline assessment included recipient’s demographics (age, principles of the declaration of Helsinki, only relying on the of- sex, ethnicity and BMI) and clinical data (comorbidities such as ficial center database. hypertension, diabetes mellitus, chronic cardiac and pulmonary diseases, type of renal replacement therapy, etc.), as well as donor characteristics and immunosuppression received. Frailty assessment Regarding outcomes, we analyzed the rate of delayed graft Frailty was assessed according to PFP scale [9] at the time of function (defined as the need of dialysis within the first week KT waiting-list evaluation. Median time from evaluation to list- after transplantation); the length of the first transplantation 2092 M.J. Pérez-Sáez et al. Table 2. Comparisons in clinical outcomes after kidney transplantation between different PFP classification strategies: (a) grouping 0–1 as a single category of robust patients, (b) considering PFP 0 patients as robust and grouping all patients with any frail PFP, and (c) comparing PFP 0 vs 1 subsets PFP 0–1 (n = 217) PFP ≥2 (n = 79) P-value Grouping PFP 0–1 as a single category of robust patients Delayed graft function (n, %) 56 (26.8) 27 (35.5) .151 Length of first inpatient (days, mean ± SD) 10.3 ± 7.8 12.3 ± 10.7 .219 Early readmission (≤90 days, n, %) 52 (24.0) 18 (22.8) .833 Surgical complications (n, %) 10 (8.9) 10 (12.6) .202 Opportunistic infections (n, %) 74 (34.1) 34 (43.0) .158 Acute rejection (n, %) 14 (6.5) 3 (3.8) .385 3-month serum creatinine (mg/dL, mean ± SD) 1.97 ± 1.12 1.79 ± 1.00 .252 1-year serum creatinine (mg/dL, mean ± SD) 1.94 ± 1.62 1.66 ± 1.26 .271 1-year graft loss (n, %) 23 (10.6) 17 (21.5) .020 1-year death-censored graft loss (n, %) 19 (8.7) 9 (11.3) .504 1-year patient mortality (n, %) 4 (1.8) 8 (10.1) .003 PFP 0 (n = 90) PFP ≥ 1 (n = 206) P-value Considering PFP 0 patients as robust and grouping any frail score Delayed graft function (n, %) 22 (25.3) 61 (30.8) .345 Length of first inpatient (days, mean ± SD) 10.1 ± 6.9 11.1 ± 9.4 .359 Early readmission (≤90 days, n, %) 20 (22.2) 50 (24.3) .703 Surgical complications (n, %) 3 (3.3) 16 (7.7) .704 Opportunistic infections (n, %) 30 (33.3) 78 (37.9) .456 Acute rejection (n, %) 6 (6.7) 11 (5.3) .652 3-month serum creatinine (mg/dL, mean ± SD) 2.06 ± 1.15 1.87 ± 1.07 .194 1-year serum creatinine (mg/dL, mean ± SD) 2.09 ± 1.69 1.79 ± 1.48 .197 1-year graft loss (n, %) 7 (7.7) 33 (16) <.001 1-year death-censored graft loss (n, %) 6 (6.6) 22 (10.7) <.001 1-year patient mortality (n, %) 1 (1.1) 11 (5.3) <.001 PFP 0 (n = 90) PFP 1 (n = 127) P-value Comparing patients PFP 0 vs 1 Delayed graft function (n, %) 22 (25.3) 34 (27.9) .678 Length of first inpatient (days, mean ± SD) 10.1 ± 6.9 10.4 ± 8.4 .773 Early readmission (≤90 days, n, %) 20 (22.2) 32 (25.2) .613 Surgical complications (n, %) 3 (3.3) 7 (5.5) .884 Opportunistic infections (n, %) 30 (33.3) 44 (34.6) .841 Acute rejection (n, %) 6 (6.7) 8 (6.3) .914 3-month serum creatinine (mg/dL, mean ± SD) 2.06 ± 1.15 1.91 ± 1.11 .366 1-year serum creatinine (mg/dL, mean ± SD) 2.09 ± 1.69 1.85 ± 1.58 .370 1-year graft loss (n, %) 7 (7.7) 16 (12.6) <.001 1-year death-censored graft loss (n, %) 6 (6.6) 13 (10.2) <.001 1-year patient mortality (n, %) 1 (1.1) 3 (2.4) .019 PFP, Physical Frailty Phenotype; SD, standard deviation. inpatient stay; the rate of early readmission to the hospital present, Chi-square test was also used to compare categorical (defined as admissions within 90 first days after transplant variables, ANOVA test to compare quantitative variables with date); surgical complications (including wound infection and/or normal distribution, and Kruskal–Wallis test for quantitative dehiscence, lymphocele, vascular complications, ureteral variables without normal distribution. Patient and graft survival stenosis and ureteral leak); opportunistic infections (including were estimated using Kaplan–Meier curves, applying the log- cytomegalovirus, BK virus and others); biopsy-proven acute re- rank test. Among those KT recipients who had 0 or 1 PFP criteria, jection; graft function (defined as serum creatinine at 3 months univariable and multivariable Cox regressions were performed and 1 year after transplantation); and both graft survival to estimate the hazard ratio (HR) and confidence intervals (CI) for (including death-censored graft survival) and patient survival. patient survival. Clinical variables who resulted statistically sig- nificant in the univariable analysis were included in the multi- variable analysis. Statistical analysis was performed using SPSS Statistics version 27 software (IBM, Armonk, NY, USA). P-values < .05 were Continuous variables were expressed as mean ± standard de- considered statistically significant. viation (SD), or median and IQR, according to normal distri- bution. Categorical data were expressed as absolute numbers and percentages. Comparisons of baseline characteristics be- RESULTS tween two groups were made using Chi-square or Fisher’s exact tests to analyze categorical variables, Student’s t-test for con- From a total of 296 KT recipients, 90 (30.4%) had 0 PFP crite- tinuous variables with normal distribution and Mann–Whitney ria at the time of KT waiting-list admission, 127 (42.9%) had 1 test for non-parametric variables. When three categories were criterion, 57 (19.3%) had 2 criteria, 16 (5.4%) had 3 criteria and Mortality in mildly frail KT recipients 2093 A age was not significantly different between groups. Female sex was progressively more frequent among those recipients with advanced stages of frailty (18.9% among patients with 0 crite- ria and 50.6% among those with ≥2 PFP criteria). Cerebrovas- cular disease was more frequent among recipients scoring ≥2 (15.2% vs 8.9% among patients with 0 criteria). Preemptive KT was a less frequent option among recipients scoring ≥2 (6.3% vs 17.8% among patients with 0 criteria). No differences were found in terms of time on dialysis prior to transplant, type of donor or donor age between groups. Thymoglobulin was more frequently used in those recipients with ≥2 PFP criteria. Follow-up was sim- ilar between groups. Regarding transplant outcomes, Table 2 shows results ac- cording with three different PFP classifications: (i) merging 0 and 1 criteria and considering both as robust patients (n = 217, 73.3%) vs ≥2 as pre-frail or frail (n = 79, 26.7%); (ii) considering only 0 cri- teria for robust patients (n = 90, 30.4%) and comparing with ≥1 as pre-frail or frail (n = 206, 69.6%); and (iii) considering 0 criteria for robust patients (n = 90, 30.4%) and comparing with the subset of B KT recipients who scored 1 (n = 127, 42.9%). In all the three classi- fication models, no differences were found in delayed graft func- tion rate, length of first inpatient stay, early readmission to the hospital, surgical complications, opportunistic infections, acute rejection rate, or graft function at both 3 and 12 months after transplantation. However, univariable analysis showed that graft and patient survival were different between groups, although death-censored graft survival was similar. Figure 1 represents the Kaplan–Meier survival curves of patient death (Fig. 1A), graft loss (Fig. 1B) and death-censored graft loss (Fig. 1C) according to PFP criteria. To disaggregate data about patient survival, a subset of pa- tients who had 0 PFP criteria (n = 90) and 1 PFP criteria (n = 217) were considered. Table 3 shows differences among those recipi- ents who were alive and those who were deceased at the end of follow-up. The multivariable analysis reveals that recipient age (HR 1.1 per year), the presence of peripheral vascular disease (HR 9.0) and having 1 PFP criterion (HR 3.52) were independent fac- C tors for patient mortality (Table 4). DISCUSSION In this study, we present transplant outcomes in a cohort of KT recipients according to different PFP classifications. Non- frail patients ranged from 30.4% to 73.3% if only those who had none PFP criteria vs those who had 0 or 1 were considered. Both graft and patient survival diverged among groups, being poorer among pre-frail and frail patients. Presenting with one PFP cri- terion conferred higher rate of patient death than having none. Frailty has been well established as a risk factor for poorer outcomes after transplantation [1–6]. Most of the studies carried out in KT recipients have used the Fried phenotype [9] to eval- uate frailty, considering as frail those patients who have ≥3 cri- teria [7]. However, intermediate frail status, originally defined as the presence of 1 or 2 PFP criteria, has been modified in renal population to the presence of 2 criteria, leaving 0 and 1 criteria to define non-frail patients. These thresholds have been recently FIGURE 1: Kaplan-Meier survival curves of (A) patient death, (B) graft loss and (C) death-censored graft-loss among 296 KT recipients according to PFP score. described as ‘specific’ to the end-stage renal disease population [19], but no studies have evaluated whether patients with 0 cri- 6 (2%) had 4 criteria. No patient presented with 5 criteria. Overall, teria vs 1 criterion behave the same after transplantation. 79 (26.7%) had ≥2 criteria. The PFP criteria distribution of those In the general population, the original definition of pre- recipients who presented with 1 PFP criterion was as follows: frailty has been associated with an increased risk of cardiovas- 66.1% weakness, 15.7% weight loss, 15% exhaustion, 1.6% slow cular events [11], as well as disability, low quality of life and walking speed and 0.8% low physical activity. Table 1 displays the mortality [10, 12]. In the KT scenario, several studies have ex- baseline characteristics of the cohort based on PFP criteria. Mean plored the relationship between the excess risk of mortality after 2094 M.J. Pérez-Sáez et al. Table 3. Comparison among patients who survived (n = 202) and those who died (n = 15) at the end of the follow-up (considering only those with 0 or 1 PFP criteria, n = 217) Alive (n = 202) Deceased (n = 15) P-value Age at transplantation (years, mean ± SD) 60.4 ± 12.7 69.9 ± 7.2 <.001 Sex (female, n, %) 44 (21.8) 3 (20) .585 Hypertension (n, %) 195 (96.5) 15 (100) .610 Diabetes mellitus (n, %) 62 (30.7) 7 (46.7) .160 Coronary artery disease (n, %) 26 (12.9) 4 (26.7) .135 Heart failure (n, %) 5 (2.5) 1 (6.7) .353 Peripheral vascular disease (n, %) 12 (5.9) 5 (33.3) .003 Cerebrovascular disease (n, %) 8 (4) 2 (13.3) .145 PFP = 1 114 (56.4) 13 (86.6) .018 Time on dialysis prior to KT (months, mean ± SD) 24 ± 22 22 ± 16 .787 Thymoglobulin induction (n, %) 19 (9.5) 2 (13.3) .438 Tacrolimus C0 /D during first year after transplant (ng/mL/mg, mean ± SD)a 1.44 ± 0.57 1.52 ± 0.43 .490 SD, standard deviation; PFP, Physical Frailty Phenotype; KT, kidney transplantation; C0 /D, concentration/dose. a C0 /D ratio = tacrolimus trough concentration (ng/mL)/daily dose (mg). Table 4. Univariable and multivariable analysis of risk factors for patient mortality after transplantation (considering only those with 0 or 1 PFP criteria, n = 217) Univariable Multivariable HR 95% CI P-value HR 95% CI P-value Age at transplantation (years) 1.09 1.04–1.14 <.001 1.10 1.03–1.17 .002 Sex (ref: female) 1.49 0.69–3.23 .302 Hypertension (n, %) 1.70 0.22–12.6 .602 Diabetes mellitus (n, %) 2.51 0.90–5.69 .070 Coronary artery disease (n, %) 1.72 0.64–4.57 .275 Heart failure (n, %) 1.29 0.17–9.5 .801 Peripheral vascular disease (n, %) 3.36 1.35–8.35 .009 9.0 2.80–29.4 <.001 PFP = 1 (ref: 0) 4.49 1.01–19.9 .048 3.52 1.03–15.9 .048 Time on dialysis prior to KT (months) 1.00 0.98–1.02 .800 1-year serum creatine 1.29 0.70–2.38 .405 Tacrolimus C0 /D ratio during first year after transplanta 1.39 0.72–2.68 .313 HR, hazard ratio; CI, confidence interval; PFP, Physical Frailty Phenotype; KT, kidney transplantation; C0 /D ratio, concentration/dose. a C0 /D ratio = tacrolimus trough concentration (ng/mL)/daily dose (mg). transplantation and the recipient frailty status [4–6]. These stud- among these pre-frail patients (10.1% vs 1.8%). Patients scoring ies considered as pre-frail patients those who presented with for 2 or more PFP domains were at high risk of a premature two PFP domains, merging 0 and 1 criteria to define robust pa- death after transplantation. On the other hand, when consid- tients. One study found that frailty increased 2-fold the risk of ering original Fried classification for robustness (0 criteria) and death after KT [4], while adjusted models were unable to show pre-frailty or frailty (≥1 criteria), we also observed differences an association between pre-frailty and mortality. Patients who in early patient mortality (higher in pre-frail or frail recipients), had 1 PFP criterion were not analyzed in detail in these studies, as well as a higher rate of global graft loss during the first as they were considered as robust, and therefore they served as year after transplant. Finally, we decided to analyze whether the control group. they were any differences in terms of transplant outcomes In our cohort, only 30.4% of the recipients accounted for 0 between those patients who had been merged in previous PFP criteria and they may be considered as non-frail. In addition, studies from other groups as non-frail patients (0 criteria vs 1 we found a high percentage of patients (42.9%) scoring one PFP criterion). Patients with 1 PFP criterion presented with higher criterion. This PFP 1 subgroup would have been considered 1-year patient mortality (2.4% vs 1.1%) and higher 1-year rate of as non-frail in some studies, or pre-frail in others. Aiming to death-censored graft loss (10.2% vs 6.6%) than patients with 0 disaggregate transplant results in this group of robust vs pre- criteria, and the multivariable analysis confirmed that having frail patients, we analyzed clinical outcomes from different PFP 1 PFP criterion increased almost 4-fold the risk of death after classification perspectives. Although minor clinical outcomes transplantation. As the Kaplan–Meier survival curves show, were similar among patients with 0, 1 or ≥2 criteria, significant these patients do not seem to die as early as those who were differences were found in terms of both graft and patient frailer, but still, their risk of death is higher than the ones who survival. When considering the PFP ‘renal’ classification, differ- did not score for any PFP criteria before transplantation. ences in early patient mortality (within the first 12 months after Our study has the inherent limitations that an observational transplantation) were found among robust (0–1 criteria) and study implies and probably the sample size is not large enough pre-frail (≥2 criteria) patients, which was significantly higher to show statistically significant results in terms of other clinical Mortality in mildly frail KT recipients 2095 outcomes after transplantation. However, to our knowledge, this Medical Care), Laura Ribera, Margarita Guino (Glories Hemodial- is the first study that explores the role of having one Fried cri- ysis Center, Fresenius Medical Care), Ramón Roca, Jordi Calls, terion in transplant outcomes and presents the association be- Alicia Rovira (Department of Nephrology, Hospital de Mollet), tween different PFP scores and these outcomes. Giving the large Josep Mora, Omar Ibrik, Florentina Liria (Granollers Hemodial- number of frailty scales, it is important to validate them in the ysis Center, Fresenius Medical Care), Thaïs López, Jaume Almi- renal population, clarifying their predictive value for bad out- rall, Carmen Moya (Department of Nephrology, Hospital Parc comes [20] or analyzing the clinical significance of their defini- Taulí), Fátima Moreno, Manel Ramírez de Arellano, Sandra Ru- tion for frailty. In addition, identifying pre-frail or frail patients bio (Department of Nephrology, Consorci Sanitari de Terrassa), may allow clinicians to establish proper strategies to revert the Ignacio Cidraque, Carlota Pájaro (Cetirsa Terrassa Hemodialysis situation [21]. Center, Fresenius Medical Care), Núria Garra, Josep Galcerán, Ma- In accordance with the original PFP scale [9], being pre-frail rina Fenollar (Department of Nephrology, Hospital de Manresa), has an impact on both graft and patient survival, and it starts Sara Outón, Fabiola Dapena, Josep Jara (Department of Nephrol- with the presence of one Fried criterion. More studies are needed ogy, Consorci Sanitari del Garraf), Rosa García, Mònica Manresa to corroborate these results and establish proper risks after (Department of Nephrology, Hospital de Palamós). transplantation according to frailty status before transplant. REFERENCES ACKNOWLEDGEMENTS 1. Quint EE, Zogaj D, Banning LBD et al. Frailty and kid- The authors of this study appreciate the contribution of all the ney transplantation: a systematic review and meta- members of the FRAIL-MAR Study Group. analysis. Transplant Direct 2021; 7: e701; doi:10.1097/ TXD.0000000000001156 2. Goldfarb DA. Re: frailty and early hospital readmission after FUNDING kidney transplantation. J Urol 2014; 191: 1366–1367 3. McAdams-Demarco MA, Law A, Tan J et al. Frailty, mycophe- Frail-MAR project is currently supported by a FIS-FEDER grant nolate reduction, and graft loss in kidney transplant recipi- PI19/00037 Instituto de Salud Carlos III (ISCIII) a ‘Proyecto Es- ents. Transplantation 2015; 99: 805–810 trella de Mejora de la Calidad’, Parc de Salut Mar, Barcelona, 4. McAdams-Demarco MA, Law A, King E et al. Frailty and mor- Spain. M.J.P.-S. has received a grant from the Spanish Society of tality in kidney transplant recipients. Am J Transplant 2015; Transplant. A.Buxeda has support from a Rio Hortega contract 15: 149–154 CM19/00004, ISCIII. 5. McAdams-DeMarco MA, Ying H, Olorundare I et al. Individ- ual frailty components and mortality in kidney transplant recipients. Transplantation 2017; 101: 2126–2132 DATA AVAILABILITY STATEMENT 6. McAdams-Demarco MA, King EA, Luo X et al. Frailty, length The data that support the findings of this study are available on of stay, and mortality in kidney transplant recipients. Ann request from the corresponding author. Surg 2017; 266: 1084–1090 7. Harhay MN, Rao MK, Woodside KJ et al. An overview of frailty in kidney transplantation: measurement, management and CONFLICT OF INTEREST STATEMENT future considerations. Nephrol Dial Transplant 2020; 35: 1099– 1112 The authors declare no conflict of interest. 8. Pérez-Sáez MJ, Gutiérrez-Dalmau Á, Moreso F et al. Frailty and kidney transplant candidates. Nefrologia 2021; 41: 237– 243 APPENDIX 9. Fried LP, Tangen CM, Walston J et al. Frailty in older adults: FRAIL-MAR Study Group members: evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001; María José Pérez-Sáez, Carlos E. Arias-Cabrales, Dolores 56: M146–M156 Redondo, Francesc Barbosa, Higini Cao, Silvia Collado, Anna 10. Hanlon P, Nicholl BI, Jani BD et al. Frailty and pre-frailty in Buxeda, Carla Burballa, Marta Crespo, Julio Pascual, Anna Faura, middle-aged and older adults and its association with mul- María Vera, Anna Bach, Guillermo Pedreira, Ernestina Jun- timorbidity and mortality: a prospective analysis of 493 737 yent, Montserrat Folgueiras, Yolanda Castillo, Aida Martínez, UK Biobank participants. Lancet Public Heal 2018; 3: e323– Marisol Fernández, Eva Barbero, Rosa Causadías (Department e332 of Nephrology, Hospital del Mar), Alicia Calvo (Department 11. Sergi G, Veronese N, Fontana L et al. Pre-frailty and risk of Cardiology, Hospital del Mar), Jesús Carazo (Department of of cardiovascular disease in elderly men and women: the Anesthesiology, Hospital del Mar), Albert Frances, Lluis Cec- Pro.V.A. study. J Am Coll Cardiol 2015; 65: 976–983 chini (Department of Urology, Hospital del Mar), Ester Marco, 12. Feng L, Zin Nyunt MS, Gao Q et al. Cognitive frailty and ad- Delky Meza de Valderrama, Andrea Morgado, Elena Muñoz verse health outcomes: findings from the Singapore Longi- (Department of Rehabilitation and Physical Medicine, Hospital tudinal Ageing Studies (SLAS). J Am Med Dir Assoc 2017; 18: del Mar), Xavier Nogués (Department of Internal Medicine, Hos- 252–258 pital del Mar), Leocadio Rodríguez-Mañas (Department of Geri- 13. McAdams-DeMarco MA, Olorundare IO, Ying H et al. Frailty atrics, Hospital Universitario de Getafe, Madrid), Olga Vázquez and postkidney transplant health-related quality of life. (Department of Geriatrics, Hospital del Mar), María Dolores Muns Transplantation 2018; 102: 291–299 (Dietary Unit, Department of Endocrinology and Nutrition, Hos- 14. McAdams-Demarco MA, Law A, Salter ML et al. Frailty as a pital del Mar), Miguel Gárriz, María Polo Gómez (Psychology De- novel predictor of mortality and hospitalization in individ- partment, Neuropsychiatric Institute, Hospital del Mar), Sara uals of all ages undergoing hemodialysis. J Am Geriatr Soc Hurtado, Maite López (Diagonal Hemodialysis Center, Fresenius 2013; 61: 896–901 2096 M.J. Pérez-Sáez et al. 15. McAdams-Demarco MA, Suresh S, Law A et al. Frailty and 19. McAdams-DeMarco MA, Chu NM, Segev DL. Frailty and long- falls among adult patients undergoing chronic hemodialy- term post-kidney transplant outcomes. Curr Transplant Rep sis: a prospective cohort study. BMC Nephrol 2013; 14: 224 2019; 6: 45–51 16. Pérez Fernández M, Martínez Miguel P, Ying H et al. Comor- 20. Pérez-Sáez MJ, Dávalos-Yerovi V, Redondo-Pachón D et al. bidity, frailty, and waitlist mortality among kidney trans- Frailty in kidney transplant candidates: a comparison be- plant candidates of all ages. Am J Nephrol 2019; 49: 103–110 tween physical frailty phenotype and FRAIL scales. J Nephrol 17. Pérez-Sáez MJ, Arias-Cabrales CE, Dávalos-Yerovi V et al. Published online January 3, 2022; doi: 10.1007/s40620-021- Frailty among chronic kidney disease patients on the kidney 01234-4 transplant waiting list: the sex-frailty paradox. Clin Kidney J 21. Pérez-Sáez MJ, Morgado-Pérez A, Faura A et al. The FRAILMar 2021; 15: 109–118 study protocol: frailty in patients with advanced chronic 18. Pérez-Sáez MJ, Redondo-Pachón D, Arias-Cabrales CE et al. kidney disease awaiting kidney transplantation. A random- Outcomes of frail patients while waiting for kidney trans- ized clinical trial of multimodal prehabilitation. Front Med plantation: differences between physical frailty phenotype 2021; 8: 675049 and FRAIL scale. J Clin Med 2022; 11: 672