The coronavirus disease 2019 (COVID-19) pandemic raised unprecedented concerns in the hematopoietic stem cell transplant community. The diagnosis of COVID-19 in these transplant recipients may require extensive laboratory testing and high clinical suspicion, as atypical clinical manifestations or other respiratory viral infections are common in this patient population. The underlying malignancies, immunosuppressed state, frequently observed coinfections, and advanced age in some patients may also predispose them to worse clinical outcomes. Similar outcomes have been previously described with other human coronaviruses, including the severe acute respiratory syndrome coronavirus and the Middle East respiratory syndrome coronavirus. Many hematopoietic stem cell transplant organizations have issued elaborative guidelines that aim to prevent transmission and hence adverse patient outcomes. All potential donors are thoroughly screened, and donated products are cryopreserved in advance. Potential hematopoietic stem cell transplant recipients are also screened, and most nonurgent transplant cases with low risk of progression and/or death are deferred. Current hematopoietic stem cell transplant recipients should adhere to precaution and isolation measures, while their transplant units should also follow strict safety protocols, similar to other infectious outbreaks. The prolonged susceptibility of hematopoietic stem cell transplant recipients to respiratory viral infections might necessitate extending these measures even after the peak of the outbreak until a gradually return to normality is possible.
Key words : Bone marrow transplantation, Coronavirus, MERS, SARS, Super-spreading events
In December 2019, multiple cases of pneumonia of unknown origin were recorded in the city of Wuhan, China.1 A novel coronavirus, later named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was soon identified to be the causative agent of the so-called coronavirus disease 2019 (COVID-19).1,2 After a rapid worldwide spread, the viral outbreak was officially declared a pandemic by March 11, 2020.3 COVID-19 has had a significant impact on all health care aspects to date. Predicting the consequences of the unfolding pandemic on certain patient populations can be challenging, particularly for hematopoietic stem cell transplant (HSCT) recipients.
Clinical Manifestations, Disease Course, and Prognosis of COVID-19 in Hematopoietic Stem Cell Transplant Recipients
The severity of COVID-19 ranges from asymptomatic to severe disease and death and usually manifests as a combination of respiratory and constitutional symptoms.4-8 Bilateral patchy shadowing and ground-glass appearance are the most common imaging findings on chest radiography and computed tomography scans, respectively, whereas lymphocytopenia and abnormal coagulation parameters are common laboratory features.5,7,9 Increasing age and underlying comorbidities seem to be associated with worse clinical outcomes, including higher rates of severe disease and intensive care unit (ICU) admission, faster progression to death, and higher case fatality rate.4,7,10
Respiratory viral infections (RVIs) are common in HSCT recipients and are associated with increased morbidity and mortality compared with outcomes in the general population, likely due to the higher rates of lower respiratory tract (LRT) involvement.11 The significant overlap between the nonspecific clinical manifestations and imaging findings of COVID-19 and other RVIs with LRT involvement could complicate the differential diagnosis in HSCT recipients. In addition, these infections often present atypically in immunocompromised patients, including HSCT recipients.12 The same might be true for COVID-19, as atypical and delayed presentations have already been described in solid-organ transplant recipients.13 In the first case of COVID-19 in an allogeneic transplant recipient described in the literature, symptoms developed 22 days after exposure, far exceeding the usual incubation period of the virus.14,15 As a result, the diagnosis of COVID-19 in HSCT recipients could be particularly challenging and may require high clinical suspicion.
Laboratory testing for RVIs is usually limited to symptomatic patients to maintain a balance between cost and benefit.16 For this reason, studies reporting RVIs in HSCT recipients are thought to underestimate their true incidence in this population.17 Delayed or missed COVID-19 identification due to asymptomatic, atypical, or delayed presentation could set the stage for super-spreading events, similar to other viral outbreaks.18,19 The high clinical suspicion that might be required for COVID-19 diagnosis, combined with the potentially catastrophic outcomes of missed cases in the community, might necessitate extensive molecular testing of all HSCT recipients, which could put a strain on already overwhelmed health care facilities. Indeed, the American Society of Transplantation and Cellular Therapy (ASTCT) suggests that a wide viral polymerase chain reaction (PCR) diagnostic panel should be used, whereas the European Society of Blood Marrow Transplantation (EBMT) stresses the need for repeated testing in cases of negative SARS-CoV-2 to exclude false negative results.20,21
Most hematologic malignancies, many of which require allogeneic or autologous HSCT as a definitive treatment, tend to show an increasing incidence with advancing age. The median age of adult patients with hematologic malignancies is approximately 60 to 65 years.22 Furthermore, increased utilization of allogeneic HSCT for the treatment of elderly patients has been observed over the past 2 decades, due to the widespread use of reduced intensity and toxicity conditioning regimens and the improvement of supportive therapies.23 In a cohort of 1590 Chinese patients diagnosed with COVID-19, malignancy was found to be a significant risk factor for ICU admission, mechanical ventilation, and death.24 Similarly, recent reports from COVID-19 patients with hematologic malignancies, for which HSCT was indicated, also showed worse outcomes compared with those shown in the general population.25,26 Therefore, we should be expecting worse clinical outcomes for COVID-19 patients undergoing HSCT due to their advanced age, comorbid malignancies, and transplant-related immunosuppression. Preliminary data, although limited, seem to support this hypothesis. In an early case series from a stem cell transplant center in the United Kingdom, 4 of 7 HSCT recipients with COVID-19 died as a result of the infection.27 In another report from multiple pediatric hematology/oncology centers in France, 2 of 5 COVID-19 patients in the ICU were allogeneic HSCT recipients.28
Coinfections and superinfections by multiple respiratory viruses have also been described in HSCT recipients and are associated with an increased mortality risk.17,29 Lymphocytopenia has been repeatedly associated with a higher risk of LRT involvement and worse clinical outcomes and, as previously mentioned, is also a common laboratory finding in COVID-19.5,17,29-32 Thus, we can hypothesize that COVID-19 may occur together with other respiratory viral coinfections and superinfections in HSCT recipients, indirectly leading to worse clinical outcomes by promoting the development of severe lymphopenia on the frail HSCT-associated hematopoietic background, especially during the first months after transplant. Although a standard effective treatment for COVID-19 is yet to be established, the combination of hydroxychloroquine and azithromycin has recently demonstrated promising results, and many hospitals have already implemented this regimen to routine COVID-19 patient care.33 However, cumulative antibiotic exposure has been associated with higher rates of RVIs with LRT involvement, and thus worse clinical outcomes, possibly by altering the natural upper airway flora.34 As a result, the current COVID-19 treatment practices along with the standard antibiotic prophylaxis post-HSCT may be a double-edged sword in case a superinfection occurs. In addition, the ASTCT also stresses the risk of severe drug interactions between COVID-19 treatment regimens and immunosuppressants often used in the setting of allogeneic HSCT, such as when graft-versus-host disease develops.35
Posttransplant allogeneic immunoreactivity can promote mechanisms of graft-versus-host disease that may affect the lungs. Bronchiolitis obliterans organizing pneumonia is considered a well-known HSCT complication associated with increased mortality.36 Certain RVIs, particularly those that involve the LRT, have been identified as predisposing factors.37 It has been hypothesized that the prolonged subclinical presence of the virus in immunocompromised patients may trigger or sustain a continuous inflammatory process in the bronchioles.37 Currently, there are inadequate data about the long-term outcomes in patients who have had COVID-19, but we cannot exclude the possibility of a similar process in the respiratory tract of allogeneic HSCT recipients diagnosed with COVID-19.
Previous Experience With Related Viruses
Severe acute respiratory syndrome coronavirus 1 The severe acute respiratory syndrome coronavirus 1 (SARS-CoV-1) was first identified following an outbreak of viral pneumonia in Guangdong Province, China in November 2002 and quickly spread to multiple countries. The virus shares a common phylogenetic origin with SARS-CoV-2 and is primarily transmitted through respiratory droplets. The disease associated with SARS-CoV-1, named severe acute respiratory syndrome (SARS), shares remarkably similar clinical, imaging, and laboratory features with COVID-19.38 Clinical data on the course of SARS in HSCT recipients are limited. Lam and associates39 reported the case of an acute myeloid leukemia patient who underwent allogeneic HSCT and presented with indolent SARS. Fever, chills, and malaise were the only manifestations, and respiratory symptoms were absent. Radiographic findings were only observed after the ninth day after admission.39 During a SARS-CoV-1 outbreak in Toronto, Canada, an immunocompromised liver transplant recipient was identified as a super-spreader and thus the main source of transmission to health care workers.18 As previously stated, atypical presentations and the risk for super-spreading should also be expected for SARS-CoV-2 infections in immunocompromised populations, such as HSCT recipients.
Middle East respiratory syndrome coronavirus
The Middle East respiratory syndrome (MERS) coronavirus is a highly lethal respiratory virus with a genomic structure similar to that of SARS-CoV-1. It was first identified in Jeddah, Saudi Arabia, in 2012.40 Transmission is sporadic, often health care-associated, and is mediated through respiratory droplets. As with SARS, MERS has common clinical, imaging, and laboratory features with COVID-19.40 In a study during the 2015 Korean MERS outbreak, autologous HSCT was associated with increased host susceptibility, more likely due to the underlying immunosuppression.41 Kim and associates42 described 2 cases in HSCT recipients during the same outbreak. A 49-year-old female patient, who had received autologous HSCT 2 months prior to the infection, had a delayed diagnosis of MERS, following multiple negative tests, due to a prolonged incubation period of 20 days. Another patient, a 34-year-old male who had received autologous HSCT 6 months prior to the infection, presented with prolonged viral shedding and high viral loads.42 These findings suggest that HSCT recipients could act as super-spreaders during a MERS outbreak. Coinfection with other viral, bacterial, or fungal agents can complicate MERS infections.43 As previously mentioned, coinfections and super-infections are associated with worse outcomes in other RVIs. The presence of coinfections in MERS strengthens our assumption that they may also occur in COVID-19, which could further increase its morbidity and mortality in HSCT recipients.
Other human coronaviruses
Human coronaviruses (HCoVs) are members of the Coronaviridae family and include 4 strains: HCoV 229E, OC43, NL63, and HKU1.44 The HCoVs circulate throughout the year with a slight predominance during the winter, causing 10% to 30% of the cases of the common cold. Upper respiratory tract infection is most commonly observed, whereas cases of LRT infection have been reported in the very young (age <1 year) and/or immunocompromised patients.45 Among allogeneic HSCT recipients, HCoVs are one of the common community-acquired RVIs, with a reported incidence of 11% to 17%.46,47 The common taxonomy and clinical manifestations shared between HCoVs and SARS-CoV-2 may have significant implications for the prognosis of COVID-19 in HSCT recipients. Human coronaviruses have been associated with high rates of LRT progression, oxygen requirement, ICU admission, and fatal events in HSCT recipients.47-49
In a prospective study of hospitalized adult patients, approximately 50% of the HCoV-infected patients requiring bronchoalveolar lavage (BAL) for an acute respiratory event were transplant recipients. Among them, only 62% had a new infiltrate on chest radiography.50 Similar to other HCoV infections, COVID-19 might lack imaging findings in a significant number of HSCT recipients and might require disproportionately frequent BAL sampling to confirm the infection. However, the latter might lead to widespread transmission of COVID-19 to health care workers, as aerosol remains a potential route of SARS-CoV-2 transmission and BAL is considered an aerosol-generating procedure.51
Prolonged shedding is a common characteristic of HCoV in HSCT recipients. In a report of 44 HSCT recipients with respiratory HCoV infection, the median duration of shedding was 14 days (range, 4-60 days), with 17 patients showing expression of prolonged shedding (≥ 21 days) of the viruses. The median shedding duration of HCoV in nasal samples did not differ between strains, but prolonged shedding was rather associated with high initial viral loads, steroid dosing of > 1 mg/kg/day, and myeloablative conditioning.52 Another report also identified prolonged shedding of HCoVs in 3 of 22 recipients.46 Therefore, it is likely that HSCT recipients may also shed SARS-CoV-2 for prolonged periods of time if they become infected. The possibility of atypical or delayed presentations, combined with prolonged viral shedding, could set the stage for super-spreading events, as previously described. Emerging data from the RECOVERY trial, the largest randomized controlled trial on the treatment of COVID-19 to date, suggested that dexamethasone reduced mortality by 35% in patients requiring mechanical ventilation.53 It is not unlikely that dexamethasone might soon be implemented into the standard treatment regimens for COVID-19. The association of prolonged viral HCoV shedding with corticosteroid treatment, especially with high dosages, creates similar implications for SARS-CoV-2 and further complicates the management of transplant recipients with COVID-19 with regard to immunosuppression.54 A report by Greninger and associates55 presented the emergence of a mutated HCoV strain as a result of a myeloablative regimen used in allogeneic HSCT. The hypothesized mechanisms behind this process include the toxicity of the chemoradiotherapy itself and the impaired immune surveillance following myeloablation.55 Recent data suggested a high mutation potential in SARS-CoV-2.56 The unknown effects of emerging mutated SARS-CoV-2 strains raise significant concerns about the risks of HSCT during this pandemic.
Other respiratory viruses
Aside from coronaviruses, a number of other RVIs affect HSCT recipients. Among them, the most common causative agents are the rhinovirus, the respiratory syncytial virus (RSV), the parainfluenza virus, the influenza virus, and the human metapneumonovirus.57 Despite their diverse phylogenetic origin and virologic properties, all these viruses share similar characteristics with regard to their clinical manifestations and outcomes in HSCT recipients, thus providing a framework for the expected effects of SARS-CoV-2 on this population. As previously stated, multiple concerns arise due to the nonspecific and atypical presentations, the frequent need for BAL sampling, and the high rates of LRT progression and mortality. In addition, asymptomatic and prolonged shedding has been consistently documented for many respiratory viruses and could contribute to the rapid spread of an infectious agent during a pandemic setting, with direct implications for the survival of immunocompromised patients, such as HSCT recipients.32,46,57
The risk of infection and the adverse outcomes vary according to multiple patient-related factors. Underlying comorbidities, lymphopenia, and myeloablative conditioning regimens directly impair the effectiveness of the immune system to prevent and control viral infections, predisposing patients to LRT infections and worse outcomes as a result.58 Graft-versus-host disease indirectly increases the risk for RVIs, due to the immunosuppressive drugs used for its treatment.57 Importantly, the risk for infection is dynamic through time; viral carriage in the community displays a seasonal fluctuation distinct for each virus, while the immune reconstitution period after HSCT is split into multiple phases, each showing a unique pattern of different RVIs.57 Viral exposure mainly occurs in the community, but it can also take place in the nosocomial setting, often in the form of outbreaks.57,58 Taking all these factors into account, it becomes clear that the incidence, outcomes, and potential for community spread of RVIs may widely vary among HSCT recipients. As a result, decision-making with regard to HSCT during the SARS-CoV-2 pandemic may have to become individualized.
An important distinction between SARS-CoV-2 and other respiratory viruses is that there are no effective preventive measures available for the former, apart from typical infection control practices. In contrast, influenza can be primarily prevented by vaccination or chemoprophylaxis with oseltamivir, RSV in children can be prevented with palivizumab, and letermovir has recently been approved for the prevention of cytomegalovirus infection in select HSCT recipients.58,59 So far, no effective treatment has been established for COVID-19, but there are many effective antivirals for the treatment of many RVIs. Examples of these include oseltamivir, ribavirin, and ganciclovir for the treatment of influenza, RSV, and cytomegalovirus, respectively.58 With the lack of effective prevention and treatment measures for SARS-CoV-2 infection, mandates are needed for health practitioners that would outline the risks and benefits of HSCT when they assess potential recipients.
Response of the Hematopoietic Stem Cell Transplant Community to COVID-19
As the threat of COVID-19 quickly evolved into a pandemic, the HSCT community mounted a rapid response to ensure the safety of both patients and the community. The extensive and rigorous measures, which various HSCT organizations have proposed as a way to minimize transmission, heavily reflect the severity and the potential consequences of this outbreak. Essentially every part of the stem cell transplant process has been affected, including the donor and patient selection, the management of current recipients, and the organization of health care facilities.
There is currently limited evidence surrounding the ability of the virus to spread through the parenteral route. In a small-scale case series, the viral RNA of SARS-CoV-2 was detected in the plasma of 15% of patients (6 of 41).6 This has also been found to be true for SARS-CoV-1 and MERS-CoV.60 However, there is no definitive evidence of donor-recipient transmission through blood products. Despite this uncertainty, current recommendations about potential donors are guided by the presumption that donor-recipient transmission is possible and additionally consider the risk of transmission to health care workers involved in the donation process. The World Donor Marrow Association and the EBMT have established strict guidelines for the selection of donors, suggesting that collection should be deferred for 4 weeks in case of contact with a confirmed case or 3 months if the donor tests positive for SARS-CoV-2. In the weeks preceding the donation, donors should ideally be asymptomatic, avoid travelling, and practice isolation and good personal hygiene.21,61 These recommendations are also endorsed by the ASTCT.20 Due to the massive scale of the pandemic, it is expected that these restrictions will exclude many donors from the donation process; in addition, when combined with the imposed travel and transportation restrictions, significant shortages of stem cell availability are expected. The Cord Blood Bank of the United Kingdom has temporarily halted all cord blood donations due to transmission concerns.62 Similar actions toward this direction may further exacerbate the problem. Meanwhile, the “Anthony Nolan” Charity and the National Marrow Donor Program urge for all products to be cryopreserved, as an effort to safeguard their availability for the upcoming months.63,64 This strategy has raised concerns for cross-contamination from potentially infected samples, and thus extensive screening of all stem cell transplant products may be necessary to ensure safety.65 The outcomes of cryopreserved grafts remain controversial. Recent evidence has shown that cryopreserved grafts do not seem to exert any negative impact on patients with hematologic malignancies in terms of hematopoietic reconstitution, acute graft-versus-host disease risk, and overall survival,66 although another study on aplastic anemia patients showed a higher rate of graft failure and lower 1-year survival rate.67
The unknown consequences of COVID-19 in HSCT recipients have also resulted in severe restrictions on the selection of potential recipients. The EBMT proposed that negative SARS-CoV-2 testing should precede any conditioning regimen, regardless of clinical symptoms. If contact with a COVID-19 patient is confirmed prior to HSCT, then the procedure should be deferred for 2 to 3 weeks; if the candidate tests positive for SARS-CoV-2, this period should be extended up to 3 months, depending on the risk of disease progression.21 The British Society of Blood Marrow and Transplantation and Cellular Therapy (BSBMCT) has categorized HSCT candidates according to their risk-to-benefit ratio and proposed that all nonurgent autologous HSCTs, as well as allogeneic HSCTs for nonmalignant or low-risk chronic conditions, should be deferred regardless of any precaution measures or negative testing. Instead, only allogeneic HSCT in urgent cases with high risk of progression and/or death may be considered. They also recommended that additional patient characteristics should be considered during decision-making, and specialized tools such as the Hematopoietic Cell Transplantation-Comorbidity Index could be helpful for this purpose.68 The BSBMCT’s suggestions reflect the risk-to-benefit balance that should be considered during decision-making. For example, long-term deferment of nonurgent autologous HSCT is recommended despite the much lower risk of RVIs compared with that shown for allogeneic HSCT.30 This proposal implies that the potential for disease transmission in the community is considered equally significant to the risks for adverse outcomes in these patients and strongly outweighs the risk of progression in many cases, such as low-grade lymphoproliferative conditions. Multiple committees have issued guidelines about the management of various hematologic malignancies eligible for HSCT; all are in line with the previous recommendations, emphasizing that decision-making should be made on an individual basis.69-71
As previously mentioned, patients who have already undergone HSCT are presumed to be at increased risk of worse clinical outcomes and death due to COVID-19. The EBMT has emphasized that these patients should strictly adhere to personal precaution measures, including restricting travel, minimizing exposure to potentially infected individuals, and maintaining proper hand hygiene. Both the EBMT and the ASTCT have called for clinicians to remain vigilant when respiratory symptoms develop in these patients and have advised that extensive molecular testing should be pursued because of the overlapping presentations with other common RVIs. Retesting should also be considered to exclude the possibility of false negative results. In addition, the benefits of increased testing sensitivity with more invasive testing, like BAL sampling, should be carefully weighed against the severe risk of transmission to health care workers. Clinic visits by both patients and visitors should also be minimized, and alternative practices such as telemedicine should substitute any unnecessary patient contact.20,21
Measures in the Nosocomial Setting
Hematopoietic stem cell transplant units are familiar with infectious outbreaks. Nosocomial outbreaks of respiratory viruses have repeatedly occurred long before the COVID-19 pandemic and are responsible for a significant proportion of RVIs in this patient population.57 These outbreaks are associated with high mortality and often develop despite precautionary measures, such as the use of high-efficiency particulate air-filtered rooms.72 Despite these extra precautions, the cornerstone for the prevention of all RVI outbreaks in HSCT units has always been the implementation of strict infection control measures, which include the isolation of patients, use of appropriate personal protective equipment, meticulous hand hygiene, and careful screening of visitors.73 Preventing interactions between symptomatic individuals and patients is equally important, as cross-infection by health care workers is thought to facilitate the rapid spread of viral infections across patients in this setting.72 A pediatric hematology/oncology department in Italy recently published their protocol for screening patients, visitors, and health care workers, giving an example of how these measures can be implemented into daily practice.74 However, dilemmas arise regarding the optimal strategy to minimize cross-infection. Removing symptomatic health care workers (even those with mild upper respiratory symptoms) from patient care could create logistical problems, as the already overwhelmed health care facilities may quickly become understaffed.72 Buchtele and associates75 described a SARS-CoV-2 outbreak among health care workers in a HSCT unit. Through contact tracing combined with broad testing and isolation of infected individuals, they effectively contained the outbreak, without interrupting the unit’s normal operations.75
Recent evidence has suggested that the percentage of asymptomatic carriers (and so the potential for asymptomatic transmission) may be much higher than initially estimated.76 Therefore, it may be prudent to screen every individual directly or indirectly involved in the transplant process (including donors, recipients, and all health care workers), regardless of symptoms or contact history. Both reverse transcriptase PCR and antibody tests are utilized as standard methods for SARS-CoV-2 screening. Both types of tests have distinct limitations in terms of their sensitivity and specificity, and combined testing may be necessary to optimize the accuracy of the results.77 Other factors, such as the type of sample and the time period from infection to testing, may also significantly influence the accuracy of the results.78 Therefore, negative testing should not provide a false sense of reassurance, and appropriate safety measures should be taken at all times. It may also be necessary to screen all individuals at regular intervals to minimize the impact of false negative results.
Given the high transmission potential of SARS-CoV-2 and the presumed prolonged shedding of the virus by HSCT recipients as previously mentioned, an active HSCT unit during the COVID-19 outbreak could act as the epicenter for super-spreading events, which could severely endanger both the community and the patients. The lack of reliable data on certain characteristics of the virus and its potential effects on this special patient population have complicated the decision-making process. There remains the likelihood that some prevention measures may need to be escalated, including a universal temporary discontinuation of all nonurgent HSCTs. In contrast to other infectious complications linked to HSCT, RVIs can occur at any time during the immune reconstitution process, and the risk may actually increase when patients are discharged. The risk remains high even at 100 or more days post-HSCT.73 In addition, with no reliable markers suggestive of effective immune reconstitution following HSCT currently available, recommendations on a per patient basis are limited.73 It may therefore be necessary for these extreme measures to be extended long after the peak of the outbreak or until significant immunity develops in the general population, as would occur if an effective vaccine is developed in the following months. As more data emerge, restrictions may be appropriately adjusted, while also keeping in mind that a balance should be maintained between the patients’ treatment benefit and the risks for community transmission and adverse patient outcomes.
COVID-19 in HSCT recipients requires high clinical suspicion and could present with additional challenges compared with those in the general population, including atypical presentations, worse outcomes, and high risk for super-spreading events. Data from other RVIs in these patients support these assumptions. During the pandemic, HSCT organizations promptly reacted and proposed widespread limitations that were aimed at preventing adverse patient outcomes and minimizing disease transmission in the community. These proposals affect donors, potential recipients, current recipients, and HSCT units and may create ethical dilemmas, shortages in donation products, and understaffing of HSCT units. Extending these measures long after the peak of the outbreak may be required to ensure the safety of HSCT recipients.
DOI : 10.6002/ect.2020.0326
From the 1Surgery Working Group, Society of Junior Doctors, Athens, Greece; the 2Institute of Health Innovations and Outcomes Research, The Feinstein Institutes for Medical Research, Manhasset, New York, USA; and the 3Department of Internal Medicine ΙΙΙ, Hematology, Oncology, Palliative Medicine, Rheumatology and Infectious Diseases, University Hospital Ulm, Ulm, Germany
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Author contributions: SME and DG provided conceptualization, investigation, and writing (original draft, review, and edits); IAZ provided conceptualization, methodology, writing (review and edits), and supervision; PG provided conceptualization, investigation, writing (original draft), and visualization; and ES and HD provided methodology, writing (review and edits), supervision, and project administration.
Corresponding author: Dimitrios Giannis, Institute of Health Innovations and Outcomes Research, Feinstein Institutes for Medical Research, 600 Community Drive - 4th Floor, Manhasset, NY 11030, USA
Phone: +516 225 6397