Key words
coronavirus disease 2019 - COVID-19 - tocilizumab - immunomodulator - treatment
Introduction
Until now, the number of positive and death cases from coronavirus disease 2019
(COVID-19) is still increasing. This disease has caused significant health and
economic burden across the world. The manifestations of the disease may vary from
mild respiratory symptoms such as fever, nasal obstruction, and cough to severe
life-threatening symptoms such as respiratory distress, shock, arrhythmia, and heart
failure [1]. Several comorbid diseases has
been demonstrated to be associated with severe COVID-19 infections, such as
hypertension, diabetes, dyslipidemia, thyroid disease, cardiovascular disease,
anemia, and pulmonary disease [2]
[3]
[4]. Currently, there are no widely accepted
drugs for the management of COVID-19 patients. Several potential agents have been
proposed to help in achieving faster recovery time and reducing the mortality rate
in COVID-19 patients, and one of the agents is tocilizumab, an IL-6 inhibitor.
Tocilizumab has been approved for the treatment of rheumatoid arthritis, juvenile
idiopathic arthritis, and giant cell arteritis [5] Recently, tocilizumab has been offered to help in reducing the
pro-inflammatory cytokines in COVID-19 and preventing the cytokine storm syndrome
that could contribute to the development of the severe outcome. Unfortunately, the
evidence regarding the potential benefit and safety of tocilizumab in COVID-19
patients is still conflicting. Therefore, a meta-analysis is required to aid in
solving this problem. This article aims to explore the efficacy and safety of
tocilizumab administration in patients with COVID-19.
Materials and Methods
Eligibility criteria
Studies were included in this review if met the following inclusion criteria:
representation for clinical questions (P: positive/confirmed cases of
COVID-19; I: receiving tocilizumab as their treatment; C: did not receive
tocilizumab or receive only standard of care treatment; O: efficacy of
tocilizumab (rate of severe COVID-19, mortality, and length of hospital stay)
and serious adverse events of tocilizumab (thromboembolism incident and
secondary infection); S: type of study was a randomized control trial, cohort,
clinical trial, case-cohort, and cross-over design) and if the full-text article
was available. The following types of articles were excluded: articles other
than original research (e. g., review articles or commentaries); case
reports; articles not in the English language; articles on research in pediatric
populations (17 years of age or younger); and articles on research in pregnant
women.
Search strategy and study selection
A systematic search of the literature was conducted on PubMed and Europe PMC
using the keywords “tocilizumab” OR “anti-IL-6”
OR “IL-6 inhibitor” AND “coronavirus disease
2019” OR “COVID-19”, between 2019 and present time
(November 1st, 2020) with language restricted to English only.
Duplicate results were removed. The remaining articles were independently
screened for relevance by its abstracts with two authors. The full text of
residual articles was assessed according to the inclusion and exclusion
criteria. The references of all identified studies were also analyzed (forward
and backward citation tracking) to identify other potentially eligible articles.
The study was carried out per the Preferred Reporting Items for Systematic
Reviews and Meta-Analyses (PRISMA) guidelines [6]
Data extraction and quality assessment
Data extraction was performed independently by two authors, we used standardized
forms that include author, year, study design, number of participants, age,
gender, number of patients who receive tocilizumab and who did not, tocilizumab
dose, and proportion of patients with each outcome of COVID-19.
The outcome of interest was severe COVID-19, mortality, length of hospital stay,
and serious adverse events which comprised of thromboembolism incident and
secondary infection. Severe COVID-19 was defined as patients who had any of the
following features at the time of, or after, admission: (1) respiratory distress
(≥30 breaths per min); (2) oxygen saturation at
rest≤93%; (3) ratio of the partial pressure of arterial oxygen
(PaO2) to a fractional concentration of oxygen inspired air
(fiO2)≤300 mmHg; or (4) critical complication (respiratory
failure, septic shock, and or multiple organ dysfunction/failure) or
admission into ICU. Mortality outcome from COVID-19 was defined as the number of
patients who were dead because of COVID-19 infection.
Two investigators independently evaluated the quality of the included cohort and
case-control studies using the Newcastle–Ottawa Scale (NOS) [7]. The selection, comparability, and
exposure of each study were broadly assessed and studies were assigned a score
from zero to nine. Studies with scores≥7 were considered of good
quality. They also independently evaluated the quality of the included clinical
trial studies using the Revised Cochrane risk-of-bias tool for randomized trials
(RoB 2) [8].
Statistical analysis
A meta-analysis was performed using Review Manager 5.4 (Cochrane Collaboration)
software. Dichotomous variables were calculated using the Mantel-Haenszel
formula with a random-effects model regardless of heterogeneity. The effect
estimate was reported as risk ratio (RR) along with its 95% confidence
intervals (CIs) for dichotomous variables, respectively. For continuous
variables, the inverse variance method was used to obtain mean differences (MDs)
and its standard deviations (SDs). P-value was two-tailed, and the statistical
significance was set at≤0.05. A funnel plot, Begg’s rank
correlation method [9], and
Egger’s weighted regression method [10] were adopted to statistically assess publication bias
(P<0.05 was considered statistically significant). When data were
reported as medians and interquartile ranges, we would convert them to means and
standard deviations for meta-analytical pooling using the formula by Wan X, et
al [11].
Results
Study selection and characteristics
A total of 4274 records were obtained through systematic electronic searches and
other ways. After the removal of duplicates, 4050 records remained. A total of
3956 records were excluded after screening the titles/abstracts because
they did not match our inclusion and exclusion criteria. After evaluating 94
full-texts for eligibility, 54 full-text articles were excluded because they do
not have the control/comparison group, 2 full-text articles were
excluded because the articles were not in English, and finally, 38 studies [12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49] with a total of 13 412 COVID-19
patients were included in the meta-analysis ([Fig. 1]). Of a total of 38 included studies, 3 were double-blind
randomized-controlled trial (RCT), 4 were open-label RCT, 23 were retrospective
cohort, 3 studies were prospective cohort, while the remaining 5 studies was a
case-control study. The dose and preparation of tocilizumab used were varied
among the included studies. Most of the included studies (24 studies) use
intravenous tocilizumab at dosage 8 mg/kg, 1–2 doses,
while the remaining studies use tocilizumab at 400 mg, 1–2
doses, and subcutaneous tocilizumab at a dosage of 324 mg given as two
consecutive injections. The essential characteristics of the included studies
are summarized in [Table 1].
Fig. 1 PRISMA diagram of the detailed process of selection of
studies for inclusion in the systematic review and meta-analysis.
Table 1 Characteristics of included studies.
|
Study
|
Sample size
|
Design
|
Tocilizumab dose
|
Tocilizumab patients
|
Non-tocilizumab patients
|
|
n (%)
|
Age (years)
|
n (%)
|
Age (years)
|
|
Campochiaro C et al. [12] 2020
|
65
|
Retrospective cohort
|
IV: 400 mg, 1–2 doses
|
32 (49.2%)
|
64±16.2
|
33 (50.8%)
|
63.5±15.1
|
|
Canziani LM et al. [13]
2020
|
128
|
Case-control
|
IV: 8 mg/kg, 1–2 doses
|
64 (50%)
|
63±12
|
64 (50%)
|
64±8
|
|
Capra R et al. [14]
2020
|
85
|
Retrospective cohort
|
IV: 400 mg, 1 dose
|
62 (72.9%)
|
63.3±14.1
|
23 (27.1%)
|
68.3±18.5
|
|
Chilimuri S et al. [15]
2020
|
1225
|
Retrospective cohort
|
IV: 400 mg, 1–2 doses
|
87 (7.1%)
|
61.6±15.5
|
1138 (92.9%)
|
63±14.8
|
|
Colaneri M et al. [16]
2020
|
112
|
Retrospective cohort
|
IV: 400 mg, 1 dose
|
21(18.7%)
|
62.3±18.6
|
91 (81.3%)
|
63.7±16.3
|
|
De Rossi et al. [17]
2020
|
158
|
Retrospective cohort
|
IV: 400 mg, 1 dose
SC: 324 mg, 1
dose
|
90 (56.9%)
|
62.9±12.5
|
68 (43.1%)
|
71±14.6
|
|
Eimer J et al. [18]
2020
|
87
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
29 (33.3%)
|
56.6±10.3
|
58 (66.7%)
|
57.2±9.4
|
|
Enzmann MO et al. [19]
2020
|
150
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
12 (15.3%)
|
N/A
|
66 (84.7%)
|
N/A
|
|
Gokhale Y et al. [20]
2020
|
269
|
Retrospective cohort
|
IV: 400 mg, 1 dose
|
151 (56.1%)
|
52.3±11.8
|
118 (43.9%)
|
55.3±12.5
|
|
Guaraldi G et al. [21]
2020
|
544
|
Retrospective cohort
|
IV: 8 mg/kg, 2 doses
SC:
162 mg, 2 doses
|
179 (32.9%)
|
63.3±13.3
|
365 (67.1%)
|
68±15.5
|
|
Gupta S et al. [22]
2020
|
3924
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
433 (11%)
|
57±12.5
|
3491 (89%)
|
62.3±14.8
|
|
Hermine O et al. [23]
2020
|
130
|
Open-label RCT
|
IV: 8 mg/kg, 1–2 doses
|
63 (48.6%)
|
65.1±12.7
|
67 (51.4%)
|
64.2±11.2
|
|
Holt GE et al. [24]
2020
|
62
|
Retrospective cohort
|
IV: 400 mg, 1 dose
|
32 (51.6%)
|
N/A
|
30 (48.4%)
|
N/A
|
|
Ip A et al. [25]
2020
|
547
|
Retrospective cohort
|
IV: 400 mg, 1 dose
|
134 (24.4%)
|
61.6±12.5
|
413 (75.6%)
|
68±14.1
|
|
Kewan T et al. [26]
2020
|
51
|
Retrospective cohort
|
IV: 8 mg/kg, 1 dose
|
28 (54.9%)
|
62±13.3
|
23 (45.1%)
|
66.6±14.8
|
|
Klopfenstein T et al. [27] 2020
|
206
|
Case-control
|
IV: 8 mg/kg, 1–2 doses
|
30 (14.5%)
|
75.6±11.3
|
176 (85.5%)
|
74.3±11
|
|
Lengnan X et al. [28]
2020
|
19
|
Retrospective cohort
|
IV: 400 mg, 1 dose
|
5 (26.3%)
|
73.2±4.4
|
14 (73.7%)
|
66.2±5
|
|
Masia M et al. [29]
2020
|
138
|
Prospective cohort
|
IV: 400 mg if<75 kg and
600 mg if≥75 kg
|
76 (55%)
|
65.2±14.9
|
62 (45%)
|
65.9±16.8
|
|
Martinez-Sanz J et al. [30] 2020
|
1229
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
260 (21.1%)
|
65.3±15.5
|
969 (78.9%)
|
68.3±17
|
|
Menzella F et al. [31]
2020
|
79
|
Prospective cohort
|
IV: 8 mg/kg, 2 doses
SC:
162 mg, 2–4 doses
|
41 (51.8%)
|
63.3±10.6
|
38 (48.2%)
|
70.3±11.3
|
|
Mikulska M et al. [32]
2020
|
196
|
Prospective cohort
|
IV: 8 mg/kg, 1–2 doses
SC:
162 mg, 1–2 doses
|
130 (66.3%)
|
64.5±12.4
|
66 (33.7%)
|
73.5±14.4
|
|
Moiseev S et al. [33]
2020
|
137
|
Retrospective cohort
|
IV: 400 mg, 1 dose
|
83 (60.5%)
|
55.6±11.1
|
54 (39.5%)
|
56.3±14
|
|
Moreno-Perez O et al. [34] 2020
|
236
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
77 (32.6%)
|
62.3±14
|
159 (67.4%)
|
57±19.2
|
|
Perrone F et al. [35]
2020
|
301
|
Open-label RCT
|
IV: 8 mg/kg, 1–2 doses
|
180 (59.8%)
|
N/A
|
121 (40.2%)
|
N/A
|
|
Potere N et al. [36]
2020
|
80
|
Case-control
|
SC: 162 mg, 2 doses
|
40 (50%)
|
59.8±16.9
|
40 (50%)
|
59.1±17
|
|
Price CC et al. [37]
2020
|
239
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
153 (64%)
|
N/A
|
86 (46%)
|
N/A
|
|
Rodriguez-Bano J et al. [38] 2020
|
432
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
88 (20.3%)
|
64.6±11.8
|
344 (79.7%)
|
68±12.5
|
|
Rojas-Marte G et al. [39] 2020
|
193
|
Case-control
|
IV: 8 mg/kg, 1–2 doses
|
96 (49.7%)
|
58.8±13.6
|
97 (50.3%)
|
62±14
|
|
Roomi S et al. [40]
2020
|
176
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
134 (78.8%)
|
65.4±10.5
|
36 (21.2%)
|
58±13.2
|
|
Rosas I et al. [41]
2020
|
438
|
Double-blind RCT
|
IV: 8 mg/kg, 1–2 doses
|
294 (67.1%)
|
60.9±14.6
|
144 (32.9%)
|
60.6±13.7
|
|
Rossi B et al. [42]
2020
|
246
|
Case-control
|
IV: 400 mg, 1 dose
|
106 (43%)
|
64.3±13
|
140 (57%)
|
70.1±16.5
|
|
Roumier M et al. [43]
2020
|
59
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
30 (50.8%)
|
58.8±12.4
|
29 (49.2%)
|
71.2±15.4
|
|
Ruiz-Antoran B et al. [44] 2020
|
506
|
Retrospective cohort
|
IV: 8 mg/kg, 1–2 doses
|
268 (52.9%)
|
65±11.7
|
238 (47.1%)
|
71.3±14.2
|
|
Salama C et al. [45]
2020
|
377
|
Double-blind RCT
|
IV: 8 mg/kg, 1–2 doses
|
249 (66%)
|
56±14.3
|
128 (34%)
|
55.6±14.9
|
|
Salvarani C et al. [46]
2020
|
126
|
Open-label RCT
|
IV: 8 mg/kg, 1–2 doses
|
60 (47.6%)
|
62.1±16.2
|
66 (52.4%)
|
61.6±14
|
|
Somers EC et al. [47]
2020
|
154
|
Retrospective cohort
|
IV: 8 mg/kg, 1 dose
|
78 (50.6%)
|
55±14.9
|
76 (49.4%)
|
60±14.5
|
|
Stone JH et al. [48]
2020
|
243
|
Double-blind RCT
|
IV: 8 mg/kg, 1 dose
|
161 (66.2%)
|
59.2±17.2
|
82 (33.8%)
|
56.3±17.1
|
|
Wang D et al. [49]
2020
|
65
|
Open-label RCT
|
IV: 400 mg, 1–2 doses
|
34 (52.3%)
|
64.1±9.6
|
31 (47.7%)
|
62±11.1
|
Quality of study assessment
Studies with various study designs including a clinical trial, cohort, and
case-control were included in this review and assessed accordingly with the
appropriate scale or tool. Newcastle Ottawa Scales (NOS) were used to assess the
cohort and case-control studies ([Table
2]). All included studies were rated ‘good’. For
clinical trial studies, the Revised Cochrane risk-of-bias tool for randomized
trials (RoB 2) was used and all of the included trials showed a low risk of bias
([Table 3]). In conclusion, all
studies were seemed fit to be included in the meta-analysis.
Table 2 Newcastle-Ottawa quality assessment of observational
studies.
|
First author, year
|
Study design
|
Selection
|
Comparability
|
Outcome
|
Total score
|
Result
|
|
Campochiaro C et al. [12] 2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
Canziani LM et al. [13]
2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
Capra R et al. [14]
2020
|
Cohort
|
***
|
**
|
**
|
7
|
Good
|
|
Chilimuri S et al. [15]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Colaneri M et al. [16]
2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
De Rossi N et al. [17]
2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
Eimer J et al. [18]
2020
|
Cohort
|
**
|
**
|
***
|
7
|
Good
|
|
Enzmann MO et al. [19]
2020
|
Cohort
|
**
|
**
|
***
|
7
|
Good
|
|
Gokhale Y et al. [20]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Guaraldi et al. [21]
2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
Gupta S et al. [22]
2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
Holt GE et al. [24]
2020
|
Cohort
|
**
|
**
|
***
|
7
|
Good
|
|
Ip A et al. [25]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Kewan T et al. [26]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Klopfenstein T et al. [27] 2020
|
Cohort
|
***
|
**
|
**
|
7
|
Good
|
|
Lengnan X et al. [28]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Masia M et al. [29]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Martinez-Sanz J et al. [30] 2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
Menzella F et al. [31]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Mikulska M et al. [32]
2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
Moiseev S et al. [33]
2020
|
Cohort
|
**
|
**
|
***
|
7
|
Good
|
|
Moreno-Perez O et al. [34] 2020
|
Cohort
|
**
|
**
|
***
|
7
|
Good
|
|
Potere N et al. [36]
2020
|
Cohort
|
***
|
**
|
**
|
7
|
Good
|
|
Price CC et al. [37]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Rodriguez-Bano J et al. [38] 2020
|
Cohort
|
****
|
**
|
***
|
9
|
Good
|
|
Rojas-Marte G et al. [39] 2020
|
Case-control
|
***
|
**
|
***
|
8
|
Good
|
|
Roomi S et al. [40]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Rossi B et al. [42]
2020
|
Case-control
|
***
|
**
|
***
|
8
|
Good
|
|
Roumier M et al. [43]
2020
|
Cohort
|
***
|
**
|
**
|
7
|
Good
|
|
Ruiz-Antoran B et al. [44] 2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Salama C et al. [45]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
|
Somers EC et al. [46]
2020
|
Cohort
|
***
|
**
|
***
|
8
|
Good
|
Table 3 Risk of bias assessment for clinical trial studies using RoB-2 tool.
 
|
Tocilizumab and outcomes
Tocilizumab efficacy
Our pooled analysis showed that tocilizumab administration was associated
with reduction of mortality rate from COVID-19 [OR 0.54 (95% CI
0.42–0.71), p<0.00001, I
2=79%, random-effect modelling] ([Fig. 2a]). However, tocilizumab
administration did not alter the severity of COVID-19 [OR 1.05 (95%
CI 0.92–1.20), p=0.47, I
2=84%, random-effect modelling] ([Fig. 2b]) and length of hospital stay
[Mean Difference 1.77 days (95% CI −0.61–4.14 days),
p=0.15, I
2=97%, random-effect modelling] ([Fig. 2c]).
Fig. 2 Forest plot that demonstrates the association of
tocilizumab with the mortality a, severe COVID-19 b,
length of hospital stay c, and serious adverse events
d in COVID-19 infection.
Tocilizumab safety
Our meta-analysis showed that tocilizumab administration was not associated
with serious adverse events [OR 0.91 (95% CI 0.71–1.15),
p=0.42, I
2=46%, random-effect modelling] ([Fig. 2d]). Subgroup analysis showed
that tocilizumab administration was not associated with thromboembolism
incident [OR 1.02 (95% CI 0.69–1.50), p=0.93,
I
2=12%, random-effect modelling], nor secondary
infection [OR 0.86 (95% CI 0.63–1.18),
p=0.36, I
2=57%, random-effect modelling].
Subgroup analysis
Subgroup analysis for clinical trial studies showed a higher OR for mortality
rate outcome [OR 0.90 (95% CI 0.64–1.26), p=0.54,
I
2=0%, random-effect modelling] compared to
observational studies [OR 0.50 (95% CI 0.38–0.67),
p<0.00001, I
2=80%, random-effect modelling]. Subgroup analysis
for clinical trial studies showed a lower OR for severe COVID-19 outcome [OR
0.81 (95% CI 0.53–1.23), p=0.32, I
2=23%, random-effect modelling] compared to
observational studies [OR 1.11 (95% CI 0.96–1.28),
p=0.15, I
2=86%, random-effect modelling]. Subgroup analysis
for clinical trial studies showed a lower Mean Difference for length of hospital
stay outcome [Mean Difference −1.43 days (95% CI
−5.13–2.26 days), p=0.45, I
2=95%, random-effect modelling] compared to
observational studies [Mean Difference 2.70 days (95% CI
−0.59–5.99 days), p=0.11, I
2=97%, random-effect modelling]. Subgroup analysis
for clinical trial studies showed a lower OR for serious adverse events outcome
[OR 0.52 (95% CI 0.29–0.92), p=0.02, I
2=38%, random-effect modelling] compared to
observational studies [OR 1.04 (95% CI 0.80–1.35),
p=0.76, I
2=41%, random-effect modelling].
Publication Bias
The funnel-plot analysis showed a qualitatively symmetrical inverted funnel-plot
for the association between tocilizumab administration and mortality ([Fig. 3a]), severe COVID-19 ([Fig. 3b]), length of hospital stay ([Fig. 3c]), and serious adverse events
([Fig. 3d]). Meanwhile,
rank-correlation Begg’s test and regression-based Egger’s test
were not statistically significant for all outcomes, showing no indication of
publication bias ([Table 4]).
Fig. 3 Funnel plot analysis for mortality a, severe
COVID-19 b, length of hospital stay c, and serious adverse
events d outcome.
Table 4 Summary of meta-analysis.
|
Outcomes
|
Effect size (95% Confidence Interval),
p-value
|
Heterogeneity (I2), p-value
|
Begg’s test
|
Egger’s test
|
Number of Studies
|
|
Mortality
|
OR=0.54 [0.42–0.71],<0.00001
|
79%,<0.00001
|
0.968
|
0.284
|
37
|
|
Severe COVID-19
|
OR=1.05 [0.92–1.20], 0.47
|
84%,<0.00001
|
0.464
|
0.150
|
30
|
|
Length of hospital stay
|
Mean Difference=1.77 [−0.61–4.14],
0.15
|
97%,<0.00001
|
0.836
|
0.213
|
17
|
|
Thrombosis incident
|
OR=1.02 [0.69–1.50], 0.93
|
12%, 0.33
|
0.916
|
0.978
|
9
|
|
Secondary infection
|
OR=0.86 [0.63–1.18], 0.36
|
57%, 0.02
|
0.558
|
0.451
|
16
|
Discussion
Based on a contrite meta-analysis of available data, tocilizumab seems to be
beneficial only in reducing the mortality rate from COVID-19 infection, but it did
not alter the severity outcome of COVID-19 and the duration of hospital stay.
However, our subgroup analysis that involves only clinical trial studies showed that
tocilizumab failed to reduce the mortality rate from COVID-19 and cannot alter the
severity outcome and length of hospital stay in COVID-19 patients. Tocilizumab also
appears to be relatively safe in COVID-19 patients, compared with standard of care
treatment as it is not associated with serious adverse events such as
thromboembolism incident and secondary infection. Several reasons may be proposed
to
explain the lack of efficacy from tocilizumab administration in COVID-19 patients.
First, interleukin-6 and other inflammatory proteins that are observed to be present
at elevated levels in patients with COVID-19 represent host responses to the
infection, similar to the elevations in cytokine levels seen in patients with
endocarditis, sepsis, and other infections, rather than components of a
self-amplifying inflammatory loop that would benefit from suppression [49]. Second, severe COVID-19 symptoms may not
be caused by cytokine storm syndrome like we used to think before. Recently
published systematic review and meta-analysis showed that the descriptor cytokine
storm does not appropriately describe the milieu in COVID-19-induced organ
dysfunction. The mean IL-6 concentration in COVID-19 patients is relatively low
(36.7 pg/mL (95% CI
21.6–62.3 pg/mL), when compared with other conditions which
received benefit from tocilizumab administration such as chimeric antigen receptor
(CAR) T cell-induced cytokine release syndrome (difference
3074 pg/mL, 95% CI 325–26735 pg/mL;
p<0·0001), or when compared with other severe conditions such as
ARDS unrelated to COVID-19 (mean 460.1 pg/mL, 95% CI
216.3–978.7 pg/mL; difference 423.4 pg/mL,
95% CI 106.9–1438.1 pg/mL;
p<0·0001), and sepsis (mean 983.6 pg/mL, 95%
CI 550.1–1758.4 pg/mL; difference 947 pg/mL,
95% CI 324–2648 pg/mL; p<0·0001).
Even in patients with hypoinflammatory ARDS, the mean IL-6 was still 5 times higher
than the concentration in patients with COVID-19 [50]. Alternative mechanisms of COVID-19-induced organ dysfunction may
play a part. Therefore, IL-6 may not play such a significant role in the
pathogenesis of COVID-19, and inhibiting IL-6 through tocilizumab administration
will not significantly alter the outcomes of COVID-19.
Our study was not without limitations. First, there was significant heterogeneity
noted in our studies. One plausible rationale for this is the fact that the
therapies for COVID-19 are rapidly evolving and hence the SOC differed significantly
from one study to another. Moreover, the unaccounted confounders, especially in the
included observational studies can also explain the heterogeneity noted in our
study. Second, there was a significant variation in the follow-up of patients.
Third, the studies did not consistently measure serum IL-6 and hence a correlation
between IL-6 level and drug activity could not be established. Last, there was no
standardization in the number of medication dosage, route of administration, and
timing of administration. This can also account for the difference in outcomes noted
across studies.
Despite the limitations, our study has significant strengths. First, we included a
total of 38 studies with over 13 000 COVID-19 patients. This is by far the largest
analysis comparing the addition of tocilizumab to the standard of care treatment.
Moreover, we also included 5 recently published clinical trial studies in our
analysis and performed a subgroup analysis that only consist of clinical trial
studies to give more complete data regarding the benefit of tocilizumab
administration in COVID-19 patients.
Conclusion
In conclusion, tocilizumab is not effective and failed to improve the outcome of
COVID-19 patients compared with standard of care treatment, although it is
relatively safe and did not cause significant serious adverse events. Our study does
not support the routine use of tocilizumab for COVID-19 patients. Physicians may
hence consider giving other potential agents for the treatment of COVID-19 patients,
in addition to standard of care treatment. Future studies should focus more on other
potential therapies besides tocilizumab.
