A Randomized Clinical Trial of the Efficacy and Safety of Interferon β-1a in Treatment of Severe COVID-19

Patients and baseline features.

During the study period, 139 patients were screened, of whom 103 subjects were eligible. Considering dropouts, 81 patients (42 in the IFN and 39 in the control group) completed the treatment for further analysis (Fig. 1). Males were 54.3% of patients. The mean age ± standard deviations (SD) in the IFN and control groups was 56.0 ± 16 and 59.5 ± 14 years, respectively. Fifty-two (64.19%) patients had positive nasopharyngeal real-time PCR (RT-PCR) for SARS-CoV-2, and 29 (35.81%) patients were diagnosed according to the clinical signs/symptoms along with the imaging findings. Hypertension (38.3%), cardiovascular diseases (28.4%), diabetes mellitus (27.2%), endocrine disorders (14.8%), and malignancy (11.1%) were common baseline diseases. Endocrine disorders were dyslipidemia and hypothyroidism. There was no significant difference in terms of demographic data and baseline diseases between the groups. The most frequent chief complaints of patients were cough, fever, and dyspnea (Table 1). APACHE II score at the time of intensive care unit (ICU) admission was not significantly different between the two groups (15.3 ± 4 in the IFN group versus 14.5 ± 3 in the control group, P = 0.79).

Vital signs and laboratory data.

Median time from the onset of symptoms (according to patients’ reports) to the administration of IFN was 10 days (interquartile range [IQR], 8 to 13). The vital signs at the time of hospital admission were not statistically different, except respiratory rate was significantly higher in the IFN group (22 versus 20, respectively, P = 0.009). Comparing the baseline laboratory data at the time of hospital admission revealed that the median blood urea nitrogen level was higher in the IFN than in the control group. The median INR value also was significantly higher in the control group than in the IFN group. Other laboratory findings were comparable (Table 2).

Treatment strategies.

Hydroxychloroquine is the main medication in Iran’s national protocol for the treatment of COVID-19. Lopinavir-ritonavir or atazanavir-ritonavir may be added to hydroxychloroquine in severe cases. The antiviral regimens were not significantly different between the two groups. All antivirals were continued for 10 days. Deep-vein thrombosis prophylaxis and stress ulcer prophylaxis were considered for patients where indicated. Based on patients’ clinical conditions, azithromycin, intravenous ascorbic acid, antibiotics, intravenous immunoglobulin (IVIG), or a corticosteroid was added to the antiviral regimens. A corticosteroid (methyl prednisolone, hydrocortisone, or dexamethasone) was administered for 61.9% and 43.6% of patients in the IFN and the control groups, respectively. The corticosteroid dose was equivalent to 250 mg methylprednisolone daily for 3 days. In addition, 35.7% and 25.6% of patients in the IFN and the control group, respectively, received IVIG. The dose of IVIG was 5 g daily for 3 days. Supportive care modalities and administered medications are summarized in Table 3.

Outcomes and complications.

As a primary outcome, the time to clinical response was not significantly different between the IFN and control groups (9.7 ± 5.8 versus 8.3 ± 4.9 days, respectively, P = 0.95), which is shown in the Kaplan-Meier plot (Fig. 2). The log rank test also revealed that there was no statistically significant difference between the groups, considering time to clinical response (hazard ratio [HR], 1.10; 95% CI, 0.64 to 1.87; P = 0.72).

The six-category ordinal scale was assessed at days 0, 7, 14, and 28 (Table 4). On day 0, there was no significant difference between the groups in terms of the components of this scale. On day 7 of therapy, 19% of patients in the IFN group were discharged with no deaths. At this time, 28% of patients in the control group were discharged and 25% died. However, the difference was not statistically significant (odds ratio [OR], 0.60; 95% CI, 0.21 to 1.69). On day 14, the results were statistically significant, and 66.7% versus 43.6% of patients in the IFN group and the control group, respectively, were discharged (OR, 2.5; 95% CI, 1.05 to 6.37). When the administration of IVIG and corticosteroid was considered a probable confounding factor, the adjusted odds ratio was 4.05 (95% CI, 1.42 to 11.55).

Regarding the time of IFN initiation, the analysis showed that early administration significantly reduced mortality (OR, 13.5; 95% CI, 1.5 to 118). However, late administration of IFN did not show significant effects (OR, 2.1; 95% CI, 0.48 to 9.6).

Other secondary outcomes, such as duration of hospital stay, length of ICU stay, and duration of mechanical ventilation, were not statistically different. However, more patients were extubated in the IFN group (P = 0.019). Additionally, the 28-day overall mortality was significantly lower in the IFN than the control group (19% versus 43.6%, respectively, P = 0.015). In multivariate analysis, the effect of IFN on the reduction of mortality was shown (OR, 2.95; 95% CI, 1.08 to 8.03). When the model was adjusted for administration of IVIG and corticosteroid as confounding factors, the effect not only remained but also became stronger (OR, 6.65; 95% CI, 1.67 to 26.45). Kaplan-Meier plot for assessment of survival was also determined (HR, 0.375; 95% CI, 0.16 to 0.87; P = 0.024) (Fig. 3).

Complications during the hospitalization course, incidence of organ failure, and adverse effects were not different between the groups. Injection-related side effects (fever, chills, myalgia, and headache a few hours after injection of IFN) happened in 8 (19%) patients (Table 5).

A hypersensitivity reaction occurred in one patient who received IFN. The reaction presented with maculopapular rash on the trunk and both upper and lower limbs. However, the patient was taking herbal medicine for cough consisting of thyme and honey and other medications, like hydroxychloroquine and lopinavir-ritonavir, concomitantly. Interferon was discontinued after the fourth dose, and rashes began to disappear within 3 days. According to the Naranjo score, the reaction possibly was due to IFN.

Neuropsychiatric problems were detected in 4 patients in the IFN group. Two cases experienced severe agitation and two cases complained of mood swings (mostly depression). One of the patients with mood swings had a history of mild depressive disorder in past years. Out of 4 cases, two patients were in the hospital for nearly 1 month. Neuropsychiatric side effects of IFN are unlikely to happen in short-term use. All patients received a psychiatric consult. According to the Naranjo score, IFN possibly and probably caused neuropsychiatric problems in three and one patients, respectively.

Conclusions.

Although IFN did not change the time to reach a clinical response, added to the national protocol, it significantly increased the discharge rate on day 14 and decreased 28-day mortality. Improved survival rate was significant when patients received IFN β-1a in the early phase of the disease. Adverse effects of IFN β-1a were injection-related, neuropsychiatric problems, and hypersensitivity reactions that all were tolerable and resolved during the follow-up period.

Study design.

This open-label, randomized clinical trial was conducted to assess the efficacy and safety of IFN-β-1a in the treatment of adult (aged ≥18 years) patients diagnosed with COVID-19. Patients were admitted from 29 February to 3 April 2020 in Imam Khomeini Hospital Complex, the main central hospital in Tehran, capital of Iran.

Eligibility criteria.

The diagnosis of COVID-19 was according to either a positive real-time PCR (RT-PCR) of the deep nasopharyngeal secretions or clinical signs/symptoms plus imaging findings highly suspicious for COVID-19. RNA extraction was performed using the viral nucleic acid extraction kit (no. YVN50/YVN100) provided by RBC Bioscience, Taipei, Taiwan. After processing nasopharyngeal samples, rapid RT-PCR was performed by a novel coronavirus (2019-nCOV) nucleic acid diagnostic kit (PCR-fluorescence probing) from Sansure Biotech (S3102E) (Changsha, China) according to the instructions of the manufacturer.

Cough, fever, myalgia, dyspnea, headache, chills, anorexia, gastrointestinal problems, and chest discomfort were considered symptoms of the disease. More than 50% involvement of the field of examination with typical findings of COVID-19, including peripheral, bilateral ground glass opacities (GGO) or multifocal GGO of rounded morphology with or without consolidations or crazy paving, was defined as a positive computed tomography scan.

Patients with severe COVID-19 were included (45). These patients had at least one of the following conditions: (i) hypoxemia (need for noninvasive or invasive respiratory support to provide capillary oxygen saturation above 90%), (ii) hypotension (systolic blood pressure less than 90 mm Hg or vasopressor requirement), (iii) renal failure secondary to COVID-19 (according to KDIGO definition) (46), (iv) neurologic disorder secondary to COVID-19 (decrease of 2 or more scores on the Glasgow Coma Scale), (v) thrombocytopenia secondary to COVID-19 (platelet count less than 150,000/mm³), and (vi) severe gastrointestinal symptoms secondary to COVID-19 (vomiting/diarrhea that caused at least mild dehydration).

Allergy to IFNs, receiving IFNs for any other reason, severe depression, previous suicide attempts, alanine amino transferase (ALT) >5× the upper limit of the normal range, and pregnant women were defined as the exclusion criteria of the study.

The Ethics Committee of Tehran University of Medical Sciences approved the study (approved ID IR.TUMS.VCR.REC.1398.1052), and the protocol of the trial was registered (registered ID IRCT20100228003449N28). The protocol of the study was explained to the patients, and written informed consents were obtained from them or from their next of kin.

Sample size and randomization.

Due to the lack of a similar study at the start of this trial, for producing significant difference, the estimated time to clinical response was assumed. By estimating 15 days to reach clinical response and to detect a difference of 5 days, the sample size was estimated. A number of 46 patients was estimated for each group.

Randomization was performed for all patients. An allocation ratio of 1:1 was accounted between IFN and the control group in randomization, but the patients had to receive at least four injections of IFN to be included in the final analysis. The randomization was performed by permuted block randomization and block sizes of 2, 4, and 6, used randomly. The statistician prepared a list of numbers for randomization. Sequentially numbered opaque envelopes were prepared according to the list, and envelopes were delivered to the clinical investigators. The statistician that performed randomization and analysis was unaware of the treatment and follow-up of patients.

Treatment protocols.

Patients in the IFN group received IFN β-1a in addition to the national protocol medications. Each 44-μg/ml (12 million IU/ml) dose of interferon β-1a (ReciGen, CinnaGen Co., Iran) was subcutaneously injected three times weekly for two consecutive weeks. The control group received only the standard of care. The standard of care (the hospital protocol) consisted of hydroxychloroquine (400 mg twice a day [BID] on the first day and then 200 mg BD) plus lopinavir-ritonavir (400 and 100 mg, respectively, BD) or atazanavir-ritonavir (300 and 100 mg, respectively, daily) for 7–10 days. Primary care, respiratory support, fluid, electrolyte, analgesic, antipyretic, corticosteroid, and antibiotic treatments were recommended in the hospital protocol if indicated. The duration of the study was 2 weeks, and the patients were monitored for 4 weeks.

Demographic data, baseline characteristics, and laboratory data were recorded for each patient. APACHE II score at the time of ICU admission was calculated for critically ill patients. All patients were daily monitored in terms of vital signs, medical interventions, and clinical conditions during the study course.

The need for respiratory support (invasive, noninvasive, or not required) was assessed by a physician regularly. Patients were assessed to fit in one of the six categories of the ordinal scale at days 0, 7, 14, and 28 of inclusion (47). If discharged, the patient was followed up by phone. Readmission was surveyed until 3 May.

Outcomes assessment.

The primary outcome of the study was time to reach clinical response. Clinical response was defined according to the six-category ordinal scale (47). This scale classifies patients into six categories according to the severity of the viral pneumonia: (1) discharge; (2) hospital admission, not requiring oxygen; (3) hospital admission, requiring oxygen; (4) hospital admission, requiring noninvasive positive pressure ventilation; (5) hospital admission, requiring invasive mechanical ventilation; (6) death. Time to clinical response was considered the number of days required to at least two scores of improvement on the scale or patient’s discharge, whichever occurred sooner. Secondary outcomes were duration of mechanical ventilation, duration of hospital stay, length of ICU stay, 28-day mortality, effect of early or late (before or after 10 days of the onset of symptoms) administration of IFN on mortality, adverse effects, and complications during the hospitalization. The following adverse effects of the antiviral regimen/IFN β-1a and complications during the hospitalization course were assessed: gastrointestinal (nausea, vomiting, diarrhea, abdominal pain, and pancreatitis), anaphylaxis and allergic reactions (rash, urticaria, angioedema, bronchospasm, and dyspnea related to medication administration), IFN injection-related reaction (skin erythema and necrosis, chills, fever, and flu-like symptoms after injection), neuropsychiatric (sleep disorder, psychosis, agitation, depression, and mania), renal impairment (according to KDIGO definition) (46), hepatic impairment (hepatic aminotransferase serum levels raised more than three times the upper limit of normal or serum total bilirubin above 2 mg/dl) (48), indirect hyperbilirubinemia (direct bilirubin level less than 15% of the total bilirubin) (49), incidence of thromboembolism (deep-vein thrombosis or pulmonary thromboembolism), incidence of nosocomial infections, and diagnosis of septic shock (according to the surviving sepsis campaign guidelines) (50). The Naranjo scale was used for evaluation of adverse effects of IFN. In this standard scale, several items, including previous reports, relationship with starting the agent, improvement after discontinuation, challenge result, alternative causes, data of drug assay, dose dependency, and patient’s history of the same reaction were considered. Total scores of >8, 5 to 8, and 1 to 4 were considered definite, probable, and possible correlations between the use of IFN and the adverse drug reaction, respectively (51).

Statistical analysis.

The quantitative variables were reported as means ± standard deviations (SD) if they had normal distributions or as median with IQR if they did not pass the normality test. The qualitative ones were reported as number (percent). For comparing the quantitative variables, t test or Mann-Whitney test was used. The qualitative variables were compared by chi-square test.

The analysis was performed on a per-protocol basis, and patients who did not receive at least four doses of IFN were not included. The Kaplan-Meier plot and log rank test were used to compare the number of days to reach the clinical response and survival time between the groups. The HR and 95% CI for clinical improvement and death were estimated by Cox proportional hazards model. The odds ratio was also calculated for patients who received IFN early versus late. The multiple logistic regression model was performed for the variables that were significantly different between the groups according to the univariate analysis. In the multiple logistic regression model, corticosteroid and immunoglobulin were adjusted as confounding factors, and the adjusted odds ratio was reported. SPSS software (version 24) was used for statistical analyses.