Outcomes of COVID‑19 in hospitalized kidney and liver transplant recipients: a single‑center experience

Introduction

Following the first case of COVID-19, detected in December 2019 in China, SARS-CoV-2 has rapidly spread across the globe. The international pandemic of COVID-19 has led to a substantial healthcare crisis and over 2.8 million deaths worldwide (as of April 8, 2021).¹

A large body of evidence indicates that the course of COVID-19 is the most severe in patients with comorbidities and advanced age.^2-4 Underlying type 2 diabetes mellitus and a history of chronic obstructive pulmonary disease, chronic kidney disease, or cardiovascular diseases are associated with increased mortality.^3-6 Solid organ transplant recipients constitute a specific group of patients, who not only commonly present with comorbidities,⁷ but also receive immunosuppressive drugs, serving as a potential cause of particularly high susceptibility to a severe course of infectious diseases.⁸ Despite improvements in prophylaxis and antimicrobial therapies, infections still account for approximately 13% of deaths in kidney (KTRs) and liver transplant recipients (LTRs).⁸ SARS-CoV-2 infection in transplant recipients might contribute to a particularly high risk of respiratory failure, requiring invasive mechanical ventilation and admission to the intensive care unit (ICU); it can also lead to death. A meta-analysis encompassing 2772 solid organ transplant recipients with confirmed COVID-19 reported a very high frequency of hospital admission—exceeding 80%—and a high mortality rate of 18.6%.⁹ Notably, the subset of patients after kidney transplantation (the largest group of transplant recipients) was characterized by an even higher mortality rate, reaching 28%,¹⁰ 32%,¹¹ and 37.8%¹² in different reports. Despite several studies investigating COVID-19 in transplant recipients, the management of the disease still remains suboptimal, with no convincing evidence on the benefit of using tocilizumab and with controversy regarding the dosage of calcineurin inhibitors (CNIs) in this group of patients (especially in liver graft recipients).^9,13,14

Patients after transplantation more frequently require hospital admission (about 80% of cases) and experience death in the course of COVID-19 than the general population, in which the case-fatality rate is approximately 2%, varying from country to country.^12,15,16 The relatively high rate of acute respiratory distress syndrome (ARDS; from 37% to 60% in different studies)^10,17 and mortality from COVID-19 among KTRs necessitates an urgent investigation of effective antivirals and prioritization of vaccinations in this group.

The factors identified as independently affecting mortality among KTRs with COVID-19 differ between studies.^11,12,17 In-hospital mortality was associated with older age, a higher respiratory rate, higher lactate dehydrogenase (LDH), interleukin-6 (IL-6), and procalcitonin levels, and a lower estimated glomerular filtration rate (eGFR) in one study,¹¹ whereas another reported older age, deceased-donor transplantation, a lack of influenza vaccination, and elevated IL-6 levels as being related to increased mortality.¹² Further studies validating these prognostic factors and identifying novel predictors that may guide clinical decisions and facilitate early intensive care are warranted.

Here, we present a retrospective study of COVID-19 features among KTRs and LTRs and we identify the factors associated with an increased risk of severe ARDS and in-hospital mortality.

Patients and methods

This is a retrospective, single tertiary center study performed on the group of kidney and liver transplant recipients admitted to our designated COVID-19 unit between November 9, 2020 and February 26, 2021. All patients had a functioning allograft on admission and were over 18 years of age. All patients tested positive for SARS-CoV-2 in a real-time reverse–transcriptase polymerase chain reaction (RT-PCR) test of nasopharyngeal swab specimens. The infections could be attributed to pre-existing SARS-CoV-2 variants, since the SARS-CoV-2 B.1.1.7 variant (first detected in the United Kingdom) became prevalent in Poland in March 2021. None of the patients were vaccinated against SARS-CoV-2 before hospital admission. Patients who underwent graftectomy (n = 2) or had a chronic nonfunctioning graft (n = 1) were excluded.

The medical records of patients after transplantation were reviewed. Demographics, epidemiological data, presenting symptoms (eg, dyspnea, cough, fever, diarrhea, or myalgia), and vital signs (blood pressure, respiratory rate, heart rate, and oxygen saturation) were extracted. The data regarding comorbidities, medication, graft function, and the time and type of transplantation were collected. Perioperative COVID-19 was defined as a positive RT-PCR test up to 1 month after transplantation.

Laboratory results (complete blood count, arterial blood gases, creatinine, eGFR, ionogram, coagulation profile, C-reactive protein [CRP], procalcitonin, D-dimers, ferritin, LDH, fibrinogen, and liver function) and radiological findings (chest computed tomography [CT] scan) assessed on admission and during hospitalization were collected.

Data on treatment modalities—immunosuppressive regimen modifications and supportive therapies administered during hospitalization—were collected. Outcomes were also analyzed, including mortality, respiratory failure requiring invasive mechanical ventilation, ICU admission, ARDS (according to the Berlin criteria), multiorgan dysfunction, acute kidney injury (AKI; defined according to KDIGO 2012), and sepsis (defined as a positive blood culture and systemic inflammatory response syndrome). Severe ARDS was diagnosed when the ratio of arterial partial pressure of oxygen (PaO₂) to the fraction of inspired oxygen (FiO₂) was lower than 100 mm Hg. At the time of analysis, all patients had been discharged home or were deceased.

The pharmacological management of patients with COVID-19 was in accordance with the guidelines released by the Agency for Health Technology Assessment and Tariff System (Poland) and the National Institutes of Health (United States). In general, dexamethasone was administered at a dose of 6 mg daily for 7 to 10 days in patients with an oxygen saturation greater than 93% in ambient air. Dexamethasone was usually initiated in the second week after the onset of symptoms; other regularly used steroids were withdrawn at that time. Remdesivir was administered at a dose of 200 mg on the first day, followed by 100 mg per day for another 4 days. It was initiated in patients in the first week after the onset of symptoms when oxygen saturation was at least 94% in ambient air. Patients with an eGFR of less than 30 ml/min/1.73 m² were not treated with remdesivir, as it was contraindicated according to the summary of product characteristics.

All procedures performed during the study were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki Declaration and its later amendments. The study was approved by an appropriate institutional ethics committee.

Statistical analysis

Statistical analysis was performed in SAS, version 9.4 (SAS Institute, Cary, North Carolina, United States). The results are presented as number and percentage for categorical variables and medians and interquartile ranges (IQRs) for continuous variables. The Mann–Whitney and Fisher exact tests were used for continuous and categorical variables, respectively. For univariable and multivariable analyses, logistic regression was performed. Selected variables from the univariable analysis (with P <⁠0.15) served as variables for a stepwise regression to create the multivariable model. The cutoff probability for adding variables into the model was P = 0.15 and for keeping variables in the multivariable model it was P = 0.05. Spearman correlation coefficient was evaluated to exclude correlating predictors from the multivariable analysis whenever required. In all analyses, a P value of less than 0.05 was considered statistically significant. The odds ratios (ORs) with 95% CIs were derived from logistic analysis.

Results

Overall, 41 transplant recipients admitted to the hospital with confirmed SARS-CoV-2 infection were included in the study. Thirty-two patients (78%) had received a kidney transplant, 1 patient (2.4%) had undergone a kidney-pancreas transplantation, and 8 (19.5%) had received a liver transplant. Seven patients (17%) were infected with SARS-CoV-2 in the perioperative period (the first month after transplantation).

All patients had SARS-CoV-2 infection confirmed by a PCR test performed because of a clinical suspicion of COVID-19. Twenty-seven patients (66%) had PCR tests performed on admission to the hospital. The remaining 14 patients (34%) were referred for the test by the family doctor and were initially treated in an outpatient setting. After the aggravation of symptoms, they were referred to the hospital (within a median of 6 days after the positive PCR test result).

Kidney transplant recipients

Among the KTRs, 72% were male, 78% had undergone a single transplantation, and the median (IQR) age was 54 (47–62) years. A total of 87% were deceased-donor graft recipients. COVID-19 was diagnosed a median (IQR) of 71 (26–119) months after transplantation. The reasons for transplantation included glomerulonephritis (n = 14 [42%]), autosomal dominant polycystic kidney disease (n = 9 [27%]), reflux nephropathy (n = 3 [9%]), diabetes mellitus (n = 2 [6%]), hypertension (n = 1 [3%]), and others (n = 4 [12%]). Twenty-eight patients (85%) were receiving a triple immunosuppressive regimen (a corticosteroid, a calcineurin inhibitor, and an antimetabolite). Specifically, 29 KTRs (88%) were on mycophenolate mofetil, 27 (82%) were on tacrolimus, and 32 (97%) were administered corticosteroids. One patient had received basiliximab as an induction immunosuppressive therapy 4 weeks before SARS-CoV-2 infection. None of the patients underwent depletion therapy with rituximab or antithymocyte globulins in the 12 months prior to COVID-19.

The most prevalent comorbidities were hypertension (93%), diabetes mellitus (33%), coronary artery disease (21%), and obesity (21%). On admission, the most common symptoms were fever (72%), cough (66%), dyspnea (57%), diarrhea (36%), and myalgia (30%).

During a median (IQR) of 12 (8–18) days of hospital stay, 15 patients (45.5%) developed ARDS and 24 (72.2%) experienced AKI. The in-hospital mortality rate among the KTRs was 30.3% (n = 10). A total of 8 KTRs (24%) were intubated, invasively mechanically ventilated, and transferred to the ICU. A comparison between the groups of survivors (n = 23) and nonsurvivors (n = 10) is presented in Tables 1, 2, 3, 4.

Table 1. Baseline characteristics, presenting symptoms, and comorbidities according to COVID-19 mortality outcomes in kidney transplant recipients

Characteristics		Overall (n = 33)	Survivors (n = 23)	Nonsurvivors (n = 10)	P value
Male sex		24 (73)	18 (78)	6 (60)	0.4
Age, y		54 (47–62)	53.9 (48–61)	55.4 (47–68)	0.5
BMI, kg/m²		24.4 (22.9–28.5)	24.2 (23.4–29.5)	24.6 (22.2–26.6)	0.26
Time since transplantation, mo		65 (26–123)	71 (26–119)	59 (6–204)	0.39
Number of transplantations	1	26 (79)	18 (78)	8 (80)	>0.99
	2	7 (21)	5 (21)	2 (20)
Perioperative COVID-19a		3 (9)	2 (9)	1 (10)	>0.99
Deceased-donor transplantation		29 (88)	20 (86)	9 (90)	>0.99
Hospital-acquired COVID-19		4 (12)	2 (9)	2 (20)	0.56
Time from symptom onsetb, d		10 (7–14)	13 (7–16)	7.5 (6–10)	0.04
Dyspnea		19 (57)	11 (47)	8 (80)	0.08
Cough		22 (66)	14 (60)	8 (80)	0.43
Fever		24 (72)	17 (74)	7 (70)	>0.99
Anosmia / ageusia		2 (6)	2 (9)	0	>0.99
Diarrhea		12 (36)	9 (39)	3 (30)	0.68
Myalgia		10 (30)	8 (35)	2 (20)	0.71
Headache		3 (9)	3 (13)	0	0.53
Oxygen use on admission		17 (52)	9 (39)	8 (80)	0.04
Compromised lungs on CT, %		35 (25–65)	35 (20–50)	62.5 (30–75)	0.04
Comorbidities
Hypertension		31 (93)	21 (91)	10 (100)	>0.99
Coronary artery disease		7 (21)	3 (13)	4 (40)	0.16
Myocardial infarction		2 (6)	0	2 (20)	0.09
DM		11 (33)	5 (22)	6 (60)	0.049
DM treated with insulin		6 (18)	1 (4)	5 (50)	0.005
History of cancer		2 (6)	1 (4)	1 (10)	0.52
Stroke		1 (3)	1 (4)	0	>0.99
Atrial fibrillation		2 (6)	1 (4)	1 (10)	0.52
Heart failure		1 (3)	1 (4)	0	>0.99
Pulmonary disease		2 (6)	1 (4)	1 (10)	0.52
Thromboembolism		3 (9)	2 (9)	1 (10)	>0.99
Number of comorbiditiesc		1 (1–2)	1 (1–2)	2 (1–4)	0.02
Data are presented as number (percentage) or median (interquartile range). a COVID-19 in the first month after transplantation b Time from onset of symptoms to ward admission (days) c Each of the following diseases was counted: hypertension, diabetes, coronary artery disease, pulmonary disease, thromboembolism, atrial fibrillation, heart failure, history of stroke, myocardial infarction, and cancer. Abbreviations: BMI, body mass index; CT, computed tomography; DM, diabetes mellitus

Table 2. Laboratory findings and vital signs on admission according to COVID-19 mortality outcomes in kidney transplant recipients

Parameter	Overall (n = 33)	Survivors (n = 23)	Nonsurvivors (n = 10)	P value
Laboratory findings
White blood cells, × 109/l	5.78 (4.5–8.8)	5.55 (4.5–8.28)	6.48 (4.1–8.84)	0.5
Neutrophils, × 109/l	4.85 (3.43–6.24)	4.3 (3.5–6.5)	5.24 (3.19–7)	0.33
Lymphocytes, × 109/l	0.74 (0.43–0.96)	0.74 (0.53–0.95)	0.55 (0.33–1.08)	0.14
Platelets, × 109/l	202 (159–250)	202 (163–250)	208.5 (122–259)	0.38
Red blood cells, × 1012/l	4.14 (3.5–4.63)	4.14 (3.57–4.76)	4.05 (3.39–4.45)	0.09
Hemoglobin, g/l	119 (99–133)	119 (101–135)	117.5 (98–129.5)	0.34
Creatinine at baselinea, mg/ml	1.49 (1.3–1.82)	1.47 (1.3–1.7)	1.75 (1.39–1.98)	0.09
eGFR at baselinea, ml/min/1.73 m2	42.7 (36.2–58.2)	44.6 (38.8–58.7)	36.8 (30.3–48.1)	0.056
Creatinine on admission, mg/ml	2.0 (1.7–2.6)	2.0 (1.56–2.41)	2.17 (1.85–3.07)	0.13
eGFR on admission, ml/min/1.73 m2	29.8 (22.2–39.3)	34.7 (24.7–44.3)	25.7 (20.1–33.4)	0.056
Urea, mg/dl	81 (55–123)	70 (53–88)	122.3 (83–176)	0.09
C-reactive protein, mg/l	64 (40–85)	53 (35–73)	88.9 (69–110)	0.01
Procalcitonin, ng/ml	0.27 (0.09–0.47)	0.17 (0.08–0.4)	0.41 (0.29–2.7)	0.04
Lactate dehydrogenase, U/l	278.5 (212–301)	252 (212–283)	435 (313–526)	0.13
Serum ferritin, µg/ml	894 (372–1973)	870 (324–1083)	1620 (435–2872)	0.16
Fibrinogen, mg/dl	5 (4.1–6.4)	4.7 (4.0–6.15)	5.45 (5.0–6.75)	0.11
D-dimer, ng/ml	1000 (623–1450)	991 (622–1500)	1067.5 (712–1410)	0.39
Aspartate transaminase, U/l	23 (16–31)	24.5 (18–30)	18 (14–31)	0.29
Vital signs
Respiratory rate, bpm	15 (15–20)	15 (14–17)	20 (16–25)	0.12
SBP, mm Hg	127 (118–134)	125 (118–135)	129 (115–132)	0.42
Heart rate, bpm	83 (75–90)	83 (74–89)	89 (81–90)	0.007
Oxygen saturation, %	96 (94–97)	96 (95–97)	92.5 (86–98)	0.16
Data are presented as median (interquartile range). a Baseline creatinine concentration and eGFR were assessed during a regular visit to a transplantation outpatient clinic before SARS-CoV-2 infection Abbreviations: eGFR, estimated glomerular filtration rate based on the Modification of Diet in Renal Disease Study equation; SBP, systolic blood pressure

Table 3. Outcomes of COVID-19 in hospitalized kidney transplant recipients

Outcome			Overall (n = 33)	Survivors (n = 23)	Nonsurvivors (n = 10)	P value
Invasive mechanical ventilation			8 (24)	0	8 (80)	<⁠0.001
ICU admission			8 (24)	0	8 (80)	<⁠0.001
ARDS overall			15 (45)	5 (22)	10 (100)	<⁠0.001
ARDS	Stage 1		1 (3)	1 (4)	0	<⁠0.001
	Stage 2		2 (6)	2 (9)	0
	Stage 3		12 (36)	2 (9)	10 (100)
AKI overall			24 (72)	15 (65)	9 (90)	0.003
AKI		Stage 1	17 (51)	13 (56)	4 (4)	0.003
		Stage 2	2 (6)	2 (9)	0
		Stage 3	5 (15)	0	5 (50)
Multiple organ dysfunction			10 (30)	0	10 (100)	<⁠0.001
Urinary tract infection			5 (15)	4 (17)	1 (10)	1.0
Sepsis			4 (12)	1 (4)	3 (30)	0.07
Renal replacement therapy			4 (12)	0	4 (49)	0.005
Pulmonary embolism			2 (6)	2 (9)	0	>0.99
Length of stay, d			12 (8–18)	11 (7–16)	20 (15–41)	0.007
Data are presented as number (percentage) or median (interquartile range). Abbreviations: AKI, acute kidney injury; ARDS, acute respiratory distress syndrome; ICU, intensive care unit

Table 4. Management of hospitalized kidney transplant recipients with COVID-19

Variable		Overall (n = 33)	Survivors (n = 23)	Nonsurvivors (n = 10)	P value
Immunosuppressants
Corticosteroids		32 (97)	23 (100)	9 (90)	0.3
Tacrolimus		27 (82)	20 (87)	7 (70)	0.33
Cyclosporin		5 (15)	3 (13)	2 (20)	0.63
MMF		29 (88)	21 (91)	8 (80)	>0.99
Triple regimen		28 (85)	20 (87)	8 (80)	0.63
Treatment
MMF withdrawal		27 (96)	19 (95)	8 (100)	0.57
Tacrolimus / cyclosporin withdrawal		12 (38)	5 (22)	7 (78)	0.006
Tacrolimus / cyclosporin dose reduction		15 (47)	13 (57)	2 (22)	0.12
Remdesivir		10 (30)	6 (26)	4 (40)	0.44
Dexamethasone		25 (76)	15 (65)	10 (100)	0.07
Tocilizumaba		2 (6)	1 (4)	1 (10)	0.52
Antibioticsb		32 (97)	22 (96)	10 (100)	>0.99
ACEI / sartan		6 (18)	5 (22)	1 (10)	0.64
Statins		11 (33)	10 (43)	1 (10)	0.11
LMWH	Prophylactic dose	13 (42)	10 (48)	3 (30)	0.29
	Intermediate dose	11 (35)	8 (38)	3 (30)
	Therapeutic dose	7 (23)	3 (14)	4 (40)
Oxygen delivery devicec	None	2 (6)	2 (9)	0	0.004
	Nasal cannula	10 (30)	10 (43)	0
	Simple face mask	3 (9)	3 (13)	0
	MR	6 (18)	4 (17)	2 (20)
	HFCN	6 (18)	3 (13)	3 (30)
	HFNC + MR	5 (15)	0	5 (50)
Data are presented as number (percentage). a Patients treated with tocilizumab had IL-6 concentrations of 234 and 750 pg/ml, respectively. b Antibiotics were administered in most cases, as the bacterial infection was highly suspected based on clinical symptoms and elevated values of inflammatory parameters (procalcitonin >0.25 ng/ml in 66% of patients) or when clinical symptoms persisted or aggravated and bacterial infection could not be ruled out. c Noninvasive devices used at the moment of the highest requirement for oxygen (before intubation). Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; CNI, calcineurin inhibitor; HFNC, high-flow nasal cannula; LMWH, low-molecular-weight heparin; MMF, mycophenolate mofetil; MR, mask with reservoir bag

Prediction of the risk of mortality and severe acute respiratory distress syndrome in the kidney transplant recipients

In the univariable analysis, the predictors of in-hospital mortality were as follows: a higher volume of compromised lungs on a CT scan, diabetes mellitus (especially treated with insulin), higher CRP, procalcitonin, LDH, and urea levels, lower oxygen saturation, and higher respiratory rate on admission (Table 5). In the multivariable analysis, baseline eGFR, respiratory rate, and diabetes mellitus constituted independent predictors of in-hospital death (Table 6).

Table 5. Risk factors for in-hospital COVID-19 mortality in kidney transplant recipients, univariable analysis

Variable	OR	95% CI	P value
Male sex	0.42	0.08–2.07	0.28
Age	1.01	0.94–1.08	0.71
BMI	0.91	0.74–1.12	0.4
Number of transplantations	0.90	0.14–5.66	0.91
Deceased-donor transplantation	1.35	0.12–14.8	0.8
Perioperative COVID-19	1.17	0.09–14.5	0.9
Hospital-acquired COVID-19	2.63	0.31–21.9	0.37
Time from symptom onset	0.87	0.74–1.01	0.08
Compromised lungs on CT	1.04	1.00–1.07	0.046
Dyspnea	4.36	0.75–25.1	0.09
Cough	2.57	0.44–14.9	0.29
Fever	0.82	0.15–4.25	0.81
Diarrhea	0.67	0.13–3.27	0.61
Coronary artery disease	4.44	0.77–25.6	0.09
Diabetes mellitus	5.40	1.08–26.9	0.04
Diabetes mellitus requiring insulin	22.0	2.08–232	0.01
Pulmonary disease	2.44	0.13–43.4	0.54
Number of comorbidities	1.93	1.02–3.65	0.04
White blood cells	1.00	0.74–1.34	0.99
Neutrophils	1.10	0.79–1.52	0.55
Lymphocytes	0.37	0.06–2.27	0.28
Platelets	1.00	0.98–1.00	0.65
Hemoglobin	0.99	0.96–1.02	0.58
Creatinine on admission	1.74	0.90–3.33	0.09
eGFR on admission	0.95	0.88–1.01	0.09
Urea	1.02	1.00–1.03	0.02
C-reactive protein	1.02	0.99–1.03	0.06
Procalcitonin	3.67	0.90–14.8	0.07
Lactate dehydrogenasea	1.02	0.99–1.04	0.06
Serum ferritin	1.00	1.00–1.00	0.22
Fibrinogen	1.24	0.76–2.01	0.37
Aspartate transaminase	0.99	0.94–1.03	0.52
Creatinine at baseline	1.78	0.64–4.92	0.26
eGFR at baseline	0.96	0.91–1.01	0.11
Heart rate	1.03	0.95–1.12	0.43
SBP	0.99	0.94–1.04	0.78
Respiratory rate	1.26	1.04–1.52	0.02
Oxygen saturation on admission	0.83	0.69–0.99	0.04
Tacrolimus concentrationa	0.99	0.93–1.05	0.74
Statins	0.14	0.02–1.34	0.09
ACEI / sartan	0.40	0.04–3.96	0.43
a Lactate dehydrogenase activity and tacrolimus concentration on admission were not assessed in all patients. Abbreviations: OR, odds ratio; others, see Tables 1, 2, and 4

Table 6. Multivariable analysis of independent risk factors for in-hospital COVID-19 mortality in kidney transplant recipients

Variable	OR	95% CI	P value
Respiratory rate	1.43	1.03–1.98	0.008
eGFR at baseline	0.90	0.82–0.99	0.02
Diabetes mellitus	31.83	1.86–545	0.04
Abbreviations: see Tables 2 and 5

In the univariable analysis, the patient’s sex, age, body mass index, and time after transplantation were not associated with severe ARDS. The risk factors for severe ARDS were a higher CRP level (OR, 1.03; 95% CI, 1.00–1.06; P = 0.03), a higher respiratory rate (OR, 1.29; 95% CI, 1.05–1.57; P = 0.02), a greater volume of compromised lungs on chest CT imaging performed on admission (OR, 1.05; 95% CI, 1.01–1.10; P = 0.009), dyspnea on admission (OR, 6.67; 95% CI, 1.16–38; P = 0.03), diabetes mellitus (OR, 5.95; 95% CI, 1.22–29; P = 0.03), and a higher urea concentration (OR, 1.02; 95% CI, 1.00–1.03; P = 0.04). Treatment with statins had a protective effect (OR, 0.10; 95% CI, 0.01–0.92; P = 0.04). Independent risk factors for severe ARDS included the extent of compromised lungs on chest CT and the presence of diabetes mellitus (Table 7).

Table 7. Multivariable analysis of independent risk factors for severe acute respiratory distress syndrome in kidney transplant recipients

Variable	OR	95% CI	P value
Compromised lung volume on CT	1.05	1.01–1.10	0.005
Diabetes mellitus	8.94	1.13–70.9	0.02
Abbreviations: see Tables 1 and 5

Liver transplant recipients

Among the 8 LTRs included in the study, 5 (62.5%) were male and 4 (50%) acquired in-hospital SARS-CoV-2 in the perioperative period. The most common symptoms on admission were cough (50%), fever (38%), myalgia (38%), and dyspnea (25%). Two patients (25%) with perioperative COVID-19 were asymptomatic. Only 2 patients (25%) needed oxygen supplementation during their hospitalization, 1 of whom (12.5%) required invasive mechanical ventilation and died. The deceased patient was 70 years old with multiple comorbidities including coronary artery disease, a history of stroke, diabetes mellitus, and pulmonary disease. None of the 4 LTRs infected with SARS-CoV-2 in the perioperative period experienced ARDS; all were discharged home after a median of 14 days. However, 2 patients developed wound infections requiring surgical intervention.

All 8 LTRs were administered CNIs (7 patients were on tacrolimus and 1 on cyclosporine); 7 patients (87.5%) received prednisone and 2 patients (25%) received mycophenolate mofetil as a baseline immunosuppressive regimen. Four LTRs (50%) were maintained on the standard dosage of a CNI, while the other 4 (50%) received a reduced dosage during their hospital stay. Mycophenolate mofetil therapy was withdrawn on admission in all cases.

Compared to the KTRs, the LTRs were more frequently infected with SARS-CoV-2 in the perioperative period (50% vs 9%; P = 0.02) and more frequently presented with diabetes mellitus (75% vs 33%; P = 0.04). Hypertension was more prevalent in the KTRs than in the LTRs (94% vs 50%; P = 0.009). The 2 groups did not differ in terms of median age (54 vs 58 years; P = 0.25) or the prevalence of other comorbidities. The LTRs less frequently had CNI therapy withdrawn (0% vs 37.5%; P = 0.04) but there were no differences in terms of the frequency of reduced CNI dosage. Dexamethasone treatment withdrawal was also less frequent in the LTRs (25% vs 76%; P = 0.01), as they less often experienced rapidly progressive clinical deterioration or required oxygen supplementation. Notably, the LTRs were less frequently treated with mycophenolate mofetil (25% vs 88%; P = 0.002) as part of the baseline immunosuppressive regimen.

Discussion

This study presents the outcomes of SARS-CoV-2 infection in KTRs and LTRs who required hospitalization and identifies the predictors of severe ARDS and in-hospital mortality. Our study established a mortality rate of 30.3% in the hospitalized KTRs, which is similar to that reported by others.^9,10 Notably, the in-hospital COVID-19 mortality of patients after kidney transplantation is similar to that observed in the general population of individuals hospitalized with COVID-19.^18,19 On the contrary, one of the case-matched studies suggests a higher mortality among single-organ transplant recipients.²⁰ It is worth mentioning that COVID-19 patients after kidney transplantation are younger than the average patient hospitalized with COVID-19. Therefore, the estimated age-standardized mortality would be much higher among KTRs, highlighting the need for special protection of transplant recipients and prioritization of the vaccination program in this group. In our cohort, age did not influence survival and the median (IQR) age in the deceased patients was barely 55 (47–68) years. The cause of the particularly severe course of COVID-19 in the KTRs remains quite enigmatic and requires studies which would assess the pathophysiological mechanisms leading to rapidly progressive clinical deterioration and resulting in compromised patient survival. Comorbid diseases and chronic immunosuppression seem obvious, but offer only a partial explanation. Our study validates the former assumption since we showed that the presence of comorbidities (especially diabetes mellitus) constituted a risk factor for mortality in the study cohort. However, we did not find any relationship between survival and immunosuppressive regimen or CNI concentration on admission, which has also been mentioned by others.^11,16 Mycophenolate mofetil was withdrawn in 96% of the patients on admission, whereas CNIs were withdrawn in 78% of the nonsurvivors and in 22% of the survivors (CNI dosage was reduced in 22% and 56% of the groups, respectively). In general, CNIs were withdrawn after clinical deterioration and the present study, due to its design, cannot draw any conclusions in terms of the effect of CNIs on the disease course. Further studies investigating the effect of CNIs which would also consider the phase of COVID-19 (eg, viral replication or cytokine storm) in which CNIs are being administered are warranted. A meta-analysis reported the pooled incidence of CNI reduction as only 38.7%.⁹ So far, there is no evidence of a survival advantage from early CNI withdrawal. Perhaps, the effect of drug-induced immunosuppression might facilitate the early phase of infection, in which viral replication occurs. On the other hand, immunosuppressive treatment during the late phase of COVID-19, namely the hyperinflammatory state, might be beneficial as immunomodulatory drugs (eg, corticosteroids or tocilizumab) are proposed for that indication.²¹ The phase of hyperinflammation and acute respiratory failure syndrome most frequently precedes multiple organ dysfunction and death.²² Effective drugs targeting hyperinflammation could possibly improve COVID-19 outcomes, which remain poor in severe and critical illness.

In our experience, lower baseline eGFR, higher respiratory rate on admission, and diabetes mellitus were independently associated with increased COVID-19 mortality in KTRs. Cravedi et al¹¹ also reported that respiratory rate and eGFR constitute independent predictors of death. The respiratory rate, as a crucial vital sign with proven prognostic value, guides clinical decisions and should therefore be monitored carefully.¹¹ Thus, only prognostic tools that include respiratory rate as an important component can be useful for risk stratification and the prediction of ICU admission in COVID-19 patients. Our observations indicate that a lower baseline eGFR (before SARS-CoV-2 infection) appears to be an independent prognostic factor and increases the risk of the fatal course of COVID-19. This might suggest that better renal graft function reflected by a higher eGFR might reduce the patient’s vulnerability to a severe course of SARS-CoV-2 infection. A recent analysis demonstrates that mortality risk increases as eGFR decreases in patients with chronic kidney disease,⁶ which indirectly confirms our observations made in the group of KTRs. Our study shows that diabetes mellitus conveys an additional independent risk factor for mortality in patients after renal transplantation with COVID-19. The issue of diabetes as a factor that increases the COVID-19 mortality risk has already been discussed^4,23,24 and the need for insulin has been reported to further worsen the prognosis.²⁵ A meta-analysis demonstrated an association between diabetes mellitus and severe COVID-19, ARDS, and mortality, which is even more pronounced in the younger subgroup (<⁠55 years old).²⁴ A higher prevalence of diabetes mellitus among deceased KTRs was also reported by Azzi et al,¹² but not by Cravedi et al.¹¹ In our study, diabetes mellitus and the volume of compromised lungs on CT imaging constituted independent risk factors for the development of severe ARDS in the course of COVID-19 in KTRs. A clinical assessment based on the above factors may contribute to earlier preparation of advanced oxygen support and may indicate early ICU admission of patients who can benefit from intensive care. Chest CT imaging was performed in every patient admitted to our unit and our study demonstrates its utility for accurately assessing a patient’s respiratory condition and prognosis. It has been already shown that in the general population, compromised lung volume constitutes a risk factor for oxygen use, intubation, and death.²⁶ Another study indicated that disease extent exceeding 25% of the lung parenchyma is related to poor outcomes.²⁷ Our results indicate that the volume of compromised lung parenchyma on admission was associated with the severity of ARDS and mortality (although not independently), which provides a rationale for the routine use of chest CT imaging.

Unlike in previous reports about COVID-19 in KTRs,^11,12 we found that CRP (assessed on admission) was higher in nonsurvivors and was associated with increased risk of severe ARDS, but not as an independent factor. CRP, a valuable and widely used inflammatory parameter, was previously reported to be elevated in patients with COVID-19 and predictive of the need for mechanical ventilation in the general population.²⁸ An elevated CRP level might suggest a COVID-19–related hyperinflammatory state, which might be pathogenetically responsible for clinical deterioration and acute respiratory failure, leading to fatal outcomes.

In our cohort, the KTRs treated with statins seemed to have a lower risk of severe ARDS (P = 0.042) and death from COVID-19 (P = 0.088). In another retrospective study including 2626 patients, in which propensity-score matching was implemented, previous statin use was associated with lower in-patient COVID-19 mortality.²⁹ The potential explanation of the observed benefit from the use of statins might be their anti-inflammatory properties, their effect on endothelial cell function, or their interference with SARS-CoV-2 replication.^29,30 In a multicenter, randomized clinical trial, the hyperinflammatory subtype of ARDS was more successfully treated (improved survival) with simvastatin than with a placebo.³¹ Perhaps statins reduce the risk of the COVID-19–related hyperinflammatory phenotype of ARDS, but current evidence remains limited. Further randomized clinical trials should validate the efficacy of statins in such cases.

Intriguingly, over 70% of hospitalized patients with a kidney allograft experienced AKI, including 15% who were diagnosed with severe AKI (stage 3). The very high incidence of AKI might be a consequence of dehydration, direct viral-induced cytopathic changes, graft rejection, or a hyperinflammatory state. Another study showed that AKI occurred in 52% of COVID-19 patients after kidney transplantation¹¹ and suggested a high risk of renal function impairment in COVID-19.³² In our cohort, all 5 patients who developed severe AKI required renal replacement therapy and ICU admission and had coexisting severe ARDS resulting in multiple organ dysfunction and death. The impact of SARS-CoV-2 infection on the long-term functioning of a graft remains unknown, but early reports suggest that graft function returns to baseline within 1 month of discharge in the majority of patients (62%).³²

The data regarding the course of COVID-19 in the early post-transplant period are still limited and the risk of perioperative in-hospital SARS-CoV-2 infection during transplantation must be considered when qualifying recipients.³³ In our study, 3 KTRs and 4 LTRs were infected with SARS-CoV-2 in the perioperative period; 1 of the KTRs died within 43 days of transplantation. On the other hand, Pascual et al³⁴ reported a 45.8% fatality rate in a group of 24 KTRs diagnosed with COVID-19 within 60 days of transplantation. Of note, the median age in the study group was relatively high (66.5 years) and age was a factor associated with mortality. In sum, immunosuppressive induction therapy, age, comorbidities, a shortage of ICU beds, and the current regional epidemiological situation should always be taken into account, as they might compromise the result of transplantation (including graft and patient survival).

Liver transplant recipients infected with SARS-CoV-2 seem to have a better prognosis than KTRs. A meta-analysis reports a mortality rate of 11.8%,⁹ whereas 2 multicenter studies found a mortality rate ranging between 22% and 25%.^10,13 We presented a case series of 8 patients, of whom only 2 required oxygen supplementation (25%), while 1 died (12.5%) of SARS-CoV-2–induced ARDS. We found relatively mild infection in most patients with liver transplants, in the early postoperative period as well. A multicenter study on 151 patients provided evidence that liver transplantation was not independently associated with COVID-19 mortality, but that the presence of comorbidities and advanced age were responsible for unfavorable outcomes.³⁵ Belli et al¹³ reported better survival of LTRs treated with tacrolimus and therefore discouraged withdrawing this treatment in patients with COVID-19. In our study, 87.5% of the patients were on tacrolimus, while 50% had the dosage reduced during hospitalization. Notably, the LTRs received mycophenolate mofetil significantly less often than the KTRs. Despite withdrawal of mycophenolate mofetil on admission, in almost all cases (except 1 kidney recipient), its immunosuppressive effect could still persist and compromise viral clearance. This could potentially be responsible for the significantly higher mortality in the KTRs compared to the LTRs. However, the analyzed sample was small and such an observation must be treated very cautiously.

Strategies to minimize the risk of SARS-CoV-2 infection in KTRs should be undertaken, because this group of relatively young patients is characterized by a particularly high risk of respiratory failure and death. In our cohort, in which the median age was only 54 years, age was not found to compromise survival. Crucially, all graft recipients should be enrolled for early vaccination against COVID-19, since this relatively small group of patients very frequently develops critical COVID-19–related ARDS, leading to fatal outcomes. The long-term consequences of COVID-19 and its impact on kidney and pulmonary function are yet to be observed in convalescent transplant recipients.

In summary, kidney transplant recipients often experience AKI and ARDS in the course of COVID-19, leading to a substantially high mortality rate. It is absolutely crucial to implement strategies to protect transplant recipients from SARS-CoV-2 infection by prioritizing vaccination in that group. Further studies regarding immunosuppression adjustment and novel antivirals are warranted to improve the prognosis of COVID-19 patients after transplantation. The prognostic factors identified in our study might facilitate risk stratification in patients, especially if there is a shortage of intensive care resources and clinical decision-making is particularly challenging.

Study limitations

The limitations of the present study result from its relatively small sample size and retrospective design. In addition, we focused on a homogeneous group of hospitalized patients after transplantation; therefore, our observations do not reflect the outcomes of nonhospitalized transplant recipients with COVID-19.