ABSTRACT
Background There is an urgent need of active treatment for coronavirus disease 2019 (Covid-19). Although efficacy have not been proven, lopinavir/ritonavir 400 mg/100 mg twice daily has been proposed as a treatment of moderate to severe Covid-19. Previously published cohorts showed Covid-19 is associated with major inflammation. To date, no data are available regarding lopinavir/ritonavir plasma concentration and its safety in Covid-19 patients.
Methods Real-world Covid-19 experience based on a retrospective cohort study.
Results On the cohort of 31 patients treated by lopinavir/ritonavir for Covid-19, we observed very high lopinavir plasma concentrations, increased of 4.6-fold (IQR 2.9-6.4), with regards to average plasma concentrations in HIV treatment. All except two patients were above the upper limit of the concentration ranges of HIV treatment. In this cohort, about one over four to five patients prematurely stopped lopinavir/ritonavir therapy due to a moderate adverse drug reaction, mainly hepatic and gastrointestinal disorders.
Conclusion Patients with Covid-19 pneumonitis treated with lopinavir/ritonavir have plasma concentrations dramatically higher than expected. Owing to that high plasma concentration may be required for antiviral activity against SARS-CoV-2, it appears that lopinavir dosage should not be reduced in the absence of adverse effect. About 80% of the patients well tolerated lopinavir/ritonavir therapy under these plasma concentrations. However, cautious is necessary as drug repurposing can be associated with a new drug safety profile.
Funding None
INTRODUCTION
Since early December 2019, a pandemic infectious disease due to a novel coronavirus, called severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is spreading all over the world. To date, no specific therapeutic agent has proven its clinical efficacy against this outbreak. However, due to the urgent need for a treatment, several antiviral drugs are being repurposed and used off-label to treat Covid-19, including lopinavir/ritonavir (LPV/r) (1). The latter is a human immunodeficiency virus (HIV) protease inhibitor approved since years. LPV is prescribed in patients with HIV in association with another protease inhibitor, ritonavir (RTV), used as a potent P450 3A4 cytochrome (CYP) inhibitor in order to dramatically increase LPV plasma exposure. LPV/r has been proposed in previous coronavirus outbreaks in 2003 and 2012 due to SARS-CoV-1 and to Middle East respiratory syndrome coronavirus (MERS-CoV), respectively (2, 3). In vitro activity and data from rodent models showed LPV has an antiviral activity against MERS-CoV(4, 5). It has also proven its potency to inhibit SARS-CoV-2 replication in vitro (6). Furthermore, in an open-label study comparing the addition of LPV/r to ribavirin in patients with SARS-CoV, the authors showed a reduced risk of acute respiratory distress syndrome (ARDS) or death by comparison with a historical control group treated by ribavirin alone (2). Viral load was also reduced in the LPV/r treated patients.
Based on these results, this drug has been considered to be potentially useful in patients with SARS-CoV-2. To date, efficacy of LPV in SARS-CoV-2 patients has been assessed in only one clinical trial in China, the LOTUS trial (7). This open-label randomized controlled clinical trial was performed in order to evaluate the efficacy and safety of LPV/r in severe Covid-19 patients. No benefit was observed with LPV/r treatment beyond standard care. However, in these patients with severe Covid-19, LPV/r treatment was started lately after a median time of 13 days after the first symptoms. Furthermore, no therapeutic drug monitoring was performed for LPV/r to assess an ideal drug exposure. Nevertheless, due to the absence of specific treatment for Covid-19, LPV/r is still a therapeutic option used for patients with moderate to severe Covid-19 pneumonia. To date, more than 40 clinical trials assessing the efficacy of LPV/r against Covid-19 are ongoing around the world (8).
In order to assess LPV/r plasma concentration and its safety in Covid-19 patients, we conducted a retrospective cohort study as part of routine care in adult patients hospitalized for a Covid-19 infection and requiring oxygen.
RESULTS
Patients
A total of 31 patients have been included during a 14 day period, all receiving at least one dose of LPV/r for a Covid-19 pneumonia. The median age of patients was 63 (interquartile range [IQR], 51-78) years and 71% were men (Table 1). All patients had a typical Covid-19 pneumonia according to radiological pattern and required oxygen with a median oxygen saturation of 92.5% (IQR, 90-96%) in room air. Pulmonary injury at CT scan was mostly moderate to extensive. Four (12%) patients had a doubtful or negative nasopharyngeal swab SARS-CoV-2 PCR, contrasting with a typical chest CT scan for Covid-19. Median C-reactive protein (CRP) and IL-6 levels at admission were 94.1 (45.4-176.0) mg/L and 60.4 (29.7-164.5) ng/mL. Twenty-one (68%) patients did not receive any antibiotic at admission or before being hospitalized. LPV/r was started within the first 48 hours after hospitalization and median time to starting LPV/r after symptom onset was 8 days (IQR, 7-10 days). No patient had drug-drug interaction involving RTV and 13 (42%) patients did not received concomitant therapy. LPV/r treatment duration was 7 (IDR, 3-8) days. At the end of the LPV/r course, five patients had recovered, eight patients were transferred to intensive care unit or died, 17 were still hospitalized with oxygen dependency. Clinical outcome is unknown for one patient that has been transferred to another hospital during treatment.
LPV/r plasma concentrations
Of the 31 patients who received LPV/r for Covid-19, plasma assays were not available for three patients due to technical issues and a total of 28 patients were analyzed. LPV plasma concentrations ranges from 4,732 to 35,012 ng/mL. Median level for trough (Cmin) and peak (Cmax) were 20,153 (IQR, 16,633-26,505) ng/mL and 22,550 (IQR 12,720-33,612) ng/mL, respectively. In all patients, considering either trough or peak for LPV and RTV, plasma concentrations were above the average concentrations observed during HIV treatment (4,000 ng/mL and 9,000 ng/mL for trough and peak concentrations, respectively) (Figure 1A). All except two patients had plasma concentration over the upper limit of the concentration ranges observed in HIV. With regards with these plasma concentration ranges, LPV concentrations in Covid-19 patients were increased of about 4.6-fold (IQR 2.9-6.4) (Figure 1B). Extent of LPV plasma concentration increase was not associated with CRP or IL-6 levels (data not shown).
Safety
LPV/r was discontinued before the end of scheduled course for 14 (45%) patients (Table 2). Reasons for early ending LPV/r therapy were the occurrence of adverse drug reaction, therapeutic limitation or patient deceased and poor efficacy in seven (22%), four (13%) and three (10%) patients, respectively. Suspected adverse drug reaction were assessed as possibly related to LPV/r therapy. They consisted in four cases of liver injuries (three cases moderate cytolytic hepatitis between three and six times above the upper limit of normal range and one case of isolated hyperbilirubinemia), two cases of gastrointestinal disorders (nausea/vomiting and diarrhea) and one case of psychiatric disorders (agitation/anxiety). These adverse effects were mild and all patients recovered after drug withdrawn. In patients with cytolytic hepatitis, LPV plasma concentrations were 6.1-fold above average plasma concentrations in HIV patients.
DISCUSSION
To our knowledge, this is the first study showing that Covid-19 patients had very high LPV plasma exposure with regards to the therapeutic levels in HIV. Although LPV/r regimen was 400/100 mg bid as in HIV treatment, LPV plasma concentrations were in median about 4.6-fold higher, and up to 8-fold higher in some patients. All patients of this retrospective cohort had a LPV plasma concentration over the average concentrations observed in HIV patients. As LPV is extensively metabolized by the hepatic cytochrome P450 system, almost exclusively by the CYP3A isozyme, LPV/r therapy was not started in the event of drug-drug interaction. None was detected during therapy and no patient had liver cirrhosis, which could have impaired LPV/r clearance. Only two patients had a medical history of hepatitis, including one patient with an active hepatitis B without hepatic insufficiency.
Some clinical evidence show that patients with moderate to severe Covid-19 have a major inflammation. Elevated levels of blood IL-6 are commonly seen in severe Covid-19 illness and could led to a cytokine storm syndrome (9). Previous studies in macaques inoculated with SARS-CoV found that defects in T-cell and B-cell function and the excess production of type 2 cytokines could lead to a deficiency in control of viral replication and prolonged proinflammatory responses (10). In Covid-19 patients, a persistent blood virus load and an exacerbated inflammatory response has been shown to be partially driven by the transcriptional factor NFκB (11). These results are in line with our cohort where patients had very high CRP and IL-6 levels at starting therapy. Besides, it is long-standing known that inflammatory responses and infections decrease drug metabolism capacity in human and experimental animals (12). In cultured human hepatocytes, direct treatments with tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), interleukin-6 (IL-6), interferon-γ (IFN-γ) and transforming growth factor-β (TGF-β) can reduce the expression of CYP1A2, CYP2C8 and CYP3A4 (13, 14). In animals and humans, several studies or reports have shown the role of inflammation in reducing drug metabolism and CYP450 inhibition (15, 16). CYP3A4 expression is transcriptionally regulated by pregnane X receptor (PXR), which is a ligand-dependent transcription factor (17). NF-κB activation by tumor necrosis factor-α, in a setting of inflammatory response, has been shown to suppress PXR activation and PXR-mediated gene expression (17, 18). This results in a reduced CYP450 expression and a decreased drug metabolism, especially CYP450-mediated metabolism. Other genes regulated by PXR, including ABCB1, coding for P-glycoprotein transporter, could also have their expression decreased, contributing to decrease non-renal drug clearance, such as LPV/r. Owing to its low oral bioavailability and extensive metabolism by CYP3A4 isoenzyme, LPV needs to be co-administered with low doses of RTV, a potent CYP3A inhibitor, to achieve drug concentrations high enough to inhibit viral replication (19). An added reduced expression of CYP3A4 could dramatically increase LPV plasma concentrations in Covid-19 patients.
Interestingly, nevertheless very high plasma concentrations in all assayed patients, LPV/r was overall well tolerated. The main adverse effects were moderate hepatobiliary disorders that have been attributed to LPV/r therapy. In HIV, moderate-to-severe elevations in serum aminotransferase levels (>5 times the upper limit of normal) are found in 3% to 10% of patients, although rates may be higher in patients with HIV-HCV coinfection (20). These elevations are usually asymptomatic and self-limited and can resolve even with continuation of the medication. Drug causality could be hardly assessable as about 20% of the patients with Covid-19 have been reported to have increased transaminases (9). However, in our cohort, transaminases quickly decreased after stopping lopinavir, making drug causality probable. Overall, adverse effects reported in this Covid-19 cohort are in line with LPV/r safety profile (21). Adverse effects led to stop LPV/r treatment in seven (22%) patients, which could appear elevated, even though, due to the relative limited number of patients treated, conclusions are difficult to draw. In the LOTUS trial, assessing LPV/r efficacy in Covid-19, treatment had been stopped for adverse events in 13.8% of the patients in the LPV/r group.(7) As a comparison, in previous clinical trials with LPV/r in HIV patients, about 11% to 14% of the patients experienced grade 2-4 adverse effects, mainly gastro-intestinal or hepatic disorders (22, 23). However, physicians should be aware that drug repurposing is a high risk condition, which can be associated with a different drug safety profile (24). The drug is used in a setting not having been correctly assessed, that could led to an increase of adverse effects and an unfavorable risk-to-benefits ratio (25). Finally, in other words, 78% of the patients in our cohort did not experienced adverse drug reaction, which could be surprisingly in this context of very high plasma concentrations. One could hypothesize that the relationship between the upper threshold of 8,000 ng/mL observed in HIV patients and lopinavir safety is uncertain, at least during a short treatment as in Covid-19.
Keeping high LPV plasma concentration appears essential considering a possible clinical efficacy of LPV in Covid-19. Antiviral activity assays on cultured cells showed that the 50% effective concentration (EC50) of LPV varied between 8.0 and 11.6 μM (i.e. between 5,030 and 7,290 ng/mL) for MERS-CoV.(5, 26, 27) Considering SARS-CoV, in vitro EC50 was 17.1 μM (i.e. 10,800 ng/mL), and a concentration as high as 25,000 ng/mL has been reported to inhibit viral replication depending on cell type used (5, 28). For SARS-CoV-2, an intro study showed that LPV but not RTV inhibits viral replication with EC50 at 26.6 μM (i.e. 16,800 ng/mL).(6) In our cohort, median trough plasma level was 20,153 (IQR 16,633-26,505) ng/mL barely at these concentrations or just above. The question of whether these plasma concentrations are effective to inhibit in vivo SARS-CoV-2 replication, especially in lungs, is unknown. Furthermore, Covid-19 associated vasculopathy or thrombosis could limit pulmonary diffusion. Altogether, this suggest that high plasma concentrations are required to have an antiviral activity. Finally, as most HIV protease inhibitors, LPV has very high protein binding of about 98.5% (27). Therefore, in considering whether a dose is appropriate for SARS-CoV-2, it is critical to consider free drug concentrations. For standard HIV-1, Cmin after 75 ng/mL that is substantially below EC50 reported for coronaviruses (27). Therefore, targeting SARS-CoV-2 antiviral activity in patients is challenging. It appears that LPV/r dosage should not be reduced on the basis of therapeutic drug monitoring in the absence of adverse drug reaction.
Our study has several limitations. First, due to the retrospective design using data from routine care, plasma assays were drawn at different times that prevent from an accurate pharmacokinetic estimation. Second, according to different practices in wards, Covid-19 severity at baseline was heterogeneous between patients. However, most importantly, in all except three due to drawing issue, LPV plasma concentrations regardless assays timing were unexpectedly high comparing HIV experience. Finally, unbound concentrations, which are the active ones, were not estimated.
In conclusion, in our single center Covid-19 experience, we found that LPV/r treated patients had unexpected very high plasma concentrations, well above those observed in HIV patients. However, LPV dosage should not be reduced in the absence of adverse effect, owing that to high plasma concentration may be required for an in vivo antiviral activity as suggested by previous studies. We observed that about one over four to five patients stopped LPV/r therapy in relation with moderate adverse drug reactions attributed to LPV therapy. Cautious is needed in this context of drug repurposing, which can be associated with a new drug safety profile. These early data need to be confirmed with prospective studies.
MATERIALS AND METHODS
Study population
We used data collected from routine care of patients hospitalized for Covid-19 in a single center setting (Cochin Hospital, Paris). Patients hospitalized in medicine wards and starting a treatment by LPV/r for Covid-19 pneumonitis between March 18th and April 1st have been included in the cohort. Covid-19 pneumonia was considered in the presence of CT scan injuries evocating a SARS-CoV-2 infection and in the absence of other diagnosis.(29) Covid-19 patients were considered as confirmed or suspected if they had a positive or negative PCR, respectively. Patients with ARDS (defined by the need for an invasive or non-invasive mechanical ventilation) at admission were not included.(30) Data regarding patient demographics, type of Covid-19 infection, patient outcome and LPV/r clinical and biological safety were collected. Patients have been identified according to LPV/r dispensing by the hospital pharmacy. The study has been performed in accordance with the declaration of Helsinki and received approval by the Cochin Hospital Institutional Review Board (number 2020-08019).
Treatment strategy
According to our local protocol at the time of the study, patients were eligible to receive a LPV/r treatment as specific anti-Covid-19 therapy if they (i) had a pneumonitis evocating a SARS-CoV-2 infection at CT scan and (ii) were requiring oxygen (minimal flow rate 2-3 liters per minute). LPV/r was started within 24-48 hours after admission. LPV/r therapy consisted in a dose of 400/100 mg twice a day in tablets during seven to ten days, according to ward’s practices. Drug-drug interactions were carefully monitored before initiation and during LPV/r treatment. Patients were follow-up until the end of LPV/r treatment. For each patient, a therapeutic drug monitoring was performed in a routine care setting within the first three days of therapy.
LPV and ritonavir assays
Plasma assays were drawn in a routine care setting, according to ward’s practices. Plasma concentrations of LPV and ritonavir were quantified using high-performance liquid chromatography tandem mass spectrometry (Xevo TQD, Waters®) by the department of clinical pharmacology (Cochin Hospital, Paris, France). A volume of 100 μL of plasma sample was mixed with 300 μL of methanol. Then, the mixture was centrifuged at 21,000 g during 10 minutes. A volume of 50 μL of supernatant was diluted by 50 μL of waters (0.05% formic acid, v/v), and 10 μL was injected into the chromatographic system. The separation was carried out on a BEH C18 analytical column (1.7μm, 1.7μm, 50*2.1 mm, Waters, Saint-Quentin, France). The phase mobile composed of 60% of water (0.05 % formic acid, v/v) and 40% of methanol (0.05% formic acid, v/v). The method was validated according to the EMA guideline for bioanalytical method validation (https://www.ema.europa.eu/en/ich-m10-bioanalytical-method-validation). The calibration range was from 160 (limit of quantification, LOQ) to 40,000 ng/mL for LPV and 10 (LOQ) to 2,500 ng/mL for ritonavir. The inter- and intra-assay bias and CV were inferior to 15% for LPV and ritonavir. No significant matrix effect on quantification of LPV and ritonavir was found in this method. LPV plasma concentrations were considered as peak level (Cmax) and as through level (Cmin) for assays drawn between one hour to five hours, and from eight hours after tablet intake, respectively. In HIV treatment, observed Cmin and Cmax ranges are 1,000-8,000 ng/mL (target > 4,000 ng/mL for non-naïve patients) and 7,000-11,000 ng/mL, respectively (31, 32).
LPV/r treatment safety
LPV/r treatment safety was assessed in a routine care setting. Cases of suspected adverse drug reaction were spontaneously notified to the Pharmacovigilance regional center. After case-by-case review assessment and review by a senior pharmacologist, cases were reported to the French pharmacovigilance system.
Statistics
Quantitative and qualitative data are presented using non-parametric settings. Descriptive analyses have been performed using Prism GraphPad 7.0 and R statistical package version 3.6.1 or later (The R Foundation for Statistical Computing, https://www.R-project.org/).
Data Availability
Data are available upon request
FOOTNOTES
Adverse effects cases have been reported to the French pharmacovigilance system under the numbers PV20200174, PV20200224 and PV20200250-4.
AUTHORS’ CONTRIBUTIONS
Study design: LC, SBo, IG, JMT, SBe
Data collection: LC, SBo, IG, LHP, LR, PL, CA, EC, JZ, NC, BT, SK, RB, JMT, YZ, SBe
Data analysis: LC, SBo; LHP, SBe
Draft the manuscript: LC, SBo, SBe
Critically review the manuscript: LC, SBo, EC, NC, BT, SK
Approved the manuscript: LC, SBo, IG, LHP, LR, PL, CA, EC, JZ, NC, BT, SK, RB, JMT, YZ, SBe
CONFLICT OF INTEREST STATEMENTS
None
ROLE OF FUNDING SOURCE
None