Discussion
Infection control measures and supportive cares are currently the main clinical management for COVID-19. However, some medications are being used based on the results of laboratory, animal and clinical studies [7]. The current study is the first to retrospectively identify severity, frequency, mechanism, and clinical management of pDDIs in hospitalized patients with COVID-19. We found that 37.9% of COVID-19 patients were exposed to at least one major or contraindicated pDDI during 24-48 h of hospital admission. More than half of pDDIs were occurred between lopinavir/ritonavir (branded as Kaletra or Aluvia) and medications metabolized by CYP3A4. Additive interactions between antiplatelets/nonsteroidal anti-inflammatory drugs (NSAIDs) and anticoagulants were the second common pDDI in COVID-19 patients.
Lopinavir, a protease inhibitor, is widely used for the treatment of HIV. It is combined with ritonavir, another protease inhibitor, which inhibits CYP3A4 enzyme and increases the half-life of lopinavir [14]. Lopinavir/ritonavir has activity against severe acute respiratory syndrome coronavirus (SARS-CoV) and middle east respiratory syndrome coronavirus (MERS-CoV) both in vitro and in animal models [15, 16]. Recently, an open-label, randomized, phase 2 trial revealed efficacy and safety of lopinavir/ritonavir combined with interferon beta-1b and ribavirin in the treatment of hospitalized patients with COVID-19 [17]. Both lopinavir and ritonavir have important DDIs due to inhibition of CYP3A4. It is essential to identify commonly used medications in patients with COVID-19 which are metabolized with this enzyme and may have severe interactions with lopinavir/ritonavir.
Our results show that HMG-CoA reductase inhibitors (atorvastatin, rosuvastatin) are the most frequent interacting medications with lopinavir/ritonavir in our patients. Some studies showed that co-administration of different protease inhibitors with atorvastatin increased the area under the concentration-time curve (AUC) of atorvastatin from 1.7 to 9.4 folds (depending on the type of protease inhibitor). Reducing atorvastatin dose to 20 mg/day and monitoring for signs and symptoms of its toxicity (myalgia, rhabdomyolysis, liver function test (LFT) abnormalities) should be taken into account when a protease inhibitor is used concomitantly [18]. However, differentiation between signs and symptoms of HMG-CoA reductase inhibitors toxicity and COVID-19 is challenging because myalgia and abnormal LFT can also be seen in patients with COVID-19 [9, 19]. Rosuvastatin also have the same interaction with protease inhibitors and its dose should not be exceeded than 10 mg/day when co-administered with lopinavir/ritonavir [20]. Rhabdomyolysis is a serious consequence of this interaction which was reported in patients who received lopinavir/ritonavir with atorvastatin or rosuvastatin [21, 22].
Anxiolytic/hypnotic/sedative agents such as alprazolam, chlordiazpoxide and midazolam were administered in our patients. According to the Lexi-Interact database, interaction of alprazolam or midazolam with lopinavir/ritonavir is rated as good and excellent; respectively, in terms of reliability. Other benzodiazpine agents that are less expected to interact with protease inhibitors (e.g. lorazepam, oxazepam, temazepam) can be administered instead of alprazolam [23]. Severe prolonged sedation and increased length of hospital stay were reported among patients receiving a combination of a protease inhibitor and intravenous midazolam [24]. Therefore, this interaction is categorized as contraindicated in terms of severity rating [25]. However, clinical outcomes of HIV patients (on protease inhibitors regimen) who received intravenous midazolam for procedural sedation and those who received diazepam were similar [26]. This combination should be administered with caution and reduced doses of intravenous midazolam should be considered [25].
SARS-CoV-2 may affect cardiovascular system and lead to myocardial injury, acute heart failure, or worsening pre-existing cardiovascular disease [27]. Therefore, interaction of cardiovascular agents with investigational medications for COVID-19 should be mentioned. In the current study, nondihydropyridine calcium channel blockers (diltiazem), cardiac glycosides (digoxin), antiarrhythmic agents (amiodarone, lidocaine), and isosorbide dinitrate had major or contraindicated interactions with lopinavir/ritonavir. Metabolism inhibition of nondihydropyridine calcium channel blockers can lead to increase their serum concentration and the risk of AV nodal blockade. Monitoring the toxicity of calcium channel blockers and a 50% dose reduction of diltiazem may be required [28]. The mechanism of major interaction between lopinavir/ritonavir and digoxin is probably due to inhibition of P-glycoprotein transporter and reduction of digoxin renal clearance by ritonavir [29]. Reducing digoxin dose (30% to 50%) or dosing frequency as well as monitoring digoxin levels is recommended [30]. Lopinavir/ritonavir may enhance the serum concentration of amiodarone and subsequently its QTc-prolonging effect. This combination should be avoided when possible, if not, a dose reduction and serial ECG monitoring is recommended [31]. Combination of amiodarone and azithromycin should also be cautioned in COVID-19 patients due to QTc prolongation synergism effect. Our previous study showed that DDIs are risk factor for prolonging QTc in ICU admitted patients [32]. Nevertheless, the interaction between hydroxychloroquine and azithromycin, two medications with QTc prolongation effect co-administered in some COVID-19 patients, is graded as minor (B) interaction [33]. For B interactions, no intervention is needed according to the Lexi-Interact database but calculating the risk score of QTc prolongation before administration of this combination to COVID-19 patients is a wisely measure.
Excessive inflammation, platelet activation, endothelial dysfunction, and stasis may lead to arterial and venous thrombotic disease in COVID-19 patients [34]. Drug interactions of common anticoagulants should be considered in these patients. In the current study apixaban and rivaroxaban were administered to the patients. Among novel oral anticoagulants (NOACs), apixaban and rivaroxaban are substrates for both P-glycoprotein transporter and CYP3A4. Therefore, concomitant administration with inhibitors of CYP3A4 and P-glycoprotein transporter (e.g. lopinavir/ritonavir) should be avoided due to increased serum concentration of apixaban and rivaroxaban and risk of bleeding [35].
Additionally, co-administration of antiplatelets or NSAIDs with anticoagulants may increase the risk of bleeding [36]. Because of underlying cardiovascular diseases in some COVID-19 patients, they may regularly use antiplatelet agents. NSAIDs may also be prescribed to relieve myalgia [37] or for their probable effect on SARS-CoV-2 (specifically indomethacin) [38]. Anticoagulants side effects should be monitored in COVID-19 patients receiving these combinations.
Administration of salmeterol or combination medications containing salmeterol may be indicated in COVID-19 patients with underlying chronic respiratory diseases. Lopinavir/ritonavir increases serum concentration of salmeterol [39] and may predispose patients to cardiovascular adverse effects of beta2-agonists.
Another important interaction that may occur in COVID-19 male patients with older age is between tamsulosin and lopinavir/ritonavir. Although the adverse consequences of this interaction has not been evaluated in the clinical studies, this combination is contraindicated according to the tamsulosin drug label [40]. When the combination must be used, monitoring tamsulosin adverse effects (particularly orthostatic hypotension) is recommended.
Logistic regression analysis confirmed that two comorbidities (IHD and CRDs) and ICU admission were significantly associated with occurrence of major and contraindicated pDDIs. Patients with IHD or CRDs may receive some medications which have severe interactions with investigational drugs for COVID-19 especially protease inhibitors. The current study shows other investigational medications including hydroxychloroquine, ribavirin, remdesivir, favipiravir, interferon beta or intravenous immunoglobulin do not interact with commonly used medications for comorbidities. Therefore, these medicines are much safer than protease inhibitors in terms of severe DDIs. We also found critically ill patients are more prone to DDIs because they have complicated conditions and receive more medications.
The first limitation of the current study is that it cannot predict the occurrence rate of actual DDIs. Because subjective and objective evidence for an actual DDI usually are not accurately recorded specifically during a critical pandemic condition and it is a limitation of the retrospective studies. These evidence can be collected with a prospective study in order to report actual DDIs. The second limitation is using one drug interaction database. Evaluating severity rating of some pDDIs with another drug interaction database (except Lexi-Interact) showed a little variation. It would be better to check the interactions with more than one database in order to report more accurate results.