Clinical characteristics of Coronavirus Disease 2019 in Hangzhou, China: the combination of lopinavir/ritonavir, interferon, and arbidol may be a well choice for antiviral therapy in common cases

A retrospective, single-center case series study was conducted in a designated hospital for special treatment of COVID-19 in Hangzhou, China. Clinical characteristics of COVID-19 patients were described and compared between the common and severe groups. Multiple linear regression and sensitivity analysis was conducted to evaluate the antiviral effects of the 3-drug combination therapy (lopinavir/ritonavir, interferon, and arbidol).

throughput sequencing assay for nasal and pharyngeal swab specimens in Hangzhou Center for Disease Control (CDC) [2]. Demographic information, epidemiological characteristics (including exposure history), clinical characteristics (including comorbidities, initial signs, duration of fever, length of hospital stay, and duration with pharyngeal swab nucleic acid positive), chest computed tomographic (CT) scan, and laboratory ndings of each patient were extracted from the electronic medical record system of Xixi Hospital of Hangzhou. Two independent reviewers (D.Y. and J.Y.) extracted the data. All differences were resolved by evaluating the eligibility of the original data and discussion prior to the nal analysis.
All the 110 patients were included in the CT imaging analysis, which was performed according to a recent study by two experienced radiologists with over 10-year experience (Y.Z. and Z.H.C.), and in case of disagreement, they would consult to reach an agreement [11]. Brie y, lesion distribution, including left lung (upper or lower lobe), and right lung (upper, middle or lower lobe), lesion location, including peripheral, central or involving both peripheral and central locations, and lesion density, including ground glass opacity, consolidation, thickness of interlobular and intralobular septa, enlarged lymph nodes within the mediastinum and pleural effusion, were analyzed.

Study de nitions
The incubation period was de ned as the interval between the potential earliest date of contact of the transmission source (wildlife or person with suspected or con rmed case) and the potential earliest date of symptom onset (i.e., cough, fever, fatigue, or myalgia). And the summary statistics of incubation periods were calculated based on patients who had clear information regarding the speci c date of exposure.
The status of negative pharyngeal swab specimen was de ned as that two consecutive pharyngeal swab specimens were both negative, and the interval time of the two specimens should be at least 24 hours.
Based on this de nition, the duration with pharyngeal swab specimen positive was de ned as the interval between the earliest date with a pharyngeal swab specimen positive and the rst time with a pharyngeal swab specimen negative.
For the clinical cure criteria in China [12], if the patient's body temperature has returned to normal for more than 3 days, the respiratory symptoms have improved signi cantly, the lung imaging has shown well absorption of in ammation, and the nucleic acid test for respiratory pathogens has been negative for two consecutive times with sampling interval at least 24 hours, the patient could be discharged or transferred to the corresponding departments to treat other diseases.
For antiviral therapy, all patients of 18 years or older were treated by the combinations of three drugs, if there were no contraindications: lopinavir and ritonavir tablets (lopinavir 200 mg/ritonavir 50 mg), 2 tablets at a time, twice a day; recombinant interferon α-2b, 5 million units at a time, adding 2 milliliter normal saline, spray inhalation, twice a day; and arbidol hydrochloride tablets, 200 mg at a time, three times a day. The course of the antiviral therapy was at least 10 days. If the patient could not complete the 10-day, three-drug antiviral therapy, because of the side effects of any of the 3 drugs or other reasons, it was de ned as incomplete antiviral therapy. Side effects of the antiviral therapy were de ned as common side effects of any of the 3 drugs, after excluding other possible causes.
A non-pneumonia case was de ned as a con rmed case by RT-PCR with fever and/or respiratory symptoms, but no radiographic evidence of pneumonia. An asymptomatic case was de ned as a con rmed case with normal body temperature or minor discomfort. A mild case was de ned as a con rmed case with fever, respiratory symptoms and radiographic evidence of pneumonia, while a severe case was de ned as a mild case with dyspnea or respiratory failure. Common cases included all patients except the severe cases.

Statical analysis
Continuous variables were expressed as the means and standard deviations or medians and ranges as appropriate. Categorical variables were summarized as the counts and percentages in each category. To compare the continuous variables for data of different patient groups, a two-tailed t-test or Mann-Whitney U test was used as appropriate. Ch-square tests or Fisher's exact tests were used for categorical variables as appropriate. Multiple linear regression analysis was used to estimate the independent effects of the 10-day, 3-drug antiviral therapy on hospital staying time. To conduct sensitivity analysis, two models, Model 1 and Model 2, were established. Model 1 (n = 93) included all the discharged cases of 18 years or older, while Model 2 included 51 cases after excluded important confounding factors from Model 1. All statistical analysis was performed with SPSS software version 19.0. P-value (two-sided) less than 0.05 was considered statistically signi cant.

Results
Demographic, epidemiological and clinical characteristics As seen in Figure 1, 113 patients with COVID-19 had been hospitalized in our hospital up to March 9, 2020. The 113 patients included 13 patients currently in the hospital and 100 patients discharged. Especially, there were 3 patients whose SARS-CoV-2 RNA was found positive again after discharge among the 13 patients currently in hospital, and these 3 patients were also included in the 100 patients discharged, given their rst hospitalization. We excluded the 3 repeatedly hospitalized cases in the analysis. The total number of severe patients was 11 (11/110, 10%), and the total number of common patients was 99 (99/110, 90%), which included 89 patients discharged and 10 patients currently in the hospital. The total number of non-pneumonia cases was 14 (14/110, 12.7%). The total number of asymptomatic cases was 4 (4/110, 3.6%) that included 2 patients (2/110, 1.8%), asymptomatic and with no radiologic evidence.
The median incubation period was 6.5 (0-26) days. Coexisting disorders could be seen in 24.5% of all patients (27/110). The most common initial manifestation was fever (63/110, 57.3%). The most common temperature range was 37.5-38.0 degrees centigrade (42/110, 38.2%). The median duration of fever during hospitalization was 7 (1−18) days. There were signi cant differences between the common group and the severe group in age and coexisting disorders (both P < 0.01).

Treatment and clinical outcome
As seen in Table 2, more patients had completed the 3-drug combination antiviral therapy in the 93 discharged patients of 18 years or older (74/93, 79.6%). Side effects of any of the 3 antiviral drugs seemed more common in the common cases group (55.6% in the common group vs. 9.1% in the severe group, P < 0.01). Seventeen (17/99, 17.2%) patients interrupted antiviral therapy owing to side effects of any of the 3 drugs in the common group, while 1 patient in the severe group (1/11, 0.9%). The use of systemic corticosteroids, immunoglobin, and antibiotics were all more common in severe patients (all P < 0.01). Hospital staying time seemed longer in the common group (median, 14, range, 3-29, in the common group, and median, 5, range, 1-25, in the severe group, P = 0.014), with potential causes that there might be more severe patients transferred to the superior hospital (4/99, 4.0%, in the common group, and 6/11, 54.5%, in the severe group) when they got worse. The median duration with pharyngeal swab specimen positive was 11.5 (2-28) days.
To investigate the associations between the 3-drug combination antiviral therapy (complete vs. incomplete) and hospital staying time (days), multiple linear regression was performed. And to conduct sensitivity analysis, two models, Model 1 and Model 2 (see Table 4 and Supplementary Table 1), were established. Model 1 included 93 discharged patients of 18 years or older (3-drug combination antiviral therapy, complete, n = 74, and incomplete, n = 19) with complete data of hospital staying time, and the adjusted potential confounding factors included gender, age, coexisting disorders, COVID-19 status (common or severe), use of systemic corticosteroids, use of intravenous immunoglobin, and use of intravenous antibiotic. In model 2 (3-drug combination antiviral therapy, complete, n = 42, and incomplete, n = 9), patients treated with systemic corticosteroids, intravenous immunoglobin, and intravenous antibiotic were excluded from Model 1, and the adjusted potential confounding factors included gender, age, and coexisting disorders. Because all the 51 patients were common cases, the factor of COVID-19 status was not included in Model 2 as a covariate. The distribution of potential confounding factors in the two therapy groups (complete vs. incomplete) in Model 1 and Model 2 was displayed in Supplementary Table 1. As seen in Table 4, in both Model 1 and Model 2, signi cant and independent associations between the complete 3-drug combination antiviral therapy (vs. incomplete) and a shorter hospital staying time could be seen (B = -5.970, 95% CI, -9.222, -2.718, β = -0.369, P < 0.01, in Model 1; and B = -5.948, 95% CI, -10.622, -1.274, β = -0.349, P = 0.014, in Model 2). These data suggested the robustness of our models.

Discussion
Given the limited knowledge of the novel pathogen, SARS-CoV-2, and the current worldwide outbreak of COVID-19, effective treatment options are urgently needed. In the present study, we summarized the clinical characteristics of COVID-19 patients in Hangzhou, China, an important imported metropolis outside Wuhan. Our data indicated the effectiveness of the 3-drug combination of LPV/r, recombinant interferon α-2b, and arbidol on COVID-19, especially in adult common cases.
As seen in Fig. 1, the total number of non-pneumonia cases was 14 (14/110, 12.7%). And especially, 4 cases were asymptomatic, including 2 patients with radiologic evidence and 2 patients without radiologic evidence. Under the same de nition of non-pneumonia cases, the ratio of non-pneumonia cases in all symptomatic cases (except severe patients) was about 14.7% (14/95, see Fig. 1) in our study, higher than another research in Beijing, China, conducted by Tian and colleagues (5.4%, 11/203) [13]. And a study from Shenzhen, China, including 55 asymptomatic cases at the time of hospital admission, indicated that the majority of them developed to be mild or ordinary COVID-19 during hospital [14]. Altogether, these data might re ect the characteristics of patients at different stages of SARS-CoV-2 infection, or very few patients might have a subclinical infection. From the perspective of controlling the epidemic, these ndings suggested that the absence of clinical symptoms could not rule out the diagnosis of SARS-CoV-2 infection. Persons with a clear history of exposure to SARS-CoV-2, regardless of clinical symptoms, should be considered for medical observation, home isolation, and further examination [15].
There were 3 cases (3/110, 2.7%) whose SARS-CoV-2 RNA was found to be positive again after discharge in our study. Some case reports also suggested this phenomenon [16,17]. Zhou and colleagues suggested some possible reasons to explain this status in their recent review [18]: 1) Virological properties of SARS-CoV-2 were not su ciently understood given that it was just discovered recently; 2) Some host factors, such as the immune function status or coexisting disorders, might in uence the control of the virus in the body; 3) Use of systemic corticosteroids might be related to the recurrence of SARS-CoV-2 RNA; 4) The tests results had been false negative when the patients were discharged; 5) Secondary infection of SARS-CoV-2 might occur in some patients. In addition, another noteworthy phenomenon was that some patients had a long duration of viral shedding. In our study, the longest time with pharyngeal swab specimen positive was 28 days. And the longest duration of viral shedding was 37 days in a recent study by Zhou and colleagues [19]. People with positive SARS-CoV-2 RNA of respiratory tract specimens are the infectious source of COVID-19. During treatment or follow-up after discharge, the level and duration of infectious virus replication are important factors in assessing the risk of transmission and guiding decisions regarding the isolation of patients and the length of antiviral treatment. Data above indicated that SARS-CoV-2 RNA of respiratory tract specimens might be persistent or recurrent positive during the course, calling for further researches on dynamic pro le and infectivity assessment of SARS-CoV-2 infection.
Severe patients seemed older and to have more coexisting disorders, lower levels of lymphocyte counts, sodium and chloride, and higher levels of C reactive protein and procalcitonin (see Table 1 and Table 2). These ndings were consistent with the results of other studies with larger sample sizes [19][20][21].
Researches from Wuhan, China, indicated that older age (odds ratio 1.10, 95% CI 1.03-1.17, per year increase; P = 0·0043) was an independent risk factor for in-hospital death [19] and showed more comorbidities and higher plasma levels of C reactive protein in severe cases [22]. These data indicate that older persons, particularly those with multimorbidity, are at high risk of being severer or death if infected by SARS-CoV-2. The elderly should be one of the key groups for infection prevention.
Judging from the CT imaging data, lesions being simultaneously present in more than one lobe or with more than one radiologic feature could be seen in a considerable number of patients (see Table 3). And within lobes, peripheral and peripheral involving central lesions were more common. These radiologic characteristics were similar to other reports [15,23]. However, we must note that the CT ndings of COVID-19 are variable. Some patients have no initial abnormal lung ndings and can be misdiagnosed with other common diseases. Sixteen patients in our study had no CT evidence of pneumonia, including 14 non-pneumonia cases and 2 cases, asymptomatic and no radiologic evidence (see Fig. 1). Therefore, epidemiological history, clinical manifestations, and the RT-PCR detection of the viral RNA from a respiratory tract sample can be considered to represent the comprehensive analysis necessary to diagnose SARS-CoV-2 infection [24]. Especially, for cases of high clinical suspicion of SARS-CoV-2 infection with a negative RT-PCR result, the combination of repeated swab tests and CT examinations can be helpful for diagnosis. CT images and the detection of nucleic acid can mutually complement each other in the diagnosis of COVID-19.
Up to now, no speci c treatment has been recommended for coronavirus infection [8]. Given the urgency of the COVID-19 outbreak and the absence of de nitive antiviral drugs or vaccines, diverse treatment regimes have been explored for the treatment of COVID-19, such as oseltamivir, ganciclovir, LPV/r, remdesivir, chloroquine phosphate, herbal treatments, and passive immunization [8,[25][26][27][28][29]. Some of these treatments may have been tried out of desperation, and among these, some show initial promise.
Approved protease inhibitors including lopinavir and ritonavir have been reported to be active against SARS-CoV and Middle East respiratory syndrome coronavirus (MERS-CoV). They were also potential virally targeted agents for SARS-CoV-2 [7]. A study in the early stage of the SARS-CoV-2 epidemic in Wuhan, China, showed the initial therapeutic effect of arbidol, with a higher discharge rate in the arbidoltreated group compared to the arbidol-untreated group [22]. In addition, a recent retrospective cohort study showed that arbidol and LPV/r combination therapy was associated with a signi cant elevated negative conversion rate of coronavirus test in 7-day and 14-day and signi cantly improved the chest CT scans in 7-day, compared with LPV/r monotherapy [30]. Pegylated interferon α-2a and − 2b, approved for the treatment of hepatitis B virus (HBV) and hepatitis C virus (HCV) infection, had the potential activity to stimulate innate antiviral responses in patients infected with SARS-CoV-2 [7]. In our hospital, since the beginning of the SARS-CoV-2 epidemic in Hangzhou, all COVID-19 patients of 18 years or older were treated with the combination of LPV/r, recombinant interferon α-2b, and arbidol, if no contraindications.
The course was at least 10 days. Our results showed that completing the 3-drug combination antiviral therapy was an independent related factor to a shorter hospital staying time, both in the whole group including common and severe patients and in the group excluding important confounding factors (see Table 4). Especially, given that 82 cases in Model 1 (82/93, 88.2%) were common cases and the 51 cases in Model 2 were all common cases, we draw the conclusion cautiously that the adult common COVID-19 patients might bene t from the 3-drug combination antiviral therapy. Meanwhile, another issue that needed attention was the side effects of antiviral drugs. In our study, 50.9% of the patients reported side effects of any of the 3 antiviral drugs, while 16.4% of the patients discontinued at least one of the 3 antiviral drugs because of side effects. Overall, prospective or even randomized researches with larger sample sizes are called for to evaluate the e cacy and tolerance of the 3-drug combination antiviral therapy in COVID-19 patients. And phase 3 or 4 clinical trials for LPV/r, arbidol or recombinant interferons in the treatment of COVID-19 are currently in progress [7,19]. These might help to stratify patients to improve e cacy and reduce side effects.
Notably, a recent randomized clinical trial comparing LPV/r monotherapy plus standard care with standard care alone indicated that no bene t was observed with LPV/r treatment beyond standard care in hospitalized adult patients with severe COVID- 19 [29]. This trial included only the severe cases with "an oxygen saturation (Sao 2 ) of 94% or less while they were breathing ambient air or a ratio of the partial pressure of oxygen (Pao 2 ) to the fraction of inspired oxygen (Fio 2 ) (Pao 2 :Fio 2 ) at or below 300 mg Hg". In our study, there were more common cases. As seen in the "Treatment and clinical outcome, Results" section and Supplementary Table 1, Model 1 included 93 patients, including only 11 severe cases and 51 cases in Model 2 were all common cases. In addition, in our study, patients were treated with 3-drug combination therapy, not with one drug monotherapy. There was a hypothesis that reducing the viral load as soon as possible might bene t the delay of the progression of lung lesions. A combination of two or three antiviral drugs might facilitate the rapid decline of viral load. These two differences between the resent clinical trial and our study might cause differences in treatment effects. And in the clinical trial conducted by Doctor Cao and colleagues, in the modi ed intention-to-treat analysis, which excluded three patients with early death, the between-group difference in the median time to clinical improvement (median, 15 days vs. 16 days) was signi cant, although modest [29]. As the authors discussed in their article [29], further studies should be done to answer these questions: "whether earlier LPV/r treatment in COVID-19 could have clinical bene t?", "whether LPV/r treatment given at a certain stage of illness can reduce some complications in COVID-19?", and "whether combining LPV/r with other antiviral agents, as has been done in SARS and is being studied in MERS-CoV might enhance antiviral effects and improve clinical outcomes?". Corresponding to these questions, in our study, most patients were common cases and combination therapy including LPV/r was evaluated. So the results of our research might be a well complement to the research conducted by Doctor Cao and colleagues.
There were several notable limitations in our study. Firstly, respiratory tract specimens were detected by RT-PCR, while results reporting is qualitative rather than quantitative in the clinical practice. This would cause the loss of some information. For example, the dynamic pro le of SARS-CoV-2 RNA load could not be carefully observed. Secondly, the viral load in the serum was not detected in the clinical practice. That was a potentially useful marker related to the severity of COVID-19. Thirdly, this was a retrospective study, so recall bias might exist. That would in uence the accuracy of the evaluation of some parameters, such as the incubation period. Meanwhile, this retrospective character also had advantages. For example, most patients had been discharged, which allowed us to obtain relatively complete data including the hospital staying time.

Availability of data and materials
The datasets used and analyzed during the current study available from the corresponding author on reasonable request.

Ethics approval and consent to participate
This study was approved by the Ethics Committee of Xixi Hospital of Hangzhou and the written informed consent was waived because of the retrospective nature of the study and the urgent need to collect data.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no competing interests.    (male as reference), age (< 60 years old or >= 60 years old, < 60 years old as reference), coexisting disorders (none as reference), COVID-19 status (common or severe, common as reference), use of systemic corticosteroids (no as reference), use of intravenous immunoglobin (no as reference), use of intravenous antibiotic (no as reference), and the 3-drug therapy (incomplete or complete, incomplete as reference). 2 In model 2 (3-drug combination antiviral therapy, complete, n = 42, and incomplete, n = 9), patients treated with systemic corticosteroids, intravenous immunoglobin, and intravenous antibiotic were excluded. All 51 patients were common cases. Covariates included gender (male as reference), age (< 60 years old or >= 60 years old, < 60 years old as reference), coexisting disorders (none as reference), and the 3-drug therapy (incomplete or complete, incomplete as reference). 3 Three-drug combination antiviral therapy: all the patients of 18 years or older were treated by the combinations of three drugs, if there was no contraindication, which included lopinavir and ritonavir tablets (lopinavir 200mg/ritonavir 50 mg), 2 tablets at a time, twice a day; recombinant interferon α-2b, 5 million units at a time, spray inhalation, twice a day; and arbidol hydrochloride tablets, 200 mg at a time, three times a day. The course of the antiviral therapy was at least 10 days. If the patient could not complete the 10-day, three-drug antiviral therapy, because of the side effects of any of the 3 drugs or other reasons, it was de ned as incomplete antiviral therapy. Abbreviation: B, regression coe cient; β, standard partial regression coe cient; CI, con dence interval for each regression coe cient. Figure 1 Patients ow and enrollment. A non-pneumonia case was de ned as a con rmed case by RT-PCR with fever and/or respiratory symptoms, but no radiographic evidence of pneumonia. An asymptomatic case was de ned as a con rmed case with normal body temperature or minor discomfort. A mild case was de ned as a con rmed case with fever, respiratory symptoms and radiographic evidence of pneumonia, while a severe case was de ned as a mild case with dyspnea or respiratory failure. Common cases included all the patients except the severe cases.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Supplementarydata.docx