Evolving spectrum but persistent high mortality of COVID-19 among patients on kidney replacement therapy in the vaccine era: the Spanish COVID-19 KRT Registry

ABSTRACT

Background

Kidney replacement therapy (KRT) conferred a high risk for coronavirus disease 2019 (COVID-19) related mortality early in the pandemic. We evaluate the presentation, treatment and outcomes of COVID-19 in patients on KRT over time during the pandemic.

Methods

This registry-based study involved 6080 dialysis and kidney transplant (KT) patients with COVID-19, representing roughly 10% of total Spanish KRT patients. Epidemiology, comorbidity, infection, vaccine status and treatment data were recorded, and predictors of hospital admission, intensive care unit (ICU) admission and mortality were evaluated.

Results

Vaccine introduction decreased the number of COVID-19 cases from 1747 to 280 per wave. Of 3856 (64%) COVID-19 KRT patients admitted to the hospital, 1481/3856 (38%) were admitted during the first of six waves. Independent predictors for admission included KT and the first wave. During follow-up, 1207 patients (21%) died, 500/1207 (41%) during the first wave. Among vaccinated patients, mortality was 19%, mostly affecting KT recipients. Overall, independent predictors for mortality were older age, disease severity (lymphopaenia, pneumonia) and ICU rejection. Among patient factors, older age, male sex, diabetes, KT and no angiotensin receptor blockers (ARB) were independent predictors of death. In KT recipients, individual immunosuppressants were independent predictors of death. Over time, patient characteristics evolved and in later pandemic waves, COVID-19 was mainly diagnosed in vaccinated KT recipients; in the few unvaccinated dialysis patients, ICU admissions increased and mortality decreased (28% for the first wave and 16–22% thereafter).

Conclusions

The clinical presentation and outcomes of COVID-19 during the first wave no longer represent COVID-19 in KRT patients, as the pandemic has become centred around vaccinated KT recipients. Vaccines lowered the incidence of diagnosed COVID-19 and mortality. However, mortality remains high despite increased access to ICU care.

Graphical Abstract

Graphical Abstract.

INTRODUCTION

Patients with chronic kidney disease (CKD) and, specifically, those needing kidney replacement therapy (KRT) were among those having the highest mortality during the first wave of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, a time when healthcare systems were frequently overwhelmed, limiting access to intensive care units (ICU), empirical therapy was later found to be ineffective and vaccination was not available [1–3]. However, dynamic factors such as new virus variants, decreased saturation of healthcare systems, approval of effective treatments and more importantly, SARS-CoV-2 vaccines may have modified the nature and outcomes of coronavirus disease 2019 (COVID-19) in patients on KRT [4].

Specific characteristics of KRT patients may have modified SARS-CoV-2 exposure and outcomes differentially from the general population over time through the pandemic [5]. On one hand, KRT patients present multiple comorbidities and poly-medication that may increase fragility and haemodialysis (HD) patients have limited ability to shield due to their need to access healthcare facilities thrice weekly [6, 7]. On the other hand, KRT patients have been excluded from the prescription of certain drugs and even from mechanical ventilation or ICU admission because of their worse prognosis, especially when healthcare systems were overwhelmed [8]. Finally, the response of kidney transplant recipients (KT recipients) to vaccines may be suboptimal, at least from the point of view of antibody development [9]. Only an analysis of presentation and outcomes of COVID-19 in KRT patients over the successive pandemic waves may identify persistent risk factors for adverse outcomes as well as characterizing the current prognosis of patients with CKD on KRT.

The aim of the present study was to evaluate the dynamic presentation features and outcomes across the six waves of the SARS-CoV-2 pandemic in Spain in a registry-based population of patients on KRT. As Spain was one of the earliest and hardest-hit countries and currently is one of the countries with the highest uptake of vaccines (80% of the total population on 29 December 2021), changes over time in Spain may be of special interest to understand the interaction between SARS-CoV-2 and healthcare measures.

MATERIALS AND METHODS

The Spanish COVID-19 KRT Registry is an initiative of the Spanish Society of Nephrology open to all nephrology and dialysis centres in Spain. It was set up on 18 March 2020 with a prospective design. It includes all patients treated with in-centre HD, home HD (HHD), peritoneal dialysis (PD) or kidney transplant recipients (KT recipients) that have developed a confirmed SARS-Cov-2 infection. Diagnosis of COVID-19 was based on the positivity of one of the following tests: reverse transcription-polymerase chain reaction (RT-PCR), rapid antigen tests or SARS-CoV-2 anti-spike antibodies before vaccination.

This analysis included the full cohort until 12 December 2021. During follow-up, local coordinators were asked to immediately report cases from their centres through a centralized web page (www.senefro.org) only accessible to registered users. Until now, the Spanish Public Health Institute has defined six pandemic waves [10], according to the following dates: first wave from 31 January 2020, second wave from 1 July 2020, third wave from 1 December 2020, fourth wave from 1 March 2021, fifth wave from 1 June 2021 and sixth wave from 1 November 2021. These dates were used for the present analysis.

Variables

At baseline, epidemiological factors (sex and age), KRT (KT, HD, PD and HHD), KRT vintage and cause of CKD, concomitant treatments [angiotensin-converting-enzyme inhibitors (ACEi), angiotensin receptor blockers (ARB) and non-steroidal anti-inflammatory drugs (NSAID)] and immunosuppressive drugs were recorded.

At diagnosis, symptoms, lymphopaenia (total lymphocyte count <500/mm³) and pneumonia were registered. Prescribed medications for SARS-CoV-2 infection and, in KT, the need for adjustment of immunosuppressants were also collected. Invasive or non-invasive mechanical ventilation was registered.

Vaccination

In Spain, the SARS-CoV-2 vaccination started on 27 December 2020. The following four vaccines have been used: BNT162b2 (Pfizer-BioNTech®), mRNA-1273 (Moderna®), ChAdOx1-S (AstraZeneca®) or Ad26.COV.2 (Janssen®) [9]. Prescription of individual vaccines was guided by local public health authorities and mainly driven initially by vaccine availability and later by successive modifications of the age range for ChAdOx1-S. Vaccination status was collected during the present study and considered when the full vaccination schedule was completed before SARS-CoV-2 infection. The course of the first full (i.e. two doses except for Ad26.COV.2) vaccine schedule was uneven, as decisions were taken at the regional level and not all regions assigned the same priority to patients on KRT or to the different KRT modalities. However, for KRT patients, it was generally completed in the first semester of 2021 [9]. The booster dose was generally administered to KRT patients from October to early November 2021, i.e., before the sixth wave.

Outcomes

During follow-up, the following outcomes were registered in patients infected by SARS-CoV-2: admission to hospital, intensive care unit (ICU) admission and mortality. Investigators could report that patients had been rejected for ICU admission.

Ethics

The registry was approved by the Regional Ethical Committee of Asturias.

Statistics

Data are expressed as mean (standard deviation) or median [interquartile range (IQR)] depending on the distribution of the variables, tested with the Shapiro–Wilk test. Categorical variables were compared using the Fisher test and continuous variables with t-test or Mann–Whitney, according to the variable distribution. For comparison of continuous variables from more than two groups, ANOVA or Kruskal–Wallis tests were used.

Univariate logistic regressions were performed for assessing associations between the outcomes (admission, ICU admission and mortality) and the registered variables. To identify factors associated with admission, ICU admission and mortality differentially over time, two different analyses were performed. The first analysis compared the first with subsequent waves. The rationale is that the first wave differed from other waves in several aspects: the healthcare system was overwhelmed, health authorities were advising against the widespread use of face masks, and there was use of potentially toxic ineffective drugs and a lack of awareness about the need to prevent thrombosis. The second analysis compared the first three waves with the subsequent waves. The rationale is to compare the waves before and after vaccine availability.

Multivariate regression models were adjusted for variables with P < .1 in univariate analyses and those considered confounders.

The statistical analysis was performed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA).

RESULTS

Baseline characteristics

A total of 6080 KRT patients with SARS-CoV-2 infection were included (36% female, age 63 ± 15 years). Among them, 2832 (46%) were KT recipients, 3063 (50%) were on in-centre HD, 173 (3%) on PD and 22 (<1%) on HHD. Prior to infection, 501 (8%) patients had received a SARS-CoV-2 vaccine (Table 1). Baseline characteristics for the successive pandemic waves are shown in Table 1. SARS-CoV-2 vaccination was common among those infected in the fifth and sixth waves. Differences between KRT modalities are shown in Supplementary data, Table S1. SARS-CoV-2-infected KT recipients were younger but had longer KRT vintage than HD patients and also had a higher rate of SARS-CoV-2 vaccination.

Table 1.

Baseline characteristics, infection related variables and outcomes in different waves

	Total (n = 6080)	First (n = 1914)	Second (n = 1930)	Third (n = 1397)	Fourth (n = 330)	Fifth (n = 471)	Sixth (n = 38)	P
Sex (male), n (%)	3901 (64)	1236 (64)	1226 (63)	901 (65)	204 (62)	312 (66)	22 (58)	.719
Age (years)	63 ± 15	67 ± 14	62 ± 15	64 ± 15	60 ± 14	58 ± 15	60 ± 15	<.001
Diabetic kidney disease, n (%)	1199 (21)	408 (25)	378 (21)	282 (21)	56 (17)	72 (16)	3 (8)	<.001
KRT, n (%):    KT    HD    PD    HHD	2831 (47)   3054 (50)   173 (3)   22 (<1)	648 (34)   1198 (63)   62 (3)   6 (<1)	922 (48)   946 (49)   52 (3)   10 (<1)	644 (46)   703 (5)   46 (3)   4 (<1)	217 (66)   108 (33)   5 (1)   0 (0)	369 (78)   93 (20)   7 (1)   2 (0)	31 (81)   6 (16)   1 (3)   0 (0)	<.001
Haemodialysis unit location (external), n (%)	1460 (53)	488 (48)	462 (52)	392 (58)	55 (52)	61 (68)	2 (40)	<.001
KRT vintage (months)	44 (17–95)	38 (15–83)	45 (17–97)	45 (18–97)	48 (18–110)	53 (19–111)	60 (61–101)	<.001
Symptoms at onset, n (%)   Asymptomatic   Fever   Cough/rhinorrhoea   Dyspnoea   Gastrointestinal	1225 (22)   3525 (60)   3512 (60)   2196 (39)   1279 (23)	259 (15)   1300 (71)   1161 (63)   742 (43)   424 (24)	498 (28)   1037 (56)   1004 (54)   629 (35)   419 (23)	339 (26)   677 (50)   792 (59)   492 (38)   236 (18)	58 (19)   188 (59)   212 (67)   130 (42)   71 (23)	68 (16)   298 (66)   319 (71)   193 (44)   120 (27)	3 (9)   25 (68)   24 (69)   10 (29)   9 (26)	<.001   <.001   <.001   <.001   <.001
Lymphopenia, n (%)	3584 (63)	1358 (75)	1037 (57)	725 (56)	196 (63)	257 (58)	11 (38)	<.001
Pneumonia, n (%)	3148 (53)	1260 (68)	843 (45)	616 (45)	179 (56)	238 (52)	12 (39)	<.001
Hospital admission, n (%)	3856 (64)	1481 (78)	1104 (58)	757 (55)	207 (64)	292 (63)	15 (47)	<.001
ICU admission, n (%)a	561 (15)	127 (9)	179 (17)	123 (17)	39 (19)	90 (32)	3 (21)	<.001
ICU rejection, n (%)a	431 (12)	208 (16)	77 (7)	109 (16)	15 (8)	19 (7)	1 (8)	<.001
Mechanical ventilation, n (%)a	656 (19)	210 (17)	170 (17)	143 (22)	46 (25)	84 (31)	3 (23)	<.001
Treatment at onset, n (%)   lopinavir/ritonavir   Hydroxychloroquine   Interferon   Tocilizumab   Steroids   No treatment	611 (12)   1425 (27)   60 (4)   468 (9)   2565 (48)   1740 (35)	596 (35)   1419 (80)   5 (<1)   172 (11)   587 (35)   253 (16)	9 (1)   5 (<1)   0 (0)   132 (8)   862 (50)   705 (44)	1 (<1)   1 (<1)   0 (0)   79 (7)   661 (53)   539 (45)	0 (0)   0 (0)   0 (0)   33 (11)   183 (61)   92 (32)	5 (1)   0 (0)   0 (0)   52 (14)   261 (63)   135 (34)	0 (0)   0 (0)   0 (0)   0 (0)   11 (41)   16 (57)	<.001   <.001   <.001   <.001   <.001   <.001
Baseline ACEi, n (%)	768 (13)	231 (12)	246 (13)	158 (12)	49 (15)	80 (18)	4 (11)	.014
Baseline ARB, n (%)	1343 (23)	384 (21)	455 (24)	317 (24)	72 (23)	106 (24)	9 (25)	.181
Previous NSAID, n (%)	181 (3)	63 (3)	61 (3)	43 (3)	6 (2)	7 (2)	1 (3)	.326
Immunosuppression, n (%)   Prednisone   Ciclosporin   Tacrolimus   Mycophenolate   mTORi   Azathioprine	2268 (81)   3 (<1)   2449 (95)   1945 (76)   412 (16)   29 (1)	461 (74)   0 (0)   511 (92)   382 (69)   124 (22)   8 (1)	745 (82)   0 (0)   814 (96)   654 (77)   114 (13)   11 (1)	531 (83)   2 (<1)   564 (96)   446 (76)   93 (16)   3 (<1)	181 (83)   1 (<1)   198 (97)   156 (76)   35 (17)   3 (1)	323 (89)   0 (0)   333 (96)   281 (81)   44 (13)   3 (1)	27 (90)   0 (0)   29 (100)   23 (79)   2 (7)   1 (3)	<.001   .219   .006   .001   <.001   .470
SARS-CoV-2 vaccination, n (%)	500 (8%)	0 (0)	0 (0)	8 (1)	66 (20)	396 (84)	30 (79)	<.001
Immunosuppressants adjusted, n (%)	458 (65)	13 (72)	16 (42)	78 (60)	88 (69)	246 (67)	17 (81)	.015
Mortality, n (%)	1207 (21)	500 (28)	292 (16)	265 (20)	52 (17)	94 (22)	4 (19)	<.001
Mortality in vaccinated, n (%)	86 (19)	NA	NA	1 (12)	8 (12)	75 (21)	2 (15)	.382
Mortality in symptomatic patients, n (%)	1010 (25)	412 (30)	244 (20)	225 (25)	46 (20)	80 (25)	3 (19)	<.001

^a% of admitted.

Data are displayed as mean ± standard deviation or median (interquartile range) unless otherwise indicated. KRT; kidney replacement therapy, KT, kidney transplant; HD, in-centre haemodialysis; PD, peritoneal dialysis; HHD, home haemodialysis; ICU, intensive care unit; ACEi, angiotensin-converting enzyme inhibitors; ARB: angiotensin receptor blockers; NSAID, non-steroidal anti-inflammatory drugs; mTORi, mammalian target of rapamycin inhibitors; NA, not applicable.

Regarding immunosuppressive drugs, KT recipients were mostly receiving prednisone (81%), tacrolimus (95%) and mycophenolate (76%) for chronic immunosuppression before developing COVID-19 (Table 1). The chronic immunosuppressive medication of KT recipients that were diagnosed with COVID-19 evolved during the pandemic: over time the percentage of patients on prednisone (P < .001), tacrolimus (P = .006) and mycophenolate (P = .001) increased and the percentage of patients on mTORi inhibitors (P < .001) decreased. Most of these differences were evident between the first and subsequent waves (Supplementary data, Table S2). However, only prednisone was significantly more frequent for chronic immune suppression during the vaccine era (Supplementary data, Table S2). In fact, chronic immune suppression with prednisone was observed in 80% of KT recipients before vaccines became available (waves 1–3) and in 88% in waves 4–5, when over 85% of KT recipients with SARS-CoV-2 infection were vaccinated.

Changing patterns of SARS-CoV-2 infection and treatment

Figure 1 shows the incidence of SARS-CoV-2 infection in the Spanish COVID-19 KRT Registry and in the general Spanish population [3]. In the KRT cohort, more SARS-CoV-2 infections occurred during the first wave than in subsequent waves (P < .001), unlike in the general population. The pandemic affected more frequently in-centre HD patients in the initial wave and KT recipients from the fourth wave, i.e. after the vaccine had become available (Fig. 2) (Supplementary data, Tables S2 and S3). Following the availability of vaccines, the number of COVID-19 cases among KRT patients decreased dramatically. However, an interesting pattern is observed in which most COVID-19 cases in KT recipients in the fifth and sixth waves occurred among vaccinated persons, while in the sixth wave, i.e. after the booster had been offered, most infections in HD occurred in unvaccinated patients.

FIGURE 1:

SARS-CoV-2 infections over time in the COVID-19 Spanish KRT Registry and in the general Spanish population during the pandemic.

FIGURE 2:

SARS-CoV-2 cases and vaccination status among kidney transplant recipients and haemodialysis. Continuous lines refer to total cases in the different cohorts. Bars indicate the percentage of vaccination among infected patients. KT, kidney transplant; HD, haemodialysis.

Symptoms varied across waves. A total of 1025 (22%) patients were asymptomatic and the frequency increased from the first wave (15%) to the second and third wave (28% and 26%, respectively), likely representing changes in HD units screening patterns, and then decreased to 9% (P < .001). In this regard, KT recipients and patients on PD were less likely to be asymptomatic (P < .001) and had higher rates of pneumonia (P < .001) (Supplementary data, Table S1). A total of 3148 (53%) patients developed pneumonia (Table 1). The most common symptoms were fever (60%) and cough/rhinorrhoea (60%).

The prescription of drugs to treat COVID-19 changed during the pandemic. Lopinavir/ritonavir (35%) and hydroxychloroquine (80%) were used only in the first wave and then became obsolete (P < .001 for both). Steroids (48%) were the most frequent treatment initiated for the cytokine storm phase of COVID-19, while 35% of patients were untreated (Table 1). KT recipients were treated more frequently with drugs targeting the cytokine storm (i.e. steroids or tocilizumab) than the other cohorts (P < .001 for both) (Supplementary data, Table S1).

Among KT recipients, 458 (65%) required an adjustment of immunosuppressive treatment during the infection, the most common being a reduction or withdrawal of mycophenolate in 384 (98%), followed by a reduction in calcineurin inhibitors [tacrolimus in 232 (68%) patients, cyclosporine in 5 (20%)] and an increase in steroid dose in 289 (82%) patients. A decrease/withdrawal of mTOR inhibitors was uncommon (3 patients, 5%).

Factors associated with admission

Among infected patients, 3856 (64%) were admitted to hospital. Factors associated to admission are summarized in Table 2.

Table 2.

Factors associated with outcomes after SARS-CoV-2 infection in univariate analysis

	Hospital admission		ICU admission		Mortality
	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P
Sex (male)	1.30 (1.17–1.45)	<.001	1.13 (0.93–1.36)	.226	1.25 (1.10–1.43)	.001
Age (per year)	1.02 (1.02–1.03)	<.001	0.97 (0.96–0.97)	<.001	1.06 (1.05–1.06)	.001
Diabetic kidney disease (yes)	1.39 (1.21–1.59)	<.001	0.82 (0.65–1.03)	.093	1.43 (1.23–1.66)	<.001
KRT (KT)	1.15 (1.03–1.28)	.010	3.75 (3.07–4.59)	<.001	0.73 (0.64–0.83)	<.001
Haemodialysis unit location (in-centre)	0.76 (0.65–0.89)	.001	0.98 (0.68–1.42)	.932	1.02 (1.00–1.44)	.045
KRT vintage (per month)	1.00 (1.00–1.00)	.331	1.00 (1.00–1.00)	.055	1.00 (0.99–1.00)	.871
Lymphopaenia (yes)	15.1 (13.3–17.3)	<.001	3.75 (2.59–5.42)	<.001	6.52 (5.35–7.95)	<.001
Pneumonia (yes)	104 (82–132)	<.001	11.6 (6.80–19.9)	<.001	10.7 (8.8–13.07)	<.001
Treatment during admission (yes):   Lopinavir/ritonavir   Hydroxychloroquine   Interferon   Tocilizumab   Steroids	–   –   –   –   –	–   –   –   –   –	28.0 (16.1–48.7)   7.05 (5.84–8.51)   33.7 (4.67–243)   93.3 (29.9–290)   26.1 (21.5–31.7)	<.001   <.001   <.001   <.001   <.001	2.34 (1.95–2.83)   1.83 (1.58–2.12)   3.52 (2.15–5.76)   2.98 (2.43–3.65)   3.49 (3.02–4.04)	<.001   <.001   <.001   <.001   <.001
Baseline ACEi (yes)	1.09 (0.93–1.28)	.265	1.11 (0.85–1.44)	.431	0.92 (0.76–1.12)	.39
ACEi withdrawal at admission (yes)	–	–	2.94 (1.68–5.15)	<.001	5.09 (3.17–8.19)	<.001
Baseline ARB (yes)	0.88 (0.77–0.99)	.047	1.47 (1.19–1.81)	<.001	0.75 (0.63–0.88)	<.001
ARB withdrawal at admission (yes)	–	–	3.43 (2.28–5.17)	<.001	4.38 (3.08–6.21)	<.001
Previous NSAID (yes)	1.06 (0.77–1.45)	.701	0.19 (0.07–0.53)	.001	0.95 (0.65–0.14)	.805
Immunosuppressive medication (yes)   Prednisone-free regimen   CNI-free regimen   mTORi-free regimen   Mycophenolate-free regimen	0.82 (0.69–0.99)   1.39 (0.93–2.10)   0.86 (0.68–1.08)   1.15 (0.95–1.39)	.046   .110   .199   .148	0.71 (0.52–0.97)   0.78 (0.45–1.36)   1.56 (1.11–2.19)   0.56 (0.42–0.75)	.030   .380   .011   <.001	0.78 (0.59–1.02)   1.18 (0.74–1.89)   1.51 (1.10–2.07)   0.87 (0.67–1.11)	.075   .480   .011   .260
SARS-CoV-2 wave (first)	2.61 (2.30–2.96)	<.001	0.41 (0.33–0.51)	.001	1.74 (1.53–1.99)	<.001
ICU admission (yes)	–	–	–	–	3.78 (3.13–4.58)	<.001
ICU rejection (yes)	–	–	–	–	17.0 (12.9–22.3)	<.001
Vaccination status (yes)	0.89 (0.74–1.09)	.261	2.74 (2.09–3.58)	<.001	0.88 (0.69–1.13)	.315

The increased OR of UCI admission in univariate analysis among the vaccinated appears to be related to the availability of vaccines in most recent pandemic waves, in which ICU access was easier than in previous waves when the healthcare system was overwhelmed.

KRT; kidney replacement therapy; KT, kidney transplant; ACEi, angiotensin-converting enzyme inhibitors; ARB; angiotensin receptor blockers; NSAID, non-steroidal anti-inflammatory drugs; CNI, calcineurin inhibitors; mTORi, mammalian target of rapamycin inhibitors; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; OR (95%CI), odds ratio (95% confidence interval).

A total of 658 (11%) patients required mechanical ventilation. Mechanical ventilation use increased over time (P < .001) (Table 1), potentially reflecting improved capacities of healthcare facilities to absorb the case load.

In multivariate analysis, patient (older age, male sex, KT recipient), COVID-19 severity (baseline lymphopaenia, pneumonia) and pandemic (first wave) factors were independent predictors for admission (Table 3). Sensitivity analyses exploring different pandemic wave categories yielded similar results (Supplementary data, Tables S4 and S5).

Table 3.

Multivariate logistic regression for outcomes after SARS-CoV-2 infection

	Hospital admission		ICU admission		Mortality
	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P
Age (per year)	1.01 (1.01–1.02)	<.001	0.97 (0.96–0.98)	<.001	1.05 (1.04–1.06)	<.001
Sex (male)	1.21 (1.01–1.46)	.034	–	–	–	–
KRT (KT)	1.81 (1.46–2.23)	<.001	2.90 (2.28–3.68)	<.001	–	–
Lymphopenia (yes)	5.58 (4.64–6.65)	<.001	3.23 (2.16–4.82)	<.001	1.98 (1.48–2.64)	<.001
Pneumonia (yes)	50.7 (39.4–65.4)	<.001	16.9 (8.64–33.4)	<.001	2.87 (2.15–3.83)	<.001
KRT vintage (month)	–	–	0.99 (0.99–0.99)	.008	–	–
SARS-CoV-2 wave (first)	1.57 (1.28–1.93)	<.001	0.43 (0.34–0.54)	<.001	–	–
ICU rejection (yes)	–	–	–	–	13.1 (9.81–17.5)	<.001

Models adjusted for diabetic kidney disease. KRT, kidney replacement therapy; KT, kidney transplant; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; OR (95% CI): odds ratio (95% confidence interval).

Factors associated with ICU admission

Among infected patients, 562 (9%) were admitted to ICU, corresponding to 15% of patients admitted to hospital. In 439 patients (7%) were rejected from ICU admission despite severity criteria for ICU admission. The pattern of ICU admissions also evolved during the pandemic, being lower during the first wave (in which rejection from ICU was higher), likely reflecting overwhelmed healthcare systems (Supplementary data, Table S2).

Factors associated with ICU admission are summarized in Table 2. Multivariate analysis identified patient (younger age, KT recipient, shorter KRT vintage), COVID-19 severity (baseline lymphopaenia, pneumonia) and pandemic (infection outside the first wave) factors as independent predictors for ICU admission (Table 3). Sensitivity analyses exploring different pandemic wave categories yielded similar results (Supplementary data, Tables S6 and S7).

Factors associated with mortality

During follow-up, 1207 patients (21%) died, 4539 (74%) recovered and 344 (6%) were lost to follow-up. Figure 3 shows the mortality in the Spanish COVID-19 KRT Registry and COVID-19 mortality in the general Spanish population [3]. COVID-19 mortality among KRT patients was more frequent during the first wave (28%) than in other waves (18%) (P < .001) (Supplementary data, Table 2) and grossly paralleled those of the general population. From the fourth wave, i.e. after vaccines became available, the number of deaths decreased dramatically and occurred mainly in KT recipients (Fig. 4). However, among breakthrough diagnosed with COVID-19 in the vaccinated, mortality was 19% (Table 1). Of the 86 vaccinated KRT patients that died, a majority (77, 89%) were KT recipients (Supplementary data, Table 1).

FIGURE 3:

COVID-19 mortality over time in the COVID-19 Spanish KRT Registry and in the general Spanish population during the pandemic.

FIGURE 4:

SARS-CoV-2 deaths and vaccination status among kidney transplant recipients and haemodialysis patients. Continuous lines refer to total deaths in the different cohorts. Bars indicate the percentage of vaccination among patients dying from COVID-19. KT, kidney transplant; HD, haemodialysis.

COVID-19 mortality was higher in HD and PD (24 and 22%, respectively) than in KT (18%) or HDD (14%) patients (P < .001). Patient (male sex, older age, diabetic kidney disease, in-centre HD), COVID-19 severity (baseline lymphopaenia, pneumonia), COVID-19 treatment (ACEi and ARB withdrawal, ICU rejection) and pandemic (first wave) factors were also associated to mortality.

In multivariate analysis, patient factors (older age), COVID-19 severity (baseline lymphopaenia, pneumonia) and COVID-19 treatment (ICU rejection) factors were independent predictors of mortality (Table 3). Sensitivity analyses exploring different pandemic wave categories yielded similar results (Supplementary data, Tables S8 and S9).

While COVID-19 severity is an expected driver of mortality, the identification of additional patient factors associated with mortality, potentially by increasing disease severity, may be of interest. Focussing only on patient factors, multivariate analysis identified older age, male sex, diabetes and KT as independent risk factors for death, while ARB use was protective (Supplementary data, Table S10). An analysis focussed on KT recipient factors confirmed the higher risk conferred by older age, male sex and diabetes and, additionally, identified an association with immunosuppressants: mTORi-free and calcineurin inhibitor-free regimens were associated with mortality, while prednisone-free regimens were associated with lower mortality (Supplementary data, Table S11).

DISCUSSION

The key finding of our study is that, although lower than in the first wave, the mortality of KRT patients diagnosed with COVID-19 remains high (around 19% for vaccinated KRT patients) despite advances in treatment and higher availability of ICU care in the era of vaccines. This is 13-fold higher than the 1.55% mortality of confirmed COVID-19 in the general population in Spain [11]. However, the vaccine era witnessed a dramatic decrease in the number of diagnosed COVID-19 cases in the KRT population as compared with the general population, as well as in the absolute number of deaths, and mortality of vaccinated KRT patients was for the most part limited to KT recipients. Thus, a shift was observed from HD patients to KT recipients, for both cases and deaths, likely reflecting the lower efficacy of vaccines in KT recipients [9]. In this regard, the increased prevalence of chronic immunosuppressive regimens containing steroids among COVID-19-infected KT recipients over time should be considered in the context of reports that associate chronic immunosuppressive regimens containing steroids with poorer humoral responses to vaccines [9]. In contrast, almost half of the HD patients infected in the fifth wave and most of the few HD patients infected in the sixth wave were unvaccinated, supporting the efficacy of booster vaccine doses in this population.

Several changes in the presentation, treatment and outcomes of COVID-19 were evident over time and through the successive pandemic waves. One of the most striking features was the shift in the KRT patient population with COVID-19. The first wave mainly affected HD patients while the fifth and sixth waves, following vaccination, were the KT recipient waves. In the first wave, the healthcare system was unprepared to prevent the spread of the virus, but HD patients needed to access healthcare facilities thrice weekly and could not shield themselves [7, 12, 13, 14]. The more recent shift to KT recipients appears related to their suboptimal response to vaccines [9]. In this regard, although the case fatality ratio was similar among vaccinated KT recipients and vaccinated HD patients, there were 8.5-fold more deaths among vaccinated KT recipients than among vaccinated HD patients, likely reflecting the higher number of breakthrough SARS-CoV-2 infections despite vaccination in KT recipients. To put these data into perspective, the number of KT recipients with functioning grafts in Spain is 1.35-fold higher than the number of patients on HD [15].

A key finding is a potential role of therapeutic nihilism in the high mortality of KRT patients during the first COVID-19 wave, especially among HD and PD patients. In this regard, KT recipients were admitted more frequently, with more symptoms and more frequent pneumonia and received more ICU care, but mortality was lower than in PD or HD patients. Multivariate analysis identified refusal of ICU care as an independent risk factor for death, among other patient and disease severity factors. Difficult decisions had to be made early in the COVID-19 pandemic when ICUs were overwhelmed. This observation is confirmed in our data, where young patients were more frequently accepted for ICU admission. However, we should strive to prevent the rejection of KRT patients during the first pandemic wave becoming a self-fulfilling prophecy and KRT patients being refused advanced life support in the future because of that earlier high mortality. Indeed, our data show that KRT patients with COVID-19 during the first wave were 2.3-fold less likely to be admitted to ICUs than in subsequent waves. In this regard, in the most recent pandemic waves, when vaccines are available and therapies such as dexamethasone are widely used [16], most KRT patients with breakthrough COVID-19 survived. Moreover, in KRT patients, unlike in the general population, additional comorbidities, such as diabetes, hypertension, coronary artery disease and chronic lung disease, do not further add to the risk of death and the impact of older age on the risk of death is up to 6-fold lower than in the general population [1, 3].

Beyond access to advanced life support, other treatment factors may influence the risk of death. In addition to the prescription of antivirals or steroids, adjustment of chronic immunosuppression could be a successful strategy in KT recipients. In this regard, mTORi-free and calcineurin inhibitor-free regimens were associated with higher mortality, while steroid-free regimens were consistently associated with lower hospital admissions, lower ICU admissions and lower mortality. Data on the prevalence of chronic immunosuppressive regimens containing steroids among infected KT recipients over time also points to the negative impact of steroids-containing immune suppressive regimens on sensitivity to COVID-19 and also, to severe and potentially lethal COVID-19. This may be, at least in part, related to reports that associate steroids-containing immunosuppressive regimens with poorer humoral responses to SARS-CoV-2 vaccines [9]. Interestingly, steroids are beneficial when used acutely in COVID-19 [17]. This controversy can be understood in light of the anti-inflammatory effect of high doses of corticosteroids on the cytokine storm phase of COVID-19, in contrast to the chronic immunosuppression with steroids prescription that may dampen antiviral and vaccine responses. Regarding acute steroid prescription, an indication bias should be noted, as they were prescribed to treat severe COVID-19, and we believe that this underlies the association with mortality. Among other pre-existent therapies, baseline use of ARBs was a patient factor associated with lower mortality in the full KRT population, albeit not in KT recipients. While this should be confirmed in a wider KRT population, there is biological plausibility for the relationship, since angiotensin receptor-1 (AT-1), the target of ARBs, was recently shown to facilitate SARS-CoV-2 entry into cells mediated by soluble angiotensin-converting enzyme 2 (sACE2) [18]. Thus, ARBs could potentially interfere with this port of entry of SARS-CoV-2 into cells and we can only speculate that, based on the clinical observation, this viral entry pathway may be more relevant in dialysis patients. This hypothesis merits being pursued in further research as it may be clinically relevant.

The main impact of vaccines appeared to be the reduction in the number of cases, as mortality among vaccinated KRT patients diagnosed with COVID-19 was similar to among non-vaccinated KRT patients diagnosed with COVID-19. However, mortality among vaccinated COVID-19 patients was almost exclusively found in KT recipients, pointing to a suboptimal response to vaccination in those patients who died despite vaccination. Thus, as the pandemic shifts from the wider KRT population to vaccinated KT recipients, the urge to optimize the vaccination regimens increases. This may imply transient adaptation of immune suppression regimens, more common use of the mRNA-1273 vaccine, additional booster doses or novel vaccines [9].

The presentation of SARS-CoV-2 infection and the case fatality ratio may have been influenced by the criteria used to perform diagnostic tests. Thus, early in the pandemic, testing seemed to be reserved for more symptomatic patients, given supply chain issues. Later, testing of even asymptomatic patients was performed in HD units for outbreak monitoring, but not in at-home dialysis or KT recipients [19, 20]. This may explain the lower likelihood of being asymptomatic and higher rates of pneumonia in KT recipients and PD patients. Indeed, asymptomatic infections in the second and third waves were 26–28% and decreased to ˂20% when the epidemiological context changed and preventive screening protocols became outdated. However, the combination of strictly observed isolation measures (e.g. masks during transport to and from dialysis centres and during HD sessions) and vaccines appears to be more effective in the KRT population than vaccines in combination with more relaxed measures in the general population, as judged by the lower number of infections recorded upon vaccine implementation in KRT patients than in the general population. Thus, vaccination and booster doses should continue to be promoted in KRT patients and isolation measures should be maintained. In this regard, early in the pandemic lack of face masks during transport to and from dialysis and suboptimal personal protective equipment were identified as risk factors for infection in KRT patients and healthcare personnel [7, 11, 12].

Some limitations should be acknowledged. In this registry-based study, inherent bias includes some incomplete data, limited variables (such as body mass index or treatment duration) or loss of follow-up. However, first, this is a multicentric prospective study involving a large cohort of patients on KRT infected by SARS-CoV-2 in Spain, corresponding to roughly 10% of Spanish patients on KRT [12], so data and conclusions are thought to be representative. Second, information on vaccination dates or antibody response is not available. In Spain, the first vaccine dose was administered at the end of December 2020, but widespread vaccination started in January 2021 and reached KRT patients at some point between February and April 2021. Third, subclinical or asymptomatic infections probably went unnoticed, especially beyond the third wave, when systematic periodic screening of HD units was no longer performed, so the studied outcomes could be overestimated. However, this makes the data more comparable to the general population, for which widespread screening of asymptomatic individuals is not performed [15]. Fourth, the sixth wave data predate the introduction and expansion of the Omicron variant in Spain and thus, the impact of this strain is not reflected in the data. Finally, some variables, such as rejection from ICU admission, did not have a universal definition and should be analysed from a subjective point of view.

In conclusion, the features of the COVID-19 pandemic in KRT patients (patient characteristics, treatment and outcomes) keep evolving, shifting towards vaccinated KT recipients and unvaccinated dialysis patients. Vaccines are associated with a lower incidence of diagnosed COVID-19, but mortality remains high despite ICU care, although most patients survive. Further improvement in KRT patient outcomes may be obtained by optimizing vaccination and immune suppression protocols in KT recipients, promoting vaccination and boosting dialysis patients and providing access to life support care if needed.

APPENDIX

The Spanish COVID-19 KRT Registry collaborative group:

Victoria Oviedo

Antoni Bordils Gil

Maria Luisa Navarro López

María Isabel Martínez Marín

Amparo Bernat García

Isabel Berdud Godoy

Mª Victoria Guijarro Abad

María del Mar Rodríguez de Oña

Elena Vaquero Párrizas

María Antonia Fernández Solís

Jose Antonio Gomez Puerta

Alvaro Ossorio Anaya

Elena Calvo

Carmen Cabré Menéndez

Nuria García Fernández

Paloma Leticia Martín Moreno

Mª José Fernández-Reyes Luis

Cristina Lucas Álvarez

Guadalupe Tabernero fernández

Manuel Heras Benito

Pilar Fraile Gómez

Joaquin Manrique Escola

Enrique Pelaez Perez

Luz María Cuiña Barja

Teresa Cordal Martinez

Maria Jesus Castro Vilanova

Silvia Moreno Loshuertos

Francisco Javier Ahijado Hormigos

Ana Roca Muñoz

Carlos Jesús Cabezas Reina

Jesús Grande Villoria

Ana Isabel Díaz Mareque

Anabertha Narváez Benítez

Argimiro Gándara Martínez

Mª Gloria Rodríguez Goyanes

Maria Crucio Lopez

Alfonso Iglesias

Laureano Perez Oller

Carlos Antonio Soto Montañez

Juan Villaro Gumpert

Ana Beatriz Muñoz Díaz

Gustavo Useche Bonilla

Maria Noel Martina Lingua

Diana Pazmiño Zambrano

Sandra Castellano Gasch

Guadalupe Caparrós Tortosa

Silvia Soto Alarcón

Mercedes Albaladejo Pérez

Ignacio Cidraque Vella

Cristina Canal Girol

Silvia Benito García

Montserrat Picazo Sánchez

Anna Patrícia Balius Matas

Núria Garra Moncau

Rodrigo Avellaneda Campos

Ana María Ramos Verde

Laura Muñiz Pacios

Alfonso Pobes Martínez de Salinas

Maria Jimenez Herrero

Liliana Morán Caicedo

Mª Dolores Prados Garrido

Mariana Garbiras

Laura Sanchez Rodríguez

Maria Dolores Albero Molina

Elena Gutiérrez Solis

Ana Hernández Vicente

Angel.M Sevillano Prieto

Teresa Bada Bosch

Hernando Trujillo Cuéllar

Helena Díaz Cambre

Cassandra Emma Puig Hooper

Adriana Maria Cavada Bustamante

Boris Gonzales Gandía

Miguel Rodeles del Pozo

Miquel Blasco Pelicano

José Jesús Broseta Monzó

Jaime Sanz García

Jose Antonio Herrero Calvo

Isabel Maria Pérez Flores

Virginia López de la Manzanara Pérez

Javier Vian Pérez

Sara Huertas Salazar

Armando Coca Rojo

R. Alvarez

Miguel Angel González Rico

Ana Isabel Martinez Diaz

Elena Giménez Civera

Ignacio Lopez Alejaldre

Claudio José Hornos Hornos

Zakariae Koraichi Rabie Senhaji

Enriqueta González Rodríguez

Andrea Patricia Zapata Balcázar

Cecilia Montoyo Castillo

David Tura Rosales

Raúl Edilberto Alvarado Gutiérrez

Rafael Garcia Maset

Rosa Garcia Osuna

Maria Eugenia Palacios Gómez

Sergio García Marcos

Francisco Roca Oporto

Manuel Ramirez de Arellano Serna

Elena Olivar Pérez

Oihana Larrañaga Zabaleta

Mº Dolores Arenas Jiménez

Marta Crespo Barrio

Silvia Collado Nieto

María Dolores Redondo Pachón

Anna Buxeda i Porras

Laura Llinàs Mallol

Alberto Mendoza-Valderrey

Carola Arcal Cunillera

Basilio Martin Urcuyo

Adelaida Morales Umpiérrez

Aránzazu Márquez Corbella

Eva Gavela Martínez

Julia Kanter

Sandra Beltrán Catalán

Mercedes Gonzalez Moya

July Vanessa Osma Capera

Alejandro Valero Anton

Elena Castillón Lavilla

Juan Casas Todolí

Amir Shabaka Fernández

Ainhoa Hernando Rubio

Inmaculada Lorenzo Gonzalez

Francisco Llamas fuentes

Francisco Javier Centellas Pérez

Ana Pérez Rodríguez

Alejandra Rodriguez

Jorge Reichert

Rosa Sanchez Hernández

Luis Fernando Domínguez Reina

Antonio Franco Esteve

Eduardo Muñoz de Bustillo Llorente

Dioné González Ferri

Alejandro Pérez Alba

Luis Guillermo Piccone Saponara

María Esperanza Moral Berrio

Silvia Ros Ruiz

Leonidas Cruzado Vega

Eduardo Bosque Muñoz

Joaquín de Juan Ribera

Josefa Martin Rivas

Inés Aragoncillo Sauco

Luis Alberto Sánchez Cámara

Antonio Pérez Pérez

Mª José Navarro Parreño

Gracia Mª Alvarez Fernández

Marisol Ros Romero

Diana Manzano Sánchez

Maria Molina Gomez

Javier Juega

Andrés Villegas Fuentes

Laura Espinosa Román

Cristóbal Donapetry García

Pablo Castro de la Nuez

Antonio Crespo Navarro

Neus Rodriguez Farre

Jesus Martin Garcia

Antonio Gascón Mariño

Secundino Cigarran Guldris

Margarita Montserrat Pousa Ortega

Roberto Holgado Salado

Margarita Delgado Cordova

Natalia Blanco Castro

Elvira Esquivias de Motta

Veronica Lopez Jimenez

Belen Gómez Giralda

María Verónica Torres Jaramillo

Juan Cristobal Santacruz Mancheno

Beatriz María Durá Gúrpide

Paula Munguía Navarro

Belen Moragrega

Marta Luzon Alonso

Fernando Gil Catalinas

Emma Huarte Loza

Cecilia Dall'Anese Siegenthaler

Pedro Jesús Labrador Gómez

Silvia González Sanchidrián

Yanina García Marcote

Bruna Natacha Leite Costa

Joan Manuel Gascó Company

Juan Rey Valeriano

Sheila Cabello Pelegrin

Paloma Livianos Arias-Camisón.

Maria Teresa García Falcón

Ana Rodriguez-Carmona

Antía López Iglesias

Felipe Sarró Sobrín

Maria Luisa Martin Conde

Javier Arrieta Lezama

Iñigo Moina Eguren

Olga Gonzalez Peña

Maria del Carmen Díaz Corte

Maria Luisa Suarez Fernandez

Pedro Vidau Arguelles

Elena Astudillo Cortés

Lucía Sobrino Díaz

Alba Rivas Oural

Luis Fernando Morán Fernández

Oscar Rolando Durón Vargas

Clara Sanz García

Adriana Maria Cavada Bustamante

Carmen Robledo Zulet

Oscar García Uriarte

Maria Begoña Aurrekoetxea Fernandez

Maria Isabel Jimeno Martín

Guillermo Alcalde Bezhold

Rosa Maria Ruiz-Calero Cendrero

Roman Hernandez Gallego

Maria Victoria Martin Hidalgo-Barquero

Pedro Abáigar Luquin

Emilio Sanchez

Sagrario García Rebollo

María Lourdes Pérez Tamajón

Alejandra Maxorata Alvarez Gonzalez

Greissi Jeniree Garcia Bonilla

Eva Alvarez

Sofía Zarraga Larrondo

Alfonso Cubas Alcaraz

Mª Teresa Naya Nieto

Fernando Henriquez Palop

Raquel Santana Estupiñán

Ernesto Francisco Valga Amado

Gabriel De Arriba De La Fuente

Marta Sánchez Heras

Concepción Alamo Caballero

Maria Pilar Perez del Barrio

Clara María Moriana Domínguez

Sara Blázquez Roselló

Gema Velasco Barrero

Jary Perello Martinez

Manuel Ramos Díaz

Marina Almenara Tejederas

Martín Giorgi González

Beatriz Diez Ojea

Vicente Paraíso Cuevas

Fernando Tornero Molina

David rodríguez Santarelli

Jessica Isabel Urdaneta Colmenares

Luis Alberto Blázquez Collado

Maria Teresa Rodrigo De Tomas

Fernando Simal Blanco

Marta Albalate Ramón

Patricia de Sequera Ortiz

Rafael Lucena Valverde

Raquel Diaz Mancebo

Rocio Echarri Carrillo

Gabriel Ledesma Sanchez

Ernesto Jose Fernandez Tagarro

Iván Chamorro Bucheli

Joan Albert Fernández Roig

Isabel Garcia Mendez

Carlos Jimenez Martin

María Ovidia López Oliva

Mª Elena González García

Maria Auxiliadora Bajo Rubio

Laura Alvarez García

Jesus Calviño Varela

Juan Carlos Ruiz San Millán

Celestino Piñera Haces

Rosa Palomar Fontanet

Rosalía Valero San Cecilio

M. Jose Aladrén Regidor

Alejandro Soria Villén

Cristina Medrano Villarroya

Orlando Siverio Morales

Diego Rodríguez Puyol

María Pérez Fernández

Jose Portoles Perez

Charo Llopez Carratala

Auxiliadora Mazuecos

Juan Manuel Cazorla Lopez

Maríaa Gabriela Sánchez Márquez

Carolina Lancho Novillo

Lien Winderickx

Carlos Íñiguez Villalón

Cristina Galeano Álvarez

Sara Jiménez Álvaro

Esmeralda Castillo Rodríguez

Daniel Eduardo Villa Hurtado

Milagros Fernandez Lucas

Alberto Rodríguez Benot

Sagrario Soriano Cabrera

Raquel Ojeda López

Jose Luis Pérez Canga

Ana Cristina Andrade López

Anna Gallardo Pérez

Aniana Oliet Palá

Maria Sanchez Sanchez

Juan Manuel Buades Fuster

Maria Eugenia Palacios Gómez

Felisa Martínez Sanchez

Cristina Jimeno Griñó

Adoración Martínez Losa

Isabel María Saura Luján

Luis Gil Sacaluga

Gabriel Bernal Blanco

Maria Jose Marco Guerrero

Ana Isabel Martínez Puerto

Mª Jesús Moyano Franco

Elena Araceli Jiménez Víbora

Nuria Aresté Fosalba

Maria de los Ángeles Rodríguez Pérez

Ramos Escorihuela

Ramón Jesús Devesa Such

Isabel Beneyto Castelló

Edoardo Melilli

María José Soler Romeo

Francesc Moreso Mateos

Nestor Gabriel Toapanta Gaibor

Alfonso Pobes Martínez de Salinas

Begoña Rincón Ruiz

Mªcarmen Ruiz Fuentes

Ramón López-Menchero Martínez

Miguel Angel Suarez Santistebam

Carlos Jarava Mantecón

Fernando Fernandez Giron

Omar Miranda Espinal

Juan Carlos Martínez Ocaña

Raúl García Castro

Florentino Villanego Fernández

Montserrat Belart Rodríguez

Camino García Monteavaro

DATA AVAILABILITY STATEMENT

Data are available upon a reasonable request.

CONFLICT OF INTEREST STATEMENT

B.Q. has received honoraria for conferences, consulting fees and advisory boards from Vifor-Pharma, Astellas, Amgen, Bial, Ferrer, Novartis, AstraZeneca, Sandoz, Laboratorios Bial, Esteve, Sanofi-Genzyme and Otsuka. He is the present secretary of the Spanish Society of Nephrology. A.O. has received grants from Sanofi and consultancy or speaker fees or travel support from Advicciene, Astellas, AstraZeneca, Amicus, Amgen, Fresenius Medical Care, GSK, Bayer, Sanofi-Genzyme, Menarini, Kyowa Kirin, Alexion, Idorsia, Chiesi, Otsuka, Novo-Nordisk and Vifor Fresenius Medical Care Renal Pharma, and is Director of the Catedra Mundipharma-UAM of diabetic kidney disease and the Catedra AstraZeneca-UAM of chronic kidney disease and electrolytes. He is the former Editor-in-Chief of CKJ. M.M.G. has received honoraria for conferences, consulting fees and advisory boards from Astellas and Chiesi. P.deS. reports honorarium for conferences, consulting fees and advisory boards from Amgen, Astellas, AstraZeneca, Baxter, Braun, Fresenius, Nipro, Vifor-Pharma. She is the present president of the Spanish Society of Nephrology (S.E.N.). E.S.-Á. has received honoraria for conferences, consulting fees and advisory boards from AstraZeneca, Vifor, Astellas, Novo Nordisk and Baxter.

C.J.C.-R., M.C.R.F., V.L.J., N.T. and S.Z.L. do not present any disclosure.