COVID-19's Effect on Infective Endocarditis in People Who Inject Drugs

This case demonstrates the overlap between IE and COVID-19, which makes accurate clinical diagnosis difficult and results in delayed treatment.

The patient is a 28 year-old woman with a history of opiate use disorder (OUD) on methadone who has recently relapsed. She suffers from untreated chronic hepatitis C virus (HCV) infection and polycystic ovary syndrome (PCOS) and is on oral contraception. The patient presented to the emergency department 12 days after the onset of symptoms including low grade fever and fatigue. In the emergency department, she tested positive for COVID-19. Within 3 days after admission to the hospital, and despite high dose dexamethasone, baricitinib therapy, and broad spectrum antibiotics, she progressed to severe acute respiratory distress syndrome (ARDS) that required intubation. She remained febrile. A D-dimer level was 100K ng/mL. The computed tomography pulmonary embolus (CT-PE) examination demonstrated bilateral subsegmental pulmonary emboli. No vegetations were identified on transthoracic echocardiogram (TTE). That night, a venous ultrasound of the right leg showed findings consistent with a new deep venous thrombosis (DVT). Testing was unable to determine whether the source was subsegmental pulmonary emboli, DVT, COVID-19 infection, OUD-IE, or was associated with OCPs. A heparin drip was started, cultures of blood and sputum were negative, and a transesophageal echocardiogram (TEE) was ordered. The patient continued to be febrile, and septic emboli on the pulps of her fingers were identified. Empiric antibiotics were restarted. Repeat blood cultures remained negative. Should there be further work-up?

The challenge with patients with OUD-IE with concurrent COVID-19 infection is the significant overlap between symptoms, physical findings, laboratory, and imaging findings between both these conditions. The purpose of this article is to provide a time-based approach to recognize the predominant pathologic process at any one given time.

Please see Personalized Therapy.

The development of OUD-IE in normal heart valves may be related to the scarring effect of various materials that are contained within the injected drug. A more likely explanation is the ability of Staphylococcus aureus (S aureus) to directly invade and damage endothelial cells of the valve.

COVID-19 infection may produce a systemic inflammatory state that may scar the endothelium with resultant loss of anticoagulation properties. This pro-thrombotic state of COVID-19 allows for the formation of thrombi on valvular surfaces, enabling S aureus or other infectious pathogens to adhere to valvular surfaces, intravascular devices, and extravascular devices.

The pro-thrombotic property of COVID-19 infection is hypothesized to be the major factor for the development of IE among otherwise healthy individuals, those with a history of OUD, or an immunocompromised state for up to several months after being infected with COVID-19.[1, 2, 3]  Before the COVID-19 pandemic, 70% of OUD-IE was right-sided and usually involved the tricuspid valve. More than one valve may be involved during the COVID-19 era. There are cases that recognize infection by multiple organisms.[4, 5, 6]  COVID-19 coinfection may predispose to multiple valve involvement in all types of IE.[4]  

There appear to be three major pathological mechanisms of COVID-19 infection: thrombosis, inflammation, and coinfection. The inflammatory response to COVID-19 is produced by excessive production of various cytokines particularly in the pattern of interleukin 1 (IL-1) and IL-6 pathways (see Table 1). Immunologic studies have shown that elevated C-reactive protein (CRP) reflects IL-6-mediated pro-inflammatory pathways, whereas elevated ferritin reveals immune dysregulation in IL-1 and interferon-gamma (IFN-γ).[7, 8, 9]  The thrombotic pattern is defined as a CRP less than 10 μg/dL, ferritin less than 3,000 mg/mL, and a D-dimer greater than 5,000 mg/mL, which would require evaluation for pulmonary emboli (PEs) and DVTs before determining the need for anticoagulation. There is evidence that COVID-19 infection may lead to persistent microthrombi that put patients at increased risk for further endothelial damage, myocarditis, and DVT formation even after resolution of clinical symptoms.  

It appears that the involvement of the endothelium leads to activation of platelets. This is brought about by the interaction of the virus' spike proteins with endothelial associated angiotensin converting enzyme 2-receptor (ACE2-R). The end result is recurrent cycles of intravascular coagulation and inflammation of the cardiac endothelium, both of which can lead to the development of OUD-IE.[10]

Table 1: Clinical Complication Patterns of COVID-19 Infection (Open Table in a new window)

Adapted from Garcia-Vidal et al[7]

Garcia-Vidal et al applied these classifications to individual patients in a systematic effort to deliver personalized care. They demonstrated that doing so resulted in a statistically significant improvement in clinical status (see below for further discussion).[7]  

The incidence of OUD-IE cases has risen locally and internationally, with a twofold increase internationally between 2000 and 2018.[11]  Historically, OUD-IE affects older populations due to comorbidities, cardiac hardware, and immunocompromised states, but more recent data documents that OUD-IE is involving younger marginalized populations with low socioeconomic status, injection drug users, the uninsured, and individuals with comorbidities including HIV or HCV infection.[12] Infectious complications among people with OUD-IE have been localized to urban centers; however, between 2016 and 2019, rates increased by 85.8% in urban centers compared with 81.7% in rural locations.[12]

The hallmark of any type of IE is that of a continuous bacteremia. 

A continuous bacteremia is the hallmark of endovascular infections such as OUD-IE. This is defined as a minimum of three sets of blood cultures positive for the same organism that are drawn from different sites at least 15 minutes apart.[16, 17]

Various molecular techniques, such as MALDI-TOF and 16SrDNA PCR PhenoTest BC Kit hold great promise in significantly reducing the turnaround time for the final identification of pathogens and antibiotic sensitivities from blood samples.[18, 19]

Because CRP levels increase during acute and chronic inflammation, including deep-seated infections, sepsis, autoimmune conditions, COVID-19 infection, and IE, serial CRP measurements are a convenient method of documenting therapeutic responsiveness of both diseases.

D-dimer is another marker used to determine the probability of a DVT or PE in a person with COVID-19 and IE given their pro-thrombotic state. Due to its high negative predictive value (NPV), a negative test provides reassurance that the probability of a DVT or a PE is low.

See Infective Endocarditis.

Some studies emphasize the importance of a full cardiac evaluation for COVID-19 positive OUD-IE. The purpose would be to rule out associated viral myocarditis as well as to detect valvular microthrombi that could increase the risk of a patient with OUD developing IE due to a variety of intravascular devices.[16, 17, 20, 21]  Guidelines from the American College of Cardiology/American Heart Association and the Infectious Diseases Society of America recommend that all patients with S aureus bacteremia undergo evaluation with a transthoracic echocardiogram (TTE). We recommend a transesophageal echocardiogram (TEE) to follow a TTE in all patients with OUD who have positive blood cultures. This was not feasible during the COVID-19 pandemic owing to concern for transmission to healthcare workers as well as limited access because of the surge of inpatients. During the pandemic, the number of TEEs declined by 50%. This increased the risk of failing to diagnose or document infection of small endocardial thrombi associated among patients with OUD-IE and COVID-19 infection.[16, 17, 22]  The amount of echocardiograms performed likely has been limited because of the high hospital census. Such has led to significant delays in diagnosis leading to increased morbidity and mortality.[23, 24]

Alternative considerations such as cardiac magnetic resonance imaging and positron emission tomography scanning may be considered as they do not require close contact with personnel for extended periods of time while still documenting valve function and evidence of vegetations.[25] These alternative imaging modalities may reduce the number of cases of occult thrombi that were identified after poor outcomes from acute onset PEs or DVTs.

Challenges of diagnosing and managing of opioid use disorder infective endocarditis (OUD-IE) in the age of COVID-19 are immense. There has been a movement toward establishing multidisciplinary endocarditis teams to better coordinate care and post-hospitalization follow-up.[22, 26]  The European Society of Cardiology has published guidelines for multidisciplinary endocarditis teams, which have lowered mortality rates.[26]  In the United States, two teams (at the University of Michigan and The Massachusetts General Hospital) have lowered in-hospital mortality by expediting consults and post-hospitalization adverse events by engaging addiction medicine services and early initiation of medications for OUDs.[26, 27]  Furthermore, with a multidisciplinary endocarditis team in place, consults and work-up are streamlined owing to established protocols that can improve diagnostic capabilities. Successful IE management requires collaboration among specialties not limited to infectious disease, cardiology, cardiac surgery, neurology, neurosurgery, and addiction medicine providers.

Empiric antibiotic treatment for OUD-IE should cover for methicillin-resistant S aureus (MRSA), methicillin-sensitive S aureus (MSSA), and streptococcal species. Two important points of therapeutic considerations are that vancomycin should no longer be considered empiric treatment and that ampicillin and cefotaxime should be the preferred treatment for E faecalis. 

Prolonged use of empiric antibiotics due to decreased diagnostic certainty has risen in the COVID-19 era and led to heightened concern regarding antimicrobial resistance.[28]  Excessive IV antibiotic use during COVID-19 has led to shifts in multi-drug resistant organisms (MDROs), especially against coagulase-negative S aureus species (CoNS).[19] This has had a particular effect on cephalosporins and fluoroquinolones.[29]

A 4- to 6-week course of IV antibiotics is recommended for OUD-IE. The need for long-term hospitalizations for therapy and disease severity can unnecessarily increase the risk for nosocomial infections.[23, 28, 30, 31, 32, 33, 34]  A single-center retrospective study at a tertiary center identified a rise in MDROs and positive blood cultures on admission for hospitalized patients with COVID-19 at 8.6% compared with 2.5% to 2.9% previously.[35, 36]  A newer glycopeptide, dalbavancin, permits weekly IV antibiotic therapy outside of healthcare facilities for off-label treatment of OUD-IE, which may be beneficial in the COVID-19 era.[35, 36]  If the patient is clinically improving, repeat one to two blood cultures daily for 5 days to ensure clearance of bacteremia. If clearance does not occur, a complete reassessment is required. A steady downward trending CRP indicates improvement in the inflammatory processes of either or both the COVID-19 and OUD-IE to guide the antibiotic course. Hospitalized patients with OUD-IE should be followed by a multidisciplinary team on a regular basis.

The role of antibiotics has not yet been established in the treatment of OUD-IE. In 2019, the “POET” trial compared a partial oral antibiotic regimen with a full course of IV antibiotic treatment for endocarditis showing non-inferiority between the two treatment arms for left sided IE.[11, 37]  A newer study revealed limited evidence in the efficacy of oral antibiotics due to issues with 7% documented failure, 23% lost to follow-up, and 67% with presumed success.[38]  This study does not support the use of oral antibiotics in treating OUD-IE because only 67% of patients were proven to respond to oral therapy. On the basis of current evidence, switching over to oral therapy should be avoided. 

See Infective Endocarditis for details on antibiotic therapy.

Insertion of any prosthetic material among patients with OUD should be avoided as it may lead to increased rates of cardiac or recurrent valvular infection. Right-sided IE accounts for 5% to 10% of all IE cases though it is no longer pathognomonic for OUD-IE.[39]  If a tricuspid valve OUD-IE does not respond adequately to antibiotic therapy, it would be preferable to remove the vegetation and repair the valve instead of inserting a prosthetic valve. Repair may lead to development of a pulsatile liver due to tricuspid regurgitation. Most patients tolerate these conditions. For the same reasons, valve repair is preferred among OUDs who are prone to relapse or continue to use drugs.[39, 40]  Avoid implants if at all possible among OUDs because of injection drug use relapse (eg, pacemakers, cardiac defibrillators, transcatheter aortic valve replacement). This is where a multidisciplinary team approach would be helpful in formulating appropriate plans for patients and risk stratifying based on risk for relapse in high risk patients. An example would be the involvement of the infectious disease team to determine if a patient might benefit from complete valve removal as opposed to valve replacement in a high risk patient where hardware might be promote recurrent IE.[41, 42]

Choosing to anti-coagulate patients with a coinfection of COVID-19 and OUD-IE is a difficult decision and is best handled via a multidisciplinary team approach. Although both harbor prothrombotic properties, as described above, the risks versus benefits for intracranial bleeding due to previously unidentified cerebral lesions is not trivial. Among patients with IE, 8% develop hemorrhagic lesions whereas 60% develop single or multiple microbleeds.[1, 42]

Garcia -Vidal et al published a report describing their experience employing "personalized therapy" of 75 hospitalized patients with COVID-19 infection.[7]  Specific therapy was based on the immune markers presented in Table 1. At the 14th day of treatment, the mortality rate of the study group was 20% as compared with 43.6% for those receiving usual care. At the 28th day of treatment, the mortality rates were 20% and 44.2%, respectively. None of the personalized patients had evidence of OUD-IE; however, this study is notable in providing a framework on which to base initial therapeutic approaches.  

 

The outlook of OUD-IE with concurrent COVID-19 infection is quite variable. Timely identification of the active disease process and institution of appropriate treatment is the most beneficial approach. The overlapping clinical symptomatology of both conditions has led to marked delays in providing appropriate treatment. In addition, infrastructural disarray during the COVID-19 pandemic resulted in delays to the delivery of healthcare (eg, limited follow-up appointments), exacerbated safety nets in place for mental health support, and decreased community-sponsored support groups (eg, food pantries, shelter space).

 

Patient education, especially for those suffering from OUD-IE, should focus on key warning signs of OUD-IE with and without concurrent COVID-19 infection. Such signs include persistent fever, worsening fatigue, and signs of congestive failure such as leg swelling and dyspnea. Patients should be told to avoid taking “street antibiotics” because this will usually delay arriving at the correct diagnosis and treatment. Of course, the COVID-19 vaccine should be made easily available in trying to establish or re-establish psychiatric or psychological housing in addition to other community programs. 

For discussion of specific antibiotic therapies, please see Infective Endocarditis.

Dalbavancin is a novel therapeutic that has valuable attributes for treating IE owing to its potential to reduce length of hospitalization and reduce the risk for relapse due to once weekly IV infusions that do not require ongoing intravenous access. As of early 2022, dalbavancin is approved in the United States for treatment of acute bacterial skin and skin structure infections. Preliminary evidence for use in other infections (eg, osteomyelitis, prosthetic joint infections, and IE) is emerging.[37, 38, 45]  

Vancomycin is no longer considered a preferred therapeutic as it is difficult to titrate and has limited efficacy as resistance patterns shift.  Additionally, use of vancomycin may render other antibiotics (eg, daptomycin) less effective. 

No individual with a history of or current OUD should be discharged with intravascular access present because of the danger of it being used to inject a variety of “street drugs”.[13]

To summarize, the incidence of IE has risen as a result of the opioid epidemic due to the increase in intravenous drug use. Diagnostic uncertainties owing to limited knowledge about COVID-19 have led to delayed diagnosis and treatment among patients with concomitant COVID-19 infection and IE, potentially leading to poor clinical outcomes.

With time, what has become clearer is that alternate diagnoses should be considered if a patient continues to decline despite standard of care for COVID-19 infection. IE should always be considered on the differential in a patient who has a history of IV drug use. Imaging techniques including surface and transesophageal echocardiograms should not be withheld from practice as these can provide lifesaving information throughout the clinical progression of disease.

Finally, with respect to treatment selection, intravenous antibiotics are currently the leading therapeutic for IE in the setting of COVID-19 infection as vegetations can develop later in the clinical course. While oral antibiotics are the preferred route of treatment among OUDs (POET study),[37]  they may not provide adequate coverage due to limitations in bioavailability. Indeed, this may be subject to change as more data emerge. Ongoing investigations are needed to build upon our knowledge of the natural history, epidemiology, and pathophysiology, which will help us select targeted therapies to better treat patients and improve clinical outcomes.