A Tale of Two Pandemics: Antimicrobial Resistance Patterns of Enterococcus spp. in COVID-19 Era

Although the COVID-19 pandemic has held the spotlight over the past years, the antimicrobial resistance (AMR) phenomenon continues to develop in an alarming manner. The lack of strict antibiotic regulation added to the overuse of antimicrobials fueled the AMR pandemic. This paper aims to analyze and identify the impact of the COVID-19 pandemic on antibiotic resistance patterns of Enterococcus spp. The study was designed as a retrospective observational study. Enterococcus spp. infections data were collected from one academic hospital in Cluj-Napoca, Romania over 18 months. A statistical analysis was performed to compare antibiotic resistance phenotypes identified. We recorded an increase in the isolation rates of Enterococcus spp. strains, from 26 isolates (26.53%) during Period A (November 2020–April 2021) to 42 strains (42.85%) during Period C (November 2021–April 2022). The number of strains with resistance to vancomycin increased from 8 during Period A to 17 during Period C. Of the total 36 strains with resistance to vancomycin, 25 were identified as E. faecium. SARS-CoV-2 patients (n = 29) proved to be at risk to develop an E. faecium co-infection (n = 18). We observed that strains with resistance to ampicillin (n = 20) and vancomycin (n = 15) are more often isolated from these patients. All changes identified in our study are to be considered in the light of COVID-19 pandemic, highlighting the threatening AMR phenomenon in Romania. Further studies should be performed to quantify the worldwide effects of these pandemics.


Introduction
Antimicrobials have proven to be an effective tool in the fight against bacteria and fungi over the years, becoming nearly indispensable for medical practice all-over the world. Antibiotics are an essential tool in treating human infections, especially in developing countries which may lack awareness when it comes to hygienization [1]. Unfortunately, due to the ability of the microorganisms to develop resistance to antimicrobial substances, these therapies became less and less effective in time. This phenomenon of antimicrobial resistance (AMR) was linked with the wide availability and the overuse of antimicrobial substances [2]. Moreover, the lack of proper epidemiological surveillance (especially in healthcare units) makes it impossible for specialists to develop customized protocols for high-efficiency low-risk antibiotic therapies [3]. To prevent the misuse and the further development of this AMR phenomenon, it became of utmost importance that both the commercialization and prescription of antibiotics be highly regulated [4].
Worldwide, a link was identified between the use of antibiotics and the knowledge about antimicrobial resistance. It appears that the lack of education on appropriate usage of antibiotics is strongly connected to the spread of AMR. Unfortunately, surveys on antibiotic usage show not just the level of misinformation, but also the increasing number of people targeted by this phenomenon [5].

Analysis of Enterococcus spp. Infections
A number of 98 patients were confirmed with an Enterococcus spp. infection during the time of our study (1 November 2020-30 April 2022). E. faecalis was the pathogen incriminated in 54 infections, E. faecium was identified in 42 samples, and 2 infections were confirmed with E. gallinarum. Of these patients, 29 were confirmed with a SARS-CoV-2 infection. Distribution of E. faecalis and E. faecium among these patients is presented in Table 1. Among the collected samples, E. faecalis was mostly isolated from wound drainage samples, while E. faecium was mostly reported in blood samples. A summary of the collected samples is presented in Table 2. The detected phenotypes, based on their minimum inhibitory concentrations (MICs), on VITEK-2 antibiotic susceptibility testing (AST) are presented in Table 3. The most susceptible strains were E. faecalis strains, while E. faecium harbored most vanA like (MIC for Vancomycin >64 mg/L, MIC for Teicoplanin >16 mg/L) and vanB like (MIC for Vancomycin >4 mg/L, MIC for Teicoplanin >0.5 mg/L) phenotypes. All E. gallinarum presented vanC like (low level resistance to Vancomycin) phenotype.

Analysis of COVID-19 Co-Infection
Data regarding gender distribution, survivability, and length of stay in the COVID-19 group are presented in Table 4. During this study, 29 patients admitted in our unit with a SARS-CoV-2 pneumonia developed an Enterococcus spp. co-infection. Out of these patients, 13 were females and 16 were males. Only 17 of 29 patients were deceased.   The results of AST for each period of our study are presented in Table 7. Statistically significant differences were reported for streptomycin (STR), linezolid (LNZ), and tetracycline (TET).
Survival rates of patients infected with E. faecalis and E. faecium are presented in Figure 1. No statistically significant difference was identified between the 2 groups.   The number of resistant Enterococcus spp. strains are presented in Figure 2 and Table  8. There is no resistance to Tigecycline (TGC) reported during Period A. Strains of E. faecalis with resistance to TGC are reported during period B and C. An increase in the number of resistant strains was reported for most tested antibiotics when Period A was compared to Period C. The number of resistant Enterococcus spp. strains are presented in Figure 2 and Table 8. There is no resistance to Tigecycline (TGC) reported during Period A. Strains of E. faecalis with resistance to TGC are reported during period B and C. An increase in the number of resistant strains was reported for most tested antibiotics when Period A was compared to Period C.

Analysis of Antibiotic Resistance Patterns
Ampicillin, streptomycin, ciprofloxacin, teicoplanin, and vancomycin resistance was more likely to occur in strains of E. faecium, rather than E. faecalis. The odds ratio for these associations are presented in Table 9.
COVID-19 co-infection occurred more often in patients infected with Enterococcus spp. strains with resistance to ampicillin and vancomycin. The results for all tested antibiotics are presented in Table 10.

Analysis of Antibiotic Resistance Patterns
Ampicillin, streptomycin, ciprofloxacin, teicoplanin, and vancomycin resistance was more likely to occur in strains of E. faecium, rather than E. faecalis. The odds ratio for these associations are presented in Table 9.        Regarding survivability, death was more likely to occur in patients infected with Enterococcus spp. strains with resistance to ciprofloxacin. Results of the Chi-square/ Fisher's test and the Odds ratios associated with them are presented in Table 11.  In infections produced by VRE, Linezolid and Tigecycline are two last resort antibiotics. Table 12 presents the MICs distribution for all the Enterococcus spp. isolates included in our study. Although the COVID-19 pandemic seems to have come to an end, its impact on the HAIs with MDR pathogens is still up for debate. Several studies reported a rise in antibiotic use, but the aftermath of this increase remains unclear [21][22][23]. Previous viral pandemics taught us that a secondary bacterial or fungal infection provides a poor outcome for the patient. Although there have been several mechanisms described that aimed to explain the relationship between viruses and bacteria regarding the infection pathogenesis, the SARS-CoV-2 involvement remains understudied. The existing data show that this virus seems to rely on regulating the gene expressions in monocytes. Furthermore, some studies show a gastrointestinal involvement of SARS-CoV-2. Considering the microbiome diversity at this situs, a possible virus-induced immunosuppression may lead to secondary bacterial infection. Thus, a specific guideline should be implemented for the use of antibacterial and antifungal drugs in SARS-CoV-2 infections [24].
The goal of our study was to describe the changes in antimicrobial resistance patterns in Enterococcus spp. during the COVID-19 pandemic and to analyze the relationship between these two microorganisms.
During the 18 months of our study, there an increase in the number of invasive infections with Enterococcus spp. was noticed. Although there was no statistically significant difference between the three time periods (p > 0.05), we noticed an increased number of infections with Enterococcus spp. (42.85%) in the last 6 months of our study (November 2021 to April 2022). The increasing number of Enterococcus spp. infections was described also by M. Polemis et al. in samples from ICUs in Greece [25]. Following the findings of their study, we noticed in our center that the majority of infections were caused by E. faecium, most of them being bloodstream infections (BSIs).
Regarding gender distribution, E. faecium and E. faecalis infections were evenly distributed among male and female patients. Although, there was no statistically significant difference between gender distribution (p > 0.05), a trend was noticed for E. faecalis infections affecting mainly the male patients.
In patients suffering from SARS-CoV-2 pneumonia, we noticed that an infection with E. faecium was more likely to occur (p < 0.05) when compared to the E. faecalis group. Although, no significant difference in mortality for patients with SARS-CoV-2 co-infection was found, infections with E. faecium had significant lower rates of survivability (p < 0.05) in comparison with E. faecalis and E. gallinarum. These results suggest that the severity of these infections was not the presence of SARS-CoV-2 co-infections, but the Enterococcus spp. strain that infected the patient. However, the number of E. galinarum strains was low and therefore the real impact of this species remains to be established. In the light of these findings, better management of the E. faecium infections could have provided better outcomes for these patients.
The site of infections proved to be of great importance as well. Although isolated from samples as puss (n = 2), cerebrospinal fluid (n = 1), or peritoneal fluid (n = 3), E. faecium was isolated significantly more often (p < 0.05) from blood samples (n = 21). Although our small number of strains did not allow us to perform an in-depth analysis, different authors described BSIs with E. faecium in patients with SARS-CoV-2, especially in ICUs. The alterations that COVID-19 brings to the gut microbiota may be the main culprit of the high prevalence of Enterococcus spp. BSI [20,26,27]. E. faecium seems more likely to cross the intestinal barrier when compared to E. faecalis. Moreover, all central line bacteriemia (n = 4) in our study were caused by E. faecium, findings previously described also by S. Hughes et al. [28]. Furthermore, central venous catheter (CVC) infections were previously associated with the ability of Enterococcus spp. to form biofilms. The biofilms formed by MDR bacterial strains may impact the prognosis of patients with CVC infections [29][30][31][32][33].
On the other hand, E. faecalis presented a more heterogenous distribution, being isolate most frequently from wound samples (n = 25) and urine samples (n = 4). We consider that further studies may be needed in order to better understand how the site of infections can impact the treatment plan for these patients.
When it comes to antibiotic therapies, vancomycin remains one of the used antimicrobials in Enterococcus spp. infections. This phenomenon represents a risk factor for increasing resistance of Enterococcus spp. strains to vancomycin. The vancomycin resistant Enterococci (VRE) were repeatedly reported in literature to burden the healthcare systems all-over the world, causing invasive infections resistant to usual treatment protocols. In our study, we identified that E. faecium strains had significantly higher rates (p < 0.05) to harbor a vanA like phenotype in comparison with E. faecalis. In association with the lower E. faecium survivability rates, we can consider that vanA resistance phenotype represents a factor of bad prognosis in patients with an E. faecium infection [34].
The results of our Kaplan Meier survival analysis, although not statistically significant (p > 0.05), provides useful data regarding the clinical impact of Enterococcus spp. infections. While E. faecalis infections are associated with lower survivability rates in the first 20 to 40 days of hospitalization, E. faecalis infections survivability rates dropped after this time period. Though there is no significant difference in the average of the hospitalization days (p > 0.05), E. faecium patients spent an average of 5.5 more days in our hospital than E. faecalis patients. The longer length of stay and lower survivability rates can be linked with the resistance of these strains to vancomycin and other usual antibiotics, lengthening the time needed to treat these infections. D. R. Giacobbe et al. described the role of enterococcal BSI in the clinical evolution of COVID-19 patients [35]. Their study reported an increased incidence of Enterococcus spp. BSI in critically ill COVID-19 patients and higher mortality rates for these patients.
As previously discussed, unregulated antibiotic use seems to evolve and its impact on the increased number of highly resistant strains of bacteria becomes relevant. In the light of COVID-19 pandemic aftermath analysis, we can assess the involvement of previous antibiotic use in developing resistant strains of bacteria that are able to promote infections. Considering the struggle of the silent AMR pandemic, it does not come as a surprise that Enterococcus spp. emerged as a real threat to healthcare systems all around the world. The association of COVID-19 and infections produced by Enterococcus spp. has been evaluated in several articles that showcase an increase in the total length of hospitalization. Since VRE is constantly under surveillance by healthcare professionals, several protocols have been proposed in order to properly tackle the issue of VRE spread. However, due to the increased pressure on the healthcare system that was imposed by the COVID-19 pandemic, several breaches that led to hospital spread of VRE have been described in literature. To provide a better supervision of these strains, further studies that aim to develop better guidelines are required [36][37][38][39][40][41][42].

Over-Time Analysis of Enterococcus spp. Infections
The overall results of our study provide us with data to interpret the impact of Enterococcus spp. infections during the 18 months of our study, in association with the pressure COVID-19 put on our healthcare system. In order to identify the aftermath of the SARS-CoV-2 pandemic, we needed an over-time analysis of these results. We evenly divided our study into three periods of 6 months each representing the first cold season COVID-19 pandemic (Period A), the warm season COVID-19 pandemic (Period B), and the second cold season and the aftermaths (Period C). Although not statistically significant (p > 0.05), both cold seasons of our study registered more COVID-19 positive patients (n = 8 for Period A and n = 16 for period C) than Period B (n = 5). These findings can be the result of both weather conditions and national restrictions used to lower the number of SARS-CoV-2 cases.
While the number of Enterococcus spp. infection increased and we noticed significantly more VRE E. faecium with lower survivability rates, we continued our analysis in order to identify if these findings represent the results of the management of COVID-19 pandemic. The results of our antibiotic susceptibility tests (ASTs) provided useful data if analyzed over-time. Most of the tested antibiotics reported increased resistance during Period C in comparison with Period A. Strains resistant to gentamicin, teicoplanin and vancomycin were isolated more frequently as the time passed. The increasing resistance to these commonly used antibiotics can be the result of misuse of antimicrobial therapies in patients with SARS-CoV-2 pneumonia. Although the antibiotic therapies were reported to have no beneficial effect in COVID-19 infection, it appears that their increased use selected more resistant bacterial strains incriminated in nosocomial infections.
An unusual resistance pattern was identified for linezolid resistant Enterococcus spp. strains. During our study, the resistance to linezolid reached a peak during Period B and decreased during period C. In the light of these increasing resistance patterns, we shift our attention to linezolid to provide a solution to MDR Enterococcus strains. Resistance to linezolid is defined as an MIC greater than 4 mg/L according to the EUCAST 2023 guideline (available on the EUCAST website: https://www.eucast.org [accessed on 22 January 2023]). In our study, 12 strains are reported as resistant. Due to the retrospective nature of this study, we were unable to perform molecular analysis in order to characterize the resistance mechanism.
Lowest resistance rates were reported for tigecycline (TGC). During Period A, no strains of Enterococcus spp. with resistance to TGC were identified, but resistance to TGC was identified during Period B (n = 2) and Period C (n = 2). Unusual for our study is that all TGC resistant strains were identified as E. faecalis. Although, the number of resistant strains were too low to provide statistical significance (p > 0.05), we consider this trend worth supervising. Resistance to tigecycline is defined as an MIC greater than 0.25 mg/L according to the EUCAST 2023 guideline (available on the EUCAST website: https://www.eucast.org [accessed on 22 January 2023]). Molecular analysis to identify the resistance mechanism was unavailable.
In our center, E. faecium proved to be a risk factor for harboring resistance to ampicillin (p < 0.01), streptomycin (p < 0.05), ciprofloxacin (p < 0.01), teicoplanin (p < 0.01), and vancomycin (p < 0.01). These findings highlight the importance of quick and efficient bacterial identification for Enterococcus spp. strains as there are differences when it comes to the resistance patterns of E. faecium and E faecalis. Corroborating these results with the increasing number of resistant strains during our study, we consider that E. faecalis strains present a higher susceptibility to acquire antibiotic resistance genes, representing a bigger threat for the healthcare systems in our region.
In COVID-19 patients, we identified statistically significant differences for Ampicillin (p < 0.05) and Vancomycin (p < 0.05). The presence of SARS-CoV-2 infection proved to be associated with higher chances of selecting an Enterococcus strain with resistance to ampicillin (OR = 2.569) and vancomycin (OR = 2.707).
The survivability rates in association with antibiotic resistance were not statistically significant different for most of the antibiotics tested. We registered a significant higher mortality rate for strains with resistance to ciprofloxacin (p < 0.05, OR = 3.015). This result can serve as a warning sign for medical professionals when it comes to using ciprofloxacin as a treatment alternative for patients at high-risk. Y. B. Kim et al. described Enterococcus spp. strains with resistance to ciprofloxacin isolated from chicken meat [43]. It is important highlights the possibility of contamination through food products and raise awareness when it comes to this understudied field.

Enterococcus gallinarum and the Emergence of Other Enterococci
Other Enterococci (OE) were reported in literature more frequently in the past few years [44]. As for our study, we cannot identify strains of OE during the first 12 months. A surprising discovery was the emergence of E. gallinarum infections (n = 2) during Period C. We identified no association between E. gallinarum with SARS-CoV-2 patients, and the survivability rates for patients infected with these strains were 100%. It is important to consider that both E. gallinarum strains were isolated from cerebrospinal fluid (CSF). During the 18 months of our study, we isolated only 4 strains on Enterococcus spp. from CSF, and 50% of these strains were represented by E. gallinarum. Although, the number of strains is insufficient to provide significance, we must pay attention to this trend of increasing E. gallinarum infections in neurological pathologies [45,46].

COVID-19, Enterococccus spp. and the Use of Probiotics
It is known that one of the first steps in infection pathogenesis is the ability of the bacterial strains to colonize different surfaces and tissues. From here, every disruption of the surface integrity provides the premises for the bacterial strain to emerge as etiologic agent for the infection. This basic model to infection applies to Enterococcus spp. as well, considering the wide variety of factors that are present. Recent studies show that hospital adapted linages of Enterococci are able to colonize the human gut developing a reservoir from where these strains can cross the gut barrier and produce systemic infections. Translocation of Enterococci from the gut into the bloodstream is enhanced by several factors like the extensive use of antibiotics which challenge the microbiome balance as well as inflammation. Thus, the extensive use of antimicrobials during the COVID-19 pandemic along with the systemic inflammation that the SARS-CoV-2 virus produces represents the arguments for the increased number of Enterococcus spp. infections in COVID-19 patients.
Recent studies show that commensal E. faecalis is able to destroy the VRE strain V583. This is achieved to the production of a specific pheromone peptide called cOBI. This shows that in time VRE colonization can be overcome by providing a balanced intestinal microbiome. It becomes clear that tackling dysbiosis represents an acceptable therapeutic target in patients with SARS-CoV-2 pneumonia since it can provide immune homeostasis in the gut. Choosing the right probiotic remains an issue, further studies being required in order to address the pros and cons of each valid candidate [47][48][49].

Limitations and Implications
Regarding the limitations of our study, one of the first things worth mentioning is the retrospective nature of this paper. Furthermore, it is important to mention that all the data included are from only one academic hospital in Romania. In addition, only a small number of samples met our inclusion criteria and therefore it is difficult to estimate the impact of our findings on a larger scale. Moreover, our facility provides bacterial identification only using the VITEK-2 system. A more accurate diagnostic tool was not available.
Our study was designed to highlight the danger that AMR phenomenon represents to our healthcare system and our results confirmed this emerging threat. Moreover, we identified changes in the pattern of antimicrobial resistance of Enterococcus spp. strains that may be linked to the on-going COVID-19 pandemic. Further studies should consider a multicentric approach, in order to validate our results on a larger scale.

Materials and Methods
This study was designed as an observational, retrospective study, including data that were gathered from a single academic hospital in Cluj-Napoca, Romania. We collected data regarding Enterococcus spp. infections during the COVID-19 pandemic period from 1 November 2020 to 30 April 2022.
The data included in this study were from patients admitted in our unit with at least one episode of infection with Enterococcus spp. To be included in this study, (1) the infection had to be confirmed by a bacterial identification on VITEK-2 Compact (bioMérieux, Inc., Marcy l'Etoile, France) and (2) an antibiotic susceptibility test should have been performed on VITEK-2 Compact (bioMérieux, Inc., Marcy l'Etoile, France).
Ninety-eight patients fulfilled the inclusion criteria. Medical records of these patients were reviewed, and the following data have been anonymously stored in our database: gender, date of admission, SARS-CoV-2 infection status, length of stay, survival, sample analyzed, Enterococcus spp. species identified and the results of the susceptibility tests for Ampicillin, Gentamicin, Streptomycin, Ciprofloxacin, Erythromycin, Linezolid, Teicoplanin, Vancomycin, Tetracycline and Tigecycline. We divided the timeline of our study (18 months) into 3 evenly distributed time periods: Period A from 1 November 2020 to 30 April 2021; Period B from 1 May 2021 to 30 October 2021 and Period C from 1 November 2021 to 30 April 2022.

Statistical Assesment
The data we collected were stored in a Microsoft Excel database. For our ordinal data, a Shapiro-Wilk Test for Normality was performed. Nominal data groups were compared using the Chi-square test or Fisher's test (based on sample size). Means comparison was performed using One-Way ANOVA and Student test. Log-rank analysis and Cox regression were performed for the survival analysis. We considered p-value statistically significant for values less than 0.05. The entire statistical analysis was performed using the IBM ® SPSS ® Statistics 20.