Clinical Course, Nosocomial, and Opportunistic Infections Among Kidney Transplant Recipients with COVID-19: A Retrospective Single Center Study

Abstract

Background

We report the results of an observational study, analyzing the clinical course of kidney transplant patients hospitalized for COVID-19 and comparing it with a control to determine if outcomes, nosocomial, and opportunistic infections were different between groups.

Methods

An observational, retrospective, case-control, single-center study, including a group of kidney transplant adults diagnosed with COVID-19, from March 2020 to April 2022. Transplant patients hospitalized for COVID-19 comprised the cases. The control group consisted of non-transplanted adults, without immunosuppressive treatment, hospitalized for COVID-19, and matched by age, sex, and month at diagnosis of COVID-19. Study variables were collected, including demographic/clinical, epidemiologic, clinical/biological at diagnosis, evolutive, and outcome variables.

Results

Fifty-eight kidney transplant recipients were included. Thirty required hospital admission. Ninety controls were included. Transplant recipients had a higher frequency of intensive care unit (ICU) admission, ventilatory support, and death. The relative risk for death was 2.45. When adjusted by baseline estimated glomerular filtration rate (eGFR) and comorbidity, only the risk for opportunistic infection remained high. Variables independently associated with death were dyslipidemia, eGFR at admission, MULBSTA score, and ventilatory support. Pneumonia by Klebsiella oxytoca was the most frequent nosocomial infection. Pulmonary aspergillosis was the most frequent opportunistic infection overall. Pneumocystosis and cytomegalovirus colitis were more frequent among transplant patients. The relative risk for opportunistic infection in this group was 1.88. Baseline eGFR, serum interleukin 6 level, and coinfection were independently associated with it.

Conclusions

Evolutive course of COVID-19 requiring hospitalization in renal transplant recipients was primarily determined by comorbidity and baseline kidney function. At equal comorbidity and renal function, there were no differences in mortality, ICU admission, nosocomial infection, and hospital stay. However, the risk for opportunistic infection remained high.

COVID-19, caused by SARS-CoV-2, was officially declared a pandemic by the World Health Organization in 2020. Despite the development of vaccines and novel antiviral therapies, it has not yet been eradicated [1]. Considering the viral mutation rate with the appearance of new variants with variable profiles of contagiousness, virulence, and resistance to active immunization and antiviral treatment, there is uncertainty regarding its duration and evolution.

Older patients and those affected by multiple comorbidities are at a higher risk of infection-related death [2,3], especially in immunocompromised individuals and those affected by malignant neoplasms [4].

The mortality rate of COVID-19 in kidney transplant recipients has been reported to be 20% to 40% [5], [6], [7], [8], a comparatively greater rate than that reported in the general population (10%-15%) [9], [10], [11]. This difference is mainly explained by the immunocompromised status of these patients; it is considered advisable to temporarily lower immunosuppression therapy intensity in high-risk transplant patients who contract the infection. However, this may result in a higher risk for rejection [1].

In relation to the clinical behavior of COVID-19 in kidney transplant patients, diverse studies have been published. In general, a series of baseline epidemiologic and clinical factors are associated with a higher risk of death, intensive care unit (ICU) admission, and ventilatory support; among them are age, basal comorbidity, being a recipient of a graft from a cadaveric donor [12], [13], [14], [15], [16], time since transplant ≤6 months, respiratory symptoms, pneumonia, and acute kidney injury (AKI) [17], [18], [19], [20], [21]. An analysis from the Spanish Registry of Kidney Transplants published in 2021 [21] found that patients older than 65 years who were recipients of kidney grafts in the 6 months before SARS-CoV-2 infection experienced worse survival and mortality than other transplant patients. Most studies, in this regard, focus on transplant populations only; studies with control groups conformed of non-transplant patients are less numerous. Among the latter, Aziz et al, in a preliminary report published in 2020, found that kidney transplant recipients had relative risks (RRs) for AKI and ICU admission of 15.4% and 34.1%, respectively, compared with RRs of 13.3% and 3.3% in non-transplant patients (P < .001) [11]. Current available data come mainly from registries from the first year of the pandemic when active immunization was not widely available, and certain antiviral therapies were not backed up by solid clinical evidence. Data concerning the rates of nosocomial infections (NI) and opportunistic infections (OI) in this context and in these patients is relatively limited. Both NI and OI are important factors affecting prognosis and hospital stay in the general population.

We report the results of an observational single-center study designed with the following objectives: (1) to analyze the clinical course of kidney transplant patients with diagnosis of COVID-19, those requiring hospitalization, and compare it with a control group of non-transplant patients; and (2) to determine if mortality, ICU admission, need for ventilatory support, NI, OI, hospital, and ICU stay differed between groups and which variables (clinical, epidemiologic, biochemical) were associated with these outcomes.

Patients and Methods

This is an observational, retrospective, case-control, single-center study, including a group of kidney transplant adults with diagnosis of COVID-19 based on suggestive symptomatology and confirmed through SARS-CoV-2 determination by polymerase chain reaction (either from a blood sample or nasopharyngeal swab) who were attended in Hospital Universitari Arnau de Vilanova (HUAV), Lleida, Spain, from March 2020 to April 2022. In this group, patients requiring hospitalization for COVID-19 comprised the case group. A control group was also included, consisting of adults who were non-recipients of solid or hematopoietic organs, without active immunosuppressive treatment (for any cause) in the last 6 months before inclusion in the study, with diagnoses of COVID-19 requiring hospital admission; this group was matched with the cases by age, sex, and month at diagnosis of COVID-19 in a proportion of 1:3.

In both groups, the digital clinical records from HUAV and Primary Health Service were reviewed. The study variables were collected, including demographic-clinical (including Charlson´s comorbidity index), epidemiologic, clinical-biological at diagnosis (including Horowitz index and MULBSTA score), evolutive, and outcomes (coinfection at hospital admission, treatment prescribed for COVID-19, AKI at admission, AKI during follow-up, shock, ICU admission, ventilatory support, renal replacement therapy, frequency of NI and OI, death, cause of death, serum creatinine, and estimated glomerular filtration rate at the end of follow-up and sequels). In transplant patients, de novo positivity for donor-specific antibodies after discharge was recorded.

Coinfection at diagnosis was defined as an infectious disease caused by any pathogenic agent different from SARS-CoV-2 confirmed or suspected at the time of COVID-19 diagnosis.

Nosocomial infection was defined as any infection suspected or confirmed during hospitalization that was not present, even in its incubation stages, at the time of admission.

Opportunistic infection was defined as any infectious disease caused by pathogens that usually do not produce clinically significant infections in immunocompetent individuals.

The AKI definitions used were those established by the Kidney Disease: Improving Global Outcomes guidelines, with severity staging done according to the Acute Kidney Injury Network [22].

The estimated glomerular filtration rate (eGFR) was calculated by the Chronic Kidney Disease Epidemiology Collaboration 2009 formula [23].

Statistical Analysis

Categorical variables were expressed in percentages, and continuous variables were expressed as mean and SD for normal distribution variables or as median and IQR. Statistical significance of the differences when comparing categorical variables was calculated by the Chi-square or the Fisher test. When comparing categorical with continuous variables, the t test or Mann-Whitney U test was used. Multivariable analysis was performed using a step-by-step regression logistics model. Survival analysis was performed with Kaplan-Meier curves, and the log-rank test was used to determine statistical significance. The statistical programs used were SPSS version 20.0 (IBM SPSS, Inc, Armonk, NY, United States) and MedCalc (MedCalc Software, Ltd, Ostend, West Flanders, Belguim).

The study was reviewed and approved by the bioethical committee of the center.

Results

From March 2020 to March 2022, 58 kidney transplant recipients were diagnosed with COVID-19. Most of the patients had received only one graft (94.8%). At diagnosis, 6 patients had active neoplasms (2 cases of breast cancer, 2 of skin cancer, 1 case of lung cancer, and 1 of a GIST), of which 2 were on active oncological treatment. Thirty-three patients, at diagnosis, had received at least 1 dose of a vaccine against SARS-CoV-2, of which, 39.7% had received 3 doses. Of the 58 patients with COVID-19, 30 (50.84%) required hospital admission (Fig 1 ).

Fig 1.

NI, nosocomial infection; OI, opportunistic infections.

In a proportion of 1:3 matched by age, sex, and month of diagnosis, 90 controls were included.

Among the cases (transplant recipients), 10 (33.3%) had concomitant infection at admission, 10 (33.33%) required ICU admission, 6 (20%) developed NI, 8 (26.6%) OI, and 9 (30%) died. In the control group, 23 (25.8%) had coinfection at admission, 9 (10.2%) were admitted to ICU, 8 (9%) developed NI, 3 (3.4%) OI, and 11 (12.6%) died (Fig 1).

Fig 2 represents the frequencies for both diagnosis of and hospitalization for COVID-19 in kidney transplant populations between March 2020 and March 2022. Peaks of hospitalization frequency for COVID-19 were observed between epidemiologic months 6 to 8 (June and July 2020), 11 to 12 (November and December 2020), and 24 to 25 (December 2021 to January 2022) of the pandemic.

Fig 2.

New cases and hospitalizations by COVID-19 in kidney transplant population per month. Month 1 corresponds to January 2020. Log Rank = 1.22, P = .25.

At diagnosis, all transplant patients received tacrolimus as an immunosuppressive maintenance treatment. In all patients, antiproliferative agents were suspended (mammalian target of rapamycin [mTOR] inhibitors or mycophenolic acid analogs (MAA)), tacrolimus dose was adjusted to target levels of 3 to 5 ng/mL, and prednisone was increased to 10 to 20 mg/d. In those patients who required dexamethasone treatment, prednisone was suspended.

Kidney transplant patients with COVID-19 admitted to the hospital had less transplant age, worse baseline kidney function parameters, received higher doses of prednisone as maintenance therapy, less frequency of mTOR inhibitors prescription, worse kidney function parameters at diagnosis, higher reactive C protein levels, MULBSTA score, and lower Horowitz index and serum hemoglobin. This group of patients experienced longer times from symptom onset to diagnosis (TbStDx) (Table 1 ).

Table 1.

Demographics, Clinical Variables, and Hospitalization by COVID-19 in Transplant Patients

	Hospitalized n = 30	Non-hospitalized n = 28	P Value
Age	61.93 ± 14.11	56.57 ± 14.2	.15
Sex n (%)
Male	18 (60)	21 (75)	.21
Female	12 (40)	7 (25)
Transplant age (mo)	50 (187-10)	68 (333-9)	.021†
Baseline creatinine (mg/dL)	1.5 (3.2-0.96)	1.15 (15-0.6)	.003†
Baseline eGFR (mL/min)	44.33 ± 16.48	67.72 ± 24.63	< .001†
Hypertension n (%)	28 (93.3)	24 (85.7)	.41
Active smoking n (%)	2 (6.7)	2 (50)	.66
Dyslipidemia n (%)	22 (73.3)	26 (92.9)	.051
Diabetes mellitus n (%)	18 (60)	10 (35.7)	.056
Obesity n (%)	4 (13.3)	5 (17.9)	.45
Ischemic heart disease n (%)	6 (20)	4 (14.3)	.41
Peripheral vascular disease n (%)	5 (16.7)	6 (21.4)	.44
Cerebral vascular disease n (%)	2 (6.7)	5 (17.9)	.18
COPD n (%)	3 (10)	0 (0)	.13
Active neoplasm	4 (13.3)	2 (7.1)	.36
Vaccine doses received n (%)
None	14 (46.7)	11 (39.3)	.33
1	1 (3.3)	0 (0)
2	6 (20)	3 (10.7)
3	9 (30)	14 (50)
Anti-SARS-CoV-2 titer at dx (BAU/mL)	0 (2081.2-0)	0 (2084-0)	.004†
Type of vaccine
ARNm Pfizer	4 (13.3)	2 (7.1)	.45
ARNm Moderna	12 (40)	15 (53.6)
Immunosuppressive treatment at diagnosis n (%)
Prednisone	30 (100)	24 (85.7)	.048†
AMMF	26 (87.7)	19 (67.9)	.08
Anti-calcineurinics	30 (100)	28 (100)	N/A
mTOR inhibitors	2 (6.7)	8 (28.6)	.03†
Prednisone dose (mg/d)	5 (5-5)	5 (25-0)	.15
TbStDx (d)*	4 (10-0)	0 (4-0)	< .001†
Charlson comorbidity index	4 (11-2)	4 (11-2)	.023†
Creatinine at dx (mg/dL)	1.79 (1.34-0.8)	1.21 (2.73-0.75)	.004†
eGFR at dx (mL/min)	39.14 ± 20.27	63.58 ± 24.7	.001†
IL-6 (pg/mL)	39.85 (1682-2.6)	10.2 (33.4-3.1)	.028†
Reactive C protein (mg/l)	79.25 (306-1.3)	4.8 (21.5-1)	< .001†
Leukocyte count (cells*10³/mm³)	7.55 ± 4.82	7.01 ± 2.73	.65
Hemoglobin (gr/dL)	12.48 ± 2.27	14.45 ± 2.01	.003†
Platelet count (cells*10³/mm³)	197.66 ± 127.31	198.35 ± 58.66	.98
Lactate dehydrogenase (U/L)	632 ± 227.76	412.33 ± 118.62	.056
D-dimer (ng/mL)	237.5 (3638-100)	208 (788-130)	.47
MULBSTA score	11 (19-4)	4 (8-0)	< .001†
Horowitz index	274.2 ± 109.51	396 ± 39.59	.04†
Logistic Regression Model.
	B	E.T.	P Value	Exp(B)	IC 95%
Transplant age (mo)	–0.04	0.02	.046	0.96	0.92-0.99
TbStDx (d)	1.5	0.63	.018	4.52	1.29-15.83
Baseline eGFR (mL/min)	–0.071	0.033	.029	0.93	0.87-0.993

R² Nagelkerke: 0.79.

MAA, mycophenolic acid analogs…; B, unstandarddized regression weight COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; dx, diagnosis; E.T Standard Error; CI, confidence interval; IL, interleukin; mTOR, mammalian target of rapamycin; TbStDx, time between symptom onset to diagnosis.

^⁎TbStDx.

^†P < .05.

The circulating anti-SARS-CoV-2 IgG antibody titer among transplant patients was lower (Table 1).

Upon multivariate analysis, the variables independently associated with hospital admission were transplant age, TbStDx, and the eGFR at diagnosis (Table 1).

Table 2 summarizes the clinical, epidemiologic, biochemical, and evolutive features of both cases and controls. Cases presented significantly higher cardiovascular comorbidity and Charlson index scores and lower baseline kidney function. Transplant patients had lower anti-SARS-CoV-2 IgG antibody titer and TbStDx with predominating respiratory symptomatology. At admission, transplant patients had lower eGFR, lower platelet count, and greater AKI frequency, regardless of its severity. These patients received tocilizumab and remdesevir more frequently.

Table 2.

Demographics, Clinical Features, and Outcomes Between Transplant and Non-transplant Recipients Hospitalized for COVID-19

	Cases n = 30	Controls n = 90	P Value
Age	61.93 ± 14.11	63.59 ± 13.79	.57
Sex n (%)
Male	18 (60)	52 (57.8)	.83
Female	12 (40)	38 (42.2)
Baseline creatinine (mg/dL)	1.5 (3.2-0.96)	0.84 (8.9-0.35)	.001†
Baseline eGFR (mL/min)	44.33 ± 16.48	78.35 ± 25.89	< .001†
Hypertension n (%)	28 (93.3)	50 (56.2)	< .001†
Active smoking n (%)	2 (6.7)	6 (6.7)	.64
Dyslipidemia n (%)	22 (73.3)	36 (40)	.002†
Diabetes mellitus n (%)	18 (60)	21 (23.3)	< .001†
Obesity n (%)	4 (13.3)	19 (21.1)	.34
Ischemic heart disease n (%)	6 (20)	9 (10)	.15
Peripheral vascular disease n (%)	5 (16.7)	13 (14.4)	.76
Cerebral vascular disease n (%)	2 (6.7)	11 (12.2)	.39
COPD n (%)	3 (10)	16 (17.8)	.31
Type of vaccine n (%)
ARNm Pfizer	4 (13.3)	23 (25.55)	.004†
ARNm Moderna	12 (40)	10 (11.11)
Jcovden (Janssen)	0 (0)	3 (3.33)
ChAdOx1 (AstraZeneca)	0 (0)	1 (1.1)
Vaccine doses received n (%)
None	14 (46.7)	55 (61.1)	.50
1	1 (3.3)	5 (5.6)
2	6 (20)	11 (12.2)
3	9 (30)	18 (20)
4	0 (0)	1 (1.1)
Anti-SARS-CoV-2 titer at dx (BAU/mL)	0 (2081.2-0)	218 (2081-0)	.01†
TbStDx (d)*	4 (10-0)	5 (45-0)	.04†
ICU admission n (%)	10 (33.3)	9 (10.2)	.003†
Charlson comorbidity index	4 (11-2)	2 (10-0)	.001†
Creatinine at dx (mg/dL)	1.79 (9.83-0.84)	0.86 (8.2 -0.31)	.001†
eGFR at dx (mL/min)	39.14 ± 20.27	74.31 ± 29.56	.02†
IL-6 (pg/mL)	39.85 (1682-2.6)	29.6 (744.7-1)	.266
Reactive C protein (mg/l)	102.99 ± 78.82	107.77 ± 91.3	.93
Leukocyte count (cells*10³/mm³)	6.88 (27.79-2.99)	7.54 (27.25-2.76)	.15
Hemoglobin (gr/dL)	12.48 ± 2.27	12.95 ± 1.96	.27
Platelet count (cells*10³/mm³)	170 (777-50)	220.5 (724-74)	.008†
Lactate dehydrogenase (U/L)	632 ± 227.76	619.58 ± 299.41	.84
D-dimer (ng/mL)	237.5 (3638-100)	272 (46768-80)	.55
MULBSTA score	10.76± 3.74	9.21 ± 4.17	.08
Horowitz index	274.20 ± 109.51	271.07 ± 93.97	.23
AKI at admission n (%)	12 (40)	16 (18)	.014†
AKIN n (%)
1	6 (42.9)	10 (62.5)	.35
2	4 (28.6)	3 (18.8)
3	2 (14.3)	3 (18.8)
Shock n (%)	6 (20)	5 (5.7)	.02†
Ventilatory support/O2 therapy n (%)	15	28
High flow nasal cannula O2	6 (20)	24 (27)	< .001†
Non-invasive mechanical ventilation	3 (10)	4 (4.5)
Invasive mechanical ventilation	6 (20)	0 (0)
AKI during follow-up n (%)	8 (26.7)	7 (7.9)	.007†
AKIN during follow-up n (%)
1	1 (10)	4 (57.1)	.16
2	3 (30)	1 (14.3)
3	4 (40)	2 (28.6)
CVVHD/HD n (%)	3 (10)	2 (2.3)	.07
Concomitant infection n (%)	10 (33.3)	23 (25.8)	.42
Respiratory	5 (50)	16 (69.5)	.15
Urinary	5 (50)	3 (13)
Skin	0 (0)	1 (4.4)
Abdominal	0 (0)	2 (8.7)
Joint	0 (0)	1 (4.4)
COVID-19 pharmacologic therapy n (%)
Tocilizumab	15 (50)	24 (28.2)	.03†
N° of doses	1.13 ± 0.5	0.87 ± 0.5	.1
Dexamethasone	27 (90)	66 (75)	.06
Initial dose (md/d)	6 (15-0)	7.2 (8-0)	.004†
Remdesivir	11 (36.7)	10 (11.4)	.002†
Hydroxychloroquine	3 (10)	3 (3.4)	.15
Nosocomial infection n (%)	6 (20)	8 (9)	.10
Respiratory	4 (66.6)	4 (50)	.2
Septicemia	2 (33.4)	2 (25)
Catheter related septicemia	0 (0)	1 (12.5)
Skin	0 (0)	1 (12.5)
Opportunistic infection n (%)	8 (26.7)	3 (3.4)	.001†
Pulmonary aspergillosis	3 (37.5)	2 (66.6)	.001†
Candidiasis	1 (12.5)	1 (33.4)
Invasive CMV	2 (25)	0 (0)
Pneumocystosis	2 (25)	0 (0)
Final creatinine (mg/dL)	1.4 (4.8-0.56)	0.77 (8-0.31)	.001†
Final eGFR (mL/min)	50.6 ± 22.48	78.85 ± 29.73	< .001†
Hospitalization time (d)	11.5 (61-6)	7 (9-1)	.001†
ICU time (d)	17.7 ± 9.84	14.4 ± 17.76	.61
Exitus n (%)	9 (30)	11 (12.6)	.02†
Sequels n (%)	2 (9.5)	7 (9)	.61

AKI, acute kidney injury; AKIN, Acute Kidney Injury Network; CMV, cytomegalovirus; COPD, chronic obstructive pulmonary disease; CVVHD, continuous veno-venous hemodialysis; eGFR, estimated glomerular filtration rate; HD, hemodialysis; ICU, intensive care unit; IL, interleukin; TbStDx, time between symptom onset to diagnosis.

^⁎TbStDx.

^†P < .05.

Compared with the controls, transplant recipients had worse outcomes, with higher frequencies of AKI during follow-up, shock, ICU admission, and the need for invasive and non-invasive ventilatory support. There were no significant differences in ICU stay, but transplant patients had longer hospital stays. Opportunistic infection and mortality were greater. The RR for death in transplant patients was 2.45 (confidence interval 95%; 1.12-5.34).

At the end of the follow-up, 28 (93.3%) cases featured functional grafts. None developed de novo donor-specific antibodies.

Seeing that comorbidity and baseline eGFR were worse among the cases, we matched the cases and controls by baseline eGFR, Charlson index, frequency of high blood pressure, diabetes, and dyslipidemia in a proportion of 1:1 to obtain 20 matched pairs (Table 3 ).

Table 3.

Demographics, Clinical Features, and Outcomes Between Transplant and Non-transplant Recipients Hospitalized for COVID-19 after Adjustment for Baseline eGFR and Comorbidities

	Cases n = 20	Controls n = 20	P Value
Transplant age (mo)	63 ± 47.85	-	-
TbStDx (d)*	4.3 ± 3.01	5.47 ± 3.93	.17
ICU admission n (%)	6 (30)	2 (10)	.11
Creatinine at dx (mg/dL)	2.19 (9.83-0.84	1.41 (8.9-0.5)	.38
eGFR at dx (mL/min)	44.5 ± 21.63	44.59 ± 23.69	.88
IL-6 (pg/mL)	91.05 (1682-2.6)	33.4 (210-4.4)	.38
Reactive C protein (mg/l)	96.7 ± 80.12	163.06 ± 109.56	.03†
Leukocyte count (cells*10³/mm³)	8.35 ± 5.54	10.86 ± 6.14	.18
Hemoglobin (gr/dL)	12.96 ± 2.27	12.64 ± 2.12	.89
Platelet count (cells*10³/mm³)	212.3 ± 153.02	229.5 ± 95.93	.67
Lactate deshidrogenase (U/L)	620.65 ± 254.45	605.28 ± 412.41	.12
D-dimer (ng/mL)	237 (2726-100)	352 (46768-130)	.06
MULBSTA score	9.9 ± 3.66	10.23 ± 4.22	.38
Horowitz index	287.84 ± 100.33	269.38 ± 106.56	.62
AKI at admission n (%)	7 (35)	10 (50)	.33
Shock n (%)	3 (15)	2 (10)	.63
Ventilatory support/O2 therapy n (%)	10	7
High flow nasal cannula O2	4 (40)	6 (85.71)	.18
Non-invasive mechanical ventilation	3 (30)	1 (14.28)
Invasive mechanical ventilation	3 (30)	0 (0)
AKI during follow-up n (%)	5 (25)	6 (30)	.5
CVVHD/HD n (%)	1 (5)	2 (10)	.54
Concomitant infection n (%)	6 (30)	6 (30)	.63
COVID-19 pharmacologic therapy n(%)
Tocilizumab	10 (50)	3 (15)	.02†
Dexamethasone	17 (85)	13 (65)	.14
Initial dose (md/d)	6 (15-0)	6 (7.5-0)	.61
Remdesivir	8 (40)	3 (15)	.07
Hydroxychloroquine	3 (15)	1 (5)	.29
Nosocomial infection n (%)	4 (20)	3 (15)	.67
Opportunistic infection n (%)	5 (25)	1 (5)	.04†
Final creatinine (mg/dL)	1.4 (4.8-0.7)	1.2 (8-0.7)	.37
Final eGFR (mL/min)	53.35 ± 23.83	52.47 ± 29.8	.3
Hospitalization time (d)	13.4 ± 7.01	11.65 ± 11.43	.05
Exitus n (%)	6 (30)	7 (35)	.73
Sequels n (%)	1 (7.1)	0 (0)	.32

AKI, acute kidney injury; CVVHD, continuous veno-venous hemodialysis; eGFR, estimated glomerular filtration rate; HD, hemodialysis; ICU, intensive care unit; IL, interleukin; TbStDx, time between symptom onset to diagnosis.

^⁎TbStDx.

^†P < .05.

When adjusted by these variables, at admission, transplant patients presented higher levels of interleukin (IL) 6, lower levels of D-dimer, and higher frequencies of tocilizumab prescription. Opportunistic infection frequency continued to be significantly higher. No differences were observed in mortality, ICU admission, hospital stay, or NI (Table 3). The relative risk for death was 0.89 (CI 95%; 0.44-1.77), losing its statistical significance.

Mortality

Figure 3 shows the survival curves for the cases vs controls without adjustment by comorbidity and baseline eGFR and after adjustment. Because hospital/ICU stays were not different between groups (time to death), even without adjustment, the differences were not statistically significant (Figure 3A; Log Rank: 1.22; P = .25).

Fig 3.

Survival between transplant and non-transplant recipients hospitalized for COVID-19. (A) Kaplan-Meier survival curve in unadjusted group. (B) Survival curve after adjustment for baseline estimated glomerular filtration rate, comorbidities, and Charlson score. Tx, transplant. Log Rank = 0.00, P = .99.

Figure 4 shows the causes of death among the study groups. There were no significant differences between the groups. Overall, the most frequent cause of death was due to SARS-CoV-2 infection or NI/OI.

Fig 4.

Causes of death in the study groups. Tx, transplant. P = .075.

Variables associated with death in all the study groups are shown in Table 4 . Age, hypertension, dyslipidemia, diabetes, brain vascular disease, Charlson index, and baseline eGFR were associated with higher mortality. Patients with lower eGFR, AKI, greater LDH, D-dimer, MULBSTA score, and lower Horowitz index had higher mortality at admission. The presence of coinfection at admission, NI, OI, ICU admission, invasive and non-invasive ventilatory support, septic shock, and the need for renal replacement therapy were associated significantly with death (Table 4). On multivariate analysis, variables independently associated with death were dyslipidemia, eGFR at admission, higher MULBSTA score, and invasive and non-invasive ventilatory support (Table 4).

Table 4.

Demographics, Clinical Variables, and Death in the Study Groups

	Exitus n = 20	Non-exitus n = 100	P Value
Age	70.15 ± 14.16	61.96 ± 13.47	.016†
Sex n (%)
Male	10 (50)	59 (59)	.47
Female	10 (50)	41 (41)
Group n (%)
Tx patients	9 (45)	21 (21)	.029†
Non-Tx patients	11 (55)	79 (79)
Transplant age (mo)	48 (187-19)	51 (164-10)	.92
Baseline creatinine (mg/dL)	1.4 (2.5-0.49)	0.88 (8.9-0.35)	.003†
Baseline eGFR (mL/min)	49.89 ± 23.35	73.74 ± 27.33	< .001†
Hypertension n (%)	18 (90)	59 (59)	.014†
Active smoking n (%)	0 (0)	6 (6)	.25
Dyslipidemia n (%)	17 (85)	41 (41)	< .001†
Diabetes mellitus n (%)	11 (55)	28 (28)	.024†
Obesity n (%)	4 (20)	18 (18)	.88
Ischemic heart disease n (%)	4 (20)	11 (11)	.29
Peripheral vascular disease n (%)	2 (10)	16 (16)	.46
Cerebral vascular disease n (%)	6 (30)	6 (6)	.001†
COPD n (%)	3 (15)	16 (16)	.86
Active neoplasm	3 (15)	9 (9.3)	.44
Vaccine doses received n (%)
None	9 (45)	58 (58)	.6
1	1 (5)	5 (5)
2	5 (25)	12 (12)
3	5 (25)	21 (21)
4	0 (0)	1 (1)
Anti-SARS-CoV-2 titer at dx (BAU/mL)	0	0 (2081.2-0)	.12
TbStDx (d)*	3.5 (10-0)	5 (45-0)	.04†
Charlson comorbidity index	5 (8-0)	3 (11-0)	.001†
Creatinine at dx (mg/dL)	2.1 ± 1.18	2.26 ± 8.34	.08
eGFR at dx (mL/min)	37.95 ± 22.91	70.48 ± 29.93	< .001†
IL-6 (pg/mL)	174.9 (1682-40.8)	23.2 (1004.6-1)	.002†
Reactive C Protein (mg/l)	150.01 ± 99.23	96.79 ± 83.29	.013
Leukocyte count (cells*10³/mm³)	7.8 (27.25-3.27)	7.05 (27.79-2.76)	.32
Hemoglobin (gr/dL)	12.17 ± 2.46	12.95 ± 4.46	.12
Platelet count (cells*10³/mm³)	196.5 (381-112)	202 (777-50)	.88
Lactate dehydrogenase (U/L)	820.21 ± 332.26	590.9 ± 257.05	0.004†
D-dimer (ng/mL)	346.5 (46768-148)	251 (6574-80)	.008†
MULBSTA score	13.35 ± 2.37	8.94 ± 4	< .001†
Horowitz index	202.62 ± 115.06	285.41 ± 89.6	.002†
AKI at admission n (%)	13 (65)	15 (15)	< .001†
AKIN n (%)
1	4 (30.8)	12 (80)	.013†
2	4 (30.8)	3 (20)
3	5 (38.4)	0 (0)
ICU admission	9 (45)	10 (10)	.001†
Shock n (%)	8 (40)	3 (3)	< .001†
Ventilatory support/O2 therapy n (%)	16	26
High flow nasal cannula O2	9 (56.2)	20 (76.9)	< .001†
Non-invasive mechanical ventilation	3 (18.8)	4 (15.4)
Invasive mechanical ventilation	4 (25)	2 (7.7)
AKI during follow-up n (%)	9 (45)	6 (6)	< .001†
AKIN during follow-up n (%)
1	1 (11.1)	4 (70)	.02†
2	2 (22.2)	2 (30)
3	6 (66.7)	0 (0)
CVVHD/HD n (%)	3 (15)	2 (2)	.01†
Concomitant infection n (%)	11 (55)	22 (22)	.003†
Respiratory	5 (45.45)	17 (77.37)	.005†
Urinary	4 (36.37)	3 (13.63)
Skin	0 (0)	1 (4.5)
Abdominal	1 (9.09)	1 (4.5)
Joint	1 (9.09)	0 (0)
COVID-19 pharmacologic therapy n (%)
Tocilizumab	9 (45)	29 (29)	.29
N° of doses	1 ± 0.63	0.94 ± 0.48	.56
Dexamethasone	15 (75)	77 (77)	.66
Initial dose (md/d)	6 (15-0)	6 (8-0)	.96
Remdesivir	3 (15)	18 (18)	.7
Hydroxychloroquine	2 (10)	4 (4)	.27
Nosocomial infection n (%)	5 (25)	9 (9)	.04†
Opportunistic infection n (%)	6 (30)	5 (5)	.001†
Final creatinine (mg/dL)	1.87 (4.8-0.56)	0.8 (5.8-0.31)	.003†
Final eGFR (mL/min)	42.44 ± 27.74	77.06 ± 27.91	< .001†
Hospitalization time (d)	12 (34-1)	7 (9-1)	.02†
ICU time (d)	17.75 ± 8.44	14.92 ± 17.13	.27
Functioning graft at death n (%)	7 (77.8)	21 (100)	.08
Logistic Regression Model.
	B	E.T.	P Value	Exp(B)	IC 95%
Dyslipidemia	–4.77	2.18	.029	0.008	0.00-0.61
eGFR at dx (mL/min)	–0.051	0.025	.041	0.95	0.9-0.99
MULBSTA	0.49	0.2	.016	1.63	1.09-2.43
Ventilatory support	1.32	0.64	.041	3.75	1.05-13.34

R² Nagelkerke: 0.71.

AKI, acute kidney injury; AKIN, Acute Kidney Injury Network; B, …; COPD, chronic obstructive pulmonary disease; CVVHD; dx, diagnosis; eGFR, estimated glomerular filtration rate; E.T stantdard Error, …; HD, hemodialysis; ICU, intensive care unit; CI, confidence interval; IL, interleukin; TbStDx, time between symptom onset to diagnosis; Tx, transplant.

^⁎TbStDx.

^†P < .05.

Among transplant patients with coinfection at admission, a cause for infection was identified in 6 of 10 cases; in the control group, 6 of 9 cases included identified causes for infection. No significant differences were observed between groups (Table 5 ). Urinary and respiratory tract infections were the most frequent coinfections at admission (Table 2); the pathogens isolated in most cases were Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa.

Table 5.

Etiology of Coinfection at COVID-19 Diagnosis and Transplant Versus Non-transplant Patients

Pathogen	TX patients n (%)	Non-Tx patients n (%)	P Value
Mycobacterium tuberculosis	0 (0)	1 (16.66)	.06
Escherichia coli	2 (33.34)	0 (0)
Haemophylus influenzae	0 (0)	1 (16.66)
Klebsiella pneumoniae	1 (16.66)	0 (0)
Pseudomonas aeruginosa	1 (16.66)	1 (16.66)
Staphylococcus aureus	1 (16.66)	1 (16.66)
Streptococcus pyogenes	0 (0)	1 (16.66)
Herpes simplex virus	0 (0)	1 (16.66)
Legionella pneumophila	1 (16.66)	0 (0)

Tx, transplant.

Nosocomial Infection

The nosocomial infection diagnosed with the most frequency was pneumonia (Table 2), and the most frequent etiologic agent was Klebsiella oxytoca.

Table 6 shows the variables significantly associated with NI. Among patients with NI, the baseline eGFR was lower, they had more frequency of coinfection, higher levels of IL-6, AKI, MULBSTA score, ICU admission, ventilatory support, mortality, and hospital stay. The RR for NI in transplant patients was 1.66 CI 95%; 0.6-4.5).

Table 6.

Variables Significantly Associated with NIs in the Study Groups

	NI n = 14	Non-NI n = 106	P Value
Baseline creatinine (mg/dL)	1.5 (8.9-0.6)	0.95 (8-0.35)	.03*
Baseline eGFR (mL/min)	56.79 ± 27.1	71.77 ± 27.88	.04*
IL-6 (pg/mL)	89.3 (1682-3.2)	24.2 (1004.6-1)	.01*
D-dimer	412 (46768-168)	251.5 (19313-80)	.01*
MULBSTA score	12.84 ± 3.97	9.21 ± 3.93	.002*
ICU admission	9 (64.3)	10 (9.43)	< .001*
Shock n (%)	6 (42.9)	5 (4.71)	< .001*
Ventilatory support/O2 therapy n (%)
High flow nasal cannula O2	4 (28.6)	26 (76.47)	< .001*
Non-invasive mechanical ventilation	0 (0)	7 (20.59)
Invasive mechanical ventilation	5 (35.7)	1 (2.94)
AKI during follow-up n (%)	8 (57.1)	7 (6.7)	< .001*
CVVHD/HD n (%)	3 (21.4)	2 (1.9)	.001*
Concomitant infection n (%)	8 (57.1)	25 (23.58)	.009*
Final creatinine (mg/dL)	2.02 ± 2.22	1.15 ± 0.96	0.01*
Hospitalization time (d)	25 (90-6)	7 (35-1)	< .001*
ICU time (d)	25.44 ± 16.11	8.36 ± 4.94	.004*

AKI, acute kidney injury; CVVHD, continuous veno-venous hemodialysis; eGFR, estimated glomerular filtration rate; HD, hemodialysis; ICU, intensive care unit; IL, interleukin; NI, nosocomial infection.

^⁎P < .05.

Transplant recipients with NI received significantly higher doses of MAA (953.33 ± 114.31 vs 691.43 ± 307.08 mg in 24 hours; P = .003).

Among transplant patients who developed NI, 5 of 6 microbiological isolations were documented; among controls, 2 of 8 were documented. Among transplant recipients, there was a significantly higher frequency of infection by Legionella pneumophila, P. aeruginosa, Stenotrophomona maltophilia, and Staphylococcus epidermidis (Table 7 ).

Table 7.

Etiology of Nosocomial Infection and Transplant Versus Non-transplant Patients

Pathogen	Cases n (%)	Controls n (%)	P Value
Klebsiella oxytoca	1 (20)	1 (50)	.04*
Stenotrophomona maltophilia	1 (20)	0 (0)
Legionella pneumophila	1 (20)	0 (0)
Pseudomonas aeruginosa	1 (20)	0 (0)
Staphylococcus aureus	0 (0)	1 (50)
Streptococcus epidermidis	1 (20)	0 (0)

^⁎P < .05.

Opportunistic Infection

Table 8 summarizes the variables associated with OI during the study period. Patients with OI were more likely to be diabetic; as in those with NI, their evolutive course was worse, with a higher frequency of ICU admission, shock, renal replacement therapy, mortality, and hospital stay. Upon multivariate analysis, baseline eGFR, serum IL-6 level, and the presence of coinfection were independently associated with OI (Table 8).

Table 8.

Variables Significantly Associated with OIs in the Study Groups

	OI n = 11	Non-OI n = 109	P Value
Group n (%)
Tx patients	8 (72.7)	22 (20.18)	< .001*
Non-Tx patients	3 (27.3)	87 (79.81)
Baseline eGFR (mL/min)	42.06 ± 15.65	72.86 ± 27.55	< .008*
Diabetes mellitus n (%)	8 (72.7)	31 (28.44)	.003*
eGFR at dx (mL/min)	36.28 ± 19.87	68.24 ± 30.94	.01*
IL-6 (pg/mL)	135.4 (1682-25)	23.3 (744.7-1)	.001*
MULBSTA score	13.3 ± 3.88	9.27 ± 3.94	.003*
AKI at admission n (%)	6 (54.5)	22 (20.18)	.01*
ICU admission	6 (54.5)	13 (11.92)	< .001*
Shock n (%)	6 (54.5)	5 (4.58)	< .001*
Ventilatory support/O2 therapy n (%)
High-flow nasal cannula O2	4 (50)	26 (74.28)	.001*
Non-invasive mechanical ventilation	0 (0)	7 (20)
Invasive mechanical ventilation	4 (50)	2 (5.72)
AKI during follow-up n (%)	6 (54.5)	9 (8.25)	< .001*
CVVHD/HD n (%)	2 (18.2)	3 (2.75)	.01*
Concomitant infection n (%)	6 (54.3)	27 (24.77)	.03*
COVID-19 pharmacologic therapy n (%)
Remdesevir	5 (45.5)	16 (15)	.01*
Final eGFR (mL/min)	45.35 ± 28.45	74.2 ± 29.59	.003*
Hospitalization time (d)	19 (61-7)	7 (90-1)	.001*
ICU time (d)	25 ± 7.92	12.21 ± 14.58	.02*
Logistic Regression Model.
	B	E.T.	P Value	Exp(B)	IC 95%
Basal eGFR (mL/min)	–0.097	0.042	.02	0.9	0.83-0.98
IL-6 (pg/mL)	0.005	0.002	.021	1	1-1.01
Coinfection at admission	–3.08	1.34	.02	0.046	0.003-0.63

R² Nagelkerke: 0.70.

AKI, acute kidney injury; B,unstandarddized regression weight; COPD, chronic obstructive pulmonary disease; CVVHD; eGFR, estimated glomerular filtration rate; E.T., standard Error; HD, hemodialysis; ICU, intensive care unit; CI, confidence interval; IL, interleukin; OI, opportunistic infection; Tx, transplant.

^⁎P < .05.

Overall, the most frequent OI was pulmonary aspergillosis, mainly by Aspergillus niger. Among transplant recipients, infection by Pneumocystis jirovecii and cytomegalovirus (CMV) was more frequent (Table 2). No significant differences were observed in transplant age and maintenance immunosuppressive therapy.

The RR for OI in transplant recipients, without adjustment for baseline eGFR and comorbidity, was 8 (IC 95%; 2.26-28.22). After adjusting for the aforementioned factors, transplant patients still had significant RR for OI, 1.88 (IC 95%: 1.12-3.17). In this group, the mortality for pulmonary aspergillosis was 100%.

Discussion

The results of this study showed that the clinical course and evolution of kidney transplant patients affected by COVID-19 and requiring hospital admission are generally worse, with higher mortality and ICU admission and ventilatory support than non-transplant patients. After adjustment for comorbidity and baseline eGFR, these differences lost their statistical significance, except for OI (Table 3). At equivalent magnitudes of comorbidity and kidney function, there were no differences between transplant and non-transplant patients. These findings are in accordance with those of Chavarot et al [24], who found no differences in patient survival and the composite survival/ICU admission between groups after adjusting for comorbidities and renal function. However, in practice, organ transplant recipients usually have substantial comorbidity that traces back to the pretransplant stage. This comorbidity is mainly determined by chronic kidney disease and its cause (which frequently includes high blood pressure, diabetes, and dyslipidemia). In addition, immunosuppressive therapy prescribed to prevent rejection, besides its effects on the immune response, is associated with adverse metabolic effects, with a higher risk for new-onset hypertension, diabetes, and dyslipidemia [25], [26], [27] and, consequently, greater cardiovascular morbidity.

Among kidney transplant recipients, the condition that is undoubtedly associated with better outcomes does not require hospitalization for COVID-19. In this case, immunosuppressive treatment has detrimental effects on the immunogenicity of the vaccines against SARS-CoV-2 [28]. In our study, patients with a higher IgG antibody titer against SARS-CoV-2 were hospitalized less frequently. Transplant patients who were not admitted received a lower prednisone dose and had older transplant age and, consequently, less intense immunosuppression. In addition, baseline eGFR was better, which is associated with a more robust antibody response after immunization [29], and they were receiving mTOR inhibitors more frequently (Table 1). The latter have been correlated with better antibody responses after vaccination for a diversity of infective agents [30]; furthermore, they are prescribed for controlling recurrent chronic opportunistic infections, such as CMV and BK virus [31], [32], [33]. This explains our findings on the multivariate analysis (Table 1).

Dyslipidemia, lower eGFR at admission, higher MULBSTA score, and the need for ventilatory support were independently associated with death (Table 4).

Patients treated for SARS-CoV-2 infection are at risk for developing OI because of the anti-inflammatory therapy used to treat the cytokine storm that characterizes the severe forms of the infection. Corticoid treatment and IL-6 blockade, despite improving the course of the infection by reducing inflammation and tissue damage, impair immune responses against other pathogens, such as mycobacteria, fungi, and protozoa [34,35]. Pulmonary aspergillosis, rhino-cerebral mucormycosis, candidemia, and other infectious diseases typically found in immunocompromised patients have been reported in COVID-19 patients without these conditions [35,36]. Nutritional status, comorbidity, and corticoid and tocilizumab dosage are associated with a higher risk for OI. In our study, when the groups were analyzed as a whole, pulmonary aspergillosis was the most frequent OI. The variables independently associated with OI were high levels of IL-6, coinfection, and lower eGFR (Table 8). Higher IL-6 levels were associated with greater frequency of tocilizumab prescription; however, when comparing patients who developed OI with those who did not, no significant differences were observed in the doses of dexamethasone and tocilizumab (Table 8). When the analysis was done, separating the patients into transplant and non-transplant groups, kidney recipients with OI received tocilizumab more frequently (Table 3). Higher IL-6, besides indicating a more intense inflammatory response against SARS-CoV-2, was associated with coinfection at diagnosis (predominantly bacterial), indicating a greater frequency of antibiotic therapies of variable spectrums. Wide-spectrum antibiotic therapy is a predisposing factor for OI, especially fungus. In our study, after adjusting for baseline kidney function and comorbidity, transplant patients with COVID-19 were still 1.88 times more likely to develop OI (IC 95%; 1.12-3.17). Maintenance immunosuppressive treatment is the predisposing factor, especially corticoids and anticalcineurinics. This explains the pulmonary aspergillosis (with high mortality) observed in this group and the greater frequency of pneumocystosis and invasive intestinal CMV infection (Table 2).

The present study has its limitations, mainly its retrospective design and the sample size. The virus variant responsible for each infection episode was unavailable for all cases. The treatment protocols for COVID-19 were not homogenous and varied depending on the availability of clinical evidence. Another limitation was not including non-invasive CMV and BK virus infection in the analysis. Given the study´s retrospective design, measurements of polymerase chain reaction for CMV in blood were not performed systematically; thus, only those cases of invasive infection were included. The same may be said about BK virus infection.

In conclusion, the evolutive course of SARS-CoV-2 infection requiring hospitalization in renal transplant recipients is primarily determined by comorbidity and baseline renal function. At equal comorbidity and eGFR, there were no differences in mortality, ICU admission, NI, or hospital stay between transplant recipients and controls. The risk for OI among transplant patients was significantly greater, with a higher frequency of pneumocystosis and CMV colitis. Baseline eGFR, higher levels of serum IL-6, and coinfection at admission were independently associated with OI. In general, the prognosis in kidney transplant recipients hospitalized for COVID-19 is comparatively bad; the principal measure to effectively impact these outcomes is to prevent hospital admission, either by improving the antibody response to vaccination or by early diagnosis and treatment (monoclonal antibodies; antivirals, such as remdesevir and molnupiravir) in an ambulatory dimension.

Declaration of Competing Interest

This study protocol was reviewed and approved by the Comitè d´Ètica d´Investigació I Calidad of Hospital Universitari Arnau de Vilanova.

Data availability

• Data will be made available on request.