Impact of immunosuppressive therapy on the severity of COVID-19 in solid organ transplant recipients.

Dear Editor,

We read with interest the article by Minotti et al. on immunosuppression on severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. Variable clinical courses of Coronavirus-19 disease (COVID-19) were reported 2, 3, 4 in solid organ transplant (SOT) recipients, but few data are available on the impact of immunosuppression on clinical severity. Since immune system plays an essential role in the pathophysiology of COVID-19 by developing a hyperinflammatory state, immunomodulators were found to improve clinical course. In addition, CNIs inhibits T cells activation and may reduce monocyte-macrophages activation thus preventing cytokine-release syndrome, but a suboptimal T cells response may hamper the clearance of SARS-CoV-2.

SOT recipients with confirmed SARS-CoV-2 infection admitted to two transplant centers in Northern Italy, ASST Grande Ospedale Metropolitano Niguarda (Milano) and ASST Ospedale Papa Giovanni XXIII (Bergamo), between February 21 and May 31, 2020 were retrospectively analyzed to assess the role of immunosuppressive therapy in clinical presentation and severity of COVID-19 and to describe its management during the course of the disease. Disease severity was classified as (1) mild (WHO Clinical Progression Scale grades 1–3), (2) moderate (WHO Clinical Progression Scale grades 4–5), (3) severe or critical (WHO Clinical Progression Scale grades 6–9). The immunosuppressive regimen before and after diagnosis of SARS-CoV-2 infection was recorded and its major changes after admission were categorized as (1) CNIs withdrawal/dose reduction, (2) anti-metabolite withdrawal, (3) steroid dose increase. Continuous variables were reported as median with interquartile range (IQR), categorical variables as absolute (%) values. Kruskall–Wallis and Mann–Whitney U test to compare continuous variables and Chi-square/Fisher exact test to compare categorical variables among groups were used. Survival curves were calculated by Kaplan-Meier method and compared by Log-rank test. Data were analyzed using GraphPad Prism v8.

Thirty-nine patients were analyzed. Major baseline demographic, clinical and biochemical characteristics are described in Table 1 . Twelve patients (31%) were classified as mild, 15 (38%) as moderate and 12 (31%) as severe or critical. Five patients (13%) required admission to intensive care unit. SARS-CoV-2 infection was detected 5 (1–10) days after symptoms onset, while the peak of the disease was observed 10 (7–14) days after clinical presentation. The overall survival was 82% (n=32): 92% (n=11/12) in mild, 100% (n=15/15) in moderate and 50% (n=6/12) in severe patients. Five patients died due to COVID-19-related respiratory failure and 2 to concomitant diseases (diffuse large B cells lymphoma and pancreatic cancer progression).

Table 1

Major baseline characteristics of the three severity groups.

	Overall n=39	Mild n=12	Moderate n=15	Severe n=12	p
Age years	62 (54–68)	63 (53–67)	60 (58–69)	62 (56–69)	0.813
Male gender n	28 (73%)	8 (67%)	10 (67%)	10 (83%)	0.566
Time-from-transplant > 1 year	7 (18%)	2 (17%)	3 (20%)	2 (17%)	0.680
Charlson comorbidity index	3 (2–4)	2 (2–4)	4 (3–4)	3 (2–4)	0.755
Arterial hypertension n	16 (41%)	2 (9%)	6 (40%)	8 (67%)	0.045
Diabetes mellitus n	8 (21%)	3 (25%)	3 (20%)	2 (17%)	0.878
Chronic kidney disease n	6 (15%)	2 (9%)	3 (20%)	1 (8%)	0.719
Organ transplant n
Liver	27 (69%)	10 (83%)	9 (60%)	8 (67%)	0.566
Kidney	9 (23%)	2 (17%)	5 (33%)	2 (17%)
Kidney-heart	2 (5%)	–	1 (7%)	1 (8%)
Kidney-pancreas	1 (3%)	–	–	1 (8%)
Leukocyte count 103 cells/mcL	5.5 (4.5–7.9)	5.5 (4.5–6.4)	3.2 (1.1–7.3)	4.5 (3.1–8.6)	0.205
Lymphocyte count 103 cells/mcL	0.76 (0.54–1.15)	0.81 (0.56–1.03)	1.14 (0.72–1.54)	0.62 (0.51–0.76)	0.289
C-reactive protein mg/dL	4.9 (2.2–10.7)	3.5 (1.9–6.4)	4.3 (2.2–7.9)	13.6 (8.2–23.8)	0.110
Creatinine mg/dL	1.4 (1.0–2.24)	1.4 (1.2–1.6)	2.1 (0.7–2.8)	1.37 (1.2–1.86)	0.918
Lactate dehydrogenase U/L	96 (92–99)	225 (198–250)	311 (287–457)	381 (340–432)	0.023
Alanine aminotransferase U/L	25 (15–44)	41 (21–51)	20 (13–29)	23 (15–75)	0.882
Bilirubin mg/dL	0.7 (0.5–1.1)	0.7 (0.7–1.1)	0.7 (0.5-1.0)	0.8 (0.4–1.3)	0.788
COVID-19 treatment*
Lopinavir/ritonavir	10 (26%)	1 (8%)	4 (27%)	5 (42%)	–
(Hydroxy-)Chloroquine	18 (46%)	2 (17%)	8 (53%)	8 (67%)
High-dose steroid	10 (6%)	1 (8%)	5 (33%)	4 (33%)
Tocilizumab	3 (8%)	–	–	3 (25%)
Convalescent plasma	1 (3%)	–	–	1 (8%)

Different treatment combinations were administered.

All patients were on immunosuppressants: 69% was on a combined regimen, mainly based on CNIs (n=37; 95%) and a large proportion was receiving mycophenolate (44%) and steroid (42%). No significant association of type, dose or level of immunosuppressive agents at the time of COVID-19 diagnosis with clinical severity has been observed (Table 2 ), even when the analysis was limited to liver transplant recipients (data not shown). Patients on tacrolimus showed lower lymphocyte count compared with those on cyclosporin (0.58 (0.61–0.96] vs 1.12 [0.98–1.69], p=0.033), while no significant difference was observed among those on mycophenolate compared to the others (0.81 [0.52–1.16] vs 0.75 [0.48–1.18], p=0.064).

Table 2

Immunosuppressive regimens at admission.

	Overall n=39	Mild n=12	Moderate n=15	Severe n=12	p
Single-agent vs combined regimen	12/27	5/7	4/11	3/9	0.614
CNIs
Cyclosporine n	11 (28%)	1 (8%)	7 (47%)	3 (25%)	0.041
Tacrolimus n	26 (66%)	11 (92%)	6 (40%)	9 (75%)
Mycophenolate mofetil n	17 (44%)	4 (33%)	8 (53%)	5 (42%)	0.574
Steroid n	18 (46%)	5 (42%)	8 (53%)	5 (42%)	0.777
Belatacept n	1 (3%)	–	1 (7%)	–	–
Everolimus n	1 (3%)	–	1 (7%)	–	–
Cyclosporine level mg/L	75 (46–89)	46 (46–46)	75 (43–94)	80 (24–113)	0.853
Tacrolimus level mg/L	5.5 (3.7–8.2)	6.7 (2.4–8.4)	4 (3.5–7.4)	5.1 (3.8–8.9)	0.821
Mycophenolate dose g/day	1.40 (1.00–1.75)	1.25 (0.63–1.50)	1.47 (1.02–1.94)	1.44 (1.00–2.00)	0.574
Prednisone dose mg/day	7.5 (5–10)	10 (5–17.5)	6.5 (2.5–9.4)	10 (3.75–15)	0.530

The management of immunosuppressive treatment was heterogeneous after confirmed SARS-CoV-2 infection: 18/39 (46%) underwent modifications, especially among patients with moderate and severe disease (53% and 59%) compared to those with mild disease (25%). Mycophenolate withdrawal (mild 75% vs moderate 63% vs severe 70%, p=0.882) was the most frequent change, together with increased steroid dose (mild 50% vs moderate 43% vs severe 40%, p=0.954). CNIs reduction or withdrawal were more common in patients with moderate to severe respiratory failure (mild 17% vs moderate 39% vs severe 59%, p=0.213). Overall survival was not significantly different between patients who changed immunosuppressive regimen and the others (30-day survival 81% vs 65%; p=0.368).

Overall, we did not find any association between chronic immunosuppressive regimen at the time of COVID-19 presentation and its severity. Even though we could not analyze SARS-CoV-2 viral load, which might be influenced by CNIs and it is associated with disease severity, similar plasma trough levels of CNIs were observed among severity groups.

High mycophenolate dose (>1000 mg/day) was found an independent risk of severe COVID-19 in liver transplant recipients, even though not confirmed in another study. In our sample mycophenolic use and its dose tended to be lower in mild compared to moderate and severe patients, though no significant difference was observed, and more frequent withdrawals were recorded in mild compared to severe patients, possibly supporting the previous observation.

Current assumption and doses of steroid were similar among severity groups at admission, although its use improved the outcome in SARS-CoV-2 related respiratory failure. Nonetheless, the dose of prednisone (or equivalent) at the time of infection was lower than those recommended in COVID-19.

Reduction of CNIs dose, mycophenolate withdrawal and increase in steroid dose were more frequent in patients with moderate and severe disease, as expected. , Nonetheless, the consequences of these adjustments on disease progression are difficult to assess, given the heterogeneity of patients in terms of disease severity, comorbidities, length of symptoms and the limited and variable data on the efficacy of antiviral and anti-inflammatory treatments. Moreover, the timing of immune modulation might be a relevant issue: the reduced immune defense may favor viral replication and expose patients to a severe course, while the potential effect of immunosuppression might be beneficial only after cytokine release syndrome have been elicited by SARS-CoV-2.

As already shown in the general population arterial hypertension was strikingly associated with disease severity (Table 1), suggesting that risk factors other than immunosuppression may have a major role also in SOT recipients.

Even though the retrospective nature of the analysis and the small and heterogenous study population limit the strength of the conclusions, CNIs levels do not appear to influence the course of the infection. Since inadequate immunosuppression may expose patients to an increased risk of graft rejection, their withdrawal is not encouraged, even though transient reduction could be considered if concomitant anti-inflammatory treatment for COVID-19 is administered. Together with other currently available data, transient dose reduction or withdrawal of mycophenolate appears advisable.