Abstract
Background
The Clinical Trial of Sarilumab in Adults With COVID-19 (SARICOR) showed that patients with coronavirus disease 2019 (COVID-19) pneumonia and increased levels of interleukin (IL)-6 might benefit from blockade of the IL-6 pathway. However, the benefit from this intervention might not be uniform. In this subanalysis, we sought to determine if other immunoactivation markers, besides IL-6, could identify which subgroup of patients benefit most from this intervention.
Methods
The SARICOR trial was a phase II, open-label, multicenter, controlled trial (July 2020–March 2021) in which patients were randomized to receive usual care (UC; control group), UC plus a single dose of sarilumab 200 mg (sarilumab-200 group), or UC plus a single dose of sarilumab 400 mg (sarilumab-400 group). Patients who had baseline serum samples for cytokine determination (IL-8, IL-10, monocyte chemoattractant protein–1, interferon-inducible protein [IP]-10) were included in this secondary analysis. Progression to acute respiratory distress syndrome (ARDS) according to cytokine levels and treatment received was evaluated.
Results
One hundred one (88%) of 115 patients enrolled in the SARICOR trial had serum samples (control group: n = 33; sarilumab-200: n = 33; sarilumab-400: n = 35). Among all evaluated biomarkers, IP-10 showed the strongest association with treatment outcome. Patients with IP-10 ≥2500 pg/mL treated with sarilumab-400 had a lower probability of progression (13%) compared with the control group (58%; hazard ratio, 0.19; 95% CI, 0.04–0.90; P = .04). Conversely, patients with IP-10 <2500 pg/mL did not show these differences.
Conclusions
IP-10 may predict progression to ARDS in patients with COVID-19 pneumonia and IL-6 levels >40 pg/mL. Importantly, IP-10 value <2500 pg/mL might discriminate those individuals who might not benefit from sarilumab therapy among those with high IL-6 levels.
Keywords: COVID-19, IL-6, IP-10, SARS-CoV-2, sarilumab, tocilizumab
Since the beginning of the severe acute respiratory syndrome coronavirus 2 (SARS-COV-2) pandemic, the hyperinflammatory state has been considered to play a crucial role in the progression of coronavirus disease 2019 (COVID-19) pneumonia. Severe infection forms are characterized by the development of acute respiratory distress syndrome (ARDS), which leads to respiratory failure requiring prolonged ventilatory assistance and, frequently, complicated intensive care unit (ICU) admissions and even death [1]. Immunosuppressive therapy has been proposed as a potential therapeutic approach to control disproportionate host inflammation, avoiding the resultant diffuse alveolar damage [2]. In this sense, interleukin (IL)-6 appears to be an important mediator that correlates with poorer outcomes and respiratory failure risk [1, 3]. Reinforcing this, IL-6 pathway blockade with tocilizumab (a human anti-IL-6 membrane-bound receptor monoclonal antibody) has been associated with lower mortality in patients hospitalized with COVID-19 pneumonia [4]. Thus it has become part of the backbone therapy for these patients. In line with this, the Clinical Trial of Sarilumab in Adults With COVID-19 (SARICOR) has shown that early blockade with sarilumab (a human anti-IL-6 soluble receptor monoclonal antibody) in patients with COVID-19 pneumonia and markers of systemic inflammation (IL-6 >40 pg/dL) was also associated with a trend for better outcomes [5].
Some serum markers have been evaluated as specific indicators of disease progression, including inflammatory and coagulation markers (such as C-reactive protein [CRP], ferritin, and D-dimer [DD]) and inflammatory cytokines (such as IL-6, IL-2R, IL-10, monocyte chemoattractant protein [MCP]-1 or tumor necrosis factor–α). Previous studies have suggested that CRP [6, 7] and IL-6 [8, 9] levels in earlier stages of COVID-19 predict clinical evolution and could be used for the selection of candidates for immunomodulatory therapy [7, 10]. In fact, clinical trials with anti-IL-6, in which selected patients had elevated biomarkers, have shown benefit. However, this benefit did not occur in all patients [11]. This could have to do with imprecise selection of candidates for treatment based on those biomarkers. Possibly, trial populations are composed of individuals with distinct phenotypes of immune response or in different stages of infection (viral phase vs early or later stages of inflammation), so the ability of these markers to classify patients within these different settings may be limited. Thus, the identification of specific markers that could detect not only those subjects at risk of progression but also those who benefit the most from anti-IL-6 therapy should be a priority.
The SARICOR was a clinical trial designed to investigate the efficacy and safety of early treatment with sarilumab added to usual care (UC) in hospitalized adults with COVID-19 pneumonia and IL-6 >40 pg/mL. In this trial, we demonstrated that an early blockade of the IL-6 pathway with sarilumab-400 mg was associated with lower risk of progression to ARDS requiring high-flow nasal oxygen (HFNO), noninvasive mechanical ventilation (NIMV), or invasive mechanical ventilation (IMV) during the first 28 days after randomization.
Study protocol contemplated baseline serum sample collection to determinate other cytokine levels (IL-8, IL-10, interferon-inducible protein [IP]-10, and MCP-1) at randomization. We present here a subanalysis of the SARICOR clinical trial, aimed to determine if other biomarkers of immunoactivation, besides high IL-6 levels, could help to identify which subgroup of patients would benefit most from this intervention.
METHODS
Trial Design and Patients
SARICOR was a phase II, open-label, randomized, multicenter, controlled clinical trial conducted in 10 hospitals in Andalusia, Southern Spain. Inclusion criteria in the SARICOR trial were (i) age 18 years; (ii) hospitalization due to COVID-19 with SARS-CoV-2 infection confirmed by a positive antigen detection test or a polymerase chain reaction assay; (iii) interstitial pneumonia confirmed by the presence of infiltrates on chest radiograph or a computed tomography scan; and (iv) IL-6 levels >40 pg/mL and/or DD >1500 ng/mL or >1000 if progressive increments were documented in at least 2 determinations after admission. For the current analysis, patients included in the SARICOR trial who had an available serum sample at randomization were selected.
Patient Consent
Patients’ informed consent was obtained before inclusion. Written informed consent was preferable, but initial oral consent before a witness documented in the clinical record and ratified later in writing was also an option, which was in accordance with Spanish Agency of Medicines and Medical Products exceptional measures applicable to clinical trials to manage problems arising from the COVID-19 emergency (https://www.aemps.gob.es/informa-en/exceptional-measures-applicable-to-clinical-trials-to-manage-problems-arising-from-the-COVID-19-emergency/?lang = en).
The SARICOR trial was approved by the Committee for Biomedical Research Ethics of the Reina Sofía University Hospital and was conducted in accordance with the International Conference on Harmonization E6 Guideline for Good Clinical Practice and the ethical principles of the Declaration of Helsinki. Authorization was also obtained from the Spanish Agency of Medicines and Medical Products (AEMPS, 20–0262). The trial is registered in accessible public databases such as the Spanish Clinical Studies Registry (REec), EUDRACT (2020-001531-27), and ClinicalTrials.gov (NCT04357860). A detailed description of the study protocol is available [12]
Randomization and Treatment
Patients were randomized in a 1:1:1 ratio to receive UC alone (control group), UC plus a single subcutaneous dose of 200 mg of sarilumab (sarilumab-200 group), or UC plus a single subcutaneous dose of 400 mg of sarilumab (sarilumab-400 group). Concealed randomization was carried out by means of electronic case report forms after obtaining informed consent and stratified according to the presence of an oxygen saturation (SatO2) of 90% while breathing room air and/or a partial pressure of arterial oxygen (PaO2) of 60 mmHg. Sarilumab was administered the same day as trial inclusion, generally within 3 hours after informed consent was obtained. Dexamethasone has been the preferred backbone therapy since the press release of the RECOVERY trial, but high and/or pulse doses (1 mg of methylprednisolone or equivalent per kilogram of body weight) of corticosteroids were also permitted.
Cytokine Determination
Serum was obtained by centrifugation of a 5-mL whole-blood sample and stored at −80°C until further use. Plasma or serum 1:4 diluted samples were analyzed to quantify the circulating level of inflammatory cytokines including IL-6, IL-8, IL-10, MCP-1, and IP-10, using a customized Cytometric Bead Array (CBA, ref. 558265, Becton, Dickinson and Company, San Jose, CA, USA) according to manufacturer instructions on a BD FacsCanto II. FacsDiva was used for acquisition, whereas FCAParray was used for analysis. At least 500 events of each bead were acquired.
Statistical Analysis
The primary outcome variable was the development of ARDS requiring HFNO, NIMV, or IMV during the first 28 days after randomization. The association between baseline biomarkers and the probability of the primary outcome was assessed. The predictive values of biomarkers were obtained by receiver operating characteristics (ROC) curves. Each cutoff point was selected based on the best trade-off values between sensitivity, specificity, and the percentage of patients correctly classified. Clinical status was recorded at baseline and every day during hospitalization for a total of 28 days after randomization. Outcome variables were first assessed by means of a time-to-event approach. Time was computed for the outcome event as the days elapsed from baseline, considered the day of randomization, to the date of the event or censoring date (day 28). Survival curves were compared according to the Kaplan-Meier method using the log-rank test. The impact of baseline biomarkers on treatment efficacy was assessed. The differences between the treatment groups were estimated as hazard ratios (HRs) with 95% CIs from Cox proportional hazards models stratified by baseline cytokine levels.
Categorical variables were expressed as percentages, and frequency was compared using the Pearson χ2 or Fisher exact test. Continuous variables were expressed as median and interquartile range (IQR) values. Comparison of continuous variables between 3 or more groups was performed using the Kruskal-Wallis test, and correlation analysis was performed using the Spearman test. A P value <.05 was considered statistically significant. All analyses were performed using SPSS statistical software package release 24.0 (IBM Corporation, Somers, NY, USA).
RESULTS
Features of the Study Population
One-hundred fifteen patients were enrolled in the trial between July 13, 2020, and March 5, 2021. Among them, 101 (88%) patients had available serum baseline samples and were included in this subanalysis (control group: n = 33 [32.7%]; sarilumab-200: n = 33 [32.7%]; and sarilumab-400: n = 35 [34.6%]). The main characteristics of this population are summarized in Table 1. These patients were comparable to the overall population included in the trial (Supplementary Table 1). In summary, patients were enrolled after a median (Q1 to Q3) of 9 (7 to 11) days from symptom onset and 1 (1 to 2) day from hospital admission. Ninety-three (92%) patients were receiving oxygen supplementation at randomization, 15 (15%) of them >15 L/min.
Table 1.
Features of the Study Population
| … | All (n = 101) | Control (n = 33) | Sarilumab-200 mg (n = 33) | Sarilumab-400 mg (n = 35) | P |
|---|---|---|---|---|---|
| Population characteristics | … | … | … | … | … |
| Agea | 60 (52–72) | 59 (52–74) | 67 (54–73) | 59 (50–67) | .26 |
| <65 y | 59 (58) | 18 (56) | 16 (47) | 25 (71) | .04 |
| 65–79 y | 37 (37) | 10 (31) | 17 (50) | 10 (29) | |
| ≥80 y | 5 (5) | 4 (13) | 1 (3) | 0 (0) | |
| Male gender, No. (%) | 67 (66) | 20 (63) | 20 (59) | 27 (77) | .24 |
| BMI,a,b kg/m2 | 30 (26–34) | 32 (29–35) | 29 (25–31) | 30 (26–33) | .12 |
| BMI ≥30 kg/m2 | 31 (54) | 13 (72) | 8 (42) | 10 (50) | .16 |
| Clinical characteristics and severity at baseline | … | … | … | … | … |
| Days from symptom onset to randomizationa | 9 (7–11) | 10 (6–11) | 9 (7–13) | 9 (7–11) | .7 |
| Days from hospitalization to randomizationa | 1 (1–2) | 1 (1–3) | 2 (1–2) | 2 (1–3) | .9 |
| SOFA scorea | 2 (1–2) | 2 (1–2) | 2 (1–3) | 1 (0–2) | .5 |
| Respiratory parameters at baseline | … | … | … | … | |
| Respiratory frequencya | 20 (18–24) | 20 (18–24) | 22 (18–25) | 20 (28–25) | .9 |
| Oxygen saturation while breathing room aira | 92 (89–95) | 93 (90–95) | 91 (89–95) | 92 (89–95) | .9 |
| Oxygen saturation while breathing room air <90% | 29 (29) | 6 (19) | 10 (31) | 13 (37) | .24 |
| FiO2a | 36 (28–50) | 36 (32–40) | 40 (30–60) | 34 (21–65) | .4 |
| Ratio oxygen saturation/FiO2a | 229 (172–281) | 184 (156–219) | 281 (253–291) | 276 (229–281) | .02 |
| Supplemental oxygen, No. (%) | … | … | … | … | … |
| No | 8 (8) | 2 (6) | 4 (13) | 2 (6) | .9 |
| Nasal cannula | 68 (69) | 24 (75) | 19 (59) | 25 (71) | |
| Nonrebreathing face mask | 7 (7) | 2 (6) | 3 (9) | 2 (6) | |
| Rebreathing face mask | 16 (16) | 4 (13) | 6 (19) | 6 (17) | |
| Oxygen flow, L/mina,c | 4 (3–8) | 4 (4–7) | 6 (4–8) | 4 (2–6) | .6 |
| <6 L/min | 57 (58) | 19 (59) | 14 (43) | 24 (68) | .4 |
| 6–14 L/min | 19 (19) | 7 (22) | 9 (28) | 3 (9) | |
| ≥15 L/min | 15 (15) | 4 (13) | 5 (16) | 6 (17) | |
| Laboratory parameters at baseline | … | … | … | … | … |
| Absolute lymphocyte count, 109/La | 0.92 (0.66–1.27) | 0.94 (0.79–1.28) | 1.17 (0.77–1.48) | 0.78 (0.47–1.04) | .02 |
| Platelets, 103/µL | 219 (181–286) | 231 (187–288) | 221 (186–291) | 204 (175–262) | .69 |
| Lactate dehydrogenase, μ/La | 367 (293–507) | 372 (304–481) | 343 (285–412) | 449 (293–561) | .36 |
| Ferritin, ng/mLa | 770 (433–1490) | 663 (443–1270) | 578 (344–1548) | 958 (630–1973) | .11 |
| D-dimer, ng/mLa | 950 (488–1825) | 761 (477–1285) | 681 (382–1600) | 1192 (543–3060) | .17 |
| C-reactive protein, mg/La | 74 (36–127) | 81 (30–127) | 57 (43–124) | 80 (45–127) | .78 |
| Procalcitonin, ng/mLa | 0.09 (0.01–0.15) | 0.08 (0.04–0.13) | 0.1 (0.07–0.17) | 0.10 (0.05–0.19) | .11 |
| Interleukin-6a,d | 54 (39–95) | 49 (21–76) | 60 (36–89) | 54 (42–117) | .6 |
| Concomitant therapies, No. (%) | … | … | … | … | … |
| Corticosteroids at randomization | 87 (88) | 29 (91) | 28 (88) | 30 (86) | … |
| Dexamethasone ≥6 mg/d | 43 (43) | 15 (47) | 11 (34) | 17 (49) | .55 |
| Methylprednisolone 40–125 mg/d | 4 (4) | 2 (6) | 2 (6) | 0 (0) | |
| Methylprednisolone 125–250 mg/d | 11 (11) | 5 (16) | 2 (6) | 4 (11) | |
| Methylprednisolone >250 mg/d | 29 (29) | 7 (22) | 13 (41) | 9 (26) |
Data are presented as No. (%) unless otherwise indicated.
Abbreviations: BMI, body mass index; FiO2, fraction of inspired oxygen; SOFA, Sequential Organ Failure Assessment.
Median (Q1–Q3).
Available in 57/101 patients.
In 87/101 patients receiving supplemental oxygen at randomization.
Available in 90/101 patients.
Biomarker levels at randomization in the study population are displayed in Table 2. For all the biomarkers, there were no differences at baseline between the study groups. IL-10, MCP-1, and IP-10 showed a correlation with IL-6 levels (rho = 0.526, P = .01; rho = 0.430, P = .01; rho = 0.525, P = .01; respectively).
Table 2.
Description of Baseline Cytokine Value (pg/mL)
| … | Baseline Cytokine Value | Control Group | Sarilumab-200 Group | Sarilumab-400 Group | P |
|---|---|---|---|---|---|
| IL-8 | 78 (30–172) | 68 (26–114) | 138 (34–185) | 70 (28–202) | .12 |
| IL-10 | 0 (0–5) | 0 (0–6) | 0 (0–4) | 0 (0–3) | .82 |
| MCP-1 | 284 (116–830) | 270 (105–891) | 347 (125–1097) | 271 (119–764) | .45 |
| IP-10 | 2025 (1076–3543) | 2008 (637–3544) | 2007 (934–3712) | 2286 (1395–3543) | .85 |
Data are presented as median (Q1–Q3).
Abbreviations: IL, interleukin; IP, interferon-inducible protein; MCP, monocyte chemoattractant protein.
The selection of cutoff values for IL-8, IL-10, MCP-1, and IP-10 was assessed by ROC curve analysis. The respective areas under the ROC curve (AUROCs) to predict the risk of progression to ARDS requiring HFNO, NIMV, or IMV for each biomarker were as follows: IL-8: 0.54 (95% CI, 0.40–0.68; P = .57); IL-10: 0.69 (95% CI, 0.55–0.83; P = .006); MCP-1: 0.63 (95% CI, 0.49–0.78; P = .056); and IP-10: 0.69 (95% CI, 0.57–0.80; P = .007). The cutoff points with the best diagnostic yield to predict the risk of progression for each biomarker were IL-8 = 103 pg/mL, IL-10 = 5 pg/mL, MCP-1 = 773 pg/mL, and IP-10 = 2500 pg/mL. IL-8 showed the worst profile with a poorer capacity of discrimination. Sensitivity for IL-10, MCP-1, and IP-10 was 55%, 55%, and 64%, whereas the negative predictive value was 85%, 86%, and 87%, respectively.
Probability of Presenting the Primary Outcome According to Baseline Biomarkers in the Overall Population
Twenty-two patients (21.8%) out of 101 included in this subanalysis developed the primary outcome. Figure 1 and Table 3 show the probability of progression to ARDS requiring HFNO devices, NIMV, or IMV during the first 28 days after randomization according to cytokine levels. Higher values at baseline for all cytokines, except IL-8, were associated with a worse prognosis. Consequently, IL-8 was excluded from further analysis.
Figure 1.
Kaplan-Meier survival curves according to the cutoff value of IL-8 levels (A), IL-10 levels (B), MCP-1 levels (C), and IP-10 levels (D) in the overall population. Abbreviations: IL, interleukin; IP, interferon-inducible protein; MCP, monocyte chemoattractant protein.
Table 3.
Probability of Presenting the Primary Outcome According to Baseline Biomarkers in the Overall Population
| … | No. of Patients With Event Within 28 Days, n/N | % of Patients With Event | Log-Rank P Value | |||
|---|---|---|---|---|---|---|
| D 5 | D 10 | D 15 | D 28 | |||
| IL-8 | … | … | … | … | … | .59 |
| IL8 <102.54 | 12/61 | 18 | 20 | 20 | 20 | |
| IL8 >102.54 | 10/40 | 17 | 25 | 25 | 25 | |
| IL-10 | … | … | … | … | … | .002 |
| IL10 <4.98 | 11/75 | 11 | 15 | 15 | 15 | |
| IL10 >4.98 | 11/26 | 38 | 42 | 42 | 42 | |
| MCP-1 | … | … | … | … | … | .001 |
| MCP-1 <772.61 | 10/73 | 12 | 14 | 14 | 14 | |
| MCP-1 >772.61 | 12/28 | 32 | 43 | 43 | 43 | |
| IP-10 | … | … | … | … | … | .012 |
| IP-10 <2500 | 8/60 | 10 | 13 | 13 | 13 | |
| IP-10 >2500 | 14/41 | 29 | 34 | 34 | 34 | |
Abbreviations: IL, interleukin; IP, interferon-inducible protein; MCP, monocyte chemoattractant protein.
Probability of Presenting the Primary Outcome According to Baseline Biomarkers and Treatment Group
Treatment effectiveness according to baseline cytokine levels was evaluated. Globally, 9 (27%) patients in the control group, 9 (27%) in the sarilumab-200 group, and 4 (11%) in the sarilumab-400 group had a progression to ARDS requiring HFNO, NIMV, or IMV. Figure 2 displays the probability of achieving the primary outcome according to study group and baseline cytokine levels. In patients with low levels of IL-10, MCP-1, or IP-10, no differences in the risk of progression between the study groups were found. Conversely, there was a trend toward a better prognosis in the sarilumab-400 group in patients with high cytokine levels. IP-10 showed the strongest association with treatment outcome (Table 4). Thus, patients with IP-10 ≥2500 pg/mL had a lower probability of progression with sarilumab-400 (13%) compared with the control group (58%; HR, 0.19; 95% CI, 0.04–0.90; P = .04) (Figure 2). To rule out the presence of any confounding factor when evaluating treatment effectiveness observed in patients with IP-10 >2500 pg/mL, a multivariate analysis adjusted by other prognosis covariates was performed. In this multivariate model, adjusted by age, sex, SOFA score, corticosteroid therapy, and baseline saturation at randomization, treatment with sarilumab-400 showed an independent association with the primary outcome (HR, 0.56; 95% CI, 0.007–0.484; P = .009) (Supplementary Table 2). The use of very high doses of methylprednisolone did not influence clinical outcome (Supplementary Table 3).
Figure 2.
Kaplan-Meier survival curves according to treatment group in patients with low IL-10 (A), MCP-1 (C), or IP-10 (E) levels and patients with high IL-10 (B), MCP-1 (D), or IP-10 (F) levels. Abbreviations: IL, interleukin; IP, interferon-inducible protein; MCP, monocyte chemoattractant protein.
Table 4.
Probability of Presenting the Primary Outcome According to Baseline Biomarkers and Treatment Group
| … | No. of Patients With Event Within 28 Days, n/N | % of Patients With Event | Log-Rank P Value | Hazard Ratio (95% CI) | Cox Univariate P Value | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| D 5 | D 10 | D 15 | D 28 | |||||||
| IL-10 | <5 | Control | 4/22 | 14 | 18 | 18 | 18 | .8 | Reference group | .80 |
| Sarilumab-200 | 4/26 | 8 | 15 | 15 | 15 | 0.79 (0.20–3.15) | .74 | |||
| Sarilumab-400 | 3/27 | 11 | 11 | 11 | 11 | 0.58 (0.13–2.60) | .48 | |||
| >5 | Control | 5/10 | 50 | 50 | 50 | 50 | .17 | Reference group | .29 | |
| Sarilumab-200 | 5/8 | 50 | 62 | 62 | 62 | 1.22 (0.35–4.21) | .75 | |||
| Sarilumab-400 | 1/8 | 12 | 12 | 12 | 12 | 0.22 (0.03–1.86) | .16 | |||
| MCP-1 | <773 | Control | 5/23 | 22 | 22 | 22 | 22 | .30 | Reference group | .34 |
| Sarilumab-200 | 3/23 | 9 | 13 | 13 | 13 | 0.54 (0.14–2.27) | .40 | |||
| Sarilumab-400 | 2/27 | 7 | 7 | 7 | 7 | 0.31 (0.06–1.58) | .16 | |||
| >773 | Control | 4/9 | 33 | 44 | 44 | 44 | .61 | Reference group | .66 | |
| Sarilumab-200 | 6/11 | 36 | 55 | 55 | 55 | 1.23 (0.35–4.38) | .75 | |||
| Sarilumab-400 | 2/8 | 25 | 25 | 25 | 25 | 0.58 (0.11–3.19) | .53 | |||
| IP-10 | <2500 | Control | 2/20 | 10 | 10 | 10 | 10 | .62 | Reference group | .64 |
| Sarilumab-200 | 4/20 | 10 | 20 | 20 | 20 | 1.92 (0.36–10.50) | .45 | |||
| Sarilumab-400 | 2/20 | 10 | 10 | 10 | 10 | 0.98 (0.14–6.96) | .98 | |||
| >2500 | Control | 7/12 | 50 | 58 | 58 | 58 | .05 | Reference group | .10 | |
| Sarilumab-200 | 5/14 | 29 | 36 | 36 | 36 | 0.57 (0.17–1.65) | .27 | |||
| Sarilumab-400 | 2/15 | 13 | 13 | 13 | 13 | 0.19 (0.04–0.90) | .04 | |||
Abbreviations: IL, interleukin; IP, interferon-inducible protein; MCP, monocyte chemoattractant protein.
Correlation of IP-10 With Common Inflammatory Markers
In order to explore alternative surrogate markers of IP-10 among commonly used laboratory parameters, we analyzed the correlation of IP-10 with C-reactive protein (CRP; rho = 0.68, P = .004), lymphocyte count (rho = −0.35, P < .0001), procalcitonin (rho = 0.31, P = .002), and neutrophil-lymphocyte ratio (rho = 0.25, P = .012). Although a weak correlation was found with all of these, none of the parameters individually showed a good diagnostic yield to predict progression to ARDS. Then, we elaborated classification and regression trees (CART) models to assess whether a combination of laboratory parameters and/or clinical variables could help to identify patients at risk of having an IP-10 >2500 pg/mL. Thus, we determined that the combination of respiratory rate (RR) with lactate dehydrogenase (LDH) and CRP levels (for values under their respective cutoff points; RR <23 bpm, LDH <340 UI/L, and CRP <140 mg/L) allowed us to classify those patients with a low probability of having an IP-10 >2500 pg/mL (AUROC, 0.80; 95% CI, 0.72–0.89). In fact, only 2 (6.4%) out of 31 patients fulfilling these criteria had an IP-10 >2500 pg/mL.
DISCUSSION
The present subanalysis of the SARICOR trial confirms that high values of a set of inflammatory cytokines at admission predicts clinical outcome in patients hospitalized with COVID pneumonia. Importantly, a clinical benefit of early IL-6 blockade with sarilumab-400 was only seen in patients harboring high levels of these cytokines, whereas patients with high IL-6 but normal levels in the remaining biomarkers did not seem to benefit from this therapy. According to our results, IP-10 could be a useful clinical tool to guide decisions regarding the use of anti-IL-6 agents in this scenario.
The role of IP-10 as a surrogate marker of a hyperinflammatory state in COVID-19 has been previously assessed in various studies [9, 13–15]. High IP-10 levels have been described as a specific [16, 17] and early biomarker in COVID-19 patients [13] that is associated with severe or critical disease condition [9, 14, 18]. In some cases, a correlation between high levels and viral titers or lung injury has been observed [15, 19]. Nevertheless, these reports are limited by their small sample size or their methodology, which is generally based on correlation analysis among clinical status and cytokine levels, so the observed associations do not necessarily reflect any causation. Furthermore, none of these studies have assessed the role of IP-10 in the response to anti-inflammatory therapies.
Our findings are in line with previous results, as high levels of IP-10 were associated with a higher risk of poorer outcomes. Besides, IP-10 was able to discriminate patients at low risk of progression to ARDS for whom sarilumab did not modify the clinical evolution. In fact, in patients with IP-10 <2500 pg/mL, there were no differences in the rates of the primary outcome between the control group and both sarilumab groups. By contrast, patients with IP-10 >2500 pg/mL had a 58% risk of progression to ARDS, which was notably reduced by sarilumab-400 treatment. The benefit of sarilumab-400 was independent of concomitant corticosteroid use, including the use of very high doses of methylprednisolone. On the other hand, sarilumab-200 did not provide any clinical benefit, irrespective of cytokine levels, although patients in this arm tended to be older and more frequently received very high doses of methylprednisolone, which could have influenced the observed results.
Of note, an inclusion criterion in the SARICOR trial was the presence of IL-6 >40 pg/mL, confirming that a substantial proportion of patients harboring high IL-6 levels will not benefit from IL-6 blockade. In fact, the absolute risk reduction in clinical trials supporting a role of tocilizumab [4, 20, 21] has been modest, whereas other studies failed to prove any benefit [22, 23]. The results presented herein suggest that the target population for IL-6 blockade should probably be narrower than the current recommended criteria [24] and might be better identified by additional markers such as IP-10. In line with this, a clinical trial of patients with COVID and high levels of soluble urokinase plasminogen activator receptor (suPAR) showed that anakinra was associated with better outcomes [25]. This reinforces the notion that individualized therapy guided by additional biomarkers could improve treatment results.
The potential of IP-10 in this scenario may attend to the IP-10 signaling pathways. This soluble factor is involved in Th1 cellular phenotype response, carrying out a potent pro-inflammatory chemotactic activity that induces the migration of macrophages, activated T cells, other monocytes, and natural killer to damage tissues. Th-1 response has been associated with some inflammatory diseases such as autoimmune thyroiditis, psoriatic arthritis, systemic lupus erythematosus, and many others, where the role of IP-10 as an inflammation target has been previously proposed [26, 27]. Particularly, increased circulating concentrations of IP-10 have been linked to severity in acute respiratory infections, SARS, or influenza and with the development of ARDS [28, 29]. Likewise, IP-10 has been described as a potent macrophage recruiter, whose participation in the inflammatory response seems to play a crucial role in lung injury in COVID-19 patients. In this context, it has been proposed that SARS-CoV-2 induces the secretion of IL-6 in the lung epithelium, which stimulates the production of IP-10 in the setting of a Th-1 cell response. This induces a chemotactic action that leads to a macrophage infiltration in the lungs, which stimulates the production of further IL-6, which perpetuates the recruitment of immune-active cells in a downward spiral that leads to cytokine storm onset [30]. Besides, independent IP-10 production by epithelial lung cells after SARS-CoV-2 infection has been reported, generating not only a potent chemoattractant effect but also direct damage to endothelial cells [28]. These data suggest that IP-10 is closely linked to the IL-6 pathway and is decisive in the lung hyperinflammation process in SARS-CoV-2 infection. Thus, IP-10 levels could help identify patients at risk of presenting a severe course of COVID-19 who might benefit from anti-IL6 therapy, in line with our results.
A potential limitation of IP-10 as a clinical tool is that it is not available for routine determination in most laboratories, even when it is not a complex or expensive procedure, and that interpretation and quantification are not standardized. For this reason, the role of commonly used inflammatory markers such as CRP has been assessed in previous studies [7, 10], which have shown a correlation of high CRP levels with clinical evolution. In our study, although a good correlation of CRP with IP-10 levels was observed, we were unable to identify a cutoff value for CRP or other commonly used serum or cellular markers, with a good predictive yield. Nevertheless, we found that with the combined use of RR, LDH, and CRP levels, it is possible to identify patients with a high probability of having low levels of IP-10. This could help to select a subgroup of patients who would not eventually benefit from anti-IL-6 blockade, whereas, in the remaining patients, the assessment of IP-10 should be considered to guide treatment decisions.
There are some limitations to our study. First, there was a small sample size and a relatively low number of events: First, the sample size was determined for the main analysis and not for this secondary exploratory analysis, and second, a few cases were lost from the original trial population due to lack of serum sample collection. This fact could hinder the finding of differences between subgroups in other biomarkers, rather than IP-10. Despite this, we were able to prove differences for IP-10, which emphasizes that the association was robust. Second, among the 14 patients not included herein due to lack of serum samples, 4 patients reached the primary outcome. Although a risk of bias due to missing results cannot be excluded, the present study population showed similar characteristics to the original trial population. In addition, our conclusions could be limited by the fact that we did not measure biomarker levels after sarilumab implementation. So, we do not know if the benefit of sarilumab was accomplished by a decrease in circulating IP-10 levels, which would reinforce the biological plausibility of our hypothesis. Finally, the study was conducted in unvaccinated individuals, which could impact the application of our results. Although vaccination has changed the clinical picture of the pandemic in developed countries, vaccination coverage is still poor in many parts of the world. Besides, some high-risk patients, such as those who are immunocompromised or older, are still at substantial risk for hospitalization and death in spite of vaccination. For this reason, defining the role of prognostic biomarkers and improving the effectiveness of therapy in vaccinated patients requiring hospitalization due to severe forms of COVID-19 are still needed.
In summary, the role of IP-10 in COVID-19 patients is promising as a possible prognostic biomarker and a potential tool for clinical decision-making. The usefulness of IP-10 as a risk factor for progression to ARDS and as predictor of response to immunotherapy should be further evaluated in vaccinated patients and for other agents, such as tocilizumab. Additionally, the IP-10 pathway seems to be a critical step in lung injury that leads to ARDS in patients with viral infections, which should be considered for exploring new therapeutic targets.
CONCLUSIONS
In this study, we confirmed the role of IP-10 as a predictor of progression to ARDS in patients hospitalized with COVID-19 pneumonia and IL-6 levels >40 pg/mL. Importantly, an IP-10 value >2500 pg/mL might discriminate those individuals who would clearly benefit from an early blockade of IL-6 with sarilumab 400 mg. Our findings provide a potential role of this cytokine as a marker, and possibly as a target, to optimize treatment strategies in the early stage of patients with COVID-19 pneumonia.
Supplementary Material
Acknowledgments
Author contributions. Dr. Merchante and Dr. Torre-Cisneros had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Torre-Cisneros. Acquisition, analysis, or interpretation of data: all authors. Statistical analysis: Merchante, Trigo, Gutiérrez-Gutiérrez, and Torre-Cisneros. Drafting of the manuscript: Merchante, Trigo. Critical revision of the manuscript for important intellectual content: all authors. Obtained funding: Torre-Cisneros. Study supervision: Merchante and Torre-Cisneros.
Financial support. This work was supported by the Consejeria de Salud y Familias, Junta de Andalucia, Spain (COVID-19 Research Program, project code COVID-0013-2020). B.G.G. and J.T.C. are supported by General Sub-Directorate of Networks and Cooperative Research Centers, Ministry of Science and Innovation, Spanish Network for Research in Infectious Diseases (REIPI RD16/0016/0001, RD16/0016/0008)—co-financed by the European Regional Development Fund, “A Way to Achieve Europe, Operational Program Smart Growth 2014–2020.” J.C.G. is supported by SCReN (Spanish Clinical Research Network) funded by the ISCIII-Sub-Directorate General for Research Assessment and Promotion through projects PT17/0017/0032 and PT20/0039. R.L.L., C.D.L.F., J.T.-C., and B.G.-G. are supported by the Center of Biomedical Investigation Network for Infectious Diseases (CIBERINFEC) funded by ISCIII through projects CB21/13/00049 and CB21/13/00012. The funding sources had no role in the study design; in the collection, analysis, or interpretation of the data; in the writing of the report; or in the decision to submit the paper for publication.
Contributor Information
Marta Trigo-Rodríguez, Unidad de Enfermedades Infecciosas y Microbiología, Hospital Universitario de Valme, Universidad de Sevilla, Instituto de Investigaciones Biomédicas de Sevilla, IBIS (Universidad de Sevilla, Junta de Andalucía, CSIC), Sevilla, Spain.
Sheila Cárcel, Unidad de Gestión Clínica de Cuidados Intensivos, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Hospital Universitario Reina Sofía, Universidad de Córdoba (UCO), Córdoba, Spain.
Ana Navas, Unidad de Inmunología y Alergia, Hospital Universitario Reina Sofía, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Universidad de Córdoba (UCO), Córdoba, Spain.
Reinaldo Espíndola-Gómez, Unidad de Enfermedades Infecciosas y Microbiología, Hospital Universitario de Valme, Universidad de Sevilla, Instituto de Investigaciones Biomédicas de Sevilla, IBIS (Universidad de Sevilla, Junta de Andalucía, CSIC), Sevilla, Spain.
José Carlos Garrido-Gracia, Unidad de Ensayos Clínicos, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Hospital Universitario Reina Sofía, Universidad de Córdoba (UCO), Córdoba, Spain.
María Ángeles Esteban Moreno, Servicio de Medicina Interna, Hospital Universitario Torrecárdenas, Almería, Spain.
Rafael León-López, Unidad de Gestión Clínica de Cuidados Intensivos, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Hospital Universitario Reina Sofía, Universidad de Córdoba (UCO), Córdoba, Spain.
Pedro María Martínez Pérez-Crespo, Unidad de Enfermedades Infecciosas y Microbiología, Hospital Universitario de Valme, Universidad de Sevilla, Instituto de Investigaciones Biomédicas de Sevilla, IBIS (Universidad de Sevilla, Junta de Andalucía, CSIC), Sevilla, Spain.
Eduardo Aguilar Alonso, Servicio de Medicina Intensiva, Hospital Infanta Margarita, Córdoba, Spain.
David Vinuesa, Unidad de Gestión Clínica de Enfermedades Infecciosas, Hospital Universitario Clínico San Cecilio, Granada, Spain.
Alberto Romero-Palacios, Unidad de Enfermedades Infecciosas, Hospital Universitario Puerto Real, Instituto de Investigacion Biomédica de Cádiz (INiBICA), Cádiz, Spain.
Inés Pérez-Camacho, Servicio de Enfermedades Infecciosas, Hospital Regional Universitario de Málaga, Instituto de Investigación Biomédica de Málaga (IBIMA), Málaga, Spain.
Belén Gutiérrez-Gutiérrez, Unidad de Gestión Clínica de Enfermedades Infecciosas, Microbiología y Medicina Preventiva, Hospital Universitario Virgen Macarena, Instituto de Biomedicina de Sevilla (IBIS), Seville, Spain.
Francisco Javier Martínez-Marcos, Unidad de Enfermedades Infecciosas, Hospital Universitario Juan Ramón Jiménez, Huelva, Spain.
Concepción Fernández-Roldán, Unidad de Enfermedades Infecciosas, Hospital Universitario Virgen de las Nieves, Granada, Spain.
Eva León, Unidad de Enfermedades Infecciosas y Microbiología, Hospital Universitario de Valme, Universidad de Sevilla, Instituto de Investigaciones Biomédicas de Sevilla, IBIS (Universidad de Sevilla, Junta de Andalucía, CSIC), Sevilla, Spain.
Alexandra Aceituno Caño, Servicio de Medicina Interna, Hospital Universitario Torrecárdenas, Almería, Spain.
Juan E Corzo-Delgado, Unidad de Enfermedades Infecciosas y Microbiología, Hospital Universitario de Valme, Universidad de Sevilla, Instituto de Investigaciones Biomédicas de Sevilla, IBIS (Universidad de Sevilla, Junta de Andalucía, CSIC), Sevilla, Spain.
Elena Perez-Nadales, CIBER de Enfermedades Infecciosas (CIBERINFEC), Instituto de Salud Carlos III (ISCIII), Madrid, Spain; Spanish Network for Research in Infectious Diseases (REIPI), Instituto de Salud Carlos III (ISCIII), Madrid, Spain; Grupo de Enfermedades Infecciosas, Instituto de Investigaciones Biomédicas de Córdoba, Hospital Universitario Reina Sofía, Universidad de Córdoba (IMIBIC/HURS/UCO), Córdoba, España; Departamento de Química Agrícola, Edafología y Microbiología, Universidad de Córdoba, Córdoba, Spain.
Cristina Riazzo, Servicio de Microbiología, Hospital Universitario Reina Sofía-IMIBIC, Córdoba, Spain.
Carmen de la Fuente, Unidad de Gestión Clínica de Cuidados Intensivos, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Hospital Universitario Reina Sofía, Universidad de Córdoba (UCO), Córdoba, Spain.
Aurora Jurado, Unidad de Inmunología y Alergia, Hospital Universitario Reina Sofía, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Universidad de Córdoba (UCO), Córdoba, Spain.
Julián Torre-Cisneros, Servicio de Enfermedades Infecciosas, Hospital Universitario Reina Sofía, Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Universidad de Córdoba (UCO), Córdoba, Spain.
Nicolás Merchante, Unidad de Enfermedades Infecciosas y Microbiología, Hospital Universitario de Valme, Universidad de Sevilla, Instituto de Investigaciones Biomédicas de Sevilla, IBIS (Universidad de Sevilla, Junta de Andalucía, CSIC), Sevilla, Spain.
Supplementary Data
Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
References
- 1. Del Valle DM, Kim-Schulze S, Huang H-H, et al. An inflammatory cytokine signature predicts COVID-19 severity and survival. Nat Med 2020; 26:1636–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Dorward DA, Russell CD, Um IH, et al. Tissue-specific immunopathology in fatal COVID-19. Am J Respir Crit Care Med 2021; 203:192–201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Jørgensen MJ, Holter JC, Christensen EE, et al. Increased interleukin-6 and macrophage chemoattractant protein-1 are associated with respiratory failure in COVID-19. Sci Rep 2020; 10:21697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. WHO Rapid Evidence Appraisal for COVID-19 Therapies (REACT) Working Group, Domingo P, Mur I, et al. Association between administration of IL-6 antagonists and mortality among patients hospitalized for COVID-19: a meta-analysis. JAMA 2021; 326:499–518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Merchante N, Cárcel S, Carlos Garrido-Gracia J, et al. Early use of sarilumab in patients hospitalized with COVID-19 pneumonia and features of systemic inflammation: the SARICOR randomized clinical trial. Antimicrob Agents Chemother 2022; 66:e0210721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Liu F, Li L, Xu M, et al. Prognostic value of interleukin-6, C-reactive protein, and procalcitonin in patients with COVID-19. J Clin Virol 2020; 127:104370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Martínez-Sanz J, Muriel A, Ron R, et al. Effects of tocilizumab on mortality in hospitalized patients with COVID-19: a multicentre cohort study. Clin Microbiol Infect 2021; 27:238–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Sayah W, Berkane I, Guermache I, et al. Interleukin-6, procalcitonin and neutrophil-to-lymphocyte ratio: potential immune-inflammatory parameters to identify severe and fatal forms of COVID-19. Cytokine 2021; 141:155428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Kesmez Can F, Özkurt Z, Öztürk N, Sezen S. Effect of IL-6, IL-8/CXCL8, IP-10/CXCL 10 levels on the severity in COVID 19 infection. Int J Clin Pract 2021; 75:e14970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Camon AM, Alonso R, Muñoz FJ, et al. C-reactive protein cut-off for early tocilizumab and dexamethasone prescription in hospitalized patients with COVID-19. Sci Rep 2022; 12:5250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Gallagher JC, Morgan RL. Successful immunomodulators for treatment of COVID-19 have opened the pathway for comparative trials. Clin Microbiol Infect 2023; 29:7–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. León López R, Fernández SC, Limia Pérez L, et al. Efficacy and safety of early treatment with sarilumab in hospitalised adults with COVID-19 presenting cytokine release syndrome (SARICOR STUDY): protocol of a phase II, open-label, randomised, multicentre, controlled clinical trial. BMJ Open 2020; 10:e039951. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Lu Q, Zhu Z, Tan C, et al. Changes of serum IL-10, IL-1β, IL-6, MCP-1, TNF-α, IP-10 and IL-4 in COVID-19 patients. Int J Clin Pract 2021; 75:e14462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Guo J, Wang S, Xia H, et al. Cytokine signature associated with disease severity in COVID-19. Front Immunol 2021; 12:681516. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Liu Y, Zhang C, Huang F, et al. Elevated plasma levels of selective cytokines in COVID-19 patients reflect viral load and lung injury. Natl Sci Rev 2020; 7:1003–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Petruccioli E, Najafi Fard S, Navarra A, et al. Exploratory analysis to identify the best antigen and the best immune biomarkers to study SARS-CoV-2 infection. J Transl Med 2021; 19:272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Tamayo-Velasco Á, Peñarrubia-Ponce MJ, Álvarez FJ, et al. Evaluation of cytokines as robust diagnostic biomarkers for COVID-19 detection. J Pers Med 2021; 11:681. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Wang GL, Gao HX, Wang YL, et al. Serum IP-10 and IL-7 levels are associated with disease severity of coronavirus disease 2019. Cytokine 2021; 142:155500. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Kwon JS, Kim JY, Kim MC, et al. Factors of severity in patients with COVID-19: cytokine/chemokine concentrations, viral load, and antibody responses. Am J Trop Med Hyg 2020; 103:2412–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Abani O, Abbas A, Abbas F, et al. Tocilizumab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial. Lancet 2021; 397:1637–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Gordon AC, Mouncey PR, Al-Beidh F, et al. Interleukin-6 receptor antagonists in critically ill patients with COVID-19. N Engl J Med 2021; 384:1491–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Salama C, Han J, Yau L, et al. Tocilizumab in patients hospitalized with COVID-19 pneumonia. N Engl J Med 2021; 384:20–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Salvarani C, Dolci G, Massari M, et al. Effect of tocilizumab vs standard care on clinical worsening in patients hospitalized with COVID-19 pneumonia: a randomized clinical trial. JAMA Intern Med 2021; 181:24–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.COVID-19 Treatment Guidelines. Hospitalized adults: therapeutic management. Available at: https://www.covid19treatmentguidelines.nih.gov/management/clinical-management/hospitalized-adults–therapeutic-management/. Accessed June 22, 2022.
- 25. Kyriazopoulou E, Poulakou G, Milionis H, et al. Early treatment of COVID-19 with anakinra guided by soluble urokinase plasminogen receptor plasma levels: a double-blind, randomized controlled phase 3 trial. Nat Med 2021; 27:1752–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Antonelli A, Ferrari SM, Giuggioli D, Ferrannini E, Ferri C, Fallahi P. Chemokine (C-X-C motif) ligand (CXCL)10 in autoimmune diseases. Autoimmun Rev 2014; 13:272–80. [DOI] [PubMed] [Google Scholar]
- 27. Ruffilli I, Ferrari SM, Colaci M, Ferri C, Fallahi P, Antonelli A. IP-10 in autoimmune thyroiditis. Horm Metab Res 2014; 46:597–602. [DOI] [PubMed] [Google Scholar]
- 28. Jiang Y, Xu J, Zhou C, et al. Characterization of cytokine/chemokine profiles of severe acute respiratory syndrome. Am J Respir Crit Care Med 2005; 171:850–7. [DOI] [PubMed] [Google Scholar]
- 29. Guo J, Huang F, Liu J, et al. The Serum profile of hypercytokinemia factors identified in H7N9-infected patients can predict fatal outcomes. Sci Rep 2015; 5:10942. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Coperchini F, Chiovato L, Rotondi M. Interleukin-6, CXCL10 and infiltrating macrophages in COVID-19-related cytokine storm: not one for all but all for one! Front Immunol 2021; 12:668507. [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.