2.1 Patients

We included all KTRs hospitalized for COVID-19, between March 9, 2020 and May 2, 2020, in three French academic kidney transplantation (KT) centers: Necker Hospital, Paris; Foch Hospital, Suresnes and Bicêtre Hospital, Le Kremlin-Bicêtre.

Control group is issued from the cohort of a retrospective multicenter observational study, the Critical Covid-19 France (CCF) including 2878 patients hospitalized for COVID-19 in 24 medical centers in France (NCT04344327)²⁰ (the three KT centers were not included in these centers). Immunosuppressed patients (i.e., hematologic stem cell or solid organ transplantation recipients, patients under chemotherapy or immunosuppressive treatments for systemic inflammatory diseases) and patients with missing data for one of the comorbidities needed for matching were excluded from control group.

COVID-19 was defined for all patients as the presence of a positive SARS-CoV-2 on reverse transcriptase-polymerase chain reaction (RT-PCR) testing performed on nasopharyngeal swab.

Data were collected retrospectively using hospital databases and clinical records in a similar manner for all patients (KTRs and nontransplanted patients). Each medical record was individually and manually reviewed by a medical doctor in each center. Data collection included demographic characteristics and underlying comorbidities of patients, COVID-19 infection characteristics and clinical course. For all patients, comorbidities were assessed with the same method using national recommendations. Diabetes was defined as the need for a pharmacologic anti-diabetic agent (oral or insulin therapy). Hypertension was defined as the need for at least one anti-hypertensive therapy. Cardiopathy was defined as a history of documented cardiopathy diagnosed by a cardiologist: coronary artery disease assessed by a coronarography, valvular, dilated, hypertrophic, cardiopathy, or arteriovenous fistula-associated high-output cardiac failure assessed by an echocardiography or atrial fibrillation assessed by an electrocardiogram or Holter monitoring. Chronic lung disease was defined as a history of documented lung disease on CT scan for bronchiectasis and pulmonary function tests (PFT) for COPD and asthma. Basal renal function was defined using estimated glomerular filtration rate (eGFR) according to MDRD and categorized into three categories: >60 ml/min, 30 to 60 ml/min, <30 ml/min, chronic dialysis.

For KTRs, transplantation history and characteristics including and immunosuppression regimen were assessed.

Clinical, biological and radiological COVID-19 severity criteria at admission were also collected to analyze patient's disease severity at admission in both groups. Fever was defined as a temperature ≥38°C. SpO2 was assessed using pulse oximeters oxygen saturation. Fraction of inspired oxygen (FiO2) was calculated as follow: Oxygen Flow (Liters) × 4 + 21%. Dyspnea was considered positive if NYHA score was III or IV. Moreover, disease severity at entry was evaluated using the 8-point WHO ordinal scale score.²¹ CT scan severity stages of pulmonary ground-glass opacities lesions were categorized as low (<30% involvement), moderate (30–50% involvement), or severe (>50% involvement).

2.2 Outcomes

Survival was defined as the time between the date of hospital admission and the date of death. Severe disease-free survival was defined as the time between the date of hospital admission and the date of intensive care unit (ICU) admission and/or the date of death. Follow-up delay was determined as the delay between hospital admission and ICU transfer and/or death or hospital discharge.

2.3 Propensity score matching

KTRs were matched to nontransplant patients by propensity score, using a 1:1 ratio, nearest neighbor method and a caliper of 0.4 with the following covariates: age, sex, body mass index (BMI), diabetes mellitus, preexisting cardiopathy (coronary artery disease, valvular cardiopathy and “other” cardiopathy including hypertrophic cardiopathy, rhythmic cardiopathy, and arteriovenous fistula-associated high-output cardiac failure), hypertension, chronic lung disease (chronic obstructive pulmonary disease, asthma, and bronchiectasis) and eGFR according to MDRD. The balance of covariate distribution between the two groups was assessed using standardized mean difference (SMD).

2.4 Statistical analysis

Continuous variables were described using medians and interquartile ranges (IQRs) and compared using the Mann-Whitney rank sum test. Categorical variables were described using proportions and compared using chi-square test or fisher's exact test. Survival and severe disease-free survival were analyzed using an univariable Cox regression and represented using Kaplan-Meier curves and compared using a two-sided log-rank test. Center effect was analyzed using a frailty modeling in which random center effect was included in the Cox model. Statistical analyses were performed using Stata and RStudio (Version 1.2.1335). All statistical tests were two-sided, and p values <.05 were considered significant.

3.1 Patient characteristics

Overall, 100 KTRs required hospitalization for COVID-19. Their basal characteristics are detailed in Table 1. Of them, 64 (64.0%) were male, median age was 64.7 years (55.3–73.1) and median body mass index (BMI) was 25.9 kg/m² (21.9–28.2). Thirty-five patients (35.0%) had a preexisting cardiopathy, including 18 (18.0%) and 20 patients (20.0%) with coronary artery disease and cardiac arrhythmia, respectively, while 85 patients (85.0%) had hypertension and 48 (48.0%) diabetes mellitus. Eleven patients (11.0%) had chronic lung disease. Immunosuppressive induction consisted in antithymocyte globulin (ATG) in 52 patients (62.7%). At COVID-19 episode, maintenance immunosuppressive regimen was CNI based in 78 patients (83.0%) and mTORi in 8 patients (8.5%). Median basal serum creatinine was 141 μmol/l (108–197) and median estimated glomerular filtration rate (eGFR) 43.3 mL/min/1.73 m² (29.6–59.6).

For control group, 2017 patients were included before matching (Table 1). Of them, 1167 (57.9%) were male, median age was 67.46 years (55.4–79.6) and median BMI 27.0 kg/m² (24.0–31.1). Controls had a lower rate of coexisting comorbidities: 429 patients (21.3%) had a preexisting cardiopathy, 1009 (50.0%) hypertension, 594 (29.4%) diabetes mellitus and 295 (14.6%) chronic lung disease.

3.2 COVID-19 presentation and outcome in KTRs

Table 2 shows COVID-19 presentation and management in KTRs. COVID-19 occurred after a median time of 5.1 years (3.0–12.6) after kidney transplantation.

The most common initial symptoms were fever (n = 67, 72.0%), cough (n = 55, 63.2%), dyspnea (n = 41, 46.1%), and diarrhea (n = 32, 36.4%). Anosmia and/or ageusia concerned 10 patients (11.5%).

Oxygen therapy was necessary for most patients (n = 79, 84.0%). Transfer to ICU was required for 34 patients (34.0%), in a median 1 day after hospitalization (0–3.75), and 29 of them underwent mechanical ventilation (MV). Median hospitalization time in ICU was 13.5 days (8.3–23). After a median follow-up of 13 days (7–32), 26 patients (26.0%) died including 17 (65.4%) managed in ICU, while 62 patients (62%) were discharged from hospital. Acute rejection episode occurred in 2/100 KTRs (2%) after COVID-19 episode.

3.3 Severe COVID-19 prevalence is not increased in hospitalized KTRs

Propensity score method allowed the matching of 83 KTRs (83.0%) to 83 controls.

Table 3 confirms that these two groups are highly comparable in terms of age, sex, and comorbidities and Figure 1 assesses covariate balance (Figure 1A) and distribution of the propensity score (Figure 1B) between matched KTRs and controls as well as unmatched KTRs and controls.

At admission, disease severity was similar in both groups (Table 4). KTRs tended to have more disease severity criteria with higher fever, CRP, troponin rates and more severe forms at CT scan. Median WHO ordinal scale score was similar in both groups (3 IQR [3–4] and 3 IQR [3–4], respectively, in KTRs and non-KTRs, p = .11).

When adjusted to preexisting conditions, survival and severe disease-free survival were similar between KTRs and matched nontransplant patients with respective 30-day survival of 62.9% and 71.0%, p = .38. (Figure 2A) and respective 30-day severe disease-free survival of 50.6% and 47.5%, p = .91, (Figure 2). Overall survival HR was 1.375 (95% CI [0.667–2.832], p = .388 and severe disease-free survival (HR) was 1.03 (95% CI [0.632–1.673], p = .911) and for KTRs group.

These results remain unchanged when unmatched patients (17 patients in KTRs group and 1934 in control group) were added to the analysis (Figure 3).

Results for severe disease-free survival remained equivalent after including the Center effect in the Cox model (HR = 1.03; 95% CI [0.632–1.673]; p = .911). However, the low number of deaths per center did not allow us to adjust for Center in the Cox Model for overall survival. Both overall and disease-free survivals were similar between the three transplant centers (p = .319 and p = .995, respectively).

Of note, KTRs follow-up was significantly longer than non-KTRs despite similar severity outcomes. This was due to a longer length of hospital stay in KTRs (13 [9–21] days vs. 9 [4–13] days in non-KTRs, p = .001) despite a similar discharge rate (58 [69.9%] vs. 51 [63%], p = .348). Indeed, because they were considered as vulnerable to infection, KTRs were usually hospitalized longer in transplant centers comparing to non-KTRs.