COVID-19 therapeutics: stewardship in England and considerations for antimicrobial resistance

Abstract

The COVID-19 pandemic saw unprecedented resources and funds driven into research for the development, and subsequent rapid distribution, of vaccines, diagnostics and directly acting antivirals (DAAs). DAAs have undeniably prevented progression and life-threatening conditions in patients with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. However, there are concerns of antimicrobial resistance (AMR), antiviral resistance specifically, for DAAs. To preserve activity of DAAs for COVID-19 therapy, as well as detect possible mutations conferring resistance, antimicrobial stewardship and surveillance were rapidly implemented in England. This paper expands on the ubiquitous ongoing public health activities carried out in England, including epidemiologic, virologic and genomic surveillance, to support the stewardship of DAAs and assess the deployment, safety, effectiveness and resistance potential of these novel and repurposed therapeutics.

A public health emergency of international concern was declared in March 2020 by the World Health Organization following an increasing number of patients infected with what had recently been identified as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), causing the coronavirus disease 2019 (COVID-19). In the absence of therapeutic approaches, worldwide collaborative efforts resulted in the fastest vaccine development and largest vaccination campaign in global history. In parallel, viable existing pharmaceutical agents that could be safely and effectively repurposed (dexamethasone, remdesivir) were identified and new DAAs (agents that directly target parts of the virus replication cycle and include complex biologicals such as neutralizing monoclonal antibodies) were identified. These agents were rapidly investigated in clinical trials and have now been deployed, widely in some countries, for patients who present with severe to critical illness or for clinical subgroups who are at higher risk of progression into severe illness and hospitalization (e.g. in the UK).^1–4 The first UK-approved DAA used for the treatment of COVID-19 was remdesivir in May 2020, with other DAAs having since been made available (Table 1).

Table 1.

Licensed DAAs used in the UK for SARS-CoV-2 (as of May 2022)

DAA	Clinical use ina	Available in the UK since
Remdesivir	Patients hospitalized due to COVID-19. Recently, available to patients with hospital-onset COVID-19 infection and to non-hospitalized individuals at highest risk with COVID-19 infection.	May 2020
Casirivimab with imdevimab	Briefly available for non-hospitalized patients. Remains in limited use for individuals admitted due to COVID-19 with confirmed Delta infection.	September 2021 (removed from UK clinical commissioning policy in Feb 2022)
Sotrovimab	Non-hospitalized at highest risk and the hospital-onset COVID-19 cohorts. Is being studied in admitted patients within the RECOVERY trial.	Mid-December 2021
Molnupiravir	Non-hospitalized patients at highest risk; and in those with hospital-onset COVID-19. Also being studied for its effectiveness as a community treatment through the PANORAMIC trial.	Mid-December 2021; February 2022
Nirmatrelvir plus ritonavir	Both non-hospitalized patients at highest risk and those with hospital-onset COVID-19 infection.	February 2022

^aClinical access policies for the use of DAAs for the UK are published via therapeutic central alert system (CAS alerts).⁵

The 2014 O’Neill review estimated that AMR could cause 10 million deaths a year by 2050.¹ AMR has been recognized as a major global public health threat and pre-pandemic had been a focus at the United Nations General Assembly and placed on the health agenda at the G7 and G20 summits.² However, to mitigate the compounding effects of COVID-19, AMR efforts were deprioritized in many settings.² The COVID-19 pandemic demonstrated the vast challenges to health care systems in responding to widespread infections where effective treatment are unavailable. Hence, it is important to maintain the efficacy of existing antimicrobials, and in the current context of COVID-19, prolonging the efficacy of the new, or repurposed, therapies by reducing the risk of emerging resistance. Increased use of antimicrobials, specifically anitvirals, results in prolonged virus drug exposure and subsequent selection pressure, inappropriate use (inadequate dosing, duration or adherence), monotherapy (as only one key mutation is required for the emergence of a resistant variant, versus combination therapy where simultaneous multiple mutations in different regions of the viral genome are needed), and use in prolonged infections (e.g. immunocompromised populations, where prolonged exposure to antivirals is required) are all thought to be drivers of subsequent DAA resistance.^3,4 SARS-CoV-2 is highly likely to develop resistant mutations when exposed to DAA in treated patients and where this occurs, these mutations may be transmitted to others and even have the potential to become the dominant variant with the right ecological conditions.⁶ Given the paucity of available research on antiviral resistance of SARS-CoV-2, evidence from similar antivirals and viruses proves useful in understanding potential risk factors for resistance and developing mitigation measures. Past experience from treatment of HIV, HBV and HCV, have demonstrated the emergence of resistance to first-generation nucleoside analogues and protease inhibitor monotherapies that subsequently necessitated combination therapy; while this risk is lower in SARS-CoV-2 due to short duration treatments it is nonetheless important to assess.^7–11 Influenza antivirals have shown mixed evidence, with sustained efficacy of oseltamivir (a monotherapy) whereas amantadine antiviral drugs have shown widespread resistance, and in vitro resistance to favipiravir has been noted.^3,12

Furthermore, emerging novel variants of SARS-CoV-2 raise concerns for the clinical efficacy of licensed therapeutics. This has been seen in sotrovimab, with several in vitro studies indicating reduced efficacy with BA.2 sub-lineage of Omicron compared to BA.1.^4,13,14 Similarly, while casirivimab with imdevimab were initially deployed, they had subsequently been found to have decreased efficacy against the BA.1 sub-lineage of the Omicron variant and were removed from clinical use worldwide (Table 1).^15,16

To retain activity of DAAs for COVID-19 treatment and detect or mitigate possible mutations conferring resistance, stewardship of these therapeutics and surveillance or the use of ‘information for action’ is essential. The UK Health Security Agency (UKHSA) developed a programme, comprising six workstreams, of public health activities to support the deployment and stewardship of these directly acting COVID-19 therapeutic agents.¹⁷ Proactive epidemiologic, virologic and genomic surveillance within a stewardship framework was established from the onset of roll-out in England. The commentary that follows describes ongoing AMR and antimicrobial stewardship (AMS) strategies for COVID-19 therapeutics in England.

Antiviral resistance and surveillance in England of novel COVID-19 therapeutics

Unprecedented genomic surveillance of SARS-CoV-2 in England has enabled a genome-first approach to the detection of emerging drug resistance. As well as sequencing viruses from a random sample of SARS-Cov-2 positives in the general population, the UK introduced a protocol to enhance sequence coverage of those receiving treatment (including pre-treatment and follow-up sampling) in both hospital and community settings.¹⁸ This enabled real-time analyses comparing pre- and post-treatment sequences to rapidly identify treatment-emergent substitutions. Screening for acquired mutations had also been completed using the genomics SARS-COV-2 dataset for mutations that are not present in most samples from a given lineage. These mutations were identified by the attainment of at least two of the following thresholds: abundance ≥50 samples per week; an increase ≥5% of the number of samples with a given mutation, week-on-week; represent a proportion ≥0.01 samples in a given week; or week-on-week the proportional representation of samples with the mutation changes by ≥0.0025. Mutations in genomic regions associated with therapeutic treatments were then provided to the Structural Biology Division (University of Oxford) for review.

The availability of structural data for complexes of antiviral therapeutic agents bound to their SARS-CoV-2 protein targets allowed for the design of a highly sensitive and dynamic surveillance strategy for antiviral resistance. Based on structural modelling of a large set of protein–drug interactions, an associated set of SARS-CoV-2 nucleotide positions encoding complex-stabilizing amino acids were identified. In-depth biophysical understanding of these interactions enabled rapid risk assessment of emerging mutations and new variants, which had been followed by experimental validation of resistance-conferring mutations and updates to genomic surveillance processes.

In parallel with the genomic identification of variants and mutations of potential concern for DAAs, efficacy of licensed and experimental COVID-19 therapeutics for these different variants of concern had, and continues, to be assessed. SARS-CoV-2 therapeutics were tested in cell-based assays in vitro and their ability to either neutralize or inhibit infection/replication of the virus was measured.¹⁹ The assay measures the number of plaques or foci formed when the virus infects the cell and assesses to what degree different treatments were able to reduce the number of infection events (Figure 1 provides details of the in vitro process). The inhibitory effect is quantified by determining the concentration of a compound that causes a 50% reduction in viral plaques or viral foci compared to the untreated virus control: this is otherwise known as the IC₅₀. Comparing these values allowed us to understand whether the compounds were able to inhibit virus infection and replication and at what concentration in vitro. Mutations in the antiviral or antibody target, such as those seen in the spike protein in different variants of concern, can then be assessed to see how these affect therapeutic efficacy. This experimental assessment of resistance using pseudovirus assays forms part of a wider consideration in the UK, including dosing, pharmacodynamics, pharmacokinetics and clinical trial data, that informs choices of therapies for different patient groups. The information, about changes in the IC₅₀ for different licensed therapeutics, alongside the structural modelling data, supports surveillance looking at the presence of variants in different clinical populations, with and without treatment.

Figure 1.

Overview of in vitro method (using VeroE6 cells). This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Antimicrobial stewardship in England of novel COVID-19 therapeutics

While in vitro methods and animal studies can provide assessment of the likely efficacy of therapeutics against emergent variants, confirming such findings with observational data provides added value. Using data linkage methods, surveillance of the clinical use and dispensing of COVID-19 therapeutic agents had been implemented across England, using data from Blueteq and Rx-info systems. The Blueteq system manages high-cost drugs for NHS England and contains dedicated clinical questions on rationale and indication for prescriptions of DAAs used for the treatment of COVID-19 in high-risk patients in England. These data are linked to demographic, SARS-CoV-2 vaccination, serological, hospital stay, mortality and viral genomic data to provide patient-level epidemiological data on the use of these novel therapeutic agents and the patients treated in England. Rx-info data on dispensed drugs had been used to validate Blueteq patient-level usage data and provided regional and an England wide view of coverage and use to improve and guide stewardship interventions. Combined, the enriched data provided key indicators for the assessment of effectiveness of novel therapeutics, such as the risk of hospitalization and the risk of death, informed on regional access to treatments and the success of targeted deployment. These surveillance and epidemiological assessments were routinely completed (on a weekly basis in England) to inform AMS on differences between treatment requests of different therapeutic agents by variant, age, gender, ethnicity, comorbidities, antibody or vaccination status. It should be noted that conducting these studies with routinely available data has been challenging. Ideally, assessments would involve those treated with novel therapeutics compared to a group who were not treated, to inform efficacy of use and equity of access. Eligibility for therapeutics is determined by patients’ underlying clinical characteristics, and may also change based on medication taken or recent treatments undergone. Since the provision of novel therapeutics is targeted to those at high risk of developing severe disease, identifying a group who are at great risk of severe disease and test positive for SARS-CoV-2 but do not take up therapeutic options is challenging. Furthermore, as new variants have arisen, the policy and healthcare burden context into which they have emerged has changed, introducing risk of time-varying confounders; ideal comparisons would involve infections that occur in the same period of time, limiting comparisons to between variants that are co-circulating.

The surveillance of respiratory viral, as well as bacterial and fungal lower respiratory tract and blood stream infections as preceding (−27 to −2 days before SARS-CoV-2 positive sample), coinfection (−1 to +1 days) and secondary infections (+2 to +27 days) in relation to SARS-CoV-2 patient-episodes has been underway since the first wave of the COVID-19 pandemic in England.²⁰ Findings from this surveillance are supported by recent literature and data relating to larger studies which indicate rates of bacterial COVID-19 co-infections are relatively low.^21–23 Research is underway to determine the incidence of these infections amongst persons with COVID-19 who have received DAA therapy.

Structured monitoring of the literature has been an integral part of the surveillance activities in England. Published work since January 2020 have been catalogued using keywords to aid interactive searching, and summarized to provide internal evidence in support of UKHSA’s research activities. In recognizing the importance of keeping abreast of the evidence, the bi-monthly searches include epidemiological studies, economic evaluations, case studies and summaries, in vitro studies, and future incoming data in the form of trial registrations. The literature, alongside epidemiological surveillance, inform on effectiveness studies, whether emerging resistance is affected by certain variables (e.g. age, comorbidities, ethnicity, antibody or SARS-CoV-2 vaccination status, ethnic status) and will inform on regional AMS and on any disparities seen.

AMS programmes ensure there are systems in place to improve medical outcomes by providing patients with optimal antimicrobial therapy, contingent to the five ‘rights’ of medication administration: right patient, right drug, right time, right dose and right duration.²⁴ Pre-pandemic, AMS structure and strategies were in place in numerous hospitals worldwide, they were therefore well suited to be rapidly deployed to assist with the COVID-19 pandemic, and AMS personnel have been heavily involved in the response.²⁵ However, the pandemic has required a rapid readjustment of the role and tools of AMS. In many circumstances, this has translated to a temporary diversion of effort and resources from some of the core and traditional AMS activities to focus on pressing and time-sensitive COVID-19 activities.² The negative impact of the COVID-19 pandemic on traditional AMS programmes has been documented with many AMS leads diverted, initially to other speciality areas, such as intensive care units, and then to deployment of COVID-19 vaccines.^26–28 With approved COVID-19 therapeutics, the AMS teams were then charged with ensuring optimal therapy and monitoring of usage. COVID-19 therapeutic stewardship contributions were numerous and included: actions to reduce the use of ineffective therapies against SARS-CoV-2, such as azithromycin and hydroxychloroquine, following increased use seen during the pandemic²⁴; deployment and monitoring (including formulary restriction and/or pre-authorization and implementation of COVID-19 therapeutics)^24,29,30; curation of COVID-19 focused treatment guidelines, comments on existing literature and delineating clinical trial options.³¹ Supporting the stewardship of the COVID-19 therapeutics, a Task and Finish group led by UKHSA, in collaboration with NHS England, was initiated in November 2021. The appropriate use of COVID-19 therapeutics is supported by a workstream focused on epidemiological analyses and surveillance (mentioned before), presenting and publishing clinical, epidemiological and prescription data to inform stewardship.¹⁷ The Task and Finish group also monitored prescription difficulties and challenges within stewardship practices and provided relevant education and training for infection teams leading on COVID-19 therapeutics. The National Institute of Health and Care Research (NIHR) commissioned three groups united under the Rapid Outcomes of Therapeutics for COVid (ROTH-COV): OpenSAFELY Collaborative, EAVE II study and QResearch. These groups met regularly, convened by the UKHSA COVID-19 Therapeutics programme, and were assessing effectiveness, outcomes and equity of access for patients treated with DAAs in the community setting.^32–35

Theoretically, monotherapy compared to combination therapy leads to a higher risk of emergence of resistance requiring higher treatment doses for the same therapeutic effect but with greater risk of adverse events.³ Adverse effects may curtail treatment, which can subsequently result in AMR and treatment failure in patients. Alongside analyses of intended outcomes and effectiveness of the COVID-19 therapeutics, monitoring and review of unintended outcomes such as suspected adverse reactions are occurring in the UK through patient and clinician reports to the Medicines and Healthcare Products Regulatory Agency (MHRA) via a Yellow Card Scheme for drug substances, and cumulative data are published monthly.³⁶ Analyses up to 30 June 2022, for molnupiravir, remdesivir and sotrovimab, suggest that most reported side effects (reactions/events occurring after use of drug, not necessarily causal) were mild (75% of reports, with 239/945 reported reactions being severe) and 17 reports of fatalities.³⁶ The published literature corroborates the low occurrence of serious reactions, with findings suggesting no correlated increased odds or risk of adverse events with use of DAAs.^37–39

Conclusion

COVID-19 vaccines have been a critical tool in the drive to reduce the health impact from the pandemic, however, diagnostics and COVID-19 therapeutics (DAAs and interleukin-6 inhibitors) have also been important interventions. DAAs have undeniably prevented progression and life-threatening conditions in patients with SARS-CoV-2 infection. With concerns related to AMR for DAA, extensive surveillance and AMS programmes were rapidly implemented in England and globally. Through surveillance and AMS channels, where certain DAA clinical effectiveness was likely to be reduced following emergence of variants with specific mutations conferring resistance, recommendations to halt inappropriate use were prompt. The pandemic has brought about opportunities for an even tighter collaboration between frontline clinicians and public health (i.e. exchanging data, building dashboards, knowledge mobilization, fast-paced evidence-based medicine, deployment of therapeutics and vaccines) and presents a bright future direction and networks to better inform on treatments for AMR pathogens and pipelines of alternative therapies. Current collaborations and workstreams have indeed paved a potential blueprint for monitoring all new novel antimicrobial agents, outside of COVID-19, including new antibiotics. Further research will be needed to characterize how AMS resources can be best used in a pandemic setting. Such data can be useful for consideration for the formation or expansion of AMS programmes and capacity building as a part of pandemic preparedness. Synthesizing data, as well as relevant published evidence from the literature, are important to understand shifts in AMR patterns, and to assess the safety, efficacy and resistance potential of the novel therapeutics, and how these are affected by variant, age comorbidities, ethnicity, antibody or vaccination status.

Funding

This supplementary publication is being supported by funding from Roche; Roche have had no influence on the contents of this manuscript. D.I.S. supported by the UKRI MRC (MR/N00065X/1). S.H., C.S.B. and S.G. are supported by the National Institute for Health Research (NIHR) Health Protection Research Unit (HPRU) in Healthcare Associated Infections and Antimicrobial Resistance in a partnership between UKHSA and (i) Imperial College London (NIHR200876) and (ii) the University of Oxford (NIHR200915). This paper was published as part of a supplement financially supported by an educational grant from Roche Molecular Systems.

Conflict of interest

This supplementary publication is being supported by funding from Roche; Roche have had no influence on the contents of this manuscript. D.I.S. consults for AstraZeneca.

Transparency declarations

S.B-A. was the lead writer, collated and interpreted contributions from co-authors. S.R., D.A-O., A.D., C.S.B., M.C. and S.H. led on the conception of the editorial and alongside E.S. reviewed and edited drafts. All co-authors participated in developing the draft either providing details related to areas of expertise, or edits. Figure was provided by J.T. All co-authors reviewed and agreed to the published version.